Abstract
Fossil chondrichthyan teeth played an important part in the establishment of a scientific understanding of ‘formed stones’. Following a slowly emerging taxonomy, Louis Agassiz presented the first comprehensive guide to Palaeozoic chondrichthyans in the 1830s. The next contribution of any substance was Arthur Smith Woodward's Catalogue of Fossil Fishes in the British Museum (Natural History) with a historical, descriptive and systematic review of the chondrichthyans, a group on which he already had an impressively large publication record. Initially stimulated by his observations on an articulated petalodont dentition (Climaxodus), Smith Woodward erected the Bradyodonti in 1921. Defined on the possession of dentitions with very slow growth rates, only seven or eight successional teeth produced throughout the lifetime of the fish, and retention rather than shedding of earlier teeth, primarily by fusion to later ones, the bradyodonts embraced petalodonts, psammodonts, copodonts and cochliodonts. The establishment and subsequent demise of the bradyodonts is briefly reviewed here.
Pre-scientific notions
Sharks’ teeth are one of a fairly select group of fossils that caught the attention of early humans; they were obviously selectively collected, being found in a wide range of archaeological contexts ranging from the Upper Palaeolithic to the Neolithic in many parts of the world.
In European folklore tradition, specimens – especially those belonging to Tertiary lamniform sharks – were often referred to as Glossopetrae or ‘tongue stones’ on account of their supposed morphological resemblance to the tongues of various creatures, including humankind. The earliest reference seems to be that in the Historia Naturalis (Natural History) of Pliny the Elder (AD 23–79; Book 37, Cap. 59), the Roman author, military commander and encyclopaedist who famously died whilst observing an eruption of Vesuvius (Fig. 1). Pliny indicated that glossopetrae were believed to fall from the skies during eclipses of the moon, an origin that endowed them with magical properties; they were essential tools in selenomancy and supposedly had the power to control the winds (Bostock & Riley 1857, p. 449; see also Duffin & Davidson 2011, p. 12).
The death of Pliny the Elder during the eruption of Vesuvius (79 AD). Boaistuau, P. (1560) Histoires Prodigieuses: ‘Prodigieuse mort de Pline [dont le] corps fut mis en cendre [par les flammes du Vesuve]’, Folio 64 verso. Reproduced by kind permission of the Wellcome Library.
On Malta, fossil sharks’ teeth are commonly known as linguae melitensis (‘Maltese tongues’), linguae S. Pauli (‘St Paul's tongues’) or lapidae Domini Pauli (‘Stones of St Paul’), evoking the New Testament story of St Paul and the snake in the woodpile (Acts 28 verses 1–6). Local embellishment of the Biblical account saw Paul's preaching being so erudite and forceful that his tongue was able to penetrate solid rock, leaving an image of itself behind. Perhaps unsurprisingly, this led to a belief in the healing efficacy of glossopetrae, which were collected assiduously and formed the basis of a thriving export market to apothecaries throughout Europe. Worn as pendants, they were used as amulets against poisons and, suspended from ornate items of renaissance tableware (Natternzungenbaum, Natternzungen-Kredenz or Languiers), were essential in the bid to detect poisoning attempts on the lives of various nobility by would-be assassins (Duffin 2012).
From folklore to science
As various authors began to question received wisdom from classical and medieval authorities, fossil sharks’ teeth took centre stage in debates about the origins of ‘formed stones’. Leonardo da Vinci (1452–1519) argued in his notebooks that glossopetrae must represent the remains of once-living organisms, an observation supported by the likes of Girolamo Fracastoro (1478–1553) when discussing the many fossils, including shark's teeth, then being discovered in the vicinity of Verona, his home town.
Additional focus was brought to bear by the Neapolitan nobleman, Fabio Colonna (c. 1567–1640), a member of the Academy of Lynxes (Accademia dei Lincei), a group of scientific elite founded in 1603 that harboured progressives and radicals. Working from his friend Ferrante Imperato's (1525–1615) impressive geological collections, Colonna's penetrating insight into the nature of tonguestones was published as an adjunct to a short paper on the shells of marine organisms, written in 1616; in De Glossopetris Dissertatio he methodically refutes the arguments of those opposed to the organic origins of glossopetrae (Colonna 1616).
Niels Stensen (Steno; 1638–86) famously described the cranial musculature of a Great White Shark (Carcharodon carcharias (Linnaeus, 1758)) captured at Livorno in 1667, and retained for his study by Ferdinand II de Medici (1610–70), the Grand Duke of Tuscany. In an appendix to his (1667) Canis carchariae dissectum caput, Steno argued persuasively that glossopetrae were fossilized sharks’ teeth (Steno 1667), a conclusion reached independently by the Sicilian painter, Agostino Scilla (1629–1700) (Scilla 1670).
Early taxonomies
Prior to the wholesale introduction of the binomial system of scientific nomenclature and an ordering of living things into a hierarchically ranked classification by the Swedish botanist, Carl Linnaeus (1707–78), culminating in the 10th edition of his Systema Naturae (Linnaeus 1758), extant sharks, skates, rays and chimaeras received relatively little attention in the literature. The great early modern encyclopaedic compilations of natural history, such as that of Conrad Gessner (1558), mostly provided brief descriptions of fishes with accompanying illustrations of the external morphology, primarily as an identification tool, and often with reference to potential uses in culinary or pharmaceutical preparations. Well-known exceptions include the works of Guillaume Rondelet (1507–66), Regius Professor of Medicine at Montpellier University, who studied the morphology, anatomy and habits of many fishes from the Mediterranean area on the basis of personal fieldwork, rather than compiling observations from classical authors.
