- The Paleontological Society
The Cambrian fauna of Massachusetts, characterized by Paradoxides (Hydrocephalus) harlani, is poorly preserved. Better-preserved specimens, occurring within the stratigraphic range of this trilobite in southeastern Newfoundland and Morocco, provide a better understanding of such taxa to widen the scope of correlation. The paradoxidid-bearing Braintree Formation has revealed three trilobite species to add to those recently recorded. They represent Ellipsocephaloidea not previously recognized in this Massachusetts sequence, namely Protoleninae by Hamatolenus (H.) aff. H. (H.) marocanus and H. (Myopsolenus) aff. H. (M.) magnus and Ellipsocephalidae by Holocephalina aff. H. levis, thus strengthening faunal relationships with midshelf Cambrian sequences in Morocco and Spain. Reinterpretations of Agnostida, based on Newfoundland material associated with P. (Hydrocephalus) harlani, indicate that Condylopyge eli and Kiskinella cf. K. cristata signify a stratigraphic position for part of the Massachusetts sequence a little above the first appearance of Ovatoryctocara granulata. Appearance of this latter species is under review as the basal boundary of a global Cambrian stage, and is below the sequence break within the Chamberlain's Brook Formation in Newfoundland marked by the Easter Cove Blister Bed. The problem of differentiating species with numerous variable growth stages is highlighted in the ontogenies and stratigraphic ranges of P. (H.) harlani and P. (Plutonides) haywardi and emphasizes the importance of cephalic morphology in paradoxidid classification. Details of the holotypes of Agraulos quadrangularis and Braintreella rogersi and Czech topotypes of Agraulos ceticephalus supplement generic and specific characters poorly understood, especially those involving proportional differences between tectonically/taphonomically flattened examples and undistorted high-relief specimens.
T rilobites were collected from the Quincy-Braintree-Weymouth area of eastern Massachusetts (Fig. 1) as early as the 1830s and apparently widely distributed as natural history curiosities. In 1834, Jacob Green, without reference to its source, described one of them as Paradoxides harlani, which Barrande (Ordway, 1861a) thought synonymous with P. spinosus Boeck, 1827 from Bohemia and close to other paradoxidids in Sweden and Wales. It was not until 1856 that eastern Massachusetts was recognized as the source of Green's specimen, when Rogers exhibited similar trilobites from Old Hayward Quarry, Quincy. Ordway (1861b) noted two other species in the quarry, Arionellus quadrangularis Whitfield, 1884 and Ptychoparia rogersi Walcott, 1884. More details of Paradoxides harlani were given by Rogers (1856), Ordway (1861a), and Meek (in Dana, 1863). These descriptions were accompanied by comparisons, respectively, with Newfoundland species P. bennetti Salter, 1859, Agraulos socialis Billings, 1874, and Solenopleura communis Billings, 1874. Although these species occur above the range of P. harlani in Newfoundland (Fig. 2), the comparisons stressed the closer association of this Cambrian fauna with the Avalonian-Baltoscandian-Perigondwanan Faunal Realm than with the geographically nearer Laurentian Faunal Realm in western New England.
Raymond (1914, p. 237) remarked on the lack of a detailed study of the Massachusetts material when he described P. haywardi Raymond, 1914 from the Quincy quarry, and this was endorsed by Palmer (1971, p. 210). Fletcher (1972) examined Massachusetts material in several museum collections while identifying trilobites from Newfoundland and recognized P. harlani, P. haywardi, A. quadrangularis, and Braintreella rogersi-like forms in the basal Chamberlain's Brook Formation (Howell, 1925; Fletcher, 2003, text-fig. 3), well below a P. bennetti fauna in the same formation.
As a further means of establishing the relationships of the Newfoundland and Massachusetts trilobites, G. Lord guided Fletcher over the Boston Bay sections (Fig. 1). His collection and those of G. Zeoli and W. Bowman contained ellipsocephaloids, e.g., Kingaspis Kobayashi, 1935 and protolenids, as well as younger trilobites of the Badulesia tenera, Tomagnostus fissus–Ptychagnostus atavus, and P. davidis zones. Until his death, Lord corresponded with Howell, Theokritoff, and Fletcher about a comprehensive account of the Cambrian fossil sequence in the Boston Basin. With the benefit of his notes, we are able to add to the account by Geyer and Landing (2001).
Three trilobite species have been identified, in addition to those recognized in Lord's USNM collection by Geyer and Landing (2001). These, together with reinterpretations of two of their Agnostida identifications, endorse correlations made with Cambrian sequences in Morocco, Spain, Newfoundland, Siberia (Fig. 2), and Britain.
Preservation of the Massachusetts fossils under consideration is very poor. Most are rough casts of the external surface so affected by Lower greenschist metamorphic recrystallization and some tectonic stress that surface characters have been largely obliterated. Collecting localities coincide with areas where a weak regional cleavage parallels bedding planes, and most distortion is due to slight syndepositional compaction reflected by some flattening of the convex exoskeleton (Fig. 9.4) or to lateral movements along ubiquitous vertical joint planes (see Fig. 6.4). The four trilobite species originally described from this area were considered to lack the necessary details for adequate comparison and it has been suggested that such taxa should be recognized only in their source area, thus restricting a correlation with other regions. However, an assemblage of typical Massachusetts-type casts is recognizable elsewhere in southeastern Newfoundland sections associated with better-preserved exoskeletons, internal molds, and casts. These provide a better understanding of the Massachusetts taxa, in particular, of the paradoxidids.
Paleontologically, the most important aspect of Massachusetts collections is the many growth stages of Green's Paradoxides harlani that is now assigned to Paradoxides (Hydrocephalus) Barrande, 1846. Typically, proportional differences justifying a species are the result of comparing type specimens of significantly different size or lithology, without proper regard for obvious similarities. Growth stages of P. (H.) harlani demonstrate the inadequacies of some paradoxidid diagnoses.
We employ the established Massachusetts Cambrian lithostratigraphic nomenclature (Howell et al., 1936; Palmer, 1971; Lord, 1972; Nellis and Hellier, 1976) disagreeing with Geyer and Landing (2001), who contended that the Newfoundland lithostratigraphic scheme could be extended to Massachusetts by application of sequence stratigraphic principles.
Since Lord collected most of the available material, we examined his files of correspondence from 1920 to 1980. These yielded information on fossil localities within 3.25 km of Old Hayward Quarry at Quincy; in Hayward Creek; Fulmar's Ledge on White Neck, Idlewell; across from Jonas Perkin's School, Braintree; Braintree Dam; Canton; Cunningham Estate in East Milton; Monatiquot Valley; Bent's (Ruggles) Creek; Houghs Neck; at house number 1000 on Southern Artery; Revere Beach; Hull and Wollaston, Weymouth Landing, as well as Old Hayward Quarry. Unfortunately, this information is not easy to evaluate. On one occasion at least, B.F. Howell (letter dated 2 March 1954) dismissed material as pseudofossils, so there is a possibility that some may not be authentic fossil localities. Furthermore, it is not always clear whether fossils from a given locality are from in situ rocks or from spoil or clasts in glacial drift.
The catalogued fossil identifications were mostly by Howell and Lord, and their labels, giving a taxonomic assignment, horizon, and locality, accompany many of Lord's MCZ and USNM specimens. The overwhelming majority of the fossils are stated to be from Old Hayward Quarry; a few are from more fissile rocks at Idlewell that he considered younger than those in the quarry. Geyer and Landing (2001, p. 118), albeit at a loss to identify the precise Idlewell locations where Lord collected fossils, suggested that the railroad cutting there may have yielded the P. (H.) harlani specimen collected by Lord (1972). One of us (GZ) can state with confidence that Lord's main locality in Idlewell was on private land at the northern end of Idlewell, and the specimen in question probably came from this locality. Additionally, fossils were collected from trenches dug in the marsh at the northern end of Idlewell and trilobite fragments from dark gray shale in the vicinity of Third Street and Fourth Avenue, Idlewell (letter from Lord to B.F. Howell dated 22 September 1962). Specific sites are also identified at the Quintree Plaza opposite the shipyard and at Honey Hill in East Weymouth, where material from the transitional sequence linking the Lower (Callavia-bearing) and Middle (paradoxidid) Cambrian units, formerly exposed in the quarry, was dumped. Notable specimens were also collected from the drift, e.g., gravels behind the William Seach School in East Weymouth and at the City Service Oil Company, East Braintree.
STRATIGRAPHY AND GEOLOGIC SETTING
The Cambrian sedimentary rocks in the Boston Basin were predominantly fine- to medium-grained siliciclastics with subordinate quartzose sandstones and carbonates. They were deposited on the shelf of the Avalonian Microcontinent on the edge of the Iapetus Ocean, close to the western borders of Gondwana and Baltica.
Due to extensive cover by glacial drift, forests, and buildings in eastern Massachusetts, exposed rocks are sparse and stratigraphy of the Cambrian rocks proved difficult to establish. Mapping has not resolved this problem, but over the last 80 years diligent fossil collecting by local enthusiasts has possibly revealed some order to the exposed strata in numerous ephemeral manmade excavations, e.g., trenches, foundations, quarries, and road and rail cuttings. Prominent among these collectors, the late G. Lord was virtually their coordinator and his unpublished notes and letters contain important details of the stratigraphy of fossils collected from in situ rocks. Regrettably, numbered specimens in the collections do not appear in the notes and some museum catalogues do not record stratigraphic levels. An added problem is that many specimens are known to have been collected from blocks in the glacial drift or from spoil without reference in the catalogues; many specimens labeled “Hayward's Quarry” were actually collected from the quarry spoil, dumped at various places around Quincy, Braintree, and Weymouth.
