6n. Bulletin of the British Museum (Natural History) Miscellanea Geology series Vol 32 No 1 28 June 1979 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique and ever-growing collections of the Museum, both by the scientific staff of the Museum and by specialists from elsewhere who make use of the Museum's resources. Many of the papers are works of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself, available separately, and individually priced. Volumes contain about 300 pages and are not necessarily completed within one calendar year. Subscriptions may be placed for one or more series. Subscriptions vary according to the contents of the Volume and are based on a forecast list of titles. As each Volume nears completion, subscribers are informed of the cost of the next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to: Publications Sales, British Museum (Natural History), Cromwell Road, London SW7 5BD, England World List abbreviation: Bull Br. nat. Hist. (Geol.) Trustees of the British Museum (Natural History), 1979 ISSN 0007-1471 Geology series Vol 32 No 1 pp 1-90 British Museum (Natural History) Cromwell Road London SW7 5BD Issued 28 June 1979 Miscellanea Contents Observations on Cycloclypeus (Cycloclypeus) Carpenter and Cycloclypeus (Kata- cycloclypeus) Tan (Foraminiferida). By C. G. Adams & P. Frame . . .3 The provenance of Sivapithecus africanus. By P. J. Andrews & T. I. Molleson . 19 A silicified brachiopod fauna from the Silurian of Iran. By L. R. M. Cocks. . 25 Two new condylarths (Mammalia) from the early Eocene of southern England. By J. J. Hooker 43 Miocene sharks' teeth from Ecuador. By A. E. Longbottom . . . .57 A new fossil terrestrial isopod with implications for the East African Miocene land form. By S. F. Morris 71 A re-evaluation of the fossil human calvaria from Singa, Sudan. By C. B. Stringer 77 New species of Protorthoptera and Protodonata (Insecta) from the Upper Carboni- ferous of Britain, with a comment on the origin of wings. By P. E. S. Whalley. 85 Observations on Cycloclypeus (Cycloclypem) Carpenter and Cycloclypeus (Katacycloclypeus) Tan (Foraminiferida) C. G. Adams Department of Palaeontology, British Museum (Natural History), Cromwell Road, London SW7 5BD P. Frame c/o Robertson Research (Singapore) Pte. Ltd, 55 Ayer Rajah Road, Singapore 5 Synopsis Cycloclypeus carpenteri Brady and C. guembelianus Brady are redescribed from syntypic material and the former is considered to be a subjective senior synomym of the latter. The type species of Cycloclypeus, C. mammillatus Carter, is discussed and the type sample redated as Late Oligocene. It is concluded that this name should be allowed to lapse and that C. carpenteri should be declared the type species of the genus. C. (Katacycloclypeus) annulatus Martin is redescribed from syntypic material and compared with other katacycloclypeids occurring in Indonesia, and with the largest-known cycloclypeids from the Futuna Limestone of Fiji. It is considered that only one species of Katacycloclypeus can be recognized at present. Introduction The genus Cycloclypeus has long been recognized as having considerable stratigraphical value in the late Cenozoic owing to the relative ease with which the evolution of the early or juvenile part of the shell (embryonic and nepionic chambers: see Fig. 1) can be traced. Tan (1932) published the first detailed account of its evolutionary history and in so doing established it as a prime example of Haeckel's Law of Recapitulation. Although subsequent authors (Cosijn 1938, Drooger 1955 and MacGillavry 1962) have disputed and, indeed, disproved certain of Tan's conclusions, his main thesis - that after its initial appearance, the genus underwent a progressive shortening of the initial coil (so-called nepionic acceleration) through the reduction of the operculinoid and heterosteginoid stages - has never been questioned. The purpose of this paper is to clear up some of the taxonomic confusion surrounding the living representatives of the genus and to describe Katacycloclypeus from the type section of the Futuna Limestone, Vanua Mbalavu, Fiji, where its association with planktonic foraminifera and Lepidocyclina makes it especially important for regional correlation. The early growth stages of Cycloclypeus are illustrated and labelled in Fig. 1, and it is here necessary only to comment briefly on the stolon system and annuli. Stolons are pores which connect the secondary chamberlets through the primary septa. Because the septa are fairly thick, these pores appear as short tubes. As Carpenter (1856) discovered, each secondary chamberlet possesses two stolons on the distal wall, one opening on either side of the proximal end of each secondary septum in the succeeding annulus; Hottinger (1977) has described this arrangement as 'crosswise-oblique', and so far as is known this simple pattern is constant throughout the genus. The term annulus has been used rather loosely in the past and has been a source of some confusion. Annulate cycloclypeids possess one or more rings of raised shell material between the centre of the test and the periphery (see Fig. 12). The annuli are composed of pillars which may be elevated above the test surface to form granules or pustules. When secondary shell material is laid down between the pustules, the annulus forms a thickened ring which may, or may not, show signs of external granulation. Thin sections, however, invariably show the presence of pillars in the wall. The area between two annuli is relatively depressed, as is the region between the inner annulus and the central boss when this is developed. Bull. Br. Mus. not. Hist. (Geol.) 32 (1) : 3-17. Issued 28 June 1979 C. G. ADAMS & P. FRAME Fig. la, b. Early growth stages of megalospheric and microspheric forms of Cycloclypeus to illustrate the terminology used in this paper. Fig. a, megalospheric individual of C. (C.) posteidae Tan from the Futuna Limestone, Vanua Mbalavu, Fiji. The embryonic apparatus comprises the protoconch 'p' and deuterconch 'd'. The nepionic stage, outlined in black, includes an ana-nepionic chamber 'a' (in this specimen divided by a single septum but frequently undivided) and 5 nepionic chambers. The neanic stage is represented by the three rings of annular chambers which are divided into chamberlets by numerous secondary septa. Fig. b, microspheric individual of C. (Katacycloclypeus) annulatus Martin. Topotype; ex Martin Collection, Java. P.36377. See Fig. 9 for the internal structure of C. (Cycloclypeus) carpenteri Brady. The embryonic apparatus is invariably minute (c.20 [xrri) in the microspheric form. It is followed in this specimen by 8 undivided (operculinoid) chambers and at least 17 divided (heterosteginoid) chambers before the first annular chamber appears. As this specimen shows, it can be quite difficult to decide which is the first truly annular chamber. The difference between Cycloclypeus and Katacycloclypeus is that the latter always possesses well-developed anmili in both generations, whereas the former usually lacks them. They may, however, be developed (usually feebly) in microspheric forms of C. carpenteri (Figs 2a, 4a, 5). The megalospheric forms of C. eidae and C. posteidae occasionally possess a single prominent annulus. Acknowledgements We are grateful to Mr Darwin Kadar (Geological Survey of Indonesia) for helpful information, and to Mr L. O'Herne (Rijksmuseum, Leiden) for kindly answering our numerous enquiries, Figs 2a, b, 4a, b. Cycloclypeus carpenteri Brady. Microspheric forms (syntypes) from the Sir Edward Belcher Collection. 51.1.20.42, 43. Figs 2a, 4a, xl. Note presence of incipient annuli towards periphery. Figs 2b, 4b enlargements (c. x 100) of part of shell to show the presence of granules along the primary septa (Fig. 4b) and along and between the septa (Fig. 2b). Fig. 3. Cycloclypeus posteidae Tan, x9. Heavily ornamented form (probably megalospheric) from sample F 181, Futuna Limestone, Vanua Mbalavu, Fiji. For internal details of this species see Fig. la. Figs 5a, b. Cycloclypeus carpenteri Brady. Broken microspheric form from the Sir Edward Belcher Collection, Macclesfield Bank, 30 fathoms ( = 55 m), China Sea. Fig. 5a, x 1 ; Fig. 5b, enlargement of central area, c. x 4. Note development of annuli in the central area and compare with Figs 12-14 and 16. (1893.8.10.1). Figs 6a, b. Cycloclypeus carpenteri Brady. One of three megalospheric forms in the Sir Edward Belcher Collection. This specimen possesses a multi-embryonic apparatus. Fig. 6a, x2-8; Fig. 6b, central area (c. x 7) showing well-developed rings of granules on the primary septa and interstitial calcite largely obscuring the ornament over the central boss. 51.1.20.39. Fig. 7. Cycloclypeus carpenteri Brady, x20. Immature megalospheric form from '210 fathoms ( = 384 m) off Kandavu, Fiji'. This is Brady's (1881, 1884) unfigured syntype of C. guembelianus. Small granules of calcite can be seen on the raised central area. See Fig. 1 1 for details of internal structure. 1959.5.5.311. OBSERVATIONS ON CYCLOCLYPEUS 6a 6 C. G. ADAMS & P. FRAME loaning us comparative material, and allowing us to section one of Dr K. Martin's original rock samples from Java. We also wish to thank Miss Christine Harrison and Mr R. L. Hodgkinson for technical assistance, and Dr J. E. Whittaker for his careful reading of the typescript. Material Unless otherwise stated, all specimens figured or referred to herein are preserved in the British Museum (Natural History), and bear register numbers in the Palaeontological or Zoological collections which, although catalogued separately, are curated together. Systematics Family NUMMULITIDAE de Blainville, 1825 Genus CYCLOCLYPEUS Carpenter, 1856 Subgenus CYCLOCLYPEUS Carpenter, 1856 TYPE SPECIES. Cycloclypeus mammillatus Carter, 1861 a. Cyclodypeus was originally described from material dredged from 'water of a considerable depth off the coast of Borneo'. Carpenter obtained a number of specimens from the Belcher Collection in the British Museum and prepared some thin sections. He stated, somewhat ambiguously, that they were 'no less than 1\" (=57 mm) in diameter', and showed, amongst other things, that they had raised centres, pustulate surfaces and numerous complete or incomplete annuli. There are, in fact, three large specimens (57 mm, 50 mm and 48 mm in diameter respectively) in the original collection. We now know that they are all microspheric individuals. Although Carpenter des- cribed the microspheric generation, his figures were of one monstrosity (1856 : pi. 30, fig. 3; probably megalospheric) and one normal megalospheric individual (pi. 30, fig. 1) with a diameter of about 14 mm and a large proloculus. Unfortunately, no specific name was erected, the first validly described species being C. mammillatus Carter, 186 la, from Takah, south-east Arabia. According to the Catalogue of Foraminifera (Ellis & Messina 1940 et seqq.) it was said to be of Eocene age although Carter himself (1861a : 83) seemed uncertain. Thus, when Brady (1881 : 67) proposed that Carpenter's specimens should be named C. carpenteri, he was already too late to ensure that this became the type species of the genus. Cycloclypeus (C.) mammillatus Carter, 1861 1861 a Cycloclypeus mammillatus Carter : 87. 