Pierre Belon (1517–64), a close contemporary of Rondelet's, was also rather overshadowed by him (Gudger 1934, p. 28). Belon, a French explorer, diplomat and naturalist, divided modern cartilaginous fishes into two groups (De cartilagineis viuiparis oblongis and De cartilagineis viuiparis planis), essentially, oblong forms (sharks) and flattened ones (rays and Lophius, the Monkfish) (Belon 1553, folios 58–98).
John Ray (1627–1705) and Francis Willughby (1635–72) drew heavily upon the work of Rondelet for their volume, Historia Piscium (Willughby & Ray 1686; Nellen & Dulcic 2008; Lewis 2012). At an early stage in their collaborative attempts to reform the study of natural history, Ray and Willughby had agreed upon a mutually acceptable division of labour – Ray would cover plants, while Willughby concentrated on fishes, birds and insects (Kusukawa 2000). On Willughby's untimely and early death, Ray worked from his unpublished notes, drawing heavily on support from Martin Lister (1639–1712) and other colleagues. Whilst earlier authors had commonly grouped cartilaginous fishes together, Willughby and Ray introduced the descriptive term Pisces Cartilaginei containing the Cartilaginei longi (embracing a variety of sharks) and the Cartilaginei plani (for the rays, Squatina, and the ‘Toadfish or Sea Divell’; Willughby & Ray 1686, p. 12), very reminiscent of Belon's earlier divisions.
Carl Linnaeus (1707–78) extended this classification system to include fin architecture as a characteristic, introducing the Chondropterygii (fishes with cartilage-supported fins) in the first edition of his Systema Naturae (Linnaeus 1735, p.11), into which Gronov (1754) was later able to place Callorynchus. The 10th edition of Linnaeus's Systema (Linnaeus 1758), generally hailed as the inception of modern biological nomenclature, saw him abandon the Chondropterygii in favour of the Amphibia Nantes, defined on the basis of being:
Animals that are distinguished by a body cold and generally naked; stern and expressive countenance; harsh voice; mostly lurid colour; filthy odour; a few are furnished with a horrid poison; all have cartilaginous bones, …: they are deficient in diaphragm, do not transpire (sweat), can live a long time without food, are tenatious of life. (Turton 1806, p. 638)
When Arthur Smith Woodward gave the 1921 Presidential address to the Fellows of the Linnean Society at Burlington House in Piccadilly, in the meeting room where Darwin and Wallace's papers on evolution by natural selection had been delivered some 63 years earlier, his opening statement was very diplomatic considering the group to whom he was talking:
Linnaeus was unfortunate in his treatment of the sharks, rays and chimaeras, which he ultimately removed from the class of fishes, and arranged among the Reptiles as part of the Order Amphibia Nantes. (Woodward 1921, p. X)
Thankfully, this taxonomically aberrant interlude did not last (see Lund & Grogan 1997 for more details); Cuvier (1817) resurrected the Chondropterygii, and Charles Lucien Jules Laurent Bonaparte (1803–57), first cousin of Napoleon and a man seemingly obsessed with lists and hierarchies, introduced two significant terms into the taxonomic literature: the Elasmobranchii for the sharks, skates and rays in 1838 and the Holocephali for the modern ratfishes, ghost fishes and elephant fishes (chimaeroids) in 1831 (Bonaparte 1832–41). The two groups were united together in the Chondrichthyes by Thomas Henry Huxley (1825–95), Darwin's Bulldog, in 1880 (Huxley 1880).
Fossil chondrichthyans
Whilst Bonaparte was looking at extant fishes, fossil fishes had caught the attention of a young Swiss naturalist, Louis Agassiz (1807–73; Fig. 2; Duffin 2007). Fascinated by fishes from an early age, as a child, Agassiz kept freshwater fish that he had caught in Lake Neuchâtel in a stone basin in the grounds of the parsonage in which he grew up. Following his initial studies at Lausanne, Zürich and Heidelberg, Agassiz moved to Münich where he was given the opportunity of describing the large collection of extant freshwater Amazonian fishes returned from South America by C. F. P. von Martius (1794–1868) and Professor Julius Spix (1781–1826), collected during their expedition from 1817 to 1820.
Portrait of Louis Agassiz (1807–73), taken from Agassiz (1833–43) Recherches sur les Poissons Fossiles, and made available through the kindness of David Ward.