All fossils associated with P. (H.) harlani in the Boston Basin that we have examined are preserved in a lithology unlike any in the Chamberlain's Brook Formation of Newfoundland. We concur with North (1971, p. 242) in his discussion on the formations of Avalonia in stating that “there is no clear evidence that all the Cambrian formations represented originally extended over the whole area within which the present outcrops appear, small though that area is. The Cambrian seaway may have been so shallow that at times it split up into a number of tortuously-connected sub-basins, controlled by features inherited from Pre-Cambrian erosion.” The Massachusetts rocks are metamorphosed and tectonized to a much greater degree than the more calcareous contemporaneous rocks in Newfoundland, as Crosby's rock descriptions (1880, p. 192) and photograph (1900, pl. 23) suggest, i.e., “gray and greenish-gray, somewhat siliceous, flinty, fine-grained and remarkably uniform slate with minute grains of pyrite, generally diffused and very rare development of cubic crystals. Very coarsely jointed. Stratification very massive and obscure, detected only by means of the trilobite remains which are assumed to lie in the plane of the bedding.” Despite former references to shales, slates, and mudstones, the essential matrix of the early paradoxidid Massachusetts fossils is a fine-grained psammite, betraying original coarse siltstone or fine sandstones. These are lithologies unknown in the type Chamberlain's Brook Formation on the western side of St. Mary's Bay (Fletcher, 2003, text-figs. 1, 3), thus calling into question the application of the Newfoundland lithostratigraphic nomenclature outside that province (Landing and Westrop, 1998). Lithostratigraphic codes (e.g., Hedberg, 1976) indicate that a formation is defined by its lithologic composition, the character of its base, and an upper limit defined by the character of the base of the succeeding formation. In Newfoundland, the Chamberlain's Brook Formation (Howell, 1925, p. 60) is essentially a fine-grained, calcareous mudstone sequence defined by a conspicuous manganiferous mudstone sequence at the base and by the extensive white Manuels Metabentonite clay (Howell, 1925, bed 36, p. 50, 59; Fletcher and Brückner, 1974, fig. 9) at the top marking the base of the Manuels River Formation (Hutchinson, 1962, p. 22). No comparable manganiferous or clay beds are developed in New Brunswick or Massachusetts and there is no basis for applying the Newfoundland lithostratigraphic nomenclature to those successions; the precise sequence-stratigraphy break taken by Landing (1996) to define the base of an extensive Chamberlain's Brook unit is marked by the manganiferous Easter Cove Blister Bed (Fletcher and Brückner, 1974, fig. 9) within that formation in the type area of Newfoundland (Fletcher, 2003, text-fig. 7; Figs. 3, 5). As interpreted in this account, the P. (H.) harlani fauna of Massachusetts is correlated with the Newfoundland Kiskinella cristata Zone below that main sequence break at the top of the zone. We do not recognize that there is lithologic correspondence with the Newfoundland sequence and consider the contemporaneous coarser-grained sequences of southern Morocco and Iberian Chains in northern Spain to have much more in common lithologically with the Massachusetts sequence and to have been deposited closer than the 900 miles-distant Newfoundland sequence. However, this fact does not warrant a common nomenclature and the traditional Massachusetts scheme should be maintained. It may be emphasized also that Geyer and Landing's (2001) use of Middle Cambrian refers to the time-scale for paradoxidid sequences, since the Cambrian Subcommission has yet to resolve problems of inter-realm correlation, that indicate substantial overlapping of their Middle Cambrian units (Robison et al., 1977; Fletcher, 2001). The current proposal before the Subcommission on Cambrian Stratigraphy to define the base of a Middle Cambrian unit at the incoming of Oryctocephalus indicus (Reed, 1910) would assign the P. (H.) harlani fauna to the Lower Cambrian (Fletcher, 2003, tab. 1).
Geyer and Landing outlined the pattern of lithofacies belts, differentiating the coarser-grained sediments of the inner shelf from the finer-grained, commonly calcareous, deposits of the outer shelf. The Massachusetts early paradoxidid rocks represent a more inshore setting than the much finer grained contemporaneous rocks of southeastern Newfoundland and resemble the Kingaspis and Hamatolenus-bearing rocks in the Moktar area of southern Morocco (British Geological Survey, 2002) and around Daroca in the Iberian Chains (Gozalo and Liñán, 1998) containing early paradoxidids (Fig. 2). However, one would not apply a Moroccan or Spanish lithostratigraphic nomenclature and the original Massachusetts scheme is appropriate.
Lord recognized three Cambrian formations in Massachusetts, in ascending order, Weymouth, Braintree Argillite, and Monatiquot; the first two names were employed in the account of Nellis and Hellier (1976). The Weymouth Formation represents the local pre-paradoxidid sequence and rests unconformably upon decomposed Precambrian granitic rocks in the King Oak Hill area (Fig. 1, locality 6). It comprises a lower unit of gray green mudstones and arkosic-quartzitic sandstones 38 m thick (informally named Avonia Beds) and an upper unit subdivided into three parts (the lower and middle parts informally named Mill Cove Shale) of red, purple, and green calcareous mudstones with incipient limestone lenses, concretions, and five prominent intervals of limestone concentrations (numbered 1–5). “Limestone No. 1” is 3 m thick and marks the base of the upper unit. “Limestone No. 2” is 2 m thick and lies 41 m above “Limestone 1.” “Limestone No. 3” (also informally named Guide Line Bed) contains Callavia Matthew, 1897 and is 12 m thick. It lies 60 m higher at the top of the lower part of this unit and forms the locally famous “Olenellus Ledge” at Pearl Street, North Weymouth. The middle part of the upper unit comprises green and purple shale, 153 m thick, with a 2-m concentration of thin Callavia limestones (“Limestone No. 4”) in the middle. It extends upwards to a disconformity marking the base of the upper part. This contact is overlain by thin beds of gray and reddish brown sandstone and shale at the base of “Limestone No. 5” (also informally named Attaquatock Limestone Beds); the latter comprising 30 m of interbedded, finely cross-bedded pink limestone and thin purple shale beds with hyolithids and orthothecids. The uppermost part of the unit is 76 m thick and the green and gray shales Lord informally named Great Swamp Beds.
Lithologically, the lower unit and lower two parts of the upper unit of the Weymouth Formation are similar to the Bonavista and Brigus Formations of southeastern Newfoundland (van Ingen, 1914) and reflect similar outer-shelf depositional settings, unlike the contemporaneous coarser-grained inner-shelf sedimentary rocks of the Glen Falls and Hanford Brook Formations of New Brunswick (Hayes and Howell, 1937).
The Weymouth Formation is generally considered to be unconformably overlain in the Weymouth area by the coarser-grained Braintree Formation, but, as the latter is vertical thereabouts, some fault contacts may occur locally and any unconformity is difficult to prove.
From the available notes and fossils recorded from the lowest part of the paradoxidid-bearing Braintree sequence, there may be at least three separate stratigraphic intervals involved. Disconformable to the underlying Great Swamp Beds, local collectors recognize the Ruggles (Bent) Creek Beds and this is succeeded by the transitional sequence beneath the Fore River Shales in which the Old Hayward Creek Quarry was excavated.
Higher parts of the Fore River Shales are marked by red and purple beds that yield younger paradoxidids, e.g., Paradoxides (Eccaparadoxides) eteminicus (Matthew, 1882) (Fig. 2). Generally, Braintree Formation sediments indicate a more inshore-shelf environment than the underlying Weymouth Formation. The minor breaks in sequence and coarse-grained rocks in the upper Weymouth and lower Braintree sequences indicate periodic shallowing and possible erosion associated with a widespread lowering of sea level (see Fig. 2 explanation) prior to a gradual deepening on the shelf that may have continued into P. davidis Zone time. However, the lack, in Massachusetts, of fossil assemblages between the ranges of P. (H.) harlani and P. (E.) eteminicus in Newfoundland (Fig. 2) may reflect the limited exposure of the Braintree Formation or a major nonsequence covering the interval between two well-marked sequence breaks within the Chamberlain's Brook Formation, i.e., those signified by the Easter Cove Blister Bed immediately underlying the Wester Cove Member and the prominent limestone just above the base of the Hartella Zone in the Deep Cove Mudstone Member (Fletcher, 2003, text-fig. 3). A nonsequence of such a proportion is manifest in New Brunswick between the Hanford Brook and Fossil Brook Formations (Fletcher, 1972, fig. 21).
Overlying the Braintree Formation, Lord recognized three members of the Monatiquot Formation, all of a deeper-water lithofacies. His lowest Taber's Ledge member is represented by dark gray shaly mudstone and thin quartzites with minor dark limestones containing a T. fissus–Ptych. atavus zonal assemblage marked by “Paradoxides hicksii” (Salter, 1866) and Condylopyge rex (Barrande, 1846); the middle Perkins School member by blackish shaly mudstone with thin quartzites and a P. davidis zonal assemblage; and the upper part by dark gray shaly mudstone with quartzitic ripple-bedded sandstones forming the Old Reservoir member.
Wherever recognized, the Monatiquot Formation is unconformably overlain by the Upper Cambrian Green Lodge Formation.
FOSSILS FROM OLD HAYWARD QUARRY
We have found no published record of the stratigraphic ranges of trilobites in the quarry other than Shaler's (1871, p. 174) statement that trilobites occur within a 30-m section at the top of the quarry. Lord noted that fossils occurred on the northern and southern sides of the quarry and wrote that when he first met Hayward in 1920, he was told that he got a specimen of “P. haywardi antiqua” when looking for a new place to find trilobites north of the quarry, near the mouth of Bent's (Ruggles) Creek.
Any interpretation of this information in terms of stratigraphy has to be tentative, because there are no data on folding and faulting there. Shaler's and Lord's statements can be interpreted in two ways, either one trilobite-rich level is repeated by faulting or there are two fossiliferous levels.
In discussing the stratigraphic relationship of the Massachusetts paradoxidid strata to the Mill Cove Olenellus strata, Foerste (1889) stated that “the Paradoxides of the Braintree Quarry all lie on their backs in the argillite as tilted at the present time. Observations on the larger forms of this genus elsewhere agree pretty well that this is a common method of occurrence of these trilobites.” However, Lord could not agree and told Howell that “Foerste happened to be working in a part of the quarry where they were mostly that way. Because of folding they were not all upon their backs. I sent you three on one block of rock, one above the other. One was not on its back.” This description agrees with what we see in the specimens and in the available exposures: that the lithology in the quarry is massive and uniform is confirmed by a photograph (Crosby, 1900, pl. 23).
FAUNA AND CORRELATION
Trilobites from the lowest part of the Braintree Formation include Condylopyge aff. C. eli Geyer, 1998, Kiskinella cf. K. cristata Romanenko and Romanenko, 1962, *Paradoxides (Hydrocephalus) harlani (Green, 1834), *P. (Plutonides) haywardi (Raymond, 1914), *Kingaspis avalonensis Geyer and Landing, 2001, Hamatolenus (H.) aff. H. (H.) marocanus (Neltner, 1938), H. (Myopsolenus) aff. H. (M.) magnus Hupé, 1953, Holocephalina aff. H. levis Gozalo and Liñán, 1996, *Agraulos quadrangularis (Whitfield, 1884) and *Braintreella rogersi (Walcott, 1884). An asterisk indicates species originally described from the Quincy-Braintree-Weymouth area.