1861b Cycloclypeus mammillatus Carter : 461. REMARKS. The single unfigured specimen on which this species is based has, unfortunately, been lost. However, the limestone sample whence it came is preserved in this Museum and has now been sectioned. Although Carter believed that he was dealing with an Eocene species, the newly- prepared thin sections reveal that the rock contains Lepidocyclina (Eulepidind) sp., L. (Nephrolepi- dind) sp. indet. (a small primitive form with rhombic equatorial chambers), Heterostegina sp. cf. Heterostegina borneeensis van der Vlerk and Cycloclypeus ex gr. eidae Tan (one specimen only). In the absence of any species of Nummulites, this assemblage clearly indicates a Late Oligocene age (Tei_4 in terms of the East Indies letter classification of the Tertiary; Adams 1970). Since C. mammillatus was described from a single specimen and cannot be redescribed from the original material, we believe that the name should be allowed to fall into disuse, and that the ICZN should declare C. carpenteri to be the type species. An appropriate application has, therefore, been submitted to the Commission. Cycloclypeus (C.) carpenteri Brady, 1881 Figs 2, 4-11 1856 Cycloclypeus Carpenter : 555; pi. 30, figs 1, 3, 4. 1862 Cycloclypeus Carpenter : 292; pi. 19, figs 2-7. OBSERVATIONS ON CYCLOCLYPEUS Fig. 8. C. carpenteri Brady. Megalospheric individual showing a large subdivided ana-nepionic chamber 'a' and 2 nepionic chambers. The external appearance of this specimen was illustrated by Chapman (1900 : pi. 3, fig. 4). 03.2.5.115. Fig. 9. C. carpenteri Brady. Microspheric individual from Funafuti Atoll showing 4 operculinoid and 14 heterosteginoid chambers making up the nepionic stage. 03.2.4.951-1000 (sq. 49). Cf. the microspheric form of C. (K.) annulatus, Fig. Ib. 0-5mm 11 Figs 10, 11. Internal structures of the two complete syntypes of C. guembelianus Brady. Fig. 10 is the specimen illustrated (external view only) by Brady (1884 : pi. Ill, fig. 8 a, b). Both specimens have one undivided ana-nepionic chamber, 'a'. Fig. 10 shows two nepionic chambers and Fig. 11 at least two nepionic chambers, but in this specimen it is difficult to decide at what point the first complete annular chamber begins. Fig. 10, ZF1366; Fig. 11, 1959.5.5.311. 1881 Cycloclypeus guembeliana Brady : 66 1881 Cycloclypeus carpenteri Brady : 67. 1884 Cycloclypeus guembelianus Brady : 751; pi. Ill, fig. 8a, b. 1884 Cycloclypeus carpenteri Brady : 752. 1895 Cycloclypeus carpenteri Brady; Lister : 437, figs 52-54. 1900 Cycloclypeus carpenteri Brady; Chapman : 22; pi. 2, figs 6, 7; pi. 3, figs 1-5. 1927 Cycloclypeus carpenteri Brady; Hofker : 71 ; pi. 24, fig. 1 ; pi. 37, figs 1-10; pi. 38, figs 1-9, 12, 13. 1951 Cycloclypeus guembelianus Brady; Hanzawa : 1 ; pi. 1, figs 1-5; pi. 2, figs 1-7; text-figs 1-10. REMARKS. Early descriptions of this species are confusing because authors have not distinguished between megalospheric and microspheric forms and because Brady's original description of C. guembelianus was inaccurate. The following brief descriptions are based upon Brady's syn- types, supplemented, where necessary, by specimens from Chapman's large collection from Funafuti Atoll. They are intended to distinguish clearly between the A and B generations, and between C. guembelianus and C. carpenteri as originally described. 8 C. G. ADAMS & P. FRAME Megalospheric generation (=A form) of C. carpenteri s.s. Figs 6a, b, 8 MATERIAL. Three syntypes in the Belcher Collection, referred to by Carpenter (1856, 1862). A further 131 specimens, of which 6 have been sectioned, from Funafuti Atoll (Chapman 1900) have also been examined. DESCRIPTION. The three well-grown syntypes (1 sectioned) range in diameter from 13 to 15 mm. All are ornamented by concentric rings of pustules which, except over the central area, arise from the junctions of the secondary and primary septa ; these are broadest over the raised central boss where they are rather difficult to see owing to the presence of secondary shell material. The specimen (Fig. 6a, b) proved on sectioning to be abnormal in that it comprised 3 megalospheric embryonts having proloculus diameters of about 200 um, 250 um and 300 um respectively. There is one other probable megalospheric specimen, one of the two monstrosities mentioned by Carpenter (1856 : 560). This appears to have developed from two megalospheric embryonts which grew their fused tests at right angles. Chapman's large collection (155 specimens) from Funafuti Atoll contains 131 individuals that are indistinguishable externally from Brady's syntypes except insofar as they include a considerable number of small, immature individuals. Test diameter 1-13 mm (8-13 mm if obvious juveniles are excluded). Ornament is as described above. The six sectioned individuals have proloculi ranging from 190 to 280 um in internal diameter followed by one ana-nepionic chamber and 1-3 (or ?4) nepionic chambers. Even the smallest specimens (1-3 mm) are pustulate externally. Megalospheric generation of C. guembelianus s.s. Figs 7, 10, 11 MATERIAL. Two complete and one fragmentary specimens from 210 fathoms (= 388 m) off Kandavu, Fiji. Challenger Collection. DESCRIPTION. The complete specimens have diameters of 1-5 and 2-0 mm respectively. In the smaller of the two specimens the proloculus has an internal diameter of 200 um, and this is followed by a reniform second chamber (diameter 370 um), one ana-nepionic and two nepionic chambers. Five neanic chambers are present. The second specimen has a proloculus diameter of about 190 urn, a second chamber with a width of 360 um, one ana-nepionic and at least two nepionic chambers. At least four annular chambers are present. The early neanic chambers are shorter (radially) than wide. Contrary to Brady's second description (1884 : 751) the surface of the test is not smooth except for 'slightly raised lines' marking the 'annular and radial sutures, but is ornamented by small, irregularly-arranged pustules over the inflated embryonic apparatus. On one specimen the pustules are numerous and heavy. Five other specimens of C. guembelianus in the Heron-Allen & Earland collection (Macassar Straits, 45 fathoms (=82 m)) differ little from Brady's specimens. They range 0-75-3-0 mm in diameter. All have one ana-nepionic and 2-6 nepionic chambers. Proloculus diameter ranges from 120 to 230 um. COMMENTARY. Brady (1884) stated that C. guembelianus was much smaller than C. carpenteri, but failed to realize that he was comparing immature megalospheric individuals with mature micro- spheric specimens. The discrepancy in size is therefore irrelevant. The only other difference noted by Brady was that the former was smooth and the latter highly pustulate. However, as noted above, the syntypes of C. guembelianus are not smooth ; in fact they closely resemble the immature specimens of C. carpenteri from Funafuti. Young individuals of C. carpenteri have very delicate tests in which the large embryonic apparatus projects well above the surface of the equatorial chambers. As growth proceeds, the animal thickens the upper and lower surfaces of the shell with layers of calcite which are laid down around the primary pustules on the central area (Figs 6b, 7). OBSERVATIONS ON CYCLOCLYPEUS 9 Later, the central area becomes smooth and domed, the pustules being arranged in concentric rings over the brim. Microspheric generation (=B form) of C. (C.) carpenteri Figs 2, 4, 5, 9 MATERIAL. Three fully-grown individuals ranging in diameter from 48 to 57 mm and two broken individuals with diameters of 24 and 28 mm. Sir Edward Belcher Collection. DESCRIPTION. In the larger forms the surface is ornamented by 4-6 feebly-developed annuli which may form complete or incomplete rings. These sometimes fade into discrete granules (Figs 2b, 4b). The shell surface is otherwise smooth. The internal structure has been described by Carpenter who did not, however, mention the embryonic or nepionic chambers. A section of a juvenile microspheric individual in the Funafuti collection revealed four ana-nepionic chambers and 14 nepionic chambers following a proloculus with an internal diameter of approximately 20 (am. Hofker (1927) described a B form with a proloculus about 30 urn in diameter followed by a coil comprising 10 ana-nepionic and 14 nepionic chambers. Hanzawa (1951) reported a Recent B form (called C. guembelianus) with a proloculus diameter of 44 urn and a coil with one ana- nepionic and 15 nepionic chambers. He also described a Pleistocene form with a proloculus diameter of 40 um, about three ana-nepionic and 18 nepionic chambers. These appear to be the only microspheric forms for which we have internal details. Chapman's (1900) B form from Funafuti was said to have a proloculus diameter of 140 um but inspection of his illustration (pi. 3, fig. 2) shows that this measurement was incorrect. Not known. Microspheric generation of C. guembelianus REMARKS. Cushman (1921) used C. guembelianus for all Recent cycloclypeids under the mistaken impression that Brady (1884) had referred the larger (microspheric) individuals in the Belcher collection to this species. But the sentence quoted by Cushman (1921 : 386) actually refers to C. carpenteri. Chapman (1900) thought that C. guembelianus merely comprised very young specimens of the A generation of C. carpenteri; he adduced no evidence for this although we believe that he was intuitively correct. Hofker (1927) 'preferred' to use the name C. carpenteri for all living cycloclypeids although he gave no clear reason, and Hanzawa (1951) used C. guembelianus for Recent specimens having 2-5 nepionic septa in the Al and A2 generations regardless of the presence or absence of ornament. CONCLUSIONS. It is clear from the present observations that C. carpenteri and C. guembelianus cannot be differentiated either by size or by ornament, and there remains only the possibility that their internal structures might be different. Table 1 compares the principal internal characters of the two 'species'. As the original megalospheric specimens of C. carpenteri appear to be abnormal in that they possess multi-embryonic chambers, they are not suitable for comparative purposes; Chapman's material from Funafuti is therefore used instead. Table 1 Species Proloculus diameter Deuteroconch diameter No. of ana- nepionic chambers No. of nepionic chambers No. of neanic ch. in 0-75 mm radius C. guembelianus (syntypes) 190 [Jim 200 [xm 360 ^m 370[Jirn 1 1 2 or more 2 2 or 3 5 C. carpenteri Funafuti (6 specimens) 150-230 (Jim 220-450 [xm 1 1-6 2-5 (or ? 