Shortly after his successful publication of some 90 species of Brazilian fishes, Agassiz determined to tackle the fossil history of the group. He visited major museum collections in Germany and found a wealth of unpublished fossil material. At Münich he met two English aristocrats, William Willoughby Cole (1807–86; later the earl of Enniskillen and Baron Grinstead) and Sir Philip de Malpas Grey Egerton (1806–81) (Fig. 3a, b). They engaged in a sort of geological tour of Europe, encouraged by their teacher and mentor back home, William Buckland (1784–1856; Professor of Geology at Oxford University). The pair were inspired by Agassiz's enthusiasm for fossil fishes and his entreaties to focus their collecting in that area, which they did to great effect, sometimes deciding the distribution of part and counterpart on the amicable toss of a coin. A close friendship and lively correspondence ensued. On hearing Agassiz's plans to describe all fossil fish, Buckland invited him to see the fossil fishes of England. Agassiz arrived in England for the first time in August 1834. People were eager to make the acquaintance of ‘the foreign fishmonger’, as Adam Sedgwick (1785–1873) called him. Buckland had secured access to the manifold riches of 60 or more British geologists. Around 2000 specimens culled from diverse sources were duly deposited in the Geological Society's rooms at Somerset House, where Agassiz was able to study and compare them at leisure. The young Bavarian artist, Joseph Dinkel (born between 1800 and 1807, died sometime after 1873), whom Agassiz had befriended in Münich, was ensconced with the collection, and tasked with producing illustrations for the lithographic plates that would adorn the descriptive work.
(a) Portrait of Sir Philip de Malpas Grey Egerton; (b) portrait of William Willoughby Cole, 3rd Earl of Enniskillen, from an oil painting by John Lucas (1873). By permission of the Trustees of the Natural History Museum, London.
The results of Agassiz's monumental labours were the five volumes of the Recherches sur les Poissons Fossiles (Fig. 4), published in separate parts or livraisons over a ten-year period (Agassiz 1833–43) and paid for in advance by subscription. Amongst the 1700 species that were described, over 90% of which were new, were nine genera (26 species) based on teeth from the Carboniferous, largely of Britain, plus another eight genera (20 species) based upon isolated fin spines. Agassiz first established some of the household names of Carboniferous chondrichthyans, such as Cochliodus, Psammodus, Cladodus, Ctenacanthus, Helodus and Orodus, while from the Mesozoic (Triassic, Jurassic and Cretaceous) he defined Acrodus, Hybodus and Ptychodus, for example, added to many others from later deposits. Amongst the holocephalans, an otherwise dearth of articulated skeletal remains was rectified by the description of a beautiful specimen of Squaloraja (but now fragmented and partly lost) from the Lower Lias of Lyme Regis. All these were placed in his order Placoidea.
Title page of Louis Agassiz's (1833–43) Recherches sur les Poissons Fossiles.
Lord Enniskillen's collection of chondrichthyan remains from the Carboniferous Mountain Limestone (Mississippian,Viséan) of Armagh formed an important resource for Agassiz. Large numbers of specimens of individual taxa were represented as, for example, in the case of Helodus teeth (Fig. 5). Together with Egerton's collection, Enniskillen's collection famously became incorporated into the Natural History Museum, London collections during the 1880s, forming the core of their holdings of fossil fishes. Amongst this material, Agassiz noted that a number of fossil genera had crushing dentitions with flat surfaces to the teeth. This was reminiscent of the condition in Recent Heterodontus, the Port Jackson Shark, known at the time as Cestracion.
Teeth of Helodus didymus Agassiz, 1843 from the Mountain Limestone (Viséan) of Armagh. This tray of specimens represents less than 10% of the material belonging to this species collected by Lord Enniskillen and made available to Louis Agassiz for description. NHMUK PV P 2638. Scale bar 10 mm.
Pioneering the use of microscopy as an analytical tool in palaeontology, Agassiz incorporated detailed descriptions and images of tooth histology of numerous fossil genera in his Recherches. Using a combination of histology and morphology he placed a number of Palaeozoic chondrichthyans together with Heterodontus in his Family Cestraciontidae. This was defined on the basis of teeth possessing a crown penetrated by numerous vertical, often bifurcating medullary canals, whose lateral branches anastomose to form a longitudinal meshwork. Within this Family, he recognized three unnamed groups of genera and subgenera: (a) Cestracion (Heterodontus), Strophodus, Ptychodus and Acrodus, each of which possessed a superficial layer of enameloid; (b) Chomatodus, Cochliodus (Fig. 7a), Psammodus (Fig. 7b), Helodus (Fig. 7d) and Ceratodus, defined by the absence of a superficial enameloid layer such that the medullary canals opened freely on to the tooth surface; and (c) Ctenodus and Ctenoptychius (Fig. 7i), characterized by the presence of a well-developed canal system with few dentine tubules. The picture, as presented by Agassiz in the 1840s, was one with the Cestraciontes as the most prominent member in one of his two most diverse and long-lived groups of fossil fishes (Fig. 6).
Chart showing the genealogy of fishes (from Recherches sur les Poissons Fossiles, v. 1, opposite p.170).