The main problem in dealing with the taxonomy and correlation of the Massachusetts material is its very poor preservation. If all taxa are restricted to the region, no reliable correlation with other regions is possible. However, since biozones, representing intervals of overlapping fossil ranges, are recognized as a means of determining relative ages of rock sequences, the spectrum of associated taxa for any particular taxon limits the range of possible correlative elements. Consequently, if the full range of that taxon, e.g., Paradoxides (Hydrocephalus) harlani, is reliably identified in a continuous monofacial sedimentary sequence marked by a succession of biozones, comparisons of poorly preserved elements associated with it may be made to better-preserved forms occurring within its known range elsewhere. In the Easter and Wester coves of Branch Cove, southeastern Newfoundland (Fletcher, 2003, text-fig. 1), well-preserved specimens of various taxa occur within the stratigraphic range of P. (H.) harlani. Some of these taxa occur in the early paradoxidid sequences of Morocco, Spain, Siberia (Fig. 2), and Britain. We compare the Massachusetts material to such Newfoundland taxa to justify our taxonomy and correlations, though, initially, it is imperative that a reliable understanding of P. (H.) harlani is established.
In contrast to the relative paucity of specimens of other elements in the Massachusetts fauna, numerous specimens of all holaspid growth stages of the gigantic P. (H.) harlani display a considerable array of cephalic variation, recognition of which indicates possible synonymizing of some later described species, e.g., P. groomi (Lapworth, 1891) from England and Acadoparadoxides briareus Geyer, 1993 from Morocco, and differentiation of Raymond's P. (Plutonides) haywardi.
The preponderance of polymerid trilobites is indicative of a shelf environment. Nevertheless, the presence of Agnostida implies some contact with the open ocean. The fauna and lithology indicate that the Massachusetts sequence developed slightly seaward of an environment favored largely by ellipsocephalids and protolenids, as represented by the Hanford Brook Formation of New Brunswick (Westrop and Landing, 2000). It seems to be from the same shelf setting as the southern Moroccan and Iberian Chains sequences containing related newly recognized Massachusetts elements such as Kingaspis, Hamatolenus Hupé, 1953, and Holocephalina Salter, 1864, but well inshore of the sequence deposited in the southwestern region of the Avalon Peninsula of southeastern Newfoundland in which ocean-transgressing Agnostida like Peronopsis Hawle and Corda, 1847, Condylopyge Hawle and Corda, 1847, and Kiskinella Romanenko and Romanenko, 1962 are relatively abundant.
Precise inter-regional correlation is somewhat problematic (Fig. 2), because certain species have long stratigraphic ranges. For instance, Hamatolenus (H.) marocanus (Neltner, 1938) and H. (Myopsolenus) magnus Hupé, 1953 occur in both the Cephalopyge notabilis and Ornamentaspis frequens zones in Morocco and there is no information on any pertinent morphologic changes within their ranges to indicate a zonal position. In the case of our poorly preserved Massachusetts ellipsocephaloid specimens, we are unable to assess their precise stratigraphic positions as is possible for the different ontogenetic representatives of Condylopyge eli Geyer, 1998.
Geyer and Landing (2001, p. 123) considered Kingaspis avalonensis Geyer and Landing, 2001 close to K. maroccana (Gigout, 1951) from the Moroccan O. frequens Zone and one of us (TPF) collected a form close to K. avalonensis (Fig. 9.1) from sandstones in the basal part of the Groupe des Feijas internes (Geyer, 1989; British Geological Survey, 2002) at Moktar, northeast of Tiznit in southwestern Morocco associated with Paradoxides (Hydrocephalus) harlani and solenopleurid specimens similar to B. rogersi (Fig. 13.15, 13.16).
Holocephalina levis Gozalo and Liñán, 1996 from the Spanish P. (E.) asturianus Zone was the earliest stratigraphic record of the genus (Gozalo and Liñán, 1996, abst.). We consider the generic assignment is acceptable and that the Massachusetts material resembles the Spanish species. Although there is no precise information on the stratigraphic position of the Massachusetts specimens, the similar Newfoundland form, H. aff. H. levis (Fig. 10.4), with P. (Hydrocephalus) harlani in the Kiskinella cristata Zone of the Easter Cove Manganiferous Mudstone Member (Fletcher, 2003, text-fig. 7) indicates that the Spanish occurrence may not be the earliest Holocephalina and that the Massachusetts and Newfoundland material may represent an older species. This raises the problem of identifying the USNM specimens assigned to Dawsonia Hartt in Dawson, 1868 by Geyer and Landing (2001, p. 121). Such an assignment, however, would greatly extend the range of Dawsonia, known elsewhere only from equivalents of the Triplagnostus gibbus Zone of Scandinavia. We suggest that this material belongs to Kiskinella, for reasons given below.
We conclude that the collections made from the Braintree Formation represent elements of an assemblage correlative with the Kiskinella cristata Zone of Newfoundland [=Pagetides division (Bengtson and Fletcher, 1983, fig. 5); Pagetides n. sp. interval (Geyer and Landing, 2001, fig. 2, W. AVALONIA)] and the Moroccan O. frequens Zone, below significant breaks in sequence or disturbances marked by the Easter Cove Blister Bed in the former and the Brèche à Micmacca in the latter (Fig. 2). Correlations are possible with a late Oryctocara to early Kounamkites zonal position in Siberia and the Olenellus-bearing Pagetides elegans faunule (Rasetti, 1967, p. 20) of the New York Taconic region (Fletcher, 2001, tab. 1). These correspond to positions immediately above the first appearance of Ovatoryctocara granulata Chernysheva, 1962 as occurring in Siberia (Egorova et al., 1976), Greenland (Blaker and Peel, 1997), South China (Zhao Yuanlong, personal commun., 2001), and southeastern Newfoundland (Fletcher, 1972; Bengtson and Fletcher, 1983), currently under review as the basal boundary of a global Cambrian stage, and more or less coincident with the base of the Amgaian Stage in Siberia (Fig. 2; Fletcher, 2001, 2003). Correlation of the Kiskinella Zone with the middle of the Spanish Paradoxides mureroensis Zone containing Paradoxides (H.) harlani and P. (Plutonides) Hicks, 1895 is likely at levels correlative with faunas occurring below disturbances and sequence breaks in Morocco and Newfoundland (BM and BB in Fig. 2). However, the drift boulders yielding the ellipsocephaloids were considered by Lord to derive from the Great Swamp Beds at the top of the Weymouth Formation and, therefore, might be slightly older, i.e., possibly equivalent to beds in the lower parts of the Cephalopyge zones of Morocco and Newfoundland, and the Acimetopus [Calodiscus granulosa Egorova and Shabanov in Savitskiy et al., 1972]–bearing Toyonian sequences of Siberia (Savitskiy et al., 1972; Fig. 2).
Specimen number prefixes
AMNH (American Museum of Natural History); BMNH (The Natural History Museum, London); M-F (T. P. Fletcher/British Geological Survey 1999 Moroccan Collection, Keyworth, England); GZ (G. Zeoli private collection at the Plympton Historical Society, 189 Main Street, Plympton, Massachusetts 02189); MCZ (Museum of Comparative Zoology, Harvard University); MPZ (Museum of Palaeontology, University of Zaragoza), SM (Sedgwick Museum, University of Cambridge); ROM (Royal Ontario Museum, Toronto); USNM (Smithsonian Institution, Washington DC).
Our classification largely follows that outlined by Fortey (1997) and Whittington et al. (1997).
Class TrilobitaWalch, 1771
Order Agnostida Salter, 1864
Suborder Agnostina Salter, 1864
Family Condylopygidae Raymond, 1913a
Genus Condylopyge Hawle and Corda, 1847
Battus rex Barrande, 1846, p. 17, from the Paradoxides (Eccaparadoxides) pusillus Zone at Týřovice, Bohemia, Czech Republic.
Basal glabellar lobes transverse, separated by a median plate (rather than joined by connective band) bearing median tubercle or spine; anterior glabellar lobe laterally expanded, well rounded to slightly truncated in front. Pygidial axis parallel-sided to medially constricted, with an unfurrowed to triannulate anteroaxis bearing sagittal nodes or ridge; postaxial field undivided or divided.
Changes in pygidial structures suggest an evolutionary lineage from Condylopyge amitina Rushton, 1966 in the English equivalent of the Hupeolenus Zone (i.e., with a Chelediscus acifer and Tannudiscus balanus assemblage as in Newfoundland), through variants of C. eli in equivalents of the Cephalopyge, Kiskinella, and P. (Hydrocephalus) harlani zones (with the variant C. cruzensis Liñán and Gozalo, 1986 developed in Spain), to C. carinata Westergård, 1936 in the P. pinus Zone equivalents, to C. rex in the Tomagnostus fissus–Ptych. atavus Zone, largely manifest in the pygidium. Variation in the growth stages of the cephalon, involving reduction in sagittal length of the preglabellar field and shape change in the anteroglabellar lobe, completely masks specific recognition. The one exception is in C. amitina, in which the preglabellar field remains long throughout growth (Rushton, 1966, p. 30). Ontogenetic stages of pygidia from equivalents of the Scandinavian P. pinus and younger zones reflect the possible phylogenetic trends involving gradual change from a relatively short parallel-sided unfurrowed axis to a prominently medially constricted, furrowed axis, and from a postaxial field initially divided by a median furrow and later by the more pointed posterior extension of the axis.
Condylopyge eliGeyer, 1998Figure 3.1–3.14
Condylopyge species indeterminate Geyer and Landing, 2001, p. 120, fig. 3.1–3.4.
Family uncertain, genus and species indeterminate Geyer and Landing, 2001, p. 120–121, fig. 3.5.
USNM 506939a, 506940–506942 from Massachusetts and over 100 specimens from southeastern Newfoundland.
Comments here are restricted to features of the pygidium, since these appear to provide the best guide to species differentiation. We have not recognized any better Massachusetts specimens than those figured by Geyer and Landing (2001, fig. 3). We offer additional observations on that material in the light of similar material elsewhere.
Condylopyge specimens from the Kiskinella Zone on Cape St. Mary's Peninsula in southeastern Newfoundland (Fletcher, 2003, text-figs. 1, 7) are figured here to demonstrate growth-stage variants of C. eli that indicate proportional differences, without reference to size, are not necessarily reliable for specific identification. Despite the extremely poor preservation of Massachusetts material, the axial characters of USNM 506942 and USNM 506939a are sufficient to regard them as closer to the same-sized Newfoundland and Moroccan examples of C. eli than to any other known species. Pygidium USNM 506939a has a furrowed axis much longer than C. amitina and is more parallel-sided than C. carinata and C. rex. As in C. eli, the postaxial field is divided.
