Syrinyothyris and of Certain Carboniferous Brachiopoda Referred ~O the Genus Spiriferina

Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

162 DR. F. a. ~X'ORTtl O~ [vo1. lxxvi,

5. Ors SYR1NGOTHYRIS Wiachell, and CERTAI~ CARBONIFEROUS BRACHIOPODA REFERRED TO SPIRIFER1NA 1)'Orbiyny. By FREVEmCK J OH~ NORTH, D.Sc., F.G.S., Geological Depart- ment, National Museum of Wales, Cardiff. (Read January 7th, 1920.) [PLATES XI-XIII.] ~ONTENTS, Page L Introduction ...... 162 II. Syringothyris Winchell ...... 165 (a) History of Previous Research. (b) The Characters of the Genus. (c) The Type Species of Syringothyris. (d) The Subdivision of the Genus. (e) General Account of 8yringothyris in Great Britain. (f) Description of the British Speeies of Syringothy~s and their Mutations. (g) Pre-Carboniferous History of 8yringothyris. IlL Notes upon Certain Species inoorreetly referred to Syringothyris, and upon Spooies of 8yringothyvis supposed to possess Plieations upon the Fold and Sinus.--The Genus Syringopleura...... 189 IV. ' Spiriferina' lami',osa (M'Coy) : Tylothy~.is, gem nov.... 194 V. The Relation between Spiriferina and 8yringothyris ... 206 VI. Spir~erina D'Orbigny ...... 208 (a) Pre-Liassie Species referred to Spiriferina. (b) Pwnctospirifer, gen. nov. (c) Carboniferous Species of Spiriferina. VII. Summary of Conclusions ...... 221 VIII. Bibliography ...... 224

]~. INTRODUCTION. IN this paper an attempt is made to remove the uncertainty which has ~ existed in the naming of the British species of Syrinyothyris and of certain Carboniferous brachiopoda referred ~o the genus Spiriferina. Recent zonal work on the Carboniferous Limestone has demon- strated the desirability of (1)a more accurate knowledge of the essential characters of Syringothyrig, and (2) of distinguishing names for the forms occurring at various horizons. In the case of the Spiriferids with a lamellose surface-ornament, it has been customary to refer them all to Spiriferina, because there was no other genus for their reception--although it had long been recognized that few, if any of them, really belonged to that genus. When this investigation was commenced some years ago, it was intended to deal exhaustively with all the British Carboniferous shells referred to Syringothyris and Spiriferina, together with the relation of the latter to their pre-Carboniferous and post- Carboniferous representatives, and a preliminary paper dealing Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] SI'RtXGOr~}'RtS AND ,.~PIRII,'Is 163 with the ,generic clmracters of Syringothyris was published in 1913.1 Since that time, however, various circumstances have prevented its completion; and, when I resumed the work after returning to civil life, it seemed desirable to publish the results already obtained, leaving certain undecided matters relating to the extra-Carboniferous forms to be dealt with in a subsequent paper. As it now stands, the paper includes :-- (1) A general account of the genus 8yri~gothyris. (2) k discussion of the British species of that genus and their classification : the detailed consideration of the remarkable forms which occur in the Waulsortian knoll-phase limestones of the Caninia Zone in Ireland is deferred, until the return of favourable conditions renders further field-work possible. (3) A discussion of the generic position of ' Spi~4ferina' ~aminosa (M'Coy), and of the relation between the genus represented by that species and Syringothyris. (4) Notes on Spiriferi~m, and certain Carboniferous Spiriferids possessing a lamellose surface-ornament. The material used in the course of this work was in part collected by myself, but for the opportunity of examining a good deal of it I am indebted to the kindness of other workers, who have either lent me material of their own collecting, or have given me facilities for studying specimens in their care. Among the former are the late Mr. C. H. Cunnington, tlm late Dr. Wheelton Hind, Mr. F. Dixey, Mr. J. W. Jackson, Mr. J. F. Jackson, Mr. L. B. Smythe, and Mr. J. W. Tutcher, Prof. E. J. Garwood, Prof. S. H. Rey- nolds, and Principal T. F. Sibly; the latter include Mr. H. A. Allen, Prof. G. A. J. Cole and Mr. W. B. Wright (Geological Survey, Dublin), Dr. F. R. Cowper Reed and Mr. H. Woods (Sedgwick Museum, Cambridge), Dr. A. Smith Woodward and ]V[r. R. B. Newton (British Museum, Natural History), and Dr. R. F. Scharf[ (National Museum, Dublin). My thanks are also due to Dr. F. A. Bather for his advice upon a matter of nomenclature, to Mr. S. S. Buckman for suggestions relating to parts of the paper, to Principal T. F. Sibly for his ever- ready advice and encouragement, and to Dr. F. L. Kitehin for his kindness in reading and criticizing the manuscript.

Explauation of Cert~dn Terms used. In order to avoid ambiguity and unnecessal T repetition in the text, the precise meaning which is attached to cel~a~in frequently recurring terms is indicated in the following paragraphs :-- I. Terms applied to External Characters. Cardinal area.--A more ortless well-developed triangular area on each side of the delthyrial fissure, distinctly marked off from the general surface of the shell. ~cardinal arei~ -may be present i-n either valw-, biit that in th~ bi~achiaI valve is usually inconspicuous. In this paper ' cardinal area' always refers to that of the pediele-valve.

t North [25]. (Numbers in brackets refer to papers enumerated in the. Bibliography, wVIII, p. 224.) Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

164 vrt. V. a. ~rORTH O~ [V01. lxxvi,

Apical angle.--The angle enclosed at the apex of the cardi~l ares, Delthyrial angle.--The angle enclosed by the sides of the delthyrial flssm'e. Cardinal extremities.raThe terminations of the hinge-line. Lateral slopes.--Those portions of the surface of both valves which lie on either side of the fold in the brachial valve and the sinus in the pediole-v~lve. Costm.mThe components of the radial ornament of the shell of Syr/nfo- thyris and other brachiopoda have usually been termed 'ribs' and, less frequently,' eostm.' In order to assist in securing uniformity in terminology, it is proposed to adopt the suggestion of the late Dr. Ivor Thomas, and to use only the term ' costm' in this connexion. As that author pointed out? ' the radially disposed features hitherto called ribs appear to correspond to the " spirals" of gasteropoda,' and certainly not to their ribs, which are con. centrio in their arrangement.' On these grounds he used the term ' ribs' for the concentric rugm of the Producti, and ' costm ' for the components of the radial ornament.

II. Terms applied to Internal Characters. Delthyrlal supporting-plates.--The delthyrlal supporting-plates are a pair of lamellm in the pedicle-valve, which project from the edges of the delthyrium into the interior of the shell. These plates wm formerly termed' dental plates,' because they were supposed to exist for the support of the teeth ; but the late Dr. Ivor Thomas, discussing their function in Orthotstss, proposed the term ~ ' delthyrial supporting-plates ' : and $in~e, in the shells under con- eideration, the plates have almost disappeared long before they reach the hinge-line, it would seem that their primary function is not to support the teeth, but to strengthen and support the thin shell of the cardinal area where it is interrupted by the delthyrial fissure. Median septum.--Vertical shelly septa may exist in the longitudinal median plane of either valve ; and where, in order to avoid confusion, it is necessary to distinguish between them, the terms ' ventral median septum' and ' dorsal median septum' may be used. Ventral median septum.--In many braehiopoda the muscular scars in the pedicle or ventral valve are bisected by a distinct ridge, which in some genera becomes an elevated shelly plate or septum rising into the cavity of the shell, and extending farther towards the anterior margin ~ do the delthyrial supporting-plates on either side of it (see text-fig. 1, m, ~, o, p, p. 166). The function of the septum was presumably to afford addttiomtl m~rfa~e for the attachment of the adductor muscles, the scare of which extend over its sides. 8 The presence or absence of a ventral median septum~ and its nature when present, are important taxonomic characters in the group of shells under consideration. Dorsal median septum.--This is never more than a low ridge in the genera here discussed, and it is of no value in discriminating between species. Apical callosity.--There is frequently a tendency to an excessive deposition of shelly matter in the apical portion of the pedicle-valve, which becomes greatly thickened, or even completely solidified. Hall & Clarke described this character as the'apical callosity,' and it was referred to by the late Dr. A. Vaughan as the ' rostral callus.'

I , British Carboniferous Producti' Mere. Ge,ol. Sure. Gt. Britain, Palm- ontology, eel. i, pt. 4 (1914) pp. 225-26. s , The British Carboniferous Orthotetin~ ' Mere. Geol. Sure. Gt. Britain, Palseontology, eel. i, pt. 2 (1910) pp 100-101. 3 j. Hall & J. M. Clarke El4] p. 53, fig. 42. The occurrence of the muscular scars on the sides of the ventral median septum has been recorded in Spiri- retina walcotti by Hall & Clarke, op. cir., and in 8p. octoplicata by W. King, 'Monogr. Perm. Foss.' Pal. Soc. 1850, p. 68. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] 8YI~I~'GOTttYRIS ~kND 8PIRIFERI~VA. 165

The apical callosity usually increases with the age of the individual, and in ephebic shells completely envelops the earlier-formed portions of the delthyrial supporting-plates and associated structures. Although few of the Spiriferid shells are entirely without an apical callosity, the structmre is especially well developed in certain species. When present, it gives a very characteristic appearance to internal casts of the pedicle-valve (see P1. XIII, fig. 12). Transverse plate and syrinx.--The transverse plate is an horizontal septum between the delthyrial supporting-plates, and projecting into the interior of the shell from the anterior end of the apical callosity. It may exist as a simple horizontal lamella, or, as in 8yringothyris, it may bear on its lower surface a split tube or syrinx. (See fig;, 1 b, p. 166.) l

II. SI~RtNOOTHYRIS Winchell. (a) History of Previous Research. The earliest account of a species now referred to Syrinyothyris was given by William Martin in 1796, when he described Anemia vuspidata ~rom the 'Limestone' of Castleton (Derbyshire).~ Thirteen years later he again described and figured Conchylio- lithus anomites eusjaidatus in his ' Petrifaeta Derbiensia,' s but dealt only with the external characters of the shell, there being nothing in his specimen to suggest the existence of internal structures. Martin compared the shell to the recent -4nomiw, and suggested that the large triangular area was the surface of attachment, and that the muscles of fixation passed through the narrow fissure (that is, the delthyrium). Although several authors refera'ed to and illustrated this remarkable shell during the following thirty years or so, ~ no additional characters were discovered until 1838, when G. P. Deshayes recorded the presence of a convex plate (deltidium) covering the delthyrial fissure. That author stated that the plate was provided near the apex with an oval foremen. 5 A similar opinion was expressed by Thomas Davidson, who 9figured an internal east of ' Spirifera euspidata' 6 from the ' dolomitie Carboniferous Limestone of Breedon Hill, in which there is evidence thai, the deltidium was in reality perforated by a circular foramen.'

The words 'upper' or 'lower,' and 'above' or 'beneath,' used in connexion with the transverse plate and syrinx, refer to the structures when the shell is held with the apex of the pedicle-valve away from the observer, and, the beak of the brachial valve pointing upwards, as in the sections in fig. 1 (p. 166). They have no reference to the position of the shell during the life of the individual. ~" W. Martin [22] p. 44 & pl. iii, figs. 1-4. 3 W. Martin [23] pl. xlvi, figs. 3-4 ; pl. ~lvii, fig. 5. Martin's type-specimen isnow preserved in the Sowerby Collection, in the Geological Department of the British Museum (Natural History). 4 j. Parkinson ~26] vol. iii, pp. 234-35 & pl. xvi, fig. 17 ; J. Sowerby [33] pp. 42-44 & pl. cxx, figs. 1-3 ; J. Fleming, ' A History of British Animals ' (t82S) p. 563. b Deshayes in'Histoire Naturelle des Animaux sans Vertebras' 2nd ed. vol. vii (1838) p. 368. Davidson [4] explanation of pl. viii, footnote, and pl. ix, figs. 1-2. The specimen is now in the Davidson Collection, in the British Museum (Natural History). Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

Fig. 1.--Sections across the beak of the Tedicle-vah~es iJ~ Syringo- thyris and other .qe~era. (iYatural size, except w]lere otherwise stktted.) q:3),

a-c:Syringotliyris r rincipo~is (Kl) , 5, 11, and 15 ram. respectively from the apex. d:8. c.~spi~lata mut. exo~eta (C), 13 mm. from the apex, at a stage comparable with (b). [Specimen in the Britilh Mu~um (Natural History), Geological Department, Reg. B 316.] e-~ S. elongata (D). Section at a aCage comparable ~4th (b). f-~S. e~ongata (C), 10 mm. from the apex, at a stage comparable with (b). [Specimen in the British Museum (Natural History), Geological Department, Reg. B 393.J g-~Sp~rifer ma~yi Hall, after H. S. Williams. h~-The plaCe-system in a specimen incorrectly referred to Sp. distans, after Davidson [5~ pl. xL~iii, figs. 4 & 5. i-j=Spirifer du4a~icwosta (I).~), 3 and 6 ram. respectively from the apex ; compare with (a) and (b). k=Sp. dista~, with section of the del~idium. l=Syringospira prima, after E. M. Kindle. m-n-~ Ty~othyris ~minosa (C), 3 and 6 ram. respectively from the apex. o-~Spiriferina walcotti. Section 3 ram. from the apex : compare with (c). p----Punctospirifer scabricosta (S), 3 ram. from the apex : compare with (o). r-t~-Cyrtina heteroc~ita (rX 2); s-t, the median septum and its partitioned tube greatly enlarged: s, where completely developed; t, near its termination. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2 ] SI'R[NGOTIII-RI8 A~D SPIRIFER/NA. l{j~

As a matter of fact, however, the specimen in question shows, not It foremen in the dcltidium, but the remains of the rod-like portion: of the matrix which originally filled the cavity of the syrinx (see Pl. XI, fig. 6). On the strength of the supposed possession of a 1)erfon~ted deltidial covering, M'Coy 1 and Davidson e both assigned czc,spidata to the genus b~rtia, an opinion which Davidson, at least, abandoned later. In 1855, M'Coy referred to the discovery in S. c,esTidata of ~u ' internal deep-seated pseudodeltidium, without perforation. 'a Although, in the absence of a figure, the precise nature of this ' internal pseudodeltidium' cannot be determined, it was in all probabilit S the 'transverse plate'; since it is quite clear from the context that it was a structure distinct from the ordinary delthyrial covering, for which M'Coy used the term 'exten~al pseudodeltidium.' In 1859, L. G. de Koninck described and figured a specimen which he referred to ' Sdpir(/'er distans ' 4 from the Carboniferous Limestone of VisP, in which the tra~tsverse plate and syrinx were clearly shown, but the structure was not made the basis of a new genus. ~ After the publication of Winchell's paper deseril)ing the characters of a new genus, S.qrilt.qothffris, Davidson called attention to the fact that 1)e Koninck had already made known the existence of similar structures, and gave what he (Davidson) described as enlarged drawings of the Belgian specimens." vFhe'e illustrations depict the syrinx as being composed of two plates al)plied to the inner sides ,'ff the delthyrial supl)orting-plates and curling under along the median lira, to enclose ~l tul)ular canal, incomplete both itl)ove illl(l ][x,l()w (see rig. 1 h). This structure, as interpreted by l)avidson, differed from the syrinx of S. cu.~pidr162162162which is split along its lower surface on!v. Davidson's interpretation; was, however, ine(~rreet, for ])e Koninck's ~)wn figures, although small, clearl.v indicate a tl-ansverse plate and syrinx differing in no wa S from that of S. cuspidctter Davidson's tigures als,~ differ from De Konilaek's, and from the conditions which obtaiu in S. c~lstJidata,, in th:Lt the lamelbe which contribute to the formation of the split tube are drawn as if they were applied to the delthsrial supporting-plates at tlae level of the cardinal area, and extended all the way from the beak to tim hinge-line, while in S. ettspida/a the tlansverse plate plunges beneath the level of the area, and st~ps short some dista~lce from the hinge-line. Apart from the question of the precise nature of the structures indicated

1 M'Coy [21] p. 426. ~ I)avidson [1] p. 470. :~ M'Coy ibid. loc. cir. 4 De Koninck [18] p. 50 & pl. ii, fig. 6. s The specimen, as is shown elsewhere in this paper, was not correctly referred to Spirifer distans. 6 Davidson [2] p. 312 & pl. xiv, fig~. 7-9 ; also [5] p. 280 & pl. xxxiii, figs. 4-5. Q. J. G. S. No. 802. ~- Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

168 ~J~. F. a. ~-o~'rR o~ Jvol. lxxvi,

9by Davidson's figures, the drawings are obviously incorrect, because, although they are supposed to represent the same specimen seen from above and below, ill his fig. 4 the lateral portions of the transverse plate are convex towards the area, while in his fig. 5 they are concave, so that the upper slit occurs at the bottom of a semicircular trough. In the circumstances, we can only regard Davidson's figures as misinterpretations of true Syringothyroid characters, and it then transpires that De Koninck was the first to call attentinn to the transverse plate and syrinx. No reasonable grounds exist for supposing that there are two types of syrinx, the one occurring in Syringothyris cuspidata and the other in STirifer distans: for, in addition to the evidence already adduced, the structures indicated in Davidson's figures have not since been observed in any other specimen, and it is shown elsewhere in this paper that STirifer distans does not possess a syrinx at all. In 1863, A. Winchell described a Spiriferid shell 1 (from the yellow sandstones lying beneath the Burlington Limestone [Lower Mississippian] at Burlington, Iowa (U.S.A.), which possessed a plate-system in the ventral valve sinfilar to that already noticed by L. G. de Koninck; and, apparently unaware of that author's communication, made the character the basis of a new genus, to which he gave the name Syringothyris, with S. typa Winchell as the genotype. He described the syrinx-bearing plate in the following terms :-- ' transverse plate connecting the vertical dental lamellm, arched above, and beneath giving off a couple of median parallel lamellm which are incurred so as to nearly join their inferior edges, thus forming a slit-bearing tube which projects beyond the limits of the plate from which it originated into the interior of the shell.' Two years later, F. B. Meek announced the discovery of a similar structure in 'Spirifera cuspidata' from Millicent (Kildal~). 2 In 1868, W. King published a ' Monograph of Slairifer cuspi- darns Martin,' in which he described in detail the characters of the species, and figured a series of sections across the beak of the ventral valve at various distances from the apex, illustrating more clearly than had hitherto been done, the structure of the transverse plate and syrinx, s The precise nature of the transverse plate and syrinx, and their relation to the delthyrial supporting-plates and to the apical callosity, was described by me in 1913) The classification of the American members of the genus has been discussed by Schuchert 5 and by Weller 8 The shell-structure of Syringothyris.--Themicroscopie stl-acture of the shell of Syringothyris gave rise to a discussion Winchell [38] p. 6. 2 Prec. Acad. Nat. Sci. Philad. set. 2, vol. ix (1865) p. 276. a King [17] pp. 1-23. 4 North [25] pp. 397-99. 6 Schuchert ~29~. 6 Weller [36]. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2~ s YRI,YGOTtt YRIS AND SPIRIFERIJ~A. 169 between Carl)enter, Davidson, King, and Meek, which, so far as the three first-named authors are concerned, developed into what was little less than a personal quarrel. Owing to its mode of preservation, Martin's specimen retained no trace of the original shell-structure, for, as that author observed :-- ' it is a complete change, evidently retaining the external shape and markings of the original, which has been displaced by limestone similar to that in which the fossil was lodged.' (~22~ p. 46.) Winehell stated that in s t;vipa the shell was fibrous aud impunctate, but both Meek and Carpenter observed punctations in specimens of that shell sent by Winchell himself. Carpenter , corroborated by Davidson, maintained that the shell of S. cuspidata was impunctate; but King, on the other lmnd, claimed to have seen on the smfface of a specimen from Tuam (Galway), in the collection at Queen's College, Galway, 'patches of faint slightly-raised oval impressions' which he compared with similar markings on the surface of 1)ielasma hastata, known to be a punctate shell. King was so certain that the shell of S. cuspidata was punctate, while Carpens so emphatically denied the presence of punctm, that Meek, who had proved the existence of punctations in an example of S. cttspidata from Millicent, suggested the possibility of there being two distinct forlns included under one and the same specific name, of which one forllr possessed the internal plates of Syrinqotl~?/ris and a punetate shell-structure, while the other had no Syringothyroid characters and an impunctate shell. Later, Carpenter himself admitted the existence of punctations in some of the Milliceut specimens, and inclined to the opinion already expressed by Meek. 1 The shell of S?p'in.qothyris was undoubtedly puncta:te, aud the reason for Carpenter's assertion to the contrary is not difficult to find. In many fossils from the Carboniferous Limestone the shell- substance has been recrystallized, and in such cases the external appearance of the shell remains unaltered, but its intimate structure is obliterated. M:uly specimens frofn Ireland and Derbyshire

For this discussion, see the following papers :--~'. King,' Monograph of the Permian Fossils of England' Pal. Soc. 1850, p. 110 ; W. B. Carpenter, ' On the Intimate Structure of the Shells of the Brachiopoda' in Davidson [3~ pp. 23- 40, and' On the Minute Structure of certain Brachiopod Shells' Ann. & Mag. Nat. Hist. ser. 2, vol. xvii (1856) pp. 502-506 ; F. B. Meek, ' Observations on the Microscopic Shell-Structure of Spirifer c~spidatus Sowerby ..... ' Proc. Acad. Nat. Sci. Philad. ser. 2, vol. ix (1865)pp. 275-77; W. B. Carpenter, ' On the Perforate Structure of Spirifer cuspidah~s' Ann. & Mag. Nat. Hist. ser. 3, vol. xix (1867) pp. 29-31 ; F.B. Meek, ' On the Punctas Shell-Structure of Syringothyris ~ Amer. Journ. Sci. set. 2, rot. xtiii (t867) pp. 407-408, also Geol. Mag. 1867, p. 315 ; W. B. Carpenter, ' On the Shell-Structure of Spirifer cuJspidatus, & of Certain Allied Spiriferidm' Ann. & Mag. Nat. Hist. ser. 3, vol. xx (1867) pp. 68-73; W. King,' Notes on some Perforated Palmozoic Spiriferidm' Geol. Mag. 1867, pp. 253-56; and T. Davidson, 'Perforate & Imperforate Brachiopoda' Geol. 1Yfag. 1867, pp. 311-13. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

170 mr. F. J. N(-)RTII ()N EVO]. ixxvi, are of this nature, and show no trace of |)untrue; but, wherever the shells are suitably preserved, as (for cxaml)le ) in the Za])hrentis Zone in many parts of the South-~Vestcrn l'rovince, the i)unctate structure is quite evident.