Specimens of representative genera figuring in the debate about the Bradyodonti. (a) NHMUK PV P 2424: articulated lower tooth plates of Cochliodus contortus Agassiz from the Carboniferous Limestone of Armagh, Ireland (Enniskillen Collection), occlusal view. (b) NHMUK PV P 2572: tooth of Psammodus rugosus Agassiz, from the Carboniferous Limestone of Armagh, Ireland (Enniskillen Collection), occlusal view. (c) NHMUK PV P 2444: isolated tooth of Deltodus sublaevis (M'Coy) from the Carboniferous Limestone of Armagh, Ireland (Enniskillen Collection), lateral view. (d) NHMUK PV P 8233: associated teeth of Helodus simplex Agassiz, occlusal view. (e) NHMUK PV P 2331: isolated partial tooth of Edestus heinrichi Newberry & Worthen, lateral view. (f) NHMUK PV P 7053: isolated tooth originally assigned to Campodus rectangularis (Trautschold, 1879), lingual view. (g) NHMUK PV OR 42168: isolated tooth of Orodus ramosus Agassiz, Carboniferous Limestone of Oreton, Shropshire, occlusal view. (h) NHMUK PV P 2513: isolated tooth (HOLOTYPE) of Copodus spatulatus Davis from the Carboniferous Limestone of Armagh, Ireland (Enniskillen Collection), occlusal view. (i) NHMUK PV P 3005: isolated tooth of Ctenoptychius lobatus Etheridge from the Yoredale Series of Wensleydale, Yorkshire (Enniskillen Collection) in lingual view. (j) NHMUK PV P 3033: articulated partial dentition of Janassa bituminosa from the Marl Slate (Permian) of Midderidge, Durham (Enniskillen Collection), lateral view. Scale bar 10 mm.
Agassiz continued to consider the crushing teeth obtained from the Carboniferous, especially those in Enniskillen's collection from Armagh, and later produced a manuscript (Agassiz 1859) introducing the genera Psephodus, Streblodus, Deltodus, Deltoptychius, Tomodus and Xystrodus. The work was never published, but the included concepts and generic names were honoured and retained by subsequent workers.
As the last of Agassiz's livraisons of the Recherches sur les Poissons Fossiles were rolling off the presses in Switzerland, Sir Richard Owen (1804–92), was working on his textbook of tooth histology, entitled Odontography (Owen 1845), based upon his series of Hunterian lectures on the comparative anatomy and physiology of teeth, and delivered at the Royal College of Surgeons from 1837 to 1839. Owen concurred, ‘with unalloyed pleasure’, with the bulk of Agassiz's observations and conclusions (Owen 1845, p. xii). Owen did remark on histological differences between the teeth of Psammodus and those of Acrodus and Hybodus; Agassiz having restricted the definition of Psammodus to teeth that ‘combine a broad, flat, punctate grinding surface, with a great breadth of crown’ (Owen 1845, pp. 59–60), Owen pointed out that the ‘calcigerous tubes’ terminated just below the occlusal surface of the tooth in that genus, whereas in the teeth of Acrodus, Hybodus and Ptychodus the surface was covered by an enamel-like layer that he later referred to as ‘vitrodentine’ (Owen 1867, p. 60; now referred to as enameloid). Furthermore, this overall arrangement was also shown in the teeth of Cochliodus and Petalodus. Owen later (Owen 1867) removed Cochliodus, together with Tomodus convexus Agassiz to his newly erected Family Cochliodontidae. M'Coy (1855, p. 621) also had serious reservations about the proposed affinities between Cochliodus and ‘Cestracion’, noting that in Cochliodus ‘the succession of the teeth was not by revolution from behind as in the Plagiostomous fishes, but vertically from below upwards’. Furthermore, it soon became clear that some of the other genera that Agassiz had grouped together did not belong in these groups; Acrodus was accordingly removed to the hybodonts, and with the discovery of Lepidosiren, the South American lungfish, in 1837, and what was to become Neoceratodus forsteri, the Australian Lungfish, in 1870, Ceratodus was removed to the dipnoans.
In the decades following the completion of Agassiz's seminal work, little was done to modify this classificatory scheme, although Newberry & Worthen (1866, p. 31) did separate the ‘Cestracionts’ into two families – the Cestraciontidae and the Petalodontidae – and recognized that teeth referred to as Streblodus were elongate tooth plates from the upper dentition of Cochliodus. Rising interest in Palaeozoic fossils and their collection meant that a number of new genera were described. In the USA, Joseph Leidy (1823–91) described Edestus from the Carboniferous, the founding member of a group of enigmatic chondrichthyans characterized by symphyseal tooth whorls in the lower jaw (Leidy 1856) (Fig. 7e), although at the time of its first description, the nature of Edestus was not fully appreciated (Leidy thought it belonged to the upper jaw complex of an osteichthyan, and subsequent discussion of the tooth whorls often placed them as defensive structures on, variously, dorsal, caudal or pectoral fins; Itano 2014). However, by far the greatest contributions came from three US State Survey geologists, Orestes Hawley St John (1841–1921, a student of Agassiz), Amos Henry Worthen (1813–88) and John Strong Newberry (1822–92). Their descriptions of fossil chondrichthyans from the Coal Measures of Illinois and Ohio were not the only legacy to be later grasped by Arthur Smith Woodward. In Europe, further finds were detailed by Ghennadi Danilovich Romanowsky (1830–1906), Gustav Heinrich Ludwig Hermann Trautschold (1817–1902) and Laurent Guillaume de Koninck (1809–87, whose collection was purchased for the Museum of Comparative Zoology at Harvard by Agassiz for $4000), and in Britain by Frederick M'Coy (c. 1823–99) and James William Davis (1846–93) (see Stahl 1999 and Ginter et al. 2010 for references). In the interim period, Sir Richard Owen and Sir Philip Egerton made initial suggestions, based on the form of the tooth plates, that cochliodonts might be allied to the chimaeroids of the Mesozoic via more primitive genera such as Myriacanthus in the Jurassic (Egerton 1872, p. 236; Owen 1845, p. 65).