Fletcher (2003, p. 88) discussed the stage development of C. eli through the Cephalopyge and Kiskinella zones in southeastern Newfoundland. The Massachusetts material is considered to have affinity with the Newfoundland taxon.
The synonymous Siberian pygidium, from about 3.25 m above the base of the Kounamkites Zone (Egorova in Egorova et al., 1976, level 15/12a in River Nekekit Section, p. 25 pull-out), appears to be slightly different from the younger type material of her subspecies from the top of the zone (Egorova, 1972, p. 59, pl. 3, figs. 8–10; 1976, pl. 52, fig. 11, level 163 in River Nekekit Section [Egorova, 1972, fig. 3]). The pygidial axis is more parallel-sided and the postaxial field is conspicuously divided; features resembling a similarly sized stage of C. eli at the top of its range in southeastern Newfoundland. Note that Egorova recorded the range of her subspecies as Oryctocara–Kounamkites zones, but only a cephalon from the Oryctocara Zone is figured (1976, p. 59, pl. 43, fig. 19).
The overall aspect of the Spanish C. cruzensis compares well with C. eli, differing essentially in the swollen anterior parts of the pleural fields of its pygidium, as in the younger C. regia (Sjögren, 1872) in Sweden. However, the holotype of C. eli (Geyer, 1998, pl. 1, fig. 5) and some Newfoundland specimens (Fig. 3.7) exhibit incipient swellings in the same region of the pygidial pleural fields, perhaps to indicate a close relationship with Spanish material in the P. mureroensis Zone.
Suborder EodiscinaKobayashi, 1939
Family Eodiscidae Raymond, 1913a
Genus Kiskinella Romanenko and Romanenko, 1962
Kiskinella cristata Romanenko and Romanenko, 1962 from the Oryctocara Zone of Gorno–Altayskaya (Kiska River), Russia.
See Fletcher, 2003, p. 96.
Kiskinella cristataRomanenko and Romanenko, 1962Figure 4.1, 4.2, 4.4–4.9
Pagetides myrmecophagus n. sp.; nomen nudum, Fletcher in Egorova et al., 1976, tab. 2, p. 47.
As for the genus.
Figured specimens ROM 56106–56109 and over 30 specimens from southeastern Newfoundland.
This species is conspicuous in the Siberian Kounamkites Zone with Condylopyge eli (Egorova et al., 1976, pl. 50, fig. 12), and Peronopsis roddyi (Resser and Howell, 1938; Egorova et al., 1976, p. 66, pl. 44, figs. 22, 23; Fletcher, 2003, p. 96), all of which are associated with Paradoxides (H.) harlani and P. (Pl.) aff. Para. (Pl.) haywardi in the Kiskinella Zone at the base of the Chamberlain's Brook Formation in Branch Cove, southeastern Newfoundland. For comparative purposes, a selection of specimens of the Newfoundland form is figured here.
Kiskinella cf. K. cristataRomanenko and Romanenko, 1962Figure 4.3
USNM 506924Aa-e and USNM 506924Ba-e from Old Hayward Quarry.
Geyer and Landing (2001) compared these specimens to Dawsonia, because they lack eyes and sutures, but recognized that they are not preserved well enough to permit an unequivocal determination. Such assignment, however, would greatly extend the range of Dawsonia, known elsewhere only from equivalents of the Scandinavian Tripl. gibbus Zone. Our alternative suggestion is that they belong to Kiskinella, which differs from Dawsonia in possessing eyes and a proparian facial suture. Rasetti's (1952, pl. 54, figs. 1–9) figures of Dawsonia dawsoni Hartt in Dawson, 1868 display a coarse granulate surface ornament and their genal regions, although inflated, are not sufficiently convex and swollen to extend (exsag.) over the lateral cephalic border as is well marked in the Massachusetts material. The lack of palpebral lobes and facial sutures on the poorly preserved Massachusetts specimens is evident.
Having examined the specimens figured by Geyer and Landing (2001) and made latex casts of the external molds, we observe that the smoothness of the test cannot be attributed to abrasion, since the border scrobicules are perfectly preserved. Therefore, the lack of surface granulation is a significant distinction. Continuations of the cephalic border are not preserved in the critical areas where palpebral lobes might occur and it is not possible to be certain that these forms lack palpebral lobes and facial sutures. This being the case, it is possible that palpebral lobes have been broken away. In the basal Chamberlain's Brook Formation in southeastern Newfoundland (Fletcher, 2003, text-fig. 3), many specimens of the smooth Kiskinella cristata have their palpebral lobes broken away and resemble the Massachusetts specimens. Newfoundland specimens (Fig. 4.1, 4.2, 4.4–4.9) demonstrate that the traces of glabellar lobation of the USNM specimens match those on internal molds of the Newfoundland Kiskinella rather than the relatively featureless glabella of Dawsonia (compare Fig. 4.3 and 4.8 with Rasetti's 1952, pl. 54, figs. 1, 2, 4).
Order RedlichiidaRichter, 1932
Suborder Redlichiina Richter, 1932
Superfamily Paradoxidoidea Hawle and Corda, 1847
Family ParadoxididaeHawle and Corda, 1847
Genus ParadoxidesBrongniart, 1822
Glabellae widening forward to L4, with rounded anterior margin (sometimes semicircular) very slightly pointed or merely an arc of a circle; S0–S2 well impressed, S3 and S4 impressed, incipient or absent; anterior and posterior sections of the facial suture moderately to strongly divergent; fixigenal fields wide to narrow; cephalic border narrow (sag.) and upstanding to flat and wide; palpebral lobes wide or narrow (tr.), short or long; rostral plate and hypostome fused or unfused.
Thorax of 14–21 segments; segments ending in short or long pleural spines progressively directed more strongly backwards from front to rear; last three to five pleura may gradually increase or decrease in length and extend beyond the posterior pygidial margin.
Pygidia small, subcircular, quadrate (wider than long or longer than wide), spatulate (with or without short spines along the posterior margin) or forked backwards; axis with one to three rings.
Surface ornament smooth, variably granulate/tuberculate or finely reticulate; terrace ridges on raised anterior border, anterior glabellar lobe, rostral plate, and hypostome.
Since the genus was proposed, numerous attempts to subdivide its members have not been generally accepted. Largely following Šnajdr (1957, 1958), it has been differentiated on the basis of ontogenetic similarities of protaspid and meraspid forms, because the distinctions became less well defined in holaspids exhibiting a degree of morphological convergence. Theoretically, this approach, recognizing that related taxa commonly display a greater morphological similarity among larval stages than among adults, is sound. However, in practice, the lack of early ontogenetic specimens for the great majority of known forms renders it virtually impossible to establish species groups, especially when one considers holaspid convergences. Based upon various combinations of characters outlined by Šnajdr (1958, fig. 17), such as glabellar shape and lobation, size of palpebral lobes, and outline shape of the pygidium, there are relatively few problems in defining a paradoxidid species. The characters of the hypostome and rostral plate also vary, but specimens are rare or unknown for some forms. Except to differentiate Paradoxides (Paradoxides), in which they are characteristically fused together, such a rarity also renders it difficult to use these characters for major subdivisions.
The range of possible combinations within Paradoxides is much greater than that proposed by Šnajdr in his recognition of four genera and it is debatable whether there is a need for generic or subgeneric subdivision. However, since Šnajdr's subdivisions are based upon his ontogenetic groups, we follow Solov'yev (1969), Dean and Rushton (1997), and Geyer and Landing (2001, p. 124) in recognizing subgenera of Paradoxides.
As noted by Geyer and Landing (2001, p. 123), Dean and Rushton (1997, p. 471) modified the generic concept of Šnajdr (1957, 1958) without overstressing the significant, gradual changes that may occur through holaspid ontogeny. These are most evident in species developed to giant size, e.g., the early representatives from Massachusetts, in which certain stages of cephalic morphology coincide with shapes taken to be diagnostic of other proposed genera or subgenera. Like Geyer and Landing (2001), we consider the evidence from Massachusetts inadequate for revision of Paradoxides sensu lato and, as an interim measure, employ subgenera based upon the cephalic characters of Šnajdr (1958, fig. 17).
Intraspecific variation must be considered in a family revision. Four main aspects are important, i.e., ontogenetic variation in glabellar shape and glabellar furrows (Figs. 5, 6.1, 6.2, 6.4, 7.16, 7.17, 8), the anterior cranidial border, relative proportions of the exoskeleton, i.e., slender (Fig. 6.1, 6.3) or broad (Fig. 6.2, 6.4), and the constancy of the pygidium (Fig. 7.3–7.11) through growth following modifications of the most immature (Fig. 7.1–7.3).
Four species groups based on pygidial shape are: 1) a subcircular form typified by P. sacheri Barrande, 1852 of Šnajdr (1958, pl. 16, fig. 10); 2) a spatulate form typified by P. pusillus Barrande, 1846 of Šnajdr (1958, pl. 22, fig. 11); 3) rather quadrate forms that may be squarish e.g., P. micmac Hartt in Dawson, 1868, or longer than wide, e.g., P. eopinus Solov'yev, 1969, or wider than long, e.g., Hydrocephalus carens Barrande, 1846 as figured by Šnajdr (1958, pl. 26, fig. 11); and 4) some younger forms have a forked posterior margin, e.g., P. rugulosus Hawle and Corda, 1847 and the youngest are commonly considerably sagittally stretched like P. (P.) davidis davidis Salter, 1863 (Bergström and Levi-Setti, 1978).
Except for the subcircular pygidial group, modifications within these groups may involve an axial kinking forward of the otherwise transverse or rounded posterior margin. In early kinked representatives, this is generally slight, but later ones are more strongly marked to herald the prominently forked forms. Incipient-prominent spines may also characterize the posterior margin, e.g., P. oelandicus Sjögren, 1872 and, except for the subcircular group, there is a tendency in the pygidial outlines of the youngest representatives of a group to be sagittally stretched, e.g., P. davidis davidis Salter, 1863 and P. forchhammeri Angelin, 1852 (see also Angelin, 1878).
Other pygidial characters of possible specific rank are the segmentation of the axis and its relative length, although in some cases these may betray ontogenetic changes.