(b) The Characters of the Genus. The shell of ~.~iqJ.in.qoth!lri.s. is Spiriferoid in form and variable in size. The dorsal valve is slightly convex; while the ventral valve is more or less acutely pyramidal, owing t() the presence of a high cardinal area. The hinge-lilm is straight, and the cardinal extremities are usually rounded, l)ut arc sometinies slightly 1)reduced and angular. The cardinal area is high, and either slightly concave, flat, or even convex (see fig. 3, p. 178). In the lmiotyl)e of S. cnspidota the area is practically flat. The apical angle varies, but tlm average for the genus is about 10(P. In well-preserved specimens the al~,a is clearly differentiated into three regions 1)v a l)air of divergent lines extending from the apex to the hingc-li(m. The whole area is marked i)v horizontal striations representing lines of growth, and the centl~ll vwtion bears, in addition, vertical striations (see fig. 2).

Fig. 2.--The th~'ee/old dieisio~t of the .reu iJl S.vringothyris (dict.ep'ammcttiO, lm'sed tq)on / he .~'peeime t~ .//.(I,~'ed iil P/. XII I, fig. 2.

The central region of tile area is bisected by a triangular delthyrium about three times as high as wide. The delthvrial angle averages about '25 ~ The delthvrium is closed by a convex delti- diuln, 1 not perforated 1)y a f()r~'tmen as in C!/r/i

I In a previous paper (North [25] p. 395) I used the term 'pseudodeltidium' for this plate. I am indebted to Mr. H. Woods, F.I~.S., for pointing out that Von Bach introduced the term ' deltidium ' for the delthyrial covering-plate in the Teh)tremata, to which Sgri~tgothyris belongs, and that the use by Americ,~n paheontologists of ]3ronn's term 'pseudodeltidium' in this sense is incorrect. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] ,," rJr r, JR1,v .~ ~ 1~ ,VPmIF~:~ZX.~. 171

,causes :~ wide rounded wave ill the anterior margin. Except for a slight median depression on the fohl, extending for a short distance from the beak of the braehial valve, the fold and sinus are smooth3 The lateral slopes of each valve are typically ornamented by radial costa:, decreasing gradually in size towards the cardinal extremities. Tim cost~ are regularly broad and depressed, and a:re separated by narrow, shallow fmTOWS ; they are always siml)le, :rod originate at the hinge-line, the last four or five on either side ,being small and inconspicuous. They are frequently obsolete, especially in the pediele-valve, the lateral slopes of which may be quite smooth. The eosta~ are crossed at irregular intervals bv inconspicuous lines of growth, which are more closely crowded towards the margins of the valves. The external sm'face of the shell in suitably- preserved specimens is marked bv a minute textile-like ornamentation, and the shell-substance is ~ninutely 1)unctate.

Internal structures: Pcdiele-valve.--There is apair of delthvrial supporting-plates which extend completely across the cavitY" of the valve, to the floor of which they are attached for ab(mt one-third of its lm,gth. The transverse plate extends for from a third to half of the distance from the apex to the hinge-line, and has on its lower surface the split tube or syrinx. 9The tran.~verse plate arises in the apex of the shell as an arch-shaped plate between, and applied to, the delthyrial supporting-plates .... The syrinx occurs on the lower mu'faee of the plate, and is formed by two nearly- parallel lamellm, the free ends of which curve towards one another. As the distance from the apex of the shell increases, the lateral portions of the arch become thinner and shorter, until only its roof persists as the tube-bearing or tran,~verse plate-a slightly convex plate down the centre of which is a median ridge on the upper .~urface, and a ~lit tube or syrinx on the lower .~urfaee. The transverse plate itself then dies out. Its distal end is concave tow,~rds the apex of the valve, and the syrinx projects for ~ short distance beyond it ' -~ (see figs. 1 ,t-lc, p. 166). The earlier-formed portion of the syrinx tends to be obscured by ~he deposition of shelly matter (an apical callosity) during the growth of the shell (see li~. la, p. 166). Braehial valve.--There is a low median septum extending for about half the length of the valve. Spiral coils are known to be present, but in the specimen in which they were developed by the Rev. Nornlan Glass, they were not in sihe, and the nature of the primary lamellte and jugum is, as yet., unknown. The threefold division of tile area.--In well-preserved specimens the area is clearly divided into three triangular portions, of which the two lateral p~n'tions are marked by horizontal strim and are similar in intimate structure, namely, ill the possession 1 Two species with plieated fold and sinus have been referred to Syringo- thyris, but in neither case has the existence of a ,~yrinx been proved. These are fiflly discussed elsewhere (see p. 189). North [257 pp. 398-99. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

172 D~. F. J. ~'OI~TH ON [vol. lxxvi, of punctm, to the remainder of the shell. The central region, however, which is bisected by the delthyrium is impunctate, and ill addition to the horizontal lines of growth, is marked by distinct, but irregularly-spaced, vertical stri~e (see fig. 2, p. 170). t The vertical stri~e are the external indications of eel~ain internal peculiarities of shell-structure. In partly-exfoliated specimens this peculiarity evinces itself in a series of trough-like depressions, which suggest that the shell-tissue of the cardinal area was tra- versed by numerous canals, perl)endicular to the hinge-line. This, however, does not appear to have been the case, as no trace of a canal-system is apparent when thin sections of the area are examined under the microscope, and there is never any infilling by darker matrix such as is frequently the case with the tubular punchttions of the shell of Sl)ir(ferina walk,o/t/. The stri~t~ are connected with a series of denticles along the hinge of the pediele- valve, and articulate with a row of pits in the hinge-line of the opposite vah-c. Davidson suggested that the function **f the denticles was to steady the valves, especially when the hinge-line was long and the teeth not very large. ~ As the shell increased in size by the peripheral addition of shelly tissue, and the dentielcs recedt, d front the apex, their course was marked by a series of lines, approximately parallel, and perpendicular to the hinge-line. In a discussion of the origin of the vertical stria,, ,I. Young suggested that they were due to the presence in the hinge-line of the living animal of fibres of aragonite, which ' being harder than the ordinary calcite of the shell went to the formation of the row of dentieles." If this were the case, the different degrees of resistamce offered by the two minerals to subsequent alteration would account for the appearance of the canals in some specimens, for amgonite, although harder than calcite, is more soluble than that mineral. The vertical stri~e are only to be seen in Sl~cimens which have undergone but little mineralogical change, and they arc often completely obliterated by reerystallization or exfoliation. The denticles are very rarely seen. Occasionally, as in the case of the specimen upon which Young's observations were based, the two valves are separated in such a way as to expose them clearly. In recrystallized specimens they have quite disappeared ; but, according" to Young, in some cases they are rendered visible by scraping and etching with acid. In tl{is way he proved their existence in

1 Dr. Ivor Thomas described in some of the Orthotetinm 'a separate areal portion in the fern of a raised triangular platform, marked off around the delthyrium. This secondary area is distinguished by vertical striations in contrast to the horizontal striations of the larger surrounding primary area.' In Sg~'ingothyris, however, the plane of the 'area is not interrupted by the central portion. 2 j. Young, ' On the Denticulated Structure of the Hinge-Line of Spirifera trigotm~is' Geol. Mag. 1884, pp. 18-20. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] SYRINGOTHt'R15" AND 8PIRIFERINA. 173

Sip. striata, and it is not unreasonable to assume their presence in all cases where there is a vertically striate area. The taxonomic value of these vertical striations lies, not in their presence, bu~ in their distribution on the cardinal area. In t~hose species of Spirifer in which they have been observed, they occur over the whole area; but in Syringothyris they appear to be confined to the central portion of the area. 1 There is a similar division of the area in Spirlferina. Unfortu- nately, owing to the mode of preservation of the shell, the vertical striations are not usually visible ; but, when present, they afford a convenient means of distinguishing between Syringothyris and those species of Spir~er, such as distans, in which there is an elevated area.

(c) The Type Species of Syring~Jthyris. Three species have at different times, and by val~ous authors, been regarded as the genotype of Syrin.qothyris, namely :--S. typa Winehell, S. c~tspidata (Martin), and S. carteri Hall. Winchell, in establishing the genus, took as his type a form which he regarded as a new species, and to which he gave the name t?/pa. A year later Meek suggested that Winchell's specimens were identical with Spirifer cuspidatus (Martin)2; and in 1867 W. King wrote :~ ' It may be safely assumed that Spin,far cuspidattes and Syringothyris typa are one and the same species,' and ' it will necessarily follow that Martin's specific name, having priority, must be adopted in preference to the one proposed by Professor Winchell.' This opinion was subsequently confirmed by Thomas Davidson. '~ The two species, however, differ in the following respects :~ (1) The area of typa is relatively lower than that of cuspidata (as typified by the holotype) ; is slightly concave, but never reclined, as is frequently the case in Martin's species ; and is always symmetrical, while that of c~spidata is frequently asymmetric or twisted. 4 Winchell, in his diagnosis of typa, stated that the area was flat ; but he probably used the term in a relative sense, for Davidson gave illustrations of specimens (provided by Winchell himself) ~vhich suggest that the ~area was slightly concave, ~ and Weller has also described the area of t.ypa as concave with the curvature

l De Koninek described the threefold division of the area in a specimen of Syvingo~hyris, and gave to the central portion the name 'pseudodeltidium compl~mentaire.' (De Koninek [181 p. 50 & pl. ii, fig. 5). 2 In F. B. Meek & F. V. Hayden, ' Palseontology of the Upper Missouri' Smithson. Contrib. Knowl. vol. xiv (1865) p. 19. a Davidson [51 p. 279, The comparison is based upon Martin's figures and the original specimen of S. c,tespidata on the one hand, and on the other hand, Winehell's description of S. typa, together with figures given by Davidson and Waller. Davidson E2] pl. xiv, figs. 2 & 3. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

174 m~. F. J. NORTH ON [vol. lxxvl,

towaMs the beak, while his figures of sl)ecimcns from the same horizon as Winchell's lead to the same conclusions. 1 (2) There are slight differences in the nature of the transverse plate and syrinx. In S. cuspidata (the holotype) the delthyrial supporting-plates are nearly parallel ; the transverse plate is nearly flat, and diverges rapidly from the plane of the cardinal area, while in S. typa the delthyrial supporting- plates are distinctly divergent, and the transverse plate is convex upwards and never deeply buried. Since, in addition to the above-mentioned differences, S. t.qpa was derived from a low horizon in the Mississippian of North America (equivalent to tim Lower Avonian of this counhT), and S. c~spidal~l (the holot.vpe) from the Upper Avonian (Dib'wJ~oph!/l/l~m Zone) of England,'~ c~ it seems desirable to reg'ard theln as distinct species, in which case SyriJ~.qolllyris typa remains as the genotype. Schuchert e.~l)resse(I the el)injure that S. t!!lm is identical with a form previously described by James Hall as Spirlfi, r cartm'i ~; and, if that be so, the latter name has priority. The specimens upon which the species earleri was f()unded were fmgment~try, poorly ]~rcservc(1, and comprised brachial valves only. Schuchert wrote :-- ' I have seen nearly one hundred individuals and know of but two specimens that have both valves in articulation. Neither of these specimens shows the characteristic transverse plate of Winchell's genus. In several separated ventral valves, hc,wever, I have seen the eamfliferous plate . . . .~o that there ~-an be no doubt about S. carteri being a true Syvi~gothyvi.r Despite the uusatisfactory c(mdition of the material, Schuchert maintains that there is no visible difference between S. l;u and ,S. carteri ' other than those due t() the conditions of preservatiou, insufficient to separate theln as distinct species.' 3 Meek 4 and Herrick 5 also emi)hasized the similarity l~tween N. ca rteri and S. t//p~l, while the latter author regarded i,oth names 9"~s synonymous with S. cuspidata Martin. Hall & Clarke, however, recognize S. typa and not Hall's species as the genotype of Syrinyothyris6 ; while Wcller maintains that, since the species S. carteri is iml~l~ectly known, being found~d upon bmchial valves only (which, as is shown elsewhere in this paper, are not satisfactol:y criteria for discrilninating t)ctween species of Syrinyothyris), and ' since it seems impossible to determine surely what carter/ really is,' S. typa should still be regarded as the genotype.7 Not having had an Opl)ortunity of examining any American material, I can only cite the conflicting opinions of the

1 Weller [36] pp. 395-97 & pl. lxix, figs. 1 5. .2 'Description of Palmozoic Fossils' 10th Ann. Rep. N.Y. State Cab. Nat. Hist. 1857, p. 170. Sehuchert [29] pp. 31-32. 4 , A Report on some of the Invertebrate Fossils of the Waverley Group and Coal-Measures of the Ohio' Rep. Geol. Surv. Ohio, vol. ii, pt. 2 (1875) p. 285. 5 Bull. Denison Univ. vol. iii (1888) p. 41. 6 Hall & Clarke [14] pp. 8, 48, & 50. 7 Weller [36~ p. 397. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] ~NYIH_VGOTH Y2r I~ .t~ D ~b'IVRI I,'ERI.\\-L 175

authors mentioned, and leave the validity of the speeies carleri an open question. Fortunately, however, this does not affect the issue so far as British species are concerned, because both 8:yHngo- th!/ris ty:pa and S. carleri (if they are distinct forms) differ in certain respects from the holotype of S. cttspidcrta. The differences between the first and last of these have already been considered ; between carteri and the holotvpe of c~,sTida/r they are as follows :--- (i) In cavtevi the area is concave; in cuspidata flat or reclined. (ii) In carteri the lateral slopes of the ventral valve are distinctly tumid; in cuspidata they are usually flat, or even concave. These differences are practically the same as those which exist between the Syringoth?/ris aft. e~tspidttta auett. (for instance, the form referred to in this paper as S. euspidata rout. c.q~'to- rh!p~cha) from the Cl#~slopo~'a- and Ztt:plt~'r tis-Zones of the South-Western Province, and the typical ,b'. cuspidata (Martin), from the Upper Avonian of Castleton (Derbyshire). The shell described and figured by Dr. C. D. Walcott as S. cuspidata more closely resembles ,b'. ccr162 than the former species. 1 From the evidence adduced, the following conclusions may be drawn :-- (i) The genotype of Syri'ngott~yris is S. typct, Winchell, if ca~'teri and t,yp, are distinct, or if the former is not a valid species, which the balance of the evidence seems to favour; but, if the two names be synonymous, carteri takes priority as the type-species. (ii) Both typa and carte~4 have much in common with the S. aft. c~t,spidata of the South-Western Province : therefore, in terms of the British species, the genotype belongs to that group of Sy~ingothgris which is characteristic of the lower part of the Avonian, wherein the cardinal area is more or less concave and the lateral slopes of the ventral valve are tumid ; and not to the true cuspidata group typified by Martin's holotype from the UH)er Avonian. S. cu.spiclata has also been regarded as the genotype of ,~'pir~/'et" Sowerby. When that genus was established by J. Sowerby in a paper read before the Linnean Society in 1,ql'L he mentioned Anemia striata Martin as a typical species. That paper was, however, not published until 1819, u and in the meantime, in Mineral Conchology ' vol. ii (1816) p. r the same author gave the species ' cuspidatus' as an example of Spir~'ec. Thus, s() far as the date of publication is concerned, the name c~tspidata has priority, and this was the attitude adopted by King, s Meelr a ~md others; but Davidson has shown conclusively that Sowerby himself

1 , pal~eontolog'y of the Eureka District' Monogr. U.S. Geol. Sm'v. 8 (1884) p. 219. 'Some A:ccount of the Spiral Tubes or Ligaments in the Genus Tereb~'~ttda of Lamarck, as observed in several Fossil Shells' Trans. Linn. Soc. vol. xii (1819) pp. 514-16. King [16] p. 125. 4 In F. B. Meek & F. V. Hayden, ' Palmontology of the Upper Missouri' Smiths. Contrib. Knowl. vol. xiv (1865) p 19. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

176 mr. F. J. NORTH ON [vol. lxxvi, always regaMed S. striata as the type-species of Spiri]'er i : and since it is now known that euspidata does not belong to that genus at all, there are no reasonable grounds for disputing the claim ,~f the former species to be regarded as the genotype of Spirifer. ~"

(d) Tile Subdivision of the Genus. The species which have been referred to Syrinyotl,//ris, and the Sl~cific names which have been applied to various forms occurring in tiffs country are as follows :-- 1. Spirifer distans Sowerby. 2. ( Cyrtia) laminosa M'Coy. 3. ( Anomites) subctmicus Martin. 4. (Anomites) cuspidatus Martin. 5. (Slririfer) carteri Hall. 6. Syringothyris typa Winchell. 7. Syriugotbyvis texts Hall. (}f these, the first three are shown in this l)al~.,r to belong to other genera, while the last three are the names of American species to which certain shells from the Lower Avonian of the South-Wcsfcrn Province have been doubtfully referred. The use of these nanlcs is, however, undesirable, partly because of the absence of facilities for comparison with the type-specimens, and partly because American palmolltologists are themselvcs not in agreement as to their precise significance (see p. 174.). Therefore, so far as British Sl)Ccimens are concerned, the only valid specific name already proposed is c~ls])idala. ] f isolated specimens from various horizons and different localities were considered separately, it would be comparatively easy to divide the genus into many apparently well-defined species; but, as is often the case, the greater the nunlber of specimens examined the more difficult specific differentiation becomes. The changes ,h,pendent upon time alone are small and progressive; but, in deposits such as the Waulsortian ' knoll '-limestones, where the conditions of deposition were somewhat abnormal, the shells at any one horizon exhibit considerable variation. The characters in which variation occurs are as follows :-- (1) The height of the area.--Genorally speaking, the area increases in height from lower to higher horizons ; but, as with the other characters, the greatest amount of variation is due to the existence of a special environment, and not to considerations of time (see fig. 3, p. 178).

I Davidson [3] p. 81. Since these lines were written, Dr. J. Allan Thomson has cited the case of Syringothyris cuspidata and Spirifer as an example of the necessity for a list of nomina consercanda approved by the International Zoological Congress, so that in this and similar cases, it should not be necessary by the strict application of the law of priority to displace a name which, like Spirifer, has been in general use, in a well-defined sense, for a long period. See ' Brachiopod Nomenclature : Spb'~fer & Syri~gott~yris ' Geol. Mag. 1919, pp. 371-74 ; also S. S. Buckman, Geol. Mag. 1920, pp. 18 20. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 27 SYRI5~GOrI4Ytr X,~TD SPIIr 177

(2) The curvature of the area.--Although exhibiting a fairly-wide range of variation, the nature of the area in forms from one horizon, and in rocks formed under similar conditions, is fairly constant. In common with Spir~fer, Spiriferina, and similar forms, the cardinal area is marked by numerous parallel transverse striations, which in suitably- preserved specimens are seen to be continuations of the concentric lines of growth that appear on the remainder of the surface of the shell, and therefore have a direct bearing upon the height and curvature of the area. The addition of new shelly matter in the Brachiopoda takes place at the free edges of each valve, l If the deposition along the hinge-line is, throughout life, less than along the anterior margin of the shell, the area will be strongly curved, as in most species of Ss as the marginal growth becomes more nearly uniform the area increases in height and decreases in curvature ; while, when the growth along the hinge-lines is in excess, the area will be reclined, that is, convex. Usually, in the genera under consideration, the rate of growth along the hinge-line either increases or decreases progressively, during the life of the individual, so that, although in its apical portion the area is curved, it becomes practically flat towards the hinge-line. In a general way, the curvature of the area is of taxonomic value, but it sometimes happens that in one and the same specimen the area is convex on one side of the delthyrium, and concave on the other side (see P1, XI, fig. 7). The area, although normally symmetrical, is sometimes distinctly twisted; but that feature depends upon differences in the curvature of the area on either side of the delthyrium. The w~riations in the height of the area are, of course, accompanied by changes in the profile of the pedicle-valve, and the curvature of the floor of the sinus along a median line from the beak to the anterior margin is, especially in imperfectly-exposed specimens, a useful diagnostic character, since from it the nature of the area can be inferred (see fig. 3, p. 178). (3) The lateral slopes.- The lateral slopes may be either flattened or tumid, and from the beak to the anterior margin may be either concave or convex according to the curvature of the area, being convex when the area is concave, and concave when the area is reclined. Excessive obesity is a gerontie condition. (4) The costm.- The costm vary in size, in abundance, and in their relation to the furrows which separate them. They are sometimes nearly or quite obsolete, particularly upon the lateral slopes of the ventral valve. Obsolete costm are usually accompanied by a reclined area. (5) The transverse plate and syrinx.--These are useful diagnostic features. The transverse plate varies in its greatest depth beneath the plane of the cardinal area, and the syrinx varies in size relatively to the transverse plate ; but, as the variations will be dealt with when certain typical forms are described, further details need not be given here (see fig. 1, p. 166). (6) The minute surface-ornament and microscopic structure of the shell.--In the majority of specimens the intimate structure .f the shell is obliterated by recrystallization, or the outer layers are lost by exfoliation, so that these features are seldom available for diagnostic purposes.

J. F. van Bemmeten, ' Untersuchugen iiber den Anatomischen & Histotog- isehen Bau der Braehiopoda Testieardinia,' Jenaische Zeitschr. vol. xvi (1883) p. 465. See also 'Cambridge Natural History: Mollusca & Brachiopoda' 1895, p. 468. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

17S DR. F. J. ~O]ITH O:N EVOI. lxxvi,

It will bc seen that, apart from the last-mentioned characters (6), the only variable features are, with the exception of the eost~e, confined to the pediele-valve, and there do not appear to be any other characters by which specimens showing the brachial valve only can be specifically determined.

Fig. 3.--7'he pr~iles of Syringothyris aud Sl)irifer distans (all ha~' of the natural size). The broken lira's ,represent the ,floor of the sinus iu the pediele-valve.

a = Spir~fer dista,ts, normal type. b = Sp. distans, var. bicarinata. e = Syringothyris priacipalis (K). d = S. cuspidata, rout. cgrtorhyncha (Z). e = S. cuspidata, mat. exoleta (C), from normal deposits. f = S. cuspidata, rout. exoteta (C), Waulsortian ' knoll '-type. g = S. cnspidata, rout. c~t.~.pidata (D). h = S. elongata (C), W~rulsortian ' knoll '-type. j = S. etongata (D).