In the absence of well-preserved and articulated specimens, the morphology, number and overall disposition of the tooth plates in the dentition remained the significant characteristics in any discussion over the nature and relationships of Carboniferous chondrichthyans. Likewise, the presence of the coronal tissue described by Agassiz as having freely open medullary canals on the occlusal surface without a superficial covering of enamel was considered to be a unifying feature. This tissue would later be described under a bewildering array of names, the most common of which are pleromin and tubular dentine (see Stahl 1999, pp. 19–22 for an excellent summary).
Smith Woodward's contribution
When Arthur Smith Woodward arrived at the British Museum (Natural History), as it was then known, in August 1882, fossil fishes were ripe for review. By the time the first volume of his monumental Catalogue of Fossil Fishes in the British Museum (Natural History) was published in 1889, he had already produced over 100 papers, almost half (46) of which dealt with fossil chondrichthyans. This group seemed to be the logical choice for the place to begin his epic work. As he states in his Introduction, it was ‘the first attempt at a systematic treatment of the Palaeontology of the Elasmobranch fishes’ (Woodward 1889a, p. v), albeit a provisional one, awaiting further finds and more extensive research. Beginning with a review of earlier studies from the seventeenth century onwards, he records his debt to earlier researchers, including those cited above. So far as the record of Carboniferous chondrichthyans goes, he remarks on the paucity of articulated material – only four species represented by anything more than isolated teeth – and picks up on the observations of Agassiz that ‘Cestraciont’ dentitions have smaller symphyseal teeth and enlarged, crushing laterals, reaching their fullest development in the cochliodonts and other Carboniferous forms.
His entries are economically presented, logically laid out and still form a fundamental reference for researchers today. New taxa are described and new combinations erected. In spite of his stated opinion that there is ‘scope for endless differences of opinion’ (Woodward 1889a, p. xxviii) when it comes to identifying synonyms, the lists that he presents are the insightful product of his wide experience of both specimens and publications. Smith Woodward's descriptions are concise, accurate and lucid and he identifies the whereabouts of type material as well as referring to the holdings of the (then) British Museum (Natural History); he also usefully lists further species identified in the literature but not represented in the collection at that time. Although not extensively illustrated, those representations that appear are accurate and useful. As the first attempt at a systematic treatment of fossil chondrichthyans it is most impressive.
Amongst his subsequent publications are a number that deal with new Carboniferous material – species of Deltodus (Fig. 7c; Woodward 1900, 1912), Edestus (Woodward 1916) and Cochliodus (Woodward 1912), for example, a discussion of Helicoprion and the similarity of its tooth files to the arrangement in Cochliodus (Woodward 1903), and a suggestion that teeth originally called Campodus (Fig. 7f) were part of the dentition of Edestus (Woodward 1916, p. 4). He began to rationalize the classification of holocephalian taxa by erecting the Myriacanthidae for Jurassic Myriacanthus and Metopacanthus (Woodward 1889b). These two genera possess sets of tooth plates whose structure and disposition resemble quite closely those of Deltoptychius, with which it also shares a series of tuberculated spines in the head region (Fig. 8a, b). Smith Woodward also described a beautiful specimen of Squaloraja, reinforcing its kinship with the holocephalians (Woodward 1886). In a critique of Jaekel's (1891) description of a newly prepared specimen of Menaspis armata, Smith Woodward (1891, p. 425) wrote that the menaspids ‘represent some unknown group that yet requires elucidation’.
(a) NHMUK PV P 4664: Associated lower jaw cartilages and tooth plates of Myriacanthus paradoxus Agassiz from the Lower Lias (Lower Jurassic) of Lyme Regis, Dorset. (b) NHMUK PV P 62092: head region showing tooth plates of Deltoptychius armigerus Traquair from the Namurian of Bearsden, Glasgow. Scale bar 10 mm.
In 1919 Smith Woodward described a largely articulated dentition of a new petalodont from the Calciferous Sandstone of Scotland as Climaxodus wisei. Broken between the two jaws, the specimen was noteworthy because of the disposition of its dentition. Up to that point, the complete petalodont dentition was known only for Janassa bituminosa from the Permian (Fig. 7j). Smith Woodward remarked on the size discrepancy between the anterior and posterior-most teeth of the symphyseal row in Climaxodus, indicating the presence of only five or six teeth in the row, compared with Janassa’s equally small seven or eight symphyseal teeth, but in stark contrast to the large numbers of successional teeth found in elasmobranchs. Similarly restricted numbers of successional teeth being found in cochliodonts and psammodonts, Smith Woodward suggested that ‘It must therefore have been a characteristic of many Palaeozoic Elasmobranchs, distinguishing them sharply from those groups which survived until later times’ (Woodward 1919, p. 5). Furthermore, the presence in Climaxodus of tissues forming the crushing surface of the crown being histologically similar (made of ‘highly vascular dentine’) to those forming the tritors of later chimaeroid teeth led him to comment that ‘the presence of an apparently chimaeroid character in Elasmobranch teeth which are noteworthy for their slow and scanty succession, may therefore have some special significance’ (Woodward 1920, p. 5).