6) 10 C. G. ADAMS & P. FRAME It follows that only one living species of Cycloclypeus has been found and that this should be called C. carpenteri. Failure to distinguish an ornate and an inornate living species poses a problem for palaeon- tologists since MacGillavry (1962), working on Oligocene assemblages from Spain, recognized an inornate lineage which he referred to C. guembelianus and an ornate lineage which he called C. carpenteri; Cosijn (1938) used a similar terminology. However, neither author studied Recent populations nor did they examine the type specimens of these species. Subgenus KATACYCLOCLYPEUSTzn, 1932 TYPE SPECIES. Cycloclypeus annulatus Martin, 1880. Tan erected Katacycloclypeus for cycloclypeids which possess one or more annuli in the megalo- spheric generation. As originally defined it included the following species: C. (K.) annulatus Martin (3 annuli), C. (K.) biplicatus Tan (2 annuli), C. (K.) transiens Tan (1 annulus and surface pustules) and C. (K.) posttransiens (1 annulus and no surface pustules). C. martini van der Vlerk (1923) should also have been included since it was described as having at least one annulus in the megalospheric form. Unfortunately, the possession of a single annulus in the megalospheric generation is not absolutely diagnostic of Katacycloclypeus since certain species of Cycloclypeus, e.g. C. posteidae Tan, may, or may not, possess an annulus. Tan believed that he could recognize an evolutionary sequence within Katacycloclypeus and planned to describe it in a later paper, which, unfortunately, never appeared. There is, however, sufficient information in his original publication (1932) for us to be able to deduce the main outline of his evolutionary scheme. He stated (1932 : 119) that the most primitive form of the subgenus was C. (K.) transiens and that this evolved from C. (C.) eidae. From his table V it is apparent that his first record of Katacycloclypeus (K. sp.) was from beds a little younger than those yielding good Tes (— Aquitanian) assemblages including Eulepidina, Spiroclypeus and C. eidae. C. (K.) biplicatus was reported from slightly younger strata along with 'very seldom' specimens of C. annulatus, the latter becoming progressively more numerous until it dominated the katacycloclypeid fauna of supposed Late Vindobonian age. All the strata in which Tan found Katacycloclypeus can now be assigned to Tertiary lower/, and we may, therefore, conclude that Tan's evolutionary sequence was intended to be C. eidae ->• C. (K.) transiensj posttransiens (6-13 heterosteginoid chambers) -> C. (K.) biplicatus (unknown number of heterosteginoid chambers) -> C. (K.) annulatus (c. 3 heterosteginoid chambers in the types ; 2-4 in our additional material from Java). It is evident that forms referable to C. (K.) annulatus appeared fairly early in the sequence and came to dominate the faunas later. We may reasonably infer that C. (K.) posttran- siens followed C. (K.) transiens in the evolutionary sequence but whether it occurred together with C. (K.) biplicatus is uncertain. Van der Vlerk (1923) described C. martini from strata that are clearly of Tertiary lower /age since they contain Flosculinella bontangensis (Rutten). He illustrated four specimens, one of which (pi. 2, fig. 4) he regarded as microspheric, although in the light of later knowledge it is seen to be a megalospheric individual with a rather small proloculus (internal diameter 0-37 mm) and about six heterosteginoid chambers. His diagnosis must therefore be amended to include forms with one to six nepionic chambers, in which case it embraces C. (K.) annulatus. However, the latter is supposed to possess three annuli whereas van der Vlerk's species has only one or two. A small piece of the type sample for C. annulatus, provided by the Rijksmuseum Leiden, has enabled us to section three broken megalospheric specimens and to show that they possess one ana-nepionic and two to four nepionic chambers (Figs 19, 20). It is apparent that without a better knowledge of the internal structure and morphological variation of the seven species and varieties of Katacycloclypeus so far described, it is impossible to distinguish between them. Since C. (K.) annulatus appears at an early stage in the evolutionary history of the subgenus it seems unlikely that the number of annuli is a feature of major taxonomic significance. This being so, it is proper, for the time being at least, to regard Katacycloclypeus OBSERVATIONS ON CYCLOCLYPEUS 1 1 as a single evolving monotypic lineage largely confined to the Middle Miocene, and which, in a period of about 4 Ma, reduced the number of nepionic chambers in the megalospheric coil from 6 to 1 3 (in the most primitive forms) to 0-2 in the most highly evolved forms found in Fiji (see below). Cole (1975) advocated the abandonment of the subgeneric name Katacycloclypeus on three grounds: (1) that the annular inflations were probably ecologically produced growth phenomena, (2) that individuals belonging to the same species could sometimes be assigned to different subgenera and (3) that the use of this subgeneric name could induce a false confidence that any annulate specimens were necessarily of T/ age. He did not, however, demonstrate that the annuli represent resting stages in growth, and the fact that they are well developed only in certain Miocene species suggests that they are not. Furthermore, the irregular development of annuli in microspheric forms, and in particular the tendency for individual annuli to fade out over part of the test, is inconsistent with the resting stage hypothesis. Whether or not Katacycloclypeus is confined to the T/ stage is irrelevant to the biological validity of the name. Only the fact that individuals of the same species can sometimes be assigned to two subgenera is of substantive importance, and Tan was aware of this when he erected Katacycloclypeus. We believe that it will be necessary to examine assemblages rather than isolated individuals in order to determine whether or not this objection is valid, and as yet this has not been done. Cycloclypeus (Katacycloclypeus) annulatus Martin, 1880 Figs 12-15, 17-23 1880 Cycloclypeus annulatus Martin : 157; pi. 28, fig. 1. 1923 Cycloclypeus martini van der Vlerk : 138; pi. 1, figs 1, 2; pi. 2, figs 3, 4. 1932 Cycloclypeus (Katacycloclypeus) annulatus Martin; Tan : 39. 1932 Cycloclypeus (K) annulatus Martin; Caudri : 190; pi. 1, figs 1-3. Not 1934 Cycloclypeus annulatus Martin; Whipple : 142; pi. 19, fig. 1. 1945 Cycloclypeus (K.) annulatus Martin; Cole : 282; pis 19 A, 20G (not H). The types of C. annulatus came from 'Die Kluft des Tji Tarum bei Tjikao, nordlich vom Gunung Parang, West Java'. A small fragment of Martin's original material from the Rijksmuseum Leiden has enabled us to prepare several additional thin sections including three orientated in the equatorial plane (Figs 15, 19, 20). The following descriptions are based on topotypes. Megalospheric generation Figs 15, 19, 20 DESCRIPTION. Test moderately large, ranging in diameter from 4 to 7 mm. Surface ornamented by at least two strongly inflated annuli. Signs of surface granulation visible externally on some speci- mens but not on others, although thin sections show that pillars are always present in the annuli and over the central part of the test. Embryonic apparatus large. Proloculus subovate, thin-walled, and partly embraced by the larger, thicker-walled, reniform second chamber. One ana-nepionic chamber (divided by a single septum in one specimen) is followed by 2-4 nepionic chambers. Early neanic chamberlets shorter (radially) than wide, but lengthening after the third chamber until radially elongate chambers are dominant after the 5th to 7th annular chamber. Stolon system distinct; canal system not seen. DIMENSIONS (based on 3 specimens). Proloculus 228 urn (min) to 335 (im (max) Deuteroconch 350 um (min) to 480 urn (max) REMARKS. As this material is relatively poorly preserved and shows signs of abrasion and breakage, it is possible that well-preserved specimens would show surface pustules and might possess more than two annuli. These specimens differ from the types in that the second chamber less strongly envelops the proloculus. However, as the types of C. (K.) martini v.d. Vlerk (1923 : pi. 1, fig. 2; pi. 2, figs 3, 4) also show considerable differences, this character is obviously highly variable and requires statistical investigation. 12 C. G. ADAMS & P. FRAME Megalospheric forms from Fiji Figs 14, 17, 18, 22, 23 MATERIAL. Numerous specimens from samples F 172, F 173b, F 175 and F 181 (Ladd's layers 8, 7, 5 and just below 1 respectively; see Ladd & Hoffmeister 1945 : 36), collected in stratigraphical order through the type section of the Futuna Limestone, Vanua Mbalavu, Fiji. Brown, iron- stained specimens occur in all four samples; white (unstained) individuals in F 132 and F 181. Most specimens somewhat abraded peripherally and therefore smaller than during life. DESCRIPTION. Test moderately large (4-6 mm in diameter) with one or two prominent annuli; the broad central boss is unornamented, as is the rest of the test including the annuli. Embryonic apparatus very large, the ovate proloculus being strongly embraced by the reniform second chamber. Both chambers are thin- walled. Ana-nepionic chambers absent; nepionic chambers reduced to two or less. When the nepionic chambers are absent (Figs 18, 23), the embryonic apparatus is followed immediately by annular chambers. Early neanic chambers about twice as wide as long near the embryonic apparatus, but after the 4th to 6th chamber becoming slightly longer than wide. Canal system comprises a wide canal in each secondary septum which divides before reaching the primary septa thus producing two parallel smaller canals in the primary septum. Stolon system not observed in any of the sections studied. DIMENSIONS (based on 10 thin sections). Proloculus diameter 0-30 mm - 0-80 mm (usually 0-63 mm) Second chamber 0.59 mm - 1 -25 mm (usually 0-90 mm) Thickness of prolocular wall 10 urn - 20 jim (4 specimens) Thickness of deuteroconch wall 10 um - 40 urn (4 specimens) The smooth external surface and numerous annuli distinguish the megalospheric generation from those of C. eidae and C. posteidae which are heavily ornamented and much smaller (Tan 1932). Stolon system not visible in the material studied; canal system as in C. carpenteri. Microspheric forms from Fiji Figs 12, 13, 21 DESCRIPTION (based on 9 specimens from the type section of the Futuna Limestone). Test thin, up to 90 mm in diameter. Small but distinct umbo surrounded by concentric annuli (usually 6-7 in a radius of 20 mm) which are often coarsely pustulate, particularly in the larger specimens. Annuli often incomplete towards the periphery of the test: remainder of shell surface smooth. Figs 12, 13. Cycloclypeus (Katacycloclypeus) annulatus Martin, xl. Microspheric individuals from samples F 180 and F 177, type section, Futuna Limestone, Vanua Mbalavu, Fiji. Note presence of numerous well-developed annuli in both specimens. These tend to be replaced by discrete granules and to fade out towards the periphery in some specimens (see Fig. 13). P.50299 and P. 50298. Fig. 14. C. (K.) annulatus Martin, x5. Typical broken megalospheric form from Sample F 173b, Vanua Mbalavu, Fiji (cf. Fig. 16). Better-preserved specimens show at least two annuli. Specimen sectioned and registered as P.50466. Fig. 15. C. (K.) annulatus Martin, c. x 35. Thin section showing the embryonic apparatus and nepionic stage. Topotype from Java. Ex Martin Collection, Leiden. Note undivided ana-nepionic and 4-6 nepionic chambers. P.50464. Test damaged and repaired during life. Fig. 16. C. (C.) posteidae Tan, xlO. Sample F 181, Futuna Limestone, Vanua Mbalavu, Fiji. Introduced to illustrate that the megalospheric form of C. posteidae can closely resemble C. (K.) annulatus externally although they are quite different internally. Compare with Fig. 14; see also Fig. la. Figs 17, 18. C. (K.) annulatus Martin, x21 and x20. Thin sections of megalospheric individuals from sample F181, Futuna Limestone, Vanua Mbalavu, Fiji. Neither specimen possesses an ana-nepionic chamber; Fig. 17 (P.50462) shows two nepionic chambers and Fig. 18 (P.50463) no nepionic chambers. See also Figs 22, 23. OBSERVATIONS ON CYCLOCLYPEUS 13 14 C. G. ADAMS & P. FRAME 19 0-5mm Fig. 19, 20. Internal structure of two topotype specimens of C. (Katacycloclypeus) annulatus Martin. Part of Martin's original material from Java. The first specimen (Fig. 19) shows at least three nepionic chambers including one undivided ana-nepionic chamber. Fig. 20 shows five nepionic chambers. The ana-nepionic chamber is divided by a single septum in this specimen. Fig 19, P.50464; Fig. 20, P.50465. Fig. 21. C. (K.) annulatus Martin. Micro- spheric form from the Futuna Limestone, Fiji. The section just misses the proloculus- * but clearly shows that there ^ are at least ... six undivided (operculinoid) chambers .and '•' about 15 divided r.