(e) General Account of Syria.qothyris in Great Britain. Owing to the ease and certainty with which specimens of Syringothyris may be generically determined, on account of the absence of plications upon the fold and the sinus, the evolution of the genus can be studied without the complications which one encounters when dealing with vaguely-defined and obviously-poly- phyletic 'genera' like Spir(/~r. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2J ~S'JT~LV(;OTHYIH+"i~D Smmt,'El:I+~?~. 179

In this country the genus S#ri~yothyri,~" first appears in the Lower Cleistopor~(~ Zone (K~), where it is represented by forms in which all the essential characters ,of the genus are fully developed. No two individuals are precisely alike ; but the prevailing expression is that of a small transverse shell, with a fiat cardinal area of which the width is about twice the height. The transverse plate is not deeply bm'ied, is convex upwards, and the syrinx is relatively large. This type of shell, exemplified by the form described hereafter as s pri~,c@alis, sp. nov. ~S~/ri~.qothuris t,ypa, auett.~, prevails throughout the Cleistopora Zone, and is not found above the lowest beds of the Zaphrentis Zone. Above the Cleistopora Zone occur two series of forms derived from S. priJ~eipedis, antl here regarded as belonging to two species: naniely, S. elongata, sp. nov. and S. euspidata (Martin). The essential difference between them lies in the greater apical angle, and the more depressed appearance of the former. The shells referred to S. elon#ata are less numerous, and exhibit less variatiou than those, of the other species, of which successive mutations i are recognized. The further history of S. euspidata is as follows :--In Z~ the prevailing type, S. e~spidata mut. e~rtorh,/ta, eha i]S. te,vto auett.~, is a shell distinctly larger than ,~. priueipali.~', and further differs from that form in that the area is ~lightly concave, the braehial valve less transverse, and the eostm somewhat stouter. Although, in its typical forln, especially characteristic of the Lower Zaphrentis Subzone, forms approaching ,~'. euspidat,, rout. e#rtorh#ncha occur in the under- lying Clei.~totmra Zone. Owing to the lithological characters of the r.eks which frequently constitute the Zaphrentis Zone in the South-Western l)rovince, the shells have undergone but little recrystallization, and the minute textile ornamentation is frequently well preserved. It is on this account that they have been compared with the North American species S. texta (Hall).* In the South-Western Province the shells found in the Upper ZaphreJ~/is and Ca~t/nia Zones group themselves around S. c~mpi- ,lal~ rout. rxo/ela ~S. ca,rleri auctt.], which is most abundant in ~tnd eh~racteristic of horizon ,/ and the Lower Cani~da Zone. This form differs m:trkedlv from S. elo~zgata, but is less sharply differentiated from S. cuspidata rout. c:~/rtorhyncha. In the North and North-West of England, Syringothyris has not 1)een recorded from horizons lower than the Caninia Zone; but in rocks of that ao'e it is represented by a form which in general apl)earance, and in its stage of development, has close affinities with S. e us2)idata niut. exoleta, of which it is in all probability merely a geographical variant. The Ca~inia Zone in the central part of Ireland, and in Belgium, is characterized by the occurrence of the remarkable Waulsortian ' knoll' facies, and in those deposits S~/rinqoth.?/ris is developed in 1 In this paper 'mutation' is used in the sense of Waagen, and no~ in that; of De Vries. .2 Vaughan, ' Burring"ton Paper' [28~ p. 364. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

1S0 Dlt. F..r. ~'OR'rH O~ ~vol. lxxvi, an extraordinary manner. Tile shells vary within wide limits, but all agree in attaining iu the adult stage an unprecedented size, accompanied in most cases by an ahnost complete abortion of the costte, especially of those on the pedicle-valve. The costm, when present, are large, but poorly defined. The increase in the size of the individual is not accompanied by any thickening of the shell- substance. Usually, the area is fiat, or more or less strongly reclined; while the median line along the floor of the sinus is distinctly concave, and not convex as is the case in earlier forms, or in contemporalT forms, occurring in normal deposits. Were they not connected by intermediate forms, extreme types might well be regarded as belonging to distinct species. The variability of the shells found in the lower zones has already been mentioned, and the great diversity of form in the deposits under consideration is (in all probability) due t~) the accentuation of different variable characters in different individuals. Much careful collecting will be necessary before these varied fomns can be further classified, and although distinctive names, if available, wouhl be very convenient, no useful purpose would be se.rved by the introduction of new names until the significance of the variations is more fully understood. In the S~,mi;~ltla Zone S!/ri;t.qo/hyri~" is rare. It is represented in the South-Western Province bv forms resembling rout. exol(~ta, and in theft area does not occur ~ above that horizon. It is not known al)ove C~ iu the North-Western Province; but in the Midlands the genus persisted until Dibuuoph.yllum times, and it is there represented by S. c,tspidata mat. casTidata and S. elo;~jata in D,. in the ' brachiopod-beds' (Dibuaopll.qllam Zone) of the Mid- lands, and the ' reef-knolls' of the Clitheroe and Craven districts, ~!!ringOth:qris is represented by forms which recall those found in the abnormal rock-types of the Caniaia Zone of Ireland. The holotype of Martin's S. cuspidata was obtained from beds of D, age at Castleton, and closely similar shells are found in the 'brachiopod-beds' at Park Hill (Derbyshire), and at Axton (Flint- shire). Principal T. F. Sibly has shown that the 'brachiopod-beds' of Park Hill, Thorpe Cloud, and other localities in Derbyshire are of Middle Dibunopl~?/llum age (D:, subzone of Loasdalia .flori- jcb;'mis), and Dr. H. H. Bemrose's observations on the toadstones of Derbyshire indicate that the shell-beds of Castleton are of similar age. The age of the knolls at Craeoe and Clitheroe is less certain: Vaughan in his last paper 1 expressed the opinion that they were, in part at least, considerably older than the ' brachiopod-beds,' and included representatives of the Semi;~la Zone, and perhaps of the Caninia Zone also. It is not intended to discuss in detail the relative ages of the Upper Avonian ' brachiol)od-beds' and ' reef-knolls'; but it is

1 Vau~han' ' Dinant Paper' [35~ p. 15. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

par~ 2 ~ ,~" YIr T"It YIM, S' .t,N I) ,.b'l'I Jr l,'b.'1r 181 interesting to record in this connexion that the examples of S qrin:7oth.qri., from the Cracoe knolls much more closely resemble those fi'om the Irish Waulsm'tian knoll-phase limestones, than those from the Derbyshire 'brachiopod-beds.' Nevertheless, the special characters of the specimens from the Waulsortian knolls and the 'brachiopod-beds' are dependent upon the existence of special conditions rather than upon the time at which these conditions occur, and it is not safe to base conclusions as to the age of the rocks upon forms of Sjrin.qothyris alone. Shells exactly resembling the Castleton type occur at Little Island (Cork), in beds which, in age, probably correspond closely with those at Castleton. In the list of fossils from the Carboniferous Limestone from the vicinity of Cork given in ' The Geology of the Country around Cork & Cork Harbour,' 1 the fossils from all horizons are tabulated together: but some of them suggest the presence of beds of Dibunophyllt~Jn age, and several of the fossils enumerated are characteristic of the Middle and Upper Dibu~,ophyllum Subzones in the Midlands. In common with most of the fossils from the neighbourhood, the specimens of S. clespidata are curiously distorted by pressure; but their similarity to the Castleton shells is very apparent (see P1. XII, figs. 2a & 2b). The genus has not yet been found anywhere in rocks later in age than the Dibunol)ltyll~em Zone.

Summary of Notes on the Distribution of Syringothyris. From this account of the distribution of Syrin.qothy,.is, it appears that in the South-Western Province, where there was no marked alteration in the conditions of deposition from Cleistopora to Lower Caninia times, the changes in the genus were continuous and progressive. In common with many other brachiopod genera, it disappeared fi'om that area after the comnlencement of the special conditions which prevailed immediately before and during Seminula times, and never reappeared there. In other areas, factors such as migration and the occurrence of abnormal conditions of deposition render the sequence less clear. In Ireland, the genus becomes increasingly important from the Gleistopoza Zone, where it is very rare, to the Ca~inia Zone, and is seldom met with at higher horizons. In the Midlands and North of England it is not present in the lowest beds of Carboniferous Limestone, but appears in the Caninia Zone in the North, and in the Dibunophyllum Zone in the Midlands. Here, as in Ireland, for reasons which have already been given, variations due to time alone cannot definitely be distinguished. The genus illustrates in an interesting way some of the principles of evolution. The (phylogenetically) young stage with small and

1 Mem. Geol. Surv. 1905, pp. 29-31. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

182 DR. F. J. ~owri[ oz~ [vol. lxxvi, simple forms is represented by the earlier mutations of both cltspidala and elonqaht : the adult stage, in which the full size i.~ attained and all the ch;u'acters arc mature, is seen in the rout. exolela and in the Crtninia,-Zone representatives of S. elon.qata; while phylogenetic old age with 'exaggeration and per-elaboration,' followed quieldy by extinction, t was reached (in normal limestone- deposits') in the Dibu, nol~h,!/lhtm. Zone, but was, in certain areas, hastened by the Sl~:,cial conditions accompanying the formation of the ' reef-knolls."

(f) l)cscrit)tio~, of the British Species of Syringothyris and tl,eir Mutations.

,S~'JUS~OTU~RtS t, al.XC,PX,,Is, sp. nov. (Text-tlgs. 1 a-1 c, 3c, pp. 166, 17,S, l'l. Xl. figs. 1 & 2.) 1905. Si/rinqoth.qris all: cuspi,lat,~ (Martin)--Vaughan [33] p. 215. 1911. S:qri,,gothris t.qt, a Winchell, tiall--Vaughan [28]'p. 364. 1913. N~jrin[/oth.qris aft: carteri--North [25] p. 401. 1917. ,~qringothqris of. enspidat,t--Dixey & Sibly [6] p. 131. In this species the shell is smaller than in any later representatives of the genus: the dimensions of an average specimen being .... width along the hinge-line, 35 ram. ; length of brachial valve, 16 ram. ; height of area, 15 ram. I) e s c r i p t i o n :--P e d i c I e- v a 1 v e.-A re;~ praetic~dly flat, alncal angle al)out 100% Sinus wide, shallow, and not sharply defined by the costae on either side of it: a median line along its floor is ahnost flat, except near the apex whert' it is slightly curved. Lateral slopes m~)re or less flattened. Costw fine, but distinct right Ul> to the al)eX, and frequently smnewhat sinuous; they are rounded but very depressed, and air sel)ar~tted by n~m'ow groove-like furrows : tl~ere are "dmut eighteen or twenty on each Iater,d slope. l)elthvrial supporting-plates markedly divergent; syrinx relatively large, but not deeply buried beneath the level of the area (see text-fig. [ b, l,. 166). '['he delthyrial suppm~ing-plates fail to extend emnpletely across the valve in transverse sections whine? the syrinx is most t.yl~ieally developed. Braehial valve.--There are few distinctive features in this valve. The fohl is sharply defined and evenly convex, but not much raised ahove the general level of the valve: lateral slopes flattened towards the cardinal extrelnities. The eostae resemble those in the opposite valve, but are not sinuous. The beak is inconspicuous, and but little raised above the hinge-line. Holotype.--A pedicle-valve from K., subzone, Howle Hill, near Ross (HerefoMshire), in the collection of Principal T. F. Sibly (Pl. XI, fig. 2). Horizon and localities.---This species is especially characteristic of the Glei,stopora, Zone, but extends into, and is rarely represented in the lowest heds of, the succeeding horizon; it i,

I Vaughan [_35~ p. 3. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

p;trt 2] ,~"l't~z.v(;o rii lJtls ~:~n ,VZ'Zl~Z~'ERZiV..t. 183 almost confined to the South-Western Province. It is represented l)y fragmentary specimens recorded by Vaughan from Km in the Bristol area, and reaches its aeme in that district in K~, and in the Bm'rington Section and in the east of Glamorgan, in K 2. Other localities are: Coed yr Gang, Upper Neath Valley, K~I; Miteheldean district (Ghmeestershire), K 2. Syri~,

Srm.~'t~o'rH~r~s nt, o~-aA'r.t, sp. nov. (Text-figs. 1 e, lj', 3 1+, & 3j, pp. 166, 178; P1. XII, figs. 3--5.) Description :--Pediele-valve.--Area flat, and about twice .as wide as high, apical angle about 100 ~ beak sharp : sinus wide and well defined, the median line along its floor slightly convex. Cost~e distinct, flattened and separated l)y narrow furrows; about eighteen or twenty on each lateral slope, those nearest the sinus are about 2 ram. wide at the anterior margin: lateral slopes more or less flattened and falling away sharply from the plane of the area. Delthyrial supporting-plates subparallel; transverse plates not deeply submerged ; syrinx relatively large. (See text-figs, i e & I.)+:) Braehial valve more than twice as wide as long, cardinal extremities subangular or slightly produced: cost~e as in the opposite valve, but more clearly defined : fold wide and depressed, but distinctly raised above the general level of the valve. The dimensions of an average specimen are as follows :--width along hinge-line, 66 ram. ; length ()f brachial wflve, 24 ram. ; height of area, 28 ram. Remarks.--This species reseml)Ies in general appearance S. princ@alis from the Cleistopora Zone, but differs from it in its size, and in the small angle of divergence of its delthyrial :supl)orting-plates, which are relatively longer, and extend completely across the cavity in sections where the syrinx has nearly disappeared, while in S. 2)+'incipalis they are already incomplete in sections where the syrinx is best developed. From S. cuspidata it differs in its greater relative width, larger apical angle, and its subangular cardinal extremities. Holotype.--A specimen in the collection of the Geoh)gieal Survey of Ireland (Reg. No. 1492); horizon, Cani'ni~t Zone; locality, south of Adare (Limerick). (t'l. XII, figs. 4a-4c.) H o r i z o n s a n d 1 ,) e a 1 i t i e s.--This species is represented in rocks of both Lower and Upper Avonian age ; and, while examples from different horizons present small variations, they are similar in general characters, nor can they at present be satisfactorily differentiated. It is recorded from Z to C in Ireland: for instance, 1 Collected by the late l~Ir. C. tI. Ctmning,ton. Douglas [8] p. 548. Q. J. (I. S. No. 302. o Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

184 ]JJt. 1~. a. l~'owrn oy [-vol. lxxvi,

Z, Malahide; C, Adare (Limerick) and Rathfeale (Cork); and from the Dibuuoph~jllum Zone in the Midlands and Yorkshire ('is, for example, Conist~m Cold, Skipton 1 Do, Kniveton). As in the case of S. cusTidata, this species attains its maximum size in the Waulsortian ' knoll '-limestones of Ireland.

S~'lll.~,OTnVRiS CUSl'lD.t'rx (Martin). In this species are included shells :[rout Z to l) inclusive, in whieh the bmchial valve is about twice as wide as high, al)ieal angle about 90 ~ to 100 ~ (100 ') in the earlier, and 90 '~ in the later mutations), cardinal extl~mfities rounded. Mutationsehameteristie (if various horizons are recognized, together with variations due to geograpllieal distribution and to environmental conditions.

S~ltl~t:o'rllYlr CL'SPil)ATA ]nut. c~'R'roziltl'.XCll.t. (Text-fig. 3d, l ). 17S; P1. XI, figs. 3a &3b.) 1825. Spin'ires" cuRpi,t, tus (Martiu)--.Sowerby [32] p. 90 & pl. cecclxi, figs. 1-2. 1911. Syriuguth.qeis of. texta llalI--Vaughan, ' Burrington Paper ' ~28~ p. 365. l)cseription :-- l'ediele-valve.--Area flaltem:d near the hinge-line, but concave towaMs the beak, which is iminted ; apical angle about 100 ~ : sinus wide, evenly concave fnmi side to side, and convex from tim beak to the anterior margin. Costa,. about twenty on each lateral slope, low and rounded, and separated by narrow furrows ; lateral slopes tumid, and making an obtuse angle with the cardinal area. I)elthyrial supporting-l)lates stout, divergent, but less so than in S. priJtclpalis; syrinx relatively large, not deeply 1,uried. Brachial.valve evenly convex; fold distinct and rounded; ealxlinal extremities ahuost'right-angles ; beak small, and but little raised above the level of tlm area. The dimensions (ff an average specimen are as follows :--width along hinge-line, 40 ram. ; length of bmchial valve, 22 ram. ; height of are~, 17 ram. Owing to the lithological characters of the rocks (of the Za- Threal/s Zone) in which this mutation occurs in the South- Western :Province, the surface of well-preserved specimens shows the characteristic minute ormtment, resembling twilled cloth. E xample.---A specimen from the Avon Section, Bristol (Z), in my collection, and now deposited in the National Museum of Wales (Reg. No. 19246 (~ 2). (Figured in P1. XI, figs. 3rr &ab.) Horizons and localities.--The Zcrphrentis Zone in the South-VVestern Province and in Irelund. In the lowest beds of the zone some of the shells are more or less intermediate between this mutation ()f S. cu,s'lddatrr and S. princilmlis , and towards the t~)1) of the zone the mutation gives place to ~'. cuspidata rout. exoleter. I Specimen ill the British Mu.~eum (Natural History) Reg. No. B 20247. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] S YIHNGOTttYIr AND NI~.IIfII,'Is 185

Reluarks.--This mutation of S. cusTidata is readily distinguished from S. priucipalis by its curved area. It is smaller than S. cusTidata mut. exoleta, but has all the essential characters of that mutation developed in a minor degree.

SYR[NGOTHYRIS CUSPIDAT~k lIlUt. EXO:L:ETA. (Text-figs. 1 d, 3e & 3J; pp. 166, 178 ; P1. XI, figs. 4 a-5 b.) 1858. Spirifera cuspidata Martin sp.--Davidson [41 p. 44 & pl. viii, figs. 19-20. 1917. 5'!]ri~gothyris cf. cUS:loidata (Martin)--Dixey & Sibly [.61 p. 142. In general characters, examples of this mutation from normal limestone-deposits are intermediate between the mutation cyrto- ~'h.qnc]ta already described and the typical mutation cusjoidata. In S. c~sl)idata mat. exoleta the genus Syrinyothyris reached, in the British-Belgian area, its acme in the size and number of: individuals and in geographical distribution. Description (of the normal type characteristic of the South- 9 Western Province):--Pedicle-valve.--Area concave, especially towards the beak; apical angle about 100 ~ ; sinus wide and shallow, the median line along its floor being evenly and markedly convex. Cost~e well defined anteriorly, rounded, but depressed, and scpaluted by narrow furrows ; lateral slopes rounded and tumid. Delthyrial supporting-plates slightly divergent; syrinx well developed and relatively large. Brachial valve.---A little more than twice as wide as long, convex, fold well defined, h)w, and rounded ; beak slightly inflated and raised above the level of the area, but less so than in the later fornl S. T 5" l ) i c a 1 e x a m p 1e s.--Specim en s from We ston-super-Marc and from Sutton, Glamorgan (set, P1. XI, figs. 4 a & 4 b), in the collection of Principal T. F. Sihly. Horizons and localitics.~In the South-Western Province this mutation ranges from Z~ to C, and has also been recorded as. occurring rarely in S~. It does not everywhere attain its maximum at the same horizon, but reaches its acnle in the Bristol district in Z,, in East Glamorgan at horizon 7, and in the Mcndip are~ in C~. Normal examples of this mutation occur in the C'ani~Jtia Zone in Ireland. V~riations of rout. exolet(~. It has ah'ca(lv been stated that this mutatim~ is typically developed in the South-Western Province. In the North-West of: England it is represented by forms which have all the essential characters of the mutation, but are somewhat less tumid, and in which the i~ttening of the. ribs is more apparent and the beak of the pedicle-valve sharper. This type has been recorded by Prof. E. J. Garwood in the Brownber Pebble-Beds (sub-zone of Se~ni~utla #~'e.qaria, lower part of C,), 1 and m a ham calcareous

1 Garwood [13~ p. 464. o2 Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

186 De. F. j. ~OItTtI O~ [vol. lxxvi. grit of similar age at Kendal. 1 I)avidson figured it from Kendal.~ The difference between these northern representatives of rout. exoleta and those from the South-Western Province is probably, in part at least, explained by the fact that in the former area they occur ahnost exclusively in oolitic and arenaceous deposits, while in the latter they are usually found in the crinoidal and shelly limestones. The peculiar variations of mut. exoleta found in the Waulsortian knolls of Ireland have already been referred to. S~RINt~OTnYmS CUSmDXTXrout. cus:,1DXTA. (Text-tlg. 3.q, p. 178 ; P1. XI, figs. 7-8 & P1. XII, figs. 1 a-2 b.) 1796. Anomites cusl~idatus Martin [22] p. 44 & pl. iii, figs. 1-4. 1809. Convh.yliolithus anomites euspidatus Martin F233 pl. xlvi, figs. 34 & pl. xlvii, fig. 5. :1816. Sloirifer cuspidatus J. de C. Sowerbv [32] pl. exx, fig. 3. 1836. Spir'ifera euspidata Phillips f27] p. 216 & pl. ix, fig. 1. 1858. Spirifera cuspidata 1)avidso~i [4] pl. viii, fig~. 21-24 & pl. ix, fig. 2. 1868. Spirifera euspidata King [17] pls. ii-iii. This is the typical form of S. cuspidata represented by Martin's lmlotvpe. Its distinguishing features arc as follows :--Apical angle about 90 ~ area rather more than half as high as wide, nearly flat, but slightly reclined, and rounding over into the somewhat inflated lateral slopes; cardinal extremities rounded. Braehial valve strongly convex, with a swollen beak well raised above the level of the hinge-line. Costte about sixteen on each lateral slope, strong, depressed, separated by furrows wider and less numerous than those in the earlier forms of S. cuspidata and in contempoml\~" examples of S. elon!lala. I)elthyrial supporting-plates nem'lv parallel; transverse plate thin; syrinx perfectly developed, but relatively small. The dimensions of an average specimen are as follows :--width Mong hinge-line, 56 ram. ; length of brachial valve, 30 ram. ; height of area, 32 mm. Very rarely the mutation is represented by giant specimens measuring nearly 100 mrn. in width. Horizon and localities.--This mutation occurs only in the Dibunophyllum Zone, especially I).,, and is represented in tlae Mid- lands, in Yorkshire, and at "Axton (]ihntstnre);"" " and by the distorted specimens from Cork, already mentioned on p. 181.