The following year, in 1921, Smith Woodward gave the Presidential address to the Fellows of the Linnean Society at Burlington House in Piccadilly. His review of elasmobranch relationships, stimulated by his observations on Climaxodus, included the introduction of a new group embracing a number of the Palaeozoic forms loosely associated with Cestracion by Agassiz almost 80 years earlier. Smith Woodward's Bradyodonti was defined on the basis of dentitions with very slow growth rates, only seven or eight successional teeth produced throughout the lifetime of the fish, and retention rather than shedding of earlier teeth, primarily by fusion to later ones. His bradyodonts embraced four families – the petalodonts, psammodonts, copodonts and cochliodonts (Fig. 7). Smith Woodward made no further changes to this system, retiring three years later in 1924, and living a further 20 years in retirement before his death in 1944. Others did take up the discussion, however.
Developments since Smith Woodward
The Danes mounted spectacularly successful geological fieldwork and collecting expeditions to East Greenland during the 1920s and 1930s. The Upper Permian Beds of Cape Stosch yielded a number of specimens that were first studied by Eigil Nielsen (1910–68) from Copenhagen in 1932. Reviewing Smith Woodward's Bradyodonti, he advocated that the edestids should be included in the group, as well as the members of the Family Orodontidae, comprising Orodus (Fig. 7g) and Agassizodus. He believed that these forms possessed a specialized symphyseal tooth row showing some similarities with the edestid condition, and noted the fact that the crown was composed of tubular dentine, as in other bradyodonts, and lacked a superficial layer of enameloid (Nielsen 1932, p. 20). Nielsen described material that he believed to be allied to Agassizodus as A. groenlandicus, plus a new taxon, Fadenia crenulata (Nielsen 1932) (Fig. 9e, f).
(a) Tooth whorl of Sarcoprion edax Nielsen (MGUH. VP. 215–94) in lateral view; (b) articulated symphyseal teeth of Erikodus groenlandicus Nielsen in occlusal view (MGUH.VP. 221–99); (c) articulated symphyseal tooth whorl of Erikodus groenlandicus in lateral view (MGUH.VP. 220–97); (d) Erikodus groenlandicus symphyseal tooth whorl in section (MGUH.VP. 220–97); (e) associated parts of the dentition of Fadenia crenulata Nielsen (MGUH.VP.45); and (f) associated partial dentition of Fadenia crenulata (MGUH.VP.55). All specimens are from the Middle Permian Posidonia Shale Member of the Foldvik Creek Formation (Middle Permian, Guadalupian) of the Hird Bay area of east Greenland. Scale bar 10 mm.
Nielsen had the opportunity to accompany the Danish Geological Survey team in the field during the 1932–33 and 1936–38 expeditions, during which more material was collected. This allowed him to separate the East Greenland species from Agassizodus and place it in a new genus named after the famous Swedish palaeontologist, Erik Stensio (1891–1984), as Erikodus groenlandicus (Nielsen 1952; Fig. 9b, d). The same localities also yielded some fragmentary material of an edestid, described by Nielsen (1952) as Sarcoprion edax (Fig. 9a).
Study of the material collected from East Greenland also afforded him the opportunity to reinforce his views about the composition of the Bradyodonti, even though the inclusion of the edestids and orodontids demonstrated more rapid tooth succession; the name was retained but the major defining characteristic was abandoned, meaning that tooth histology was now the sole remaining uniting characteristic of the group.
The symmetrical symphyseal teeth of Fadenia are high centrally, with a rather blunt median ridge on the central crest of the tooth. The two lateral wings are posteriorly-directed and subparallel to each other. The ornamentation is largely confined to ridges in the tissue on the surface of the lateral wings. Corrugations along the posterior and lateral margins of the tooth, especially of the wings, assist in the interlocking of adjacent teeth (Fig. 9e, f).
The symphyseal teeth of Erikodus (Fig. 9b, d) are lower-crowned than those of the somewhat similar Fadenia, but otherwise have a similar outline in occlusal view. Each tooth has a strongly convex anterior margin and a correspondingly concave posterior margin in occlusal view; both margins are heavily folded. The lateral wings of the crown are not as long as in Fadenia, and less parallel to each other. In relatively unworn specimens (e.g. MGUH.VP.221), the centre of the crown is marked by an occasionally quite complex transverse ridge which may bifurcate strongly laterally, giving rise to a series of secondary ridges that may branch basally; almost all evidence of surface ornamentation may be obliterated in worn specimens (e.g. MGUH. VP.220).
The low-crowned lateral teeth of Erikodus are much more elongate than the symphyseals, with a straight rather than curved coronal axis. The reduced ornament of surface ridges is often removed by ante-mortem wear. Posterolateral teeth also possess a very low crown, this time oval to elliptical crowns in occlusal view. Surface ornamentation is usually removed by ante-mortem wear, but when present consists of a relatively simple set of surface ridges.