- (heterosteginoid)1 chambers in the nepionic, .stage. P.50298! V v>"> "•'••• -?$? 0-5mm 22 1 mm Figs 22, 23. C. (K.) annulatus Martin. Embryonic apparatus of megalospheric individuals from the Futuna Limestone, Fiji, illustrated in Figs 17, 18. Note absence of nepionic chambers in Figs 18, 22. Figs 17, 23, P.50462; Figs 18, 22, P.50463. The small proloculus and second chamber are followed by a coil comprising at least 8 ana- nepionic chambers and about 16 nepionic chambers. The first nepionic chamber is divided by a single septum. Early neanic chamberlets slightly shorter than wide but becoming longer than wide after the 7th or 8th chamber: towards the margin they are very elongate (2-3 times longer than wide). Stolon system not observed. The primary septa contain two canals, the secondary septa only one. REMARKS. The cycloclypeids from the Futuna Limestone of Fiji are the largest calcareous foraminifera known. Our specimens (up to 90 mm in diameter) are considerably larger than the OBSERVATIONS ON CYCLOCLYPEUS 15 types of C. (K.} annulatus from Java, and even bigger individuals have been observed in the field : Cole (1945) measured specimens up to 100 mm in diameter from the type section, while Ladd & Hoffmeister (1945 : 36) mentioned specimens of up to 6 inches (=152 mm) in diameter. The embryonic apparatus in the megalospheric form is also much larger than in the types of C. (K.) annulatus while the reduction of the ana-nepionic and nepionic chambers provides a further striking difference. However, these characters could merely reflect the fact that the proloculus itself is so much larger, in which case they are not of fundamental importance. Proloculus size is, however, a very poor taxonomic character in foraminifera and should be used only with caution. The most important feature of the Futuna Limestone specimens is that some of them have reduced the nepionic spire completely. No further evolution of this character is possible, and it may be significant that having reached this stage, the species (and subgenus) became extinct. ASSOCIATED FAUNA. Cycloclypeus(C.)posteidaehexaseptarTan, C. cf. C. (C.)posteidaedodekaseptus Tan and C. cf. C. (C.) eidae heneikaseptus Tan ; the last two are remarkable in that they possess very small proluculi - smaller than those of the microspheric generations as described by Tan - and yet in all other characters closely resemble the megalospheric forms. Lepidocyclina rutteni van der Vlerk and L. martini (Schlumberger) are also present. Planktonic foraminifera include more than 25 species, but the association of Globigerina nepenthes Todd picassiana Perconig, Globigerinoides bollii Blow, Gs. subquadratus Bronnimann, Gs. parkerae Bermudez, Globigerinita glutinata (Egger), Gt. glutinata juvenalis (Belli), Globorotalia (Turborotalia) siakensis (LeRoy), Gr. (T.) continuosa Blow and Gr. (Gr.) cultrata (d'Orbigny) limbata (Fornasini) establishes the age as N 13/1 4 (probably N14) in terms of Blow's (1969) zonation. The reduction of the nepionic coil to 0-2 chambers in C. (K.) annulatus supports this late Middle Miocene age. Evolution Tan (1932), in an elaborate and rather difficult paper, described an evolutionary series in Cyclo- clypeus beginning with C. koolhoveni Tan (21-38 nepionic chambers in the A form) in the Oligocene and ending with the Recent C. carpenteri (1-6 nepionic chambers: this paper). His numerous subspecies have not stood the test of time, most specialists preferring to recognize five or six main species. However, Cosijn (1938) and MacGillavry (1962) concluded that C. carpenteri evolved directly from C. eidae in the Miocene, thus eliminating the need to recognize C. posteidae and the C. indopacificus Tan group. They did not, however, study the living representatives of the genus on which this species is based, but assumed that the presence of a short nepionic coil in a few Miocene specimens proved their point. Notwithstanding this, all authors are agreed that the early Oligocene cycloclypeids had a long initial coil comprising some 21-38 chambers in the megalospheric form, and that this was eventually reduced to 2 or 3 nepionic chambers, a process known as nepionic acceleration. The greater part of this shortening had already taken place by Middle Miocene times. The subgenus Katacycloclypeus has been little studied since 1932, presumably because authors thought of it as comprising a different evolutionary lineage, the presence of annuli having been regarded (for no very good reason) as a character of fundamental taxonomic importance. If reduction of the ana-nepionic and nepionic chambers in Cyclodypeus is accepted as an indication of the grade of evolution, then C. (K.) annulatus s.l. is not only the largest, but also the most highly evolved member of the genus, notwithstanding the fact that it is extinct and that more 'primitive' members live on today. If the history of this subgenus has any predictive value, C. carpenteri must now be on the verge of extinction and will disappear when it loses its initial coil. We believe that Tan (1932) and MacGillavry (1962) were justified in regarding species of Katacycloclypeus as representing a separate lineage from those species of Cyclodypeus which led eventually to C. carpenteri. Drooger (1955) studied numerous specimens of Cyclodypeus s.s. from samples of early /age from Borneo, and concluded that the number of nepionic chambers depended on two factors, 16 C. G. ADAMS & P. FRAME nepionic acceleration and the size of the proloculus. He thought that the former acted to reduce the number of heterosteginoid chambers, and the latter to decrease the number of operculinoid chambers in the microspheric generation, and, above a certain size, the number of heterosteginoid chambers in both generations. Conclusions The foregoing evidence permits us to draw the following conclusions. 1. The range of the genus Cycloclypeus is Early or Middle Oligocene to Recent, C. mammillatus Carter having been shown to be from strata of Late Oligocene age and not from the Eocene as originally stated. 2. C. carpenteri Brady is a subjective senior synonym of C. guembelianus Brady, and there is no evidence to suggest that more than one species is alive today. 3. C. carpenteri should be declared the type species of Cycloclypeus, and C. mammillatus Carter, which was described from a single unfigured specimen, should be allowed to fall into disuse. 4. It is not possible to recognize more than one species of Katacycloclypeus at present. So far as can be ascertained, neither surface ornament nor number of annuli are of great taxonomic significance, and it will not be possible to use the embryonic apparatus and nepionic coil satis- factorily until an adequate number of specimens from well-dated horizons throughout its complete range have been collected and investigated statistically. It might then be possible to recognize a primitive group with some 6-13 nepionic chambers (perhaps C. (K.) transiens), an intermediate group with 2-5 nepionic chambers (C. (K.) annulatus) and an advanced group with 0-2 nepionic chambers (C. (K.) martini). This cannot, however, be done at present and it is therefore best to refer them all to C. (K.) annulatus. 5. It is currently convenient to regard Katacycloclypeus as a subgenus of Cycloclypeus, but for biostratigraphical rather than biological reasons. It is not possible to state that the mere posses- sion of annuli is a character of diagnostic importance at generic level although it has some significance at species level, at least in the megalospheric form. 6. All published references to Katacycloclypeus from strata which are not clearly of Tertiary es or/i_2 age should be treated with caution. The subgenus can be recognized only from properly orientated thin sections or from matrix-free megalospheric individuals. In random thin sections of limestone it can easily be confused with Cycloclypeus (Cycloclypeus). 7. The presence or absence of surface ornament (granules or pustules) can only be regarded as significant after vertical sections have been examined, since secondary (organic) calcification often obscures the primary ornament. References Adams, C. G. 1970. A reconsideration of the East Indian letter classification of the Tertiary. Bull. Br. Mus. not. Hist., London, (Geol.) 19 (3) : 85-137. Blow, W. H. 1969. Late middle Eocene to Recent planktonic foraminiferal biostratigraphy. In Bronniman, P. & Renz, H. H. (eds), Proceedings First International Conference on Planktonic Microfossils, Geneva 1967, 1 : 199-422, pis 1-54. Leiden. Brady, H. B. 1881. Notes on some of the Reticularian Rhizopoda of the 'Challenger' Expedition. Part 3. 1. Classification; 2. Further notes on new species; 3. Note on Biloculina mud. Q. Jl microsc. Sci., London, (N.S.) 21: 31-71. 1884. Report on the Foraminifera dredged by H.M.S. Challenger during the years 1873-1876. Rep. sclent. Results Voy. Challenger, London, (Zool.) 9. xxi + 814 pp., 115 pis. Carpenter, W. B. 1856. Researches on the Foraminifera. Part 2. On the genera Orbiculina, Alveolina, Cycloclypeus and Heterostegina. Phil. Trans. R. Soc., London, 146 : 547-569, pis 28-31. , assisted by Parker, W. K. & Jones, T. R. 1862. Introduction to the study of the Foraminifera. xxii + 319 pp., 22 pis. London, Ray Soc. OBSERVATIONS ON CYCLOCLYPEUS 17 Carter, H. J. 1861a. Further observations on the structure of Foraminifera and on the larger fossilised forms of Sind, etc., including a new genus and species. /. Bombay Brch R. Asiat. Soc., 6 : 31-96. Also (1861b) Ann. Mag. nat. Hist., London, (3) 8 : 309-333, 366-382, 446-470, pis 15-17. Caudri, C. M. B. 1932. De Foraminiferen-fauna van eenige Cycloclypeus-houdende Gesteenten van Java. Verh. geol-mijnb. Genoot. Ned., The Hague, (Geol.) 9 : 171-204, pis 1-3. Chapman, F. 1900. On some new and interesting Foraminifera from the Funafuti Atoll, Ellice Islands. /. Linn. Soc., London, (Zool.) 28 : 1-27, pis 1-4. Cole, W. S. 1945. Larger Foraminifera. In Ladd, H. S. & Hoffmeister, J. E. (eds), Geology of Lau, Fiji. Bull. Bernice P. Bishop Mus., Honolulu, 181 : 272-297, pis 12-30. 