Considering the above and specimens from Massachusetts, the shortcomings of Šnajdr's generic groupings can be highlighted. The cranidium of the holotype of P. sacheri (see Šnajdr, 1958, pl. 17, fig. 1) has a narrow, upstanding anterior border and wide (tr.) preocular fields [holaspid growth stages show that the preocular fields gradually widen (Šnajdr, 1958, pl. 16, figs. 1–7)], a widening-forward glabella with only S0, S1, and S2 well defined, palpebral lobes extending almost to the posterior border furrows, and a spatulate pygidium with only the articulating half ring and first axial ring well defined. In large individuals from Massachusetts and Newfoundland referred to P. (Acadoparadoxides) harlani and large forms from Morocco referred to A. briareus, pygidia are virtually identical to those isolated subcircular forms also assigned to sacheri by Šnajdr (1958, pl. 16, figs. 6, 8–10). However, their cranidia differ from sacheri and resemble Hydrocephalus Barrande, 1846, e.g., H. lyelli (Barrande, 1852) (Šnajdr, 1958, pl. 18, figs. 5, 6), in that their anterior borders are wide, flat, and barely differentiated by furrows, their preocular fields are narrow (tr.), their glabellae widen markedly to L4 (generally with traces of S3 and S4), and their palpebral lobes end well short of the posterior border furrows. Such combinations of Hydrocephalus-type cranidia and sacheri-type pygidia do not fit any of Šnajdr's genera.
Until this family is properly researched, we give precedence to cranidial features in recognizing Šnajdr's (1958) cephalic types as P. (Eccaparadoxides) Šnajdr, 1957, P. (Hydrocephalus), P. (Paradoxides) Brongniart, 1822, and P. (Plutonides) Hicks, 1895. The latter, incorporating both short- and long-eyed forms, is regarded as the senior synonym of Acadoparadoxides Šnajdr, 1958 and Baltoparadoxides Šnajdr, 1986.
Two paradoxidid species occur at unspecified levels in Old Hayward Quarry, Quincy. Paradoxides harlani grew to 27 cm in length, with glabellar furrows S3 and S4 variously impressed, a flat anterior cranidial border, and a subcircular pygidium; “P. haywardi” Raymond, 1914 has only S1 and S2 well impressed, a narrow, upstanding, anterior cranidial border, and more ovate pygidium. We assign the former to P. (Hydrocephalus), the latter to P. (Plutonides).
Subgenus HydrocephalusBarrande, 1846
Phlysacium Hawle and Corda, 1847, p. 16.
Paradoxides (Eoparadoxides) Solov'yev, 1969, p. 14.
Rejkocephalus Kordule, 1990, p. 55.
Hydrocephalus carens Barrande, 1846 from the Paradoxides (Eccaparadoxides) pusillus Zone near Skryje, Czech Republic.
Glabella widening markedly to L4, with anterior margin less than semicircular; S0-S2 transglabellar, S3 and S4 short, shallow; anterior cranidial border flat, not well defined by impressed border furrows, widening (sag.) markedly abaxially; in more mature stages, border furrows barely discernible.
Paradoxides (Hydrocephalus) harlaniGreen, 1834Figures 5, 6, 7.1–7.11, 7.13–7.17
Paradoxides harlani Green, 1834, p. 336; for references to the Massachusetts material (with the exception of Walcott, 1884 and Raymond, 1914, p. 237, unnumbered pl., figs. 1, 2, 7), see Geyer and Landing, 2001, p. 124.
Acadoparadoxides briareus n. sp. Geyer, 1993, p. 80, pls. 6–9, 10.1– 10.3, 11.3.
Exoskeleton narrow (Fig. 6.1) or broad (Fig. 6.4), up to 27 cm long; glabella with an evenly rounded anterior margin; palpebral lobes medium-sized; genal spines long, extending posteriorly as far as the tenth thoracic segment; thorax of 17 or 18 segments of which the last four pleural spines commonly extend beyond the posterior pygidial margin and gradually decrease in length; pygidium subcircular, subquadrate, with one well-defined ring on a relatively long, bluntly pointed, triangulate axis.
See Geyer and Landing (2001, p. 124), but since their description incorporated features of material we consider to represent P. (Plutonides) haywardi, it is necessary to disregard those diagnostic elements of haywardi, such as its upstanding anterior cranidial border and ovate pygidium in their description of Paradoxides (H.) harlani.
Smallest cranidial stage (Fig. 5.1) about 2 cm long; glabella expanding forward from a relatively narrow (tr.) occipital ring 1.3 cm wide (tr.) to 2.3 cm wide at S4. S0–S4 evident; S0 and S1 transglabellar; anterior glabellar margin a rounded arc. Anterior border not preserved; fixigenal fields conspicuously narrow, particularly opposite S4; glabellar width 82 percent of cranidial width at S4. Palpebral lobes wide, extending backwards from L4 to the level of S1; posterior limb of facial suture diverging sharply outwards. A slightly larger specimen, 3 cm long (Fig. 5.4), closely matches this general form, but with the flat anterior border typical of the subgenus preserved. However, one anomalous specimen MCZ 100021 (Fig. 5.2), with the same flat anterior border, has a glabella more like that of P. (Pl.) haywardi and is difficult to interpret; it resembles a form in the Iberian Chains associated with P. (Pl.) mureroensis (Liñán and Gozalo, 1986, pl. 11, fig. 14).
Larger cranidia up to 7 cm long, with occipital ring proportionally wider, with S3 and S4 less well impressed; preocular fields gradually widening (tr.) through ontogeny, anterior borders (Šnajdr, 1958, pl. 16, figs. 1–5) becoming longer (sag.); palpebral lobes slightly shortening, situated opposite S3 to mid L1 (Fig. 6.4), i.e., a little more posterior than in the smallest stage. Librigenae with lateral borders wide, extending back into long, slightly inbending genal spines (Fig. 6.2); posterior border bending forward to meet genal spine (Fig. 6.4). Hypostome wider than long, beneath L5 and upper half of L2 (Fig. 7.17), with posterior border slightly curved, with embayment at base of posterolateral short spine (Fig. 7.15).
Thoracic axial rings and pleura of the narrow form differing proportionally from those of the broad form, but disposition of curved pleural spines enveloping the pygidium the same in both forms; posterior pleural spines extending back almost to, or a little beyond, the posterior pygidial margin; second last extending a little more posteriorly and the next one even farther posteriorly; all other pleurae gradually extending backwards less far than the one in front; up to 18 segments.
Pygidium subcircular (Fig. 7.7); only minor modifications during growth (Fig. 7.4–7.9, 7.11); postaxial field commonly squeezed (tectonically/taphonomically) narrower giving laterally ovate pygidium; axis sometimes with two further transverse furrows below the anterior ring (Fig. 7.4). Doublure with terrace lines, widening from the lateral margin, at a level opposite the base of the anterior axial ring, backwards to the base of the axis (Fig. 7.7).
Isolated young cranidia form the majority of Paradoxides (H.) harlani specimens in the collections, ranging in sagittal length from 2 cm to 7 cm. The glabella exhibits the greatest amount of variation, mainly in different rates of forward expansion and traces of S3 and S4. We have examined over 200 specimens from the Quincy-Braintree-Weymouth area and, apart from the Walcott specimen assigned to P. (Plutonides) haywardi, recognized only one pygidial form from the lowest beds of the Braintree Formation. Initially, it was tempting to regard P. (Pl.) haywardi as a rare variant and to follow Geyer and Landing in synonymizing the two species. However, in view of the undoubted cranidial differences consistent with those recognized by Raymond, the rarity of the P. (Pl.) haywardi pygidium must be treated as an anomaly.
Other material examined
Neotype: Green's original specimen has not been traced. However, the concept of the species has been based upon the partially reconstructed MCZ 109330 (Fig. 6.1; Walcott, 1884, pl. 9) from the Braintree Formation in Old Hayward Quarry and employed by Solov'yev in proposing Eoparadoxides. Here it is selected as the neotype.
Lake (1935) remarked on the sparse material of the oldest British Paradoxides groomi Lapworth, 1891 and his illustrations of its librigenae (1935, pl. 29, figs. 4, 8) identify it as having features of P. (H.) harlani.
In his initial description of commercially fabricated Moroccan specimens of Acadoparadoxides briareus, Geyer (1993, p. 82) raised the possibility of it being P. (H.) harlani and, with Landing (2001), outlined minor differences considered specific. Most of those differences, e.g., the straight anterolateral pygidial margins, are recognizable in the suite of variation now noted in P. (H.) harlani (Fig. 7.11). Examination of other fabricated Moroccan examples of A. briareus in various rockshops and specimens housed in the Hunterian Museum, Glasgow (see http://www.hmag.gla.ac.uk/Neil/trilobite/paradoxcomp.jpg), the Dudley Museum in England, and the University of Zaragoza Department of Palaeontology, has revealed a similar range of variation to that observed in the Massachusetts and Newfoundland material. However, hypostomal differences are difficult to confirm, since only one Moroccan specimen has been described (Geyer, 1993, p. 81). As Geyer and Landing (2001) inferred, some of the differences may be stratigraphically controlled, since the oldest Moroccan specimens occur in the Cephalopyge Zone. However, we treat with caution the concept of a morphocline of pygidial types, because of the sparcity of detailed stratigraphic analysis of each species, not to mention that their initial member P. (A.) nobilis Geyer, 1998, with its distinctive anterior cranidial border, seems to belong to another subgenus, specimens of which were figured by Geyer (1983, pl. 17, figs. 7–9), Sdzuy (1995, pl. 1, fig. 6), and Liñán and Gozalo (1986, pl. 11, fig. 8; Fig. 8.13).
Subgenus PlutonidesHicks, 1895
Acadoparadoxides Šnajdr, 1958, p. 146.
Baltoparadoxides Šnajdr, 1986, p. 174.
Plutonia sedgwickii Hicks, 1871 from the Lower Solva Beds near Nun's Well, St. David's, Wales.
Glabella widening gently forwards (early holaspid with lateral furrows to L1 and L2 generally subparallel) to a globose anterior lobe, with a semicircular anterior margin; holaspis with only S1 and S2 well impressed; anterior border upstanding, widening abaxially, delineated by relatively deeply impressed furrow axially. Pygidium wider than long. Surface ornament granulate/tuberculate, with terrace ridges on glabella and borders.
The lectotype of the type species [SM A1086] selected by Dean and Rushton (1997, p. 476) was figured by Hicks (1871, pl. 15, fig. 1) and Lake (1935, pl. 31, fig. 9); it is a distorted sheared cranidium and their diagnosis appears to be largely based on Paradoxides hicksii Salter, 1866 that has a cranidium of different form, i.e., with S3 and S4 well impressed and a flat anterior border. Using the computer facilities of Photoshop, the undistorted form of the lectotype is revealed (Fig. 8.1) as one like P. sacheri Barrande, 1852, the type species of Acadoparadoxides, with only S0–S2 impressed, but with slimmer, much shorter palpebral lobes. Here, we regard the relative lengths of palpebral lobes as specific characters and Plutonides as the senior synonym of Acadoparadoxides and Baltoparadoxides (the latter representing early holaspids with subparallel lateral furrows to L1 and L2), differing from Paradoxides (Hydrocephalus) and P. (Eccaparadoxides) in having S3 and S4 barely, if at all, impressed in the holaspid and from P. (Paradoxides) in having a much more globose anterior glabellar lobe; its upstanding anterior border, well defined by a furrow, also differentiates it from P. (Hydrocephalus) and P. (Paradoxides).