(,q) Pre-Carboniferou~ l listory of S~jringotl~j/ris. In this country and in Belgium, ~qrin:fothyris first appears in the Cleistopora Zone of the Carboniferous Limestone, with all its essential characters fullv developed ; but in neither of those areas do the Devonian rocks yield any Spiriferid shells that can be regarded as possible ancestors of S#ringothyris. An apical callosity and even a transverse plate are frequently

1 A number of specimens are in the Sedgwick Museum, Table~ 639. 2 Davidson's figures [4] pl. lii, figs. 3 & 3 a, appear to be idealized, and to combine the characters of S. cuspidata and Ty~othyris s~bconica. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

pal~c 2~ ~Ivt/.v(;~7. ~J,'/,,,' ANI~ ,~v'H~//.~nr 187 developed in certain species of Spirij~r from both Devonian and Carboniferous rocks, for instance : Sp. disjunctus J. de C. Sowerby, ,ST. duplicicost(r Phillips, and Sp. striatus (Martin) ; and if, as unfortunately is seldom the case, a specimen is suitably preserved and its internal structures are not obliterated by recrystallization, the transverse plate can be seen when the beak of the ventral valve is rubbed away. Text-figs. li & tj (p. 166) represent the structure as observed in a specimen of Sp. d~tplicicosta from Park Hill (Derbyshire), the sections being 3 and 6 mm. from the beak respectively. In text-fig, li (p. 166) the structure is seen to ['e essentially the same as in the apical region of Syrinyothyris. In text-fig, l j, although the relation between the transverse plate and the delthyrial supporting-plates is the same as in Syrinyothyr~,~" (see text-fig. [ d), there is no trace of a syrinx. The extremity of the plate is concave, as in Syrinyothyris ; and therefore, in sections farther from the beak than text-fig, l j, it is incomplete, being represented by two small portions on the inner sides of the delthyrial supporting-plates. A similar transverse plate was observed by me in ~ large specimen of S T. striatus (Martin) from Kildare ; King describes a section showing the transverse plate in this species, 1 in which he supposed that there was actually a tubular canal, but this has not been verified; in any case, the structure figured hv King bcars no resemblance to a true syrinx. It has already been shown that the species in which a syrinx is known to exist have a smooth fold and sinus, and the examination of the forms jUSt considered proves that it is not present in species in which the fold and sinus arc plic'ttcd (this matter is more fully discussed in w of this pal)er , i ). 189). The whole of the evidence, therefore, points to the conclusion that it is among species of the former type that the ancestors of S~dri~Lrtothffris must be sought. In the ])evonian rocks of North America there seems to be clear evidence of the evolution of the syrinx. American palmontologists have described a nunlbcr of species of Spi,r~'~.r in which the fold and sinus are non-plicate, and in which a well-developed apical callosity and transverse plate are seen : as, for example, Sp. yranulosus Conrad and Sp. (Isper Hall, both from the Hamilton group (Middle Devonian). In many specimens the plate is said to be more or less thickened along the median line. In a variety of Sp. marcyi Hall from the Upper Devonian of New York, "described by H. S. Williams, ~ there are two parallel ridges on the under surface of the trzmsverse plate, and that author regards the structure as an incipient syrinx. If text-fig. 1 g (p. 166), which is a cross-section of the beak of St). marc!/i, bc'compared with fig. 1/,, which illustrates the fully- developed syrinx of S?/rin.qothyris, the relation between the twc~ forms is at once apparent.

' King ?17~ pl. iii. fig. 25. -" Willi~uns ~37- pp. 61 4;2 & fig. 16. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

188 DR. F..,. ~owrn ,)~ [vol. Ixxvi,

In the absence of specimens and accurate stmtigraphical evidence a detaileddiscussion of this interesting subject is not possible, and the following summary of the evolution of Syriugothyris is to a great extent based upon the descriptions given by James Hall & ,I. M. C 'larke, 1 G. H. Girty,~ and H. S. Williams.Z In many Lower Devonian Spirifers there was, as the organism reached maturity, a tendency to deposit shelly matter (an apical callosity) in the beak of the pedicle-valve. From the free, that is, dorsally-directed, end of that callosity, a plate (the transverse plate) extended forward between the delthyrial supporting-plates, anterior to, or below, the del~idium. Originally an accompaniment of lnature growth, these characters seem eventually to have become permanent, and in late ])evonian and Carl)oniferous times were essential features in many species. ]in some forms, in which the radial ornament extended over the whole shell, the plate remained unchanged, while in others, characterized by a smooth fold and sinus, it lx, camc curiously lnodified. A pair of longitudimd ridges were developed on the nether surface ,)f the plate, and these, in the course of time, 'gave rise to two lamelhe, the free ends of which curled one towmxls the other to form an incomplete tube or syrinx. The formation of the syrinx was preceded ~,nd accompanied by a flattening and great elevation of the cardinal area of the valve. A depal~ure from the usual conditions is seen in the species deseril)ed 1)y Weller as Psc~tdosyrinx .~ampsoJ~i.

l)SEU 1)O S YRI N X Weller. ' The Mi.~sissippian Brachiopoda of the Mis.~issippi Valley Basin' Illinois (~eol. Surv. Monogr. 1 (191~) pp. 4Dt-l.t)5. The forms referred to this genus resemble ,~'.qrin!]othyrL~ in all respects save one: namely, the nature of the transverse plate. In shape and size, in the absence of plications on the fold and sinus, in the differentiation of the area into a central and two lateral portions (of which only the former is marked by vertical striations), and in the punctation of the shell-substance, Psel, dosyri~t.r agrees with Syrin.qot]~?/ris; but the transverse plate is quite simple, :rod bears no trace whatever of the existence of a syrinx. The possibility that the absence of the syrinx might be due to the mode of ])reservation of the shell was carefully considered by the author of the genus, who found that the feature was of constant occurrence in specimens preserved in such a way as to show with certainty that no syrinx had ever existed. Pseudos,grinx is found in the Lower Mississippian of North America: that is, it was contemporaneous with S?/rin.qoth?/ris; and the coexistence of forms, externally indistinguishable, the 1 Hall & Clarke [14]. s Girty [12] p. 51. 3 Williams [37] pp. 61-62. 4 These two types practically coincide with two of the groups into which the Spirifers were divided by Hall & Clarke : namely, the ' aperturati' and the ' ostiolati ' respectively. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] ,,,'~-mx(;orH:rms ~t.XD S/WCZl.'L'mX.~. 189 majority of which possessed a syrinx while a few did not, is somewhat remarkable. So far as I am aware, no s]Mls referable to the genus have been found in the Carboniferous Limestone of Great Britain and h'eland. Type-species.--Pseudosyrinx saml)soni (Weller), Illinois Geol. Surv. Monogr. 1 (1914) pp. 405-406 & pl. lxvii, figs. 11-13; ~'qrinqothjp'is sampsoni Weller, Bull. Geol. Soc. Am. vol. xx (1909) p. 8Ii & pl. xiv, fig. 4. The prevailing septal arrangement in the pedicle-valve of forms with a plicated fold and sinus is that described as existing in Np. d~ffdicieosta (see fig. 1 i, p. 166), but E. M. Kindle has described a curious modification of the septa in certain Devonian fossils, for which he established the genus Sy;'inyosdoira.

Sriti.xoos PtnX Kindle. 'The Devonian Famm of the Oumy Limestone' Bull. U.S. Geol. Surv. 391 (1909) pp. 28-29. The genus was established for shells in which the pedicle-valve is pyramidal and the area flat, as in S. c~tspidata. The author in describing the genus says: 'though solnewhat closely related to S.r it represents a type generically distinct from it' ; actually, however, it resembles S!/;'i~zyoth!lris in nothing but its pyramidal shape. Type-species.--S. prima. Kindle, op. cit. pp. 29-30 & pl. vii, figs. S-Sd; pl. viii, figs. 1-1 a. This is the only species at present referred to the genus, and only two specimens were known to the author when the name was proposed. Horizon and locality,---The (Devonian) Ouray Limestone, Hillsboro' (New Mexico). S. fl;'ima differs from Sy;'inyotl~.qris in the following respects :-- (1) The whole surface of the shell, including the fold and sinus, is plicated, and the ribs are much more numerous than in any species of Syri,gothypis. (2) The septM system in the pedicle-vMve consists of a small transversq plate joining the delthyriM supporting-plates a little below the level of tho area. The delthyriM supporting-plates do not extend to the floor of the valve, but curve towards one another and unite in the median line (seo fig. le; p. 166). (3) The shell-substance is impunetate, and the whole area is marked by vertical striations.

III. NOTES UPON CEIU['XIN SPECIES ]NCORRECTI, V ]r TO SI,~I~[NGO'I'IlI'RIS, ~kND UPON SPECIES OF ~IrRINGOTHI:RIS SUPPOSEI) TO POSSESS PLIC&TIONS UPON :ritE FOLD :LND 8TNUs.--THE GEXlIS SYI{I.VGO1)LE'UICA. Spiri~,r distans 8owerby, a species in which eostte occur on the fold and sinus as well as on the l~teral slopes of the shell, has been referred to ,~q,~'i~gotILqJ'is; Sj)im:f~u" alt~ts Hall (Middle Devonian), with similar external charaete/'s, has been regarded as Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

190 ~Jt. F J..N'ORTH ON ['VO]. lxxvi, an ancestor of Syringothyris; and '~S),irifbr' r,,ndalli Simpson was stated by Simpson to have a s) rinx in a shell the fold and sinus of which were plicated ; but in none of these cases has the coexistence of the external characters mentioned, with a transverse plate and syrinx, been proved.

SYRINGOTllYIII8 RANDALLI (Simpson). 189o. Spir~fel. randalli Simpson, Trans. Amer. Phil. So(.. vol. xvl, I'. 4.11. 1910. S yrinqoplevra rctndalli Schuchert, Amer. Journ. Sci. set'. 4, vol. xxx, pp. 223-24. Horizon and locality of the type-specimens.--A little below the Sub-()lean Conglomerate, that is, fairly hio'h in the American equivalent of the Carboniferous Limestone, at Warren, Philadelphia. In 1910 Schuchert estal)lished a genus 5)/ri~!/ol, h'~,r~r to im.ludc syrinx-1)earing Spiriferid shells, in which there weir costa: on the fold and sinus, and his gcnotype was Spirif~,r rcqndalli Simpson. Girty, 1 however, rejects that genus tm the following grounds :-- S. r(~uJalli was not correctly described t)y Siml)son, who, owing to the imperfection of his material, apparentl.v confused two forms which ()ccur together in the same formation, and he attributed (o the Syringothyroid species rondalli the external characters of a true ~ir(fe:r. Simpson described his species as being externally like ~)~. disjuncius, wi~h from forty to sixty radial costa., which at once suggests that the specimen Ul)On which tlmt statcment was founded could not have been a S?/ri~.qot]~,qris, in which tim number of costa: is between twenty and thirty. Girtv maintains that of Simpson's original specimens those "which possess a transverse plate and s.yrinx have a smooth fold and sinus. If this is the case, the species is correctly regarded as a ~:qrilt!lOl/l:qris. and has :~ smooth fold and sinus.

Sl'IllIFEIt ALTUS Hall. 1836. Spirifer alta Hall Prec. Am. Phil. See. vol. x, pp. 24651, .~t'~' also' P~d. New York ' vol. i'~ (1867) p. 248 & pl. xliii, figs. 1-7. 1894. C2/rtia alta Hall & Clarke, ' Pal. New York' vol. viii, pt. 2, p. 42. This is a Middle Devonian form possessing a high area and a transverse plate, and was regarded by Hall as a direct ancestor of h~qringot]~!/ris. There is, however, no trace of a syrinx, but only a median thickening of the transverse plate, and the fold anti sinus are plieated. Hall compared his species with SpiriXer si~tp/ex Phillips, -~ but in that form the shell is smooth.

1 Girty [11] pp. 548 54. 2 j. A. Phillips, ' Figures & Descriptions of the Palaeozoic Fossils of Devon, Cornwall, & West Somerset' 1841, p. 71 & pl. xxix, fig. 124. I have examined several specimens of Sp. simplex from the Middle Devonian of Devon, but have not obtained any evidence of the existence of a transverse plate. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2~ ,S'~WtXC, OTH I'IiL'~~ AND .SI'IIHI,'ERI A'A. 191

The existence of a transverse plate in a Devonian Spirzfer does not of necessity indicate an ancestor of Syrinjoth?/vis ; for, as is shown elsewhere in this paper, such a structure occurs in a number of species of S])irz~/br, and, furthermore, the plieated fold and sinus clearly indicate that the species under consideration has no connexion with Winehell's.

SPIRIFER DIS'I'ANS J. de C. 8owerby. 1825. Spirifer distans J. de C. Sowerby [32 i p. 152 & pl. ecccxciv, fig. 3 (two figures). 18gg. Qqrtia distaus Sowerby, M'Coy E20~ p. 136. 1858. Spirifer distans Sowerby, Davidson [g~ pl. viii, figs. 1-16. 1880. Syrinqoth.yris distaus Sowerby sp., I)avidson [5~ p. 281 & p]. xxxlii, figs. 4-5. The fossils which have been described uuder this name seem t(, include more than one species, and they have been regarded as SqriJ~.qoth?/ris, although none of them- really belong to thab genus. They all agree in having more or less clearIy-defined plications on the fold and sinus, a character which in itself is sufficient to suggest that they are not species of Syrin.qolhyris. The locality of the holotytJe of the species is recorded by Sowerby as 'Dublin,' and although it has also been noticed in other parts of Ireland, there do not apl)ear to be any forms in the Carl)oniferous Limestone of F, ngland and Wales that can be definitely referred to this species. ,S'pi~'Oeet 9 dixtams" exhibits considerable variation in the length of the hinge-line. Some specimens, as, for instance, Sowerby's holot.vl)e, closely resemble a n(wmal Sl)irifer, but other examl)les have a relatively longer hinge-line and a less ineurved area. Sp. di.~laJm differs from S qrin[/olh,q~'is in ninny respects. In general al)pearanee the shell 1)cars a muclt closer resemblance to Spi'~'ifer. The lateral slopes of each valve are more tmnid, and the area is less elevated and more incurred than in S,qrin!lotlqp'is. (This more especially applies to forms like Sowerby's holotype, because in the height and curvature of the area some examples reseml)le the form of Sgri~ff/olhyris found in the Zaj)bve'Jt/i,~ Zone of the South-Western Province.) The apical angle of Sp. distans is greater, from 120 ~ to 125 ~ as against 95 ~ to 105 ~ in S. cu.sy)idala ; the delthyrial fissure is proportionately narrower, its sides diverging at an angle of about 16 ~ while in S?tri~qothyris the corresponding angle is about 25 ~ The delthyrial supporting- plates are situated ahnost at right angles to the plane of the cardinal area, and in consequence of the narrow fissure appear in sections across the beak of the ventral valve as a pair of nearly parallel plates, much closer togethea" than is the ease with Sgrinqo- thyris (see tig. t k, p. 166). There is in s distans no tf~efotd division of the area, which is marked by vertical striations throughout its entire length. The fold is very little elevated, and makes a nmeh less pronouncd wave in the anterior margin than in S?l~'i~soth.yris. The costa., Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

192 1)R. F..~..XOIt'rll OX LYe1. IXXVl. especially those near tile fold, show a tendency to bifurcation, ~ condition not found in ally shell ill which Syringothyroid characters have been proved to exist. The fold is divided by a shallow median depression into two broad, low, rounded costal, and the depression itself may or may not be slightly undulated. The sinus is marked by about eight cost~e, quite distinct, but not so well developed as those on the latcl-al slopes of the shell. I have not succeeded in obtaining any evidence of the existence of l)unctations in the shell-structure; but, for reasons which have already been given, that does not necessarily prove their absence, since all the sl)ecimens were either recrysfallized, siliciticd. or partly cxfoliated. The elongate type ()f [q/)/r~f~'r di.~/(fns is str(mgly suggestive of the shell descril)ed l)v M'Coy as ,5)~. bicari~iala, 1 1)at his single tlgure inadequately represents it. Suhsequently, however, ])avidson re-figured the origin)d specimen in another position, and his illustration,: t~)gether with M'Coy's descril)tion, leaves little doubt as to the identity of the sl)ecimen. As ah'eadv mentioned, the normal and the elongate types are connected by ~ntermediate forms, :rod there ale no grounds for regarding them as distinct species ; but the mmm. biearin~tta may convcnir be used as a varietal name to designate sl)ecimcns of the latter type. The significance ()f the two names would then 1)e as folh)ws :.... (1) Np. d i.~.t,tl,s Sowerby (text;-fig. 3 ~,. p. 178). Shell ovoid, about or rather less than twice as long as wide: cardinal extremities rounded. Example: Sowerby, 'Mineral Conchology' pl. ccccxciv, fig. 3; and Davidson [4] pl. viii, figs. 1-4. (2) Sp. dist~[,t.~, var. bicct'ri***tt~rM'Coy (text-fig. 3 b, p. 178). Shell distinctly elongated; greatest diameter at the hinge*line; more than twice as long as wide; cardinal extremities angular and pro- duce~l. In the specimen figured by M'Ooy the fold in the bra('hial valve is slightly concave, but this is only one aspect of the irregularly-plicated fold of the species, and has n,) taxonomic value. Example: M'Coy F20] p. 129 & pl. xxii, fig. 10; Davidson [4] pl. viii, figs. 11 14, 16 & 18, pl. lii, fig. 5. There does not appear to be any record of the occurrence oi: a transverse plate and syrinx in a correctly-determined specimen of Si~. ,listaas, alld in the specimens which have been rubbed down in the course of this investigation n,) trace of any such structure has been observed, although tim delthyrial SUl)porting-plates were perfectly preserved. This seems sufficient reas6n for asserting that no syrinx-bearing plate occurs, and, therefore, that the species does not belong to S!lri~.qoth?/ris, but must be referred to Spir~fer. ])e Koninck's diseovel\v of the syrinx has already been mentioned, and it has been shown that the structure which he described was exactly like that of Syrin.qothyris; it now remains

1 M'Coy [20] p. 129 & pl. xxii, fig. 10; see also Davidson [4] pl. viii, fig. 18, where M'Coy's figure is reproduced. 2 Davidson, op. cit. p. 224 & pl. lii, fig'. 5. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2~ SITII.YGOTHYRI$ ~'D SP~I~U.'L'RIX~. 193

to see whether lfis specimen was correctly named. His figures 1 certainly do not indicate anything like Sowerby's holotype of SjoiriJ'eJ" distains. " One of them (fig. 5) shows a threefold division of the area, and a delthvrium relatively wider than that of Sp. distans, while fig. 6 is obviously a Sgrin.qothy~'is. Neither of the figures is sufficiently clear to admit of specific determination ; but they most probably represent one of the Lower Avonian forms of S.qrinyoth~fris. There can, therefore, be no doubt that the possession of a non- plicate fold and sinus is an essential feature of S]tri~,yothy~'is. So far as I am aware, the three species enumerated in this section are the only species that are supposed not to conform with this rule, but in no case does the supposition hokl good. It was necessary to establish this fact clearly, because it has an important bearing upon the phylogeny of the genus, in that it limits the stock to a group of Spirifers with a smooth fold and sinus: that is, a ,group corresponding approximately to the 'ostiolati' of Hall & Clarke; for, although a high area in the pedicle-valve may be developed in several members of the Spiriferidm (as, for example, Cyrtia, Spir~fer, etc.), it is only in forms with a smooth fold and sinus that the feature is accompanied by the formation ()f a syrinx-bearing transverse plate. In his description of Sp. dLr Davidson says that the fold on the brachial valve may extend to the front margin, as in the typical form, or may soon become converted into a central rib.~ There is, in the Carboniferous Limestone of Derbyshire and Lancashire, a well-defined form which possesses the last-mentioned character, and it has usually been referred to Sp. dist(f,s. [n general appeal~nce : that is, in the ]might and curvature ()f the area, magnitude of the apical angle, width of the delthyrium, and 9tumklity of the lateral slopes, it closely resembles the elongate type of that species ; the l)eculiarity of the fohl and sinus, h~}wcver, is so striking that the form is best regarded as a dlstinet species, and it is described below as Spittle,;" Tlieatosulcat~ts, sp. nov. The presence of a median costa on the fold and sinus recalls .Sp. tria~yidaris (Martin); but the two species are totally different an all other respects.

SPIRIFEtt PI,ICATOSULCATUS, sp. nov. (P1. XIII, figs. 17 (t & 17 b.) D e sc riptio n.--Shell transversely trigonal, greatest width Mong the hinge-line, cardinal extremities acutely angular. The dimensions of the bolotype are :--width along hinge-line, t;4 ram. ; length of brachial valve, 28 ram. ; height of area, 21 lnm. Pedicle-valve sub-pyramidal, cardinal angles sharp, and lateral slopes evenly convex. Cardinal area nearly flat in the neighbourhood of the hinge-line, but becoming increasingly convex tow'u'ds

1 De Koninck [18] pl. ii, figs. 5 & 6. ~ Davidson [4] p. 46. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

19-1, Dlc ,.'. J. NOW,',, O.X" [ vo1. lxxvi, the beak, which is small, pointed, slightly ineurved, and situated well above a line perpendicular to the plane of the braehial valve. Width of the area at the hinge-line rather more than three times its height, apical angle about 130~ delthyrium narrow, its width at the base being one-third of its height, delthyrial angle about 18'. Sinus shallow, with a median angular costa in the anterior half, corresponding to the costa on the fold in the optx~site valve. The sinus is produced anteriorly into a quadrate tongue-like pro- jection; it line along the bottom of the slims from tim beak to the anterior margin is evenly curved, and is practically a quarter of it circle. Brachial valve convex, the greatest convexity being between the middle line and the anterior margin : mesial fohl distinct, but not much raised above the general level of the v;th'e, except towards the anterior margin. Tim mesial fold is divided by a shallow median depression, increasing in width froln the 1)eak to lhe margin of the valve, and on the floor of which in its anterior halt: there is a l)rOnounced sul)angular costa. Each lateral slope in 1)oth vah'es bears nineteen or twenty low rounded costa,, usually sin,ple, hut occasionally bifurcating and decreasing gradually i)l size as they al)proach the cardinal ex- fremitics. Shell-structure unknown; in none of the spechnens exaluined was the shell-substance sutticientlv wall-preserved t~ admit of microscopic examination. Internally the delthvrial sutTorting-l)lates are perl)endieular to the plane o'f the cardilml area, and in cross-sections near to the beak they appear as a l)air of close parallel plates not c(mnccted by a transverse plate. There is no median Sel)tum in the pedicle- valve. Horizon and localities.--This species occurs in the Dibu~o- plq/ll.'~. Zone, D..-D.~, 'Braehiopod-Beds.' The locality of the holotypc is Beeston, near Waterhouscs (Staffordshire) ; it is also rel)resented by smaller specimens at Park Hill and Castleton (I)erbyshi re). The holotype is in the Museum of Practical Geoh)gy, ,lermyn Street, London, and is nulnbered 284.22.

IV. ' SP[RJFER]~' ~,XM[~OSA (M'Coy) : rl~ I~'LO T I[V~S, gen. nov. The species $1~. lami~wsa M'Coy has given rise to considerable discussion. It has at various times been referred teat least four genera, ill addition to the one to which it is assigned in this paper, and in it have heen included forms which, although slightly resembling it in external characters, really belong to other g'.enera. The uncertainty in the use of the name is to be attributed to two circumstances: (1) M'Coy's original description is inadequatei, and his figures are too crude to portray clearly the characters of the species, while the h(dotype, which was originally ill the Gritfith Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] ,vY~r I7~I5" AND ,~I'IRII,'L'RINi!. 195

Collection, is no longer in existence, or at least cannot be recognized 1; (2) The various ways in which the fossiL4 have been preserved make their identification a matter of considerable difficulty. Many specimens are silicified, and show only the external appearance of the shell ; others are more or less exfoliated, and bear no trace of the imbricate structure which is an essential feature of the species, while many are represented only by internal casts or external moulds. In the course of this investigation I have examined material from various localities in the South-Western Province, and from the Irish localities cited by M'Coy. Two Irish specimens preserved in the Sedgwick Musemn, Cambridge, which were presented to that Institution by Sir Richard Griffith and were named by M'Coy himself, afforded most useful information: for one is a testaceous specimen of a brachial valve of ' Spir~feri~a' lamiuosa; from Hook Head, anal the other an exfoliated specimen of a pedicle-valve from the ' Calp ' of Bundoran,~ resembling those so frequently found in the South-Western Province. The discovery of specimens showing a cast of the interior and a mould of the exterior of the same shell, and a comparison between wax impressions of such casts and the internal structures as revealed by sections across the apex of the pedicle-valve, have established beyond doubt that the internal casts found in certain bands of dololnitc in the middle part of the Avonian also belong to the same species. The species under consideration differs from Spiri~br in the presence of a well-developed median septum in the pedicle-valve ; from Sl)iri~brina in the absence of the coarsely-punctate shell- structure which characterizes that genus; and from S~/rin.qothyris in the absence of a transverse plate and syrinx, in the possession of a median septum, and in the imbricate surface-ornamentation. In the presence of a median septum accompanied by an impunetate shell it resembles the Silurian genus Deltkvris (Dahnan), but differs from any other species referred to that genus in its much greater size and its smaller and more numerous costm, and also in the marked development of an apical callosity. There is, in the circumstances, no alternative but to regard it as belonging to a hitherto undescribed genus. It agrees in all its essentials with 'Spirifer' subconicus (Martin), and I propose to assign those two species to a new genus, 7~lothyris.