1975. Concordant age determination by larger and planktonic foraminifera in the Tertiary of the Indo-Pacific region. J. foramin. Res., Washington, 5 (1) : 21-39, pis 1-2. Cosijn, A. J. 1938. Statistical studies on the phylogeny of some Foraminifera. Cycloclypeus and Lepido- cyclina from Spain, Globorotalia from the East-Indies. Leid. geol. Meded., Leiden, 10 (1) : 1-62. Cushman, J. A. 1921. Foraminifera of the Philippine and adjacent seas. Bull. U.S. natn. Mus., Washington, 100 (4) : 1-488, pis 1-100. Drooger, C. W. 1955. Remarks on Cycloclypeus. I. Proc. K. ned. Akad. Wet., Amsterdam, (B) 58 (5) : 415-423. II. Ibid. : 424-433. Ellis, B. F. & Messina, A. 1940 et seqq. Catalogue of Formainifera. Am. Mus. nat. Hist. Spec, publ., New York. Hanzawa, S. 1951. Recent and fossil Cycloclypeus from the Ryukyu Islands and their adjacent seas. Short Pap. Inst. Geol. Paleont. Tohoku Univ., Sendai, 3 : 1-12, pis 1-2, 10 figs. Hofker, J. 1927. The Foraminifera of the Siboga Expedition 1927-30. Part I. Siboga Exped., Leiden, 4 : 1-78, pis 1-38. Hottinger, L. 1977. Distribution of larger Peneroplidae, Borelis and Nummulitidae in the Gulf of Elat, Red Sea. Utrecht micropaleont . Bull., 15 : 35-109, 36 figs. Ladd, H. S. & Hoffmeister, J. E. 1945. Geology of Lau, Fiji. Bull. Bernice P. Bishop Mus., Honolulu, 181 : 1-190, pis 1-10. Lister, J. J. 1895. Contributions to the life history of the Foraminifera. Phil. Trans. R. Soc., London, (B) 186 : 401-446, pis 6-9. MacGillavry, H. J. 1962. Lineages in the genus Cycloclypeus Carpenter (Foraminifera). I. Proc. K. ned. Akad. Wet., Amsterdam, (B) 65 (5) : 429-444. II. Ibid. : 445-458. Martin, K. 1879-80. Die Tertidrschichten auf Java, nach den Entdeckungen von Fr. Junghuhn. ix + 164 + 51+6 pp., 6 tables, 28 pis. Leiden. Tan, S. H. 1932. On the genus Cycloclypeus Carpenter, part I. Wet. Meded. Dienst Mijnb. Ned. Oost-Indie, Weltevreden, 19 : 1-194, pis 1-24, 7 tables. Vlerk, I. M. van der 1923. Een nieuwe Cycloclypeussoort van Oost-Borneo. Samml. geol. Reichsmus. Leiden, (ser. 1) 10 : 137-140, pis 1-2. Whipple, G. L. 1934. Larger Foraminifera from Viti Levu, Fiji. In Ladd, H. S. (ed.), Geology of Viti Levu, Fiji. Bull. Bernice P. Bishop Mus., Honolulu, 119 : 141-153, pis 19-23. The provenance of Sivapithecus africanus P. J. Andrews and T. I. Molleson Subdepartment of Anthropology, British Museum (Natural History), Cromwell Road, London SW7 5BD Synopsis Controversy over the taxonomic affinities of Sivapithecus africanus from Miocene deposits in Kenya has been increased by doubts about its provenance. One specimen is known to be from middle Miocene deposits on Maboko Island, and the only other specimen of 'known' provenance is the type specimen (M 16649), which is reported as coming from Rusinga Island. There is considerable doubt over this, however, so X-ray fluorescence spectrometry (XRF) of the matrix of M 16649 has been attempted to try and establish its site of origin. The XRF spectrum for M 16649 shows great differences from the range of patterns found for the Rusinga Island sites, and it shows the greatest similarity with the spectra from several different levels of the Maboko Island deposits. In some ways, however, the XRF spectrum for M 16649 is unlike any of the samples tested, so while it can be concluded that this specimen probably did not come from Rusinga Island, it is not yet clear where it did come from. Introduction Sivapithecus africanus was first described in 1950 on the basis of a maxilla (registration number M 16649) and two isolated teeth (Clark & Leakey 1950, 1951). These specimens differ morpho- logically from other East African Miocene dryopithecines, although the maxilla was initially assigned (Maclnnes 1943) to a species of Proconsul Hopwood 1933. Clark & Leakey (1950, 1951), in the first major revision of the East African Miocene Hominoidea, redescribed the specimens and recognized their differences from other specimens of Proconsul by assigning them to a new species of the Asian middle to late Miocene genus Sivapithecus. This view gained some measure of support when Simons & Pilbeam (1965) synonymized S. africanus with a previously-described species of the same genus, calling it Dryopithecus (Sivapithecus) sivalensis (Lydekker, 1879). Two years later, Leakey (1967, 1968), who never accepted Simons & Pilbeam's synthesis, referred Sivapithecus africanus to the genus Kenyapithecus, which had as its type species K. wickeri from Fort Ternan, Kenya (Leakey 1962). Pilbeam (1969) and Andrews (1971, 1973) later synonymized Kenyapithecus with Ramapithecus and referred all the 'K. africanus' specimens, including M 16649, to species of Proconsul (or as it then was, Dryopithecus). The situation still remains, however, that these specimens are very distinctive, with at least two autapomorphic characters and only one synapomorphy with later Miocene species of Ramapithecus, Sivapithecus and Dryopithecus, so that on present evidence there is no justification for retaining them either in Proconsul or in Sivapithecus. This uncertainty is increased by doubts over the provenance of two of the three known specimens. One of the isolated teeth attributed to ' Sivapithecus africanus' is the only specimen of confirmed provenance. This is an upper molar (specimen CMH 26 in Clark & Leakey 1951 ; KNM-MB 107 in Andrews 1973) and it is said to be from middle Miocene deposits on Maboko Island, Kenya. The other isolated tooth is from an unknown locality in western Kenya (Clark & Leakey 1951), and the third and only other specimen is the type specimen, M 16649, which is said by Clark & Leakey (1951) to have been recovered from early Miocene deposits at R106 on Rusinga Island, Kenya. In the first description of the specimen, however, Maclnnes (1943) did not identify the locality: the collection as a whole was made in 1932-35 by L. S. B. Leakey's East African Archaeo- logical Expeditions to Rusinga Island and Songhor, but at least one of the specimens, the mandible of a cercopithecoid primate, is almost certainly not from these early Miocene deposits but from Maboko Island, three to four million years later in time. Because the preservation of M 16649 is unlike that of other specimens known to be from Rusinga Island, and in particular R106, we Bull. Br. Mus. not. Hist. (Geol). 32 (1) : 19-23. Issued 28 June 1979 19 20 P. J. ANDREWS & T. I. MOLLESON Table 1 Tube anode current voltage crystal detector path window attenuation Tungsten 30 MA 60 kV LiF 200 flow counter air l-0-oo 22 Chromium 30 MA 60 kV L1F200 flow counter air l-0-oo 22 Table 2 20 scanning speed 2°/min. betw. 15° and 126° chart speed 25 mm/min. full scale deflection 1000 counts/sec. time constant 0-4 or 1-0 sec. Notes on Table 3 Zirconium (Zr) Ka1>2 (22-55°), KfJ, (20-07°), Ka1>2 (2nd order 46-04°), Kalf2 (4th order 102-91°). It is difficult to assess the presence of zirconium by XRF since at most angles the line is masked by strontium, which is usually the more abundant element. By checking for other lines, e.g. Kp^ (20-07°), and by com- paring the strontium Ka1>2 (25-15°) line with the zirconium Ka1>2 (22-55°) plus strontium KP} (22-42°) line, it was possible to deduce the presence or absence of zirconium in the sample. Zirconium was present in 10 of the 14 samples from Rusinga but in only 4 of the 7 samples from Maboko and 1 of 4 samples from Songhor. There was no indication of the presence of zirconium in the Sivapithecus sample. Yttrium (Y) Kalj2 (23-80°). Yttrium is clearly present in the Sivapithecus matrix and in most of the samples from Maboko and Songhor (6 out of 7 and 3 out of 4 samples respectively). On the whole it is not an important element in the samples from Rusinga although it is clearly present in M 15323 from R106. Rubidium (Rb) K2 (26-62°). Rubidium is a more important constituent of most of the Rusinga samples than it is of the Maboko ones. It was not detected in the Sivapithecus matrix. Lead(Pb) L2 (2nd order 91-04°). Zinc was present in all the samples, but the lines were particularly strong in the Sivapithecus spectrum and in the spectra for M 16632 (Maboko) and M 32949 (Mfwangano). Copper (Cu) Ka1>2 (41-80). Copper is not an important constituent of the Sivapithecus matrix nor of the samples from Maboko. In every case it was more important in the samples from Rusinga and Songhor. This was clearly further demonstrated in the chromium tube spectra. Nickel (Ni) K2 (76-94°). The vanadium Ka1>2 line at 76-94° tends to be difficult to distinguish from the titanium K3lj3'line at 77-27°, but it is clearly present in the spectrum of the Sivapithecus matrix (where the titanium content is low) and of all except one of the samples from Maboko. Vanadium was detected in only one of the Rusinga samples (M 32847). Titanium (Ti) Kalj2 (86-14°). Titanium is a minor constituent in the Sivapithecus matrix and in most of the samples from Maboko and Songhor whereas it is an important element in at least half of the samples from Rusinga. Barium (Ba) L• i>. G Ai O O G fp CM CM op ^3 \p i— c 9^^7373 NO NO l~- r(- co vb co C NO T-I/ it rM H G A r4 A Tj-Tj-TfcoOoooo-^ONOOiS-ooT-i ob "3- in r~- A do r-- A NO o do T|- i>- ro •>;}-i>-Tj-irNTj-inoor--inTj-Noincor-- NO co •* -* « r~ Tf rM^orocM co J-H V) VI VI VI VI GO V) VI VI VI f—- *— I rr— I tj-l I w— ( «— t «— ( 1 _^ H~. __J __. __* . Q.) tl) (1) tl> O O O n-4 ^.^ W (U CJ QJ GGGGGGGGCGCGGG GD,aQ,O,aa D, H fl D. CX&D. GGCGr4GAGG'*GCCC GAArjCOG GCGC CrsG T~7373°°T3?73 ^?"73 ^ °P ^ ^ ^-iCGroGAc cMG^'CM OfoO •n-^-irNCJN^rT-T-i CMO^^HCO vpt^rh oorMcoAcco-^- cot^dbt-- AAco ^•incpopocMco inrfcM^^ ONOOO oo — * C CM *— c ^^rOCOfN ^H^HCS oo co in oo to ^f oo r^ ^^ i^- ^* *"H »-H in t^» NO ^f ^^ f^ in CM ON oo in CM ON *^^ *O ^^ t^ <^5 ^O *•"* ^~^ ON co NO NO *••< I-H ^* ^^ oo cs *n ON co f*^ NO in *n «TN| *-H oo i~* t^ CMCM« <— < i— < *— i CM '— !•— — i^"CO^n'— c»— i CM >— i ro CM Tf >— c ^P 73 73 73 73 "° 73 73 73 73 73 73 73 73 ^P ^P °P 73 °P CM <* ^ ^ in 73 ^P "* CMCCCCi-HCCCCCCC^- GA—- CS'^'CCGcMfOcotN'-HCM'— 1<— icM f" O O G CM C i— c cAo-irs (S fS »— ' V>V1V1V1V1V>V>V> VI VI V) V) V> VI VI tUOjfDOCJfUCJO . n) tl) . ty (ij ^ (U (U (l> u-t-5— u-^u-^t- 73 i— i— 73 1- i— 73 73 LH '— 73 73 VH 73 73 73 cccxn,o.ao.aD.o,cD, D.GCXO.GGD. O.CGCX CGG 7373OrPc^vPlT>0PrPl7"73*— '^'"'^'"rP'?sOOPrP '^TS'-P'T" '^'TTS G G CM CM CM ^ ^— c *— ' CM '^c G co co CM CM CM Tt* CM ro O G *n co co ON G oo oo in co P OS • •S 2 22 P. J. ANDREWS & T. I. MOLLESON have examined the matrix from the specimen by qualitative X-ray fluorescence analysis and compared it with matrix samples from a number of Miocene localities in Kenya to try and determine which locality it may be from. Methods and results Matrices taken from 28 East African specimens in the collection of the British Museum (Natural History) were sampled. These represented six East African Miocene sites. We concentrated on material from Rusinga (14 samples), Songhor (4 samples) and Maboko (7 samples) and took single samples from several other sites. X-ray fluorescence spectra were recorded using a Philips PW 1140 generator, PW 1540 gonio- meter and PM 800 flat bed recorder. Samples of the finely powdered matrix of 150-160 mg were weighed out and distributed in an even layer over a Mylar film in a standard Philips sample holder before being placed in the X-ray beam. Instrument and running parameters are listed in Tables 1 and 2. A blank Mylar film was run to determine impurities in the tube. Eight of the samples, including the matrix of M 16649, were run again in the same conditions but using larger samples of powder (500-700 mg) and a chromium tube in place of the tungsten tube. This allowed a check to be made for those elements (tungsten, nickel, vanadium) that are masked by the lines of the tungsten X-ray tube. The spectra derived from the different tubes, however, cannot be compared directly since some elements are energized more readily by one tube than by the other. A rough estimate of the relative abundance of the different elements present was made by measuring the heights of the lines presented by each sample. These heights were all normalized to the tungsten Lyi line at 20=31-66°. The figures obtained, while not calibrated for quantitative determination, were tabulated (Table 3) so that comparisons can be made of the different samples with each other and with the sample from M 16649. The plot for M 16649 was read last to avoid bias in interpretation. Conclusions RUSINGA ISLAND. In general the matrix samples from Rusinga Island have relatively high con- centrations of copper, iron and titanium and contain appreciable amounts of zirconium. Zinc and manganese are