Since the thorax and pygidium of P. (Pl.) sedgwickii are not known, those of P. (Pl.) sacheri (Šnajdr, 1958) may be taken as a guide to the subgenus. The holotype of P. (Pl.) sacheri (Šnajdr, 1958, pl. 17, fig. 1) has 19 thoracic segments and an elongate (sag.) spatulate pygidium with a transverse posterior margin. Such a pygidial shape differs from isolated subcircular pygidia assigned to this species by Šnajdr (1958, pl. 16, figs. 6, 8, 10, 17) and by Dean and Rushton (1997, fig. 305, 2b). These differences may indicate intraspecific variations or that the isolated forms do not belong to the species. This observation suggests that caution must be applied to any scheme considering pygidial outlines for major subdivision of Paradoxides and justifies our decision to use only cephalic characters to diagnose subgenera.
Paradoxides (Plutonides) haywardi (Raymond, 1914) Figures 7.12, 8.2–8.9
Paradoxides haywardi sp. nov. Raymond, 1914, p. 237, unnumbered pl., figs. 1, 2, 7.
Cranidia up to 4 cm long. The holotype differs a little from its better-preserved examples of the species, because it is distorted. The right half has been crushed and the anterior border is kinked; Raymond considered the latter feature important, but none of the undistorted forms displays this character. Compared to a P. (H.) harlani cranidium of the same size, only S1 and S2 are well impressed on a glabella, whose anterior outline is semicircular and lateral furrows of the lower lobes considerably more sagittally parallel. One feature of note is the gradual widening of the fixigenal fields with growth, some of which may be tectonically or taphonomically induced. Reference to the figured specimens indicates that there is a range of widths (tr.) to the preocular and lower fixigenal fields—a range also notable in P. mureroensis Sdzuy, 1958 (Fig. 8.10, 8.11). According to Raymond (1914), Haynes's exoskeleton has 17 thoracic segments and a pygidium resembling the small transversely ovoid form of MCZ 20, which is broad and short, i.e., 15 mm wide and 9 mm long, with the pygidial axis bluntly rounded, proportionally longer, and more rounded than the more triangulate P. (H.) harlani axis. Superficially, it resembles those of the type species of Hydrocephalus figured by Šnajdr (1958, pl. 26, fig. 11; pl. 27, figs. 1–3) and differs considerably from his spatulate and subcircular forms referred to Acadoparadoxides. Walcott's specimen (Fig. 7.12) indicates that the pygidium is 20 mm wide and about 12 mm long with the last five pleural spines of the thorax all extending beyond the posterior pygidial margin and the last four gradually reduced in length posteriorly. Although the latter feature may be variable, as in P. (H.) harlani, the pygidium is unlike other small pygidia in the collections with long blunt axes (Fig. 7.1, 7.3) because they are not wider than long. They tend to be more subhexagonal, slightly elongate with an axis less triangulate, like Paradoxides mureroensis of Sdzuy (1961, fig. 24). This is an anomalous situation; either the pygidial form of any species is highly variable or each form marks a distinct species. However, since none of the three known complete exoskeletons (Liñán and Gozalo, 1986, p. 51) has been described (their 1986, pl. 11, fig. 1 is unclear), isolated Spanish subcircular pygidia may belong to another subgenus or species like P. (H.) harlani, to which they bear a very strong resemblance. Although contorted, reference to transverse Spanish pygidia with long bluntly rounded axes (Sdzuy, 1961, pl. 17, figs. 1–5; Liñán and Gozalo, 1986, pl. 11, fig. 10) suggests a close relationship between P. (Pl.) mureroensis and P. (Pl.) haywardi that is supported by their similar cranidia.
The more mature Massachusetts specimens resemble P. (Pl.) sedgwickii (Hicks, 1871) and P. (Pl.) harknessi (Hicks, 1871) from the Solva Beds of South Wales, but the palpebral lobes are intermediate in length between the much shorter P. (Pl.) sedgwickii lobes and the longer P. (Pl.) harknessi lobes (Lake, 1935, pl. 27, fig. 3). However, cranidia of P. (Pl.) haywardi form, e.g., ROM 56409 (Fig. 8.9), are common within the Kiskinella Zone in southeastern Newfoundland, associated with P. (H.) harlani, but palpebral lobe length varies with growth, i.e., it shortens proportionally with increased cranidial size. While no forms are known to have palpebral lobes as short as those of P. (Pl.) sedgwickii, the smallest Newfoundland forms have lobes as long as P. (Pl.) harknessi, but their fixigenal fields are much narrower (tr.).
Sdzuy's (1961, fig. 24) cranidial drawing of Paradoxides mureroensis has the palpebral lobes extending back to the posterior border furrows like harknessi, but this feature is not too obvious in his 1961 plate-16 figures and barely supported by more recent figures of the species by Liñán and Gozalo (1986, pl. 11, figs. 5, 6; Fig. 8.10–8.12). The posterior section of the facial suture on librigenae Sdzuy (1961, pl. 16, figs. 12, 13) assigned to the species shows that the posterior end of the palpebral lobe is more or less level with mid L1 as in P. (Pl.) haywardi. The very slightly wider fixigena of P. (Pl.) haywardi (Fig. 8.9) seems to distinguish the species from P. (Pl.) mureroensis (Fig. 8.10). Two other forms in the Iberian Chains attributed to P. (Pl.) mureroensis need clarification. One has a Plutonides-type glabella and a flat Hydrocephalus-type anterior border (Liñán and Gozalo, 1986, p. 51, pl. 11, fig. 14) like the Massachusetts specimen MCZ 100921 (Fig. 5.2); the other has a Hydrocephalus-type glabella and a Plutonides-type anterior border (Liñán and Gozalo, 1986, pl. 11, fig. 8, see Photoshop-corrected form in Fig. 8.13) resembling “Paradoxides (Acadoparadoxides) nobilis” Geyer, 1998 in Morocco. Both forms appear to be relatively rare and their full significance is unclear.
No complete exoskeleton with thorax and pygidium undoubtedly connected to the P. (Pl.) haywardi cranidium is known. However, Raymond considered that “a nearly entire specimen” held by W. P. Haynes confirmed the relationship of Walcott's (1884, pl. 8, fig. 1d) terminal thorax and pygidium (Fig. 7.12) to his holotype.
Order PtychopariidaSwinnerton, 1915
Suborder Ptychopariina Richter, 1932
Anomocare campbelli King, 1923, from the early Middle Cambrian of the Dead Sea area of Jordan.
Following Geyer, 1990, p. 102, 333. Cranidium almost uniformly convex; all furrows (except the occipital furrow) obsolete on test surface; glabella with concave sides and a kingaspidoid pattern of glabellar furrows; on internal molds the anterior glabellar lobe with distinct extensions of the anterolateral corners, or anterolateral corners at least conspicuous and visibly connected to the eye ridges; preglabellar area moderately wide. Librigenae usually with spine. Test smooth, punctuate, or with thin terrace ridges.
Kingaspis aff. K. avalonensisGeyer and Landing, 2001Figure 9.1
Holotype; cranidium USNM 50693a (Geyer and Landing, 2001, fig. 5.1–5.3) from the base of the Braintree Formation in Old Hayward Quarry and several specimens from Moktar in the Kerdous region of southern Morocco.
Geyer and Landing (2001) discussed all available Massachusetts specimens. We draw attention to a similar Moroccan form represented by M-F237A (Fig. 9.1) occurring with Paradoxides (Hydrocephalus) harlani, Hamatolenus (Myopsolenus) magnus, and a solenopleurid (Fig. 13.15, 13.16) resembling Braintreella rogersi in green sandstones of the Tamanart Formation at the base of the Feijas internes Group (Geyer, 1989) east of Tiznit.
During the survey of the Kerdous Region (British Geological Survey, 2002) internal molds of Kingaspis cranidia were found with relatively long slender occipital spines. Compared to the other Moroccan, occipital spine-bearing species, K. amouslekensis Geyer, 1990, the occipital furrow is better defined and the spine is a slender projection from a triangulate ring similar to Kingaspidoides brevifrons (Hupé, 1953) (Geyer, 1990, pl. 19, figs. 2, 5a, 7a). Geyer and Landing's (2001) figures of Kingaspis avalonensis display the external surface on which the furrows are characteristically less well impressed and indicate some variation in the glabellar form. The Moroccan internal molds exhibit deeper furrows and comparable glabellar differences. For instance, their proportionally wider glabellae resemble that of K. avalonensis (Geyer and Landing, 2001, fig. 5.5) and their narrower, tapering glabellae are close to another specimen (Geyer and Landing, 2001, fig. 5.8). The Moroccan specimens are regarded as Kingaspis aff. K. avalonensis not only for the common form of occipital spine, but also for the comparable proportions of the glabellar length to cranidial length (44.4 percent) and fixigenal width 30.6 percent of the cranidial width across the palpebral lobes.
Subfamily ProtoleninaeRichter and Richter, 1948 emend
Emended by Geyer (1990, p. 163, 336) as ellipsocephaloids with glabella strongly convex, usually with subparallel sides tapering at anterior lobe, with four pairs of deeply impressed glabellar furrows, of which the posterior three pairs are almost parallel and obliquely directed backward, front with parafrontal band (in front of S3); palpebral lobes fused to eye ridges, but separated by a incipient-impressed furrow. Librigenae with genal spine. Thorax multisegmented. Pygidium rounded in outline.
Geyer (1990) included six genera in this subfamily; relevant to the Massachusetts material, Hamatolenus with a very short depressed preglabellar field lacking caecae is distinguished from Protolenus Matthew, 1892, which has a prominent long (sag.) convex preglabellar field, usually marked by conspicuously radiating caecae.
Genus HamatolenusHupé, 1953
Hamatolenus continuus Hupé, 1953, a junior subjective synonym of Protolenus elegans var. marocana Neltner, 1938 from the Cephalopyge Zone in the Ouirgane region of the High Atlas, Morocco; see Geyer (1990, p. 165) for synonymy.
Preglabellar field very short, not exceeding the sagittal width of a flat to gently convex anterior border.
