Cornulites, SCHLOTHEIM, 1820

Herringshaw, Liam G., Thomas, Alan T. & Smith, M. Paul, 2007, Systematics, shell structure and affinities of the Palaeozoic Problematicum Cornulites, Zoological Journal of the Linnean Society 150 (4), pp. 681-699 : 683-688

CORNULITES SCHLOTHEIM, 1820

Type species. Cornulites serpularius Schlotheim, 1820 , from the Silurian of Gotland. By monotypy.

Emended diagnosis: Conical calcite shells with varying angle of taper, ranging from narrow, almost tubular forms, to widely inflated cones; transversely annulated, commonly with longitudinal costae. Shells straight or slightly sinuous, rounded or ovoid at origin; sinuous forms often straightening in later growth stages. Shell microstructure of stacked lamellae, apically straight, but becoming undulating towards aperture, finally forming apertural groove. Lamellae connected internally at apex by horizontal tabulae; separated distally by cellulae that become prominent towards aperture, particularly in large forms.

Remarks: Four species from the Much Wenlock Limestone Formation are here placed in Cornulites : C. serpularius , C. scalariformis , C. cellulosus and C. gremialis . The cellular shell structure is known from both C. scalariformis and C. cellulosus , but material of C. serpularius suitable for sectioning has not been found. Dzik (1991: 129) regarded cellulae as ‘typical only for adult Silurian Cornulites ’: if they are absent from the type species, other taxa may be better placed in a separate genus.

CORNULITES SERPULARIUS SCHLOTHEIM, 1820

( FIG. 2A, B)

* 1820 Cornulites serpularius Schlotheim , p. 378,

pl. 29, fig. 7.

? 1859 Cornulites serpularius Schloth. ; Murchison, p. 221, pl. 16, fig. 3 [only these specimens resemble Schlotheim’s syntypes of C. serpularius , all others figured by Murchison being examples of C. cellulosus ].

? 1873 Cornulites serpularius Schloth. ; Salter, pp. 85, 93, 128, 177.

? 1875 Cornulites serpularius Schlotheim ; Baily, p. xlii.

? 1882 Cornulites serpularius Schlot. ; Vine, p. 377.

1888 Cornulites serpularius Schlotheim ; Hall, p. 8.

? 1923 Cornulites serpularius Schlotheim ; Reed,

p. 269.

1962 Cornulites serpularius Schlotheim ; Fisher, p. W137, fig. 80 [reproduction of Schlotheim’s original illustration].

? 1972 Cornulites serpularius Schlotheim ; Blind, p. 5.

1979a Cornulites serpularius Schlotheim ; Larsson, p. 208 [noted dissimilarity between Schlotheim’s illustration of syntypes and material assigned to species by subsequent authors].

? 1991 Cornulites serpularius Schlotheim ; Dzik,

p. 126, pl. 2, figs 5, 6 [specimens probably examples of C. cellulosus ].

2005 Cornulites serpularius Schlotheim ; Vinn & Mutvei, p. 726.

Material: The type specimens figured by Schlotheim (1820) from the Silurian of Gotland cannot be located ( Larsson, 1979a; D. Korn pers. comm., 2003; O. Vinn pers. comm., 2005). This account is based on two specimens attached to the brachiopod BU 4376 from the Much Wenlock Limestone Formation of Dudley, England. Other specimens attached to brachiopods BMNH 58209, B8926 and B34691 (all from Dudley), B23134 (Aldridge, West Midlands), B34653 (Tickwood, Shropshire), and B3921 and B10115 (localities unknown), all of Wenlock age, are also referred to C. serpularius .

Emended diagnosis: Very closely annulated, slightly sinuous species of Cornulites without costae. Aperture shows uniform increase in size, giving shell a regular conical shape.

Description: Specimens of type series 6–7 mm long, with 20–25 annulations per shell. Annulations regular in shape, running parallel to circular aperture. Specimens on BU 4376 slightly smaller, ∼ 5 mm long; aperture more elliptical, slightly irregular; annulations slightly undulating, not always parallel to aperture margin.