not important constituents. Yttrium and barium are often not present and it is unusual for the material to contain lead or vanadium. Rubidium is usually present. MABOKO. The samples from Maboko contain relatively high levels of manganese and vanadium. Zinc, copper, iron, titanium and potassium are less important than they are in samples from Rusinga. Yttrium, lead and barium are usually present whereas rubidium is only present in very minor quantities. SONGHOR. The four samples from Songhor have relatively low concentrations of zirconium, lead, zinc, iron or vanadium. Strontium, titanium and calcium can also be rather low. Neodymium was not detected in any of the four samples nor was potassium clearly present in three of them. Sivapithecus africanus, M 16649. The matrix from this specimen contained relatively high levels of zinc, manganese, vanadium and calcium, but comparatively low levels of copper, iron and potassium. There were clear peaks for yttrium, lead, neodymium and barium. Rubidium was not detected and zirconium is probably not present. The presence of yttrium, barium, neodymium, lead and vanadium and the absence of rubidium make up a pattern that is very different from the Rusinga pattern. In many respects the XRF spectrum of M 16649 is more like those from the Maboko Island samples, but between these also there are differences, for instance in the presence of rubidium and lower levels of lead in the Maboko samples. It is concluded, therefore, that there are very strong reasons for doubting the reports that M 16649 is from R106 on Rusinga Island and that it may be from Maboko Island, but that the data are insufficiently conclusive to show definitely which site it may be from. PROVENANCE OF SIVAPITHECUS 23 The results of this analysis would seem to indicate the value of carrying out further work along these lines. In the first instance, samples of controlled provenance should be collected rather than relying on matrix samples from specimens collected many years ago. A wider variety of localities could be sampled and much larger samples would be available, and in addition to XRF their mineralogical and elemental characteristics could be determined. Acknowledgement We gratefully acknowledge the help of Mr V. K. Din, of the Department of Mineralogy, British Museum (Natural History). Postscript Since writing this report, Dr Martin Pickford has found in the collections of the British Museum (Natural History) the broken fragment of a left P3 from Maboko Island. It is clearly labelled as coming from Maboko Island and has been registered M 36370. It has the same combination of grey enamel and white bone seen on M 16649 and, although no matrix is preserved so that it cannot be compared directly with that specimen, the colour similarity, which is unique to these" two specimens from the Miocene of East Africa, must be considered additional indication that they both originated from Maboko Island. The specimen itself has a robust crown with a broad flattened anterior wear facet. Enamel thickness varies up to 1-1 mm at the break on the buccal face, and although the tip of the crown is broken it appears to have been low-crowned. M 36370 is thus consistent in size and morphology with the other specimens of ' Sivapithecus africanus'. References Andrews, P. J. 1971. Ramapithecus wickeri mandible from Fort Ternan, Kenya. Nature, Land. 231 : 192-194. - 1973. Miocene primates (Pongidae, Hylobatidae)of East Africa, xix + 511 +x pp., 82 figs. Ph.D. thesis, University of Cambridge. Clark, W. E. Le Gros & Leakey, L. S. B. 1950. Diagnoses of East African Miocene Hominoidea. Q. Jl geol. Soc. Lond. 105 : 260-262. 1951. The Miocene Hominoidea of East Africa. Fossil Mammals Afr., London, 1. v + 117 pp., 28 figs, 9 pis. Hopwood, A. T. 1933. Miocene primates from Kenya. Zool. J. Linn. Soc., London, 38 : 437-464, pi. 6. Leakey, L. S. B. 1962. A new Lower Pliocene fossil primate from Kenya. Ann. Mag. not. Hist., London, (13) 4 : 689-696, pi. 18. 1967. An early Miocene member of Hominidae. Nature, Lond. 213 : 155-163. 1968. Lower dentition of Kenyapithecus africanus. Nature, Lond. 217 : 827-830. Lydekker, R. 1879. Further Notices of Siwalik Mammalia. Rec. geol. Surv. India, Calcutta, 12(1): 33-52, plate. Maclnnes, D. G. 1943. Notes on the East African Miocene primates. // E. Africa nat. Hist. Soc., Nairobi, 17: 141-181. Pilbeam, D. R. 1969. Tertiary Pongidae of East Africa: evolutionary relationships and taxonomy. Bull. Peabody Mus. nat. Hist., New Haven, 31 : 1-185, 31 figs. Simons, E. L. & Pilbeam, D. R. 1965. Preliminary revision of the Dryopithecinae (Pongidae, Anthro- poidea). Folia primatol., Basel, 3 : 81-52. A silicified brachiopod fauna from the Silurian of Iran L. R. M. Cocks Department of Palaeontology, British Museum (Natural History), Cromwell Road, London SW7 5BD Contents Synopsis 25 Introduction ............. 25 Age and affinities of the fauna .......... 25 Ecology 27 Systematic Palaeontology ........... 27 Order Orthida ............ 27 Order Strophomenida . . . . . . . . . . -31 Order Pentamerida 35 Order Rhynchonellida 36 Order Spiriferida ............ 3^ Acknowledgements . . . . . . . . . . . .41 References 4j Synopsis A silicified brachiopod fauna has been recovered from a single bed in a limestone unit here named the Qarabil Limestone Formation, at the Robat-e-Qarabil inlier in the eastern Elburz Mountains, Iran. Twenty-three different articulate brachiopods are described, including one new genus, Xerxespirifer (Family Delthyridae), type species X. iranicus sp. nov., and two other new species, Salopina brandi and Pentamerus asiaticus. A probable Wenlock age is deduced. Introduction The Robat-e-Qarabil inlier lies to the north-east of Iran (Fig. 1), in the eastern Elburz Mountains, and was visited by the 1972 Oxford University Expedition to Iran, who made a geological map of the area, and also made extensive fossil collections, chiefly from the Devonian. The stratigraphy of the inlier has been briefly described by Brice et al. (1973 : 179), who described some of the brachiopod faunas, also chiefly from the Devonian; they list a succession, commencing with Horizon 1, 'Marnes calcaires et greseuses a fragments de Brachiopodes et nombreux gros Poly- piers', here termed the Qarabil Limestone Formation. This formation was measured by the Oxford Expedition as consisting of 52 m of limestones and interbedded marls with corals, which overlie the lavas and tuffs of an unnamed and undated formation beneath. The silicified fauna described in the present paper comes from a single bed 33 m above the base of the Qarabil Formation, 1\ km north-east of Robat-e-Qarabil ; the bed is termed 'Lower Group 4' in the unpublished expedition report. Above the Qarabil Limestone Formation there is a stratigraphical break, which is followed by 90 m of thin limestones and marls with occasional sandstones (Horizons 2 and 3 of Brice et al. 1973), succeeded by 200 m of quartzites and shales, followed by a 1300 m sequence of clastic and carbonate rocks of largely Devonian age, many of which are richly fossiliferous (but they are not known to be silicified). Age and affinities of the fauna Many of the 23 different articulate brachiopods described below represent genera with a fairly wide age range within the Silurian ; some also range down into the Ordovician and others up into Bull. Br. Mus. not. Hist. (Geol.) 32 (1) 25-42. Issued 28 June 1979 25 26 L. R. M. COCKS Fig. 1. The geographical position of the Robat-e-Qarabil inlier, Iran. the Devonian, and there is no doubt that the age of the silicified fauna lies somewhere between the late Llandovery and the early Ludlow. The key forms in the more detailed assessment of age are as follows. Glyptorthis is not known from rocks younger than late Wenlock ; Epitomyonia is not known from rocks younger than middle Wenlock; Eoplectodonta is most common in the early Silurian, and is not known from rocks younger than the middle Ludlow ; Shagamella is not known from rocks earlier than late Wenlock; Pentamerus is most common in the Llandovery, and P. asiaticus sp. nov. is most like the pre-late Wenlock forms of the genus ; Plicoplasia is not known from rocks older than the Ludlow; the large numbers of varied spiriferides present suggest an age later rather than earlier in the Silurian. Thus any age assessment would involve an extension of the range of at least some of the genera present, but a Wenlock age seems the most probable for the fauna as a whole, and probably the middle or later part of the Wenlock. The associated fauna consists of fragments of the trilobite Diacalymene, kindly identified by Dr D. J. Siveter, the bivalve Palaeoneilo, kindly identified by Dr N. J. Morris, bryozoans, and the corals listed by Lafuste (in Brice et al. 1973 : 181), none of which contradicts the Wenlock age deduced from the brachiopods. No inarticulate brachiopods or graptolites are at present known from the Qarabil Formation. The affinities of the fauna are puzzling. One of the most common brachiopods, Xerxespirifer iranicus, appears to be endemic, which is unusual in the middle Silurian, and the fauna is strikingly different from the roughly contemporary faunas described by Nikiforova (1937), and Sapelnikov & Rukavishnikova (1975) from the nearby southern parts of the Soviet Union. Diirkoop (1970) has partly described a fauna from central Afghanistan, which he ascribes to the Llandovery chiefly on the basis of an unillustrated Pentamerus, which may be comparable with the Iranian fauna. It is not certain upon which continental plate Robat-e-Qarabil lay during Silurian times. In a recent reconstruction Ziegler et al. (1977) have extended the tentative boundary of Stocklin (1974), IRANIAN SILURIAN BRACHIOPODS 27 which would place Robat-e-Qarabil on the southerly Gondwanan plate. However, this poses problems in that later Devonian and Jurassic faunas found to the north-east of Jarjarm, 50 km to the south-west of Robat-e-Qarabil, have a distinctly European as opposed to a Gondwanan aspect (Dr N. J. Morris, personal communication), which would indicate that the continental suture may have lain to the south, and that Robat-e-Qarabil would have been on the Kazakhstan plate. The question is unresolved, but if the new fauna were eventually found to have come from the Gondwanan plate then it would be of great interest, since no Silurian shelly faunas are known for over 3000 km to the west, or (apart from the possible Afghanistan fauna mentioned above) for a very long way to the east. Perhaps in Silurian times Robat-e-Qarabil lay on a smaller continental unit near the Gondwanan plate, which later drifted northwards. Ecology A total of 53 pairs of articulated valves, 143 disarticulated pedicle valves and 52 disarticulated brachial valves were recovered from the Lower Group 4 sample at Robat-e-Qarabil. However, of all the 23 different brachiopod species represented, only Plicoplasia sp. had a great dispropor- tion of one valve to the other (no articulated specimens, 47 pedicle valves and 2 brachial valves), and, as argued below, this species was probably carried into the Lower Group 4 bed from else- where. When Plicoplasia sp. is therefore discounted, only five brachiopod species are present at over 5% of the total sample: Salopina brandi (26%), Pentamerus asiaticus (12%), Lissatrypa sp. (12%), Eoplectodonta aff. bidecorata (11 %) and Xerxespirifer iranicus (10%); these five species together make up 71 % of the total brachiopod fauna in the bed. The fauna