Subgenus Hamatolenus (Hamatolenus) Hupé, 1953
Palpebral lobes longer than the anterior branches of the facial suture, more than one-third of glabellar length, reaching the posterior border furrow.
Hamatolenus (H.) aff. H. (H.) marocanus (Neltner, 1938) Figure 9.2, 9.4
Two incomplete cranidia, GZ 17 and GZ 20 from the ledge below the level of the Old Hayward Quarry, in the same lithology as in the quarry.
GZ 17 is the better specimen even though it is distorted in the anterior region. However, of the species recognized by Geyer (1990), it is closest to the type even though the fixigenal regions are narrower.
Subgenus MyopsolenusHupé, 1953
Palpebral lobes short, not reaching the posterior border.
This subgenus has also been recognized in North Wales (Bassett et al., 1976, p. 626).
Hamatolenus (Myopsolenus) cf. H. (M.) magnusHupé, 1953Figure 9.3
Glabella rounded anteriorly; preglabellar field very short, shorter than the anterior border.
Flattened cranidium GZ 9 from dumped debris derived from former excavations at the General Dynamics (1974–75) Wet Basin Slip No. 12, Quincy.
Of the three species discussed by Geyer (1990), this specimen is closer to magnus, because of its rounded glabellar front, narrower glabella, and thicker eye ridges. Differences, possibly influenced by tectonic stress, are the virtual absence of a preglabellar field and much thicker palpebral lobes.
Genus HolocephalinaSalter, 1864
Holocephalina primordialis Salter, 1864 from the Paradoxides davidis Zone at Porth-y-rhaw, South Wales.
Ovate, narrowing posteriorly to a very small pygidium; thorax longer than cephalon. Cephalon semielliptical, length exceeding half width, evenly convex, with very faint axial furrow; glabella very little raised above genae, slightly more than half as long as cephalon and one-third as wide, narrowing forwards to a slightly rounded anterior, with three pairs of barely differentiated lateral furrows; occipital furrow well impressed; facial sutures marginal; genal spines extending backwards to about fifth segment. Thorax of 14–17 segments.
Lake (1938) reviewed this genus that includes H. primordialis [teres Grönwall, 1902], H. incerta Illing, 1916 from the P. hicksii Zone Passage Beds, Hartshill, England, and H. meneviensis Hicks, 1872 from the Menevian of Porth-y-rhaw, St. David's, South Wales. A Newfoundland specimen from the P. davidis Zone was referred to H. americana by Resser (1937) and Gozalo and Liñán (1996) described Holocephalina? leve from the much older P. asturianus Zone in the Murero Villafeliche Section 1, Spain.
Specific features concern the cephalic characters of the glabella, occipital ring, border with genal spines, and ornament on the external test and internal mold.
Holocephalina sp. nov. aff. H. levisGozalo and Liñán, 1996Figure 10.1–10.4
Exoskeleton ovate, with length:breadth ratio 5: 3. Cephalon semicircular, with length:breadth ratio 7:10. Glabella defined by weakly impressed axial furrow, tapering forwards to a slightly rounded truncated front, raised slightly above genae; occipital furrow well impressed, arched slightly forward over the midline; occipital ring five-sevenths cephalic length, medially swollen, with a median node. Fixigenae flattish, narrower than the base of the glabella, sloping very gently down into the border furrows; preglabellar field long (sag.), convex; eye ridges narrow, ending as incipient palpebral lobes on the fixigenae. Posterior border furrow shallowing away from the occipital furrow, joining a barely distinguishable lateral border furrow; anterior border weakly defined, flattish, about one-third the length (sag.) of the preglabellar field. Facial suture opisthoparian, marginal. Librigena with genal spine extending as far as the fifth thoracic segment.
Thorax of 14 segments; axis bearing small median node on each segment; nodes diminishing in size posteriorly.
Pygidium, small, wider than long in ratio 11:5, width about half that of the posterior width of the cephalon; axis raised, weakly segmented, tapering backwards to a well-rounded terminus, not reaching the posterior border; anterior border well defined merging with a less well-defined lateral-posterior border; pleural regions weakly marked as five segments.
Ornament finely punctate on the cephalon.
Four cranidia (MCZ 114894, MCZ 114896, MCZ 114917, MCZ 109343B) and a damaged complete exoskeleton (MCZ 332) from Massachusetts and a cranidium (ROM 56119) from the Kiskinella Zone, Bed 2d, Easter Cove Manganiferous Member, Chamberlain's Brook Formation, Branch, Newfoundland (Fletcher, 2003, text-fig. 7).
The Spanish and Massachusetts specimens are preserved in a similar lithology so that surface ornament that might be a distinguishing character is masked. However, the main feature that differentiates these older forms is a longer glabella than in H. americana, H. incerta, or H. meneviensis and very weakly impressed glabellar furrows compared to H. primordialis. The Massachusetts occurrence is apparently the oldest for the genus, since its assemblage appears older than the Spanish P. asturianus Zone (Fig. 2) and closely matches the better-preserved Newfoundland specimen (Fig. 10.4) from the Kiskinella Zone.
Superfamily PtychoparioideaMatthew, 1888
Palpebral lobe small, upturned; anterior border of even width or very slightly swollen over the axis. Thorax of 14– 17 segments. Surface ornament finely granulate or smooth, with fine granules on the axial parts of the glabella, thorax, and pygidium; smooth areas may bear small punctae.
Family AgraulidaeRaymond, 1913b
Fortey (1997, p. 302) regarded the Agraulidae as ellipsocephaloid; this was tentatively supported by Cotton (2001). However, until all the characters of the type species are known, its superfamilial position will remain unclear. Based upon topotypes of the type species of the nominate genus and closely matched “species” from Newfoundland and Spain, we follow Matthew (1888, p. 129) in placing this possibly monogeneric family in the Ptychoparioidea.
Genus AgraulosHawle and Corda, 1847
Arion ceticephalus Barrande, 1846 from the Paradoxides (Eccaparadoxides) pusillus Zone at Skryje, Bohemia, Czech Republic.
Barrande's description is based on internal casts and important details of the external test surface have not been fully appreciated.
Many museums house numerous Bohemian specimens of this species, but like most fossils collected from the type locality, they are invariably internal molds coated with goethitic powder as a result of decalcification of the outer shell. Nobody involved in the classification of this taxon appears to have cleaned out the external molds to prepare casts exhibiting the external aspect, i.e., characters that add much to any detailed diagnosis.
This species is associated with relatively late paradoxidids and relationships with earlier “agraulids” have not been established. Following the earlier Treatise, the general concept is that Agraulos is a ptychopariid with the cephalic furrows greatly suppressed so that the cephalon is typically smooth and almost featureless. Since Růžička (1946), Skreiaspis has been considered a close relation (Harrington et al., 1959, p. O279), marked by a raised glabella with a relatively well-impressed axial furrow. Sdzuy (1967) and Fletcher (1972) followed this definition in assigning older species to Skreiaspis, i.e., S. tosali Sdzuy, 1967 in Spain and S. quadrangularis Whitfield, 1884 and S. affinis Billings, 1874 in Newfoundland, possibly implying a Skreiaspis–Agraulos phylogenetic trend. Closer examination of Czech topotype examples of the type species S. spinosa (Jahn, 1893) from beds equivalent to the Scandinavian Tomagnostus fissus–Ptych. atavus Zone indicates that, although the surface ornament is similarly punctate, the basic structures, e.g., the shapes of the glabella, preglabellar field, and anterior border, are not those of Agraulos ceticephalus (Barrande, 1846). Therefore, we agree with Geyer and Landing's decision to regard Whitfield's Massachusetts species as Agraulos.
In view of the poor understanding of Agraulos, we include here a description of Czech topotype material of the type species preserved in mudstone, represented by both internal casts and latex casts of external molds, supplemented by a specimen from southeastern Newfoundland preserved in limestone.
Agraulos ceticephalus (Barrande, 1846) Figures 11.1–11.7, 12
For synonymy, see Šnajdr, 1958, p. 174.
Cranidium wider than long in the ratio 5:4, subtriangular, with a sharply rounded pointed anterior and shallow furrows on the external surface that are well defined on the internal cast. Ornament with three sizes of punctae scattered over a background of tiny granules (Fig. 11.7), with fine terrace ridges (Fig. 11.4) along the anterior cranidial margin and posterior edge of the occipital ring, over narrow stretches.
Cranidial to glabellar (with occipital ring) length ratio 8:5. Glabella raised slightly above the fixigenae, flattish, prominently defined by a well-impressed axial furrow, tapering forwards from the occipital furrow, truncated anteriorly, with conspicuous slight axial depression at anterolateral corner; width of occipital furrow twice that of the glabellar anterior; lateral glabellar furrows four pairs, narrow, arching backward, extending over about one-third of the glabella, not reaching a faintly defined narrow axial ridge. Occipital ring short (sag.), strongly curved backward into a blunt point (Fig. 11.4), i.e., not the small spine as drawn by Šnajdr (1958, fig. 37), defined by a deep occipital furrow, unusually long (sag.) axially; posterior margin of occipital ring in three sections—a very narrow medial transverse portion, coinciding with the base of the axial glabellar ridge, and two flanks curving forwards.
Fixigenae flattish, with a narrow posterior border defined by a wide (exsag.) furrow; width across base a little less than two-thirds width (tr.) of the occipital furrow; crescentic anterior border defined by a weakly impressed border furrow weakening across the axis; border furrow arching backwards axially, separated from the glabella by a flattish preglabellar field slightly shorter (sag.) than the anterior border (sag.); eye ridges faintly defined, thin, curving outwards from the anterior glabellar lobe (the anterior lobe appears to be separated by the axial ridge) to meet prominent upstanding small palpebral lobes just above their midlengths; palpebral lobe astride the midcranidial length, widening slightly posteriorly; palpebral areas of the fixigenae deeply impressed posteriorly. Posterior section of facial suture curved strongly backward in contrast to the inward-curving anterior sections.
An internal mold of cranidium SM A72604 (Fig. 11.1–11.3) and an external mold of cranidium and partial thorax SM A1568 (Fig. 11.4) from Skryje, Bohemia, and ROM 56114 (Fig. 11.5–11.7) from the T. fissus–Ptych. atavus Zone in the lower beds of the Manuels River Formation in Deep Cove, St. Mary's Bay, Newfoundland (Fletcher, 2003, text-figs. 1 [between localities 8 and 9], 3).