TYLOTHYRIS, gen. nov. Description.--Shell spiriferoid, about twice as wide as long, greatest width along the hinge-line. Cardinal area moderately high, slightly concave or flat. Apical angle about 115~ ~. i Dr. R. F. Scharff was good enougt~ to institute a special search for this specimen in the collection in the National Museum, Dublin, and it may be definitely stated that it is no longer in that Institution. 2 The register-numbers of these specimens are 635 & 636 respectively. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

191; :Din r'. ,r. NORTtt O~ [vol. lxxvi,

Dcltllyrium nearly twice as high as wide, delthyrial angle about 35% l)elthyrial covering unknown, although its former existence is indicated by grooves on each side of the delthyrium. Pedlcle-valve sub-pyramidal, lateral slopes convex or nearly flat; mesial sinus wide, subangular, and sharply differentiated by the subangular eosta~ on each side of it. Brae h ia 1 va Ive gently convex, mesinl fold distinct, and mise~l ~Lbove the general level of the valve, especially towards the anterior margin. The lat~q'al slopes in both valves are each ornamented by sub- angul~tr costze, decreasing gradually in size towards the cardinal extl'enfities. Surface of both valves crossed by well-defined, regularly-disposed imbricating lines of growth, which are on the average :Lbout half a millimetrc apart. Shell-structure fibrous and im- l?unetate. Internal structures:-- Pediele- val*ve.~ Delthyrial supporting-plates divergent towards the floor of the valve. Median septum well developed, reaching nearly to the level of the area at the al)ex, becoming gradually lower and disappearing about ball wav between the beak and the anterior margin of the valve. The carOller-formed portion of the median septum is embedded in the apical callosity, which near the beak completoly fills the cavity between the delthyrial supptwting-lflates, and extends forward along the floor of the valve, connecting those plates with the median septum, on each side of which, at the termination of the callosity ;tb,mt half-way along the median septum, there is a pit.like depression which gives a very characteristic al)peal~ance to the internal cast. (Sec text-tigs, l m & l n, l'. 166. and Pl. XIII, fig. 12.) Brachial valve.--There is a low median crest or septum. Brachial skclctoll and jugtmt imperfectly known. Range.--Carbonifcrous Limestone, Cleistopm'a Zone to Upper DibnJwlJhyllum Zolm. I ~c n o ty pc.-- 71r laminosn (M'Coy). S y s t e m ~t t i e p o s i t i o n.---Ti le systematic position of TqlothypA" camaot be st~tte(l with absolute eert~dnty. The median septum is only a special development of a feature represented in most of the S1)ir!fcrs, and cannot be regarded as of more than generic value. The absence of punctm in the shell-substance connects it with NpiJ'~fe~" rather than with SptJ'(/'en'i~a. The ~l)iral coils, as seen fr()m the rubbing-down of the braehiaI valves of several specimens, are arranged with their apices directed laterally outwards, that is, in the way normal to Spir~fer and allied genera. There is no evidence to suggest that the coils were spinose, as iu the Jurassic species of Spiriferina. Owing to the mode of 1)reservation of the shells, indubitable evidence of the precise nature of the jugum has not been obtained. Small fragments of shelly rod were observed in some specimens in the position in which the jugmn would be expected to occur. Theywere, however, obviously disturbed, and it was not possible to determine whether the)- Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2~ +'~Ttt=vc;orH IRI,S' &ND Sl~]ltll,'L'lfl~XL.l. 197

represented a eontirmous jugum such as is present in Spiri- ferina, or mere jugal processes like those of, for instance, S1)iriJbr sl#.tata. 1 On the classification given in the second English edition of Zittel's ' Pal~eontology,' l~loth.qris would appear to belong to the sub-f~mily Trigonotretinm, Sehuehert ; it certainly does not belong to the family Suessiidm, which includes S2)irt~brina.

T~-LO'rl~lUS l~a~[I~OSX (M'Coy) emend. (Text-figs. lm, lJ~, 4a, & & b, pp. 166, 199 ; P1. XII1, figs. 1, 2, 12, 13, & 16.) 18~1:. Cyrtil~ laminosa M'Coy r20j p. 137 & pl. xxi, fig. 4. 1844. Spirifera decemcostata i~l'(2oy,op. cit. p. 131 & pl. xxii, fig. 9. 1841. Spirifera mesogonla M'Coy. op. cir. p. 137 & pl. xxii, fig. I3. 1855. Spirifera laminosa M'Coy [-215 p. 426. lb~7. ~piriferina laminosa I)e Koninck, Ann. Mus. Roy. Nat. Hist. Belg. vol. xiv, pt. 6, pp. 103-105. 1905. Sqringoth.gris aft. laminosa Vaughan, ' Bristol Paper' r33~ p. 300. 1911. S~jriugothgris laminosa (auctt.) Vaughan,' Burrington Paper' ~28j p. 383 & pl. xxxi, fig. 8. D e se l"ip t i o n.--Shell spMferoid, of medium size, about twice as wide as long ; greatest width along the hinge-line. l'e d i e 1e- va 1 v e sub-1)yramidal , cardinal area rather more than three times as wide as high, concave, nearly tlat in the neighbourhood of the hinge-line, but with increasing eurwtture towards the beak, which is very slightly incurred ; apical angle about 120 ~ The beak of the pedicle-valve lies considerably above a line perpendiculal~ to the plane of the braehial valve (see text-figs. 4a & -tb, p. 199). Delthvrium nearly twice as wide as high, delth3"rial angle about 35~'. Delth~'rial eoveril~g unknown. Lateral slopes evenly convex; mesial sinus wide, subangular, and terminating in a rounded wave at the anterior margin of the valve. A line along the tloor of the sinus front the beak to the anterior margin is evenly curved. 13 r a c. h ia 1 v a 1 v e convex, the greatest convexity being in the neighl)ourllood of the beak. Mesial fold narrow at its commence- ment, but increasing rather rapidly in width towards the anterior margin, where it is considerably raised above the general level of the valve. Each lateral slope in both valves is ornamented by ten or twelve simple eosta~ decreasing gradually in size towards the cardinal extremities ; the last four or five costm are short and inconspicuous. They are subangular, and are separated by deep furrows, each abm'lt as wide as the eostw on each side of it. Surface of both valves, except the cardinal area of the pediele- valve, crossed by well-defined, regular, imbricating lamellm of growth, which are, on the average, about half a millimetre apart. Shell-structure foliaeeous, fibrous, and impunetate. Internal struetures:--Pedicle-valve.~The delthyrial

1 The jugum is the transverse shelly band and its accessory processes which unite the primary lamell~e of the spiralia. If the band is incomplete, its parts are known as ' juga] processes.' Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

198 DR. r. J. ~OR'rH ON [vol. lxxvi, supporting-plates extend completely across the cavity of tile wdve, and are divergent, e)mlosing between them an angle of about 35 ~ ; median septmn well developed, reaching nearly to the level of the area at the apex, becoming gradually lower, and disappearing about half-way between the beak and the anterior margin. The septum is, at its base, about as thick as the shell itself ; but it tapers rather suddenly, and becomes very thin. Its apical portion is completely embedded in the apical callosity, the structure of which is typical of the genus, and has already been described. Remarks:--Shell-structure.--The majority of the Irish specimens are silicified, and afford no evidence as to the microscopic structure of the shell; but in suitably-preserved examl)les ] have found no conclusive evidence of punctm, neither when the shell is examined with a hand-lens, nor when fragments are examined under the microscope. The absence of punctte cannot be due t~, reerystallization, because the fibrous structure is clearly preserved. The species under consideration cannot be readily distinguished from S!qriJ~.qoth#ris on the minute shell-structure, but it cmmot l)Ossibly be confused with Spirife'mna. A lamelh)se Spiriferid front the Bernician of Redesdale (Northumberland), dealt with in a later section ()f this 1)aper, has a eoarsely-punctate shell, and William King, supposing it to 1)e identical with the species undee consideration, assumed tim existence of 1)uneta.' in that species, which he, in consequence, referred to Spi~'i/;,~'itm. Internal casts.--Owing to the development of the apical callosity, an internal east of the pediele-valve of 2!'. laminosa possesses distinctive characters. As will be seen from an exalnination of text-fig. 4 b (p. 199) its profile differs considerably from that of the external surface. The precise nature of the cast will be best understood by reference to PI. XIII, llg. 12, where the central conical portion (a) represents the cavity between the delthyrial supporting-plates, and the lateral cones (b) the cavities on each side of it; towards the centre of the figure are the casts (c) of the pit-like depressions at the anterior end of the apical callosity, and between them is the narrow fissure (d) left by the unembedded portion of the median septum. The significance of these st,'uctures will be at once apparent if they are compared with PI. XIII, fig. 15, which illustrates the actual plate-system and apical callosity in a specimen of T..~ubeoJzica (Martin). They should be eontl~sted with P1. XIII, tlg. 14, which depicts an int~ernal east of a typical Spiri~rina : namely, Sp. walcotti. Internal casts such as those described were found by Principal T. F. Sibly in dolomites of the Lower Caniaia Zone in the Mitehel- dean district, and they also occur in almndance in deep-re~l dolomites at Tynant Quarry, Taft's Well, near Cardiff. Holotype (of the emended speeies).--A specimen (braehial valve) from Z, Hook Head (Wexford), in the Museum of ]'ractieal Geology (Reg. No. 28425). Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] b'YRINGOTtl J:I~I8 ltlq]) SPIBIFERINA. 199

Fig. 4.--The p~'oflles of Tylothvris , Spiriferina, and Punctospirifm:. (The broken lines i~t b and h represent the contour of the interior of the shell.)

a = Tylotbyris laminosa rout./3. b = T!flothyris laminos~ rout. v. c = Tylothyris, between ~amin,o.~a and s~beo~dea (C-S). d = TgLothgeis su,bco,dza eastleto'r~e~sis (D). e -= Tglothyris s.bcoldca mrbconica (D), f = Spb'izt~eimt i.sc'.Ip(.. g = Spir~fe~'ina octoplic~tt.. h = P.netospir(fer scobric..~t..

31utatioLis ()f 71,//oth?/ris lami~tosa. The species ill its typical condition is characteristic of beds of Z~-C 1 age, but it is ~dso represented in K and in ZI. (a) '_PYLO:rH~:ItIS ~,~t~tI~OSX rout. ~,3 includes small specimens from K and Z,, in which there are indications of all ~he essen~iaI characters of the species. Description.---7: laminosa rout. /3 differs from the typical form of the species in the following" respects :--The shell is smaller and the area less elevated, the costm are more sharply defined, and the development of the apical callosity is less pronounced. The dimensions of an average specime~ are as follows :~width along hinge-line, 24 mm. ; height of are~t, 6 ram. ; length of bmchial valve, 12 mm. Horizons and localities.~K t in the Upper :Neath Valley, 1 K z at Mitcheldean and Burrington, ~ and rarely in Z~ in the Mendip m'ea generally. I Collected by the late Mr. C. H. Cunnington. " Vaughan [28] p. 384. Q. J. G. S. No. 302. e Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

'200 DR. r. J Nolvrh or~ [vol. lxxvi,

(b) T~'LOTHYRlS LXMI~OSA rout. ~.. Mutation 7 is the form which most closely approximates to the description of the species already given. Horizons and localities.--Mutation ~ is particularly characteristic of the upper beds of tile Lower Avonian in the South-Western Province. It does not everywhere attain its maximum at the same horizon. In the Mendip area it is particularly characteristic of Z., of Horizon 7 in the east of Glamorgan, and of the laminosa dolomites (C) in the Avon Section. The dimensions of an average specimen are as follows :--width along the hinge-line, 30 mm. ; length of brachial valve, 15 mm. ; height of area, 8 ram. Generally speaking, there is a slight increase in size and a greater development of the apical callosity in the shells from Horizon V and C~, but the differences are small; and, owing to the varying mode of preservation, the form characteristic of any pal~icular horizon cannot be definitely described in exact terms.

Further notes on T. laminosa.--Locally, the species reaches larger dimensions, and in such places usually seems to be particularly abundant. This is noticeable, for example, at Barry (Glamorgan), where in size, and in the intensification of certain characters the shells from T-C recall the special forms of Sj/ringo- thgris found in the knoll-limestones, and in the Midhmd ' Brachio- pod-Beds.' An unusually large specimen from the Cadoxt~n inlier, near Barry, had the following dimensions :--width along hinge-line, 40 ram. ; height of area, 12 ram. ; length of braehial valve, 20 mm) T. laminosa is not widely dish4buted above C~. Vaughan mentions it as being rare in C, in the Gower Peninsula,~ and as occurring in abundance in a band at one level in S~ in the Avon Section. .~ In Ireland it has been found in Z (? C,) at Hook Head and at Malahide; in Z~ (but not in C~) in Clare4 and at Bundoran. Large and small forms occur, as in England and Wales ; but there is not at present sufficient evidence to admit of a discussion of the zonal significance of the species in all the localities mentioned. The typical form occurs also in Belgium, as, for example, at Allain. ~ The cardinal extremities are frequently broken off, as in the :specimens from Burrington figured by Vaughan 8 : this gives to the shell a deceptive rounded appearance, but the precise nature of the cardinal extremities is usually indicated by the imbricating growth- lines, when they are preserved. The loss of the cardinal extremities

1 In this connexion it is in~eresting ~o recall Mr. E. E. L. Dixon's suggestion that knoll-reefs may be expected to occur in the Barry area, similar to those disoovered near the junction between the Lower Avonian and the Upper Avonian in Pembrokeshire. In the discussion of a paper by F. Dixey & s F. Sibly [6] p. 162. "~ Vaughan [7] p. 546. ~ Vaughan [33~ p. 195. Douglas [8] p. 549. Vaughan [35] p. 39 ; see also pl. vi, figs. 1-3. ~ Vaughan [28] pl. xxxi, fig. 8. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] SrIr AND SPIRIFERINA. 201 also results in the disappearance of the last three or four cost~e, giving the impression that there are only six or seven bold plications on each side of the field. 9In exfoliated specimens the costm appear to be stouter and less angular than when the shell is preserved. Judged by the figures, the specimens referred to by M'Coy as Spire'era Jecemcostat(r and S T. meso.qo~iia were in all probability exfoliated examples of T~/loth;yris la.~ni~osa; but, as neither of these forms is adequately defined by M'Coy's description and figures, and as the original specimens are no longer in existence, 1 the names are best regarded purely as synonyms of that species. M'Coy said of the species ' mesial hollow produced when old iato a long, tongue-shaped, flattened lobe,' o and Davidson figured a form possessing that character, which is, however, not of common occurrence. Davidson's figures are greatly restored and idealized, and appear to have been based upon three imperfect specimens now in the Sedgwick Museum. These I have been able to examine through the kindness of Dr. F. R. Cowper Reed, and they bear very little resemblance to the figures. Dr. Reed wrote concerning them:-- 'Indeed, unless we knew that these were the actual specimens which he (Davidson} used, and to which he referred, one might hesitate to accept them as the types.' (In ~itte~s.) Apart from the prolongation of the itoor of the sinus, the specimens are nearer to T. sztbco~ice~ than to ~/: l~z~i~osa i1~ the height and curvature of the area. The locality is given as Derbyshire, but the horizon is not known. Comparison with :T. subconice,--Tylolh:qris l~ztiJlosa and its mutations are readily distinguished from I'. subconica by the greater concavity of thearea (compare figs. 4 a, 4 b, & 4 d, p. 199). In T. szebco~lica the sinus and the costm are more angular, the line along" the base of the sinus less curved, and the fold less elevated than in I". laminosa.

Shells incorrectly referred to Tylothyris laminosa (M' Coy).--A number of shells which do not really belong to that species have been referred to ~/'. la~inosa, on account of their laminose ornament. Two of these are of sufficient interest to be dealt with separately :-- (1) 'Spiriferiq~a laminosa' from beds of Dib~wphytlum age at Redesdale (Northumberland) and other localities, so named by King ;

1 These specimens, together with the holotype of lV['Coy's Sl). taminosa and others, were formerly in the ' Griffith Collection,' but their whereabouts are now unknown. Dr. R. F. Scharff very kindly caused a fresh search to be made for them in the collection of the National Museum, Dublin, but wifJaout success ; and they have not hitherto been recognized in any public collection, although Davidson speaks of having examined the type of 813. decemcostata [4~ p. 43. In all probability the original figures were, as was often the case, idealized, so that actual specimens, if met with, would pass unrecognized. / M'Coy E21~ p. 426. 1'2 Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

202 m~. v. a. ~'ou'ru os [vol. lxxvi,

and mentioned by Vaughan in his Belgian paper? (See text-fig. 5f, p. 211 ; PI. XIII, fig. 6.) (2) Shells occurring in the P+'odl~ctus corr~tgato-hemisphecicus zone (S)of the Ravenstonedale district, doubtfully referred to 'Sldriferina laminosa' by Prof. E. J. Garwood? (See text-fig. 5 9, P. 211.) They are undoubtedly time-variants of one and the same species, but cannot be included in M'Coy's lumiuosa, from which they differ in shell-structure and in the non-development of the apical callosity. They differ front a typical ~S~ri]~kri~a in having numerous snmll, somewhat rounded cost~, and ira the relatively- large rounded fohl and sinus. Those forms will be referred to a new genus, P,nctospirife+', and will be considered after a general discussion of the Carboniferous ' Spiriferinas' and their relation to, the type-species of that genus. Certain other Spiriferoid shells in the Carboniferous Limestone have a strong imbricate ornament, and might, especially in fragmentary specimens, be confused with either "_1'. laminosa or .Punctospirife~'. They have, however, an impunctate shell- structure, and no ventral median septum, and must therefore remain at present in the protean ' genus' SpirO'kt'. They may be l~adily distinguished from the other lamellose genera discussed ia this paper by their transverse shape, nmcronatd eaMinal extremities, and particularly by their low incurred h~pir#~'r-likc area and numerous small costal. They include forms such as occur in the Dibunoph!/lhtm Zone in the Midlands, and externally have much in common with the North American l)cvonian species. S. pe;~n<~lus Atwater (=& +~ucronaht.~" Cornel(I).

Tk'hO'rnYRlS SUBC()Y[C~ (Martin). :~ The original description and [igures do not a~tequately detine this species, and the name has in consequence been made to include more than one form. it is therefore, as in the case of T. lamino,'~, necessa~ T to re-define the species. Unfortunately the holotype has been lost, and i am unaware of the existence of any specimen froln the original locality which could serve as a topotyl)C. Forms which can be referred to this species are found in Derby- shire and StaffoMshire, the Craven district, and in h'eland, in each case in the upl)er beds of the Dibunophyll~m Zone. The shells may*be divlded into two groups, in one of which the area is slightly incurred at the beak, while in the other the area is practically a plane triangle. These two groups are not strictly time- variant.~, neither do they entirely owe their characters to the type of deposit in which they occur, although both factors play some part in their differentiation. They are best regarded as sub-species of T. s+tbconicu. I Vaughan [35] p. 44. Garwood [13 p. 572 & pl. li, figs. 7 a-7 e, text-fig. 7. Not ,qpir(t'ero s.bc(,//ic~ Phillips, which is a synonym of Cy~'timt betero- clit~t. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

r~'L)TUYRISr ( SU~CO~'~C~ SURCONIC,~ (Martin) emend. (Text- rio.. -~e, p. 199 ; P1. XI]I, figs. 5a-5c.) 1809. Conchyliolithus am)mites s~tbcoMcus Martin, ' PetrifacttL l)crbiensia' pl. xlviii, figs. 6-8. 1~58. Spirifera subeonica I)avid~on [4j p. 48 & pl. ix, fig~. 3, 3 a, 3 b. 1906. Syriuqothyris subconiea Vaughan, 'Rush Paper' ~34,n, p. 311 & pl. xxx, fig. 7. This subspecies includes the form upon which the species was originally founded, since in the originM account Martin described his specimen as having an area ' which is straight, and not recurred '~s in cu,~id~h~s.' His figures also indicate a flat area. D e s c r i pt i o n.--Shell spMferoid, of medium size, twice as wide as long, greatest width along the hinge-line, cardinal extremities slightly produced. P e d i e 1e - v a 1 v e acutely 1)yranfidal, cardinM area practically plane triangle, a little more than twice as wide as high ; plane of ,the area at right angles to that of the brachial valve, its lateral c(t~es well defined and angular. Apical angle about 115 ~ IIelthyrimn twice as high as wide, delthyrial covering unknown. Lateral sh)pes slightly convex or nearly ttat. Sinus angular, Sharply defined by angular ridges diverging at an angle of about 30 ~ The line along the floor of the sinus from the heak to the anterior nmrgin is very slightly curved or practically flat. Brachial valve slightly convex, nearly senficircular in outline. Mcsial fold distinct, low, and r,)unded; its convexity decreases towards the anteri6r margin, while that of the lateral sh)pes increases in the same direction. Each lateral sh)pe in both valves is ornamented bv ten or eleven siml)le angular costm of which the first six or seven on either side are large and well developed, the remainder being short and incon- splm.u)us and not apparent in exfoliated sl)ecimens. In some Sl)Cemlens the eostm are slightly sinuous: thev are separated by furrows al)proximately equal in del)th and wi(Ith to the costm on each side of theln. Surface of both valves, except the cardinal area, marked by regular imbricating lamellm. Internal structures:--Pedicle-valve.--The delthyriM sul)- porting-plates extend across the cavity of the valve, and are divergent. Median septmn well developed, reaching nearly to the level of the area near the beak, becoming gradually lower, and disapl)earing about half-way between the beak and the anterior margin. The apical portion of the septum i.~ embedded in the apical c:dh)sity, which is well developed. In the brae h ia 1 val v e there is a low median septum extending for about two-thirds of the distance from the beak t() the anterior margin. Shell-structure fibrous and impunctate. The dimensions of an average specimen are as follows :--width along hinge-line, 36 mm. ; length of hmchial valve, 18 ram. ; height of cardinal area, 15 ram. Holotype (of S. subconica s~tbconica).--A specimen from Attcrmire, in my collection, now deposited in the National Museum Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

204 n~. r. J..wORTH O.~- [vol. lxxvi, of Wales. Other examples, from Settle, are on a tablet numbered 501 in the Sedgwick Museum, Cambridge. Horizon and loealities.--Upper Dibu~ophyllum Zone (D.,-D~). Principal Sibly repolCs it as being characteristic but rare in D:, in Derbyshire and North Staffordshire, 1 while in lreland i~ is characteristic of the Cyathaxo~ia Beds (D~) at Rush ~ ; Mar~in'a figured specimen was reeorded as from Middleton (Derbyshire). A numl)er of sl)ecimens from the neighbourhood of Settle are in the Sedgwick Museum. Remarks.--In the original description (,f ' A. subco~tic~s,' ~M~artin eoml)ared tiffs species with his 'Anomitcs cuspidat,~s,' from which it differed in having fewer and more angular ribs, and a fiat area. He said that, in the shape of its area, it was intermediate between that species and another uudeseribed form with a more in: curved area ; this form was, in all probability, 7!qlotttyris lu~l~i~osa. Davidson expressed a doubt as to whether Martin's species was only an ex.~eptiona] variety of ~'. cuspidttla, and reproduced Martin's figures. :~ In his appcndlx to the ' Carhoniferous lh~tchiopoda,' Davidson figured a testaeeous specimen of ~/: strbconica in which the lamellose ornament was perfectly 1)reserved, and admitted its specific distinction from S. cuspidata, but eomlmrcd it to S. lami~losu. De Koninck stated that the lamellar orn'tment occurred in young shells only, 4 and was absent or scarcely visible on adult Sl~eeimens; but that author, owing no doubt to the state of 1)reservation of his material, apparently regarded C!p'ti~ur seT)lo.ua (Phillips) as the adult form of ~: subconica, for the form which he figures as the l~ttter sl)ecics is certainly C. seT/osu. The lamellar ornamentation is not often preserved; in ex- f()liate~l specimens the eost'e are smooth, and al)pear morc rounded than in perfec~ shells.