is, however, a rela- tively diverse one, and probably indicates deposition under water that was not too shallow, perhaps comparable with the depth range of the Pentamerus or Stricklandia communities of the Llandovery of the Welsh Borderland (Ziegler, Cocks & Bambach 1968). This deduction is supported by the abundance of the associated coral fauna in Lower Group 4, since corals also reach their ecological acme in the Pentamerus and Stricklandia community depths. The relative abundance of spiriferides in the fauna also suggests a mid-shelf ecological position by comparison with the common mid-shelf spiriferide faunas of the later Silurian and the Upper Palaeozoic. Systematic palaeontology References to the authors of genera and species are given below in the list of references (p. 41), but those to the authors of higher taxa will be found in the Treatise on Invertebrate Paleontology (Williams et al. 1965). Order ORTHIDA Schuchert & Cooper, 1932 Superfamily ORTHACEA Woodward, 1852 Family HESPERORTHIDAE Schuchert & Cooper, 1931 Subfamily HESPERORTHINAE Schuchert & Cooper, 1931 Genus HESPERORTHIS Schuchert & Cooper, 1931 Hesperorthis ? sp. (Figs 4, 5) MATERIAL. Two pedicle valves, BB 93340 (Fig. 5) and BB 93338, and two brachial valves, BB 93336 (Fig. 4) and BB 93339. DIMENSIONS (in mm). length width BB 93340, pedicle valve 18-8 c. 23 BB 93339, brachial valve 13-4 17-3 REMARKS. Although the pedicle valves are poorly preserved, they contain a suggestion of an apical plate in the delthyrium which indicates their attribution to Hesperorthis, in contrast to the open delthyrium of Dolerorthis. Chiang (1972) suggested that Hesperorthis should be confined to forms like the type species H. tricenaria, which possess only simple costae, as opposed to those 28 L. R. M. COCKS like the Iranian form, whose costellae branch at mid-shell length (Fig. 5). These species with branched costellae, such as H. gotlandica and H. martinssoni from the upper Silurian of Gotland (Bassett & Cocks 1974 : pi. 1), have long been classified within Hesperorthis, but the differences observed by Chiang appear to be consistent from species to species, and hence this record of Hesperorthis is qualified by a query. Subfamily DOLERORTHINAE Opik, 1934 Genus DOLERORTHIS Schuchert & Cooper, 1931 Dolerorthis sp. (Figs 2, 3) MATERIAL. One pair of conjoined valves, BB 93341 (Fig. 2), and two pedicle valves, BB 93337 (Fig. 3) and BB 93349. DIMENSIONS (in mm). length width thickness BB 93341, conjoined valves 17-2 19.8 6-4 BB 93349, pedicle valve 18-4 c. 22 REMARKS. Unfortunately the interior cannot be seen on any of the three specimens available, but the shape, ribbing, and open delthyrium and notothyrium are all typical of Dolerorthis, a genus which is well known and relatively common in rocks of Caradoc to early Devonian ages. Subfamily GLYPTORTHINAE Schuchert & Cooper, 1931 Genus GLYPTORTHIS Foerste, 1914 Glyptorthis sp. MATERIAL. Two small pairs of conjoined valves, BB 93391-92, two pedicle valves, BB 93348 and on BB 93405, and three brachial valves, BB 93345-7. DIMENSIONS (in mm). BB 93391, conjoined valves BB 93348, pedicle valve BB 93346, brachial valve length 4-2 5-5 10-1 width 6-0 7-8 c. 12 thickness 2-3 REMARKS. Glyptorthis is rare in rocks of Silurian age, although G. whitei has been described by Bassett (1972) from the late Wenlock of Wales. The Iranian form differs from G. whitei in having relatively larger frilly growth lamellae which are also spaced further apart. In the pedicle valve interior the Welsh form has a raised muscle platform absent from the single Iranian pedicle interior. Family SCENIDHDAE Kozlowski, 1929 Genus SKENIDIOIDES Schuchert & Cooper, 1931 Skenidioides sp. (Fig. 14) MATERIAL. Three silicified pedicle valves, BB 93390 (Fig. 4), BB 93356-57. Figs 2-11. Orthida from the Qarabil Limestone. (2, 3) Dolerorthis sp. ; (2) BB 93341, conjoined valves viewed dorsally, laterally and ventrally, x2; (3) BB 93337, pedicle valve, x3. (4, 5) Hesperorthis ? sp.; (4) BB 93336, brachial valve interior, x2; (5) BB 93340, pedicle valve exterior, x 1 -5. (6-11) Salopina brandi sp. nov.; (6) BB 93333, holotype, conjoined valves viewed dorsally, laterally and ventrally, x4; (7) BB 93334, a fragmentary brachial valve viewed externally and internally, x4; (8) BB 93330, juvenile conjoined valves viewed dorsally and ventrally, x4; (9) BB 93331, pedicle valve viewed externally and internally, x4; (10) BB 93332, interior of a fragmentary pedicle valve, x5; (11) BB 93335, brachial valve exterior, x2. 30 L. R. M. COCKS DIMENSIONS. All three valves have broken edges, but they are estimated to have been approxi- mately 5 mm long and 6 mm wide when complete. REMARKS. There are many described species of Skenidioides from the Ordovician and Silurian (ten from Britain alone, Cocks 1978), and these fragmentary pedicle valves are best left under open nomenclature. Superfamily ENTELETACEA Waagen, 1884 Family SCHIZOPHORHDAE Schuchert, 1929 Subfamily DRABOVHNAE Havlicek, 1950 Genus SALOPINA Boucot, 1960 Salopina brandi sp. nov. (Figs 6-11) DIAGNOSIS. Ventribiconvex Salopina with subcircular outline, weak dorsal sulcus, fine costellae and short, divergent brachiophores. DESCRIPTION. Exterior. Ventribiconvex with subcircular outline. Weak sulcus on brachial valve. Hinge line relatively wider in juvenile individuals (Fig. 8). Small open delthyrium. Small curved apsacline pedicle interarea under incurved umbo, brachial valve interarea very small, with umbo scarcely developed. Ornament of fine costellae, hollow in cross-section. Prominent growth lines at irregular intervals. Pedicle valve interior. Strong pair of teeth, supported by dental lamellae which merge with the valve floor posterior to the anterior end of the teeth. Muscle field weakly impressed and indis- tinct. Valve interior smooth, except at the periphery where the external costellae can be seen. Brachial valve interior. Small, slender linear cardinal process. Strong brachiophores which also act as the anterior part of the socket. However, the brachiophores are relatively short for the genus and diverge from each other at approximately 90°. Muscle field of two pairs of adductor scars, separated centrally by a wide and shallow median ridge; the anterior pair larger than the posterior pair (Fig. 7b). MATERIAL. Holotype BB 93333, conjoined valves (Fig. 6). Additional material: 7 other conjoined valves, including BB 93330 (Fig. 8) and BB 93353; 31 pedicle valves, including BB 93331-2 (Figs 9, 10), some fragmentary; 18 brachial valves, including BB 93334-5 (Figs 7, 11), some fragmentary, all from Lower Group 4, Robat-e-Qarabil inlier, Iran. DIMENSIONS (in mm). length width thickness BB 93333, conjoined valves (holotype) 9-9 10-3 5-9 BB 93330, conjoined valves 4-1 7-1 2-8 BB 93353, conjoined valves 5-8 7-4 3-4 REMARKS. Walmsley et al. (1969) have given a useful review of the ten species which they ascribe to Salopina which had been described up to that time. From their table of specific differences, it can be seen that Salopina brandi falls into the group with relatively fine costellae, and it can be distinguished from other species with fine or medium costellae as follows. From S. lunata (J. de C. Sowerby, 1839), from the Ludlow of Britain, it differs in the shorter pedicle valve muscle field and the short divergent brachiophores; from S. submedia (McLearn, 1924), from the Telychian to Pridoli of North America, it differs in its convex brachial valve and shorter pedicle valve muscle field; from S. conservatrix (McLearn, 1924), from the Telychian to Eltonian of eastern North America and Europe, it differs in the valve outline, and also in the short divergent brachiophores, as opposed to the thin, erect to slightly convergent brachiophores of S. conservatrix; from S. shelvensis Walmsley, Boucott & Harper, 1969, from the Idwian of the Welsh Borderland, in the development of a dorsal sulcus, and once again in the distinctive brachiophores, which in 5". shelvensis are thin and almost parallel. The subsequently-described (Johnson et al. 1976) species S. delta differs from S. brandi in possessing coarser costellae, a relatively wider hinge line, longer dental lamellae, and distinctive supporting plates to the brachiophore in the brachial valve. IRANIAN SILURIAN BRACHIOPODS 31 Family DICAELOSHDAE Cloud, 1948 Genus DICOELOSIA King, 1850 Dicoelosia sp. (Fig. 12) MATERIAL. One pair of conjoined valves, BB 93342 (Fig. 12), length 4-8 mm, width 4-6 mm, thickness 2-4 mm. REMARKS. The single specimen of Dicoelosia found in Lower Group 4 at Robat-e-Qarabil is relatively long, and similar in this respect to D. verneuiliana (Beecher) from the Telychian and Sheinwoodian of Gotland, Sweden (Wright 1968) and D. alticavata (Whittard & Barker) from the Telychian of Britain. It is also very indented at the anterior margin, which distinguishes it from D. biloba (Linnaeus), D. paralata Bassett, D. osloensis Wright, D. oklahomensis Amsden and from other species outside the Silurian. It is also relatively longer than D. parvifrons Johnson, Boucot & Murphy, and with its very concave brachial valve probably represents an undescribed species. Genus EPITOMYONIA Wright, 1968 Epitomyonia aff. clausula Johnson, Boucot & Murphy, 1976 (Fig. 13) MATERIAL. Three pairs of conjoined valves, BB 93344 (Fig. 13), BB 93345 and BB 93354, and one pedicle valve BB 93355. DIMENSIONS (in mm). BB 93344, conjoined valves BB 93345, conjoined valves BB 93354, conjoined valves length 3-4 3-5 2-6 width 3-9 4-3 3-5 thickness 1-9 1-9 14 REMARKS. Epitomyonia was originally described from the Ashgill of Europe (Wright 1968), and is still not recorded from rocks of Llandovery age. However, it is now known by the species E. clausula from the Roberts Mountain Formation (Wenlock) of Nevada, U.S.A. (Johnson et al. 1976), and that species appears very similar to the scarce material from Iran, although the specific determination cannot be confirmed until internal details of the Iranian form are available. Order STROPHOMENIDA Opik, 1934 i Superfamily PLECTAMBONITACEA Jones, 1928 Family SOWERBYELLIDAE Jones, 1928 Genus EOPLECTODONTA Kozlowski, 1929 t Eoplectodonta aff. bidecomta (Barrande, 1879) (Figs 16, 17) MATERIAL. Three pairs of conjoined valves BB 93360, BB 93362, BB 93364; 13 pedicle valves, including BB 93361 (Fig. 17), BB 93363, BB 93365, BB 93367-8; and 13 brachial valves, in- cluding BB 93366 (Fig. 16), BB 93359, BB 93369-70. DIMENSIONS (in mm). BB 93360, conjoined valves BB 93361, pedicle valve BB 93362, pedicle valve BB 93366, brachial valve length 10-4 5-6 7-6 6-9 width c. 15-2 10-3 c. 15-5 12-5 REMARKS. Cocks (1970 : 168) listed the late Silurian species of Eoplectodonta, which consist of E. duvalii (Davidson) from the Wenlock and Eltonian of Britain and Gotland, E. sowerbyana (Barrande) and E. bidecorata (Barrande), both from the Wenlock of Czechoslovakia, and possibly 'Sowerbyella' minuta Kulkov from the late Wenlock of the Altai Highlands, Soviet Union. 