Remarks: Larsson (1979a) first noted the discrepancy between Cornulites serpularius as figured by Schlotheim (1820) and most of the specimens assigned to the type species by later authors. There are in fact very few cornulitids from the Much Wenlock Limestone Formation that closely resemble Schlotheim’s illustration of two small, closely annulated specimens without costae. As the material upon which Schlotheim based his figure cannot be found, comparisons with other specimens are necessarily somewhat speculative, but the difference between the syntypes of C. serpularius and most other Wenlock cornulitids may be due in part to the drawing not being an exact representation of the specimens. Schlotheim’s figure of C. serpularius shows the two syntypes to be unattached and resting free on an apparently homogeneous substrate, whereas all small cornulitids from the Much Wenlock Limestone Formation are found partly or wholly attached to the hard parts of another organism and, where it is preserved, in a matrix of bioclastic limestone. C. serpularius is restricted here to small specimens that most closely resemble the material originally illustrated by Schlotheim (1820), but with the qualification that the exact morphology of the type material remains uncertain. All other specimens are transferred to different species.

The profound difference in size between C. serpularius and the three other species of Cornulites recognized here raises the possibility that the specimens of C. serpularius were juveniles of one of the other, larger species. The difference in spacing of the annulations does not support such a relationship between C. serpularius and C. cellulosus – the annulations are relatively broad even in the early growth stages of C. cellulosus – whilst the absence of costae from C. serpularius precludes it being an early ontogenetic stage of C. scalariformis . C. gremialis has close annulations quite similar to those of C. serpularius , but they are often irregular, a feature not illustrated in the syntypes of C. serpularius ( Schlotheim, 1820: pl. 29, fig. 7). The presence of fine costae in C. gremialis also indicates that the taxa are not conspecific.

CORNULITES CELLULOSUS SP. NOV.

( FIGS 1B , 2C–J, 4A–M, 5A–G , 6A, B , 9A, B )

1859 Cornulites serpularius Schloth. ; Murchison, p. 221, pl. 16, figs 3a, 4, 6–10 [various specimens described as from the Much Wenlock Limestone Formation of Ledbury, Herefordshire].

? 1859 Cornulites serpularius Schloth. ; Murchison, pl. 10, fig. 2 [same figure of same two specimens in pl. 16, fig. 4, but described as from the Upper Llandovery].

1872 a Cornulites serpularius Schlot. ; Nicholson, p. 202, figs 1a, b [described and figured

‘ C. serpularius ’ as large taxon with thick, cellular shell walls].

? 1872b Cornulites serpularius ; Nicholson, p. 449 [no illustrations, but again described

‘ C. serpularius ’ as being large].

? 1873 Cornulites serpularius Schloth. ; Salter,

pp. 85, 93, 128, 177.

1875 Cornulites serpularius Schlotheim ; Baily, p. 41, pl. 19, figs 2a, b [copy of Murchison (1859)].

? 1882 Cornulites serpularius Schlot. ; Vine, p. 377 [described as solitary form with ‘very thick’ shell walls, but no specimens figured].

1888 Cornulites serpularius Schlotheim ; Hall,

p. 21, pl. 116 A, figs 5–8.

? 1888 Cornulites proprius Hall ; Hall, pl. 116, figs 15–21 [illustrated sectioned specimens showing thick shell walls formed solely of cellulae].

? 1923 Cornulites serpularius Schlotheim ; Reed, p. 269 [summary of previous work on

‘ C. serpularius ’, without illustrations].

1923 Cornulites serpularius Schlotheim ; Bather, p. 543, fig. 1.

1962 Cornulites serpularius Schlotheim ; Fisher, figs 78.1a.

? 1962 Cornulites serpularius Schlotheim ; Fisher, p. W137, figs 78.1b–e [figures described as sections of ‘ C. serpularius ’ but are copies of those figured by Hall (1888), pl. 116, figs 16–18, 20, 21) as C. proprius Hall, 1879 .

? 1972 Cornulites serpularius ; Blind, p. 5 [detailed description of shell structures similar to those seen in C. cellulosus , but without illustrations].

1974 ‘Unattached cornulitid’ Richards, p. 515, pl. 1, fig. 8.

? 1979a Cornulitidae gen. b, sp. a; Larsson, p. 210 [no illustrations, but described as conspecific with large, cellular forms previously placed in C. serpularius ].