The results of the first ever SEM study of the tissue covering the occlusal surface of the Permian teeth from East Greenland are shown in Figure 10. With the exception of Sarcoprion sp., the isolated teeth used in this study were all chosen after careful examination under a stereoscopic light microscope on the basis of their lack of ante-mortem and post-mortem wear; the occlusal surfaces were all highly lustrous and devoid of obvious scratch marks. Examination of the acid-etched occlusal and minor breakage surfaces of the teeth indicates that a clearly defined layer of tubular dentine or pleromin (e.g. Fig. 9d) forms the main component of the crushing tissues present; pillars of hypermineralized tissue (Fig. 10c) terminating just below the occlusal surface give a spotty appearance to the crown (Fig. 9b, c, e, f), which is accentuated by ante-mortem wear in larger specimens. An enameloid layer is entirely absent; the superficial layer of the crown comprises a vertically oriented, tuft-like arrangement of single crystallites (occasionally altered by minor diagenetic recrystallization; Fig. 10c) in Edestus (Fig. 10a), Erikodus (Fig. 10d), Fadenia (Fig. 10f) and Sarcoprion (Fig. 10c). No specialized arrangement of the crystallites is evident at the crown ridges (Fig. 10a). Evidence of a thin compact layer of single crystallites, each measuring 0.5 μm in length, is also present in all genera (adjacent to the apex of the crown in Sarcoprion (Fig. 10b) and Erikodus (Fig. 10d)).
SEM photographs of the compact dentine layer forming the superficial layer in eugeneodontiform teeth. The specimens chosen for SEM study were first etched in 10% HCl for the following lengths of time: 5 s – Sarcoprion edax; 8 s – Fadenia crenulata; 30 s – Edestus sp., Erikodus groenlandicus. Each was then coated in a thin layer of evaporated gold palladium and studied using either an FEI Inspect SEM at an acceleration voltage of 20 kV, or a Jeol field emission scanning electron microscope JSM-6335F using an acceleration voltage of 7 kV. (a) Surface view of a cusp apex in Edestus sp., Pennsylvanian (Upper Carboniferous) of Baldwin Mine, Sparta, Illinois, USA; (b) surface view of the compact layer adjacent to the cusp apex in Sarcoprion sp., Middle Permian Posidonia Shale Member of the Foldvik Creek Formation (Middle Permian, Guadalupian) of the Hird Bay area of east Greenland (MGUH.VP.216); (c) tufted arrangement of crystallites near the crown base in Sarcoprion sp. (MGUH.VP.216); (d) broken surface showing the junction between the compact outer and inner tufted layers in a lateral tooth of Erikodus groenlandicus, Middle Permian Posidonia Shale Member of the Foldvik Creek Formation (Middle Permian, Guadalupian) of the Hird Bay area of east Greenland (MGUH.VP.309); (e) hypermineralized dentine pillar (indicated by arrow) in a posterolateral tooth of Fadenia crenulata, Middle Permian Posidonia Shale Member of the Foldvik Creek Formation (Middle Permian, Guadalupian) of the Hird Bay area of east Greenland (MGUH.VP.GGU 29666); and (f) tufted layer in surface view, Fadenia crenulata (MGUH.VP.GGU 29666).
James Alan Moy-Thomas (1908–44) was lecturer in Zoology and Comparative Anatomy at Christ Church College, Oxford, and very active in the description of Carboniferous fossil fishes during the 1930s, before being tragically killed in a car accident at the age of only 35. His studies of associated and articulated remains of Chondrenchelys problematica in 1935 and Helodus simplex in 1936 not only permitted some of the first reconstructions of Carboniferous chondrichthyans, but details of the dentition suggested that these, too, could be allied with the bradyodonts (Moy-Thomas 1935, 1936). The punctate occlusal surfaces of the teeth of Helodus indicated the presence of the tubular dentine that was now the main defining characteristic of the group.
Moy-Thomas summarized his understanding of the bradyodonts in his 1939 review of early elasmobranchs and his textbook of the same year (Moy-Thomas 1939a, b). He placed the Bradyodonti on equal footing with the Selachii, and included within it the cochliodonts, holocephalians (Myriacanthids, Squaloraja and extant chimaeroids), petalodonts, psammodonts, edestids and chondrenchelyids. Ranking the Bradyodonti above the Holocephali was a novelty, as was the incorporation of Menaspis, a fish from the Permian of Germany, into the cochliodonts. This classification scheme was essentially accepted and repeated in the textbooks and summaries that appeared after the 1930s.
The first real challenge to the concept of the Bradyodonti came with the work of Leonard Radinsky (1937–85). Radinsky surveyed the tooth histology of a wide range of chondrichthyans in order to assess its validity as a taxonomic criterion (Radinsky 1961). Finding that the tissue called ‘tubular dentine’ is not a distinctive characteristic of taxonomic significance, but an adaptive characteristic that developed independently in several groups, he recommended that the bradyodonts be retained as a convenient repository for Smith Woodward's original four families, but based only on slow tooth replacement rates, ‘until new evidence allows a truer phylogenetic classification’ (Radinsky 1961, p. 80). Furthermore, the bradyodonts should be included with the chimaeroids in the Holocephali, thereby reducing their taxonomic rank. He suggested that the edestids be kept separate and noted that ‘major anatomical differences between petalodonts and cochliodonts suggests that the order Bradyodonti is an artificial grouping’ (Radinsky 1961, p. 81).