TYLOTHYRIS SUB(~ONICA CASTI.I~TONENSI.~, subst), nov. (Text- fig. 4d, p. 199; P1. XIII, figs. 3, 4a & 4b.) This form differs front T. subconica slebco~ico in the following respeets:~The shell is rather less transverse, and ~be cardinal extremities are blunter, 1)eing 1)tactically right angles. The cost~ are stronger, less angular, and towards the cardinal extremities are less clearly defined than in the other subspecies. The sinus is shallower and less anguhr, and the line along its floor more strongly curved than in ~/: subconica. The alva is more or less concave, and terminates in a small, incurred, pointed beak (see fig. 4. d, p. 199).

i Sibly [31] p. 51. 2 Vaughan /34] pl. xxx, fig. 7. :~ Davidson [4] p. 48 & pl. ix, figs. 3, 3 a, 3 b. These figures were apparently copied on the stone directly from Martin's drawings, and are therefore reversed. 9 4 DeKoninek [19] p. 255& pl. xiibis, figs. 5a-5~. (Roallyfigs, 4a-4'c, although in the text, as well as in the explanation of the plate, he cites them as 5 a-5 c ; fig. 5 is a species of Clrone~e.~.) Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] 8YRINGOTHYRI8 A.:ND SPIRIFERINA. 205

Horizon and localities.--Upper.Dibunophyllum Zone (D2) Wetton (Staffordshire) ; Castleton, etc. This form is more particularly characteristic of the ' Braehiopod-Bed' type of deposit, and its relation to T.ylothyris subconica subconica recalls that which exists between the forms of Syringothyris occurring in normal deposits, and the special types found in the knoll-phase limestones. H ol o ty p e (of T. subconica castletonenMs).--A specimen from Settle on Tablet 501, in the Sedg~ ick Museum, Cambridge. Comparison with Syringothyris cuspidata.--T, subconica subconica, T. subconica castletonensis, and S. cuspidata sometimes occur in the same beds, and have frequently been confused. The difference between the first two and the last is, of course, at once evident if the beak of the pedicle'valve is broken, or is rubbed down ; but there are also external differences. Com- pared to either of the species of T.glothyris with which it can occur, S. cuspidata is a larger shell, has a higher area, more numerous and less angular eostm, and a less sharply-defined sinus iu the pedicle-valve. Moreover, it never e)zhibits the lamellose ornament seen in suitably-preserved specimens of Tylothyris. The shell recorded by Phillips as a ' remarkable variety of Spirifera cttspidata' was in all probability T. subconica. 1 Comparison with T. laminosa.--The flat area of T. subeonica subconica at once distinguishes it from ~/1 laminosa. T. subconica castletonensis differs from that subspecies in its higher area, the beak of which is not much raised above a line perpendicular to the plane of the brachial valve. The fold is more elevated, and the line along the floor of the sinus more strongly curved in T. laminosa than in the later species. Certain shells occurring in the Dibunophyllum Zone which are more or less intermediate between T. subconica castletonensis and T. laminosa, but more closely resemble the latter species, are included under that name, as no useful purpose would be served by attempting to differentiate them. They include forms like that figured by Oavidson, ~" and the spechnen from Bundoran mentioned by him on p. 156 (see fig. r c, p. 199). This latter form from the ' Calp,' which includes representatives of S.~ and D, is in all probability intermediate in age between T. laminosa and T. subconica. Here also belongs a specimen referred to by Martin in his description of Anomites cuspidatus. He says : ' In the other anomite which I have not as yet named or described the hinge is still less than in A. su,bconicus, and the beak of the large v~lve instead of being straight is somewhat incurred.' J. Parkinson, in his 'Organic Remains,' figured what he believed to be the form to which Martin referred. His figure shows clearly the curved area and the laminose surface-ornament. 3 Phillips [271 p. 216. Davidson [4] p. 36 & pl. vii, figs. 21-22. :~ Parkinson [26] p. 235 & pl. xvi, fig. 16. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

206 YR. r. a. ~'O]~Tn o~ ~ Lvol. lxxvi,

Tgloth:u svbconica is, therefore, connected I,v intermc~liatc forms with the Lower Avonian species, 7: lamiaosa: [rotn which it undoubtedly descended. So far as I have been able to ascertain, there are no representatives of Tyloth?lris in Permian or newer hods, and, as in the ease of Syrinyoth?p'is, its l,re-Carb(}niferous history in Britain is at present unknown.

V. THE .REL&T'ION BETWEEN J~IqRII,'t,/ItI.V.'I AND ~ I'RIN(;OTil ] 7N,~', The late Dr. Arthur Vaughan, ill his ' Bristol P:tl)el',' eonlpared certain Spiriferoid shells from tlle ' Ul,per Tournaisi~m' of the Avon section with M'Coy's species '~S)/r(fi.ra laminosa,' and later [~35] p. 48, definitely expressed the opinion that they did indeed belong to that species, an opinion which has now been fully con- lirmed. Dr. Vaughan, however, referred the species to the germs Syringothyris, remarking, [83] p. 301, that ' w~ithin the delthyrium there is a distinct rostml c~llus ,rod a trace of a syrinx. I have not as yet obtained unequiwmal proof of the existence of syrinx, but the presence of the rostral callus is indisputable.' In a later paper, [28_] p. 383. he stated that 'wholly or partly buried within this callus is an infilled syrinx that c:n~ be seen hMf projecting Mong the middle line of the cMlus, and in weathered specimens is represented at the apex by a short narrow rod. A cross- section near the beak shows the infilled syrinx "tnd the dental-plates ; at a little distance below the beak the strong mesial septum begins to appear.' No specimen that I have examined affords any justification for assuming the existence of a syrinx cml~dded in the apical callosity, and in the absence of a figure it is impossible now to determine the precise nature of the structure to which Vaughan referred. His allusion to a short narrow rod seen in the apex of weathered specimens suggests that it may have l~en the edge ~}f the median septum projecting through the uplmr sm~ace of the apical callosity ; but, in any cast, it cannot have been comparable with the syrinx of ~:yrin.qoHtyris, for the following reasons: .... (1) The essential characters of 8yrlngothyri.s do not appear in sections immediately under the beak. (See fig. 1 a, p. 166.) (2) There is no elevated median soptum in the ventral valve of Sy~-i~gothyris. (3) The two structures: a median septum and a syrinx-bearing transverse plate, could not coexist in the same shell, since they would mutually interfere one with the other. (4) Sections (see figs. lm& I n, p. 166) across the beak of T?#oihyvis ~aminosa clearly show that it is not a Syringothyris. (5) No known specimen of Sy~,ngothyris has the imbricate ornament of T. laminostt. Vaugh~m recorded the close connexion between th~" species T. lar and T. subconica in the following terms :-- 'S. zubconica is so similar to S. [aminosa that its specific separation has been viewed with scepticism. There seems, indeed, to be no reasonable doubt Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2 ~ ,~ YR I AGOTH 11r AND ,S'I'tl~I I.'EI?I.VA. 207 that 8..s.beo~tic. is the direct descendant of S. b~mi.o.~., and differs ,,nly in the accentuation of the traits already adumbrated in tim earlier form." But he further states that ' A cross-section of the beak of S..*,bconica reveals the characters of S!lrin- 9othyris as typically developed as in S. c~spidota (Mart.) itself.' This, however, is certainly not the case. S. eusl)ida/a and S. s ub- conica frequently occur together, as, for example, at Settle; and small specimens of the forlner bear soine slight rcsembhmee, externally, to exfoliated specimens of the latter, although the difference between the two is at once manifest when the beaks are rubbed away, and the cross-section examined, it was apparently upon a small, wrongly-identified specimen of the former species that Vaughan based the statement just quoted. Unfortunately, from the supposed existence of the ssrinx in 7: /aminosa and the assigmnent of 5/'. subconiea to S.~rin!/olh,qri.~', Vaughan drew certain conclusions regarding the l>hyh~geny and relationship of the two genera Sl)iri/'eriJta and ,b'~J'i~..qolh]/,~'i.~. :- 'Hence, S. g, mi~tosa develops, phylogenetically, into a typical Syringo- th.!l,'is.... The results obtained indicate the close relationship in Carboniferous times of the two genera Spirife'rina and Syringoth !p'is .... The most probable guess is, that Spiriferi~ut and Syringothyris were both derived at much the same time from a common ancestral group, and that they developed along different lines, but that in the earliest period of their history they possessed the same essential structure.' ~I,oc. cit.] Sufficient evidence has already been adduced to shiny that this assumption is not supported by observed facts, for 7: lominos, is not a Spiri)~'erina, and that genus and b'qri~//othffris have practically nothing in common. What is already known concerning the evolution of SyJ'in.qothyris shows that at no period of its history did its ancestors show any sign of the (levelopment of the essential features of Spir~/'o~'i~ut: nalnely, a strongly-lmnctate shcll-structm'c and an elevated median septmn in the pedicle-valve. Finalls, there are early Palm()zoic Spiriferoids with an iml)ricate 4q'nalnen- tation of the shell alld a pronounced median septum, altlOl~g which h) look for the ancestors of Spi~'(/'erimt. It is evident that fnrther study of the subject w~ls causing Vaughan to modify his views, and but for his untimely death would in all iJrobability have cleared up the uncertainty which had resulted from his interpretation of T. la.minos,. In his last paper, he speaks of two forms included under that name. One of these, characteristic of the Lower Caninia Zone, he still regarded as a Syrin!lothyris, while the other, characteristic of t,he Upper Zaphrentis Zone, with :t primitive and buried syrinx and an impunetate shell-structure, he considered to be a S])ir~f?;r, or a rudimentary ,~'y'ri.n!/oth:q,'i.~. He definitely stated that neither is a 8pir~ferina, and added : ' In any ease, it i's not possible to sept, rate usefully the two forms in tield-wm'k.'

..: Vaughan ~28 -~ p. 384. "-" Vaughan ~35_" p. 44. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

208 mr. r. J. ~rORTtt o~= [vol. lxxvi,

Of these two forlns, the second is undoubtedly the Zaphrentis. Zone mutatmn of the species here described as T. laminoso, but the identity of the first is not so clear. To illustrate it, Vaughan quoted the figure of a specimen from Horizon y at Burrington Combe, l and describes it as having fewer ribs than the second form. The cardinal extremities of the tigured specimen arc broken, and this, as has already been mentioned, conveys the impression that the shell possesses fewer ribs than it actually does. The tigurc so closely resembles an exfoliated Z.: specimen as to suggest that Vaughan's two forlns arc time-variants of the same species.

VI. SPIRIFFAtINA I)'Orbigny. Shells referred to Spiri/'eriJ+a occur in Carl),niferous, PerJnian, and Liassic deposits in the British area ; and elsewhere in the Trias, where that formation includesmarine det)osits. It has, in tact, been customary to regard all Spiriferoid l}mchiol~da in which a ventral median sel)tmn coexisted, or was supposed to e(~xist wifll ~L t)unctatc shell-structure, as ~'pirife,,'ina, merely because there was no other genus to which they could conveniently be referred, and not from any failure to recognize tlle polyphylctic n'tture of the species involved. The genotyl)e of S/,ir(/;~'iarr is a Liassic shell (Sp. ~'os/~'ata Schlotheim), and before we attempt to discuss the Pah~ozoie forms, some account of the gemls in its original sense is necessary. The genus Spi~'~ferina was established t)y I)'Orbigny, in 1S4.7, ~ for certain Liassic brachiopoda previously included in Spirited', but whi~'h differed from tvpical members of that genus in tile possession of a vertical shelly septum in the ventral valve, and in the existence of abundant relatively-large canals or punctze in the shell-structure. It is somewhat unfortunate that D'Orbigny took as his genotype ~'#..>'osl4"ata Schh)theim, because in external appearance that sl)eeies i~ not ~ypical of the genus. It has a uniformly roun(lcd, or even gh)l)ose shell, the surface of which may be quite smooth, or at the most only slightly undulated ; while in the other SlX,(.ies there is usually a radial ornament consisting of well-defined angular costze. External characters:--Shal)e and dimensions.---The shell is ovate, and the hinge-line less than the greatest diameter of the shell. The cardinal extremities arc rounded, and ttle lateral slopes of each valve tumid. The size is variable, but the dimensions of an average specimen are--width along hinge-line, 28 mm. ; length of brachial valve, 21 ram. Usually, as in Sp. walcotti, the ornament consists of a few prominent angular costte separated by V-shaped furrows, on each side of ~ median fold in the brachial valve and a corresponding sinus in the pedicle-valve. The fold and sinus are more or less angular and resemble the costw and furrows, from which they differ only in size.

I Vaughan [28] pl. xxxi, fig. 8. C. R. Aead. Sci. Paris, vol. xxv (1847) p. 268. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] ,b'YIr 8 &I~D ,.~I'I Ir I FElll NA. 209

The plications of the shell impart to the anterior margin a chalac- teristic serrate appearance. The surface of the shell is marked by a few concentric imbricating lamellze of growth, widely spaced in the ulnbonal region, but closely crowded anteriorly. In well-l~'eserved specimens the outer layer is covered with title, tubular, spinosc projections, which when broken, as is usually the case, give to the shell a characteristic pustulose appearance. The area.--The area of tile pedicle-valve is variable in height, but the average apical angle is 140 ~. When the area is low it is strongly incurred, though as its height increases it becomes flatter ; the plane of the area is always inclined to the plane of the bmchial valve, so that the apex of the pedicle-valve is above that of tile opposite valve. The area i~ defined by a pair of lines diverging from the apex, attd outside of which the shell gl,~dually curves over into the late~-al slopes. The delthyrial angle is about 40 ~ In his original description l)'Orbigny stated that the 'delthyrial aperture' was without covering ; but this was due to the imperfection of his specimens. The delthyrial covering consists of a pair of imperforate deltidial plates slightly inclined (me to the other to form a roof-like structure. The delthyrial covering is rarely preserved : but its former existence is indicated by the grooved margins ()f the tissure. Shell - structure. --The shell is tibr

210 ~)~r F. J. NORTH ON Lvol. lxxvi,

(,,) l're-Liassic Species ret'errcd lo S'piriferina. A general discussion of the shells from infm-Liassic f~n'mati~,ns hitherto) referred ~ Spir(ferin, a is 1)cset with many difficulties. In this country the sequence is interrul)ted bv the m)n-fossiliferous Triassic rocks, the conditions Jwcessarv for ti~e formation of which must have profoundly affected the lifi~ of the times : and, assmning that the Liassic shells are homogmlctic with some of tile Carl).niferous species, the enforced migration in l)ost-Carl~(mifer.us times may be expected to have induced c,)nsideralde changes i1, the animals, before their descendants" returned t() this ;m,a witl~ tile expanding Liassic sea. The 1)aueity of bur knowledge -f the lwachidia and ass~wiat(,d structures ()f tim earlier species is ;ds,~ unfav(mral)le to their satisfactory (.lassification. This might, indeed, 1,. urged as a reas, m ;tgainst attempting any generic differentiati()n; but. since the opport, unities of obtaining evidence of this natm'e are s,) few. anti there are certain persistent differences in t,xt(,rna[ apl~t,:~raltce, ill- volving 1)oth contour and ornalnent.'tti(m (the ~mly clmnteters (~f any real value to the field-worker)..'t sehcm~' ~,f classiiieation 1,ascd m~dnly, but not exclusively, Ul)(m these featm'es seems justiliahlc. The laminose-punctate-scptate Spiriferids nmv I,e divilh,d ,,ll external characters into at least two w,'ll.drlilw~l s(,ru.s : (1) Shells with an almost flat, triangular, moder:rtely-hi~'h area, with anguh~r margins, by which it is sharply differentiated from the lateral slopes: l)~.;d~ inconspicuous and l)ointed, foht ~uld simls well developed, round(,d, and considerably larger than the costal, of whic'h there are about six or seven on each lateral slope. 'l:he greatest width of the shell is at the hing~;-line. The anterior margin has a large rounded medi:ul fohl, and ia wrinkled laterally. (Sea figs. 5 f-S j, p. 211.) (2) Shells with curved area. the margins of which are more or less rounded, and curve over into the la~rM Mopes; smM1 incurved beak in the pedi(rlc- valve. There are a few large angular ooshe, and ml anguhu" fold and sitars structurally like the cost,se and furrows, than which they are only a little wider. The anterior margin of the valves is strongly aerrat~,l, with :~ few sharp folds, decreasing in amplitude from and in(.lnding the median fold. (See figs. 5 n 5 e..) These distinctions apl,ly t~, shells ill both Palaeozoic and M,,soz~,ic formations. In the Carboniferous IAmestone the first t.vl,e is represented by the Ravens~ncdale and Redesdale forms aires,Iv mentioned (and hitherto referred to ~b)~/rO'~'riaa /a minosa), all(l il(e second bv Spi,r~)C'eria, o,:/oplirata, in tlle Permian, I~()th tyl~(,s occur among the shells referred to the iml)crfectly mlderst~,.d species, SI). eris/ata (fig. 5 d) and Sfl. multi/,lieata (fig. 5 h), and they both also occur in the marine Triassic deposits of S,:,utl~- Eastern and Central Europe. In the Lias the first type is re]we- sen~d by Spi~'~'erin.a o.viq.qona (fig. 5j), and the second l)v ,~)~. walcotti'( fig. 5 e). In this ])~rper it is proposed to assign the Carboniferous species belonging to the first of those two types to a new genus, l'~mcto- sl)ir(/'ei" ; those bel,mging to the second type are best regarded as Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part ')" ,s I "RI NGOT[I I "RIS t N D SP[RI I"ERINA. 21. ],

St,ir~f~;;iJ,,, until more evidence as to their internal structures i~ available. The ('arbonifer, ms' Spiriferinas ' of North America also comprise two similar groups. The first type is there represented by S_p. ken- tackff~asis, in which the plieations are small, numerous, and crossed by elosely-1)aeked concentric lamelhe ; while the seeond is represented by forms like ,b'p. spinosa, which strongly resembles the British ~,;p. octoplicata3 According to J. Hall & J. M. Clarke, the jugum of $1~. ]ceut~cl,'!/cJt~i,~ differs from that of the Liassic SI). rostrata

Fig. 5.---Fo~'m of the aule~'ior ma,~'#i~t of the yah, as of. Spiriferina and assoc;ated (je,Jtera.

d h

~--Spirife,'rina insc,tpta, Carboniferous Limestone (D). b = Spiriferi~u~ octop~icata, Carboniferous Limestone (D). e = Spi~'iferina perl~biz'ata, Carboniferous Limestone (D). d = Spiriferina eristata, Permian. e -- Spil'iferina tcalcotti, Lower Lias. f = Pat~ctospb'ifer .~cabrico.~t,t rout. redesdaLe,sis, Carboniferous Limestone (D). g = Pu~wto.~pirifer ~'cabricosb~ rout. ad~fellensis, Carboniferous: Limestone (S). h = SpiJ'(/~rina mull ipl.icata, Permiani i = Sl,iriferi,a o,,'ggmu~, Lower Li~.~. ill that its latend portions converge upwards between the primary lamella, of the spires, and form a structure resembling :m inverted V, instead of being a siml)le transverse loop as in the latter species3 This fact favours the sel)amtion here suggested, especially as the jugmn of Plc~etospirif~,,r is of a similar nature ; while, in a specimen of Sp. cNstata figured by H. B. Geinitz,~ the jugum was like that of Sp. rosfrat,. The ('arl)oniferous shells are considerably smaller than those

1 See Girty E9] pp. 371- 72. "- Hall & Clarke [14] p. 54. 9~'Dyas oder Zeehsteinformation & dt~s Rothliegende ' 1861, pl. xvi, fig. ~.. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

212 I)R. 1~..L ._x'oa'rI~ ON ~vol. lxxvi, from the Lias, and the Permian shells in this country are smaller than either. The small size of the 1)ortolan specimens is doubtless due to the unfavoural)le conditions under which they lived; while the large size of the Liassic species is in keeping with the fact that the race was nearing extinction. Although there is every reason for believing that such is the case, it is not here definitely asserted that the Permian and Liassic species mentioned are homogenetic with the Carboniferous species cited as examples of the two types. A discussion of the post- Carboniferous forms must be deferred, until it has been possible to examine the holotypcs (where they still exist) or topotypes of the species concerned. The whole series of laminose Spiriferids affoMs an interesting example of the conditions referred to by the late Dr. Ivor Thomas when discussing the Productids 1 :~ ' It is becoming more and more recognized that possession of similar features alone may be misleading and an insufficient criterion for the estimation of species. Externally the individuals may be morphologically almost indistinguishable, and yet belong to totally different lines of development and therefore to different genera.'