? 1991 Cornulites serpularius Schlotheim ; Dzik,

p. 126, pl. 2, figs 5, 6 [specimens probably examples of C. cellulosus ].

Derivation of name: Latin cellulosus , full of small chambers, in reference to the distinctive shell structure.

Holotype: BU 4372, from the Much Wenlock Limestone Formation (Silurian: Wenlock: Homerian) of Dudley, England.

Paratypes: BU 4369, 4370, 4371 (with one longitudinal and five transverse sections), 4378 (with longitudinal section), 4380−84, 4386 (longitudinal thin section of specimen), 4388−91 (longitudinal thin sections of four separate specimens) and 4392; BMNH A230 (two specimens on same slab), A450, A455 (with one longitudinal and one transverse section), A459 (with transverse section), A460 (with transverse section), A845, A846, all from locality and horizon of holotype .

Diagnosis: Species of Cornulites with very broad annulations that are rounded in profile. Shell wall lamellar at apex, cellulae appearing between lamellae in mid-region, becoming dominant in apertural region; interior of apical shell region partitioned by tabulae running approximately perpendicular to lamellae, tabulae separated by dome-shaped camerae.

Description: Costae normally absent, very fine where present; annulations ∼ 1 mm thick at apical end of shell, up to ∼ 2 mm thick in mid- and apertural regions. Shell normally a straight-sided cone, but may show some sinuosity, particularly in early stages; aperture large, round. Specimens never found attached to shells of other organisms.

Specimens range in size from approximately 40 mm to at least 80 mm long, with maximum aperture diameter of around 25 mm. The largest specimen is BU 4392, which is 80 mm long, but with the apical region missing, indicating that it had a total length of approximately 95 mm. The holotype BU 4372 ( Fig. 2G) is approximately 62 mm long and free of matrix. The shell has 29 annulations, each of which can be traced around the entire shell exterior, although those closest to the aperture are somewhat irregular, with cellulae often visible in large numbers. The shell is straight-sided, but in cross-section changes from circular apically to elliptical aperturally; the aperture is oval, with a minimum diameter of 11 mm and a maximum of 16 mm. This may be due to crushing of the apertural region of the specimen, as the shell has longitudinal fractures on the plane running through the widest part of the aperture.

Towards the aperture of some specimens the annulations become obscure: the first 20 annulations of paratype BU 4369 ( Fig. 2E) are very regular and similar, showing no sign of cellulae, but the remainder of the shell up to the aperture is extremely irregular, with annulations difficult to discern and numerous cellulae present on the external surface. A similar pattern is seen in paratype BU 4370 ( Fig. 2F), with the first 20 or so annulations obvious, but those towards the aperture much harder to distinguish.

Not all specimens have continuously straight-sided shells: BU 4370 is straight-sided for the first 11 annulations (∼ 15 mm) before changing growth direction (see Fig. 2F). The shell is then straight-sided again for the rest of its length. Paratype BMNH A846 ( Fig. 2H) is very similar, showing a distinct change in growth direction after the 12th annulation, and with the annulations becoming difficult to identify in the apertural region. Paratype BMNH A845 ( Fig. 2D, I), in contrast, has a distinctly curved apical shell region before becoming straight-sided in the mid-region and towards the aperture. It becomes expanded and extremely irregular close to the aperture (see Fig. 2D).

The apertural shell region is well preserved in paratype BU 4371 ( Fig. 2J) and has a morphology not seen in other specimens: after consistently increasing in diameter up to the annulation closest to the aperture, the shell then begins to close again, such that the aperture is narrower in diameter (∼ 11 mm) than the final annulation (∼ 13 mm). The origin and development of this apertural constriction is discussed below.

Remarks: Cornulites cellulosus is the largest species of Cornulites in the Much Wenlock Limestone Formation. This, combined with the size of the annulations, their convexity in profile, and the thick, cellular walls, makes it easy to distinguish from C. serpularius , C. gremialis and C. scalariformis . C. cellulosus is most similar to C. proprius Hall, 1879 from the Niagara Group (Silurian) of Indiana. C. proprius reached shell lengths of up to 80 mm and also had a thick, cellular shell wall, but differs from C. cellulosus in that the cellulae of C. proprius form the entire thickness of the wall (see Hall, 1888: pl. 116, figs 1–21), and are not divided into zones by the lamellae, as in C. cellulosus (see below). The annulations of C. proprius are also much finer and more irregular.