The internal cast of the cranidium displays the basic structure to be taken as characteristic of this species. Although the illustrated external cast (Fig. 11.4) has only 10 thoracic segments preserved, complete exoskeletons in the collections possess fourteen. Comparisons with A. quadrangularis, A. affinis, A. tosali (Fig. 12) and A. socialis (Billings, 1874), respectively representing stratigraphically younger species in pre-T. fissus–Ptych. atavus zonal sequences, indicate a gradual suppression in the strengths of dorsal furrows and axial median spines/nodes. The type species appears to be the penultimate member of these changes with median nodes on only the fourth, fifth, and sixth thoracic rings and the anterior three rings covered by fine terrace ridges (Fig. 11.4); the even smoother A. socialis, high in the T. fissus–Ptych. atavus Zone in Newfoundland, is the end member.
A comprehensive account of this genus will be given elsewhere, but these details of the type species appear to exclude Skreiaspis Růžička, 1946 and Proampyx Frech, 1897 as close relatives. Of other species A. longicephalus (Hicks, 1872) may be an early variant of A. ceticephalus in having a slightly more prominent occipital spine and A. tosali may be a junior synonym of A. holocephalus (Matthew, 1890).
Agraulos quadrangularis (Whitfield, 1884) Figures 11.8–11.23, 12
Glabella about two-thirds cranidial length, defined by deep axial furrow, slightly wider than fixigena, raised slightly above fixigena, tapering forwards, with a very slight constriction midlength, with a truncate front. Fixigenae gently arched exsagittally, confluent in front of glabella; crescentic anterior border barely differentiated on outer shell surface; palpebral lobe small, upstanding, situated just anterior of cranidial midlength and attached to thin eye ridge.
Holotype: an incomplete external cast of cranidium AMNH-FI 51292 (Fig. 11.8) and several other external casts from the Old Hayward Quarry, Quincy.
Other material examined
Numerous Newfoundland specimens: cranidium ROM 56115 (Fig. 11.21, 11.22), an internal cast exhibiting the modified characters of A. ceticephalus; ROM 56116 and 56117 (Fig. 11.20, 11.23) show the typical punctate test with features of the Agraulos internal structure masked.
The holotype cast shows signs of abrasion and comprises a glabella (without occipital ring), the anterior parts of the fixigenae, and a palpebral lobe. Available collections include relatively few examples of this species and most are somewhat distorted cranidia. Geyer and Landing (2001, p. 132) outlined the morphology of cranidial casts and we figure other specimens that provide some measure of variation largely caused by compression. A complete, distorted specimen (Fig. 11.15) indicates that the librigena extends backwards into a prominent genal spine, a thorax of 14 segments, and a typical ptychopariid small transverse pygidium; other specimens show the occipital ring bears a prominent median spine and that genal spines are relatively long.
Family PtychopariidaeMatthew, 1888
Preglabellar field slightly convex or flattish; genal angles with or without short spines; pygidium small, wide, with four or five segments, without a prominent border.
Genus BraintreellaWheeler, 1942
Ptychoparia rogersi Walcott, 1884 from the Paradoxides Quarry at Braintree, Massachusetts, USA; by original designation.
The holotype of the type species is a very poorly preserved fragment, with little to establish characters of Braintreella as proposed by Wheeler. The main problem is the shape of the deformed glabella as described below. However, illustrations of associated Massachusetts specimens (Geyer and Landing, 2001, fig. 9.7–9.9, 9.11–9.13; Fig. 13.4–13.7) demonstrate the rather disparate morphologies of the cranidia, not all of which can be attributed to tectonic/taphonomic distortion. Differences in the amount of glabellar tapering, the shape of the glabellar front, the convexity or sunken nature of the preglabellar field, the sharpness or bluntness of the occipital spine, the relative widths of the fixigenae and the thickness of the anterior border appear to indicate substantial variability and render it difficult to provide a concrete diagnosis of a form that here is considered one taxon. A case may be made for restricting the species rogersi and genus Braintreella to the holotype, but, by the same reasoning, several other taxa would have to be established for some of the associated clearly related variants. However, in view of an imperfect preservation that makes comparison with well-preserved material elsewhere rather subjective, we follow Geyer and Landing (2001) in restricting the genotype species name to the presently known specimens associated with the holotype in this region.
Braintreella rogersi (Wheeler, 1942) Figure 13.3–13.7
Ptychoparia rogersi n. sp. Walcott, 1884, p. 47, pl. 7, fig. 2.
Braintreëlla currieri n. sp. Wheeler, 1942, p. 570, pl. 1, fig. 3.
See Geyer and Landing (2001, p, 129). Material examined by us shows that the occipital ring bears a pronounced blunt spine (Fig. 13.4, 13.5) and that the median node mentioned in their description may refer to broken traces of a spine.
Holotype: MCZ 109314 is an incomplete cranidium with the glabella and occipital ring tectonically crushed (Fig. 13.3). The specimen is an external mold and tectonic flattening of the genae appears to have produced a greater width (tr.) than length (sag.). The fixigena is slightly wider (tr.) than the base of a well-defined upstanding anteriorly tapering glabella. Tectonic/ taphonomic distortion has also resulted in an asymmetric rounding of the glabellar front with some shell in the right anterior margin pressed down vertically into the rock and an obliteration of any lateral glabellar furrows. The occipital ring is short (sag.) and upstanding with a trace of a median spine. The better-preserved left fixigena contains a small, slightly upstanding palpebral lobe, positioned just anterior to the midlength of the cranidium, that extends anteriorly towards the anterior glabellar lobe as a curved faint thinner eye ridge. The frontal area comprises a slightly convex preglabellar field about the same sagittal width as the upturned anterior border. The posterior section of the facial suture extends outwards from the base of the palpebral lobe, in contrast to the anterior branch that is more sagittally oriented.
Other material examined
Paradoxidid specimens are abundant in the various collections from Massachusetts, but there is a paucity of other taxa and we located only nine worthwhile specimens attributable to B. rogersi additional to those figured by Geyer and Landing (2001, fig. 9.7–9.9, 9.11–9.13). Poor preservation makes complete morphological details difficult to establish; neither an entire thorax nor pygidium is known and metamorphism and tectonism have obliterated subtleties of the exoskeletal surface.
Geyer and Landing (2001) described only specimens housed in the Smithsonian Institution. We figure others from the collections of Lord, Zeoli, and the MCZ to illustrate wider cranidial variants with glabellae having more quadrate fronts and traces of three pairs of glabellar furrows (the most posterior pair being oriented slightly backward).
In southeastern Newfoundland, associated with Paradoxides (Hydrocephalus) harlani and P. (Plutonides) haywardi, slightly flattened solenopleurid specimens preserved in mudstone (Fig. 13.8, 13.9) similar to the Massachusetts material of B. rogersi, e.g., Fig. 13.6, accompany better-preserved examples retaining surface details of the exoskeleton (Fig. 13.12, 13.13). They are provisionally referred to Parasolenopleura sp. and will be more fully considered in a later publication. Unlike the Massachusetts situation, complete specimens occur having a thorax of 14 segments with short spines on the anterior 12 rings (Fig. 13.1, 13.2) and a small transverse pygidium without a well-defined posterior border and an axis of four rings and terminal portion raised slightly above posteriorly conjoined pleural fields bearing well-impressed segmental furrows (Fig. 13.14). Raised axial portions of the test bear small granules (Fig. 13.8) in contrast to smooth fixigenal areas with fine punctae. While no satisfactory Massachusetts librigena are available for comparison, we illustrate well-preserved specimens from Newfoundland showing details of the surface ornament (Fig. 13.11) and caecal impressions on the internal surface (Fig. 13.10). On the evidence available to the authors, B. rogersi and this Newfoundland form have features in common with the stratigraphically younger “Calymene aculeata,” described from Scandinavia by Angelin (1852)—the genotype of Parasolenopleura Westergård (1953, p. 21, pl. 5, figs. 8– 10). One character of difference with the Newfoundland form lies in the thoracic axis of P. aculeata, which bears fewer spines, recalling the gradual stratigraphic diminishing of thoracic axial spines noted above in Agraulos (Fig. 12).
Like B. rogersi, P. aculeata has a pronounced preglabellar field taken as one of its generic features and Braintreella may be linked phylogenetically to other Parasolenopleura species occurring below the Scandinavian Triplagnostus gibbus Zone and its equivalents. Parasolenopleura characterizes paradoxidid sequences and, in Newfoundland, is represented by four species in different stratigraphic positions: 1) the earliest species (Fig. 13.8–13.14) ranges from the top of the Cephalopyge Zone into the basal Paradoxides (Hydrocephalus) harlani Zone, and, in part, resembles Atopiaspis tikasraynensis Geyer, 1998 in the Moroccan O. frequens Zone; 2) Parasolenopleura gregaria (Billings, 1865) occurs with Condylopyge carinata in the Agraulos affinis Zone (Fletcher, in press) and is possibly the senior synonym of P. cristata (Linnarsson, 1877) and P. lemdadensis Geyer, 1998; and 3) P. aculeata (Angelin, 1852) is common in the Hartella Zone. There also appears to be a link with Badulesia Sdzuy, 1967 via the Parasolenopleura form identified as Badulesia sp. (without the diagnostic cranidial ridges) in the Spanish Paradoxides asturianus Zone (Liñán and Gozalo, 1986, pl. 23, fig. 6; Gozalo and Liñán, 1998, p. 101), i.e., contemporaneous with strata in Newfoundland and Spain containing Badulesia tenera (Hartt in Dawson, 1868) (Chirivella Martorell et al., 2003, p. 86; Fig. 2).
The P. (Hydrocephalus) harlani fauna in coarse clastic rocks of the Moroccan O. frequens Zone around Moktar includes solenopleurid forms resembling Braintreella (Fig. 13.15, 13.16), but they are similarly poorly preserved and difficult to identify. However, they more closely resemble solenopleurid forms noted in the early Paradoxides (Hydrocephalus) and P. (Plutonides) sequences than those described from younger zones.
This work is dedicated to the memory and considerable endeavors of our former colleague Stinson Lord, by common consent, the hard-working preeminent twentieth-century collector and inspirer of the many collectors of Massachusetts Cambrian fossils now available for study. It is also appropriate to acknowledge the time, help, and access to collections kindly given to us by N. Clark (Hunterian Museum, Glasgow University), F. Collier (MCZ), R. Downey Jr. (Braintree Historical Society Museum), the late B. F. Howell (Princeton University), B. Hussaini (AMNH), E. Liñán (University of Zaragoza), J. Skehan (Weston Observatory, Boston College), and J. Thompson (USNM). Our primary manuscript has been substantially improved following highly constructive comments by Adrian Rushton and Steve Westrop; we are also grateful to Peter Jell for his efforts as Associate Editor.
- Accepted 7 May 2004.