(b) l'L~Ne'rOSl'Jl~ll,'El~, gen. nov. l)escrll~ti~n. -Shell sph'iferoid, about twice as wide as long. greatest widt]~ at ,~r near the hinge-line; cardinal extremities slightly r,,unded ,,1' subangular ; area moderately high an(l concave. Pcdicle-valw~ subpyramidal, lateral slopes evenly convex, mesial sinus wide and shallow; beak small, pointed, and slightly incurred. l]mchial valve convex; mesial fohl well developi~d, rounded, dis- tinetly raised above the general level ~f the valve, and about as wide at the anterior margin as three of the adjacent costm and the furrows sel)amting them. Lateral slopes of both valves ornamented by rounded costm, separated by rounded furrows, about equal to them in width. Surface of both valves crossed by regularly-disposed imbricating lamellm, which in adult specimens may be more or less obsolescent, especially in the posterior portion of the valve. Shell-structure fibrous and strongly punc~ate. Internal structures:--P~'dicle-valve.--Delthyrial sup- pol~ing-plates slightly divergent: median septum well developed, thickened at its base, and near the beak extending about two-thirds of the distance fr,,m the floor of the valve to the area ; it terminates somewhat abruptly about half-way towards the anterior margin. There is no marked development of an apical callosity, so that the three vertical septa appear as distinct structures in sections quite near to the beak. Brachial valve.--P, low median ~.rest t>iseets the muscular sears; the spiral coils are large, ;tll,l have their apices directed 1 , British Carboniferous Producti' ]~r Ileol. Surv. : P~Llmontology,vol. i, pt. 4, p. 253. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

I~rt 23 ,~)J~/.x (;o)~JYir a~ D .S'/'H~]I.'~']r 213 laterally to a point a little anterior to the cardinal extremities; the jugu1~.~ is slender, and forms a .~hallow V-shaped process, with its apex directed ~'entr~lly and posteriorly9 Ran ge.--(As at present determined) Carboniferous Limestone. ~ Genotype.--P. sectbricosta,sp. nov. Remarks.--This genus differs from Spir~.;feri~zain the greater nmnber of its eostw, which are smaller and less angular, in the greater relative size of its fold and sinus, and in its jugum. ]~'rom Tylothj/ris it differs in its punctate sheU-substance, and in the absence of a well-developed apical callosity.

PUNC'rOSPIRIFER SCABRICOSTA, sp. lIOV. (Text-figs. lp, 4h, 5 J, " & 5g, pp. 166, 199, 211 ; P1. XIl:I, fig. 6.) 1912. sTiriferi~za laminosa (M'Coy), Garwood F13] p. 572 & p]. li, figs. 7a-7e. Description.---Shell spiriferoid, of mediuln size, greatest width a little below the hinge-line. Pedicle:valve subpymmidal, cardinal area about a thh'd as high as wide; concave, especially towards the beak which is small and pointed ; apical angle about 110 ~ delthyria] angle -~(1~ sinus shallow (width about three times the depth). Brachial valve convex, median fold narrow at the beak, rounded, and strongly defined anteriorly. Each lateral slope in both valves bears from seven to ten simple rounded costm, separated by rounded furrows, and crossed by a strong concentric lamellar ornament. Internal structures :-- Pedicle- valve. Delthw'ial supporting-plates attached to the floor of the valve for about a third of its length; median septum attached to the floor of the valve for half its length, thickened at the base, and tapering rapidly upwards; no apical callosity. The dimensions of an avel~tge specimen are--width along hinge- line, 20 mm. ; length of bmchial valve, 12 ram.; height of area, 10 ram. Holotype.--A specimen from the Ashfell Sandstone (lower part of the Productv.s-corr~tyato-hemis2~ltericzts Zone, S) of the Ravenstonedale district, in the collection of Prof. E. ,l. Garwood; figured, Q. J. G. S. vol. lxviii (1912) p1. li, figs. 7 a-7 e. Comparison with Tylothyris laminosa.---The more obvious differences between this species and T. laminosa (namely, the punctate shell-structure and the absence of an apical callosity) have already been mentioned. It is further distinguished by the fact that the shells are usually much smaller than the representatives of T. laminosa and T. subconica, which occur in the Upper Avonian strata. i The jugum as seen in a specimen rubbed down for the purpose is strikingly similar to, but rather less angular than, that described by Hall & Clarke [14j p. 54, as occurring in Spiriferina kent~eckyensis, which in external appearance also has much in common with the genus under consideration. 2 But it is highly probable that the genus includes species in Permian and even later formations. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

2t4 DR. F. a..~oR'rH ox [vol. lxxvi,

Mut~ttion,~ of l%tnctospi~'(/~r scabrieos/a. P. scal,'icosta is typic~tlly developed in the Upper Avonian of the North of England. It is found in the Productus.eorrugato. hemisphe~'ieus Zone of Westluorlttnd, and in beds of Dib,tno- l)hylhon ~ge in Northumberland. The forms characteristic of those two horizons tx)ssess certain distinctive features. The diffcrenee~ between them ~trc not sufficient to warrant specific separation ; hut the later type is undoubtedly ~L time-variant of that which occm.~ at the lower horizon, ~tnd they may be conveniently considered as mutattions of one species.

I1) l ~. SCABItICOS'rA, nlllt. ASllFFLI,FNSIS. (See text-fig. 5q, p. 21.l.) :E xa m p I e.--The holotyla' of P. seabricosta. Horizon.--P~'oductus-corru!la, to-hemi~clghericus Zone (S) of the Kirkby-Stephen and Ravenstone&de districts, where it is found in the Ashfell Sandstone in the lower l~art of that zone, :Lnd in the ' Brw)z~,a-Band' at the top of the zone (S~).

(2~ I'. ScAl~l~lCos'r.t, mat. U~:OESO,tr,~;XsrS. (,See text-fig. 5./; p. 211; Pl. Xlll, lig. 6.) Ex;tmples.--Sl)ecimens descril)ed by King, l)avidson, and ,,thers, as ,~'1,i~'~/'e.~'r 1, m i u osa. 1 Horizon. Bernician (l),, Redesdale Beds) of Redesdale, etc. { Northuml,erla nd). These two mutations differ in external fl;atures. The rout. ~tsl!/'elleJlsis :qq~roximates most closely to the specific characters ~drca(ly givel~ ; mu~. redesdalejtsis differs from it in the following respects : (1) The ~re~ is a little less elevated, its height being about a qu&r~er of the width, instead of two-fifths, as in rnut. a~h.t'elle~i~. (2) The fold in the br~ehial v~lve is leas elevated, and the sinus in the pedicle-wlve shallower. (3) The cos'~ ~re rather more numerous and narrower, there being about nine or ten on e~eh l~terul slope, insteaxl of only seven or eight, as in the earlier mutation.

(c) ('arbot~iferous Species of Sl,;riferim,. Forms still retained in Spirt.;f'eriua occur throughout the Car- l~oniferous Limestone, from the Cleistopm'a Zone to the Dibuno- phfllum Zone; they reach their aelne in the Upper Dibunophylh, m Zone, where they include s iuseulpta (Phillips), ST. oetoplicata (J. de C. Sowcrby), ~nd SP. pe~Tlicata, sp. nov., species which differ in the number and intensity of their plicati6ns. E~rly mutations of the mature forms occur at lower horizons, but they are not common, and have not hitherto been satisfactorily classified, r )e general tendency has been to refer to them all as

I King [17~ ; Davidson [4: p. 37. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

l>al~ 2] ~'~'~r )~ND 8PIIr 215

Spiriferina cf. octoTlicata ; while Vaughan compared certain of them with the Belgian species, S p. peracuta De Koninck. Since the three species mentioned are most typically developed in the Dibunophyllum Zone, and the holotypes of the species ah'eady described (namely, octoTlicata and insculpta) were obtained from beds of that age, it seems desirable to select a form froln that horizon as the holotype of the new species Terplicata, and to deal first with the three species as typically developed, before discussing their earlier mutations.

Se[RIFERINA oc'rOPLI(2A'rA (J. de C. Sowerby). (Text-figs. 4 9 & 5 b, pp. 199, 211 ; Pl. XIII, figs. 8 & 9.) 1827. Spirlf~r octoplicatus J. de C. Sowerby [32] p. 120, & pl. dlxii, figs. 2-3, no~, fig. 4. 1844. 8pirifer cristatus M'Coy [20J p. 133. 1859. Spiri.ferina cristata, var. octoplicata Davidson [4] p. 38 & pl. vii, figs. 37- 47 (probably only figs. 38-41 really belong to this species). 1915. 8pir~ferina octoplieata (Sowerby), Vaughan [35] pl. iii. The original description is as follows :- 'Transversely elongated, gibbose, semicircular, plaited; plaits 8 or 10, deep and angular; central elevation plain; beaks remote, incurred ; area triangular, curved." Description.--Shell small, transverse, cardinal extremities slightly rounded. The dimensions of a specimen of average size are : width along hinge-llne, 20 ram. ; length, 10 mm. ; thickness, 8 ram. The area of the pedicle-valve is concave, the curvature increasing t~wards the beak; it is defined by a pair of lines diverging from the ape• and gradually curves over into the lateral slopes. The beak of the pedicle-valve is small, pointed and incurved, and that of the brachial valve is tumid and incurred. Both valves are markedly and ahnost equally convex. The mesial fold is angular, but i~ frequently flattened antel~orly ; the mesial sinus is also angular, and is sharply defined by a pair of angular cost~ originating at the beak. The lateral slopes of each valve are ornamented by about four angular cost~ : those on eiSher side of the median fold are rather more than half as wide as the fold, and the l:emainder are progressively smaller. The cos~ are .~eparated by deep ~ngular furrows. Of the cost~e in the brachial valve, only three on each side of the fold originate at the beak, so that in sinai1 immature specimens each lateral slope has but three cost:e. Median septum in the pedicle-valve aRached to the floor of the valve for about a third of its width, thick at the base, it rapidly tapers and becomes knife-like; near the beak it rises about half-way to the level of the area, and becomes gradually lower anteriorly. Its posterior portion is embedded in a slightly- developed apie.~l callosit "~ Surface of the shell marked by a few concentric lamellm which, like thoae in h~p. wglcotti, are widely spaced, and occur chiefly in the anterior region. The flaking-away of the external layers often leaves the surface of the shell smooth, and is responsible for th~ characteristic appearance of many of the Derbyshire specimens. Q. J. G.S. No. 302. Q Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

216 YR. F. J. ~ORTH O~ [VOI. Ixxvi,

Sowerby's types are preserved in the British Museum (Natural History) and typical specimens are excellently illustrated in Vaughan's ' Belgian paper' (pl. iii). Horizon and localities.--Carboniferous Limestone (K-D); (of the mature form, that is, the holotype, D~D~ ; locality, Derby- shire). In this country the mature form has not been found south of the Midland area, but Vaugban records its occurrence at Yis~. l Remarks.--1)avidson and M'Coy have included in this species several forms which obviously do not belong to it, and as a result, the descriptions given by both those authors are misleading. M'Coy 2 describes it as having five or six cord-like ' ribs' on each lateral slope, a ' midrib' as wide at the margin as three lateral ones, and a surface covered with coarse scale-like lamin~ of growth. He gave no figure, but he was probably referring to. a form like that figured by Davidson [4] pl. vii, fig. 37. In a previous work, however, M'Coy stated that ST. octoplicata had four large angular 'ribs'on each side of the mesial ridge, and a smooth surface. 3 None o[ Davidson's figures adequately depict the typical mature form of this species. Those which approach nearest to it are pl. vii, figs. 38-41, said to be from Sowerby's original examples. This uncertainty as t~) the precise significance of the name ocloTlicata is probably in part to be attributed to the fact that one of Sowerby's figures (fig. 4) differs considerably from the other two. There can, however, be no doubt as to which of the two forms Sowerby himself regarded as the type of his species, for he suggested that the specimen illustrated in fig. 4, which possessed certain special characters, might prove to belong to another species. Of the two specimens figured by Davidson 4 from ' Lough Hill' (Sligo), 5 one, fig. 47, belongs to this species; while the other, figs. 42-44, more closely approaches Spiriferina Terplicata. The desirability of separating this species from the Permian ~p. cristata Schlotheim, has been discussed by I)avidson, King, L. G. de Koninck, and others, but owing to the uncertainty as to the exact significance of either name, the opinions expressed were at variance. As, however, the name cristata was originally applied to a vaguely-defined Permian shell, it seems desirable to regard the two species as distinct, although it is highly probable that from Sp. octoplicata descended some of the Permian Spiriferinas referred to cristata. In the extent of the development of the ventral median septum, Sp. cristata is intermediate between the Carboniferous and the Liassic species.

1 Vaughan [35] p. 45. M'Coy [21] p. 18. 3 M'Coy [20J p. 133. Davidson [4J pl. vii, figs. 42, 44, & 47. These specimensare now in the Museum of Practical Geology (Geological Society Collection). That illustrated in fig. 47 bears the Register No. 5705 A, while on the same ~ablet is ~hat illustrated in figs. 42-44, together with a similar specimen not figured. 6 Prof. G. A. J. Cole suggests that thi~ must be a misprint for Lough Gill. (In Ht~e~s.) Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] ,~'rR, vcorlt }-RIB AND 8PIRIFERINA. 217

Mutations of Spiriferina octoplicata. 1. SPIRIFER*~'X OCTOPLICXTX, mut. D. (P1. XIII, fig. 8.) This mutation is the typical Dibunophyllum-Zone form already described, and represented by Sowerby's type-specimens. 2. SP]RrFERI~'A oc:roPLrCXT~, mut.~, Vaughan. (P1. XIII, fig. 9.) This is a small form possessing all the essential characters of the later mutation, somewhat less highly developed. The area is lower, and the valves less convex than in the typical form, than which, in consequence, the shell has a less inflated appearance. It also differs from the later mutation, in that the eost~e are less angular and are separated by shallower furrows, while the concentric ornament is more regularly developed. Horizon s and loca lities.--Middle Avonian, Horizons y to $, in the South-Western Province. Remarks.--This mutation is the form defined by Vaughan 1 as ST. octoplieata, mut. ~. It differs from S T. i~sculpta in the number and size of its cost~e, which in the latter species are fewer in number and larger in size. Sp. insculpta, moreover, occurs at a higher horizon. It is readily distinguished from the Lower Avonian mutation of Sjo. perTlicata which has five or six costm on each lateral slope. SmRIFERI~A INSCULPTA (Phillips). (Text-figs. 4f& 5a, pp. 199, 211 ; P1. XII[, fig. 11.) 1836. 8pirifera insculpta Phillip.~ [27] p. 216 & pl. ix, figs. 2-3. 1844. Spirifera quinqueloba M'Coy [20] p. 13~ & pl. xxii, fig. 7. The original description is as follows :-- ' Spirifera insculpta. Bolland, Derbyshire. Cardinal area large ; the mesial and two or three lateral folds very bold, acute, and strongly striated across.' Although the holotype of this species, which is now in the British Museum (Natural History) Geological Depal~ment,. Gilbert- son Collection, B 30-~, is fairly well depicted in Phillips's figures, the name has been used in a loose way for small strongly-plicated Spiriferoid shells from Devonian as well as from Carboniferous rocks. Davidson's figures of the original specimen : are somewhat misleading, and convey the impression that the shell is more elongate than is actually the case. The holotype is a large (for the species) mature specimen, and is readily distinguished from adult specimens of ST. octoplicata, although the differences between young individuals of those two species are less obvious. The essential characters of the species are as follows :--Shell spirifeeoid ; moderately-trigh curved area ; ?5~fi;ks small, but w:e]] developed, and incurved in the pedicle-valve, prominent and incurved in the brachial valve. The cardinal exh'emities are rounded. i Vaughan [35] p. 45. 2 Davids0n [4] pl. vii, figs. 52-55. Q2 Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

218 D~. F. J. ~'owra ox [~vol. lxxvi,

In tile braehial valve there are, on each side of the angular fold, three lateral eost,e ; the first of these is nearly as large as tile fold, but the others decrease rapidly in size. In the pediele-valve, in- eluding those bounding the sinus, there are three costm on each h~teral slope. The fold, the sinus, and the eostm are sharply angular, and the latter are separated by deep angular furrows about as wide as the eost~e. The smffaee of both valves, except the cardinal area, is covered by strongly-ilnbricating lamelhe, especially distinct anteriorly. The shell-subslatnee is strongly and eoar~ly punctate. Nothing is known of the internal structures, beyond the fact that there is a niediau septum in the pediele-valve, like that in Spir~rina octoplic,ta. The dimensions of an average specimen are as follows: width along the hinge-line, 14 ram. ; length of bntchial valve, 9 ram. Remarks.~Small immature specimens have a lower area, and are less tumid; but the small nmnber of prominent costal, the deeply-serrated anterior margin, and the strong lamellar ornamentation enable them ta~ be readily identified: although, since the outer lateral c~,sta; ('ommence at some little distance from the beak, there are, in young shells, only two lateral eostm on each side, and these are somewhat less angular than iu the adult. Under some conditions of preservation the extermd shelt-htyers are lost, and the surface appears tq) 1)e smooth; the puneta', arc then very distinct. The immature form is well illustrated by a small st~cimen B 20570, from Gwyn, Llaufair (Auglesey), in the (teologieal Department of the British }Juseum (Natural History), and 1)y solne of Davidsou's figures, as, for instance, ~4J p1. lii, figs. 14 & 15. ,Sp. ~uinqueloba of M'Coy was, in all prohabilitv, a small, but not tyl)ieal, example of this species. h'p. insculpta .may be distinguished from b~p. oetoi~lieata by the small number of the costa~ and the deeper fulTows selxalating them; 1)y the adore uniform distribution of a strong lamellar ornament; an([ by its general shai~e. The hinge-line of i nsculpta is rela: tivelv, shorter fhan that of octolJlicata, so that in the former species the shell tends to 1)e rotund, while in the latter it is dis- tinctlv, elongated. H ol'izons and locallties.--Upl)er Avoniau (especially 1)2) in the Midlands and Yorkshire, as, for example, Parkhill, Wetton; Settle. It is last recorded from the Upper Posidonomya Beds of Loughshinny, where 9it is represented bS a typical specimen figured by Vaughan. 1 None of the Lower Avonian Spirlferin~e from the South-Western Province appear to belong to this species ; but the name has been applied to small highly-plieated Spiriferids occurring in the Devonian rocks/- In these, however, the shell-substance is impunctat(,, and they arc, in all probability, representatives of tlw 1 Vaughan [24] pl. 1, fig. 9. "0 G. F. Whidbo~nae, ' A Monograph of the Devonian Fosuils of the South of England' Pal. Soc. vol. ii (1892 97) pp. 109-10. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2~] ,s'Ynl.VVOTfl~7~I.s"XND SJ~JRI~'~'nZ_V.4. 219 Siluriau genus Delthyris (for instance, D. elevatet); but, whether ~his is a case of homceomorphy, or whether the Silurian and Car- boniferous species are genetically connected by a gradual transition during Devonian times to the punctate condition seen in the (~arb0niferous shells, cannot at present be determined.

.SPIRIFE.R1NX PERPLIC_~'rA, sp. nov. [Text-fig. 5c, p. 2ll; P1. XIII, figs. 7 a-7 c & 10.) D e s c r i p t i o n.--Shell spiriferoid, of medium size, gl~atest width at the hinge-line, cardinal extremities nearly right angles, or slightly produced. Pedicle-valve ~ubpyramidal; cardinal area well defined, about a third as high as wide, flat near the hinge-line, but concave towards the beak, which is small and pointed ; apical angle 120 ~ ; delthyrial angle 40 ~" ; lateral slopes tumid and evenly convex ; sinus angular, about twice as wide as deep. Brachial valve evenly convex, mesial fold angular, and about as wide at the margin as two of the adjacent cost~ and the furrow between them. Each lateral slope in both valves bears about six angular costm, separated by equally angular furrows slightly narrower than the cost~. Surface of both valves, except the urea, crossed by concentric ~ale-like laminm, not evenly distributed over the surface of the .*hell. Shell-subst~u~ce coarsely punctate. l)elthyrial supporting-plates and median septum as in Sp. octoplieata. The dimensions of an average specimen are as follows :--Width along hinge-line, 22 ram. ; length of bmchial valve, 10 mm. ; height ,,f area, 6 ram. Holotype.--A specimen in the collection of Mr. J. W. Jackson, F.G.S., of Manchester (figured in P1. XIII, figs. 7 a-7 c). Horizons and localities.--Carboniferous Limeshme; of the hoh)tyl)e, Upper Dibu:noph;gllum Zone, Peak's Hill. The typical form occurs only in the Dibunoph.~llum Zone. Comparison with Sp. o crop lica ta, etc.--This species differs from Sp. octoplicetta in having a more clearly defined and less curved area, and in the greater number of its cost~e, which are separated by narrower and shallower furrows than in that species. From both T.qlothyris laminosa and P~enctospiriJer scabricosta it differs in the relative uniformity of the size of its cos 't~e and the fold in the brachial valve, and it may be further distifiguished from the former species by its punctate shell-structure. One of the specimens figured by Sowerby as SpS'(fer octoplicata, namely ~32J pl. dlxii, fig. 4, and refigured by Davidson, 1 in all probability belongs to this species, although its fold is of an exceptional character, being divided anteriorly into three by two short furrows. 1 i Davidson [4] pl. vii, fig. 37; pl. lii, fig. 13. The original specimen is distorted, and Davidson's figures are idealized. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

220 DR. F. J. I~'ORTI'[ ON [vol. Ixxvi,

Mutations of STiriferina perTlicata. 1. SPIRIFERINA PERPLICATA, IIlUt. D. This is the felon represented by the holotype of the species (P1. XIII, figs. 7 a-7 c).