CORNULITES GREMIALIS SP. NOV.

( FIG. 2K–O,?2P)

Derivation of name: Latin, gremialis , growing in a cluster from a stump, in reference to the species often occurring in clusters attached to a shell fragment.

Holotype: Largest specimen of cluster BMNH A470 , attached to indeterminate shell fragment, from the Much Wenlock Limestone Formation of Dudley, England.

Paratypes: Ten other specimens in cluster BMNH A470 , BMNH A483 a (six specimens attached to?brachiopod shell fragment) and A483b (two specimens attached to cephalopod); BU 4378 (cluster of six specimens) and 4385 (two fused specimens), all from locality and horizon of holotype .

Other material: BU 4368, a solitary specimen from the Much Wenlock Limestone Formation of Wren’s Nest Hill, Dudley, is questionably assigned to the species.

Diagnosis: Closely, rather irregularly annulated species of Cornulites , annulations alternating between pronounced and less pronounced. Costae fine but prominent; closely spaced.

Description: Shell normally slightly sinuous or twisted, particularly in early growth stages. Aperture increases regularly in size; aperture sometimes ovoid rather than circular. Specimens normally found in clusters, with apices attached to shell fragments; smaller specimens often using larger ones as substrate. Pronounced annulations not normally obvious in apical region of shell.

The holotype is the largest specimen on cluster BMNH A470 (see Fig. 2K–M). It is approximately 32 mm long, with an aperture diameter of around 10 mm. The shell surface has pronounced annulations spaced approximately 1 mm apart, and around five costae per millimetre .

Remarks: BU 4368 ( Fig. 2P) is a solitary, unattached cornulitid approximately 33 mm long, with a circular aperture ∼ 13 mm across. In some respects, particularly the arrangement of annulations and costae over much of the shell surface, it resembles specimens of Cornulites gremialis . At the apex, however, the annulations are thicker, and there is a suggestion that cellulae are present, making it much more like C. cellulosus . Its specific assignation is thus uncertain, but if cellulae are a feature of all cornulitids then there are sufficient similarities to C. gremialis for it to be questionably included in the species.

The overall shell shape, and closely spaced, irregular annulations that normally alternate in prominence make C. gremialis easy to distinguish from C. scalariformis . The conical shell morphology of C. gremialis is similar to that of C. cellulosus , but the two taxa can be separated by the differences in annulation and costal morphology. As noted above, the closest similarity of C. gremialis is to C. serpularius , but the irregularity of the annulations and prominence of costae in C. gremialis suggest they are not conspecific.

CORNULITES SCALARIFORMIS VINE, 1882

( FIGS 2Q, R, 3A–D , 7A, B )

? 1859 Cornulites serpularius Schloth. ; Murchison, pl. 16, fig. 5 [illustration of three specimens similar to C. scalariformis ].

* 1882 Cornulites scalariformis Vine , p. 379, pl. 15, figs 1, 9, 10.

1974 ‘Cornulitid transitional between attached and free-living formsRichards, p. 515, pl. 1, fig. 7.

1979a Cornulites ? scalariformis Vine ; Larsson,

p. 208.

? 1979a Cornulites ? cf. scalariformis Vine ; Larsson, p. 208 [no illustrations].

Material: The type specimens of C. scalariformis have not been traced. However, Vine’s illustrations (1882: pl. 15, figs 1, 9 and 10) are sufficiently clear to enable material to be identified with greater certainty than for C. serpularius . Four specimens attached to brachiopod BU 4373, three specimens attached to brachiopod BU 4374, and two specimens attached to brachiopod BU 4375, BU 4377a–d (four separate, unattached specimens), BU 4387 (longitudinal thin section of specimen), BMNH A847, and BMNH AN1177, all from the Much Wenlock Limestone Formation of Dudley, are placed in C. scalariformis .