The further evidence that Radinsky anticipated slowly began to appear, firstly from the work of Colin Patterson (1933–98) on the Phylogeny of the Chimaeroids (Patterson 1965). Patterson reviewed chimaeroids and the fossil record of putative holocephalians with a view to commenting on the origins and evolution of the group. He combined detailed studies of the anatomy of the few specimens of fossil material consisting of more than just isolated tooth plates, with careful attention being paid to the histology of the dermal skeleton. His conclusions included the idea that the poorly known Carboniferous form, Deltoptychius, was allied with Menaspis, known from several specimens from the Kupferschiefer of the German Permian, and that these were most closely related to the Jurassic myriacanthoid holocephalians. He also suggested abandoning the bradyodonts as a formal taxonomic grouping since the two characteristics on which it was founded – slow tooth replacement and tooth histology – were unreliable (Patterson 1965, 1968).
There were certain quarters in which Patterson's conclusions did not sit well. From an initial discussion of the bradyodonts as a group, Svend Erik Bendix-Almgreen (died 2004) from the Geological Institute at Copenhagen University in Denmark concluded that it represented ‘a rather heterogeneous assemblage of late Palaeozoic elasmobranchs’ (Bendix-Almgreen 1968, p. 166). The discussion later centred around the interpretation of Menaspis – Bendix-Almgreen (1971) disagreed with Patterson's (1965, 1968) interpretation of most parts of the anatomy of the Permian fish and interpreted it as a somewhat anachronous relative and perhaps descendant of the rhenanid arthrodires, which could therefore not possibly be related to the myriacanthids or later chimaeroids. Ortlam (1985) went even further, declaring that Menaspis was the final survivor of the armoured placoderm fishes. The study of new material in the 1990s seemingly confirmed Patterson's main arguments and Menaspis is now included in its own monotypic order within the Holocephali (Schaumberg 1992; Stahl 1999).
Conclusions
In the intellectual generation following Smith Woodward's erection of the Bradyodonti, the concept was challenged and the recommendation emerged that the group should be abandoned – a not untypical state of affairs in the dynamic discipline of taxonomy. Where do things stand today? The closest equivalent to Smith Woodward's monumental systematic summary of the chondrichthyan fishes is presented in two recent volumes of the Handbook of Paleoichthyology (Stahl 1999; Ginter et al. 2010).
Firstly, in terms of diversity, the known fossil record of chondrichthyans in the Palaeozoic has increased dramatically since the work, firstly, of Agassiz and secondly of Smith Woodward (Table 1). This is only to be expected as new faunas have been uncovered and new studies carried out on existing collections. The figures indicate that an enormous increase in the number of species, genera and higher taxonomic units has taken place since the 1890s. Some of this increase is due to the use of bulk sampling and processing of Palaeozoic rocks for microvertebrate remains. These studies are in their infancy, however, and are by no means as advanced or extensive as similar sampling in the Mesozoic and Tertiary.
Table to illustrate how the perceived diversity of Palaeozoic chondrichthyans has changed from the 1830s to 2010
The discovery of new faunas, from both the Palaeozoic and the Mesozoic, has also enhanced our understanding of Carboniferous chondrichthyans. Beautifully preserved, articulated specimens, sometimes showing soft parts, have been collected from Carboniferous lagerstätten such as those of Mazon Creek (Pennsylvanian of Indiana), Bear Gulch (Mississippian of Montana) and Bearsden (Pendleian, earliest Namurian of Scotland). Specimens from these faunas have helped to solve various problems of synonymy, as well as giving some remarkable insights into the morphology of Palaeozoic species and providing more information by which to judge their relationships.
The current classification of Palaeozoic chondrichthyans retains chondrenchelyids, Helodus, cochliodonts, psammodonts, copodonts, menaspiforms and chimaeriforms within the Holocephali. The caseodontoids and edestoids have been removed to their own order, the Eugeneodontiformes, allied with the petalodontiforms and orodontiforms within the Subclass Euchondrocephali, which also embraces the Holocephali (Lund & Grogan 1997; Ginter et al. 2010).
The fact that our view of many groups still relies on isolated parts of the dentition means that the picture might well change again with the discovery of new material and careful re-examination of museum collections, especially using the technological innovations that permit imaging of specimens without the need for their destruction. Ginter et al. (2010, p. 133) have written, for example, that: ‘Despite their abundance, the petalodonts are still, with a few exceptions, poorly understood … The phylogenetic position of the petalodontiforms is not yet established with any confidence’. However, Arthur Smith Woodward might be gratified that the groups that he originally allied together in the Bradyodonti are still perceived as being closely allied, even though the Bradyodonti as a name has disappeared.
Acknowledgments
My work in Copenhagen was supported by a grant from the European Commission's (FP6) Integrated Infrastructure Initiative Programme SYNTHESYS (DK-TAF), which is very gratefully acknowledged. Dr Gilles Cuny (Lyon) kindly helped with the scanning electron microscopy and assisted in many ways during my visit to Copenhagen in June 2013. Dr Wayne Itano (Boulder, Colorado) showed a great deal of interest in the SEM project and very kindly provided specimens for the study from his personal collection. Dr David Ward (Orpington), the Trustees of the Natural History Museum and the Wellcome Library kindly allowed me to use their images in this paper. Finally, I should like to thank the editors of this volume for their patience regarding my submission.
- © 2016 The Author(s). Published by The Geological Society of London. All rights reserved
References
Cochliodonts and chimaeroids: Arthur Smith Woodward and the holocephalians