2. SPIRIFERINA PERPLIC~kTA, rout. K. (P1. XII[, fig. 10.) 1905. Spiriferina el. oetoplicata Vaughan, ' Dristol Paper' [33] pl. xxvl, fig. 6 ; and ' Burrington Paper' [-'28] p. 384. 1915. Spiriferina oo$oplicata rout. K, Vaughan, 'Belgian Paper' [35] p. 44 & pl. vii, fig. 1. This mutation has all the essential characters of the species, as already enunleratcd, but it differs from the typical (D) form in the following respects :--The cardinal area is lower and the bntchial valve less convex ; the shell is, in consequence, less gibbose. Pedic le-valvc.~The beak is more sharply incurred ; the costa~ on either side of the sinus are rather more conspicuous than the lateral cost~e, so that the sinus is sharply defined ; this character is still noticeable, although less obvious in the 'later mut~Ltion. Brachial valve.--The fold is slightly flattened anteriorly, and the costm are separ:tted by somewhat shallower furrows. Horizons and localities.--In the South-Western Province this mutation occurs in the Cleis/opora Zone and at Horizon ft. in the Avon Section it reaches its acme at the t~q) of K2, but in the Burrington Section it is rare at that horizon: at Mitcheldean it occurs in K. Remarks. --This mutation is the form cited by Vaughan as Spi,r~ferina cf. octoplicata in his earlier (that is, ' Bristol' and 'Burrington ') papers, and as ST. octoTlicat:a lnut. K in his 'Belgian Paper.' In the 'Burrington Paper' (pp. 384,-85), he also compares it to Sp. Teracuta De Koninck, which is al)undant in Z~ of Belgium. In the absence of an examination of the type-specimen of Sp. peracuta, a comparison between that species and the :germ under discussion is not possible ; but, judging from De Koninck's 1 and Vaughan's~ figures of typical specimens, the two are distinct, Sp. perplicata having more costm and a relatively-.larger median fold than the other form. In any case, this does not affect the conclusion arrived at concerning the relationship of Sp. pe~Tlicata rout. K, because, even if it should eventually prove to be identical with De Koninck's species, it would still be the Lower Avonian representative of Sp. perplicala. The two fdllowing forms figured by Porth)ek l)robahly helong to this species :--- (1) The obscure specimen described as Spirifera partita,:~ and now in the Museum of Practical Geology (Reg. No. 28427). Owing to its

1 Ann. Mus. Roy. Hist. Nat. Belg. eel. xiv, pt. 6 (1887) pl. Xxii, figs. 56-61.' ~" Vaughan [35] pl. vii, fig. 2. a j. Portlock, 'Geol. Rep. Londonderry, &c.' 1843, p. 567 & pl. xxxvi'd, fig. 3 ; see also Davidson [4] pl. vii, figs. 60-61. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

l~rt 2] 8YRINGOTHYRI8 AND 8PIRIFERI~VA, 221

imperfect nature and state of preservation, the atrmities of the specimen cannot with certainty be determined. The locality of the figured specimen is given as Kildress (Tyrone). 2. A specimen described as ' near to' oc~op~icata, but having more coshe than that form i : in this case, however, the specimen does not appear to have been preserved.

Additional Note upon Spirliferina. It is, perhaps, scarcely necessary to mention that, although the three Carboniferous species of Spiriferina enumerated can be readily recognized even in their early mutations, as well as in their mature condition, there are no hard-and-fast lines of distinction between them. Regarded as a whole and as members of a genus, they are sharply differentiated from contemporary forms, and are an interesting example of evolution in a ' sheaf of parallel forms.' The young stage (in a phylogenetic sense), with smaU and simple forms, occurs in the Lower Avonian, and the adult stage, in which the full size was reached and the essential characters fully developed, was attained in Dib~nophyllum times. There are other forms ~which represent species additional to those enumerated here, but they are less common, and a discussion of their characters is reserved for a future paper.

VII. SUMMXRY OF CONCLUSIONS. 1. The special characters of Syringothyris were first described by L. G. de Koninck in a Belgian specimen wrongly identified as Spir~fer distans. The name Syringothyris was proposed by A. Winchell, who discovered in a North American Carboniferous shell, characters similar to those already noticed by De Koninek. The existence of the syrinx in S. cusTidata (Martin) was first recorded bv F. B. Meek. 2. The essential characters of the genus are :--its strongly cuspi- date outline, a fold and sinus devoid of plications, an imperforate deltidium, a finely and sparsely punetate shell-substance, and a syrinx-bearing transverse plate between the delthyrial supporting- plates. 3. The satisfactory classification of the forms included in the genus is a matter of some difficulty, owing to the nature of the variations which occur. Variations due to time, to environmental conditions, and to distribution in space are recognized; and distinctive names are given to the mutations characteristic of c@rtain horizons. 4. The syrinx and the transverse plate to which it was attached originated as a modification of an apical callosity ~neh .~t existed in many Spiriferoid shells. It was initiated in Middle Devonian

I j. Portlock, ibid., pl. xxxviii, fig. 8. See, for example, Vaughan [35] p. 3. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

< .... ~.~ m

IU'I !

o "~ ]

. 9 ~ ~

-~. ~~

jr i ~o , ~

r~ ~ .~ Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

li~l~5 2 ~ 8YRI NGOTttY2~I S &ND SPIRI FEBI I~24. 2~.3 times, and reached its acme early in the Carboniferous Period. Our knowledge of the pre-Carboniferous history of the genus is, up to the present, derived entirely from its North American representatives. 5. In all known species of Syrinyothyris the fold in the bruchial valve and. the sinus in the pediele-valve are smooth. Species such as Spirifer distans, in which the ~old and sinus are plicated, do not possess a syrinx, and have-been incorrectly referred to S3tr/~tgo- thyr/s. The genus Syringopleura, founded to include shells with the infernal characfel~ of }~yrinyothyris, accompanied by a plicated fold and sinus, is not valid, since the type-species S. randalli was imperfectly described, and the syntypes inglude two distinct forms : ~f these, those in which the fold and sinus were plieated do not possess the internal characters of Syrinyothyris. 6. The form referred to by M'Coy as Spirifera laminosa, which has also been regarded as belonging to Oyrtia, Spiriferina, and Syrinyothyris, is here assigned to a new genus, T.~lotl~yris, since it has not the puncture shell-structure of STiriferina, nor the internal characters of Syringothyris. The essential characters of Tylotkyris are: a laminose surface-ornament, impunetate shell- substance, a ventral median sepham, and a peculiarly developed apical callosity. The genus is represented in the Lower Avonian by mutations of T. laminosa (M'Coy), and in the Upper Avonian by subspecies of T. subeoniea (Martin). 7. S.~/rin.qothyris and S2iriferina are in no way related, either mol~hologically or phylogenetically. 8. The small Carboniferous shells with a puncture shell-structure, laminose surface-ornament, and a ventral median septum, which have hitherto been referred to STiriferina, include two types characterized by external differences and by slight variations in the jugum. Of these two types, one, in which there are numerous costm and a relatively-large rounded fold and sinus, is relegated to a new genus, Punctospirifer; while the members of the other type, which include shells with a few large angular cos "tin but little smaller than the angular fold and sinus, are, for the present at least, retained in Spir(ferina, although the type-species of that genus is a Liassic form. Both of these types belong to the family Suessiidse which is represented also in Permian and Triassic rocks. The subdivision here suggested for the Carboniferous forms will, it is believed, prove to be applicable to the later species also. The investigation of these forms, and 0f the relation, if any, which exists between them and the Devonian and Silurian species which resemble them in all but the puncture shell-structure, is not yet completed. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

224 D~. ~. J. NORTH ozr [vol. lxzvi,

VIII. BIBLIO6R~I~HY. [Thi* list includes only the more important works i** which the characters or the distribution of species of 8yt'i~.'lothyris and Spiriferi~ are mentioned or di~cussbd. References to other papers aru given in the text.~ 1. T. DAVZVSON.--'On the Genera & Sub.Genera of Brachiopoda that are 'ovided with Spiral Appendages for the Support of the Oral Arms.' ~ eologist, col. i (1~ff~8) pp. 409-16, 457-72. 9. T. DXVXDSON.--' Perforate & Imperfo,~te Brachiopoda' Geol. Mug. 1867, pp. 311-15 & pl. xiv. 3. T. DAv[vso~.--' British Fossil Brachiopoda ' Men. Pal. Sac. col. i (1851-55) Pt. [--' Introduction & Tertiary Species; Pt. Ill--The Oolitic & Line sic Species.' 4. T. DAVIDSON.--' British Fossil Brachiopoda' col. ii (1858-1863): Pt. V--' The Carboniferous Brachiopoda.' 5. T. DaviDso~.--' British Fossil Brachiopoda' col. iv, pt. 3 (1880) 'Supple.- ment to the Carboniferous Brachiopoda.' 8. F. DlxBY & T. F. SIBL~r.--' Ou the Carboniferous Limestone Series of the South-Eastern Margin of the South Wales Coaltield ' Q.J.G.S. col. ixxiii (1917) pp. 111-64. 7. E. E. L. 1)rxos & A. VAU~A~.~' The Carboniferous Succession in Gower (Glamor,ganshire), with Notes ou its Fauna & Conditions of Deposition' Q. J. G.S. col. lxvii (1911) pp. 477-571. 8. J. A. DOUGLAS.--' The Carbonit'erous Limestone of County Clare (Ireland)' Q. J. G. S. vol ixv (1909) pp. 538-86. 9. G. H. GIR~r.--'The Guadalupiall Fauna' U.S. Geol. Sure. Prof. Paper 58 (19o~). 10. G. H. GXRT~.--' The Carboniferous Formation & Faunas of Colorado' U.S. Geol. Sure. Prof. Paper 16 (1903). 11. G. H. GIaTr.--' O,, the Gel,us 8yrin#o~I~r~ Schuchert' Journ. Geol. col. xix (1911) pp. 548-54. 12. G. H. GIRTI-.--'The Fttuna of the Ouray Limestone' 20Lh Ann. Rap. U.S. Geol. Sure. (1900) pp. 31-63. 13. E. J. GARWOOD.--' The Lower Carboniferous Succession in the North-West of England' Q. J. G. S. col. lxviii (1912) pp. 4~,9-586. 14. J. HALL & J. M. CLAR~rE.--' The Natural History of Now York, Palm, ontology ' col. viii, pt. 2 (1894,). 15. E. M. KINDL~.--' The Devonian Fauna of the Ouray Limestone' Bull. 391~, U.S. Geol. Sure. 190~ (S.~ri~gospira, pp. 28-29). 16. W. Kr~o.--'A Monograph of the Permian Fossils of England' Pal. Sac. 1850. 17. W. K~NO.--'A Monograph of Spirifsr euspid~t~s Martin' Ann. & Mag~ Nat. Hist. ser. 4, col. ii (1868) pp. 1-23. 18. L. G. v~ KON[NC~.~' M~mo~ra sur les Genres& les Sous-Genres des Brachi0- podes Munis d'Appeudices Spiraux, ..par T. Davidson, traduit & augment~ de quelques Notes par le Dr. L. G. de Koninck (1859)' M~m. Soc~ Roy. Scl. Liege, col. xvi (1861) pp. 1-51. 19. L. G. vs KONINCK.~' Description des Animaux Fossiles qui se trouvent dana le Terrain Carbonit~re de Belgique' 184,2--44. Li~e. 20. F. M'CoY---' Synopsis of the Carboniferous Fossils of Ireland' 1844. 21. F. M'CoY in Sedgwick & M'Coy : 'Synopsis of the Classification of th# British Pal~ozoic l~cks, with a Systematic Description of the British Pabeozoic Fossils in the Geological Museum of the Unive~mity o~ Carat bridge' 1855. 22. W. MA~T~.--' Account of some Species of Fossil Anomi~e found in Derby; shire' Trans. Linn. Sac. col. iv (179~) pp. 44-50 & pl. iii. 23. W. M ~a~rtN.--'Petrifacta Derbiensia' 1809. Wigan. 24. C. A. MA~L~Y & A. V~tuoH~.~' The Carboniferous Rocks at Loughshi~m~ (County I)nblin), with an Account of the Faunal Succession & Correla- tion' Q. J. G. S. col. lxiv (1908) pp. 413-74,. 25. F. J. NORTH.--' On the Genus 8yrin#othyris Winchell' Geol. Mug. 1913, pp. 393-401. 26. J. PARK~$SO~.--' The Organic Remains of a Former world' col. ii (1811). 27. J. P~L~XPS.--' Illustration,s of the Geology of Yorkshire' pt. 2 (1836). 28. S. H. REY~rOLVS & A. V~u~[AN.--'TI~e Faunal & Lithological Sequence in the Carboniferous Limestone Series (Avonian) of Burring~on Combe (Som~.rset) ' Q. J. G. S. col. lxvii (1911) pp. 342-92. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

part 2] sY~'6or~r~ " AN]} SPII~IFERINA. 225,

29. C. SCHUOnBRT.--' On 8yringothyris Winchell, & its American Species' 9th Anu. Rep. N.Y. State Geologist (1890) pp. 28-37. 30. T. F. ~IBLY.~'OU the Carboniibrous Limestone (Avonian) of the Mendip Area (Somerset), with especial re[erence to the Palmontological Sequence # Q. J. G. S. vol. lxii (1906) pp. 32#-80. 31. T. F. SISLY.~' The Fauual Succession in the Carbonitbrous Limestone (Upper Avonian) of the Midland Area' Q. J. G. S. vol. lxiv (19u8} pp. 34-82. 32. J. & J. DR C. SOW~RBY.~'The Mineral Conchology of Great Britain' 1812-18~5. 33. A. VAu~HA~.--'The Pal~eontological Sequence in the Carboniferous Lime- stone of the Bristol Area' Q. J. G. S. vol. lxi (1905) pp. 181-307. 34. A. VAUGHAN.~'An Account of the Faunal Succession & Correlation of the Carboniferous Rock., at Rush (County Dublin)' Q. J. G. S. vol. lxii (1906) pp. 295-323. 35. A. V^VOHAN.--' Correlation of Dinantian & Avonian' Q.J.G.S. vol. lxxi (1915) pp. 1-52. 36. S. WEL~.ER.--'The Mississippian Brachiopoda of the Mississippi Valley Basin' 9Iilinoi~ G. ol. Surv., Monogr. I (1914) : (Syrin#othyris & Pseudosyrinx). 37. H. S. WILLIAMS.--' Recurrent Tropidolelotus Zones in the Upper Devonian ot New York ' U.S. Geol. Surv. Prof. Paper 79 (1913) pp. 61-62. 38. A. WX~CHELL.--' Description of the Fossils of the Yellow Sandstones lying beneath the Burlitlgton Limestone, at Burlington, Iowa' Prec. Acad. Nat. Sci. Philad. ser. 2, vol. vii (1863) pp. 2-25.

EXPLANATION OF PLATES XI-XIII. l [All figures are of the natural size, unless otherwise indicated.~ PLATE XI. Fig. 1. Syringothyris principal,s, sp. nov. Pedicle-valve. K zone; Trot's Corner, Mendip area. (Coll. of Principal T. F. Sibly, 20474 G l.)S 2. Suringothyris princi~aLis, sp. nov. Pedicle-valve. K 2 subzone ; Howle Hill, Miteheldean district. (Coll. of Principal T. F. Sibly, 20474 G2.) 3. Syringothyr~s cusp~da~a, rout. cyr~orhyncha. (a) Brachial valve; (b) profile. Z l subzone ; Avon Section (south side), Bristol. (Coil, by the writer, now in the National Museum of Wa!es. Reg. No. 19246 G 2.) 4. Syringothyris c~spidata, rout. e~o~eta. (a) Brachial valve; (b) profile. C~ subzone ; Black Rocks, near Sutton (Glamorgan). (Coll. of Prin- . cipal T. F. Sibly, 20474 G 3.) 5. Syringothyris cu~pida~a, mat. e$ole$a. Waulsortian ' knoll '-t~ype. (a) Brachial valve; (b) profile. C zone; Momane (Limerick). (Coll. of the Geological Survey of Ireland. No. 1142. Figured by permission of the Director.) 6. Syringothyris euspidata, mut. cyrtorhyncha. Internal cast of the pedicle-valve, showing the infilling of the syrinx and its slit. Z zone; Mitcheldean Quarry (Gloueestershire). (Coll. of Principal T. F. Sibly, 20474 G 4.) 7. Syringothyris cuspidata, mat. cuspidata, Specimen with asymmetric area (pedicle-valve, looking towards the apex). Upper Avonian; Ell Bolton, Cracoe (Yorkshire). (Coll. by the late Dr. Wheelton Hind.) 8. Sy~ingothyris cuspidata, rout. c~espidata. Peclicle-valve, looking towards the apex. D 2 subzone; Castleton (Derbyshire). (Coll. by the late Dr. Wheelten Hind.)

i The cost of the draw~gs from which these Plates and other illustrations are reproduced has been defrayed by a grant from the Dixon .Fund of the University of London. 2 Principal Sibly has now presented to the National Museum 9of Wales, Cardiff, all his specimens that have been figured or mentioned in this paper, and their re~stration-numbers are given above. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

~26 DR. F. J. NORTH ON [vol. lxxvi,

PLATE XII. I~. 1. 8yringothyris euspidata, rout. c~t~pidata. (a ) Brachial valve ; (b) profile. (The same as P1. XI, fig. 8.) D.~ subzone ; Castleton (Derby- shire). (Coll. by the late Dr. Wheelton Hind.) 2. Sytiagothyri~ cuspidata, rout. cuspidata, distorted specimen. (a) Profile ; (b) braehial valve. D zone ; Little Island (Cork). (S~lg- wick Museum Coll., on Tablet 945.) 3. 8yringothy~.is dongata, sp. nov. The area of the pediele-valve, slightly crushed. Z zone ; Malahide (County Dublin). (Geological Survey of Ireland Coll. Figured by permission of thv Director.) 4. 8yei'ngothyris dongata , sp. nov. (a) The area of the pedicle-valve ; (b) profile; (c) the brachial valve. C zone; Adare (Limerick). (Geological Survey of Ireland Coll. No. 1492. Figured by permission of the Director.) 5. 8b,t~agothy~'is elon.qata, sp. nov. (a) The area of the pedicle-valve; (b) brachial valve. D zone ; Kniveton (Derbyshire). (Coll. by the late Dr. Wheelton Hind.)

PLATE XIH. (No~z.mFip. 1-5 are from exfoliated specimen, and show no trace of the imbricate structure (fig. 16) seen in well-preserved examples of TyZo- ~hy~: such specimens are rare.) l~ig. 1. Ty~othyris ~aminosa, mut. ~. Imperfect pedicle-valve. Z, subzone ; Twyn-yr-odin (Glamorgan). (Coll. of Principal T. F. Sibl~, 20474 G 5.) 2. Tylvthyris lamino.~a, mut. T. Brachial valve. Horizon T ; Alps Quarry, Wenvoe (Glamorgan). (Coll. by the writer, now in the National Museum of Wales. Reg. No. 19246 G 3.) 3. Ty~othy~is s~tbconica cast[etonensis, subsp, nov. Brachial valve. D.2 snbzone ; Castleton (Derbyshire). (Coll. by the late Dr. Wheelton Hind.) 4. Ty~oth!p'is sT~bconica cast~etonensis. (a) The area ; (b) profile. D~ subzone; Castleton (Derbyshire). (Coll. by the late Dr. Wh~elton Hind.) 5. Ty~oth!lris s~tbconica s~bconica (Martin). (a) Braehial valve ; "(b) area ; (c) profile. D,. subzone ; Settle (Yorkshire). (Sedgwiok Museum Coll., Tablet 501.) [Fig. 5 c slightly exaggerates the curvature of the area : it is usually almost fiat.] 6. PunctosF~rifer scabricosta, sp. nov., rout. t~desda~en~s. Brachial valve. D l sabzone; Redesdale (Northumberland). (Coll. of Prof. E. J. Garwood.) 7. Spiriferina perplicata, sp. nov. (a) Brachial valve; (b) profile: (c) pedicle-valve. D, subzone; Treak Cliff, Castleton. (Coll. of Mr. J. W. Jackson, F.G.S., Manchester.) 8. t~piriferina octoplicata, rout. D (J. de C. Sowerby). Braehial valve. Parkhill (Derbyshire). (National Museum of Wales, 19246 G 10.) 9. Spirifarina octoplicata, rout. ~. Portion of brachial valve. Horizon T ; Little Island, Barry (Glamorgan). (Coll. of Principal T. F. Sibly, 20474 G 6.) 10. 8~i~'iferina perp~icata, rout. K. Brachial valve. K 2 subzone ; Walton Castle, Clevedon (Somerset). (Coll. of Mr. J. W. Tntcher.) 11. l~piriferina inscu~pta (Phillips). D~ subzone ; Treak Cliff, Castleton. (Coll. of Mr. J. W. Jackson, F.G.S.) 12. Tylothyris lami/nosa, mat. ~,~ Cast of the interior of the pedicle-vaive. C I subzone (laminbsa dolomite); Collafield, north of Littledcan (Gloucestershire). (Coll. of Principal T. F. Sibly, 20474 G 7.) FNoveraber 30th, 1920.1 Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015

pal~ 2J SYI~INGOTHYRIS" AN]~ ,?PIRIF1~'RINA. 227

Fig. 13. Ty~othyris Laminosa, rout. 7. Interior of the pedicle-valve. Horizon 7 ; near Lavernock (Glamorgan). (Coll. of Principal T. F. Sibly, 20474 G 8.) 14. Spir~ferina waLcottiD'Orbigny. Internal cast of pedicle-valve. Lower Lias ; Radstock (Somerset). (Coll. of the writer, now in the Na- tional Museum of Wales. Reg. No. 19246 G 9.) 15. Ty~othyris subconica subconica. Interior of pedicle-valve. D~ subzone; Waterhouses (Staffordshire). (Coll. of' Principal T. F. Sibly, 20474 G 9.) 16. Ty~o~hy,ris Laminosa (M'Coy). Portion of the surface of a specimen retaining the imbricate structure, enlarged. Z zone; Hook Head (Wexford). (Coil.of Geological Survey, London. Reg. No. 28425.) 17. STirifer p~icatosu~catus, sp. nov. (a) Brachial valve ; (b) the area. D, probably D 8 subzone ; Beeston, near Waterhouses (Staffordshire). (Coll. of Geological Survey; London. Reg. No. 28422.)

DISCUSSION. Dr. F. A. B~!r~Lm~ congratulated the Author on the thorough- ness of the methods used ill his study. He had understood the Author to say that the forms of Syringoth?/ris merged one into the other, and hc enquired whether that statement referred to contemporaneous varieties and species, o1' to successive mutations. Was it possible to observe recapitulation of previous stages in the ontogeny of the later forms ? Principal T. F. SIBt,y welcomed this important contribution to Carboniferous pahcontology, and expressed his api)reciation of the value to zonal stratigraphy of the Author's careful and detailed researches. The AUTHOR stated, in reply to Dr. Bather, that the variations of S:qrin.qoth~./ris in a vertical sense were gradual and continuous, while at any one horizon forms belonging to either of the two species enumerated could be easily distinguished, when they occurred in normal deposits. Owing to the absence of suitable material, a study ,~f recapitulation had not been possibles. In conclusion, the Author expressed his appreciation of the encouraging remarks made by the speakel:s in the Discussion, and thanked those who, by the loan of material, had made the work possible. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015 QUART. JOURN. GEOL. Soc. VOL. LXXVI, PL. XI.

G. M. Woodward, del, Bemrose, Gollo. [~r,b~, SYRINGOTHYRIS. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015 qUART, JOURN. GEOL.SOC. VOL. LXXVI, PL.Xll.

G.M, Woodwardj de.I. Bemrose~ Collo,, DePl~, SYRINGOTHYRIS. Downloaded from http://jgslegacy.lyellcollection.org/ at New Mexico State University on January 26, 2015 qUART, JOURN. GEOL.SOC. VOL. LXXVI, PL.Xll.

G.M, Woodwardj de.I. Bemrose~ Collo,, DePl~, SYRINGOTHYRIS.