Emended diagnosis: Species of Cornulites with widely spaced, narrow annulations separated by concave areas with thin, prominent costae. Aperture width expands slowly, particularly in later growth stages, producing essentially tubular shell morphology. Shell wall predominantly lamellar, but with occasional cellulae present adjacent to external annulations.

Description: Cornulites scalariformis has a shell length of up to at least 45 mm, as seen in BMNH AN1177 ( Fig. 2Q) and BMNH A847 ( Figs 2R, 3D ), with an aperture diameter of up to 5 mm. Annulations of BMNH A847 approximately 0.5 mm thick, separated by 1-mm bands in which costae are present; costal density around 5 per mm. The type specimen illustrated by Vine (1882: pl. 15, fig. 1) is incomplete, being around 15 mm long, with a diameter of ∼ 3 mm at apical end and ∼ 4 mm at apertural end, and shows eight annulations. The shell microstructure of C. scalariformis seen in thin section BU 4387 is described below.

Remarks: Cornulites scalariformis is easily distinguishable from C. serpularius and C. cellulosus by having a narrow, almost tubular shell even when large (aperture width approx. 3 mm at specimen length 12 mm, and 5 mm at length 45 mm) and well-spaced, prominent annulations separated by areas containing pronounced costae. Some smaller specimens, such as those attached to brachiopod BU 4374 ( Fig. 3C ), show similarities to C. gremialis , but the two species can be separated by the prominence of the costae and consistency of annulation morphology: C. scalariformis always has more pronounced costae than C. gremialis , whilst its annulations are generally very similar, and do not show the variation of C. gremialis .

The possibility cannot be discounted entirely that the species of Cornulites described here are ecophenotypes, with the variations in morphology reflecting adaptations to different ecological niches. However, the differences in ornamentation and shell shape appear to be significant and consistent, and the morphologies are therefore regarded as representing species.

Figures

Figure 1. Morphology of Cornulites. A, external morphology, showing apertural groove (ag), annulations (ann) and costae (cos). B, internal morphology in longitudinal section. BU 4390, C. cellulosus, apical and mid-region, showing lamellar shell wall (lam), with internal tabulae (tab) and camerae (cam) in apical region, ×6. Cellulae (cel) occur towards the apertural end of the specimen.

Figure 3. Cornulites scalariformis Vine, 1882, Much Wenlock Limestone Formation, Dudley. A, B, four specimens attached to brachiopod BU 4373. A, ×2.25, B, ×3. C, three specimens attached to brachiopod BU 4374, ×3. D, NHM A847, ×0.75.

Figure 5. Shell structures in Cornulites cellulosus sp. nov. A–C, BU 4371, pseudopuncta. A, transverse view, ×45, B, longitudinal view, ×45. C, ESEM image of transverse section (scale bar = 20 µm), shell interior towards top in all images. D, BU 4371, ESEM image of transverse section across mid-region shell wall, showing bipartite structure within lamellae, scale bar = 10 µm. E, BMNH A455, transverse section across apertural region showing concentric lamellae and apertural groove filled with cellulae, ×4. F, G, BMNH A459. F, transverse section close to aperture, ×3.25. G, close-up of concentric lamellae, separated by zones of cellulae, ×4.

Figure 6. Shell structures in Cornulites cellulosus sp. nov. A, B, BU 4378, ×60. Longitudinal section through apertural version of shell showing continuity between lamellae (lam) in the shell wall and partitions bounding cellulae (cel): selected lamellae are highlighted in B, ‘ag’ indicates the apertural groove. The pattern of overlap in the lamellae shows that growth occurred in the direction indicated by the arrows.

Figure 7. A, B, BU 4387, C. scalariformis, longitudinal

Figure 9. Development of cellulae in Cornulites cellulosus sp. nov. A, B, BU 4378, ×60. Longitudinal section through annulation in apertural region of shell, showing stacking of cellulae (cel) and lamellae (lam). Lamellae deposited at selected time intervals t0–t5 are highlighted in B; black arrows represent growth directions within and between cellulae. L1 and L2 are thickened series of lamellae without cellulae, and represent later shell layers that closed off and overgrew the lamellae and cellulae of an earlier growth stage.