32 L. R. M. COCKS '£.' minuta is poorly known, but differs from the Iranian form in its very wide spaces between the primary costae. E. duvalii, E. sowerbyana, E. bidecorata and the Iranian form are all very similar and clearly closely related to each other, except that E. sowerbyana has notably fewer costae than the rest (Havlicek 1967 : 60), and in fact the Iranian form is most similar to E. bidecorata in this respect, hence its tentative attribution. However, the Iranian form diifers from bidecorata in at least two minor points, firstly in the common occurrence of low but prominent anterior plicae, which Havlicek describes as rare in bidecorata, and secondly by the development of a small but persistent median septum in the brachial valve (Fig. 16), consisting of scarcely more than a conjoined row of prominent tubercles. This septum has not been described or illustrated for E. sowerbyana or E. bidecorata by Havlicek (1967), but can be seen in the additional material of sowerbyana illustrated by Cocks (1970 : pi. 13, figs 5, 8) and also in the material of E. duvalii illustrated by Bassett (1974 : pi. 22, fig. 5), and perhaps occurred sporadically in all late Silurian populations of Eoplectodonta. However, its occurrence seems more prevalent in the Iranian form. Superfamily STROPHOMENACEA King, 1846 Family STROPHOMENIDAE King, 1846 Subfamily STROPHOMENINAE King, 1846 Genus KATASTROPHOMENA Cocks, 1968 Katastrophomena sp. MATERIAL. One pedicle valve, BB 93376, length 14 mm approx., width 19 mm approx., and one brachial valve, BB 93377, length 11-1 mm, width 17-8 mm. REMARKS. The interior of neither specimen is visible, but the shape and ornament of these valves are typical of Katastrophomena, which is known from the Ashgill to the early Ludlow of Europe and North America. Subfamily LEPTAENINAE Hall & Clarke, 1893 Genus LEPTAENA Dalman, 1828 Leptaena sp. (Fig. 15) MATERIAL. Two pedicle valves, BB 93371 (Fig. 15) and BB 93372. DIMENSIONS (in mm). length of disc length of trail width BB 93371, pedicle valve 8-3 7-3 c. 14 BB 93372, pedicle valve 7-5 4-1 c. 12 REMARKS. This relatively small species of Leptaena is represented in the Iranian collection from Lower Group 4 by pedicle valves only. It is generally similar to a number of previously-described species, such as L. ziegleri and L. purpurea from the late Llandovery of the Welsh Borderland (Cocks 1968), but the outline of the disc and valve of the Iranian specimens is more rounded than the rather squarer or more quadrilateral outline of the British forms. Figs 12-21. Orthida and Strophomenida from the Qarabil Limestone. (12) Dicoelosia sp., BB 93342, conjoined valves, viewed ventrally, dorsally and laterally, x5. (13) Epitomyonia aff. clausula, BB 93344, conjoined valves viewed laterally and dorsally, x5. (14) Skenidioides sp., BB 93390, pedicle valve interior, x6. (15) Leptaena sp., BB 93371, pedicle valve exterior, interior, and viewed laterally, x3. (16, 17) Eoplectodonta aff. bidecorata; (16) BB 93366, brachial valve interior, x4; (17) BB 93361, pedicle valve exterior, x4-5. (18) Shagamella ? sp., BB 93373, pedicle valve interior, x5. (19) Chilidiopsid gen. et sp. nov., BB 93385, brachial valve interior and exterior, x4. (20) Coolinia cf. pecten, BB 93383, brachial valve interior and exterior, x9. (21) Brachyprion sp., BB 93378, pedicle valve interior, x 2. IRANIAN SILURIAN BRACHIOPODS 33 20a 20b 34 L. R. M. COCKS Family STROPHEODONTIDAE Caster, 1939 Genus BRACHYPRION Shaler, 1865 Brachyprion sp. (Fig. 21) MATERIAL. Two pedicle valves, BB 93378 (Fig. 21) and BB 93379. DIMENSIONS (in mm). length width height BB 93378, pedicle valve 164 c. 22 c. 5 BB 93379, pedicle valve 14-1 c. 19 c. 5 REMARKS. The degree of convexity is not properly conveyed by the photograph (Fig. 21). This species of Brachyprion is very comparable with B. arenaceus from the late Llandovery of the Welsh Borderland (Cocks 1967). Genus LEPTOSTROPHIA Hall & Clarke, 1892 Leptostrophia sp. MATERIAL. Four pedicle valves, BB 93380-2 and on BB 93372. DIMENSIONS (in mm). length ' width BB 93380, pedicle valve c. 10 c. 12 BB 93381, pedicle valve 3-9 c. 6 REMARKS. It is uncertain whether these four rather poorly preserved pedicle valves represent a small species of Leptostrophia, or whether they are juveniles. Their interiors show muscle fields typical of Leptostrophia, with muscle-bounding ridges diverging at approximately 70°. The ribbing style appears to be multicostellate, rather than the unequal parvicostellae more common in Silurian species of Leptostrophia (Cocks 1967). Superfamily DAVIDSONIACEA King, 1850 Family CHILIDIOPSIDAE Boucot, 1959 Genus COOLINIA Bancroft, 1949 Coolinia cf. pecten (Linnaeus, 1758) (Fig. 20) MATERIAL. One brachial valve, BB 93383 (Fig. 20), and two pedicle valves, BB 93384 and on BB 93372. REMARKS. All three specimens are fragmentary, not extending to the anterior or lateral margins, and the largest is preserved to a length of only 6 mm (large specimens of Coolinia from Europe reach lengths of over 40 mm). However, these small shells seem similar to the widespread C. pecten, which occurs abundantly in the Silurian of Sweden, Norway, Britain and Czechoslovakia. Chilidiopsid gen. et sp. nov. (Fig. 19) MATERIAL. Two brachial valves, BB 93385 (Fig. 19) and BB 93386, and a fragment of pedicle valve, BB 93387. DIMENSIONS (in mm). length width BB 93385, brachial valve 5-2 7-1 BB 93386, brachial valve 5-1 c. 8 REMARKS. This species differs from other related and contemporary chilidiopsids, such as species of Coolinia, Fardenia and Morinorhynchus, in possessing an ornament of differentiated parvicost- ellae, not unlike the stropheodontid Leptostrophia. The only other Silurian chilidiopsid with such an ornament is Valdaria, but in that genus the costellae swing round to merge with the hinge-line laterally (Bassett & Cocks 1974 : pi. 5), and the socket ridges are poorly developed in contrast IRANIAN SILURIAN BRACHIOPODS 35 with the prominent very divergent socket plates of the Iranian form. The Robat-e-Qarabil brachial valves are very gently convex, and the single fragment of pedicle valve (which lacks the umbonal region) is rather more convex ; it is not known whether the species is naturally small for the family, or whether the rare material is of juveniles. Unfortunately the small number of specimens to hand is inadequate to erect new generic and specific names. Superfamily CHONETACEA Bronn, 1862 Family ANOPLHDAE Muir-Wood, 1962 Genus SHAGAMELLA Boucot & Harper, 1968 Shagamella ? sp. (Fig. 18) MATERIAL. One pedicle valve, BB 93373 (Fig. 18). DIMENSIONS. Length 3-5 mm, width 4-9 mm, height 1-9 mm. REMARKS. The shape and form of this single valve recall Shagamella, which is known from the late Silurian of Britain and Venezuela (Boucot & Harper 1968). However, no spines can be seen on the valve edge, which is imperfectly preserved. A form approaching this is also seen in the contemporary plectambonitacean Leangella, but the distinctive muscle platform of the latter is not preserved on the Iranian specimen; hence the attribution to the anopliid. Order PENTAMERIDA Schuchert & Cooper, 1931 Superfamily PENTAMERACEA M'Coy, 1844 Family PENTAMERIDAE M'Coy, 1844 Subfamily PENTAMERINAE M'Coy, 1844 Genus PENTAMERUS J. Sowerby, 1813 Pentamems asiaticus sp. nov. (Figs 22-26) DIAGNOSIS. Very weakly trilobate Pentamerus, with relatively small and tightly incurved beak, and with relatively wide apical angles. DESCRIPTION. Biconvex with weak trilobation, apical angle wide for the genus, approximately 90° for the pedicle valve and 110° for brachial valve. Pedicle umbo small and incurved (Fig. 22c), brachial umbo also small with orthocline interarea. Ornament absent, apart from growth lines. Pedicle valve interior has a simple spondylium and supporting septum (Fig. 24), extending to about valve mid-length. A deltidial plate is preserved. The brachial valve interior has discrete outer plates (Fig. 25), extending to just under valve mid-length in adult specimens. MATERIAL. Holotype BB 93388, conjoined valves (Fig. 22). Additional material, one other pair of conjoined valves, BB 93396 (Fig. 26), 5 brachial valves, BB 93397-401, including BB 93399 (Fig. 25), and 16 pedicle valves, BB 93402 (Fig. 23), BB 93403 (Fig. 24) and BB 93404-17. DIMENSIONS (in mm). length width thickness BB 93388, holotype, conjoined valves 47-3 48-2 22-1 BB 93397, brachial valve 61 -4 21 -5 BB 93404, pedicle valve 33-2 c. 37 18-0 REMARKS. The diagnosis separates P. asiaticus from the common Llandovery species P. oblongus, which is more trilobate; from the Wenlock P. esthonus (revised by Bassett 1977), which has a larger pedicle umbo and a relatively smaller cardinal margin ; from P. scalaris, P. subquadratus and P. issensis, all from the Wenlock of the Urals (Sapelnikov 1972), by the small, tightly incurved umbo ; from P. longiseptatus from the Llandovery of Kazakhstan and P. oblongiformis from the Ludlow of Kazakhstan, which both have a larger beak and narrower apical angle (Nikiforova 36 L. R. M. COCKS 1937, Sapelnikov & Rukavishnikova 1975); from P. magianicus from the Lower Silurian of the Gornoi Altai (Menakova 1964) and Tian-Shan, which has several lobations and a narrower apical angle. There are also several nominal species of Pentamems from North America, but these are either generically unrevised, junior synonyms of P. oblongus, or otherwise dissimilar to P. asiaticus. Subfamily CLORINDINAE Rzonsnitskaya, 1956 Genus CLORINDA Barrande, 1879 Clorinda ? sp. (Fig. 27) MATERIAL. Two pedicle valves, BB 93418 (Fig. 27) and BB 93419. DIMENSIONS (in mm). length width thickness BB 93418, pedicle valve 12-1 c. 14 5-0 BB 93419, pedicle valve c. 15 c. 18 4-7 REMARKS. Since no brachial valves are available, the generic attribution must remain uncertain, but the form and structure of these two pedicle valves are consistent with Clorinda or Brevilamula. Order RHYNCHONELLIDA Kuhn, 1949 Superfamily RHYNCHONELLACEA Gray, 1848 Family TRIGONIRHYNCHHDAE Maclaren, 1965 Genus ROSTRICELLULA Ulrich & Cooper, 1942 Rostricellula sp. MATERIAL. Seven articulated specimens, BB 93420-6, one other pedicle valve, BB 93427, and one other brachial valve, BB 93428. DIMENSIONS (in mm). length width thickness BB 93420, conjoined valves 4-6 4-8 1-9 BB 93421, conjoined valves 3-3 5-4 3-2 REMARKS. These small sulcate rhynchonellaceans lack a cardinal process, precluding their inclusion in the two most common Silurian genera, Stegerhynchus and Ferganella. They are attributed here to Rostricellula, which lacks a cardinal process. It is not known whether these small specimens represent a small species, or whether they are juveniles. Order SPIRIFERIDA Waagen, 1883 Superfamily ATRYPACEA Gill, 1871 Family ATRYPIDAE Gill, 1871 Genus ATRYPA Dalman, 1828 Atrypa ? sp. MATERIAL. Two pedicle valves, BB 93429, length 12-8 mm, width 14-4 mm, and BB 93430. Figs 22-31. Pentamerida and Spiriferida from the Qarabil Limestone. (22-26) Pentamerus asiaticus sp. nov.; (22) BB 93388, holotype, conjoined valves viewed ventrally, dorsally and laterally, x 1 ; (23) BB 93402, pedicle valve viewed dorsally, x 1 ; (24) BB 93403, silicified pedicle valve showing the internal structures, x2; (25) BB 93399, silicified brachial. valve interior, x5; (26) BB 93396, conjoined valves viewed ventrally and dorsally, x2. (27) Clorinda sp., BB93418, pedicle valve exterior, x2. (28-31) Lissatrypa sp.; (28) BB 93431, conjoined valves viewed ventrally, dorsally, laterally and anteriorly, x 3 ; (29) BB 93447, pedicle valve interior, x 3 ; (30) BB 93432, broken conjoined valves viewed from the posterior showing spiralia, x 3 ; (31) BB 93443, broken conjoined valves viewed dorsally and laterally, showing spiralia and brachiophores, x3. IRANIAN SILURIAN BRACHIOPODS 37 31b 38 L. R. M. COCKS REMARKS. Although certainly atrypid, the material is not sufficient for a positive generic attribu- tion especially since the interiors are obscured by matrix. The growth lines are pronounced, even verging on frilly, over some of the costae. Family LISSATRYPIDAE Twenhofel, 1914 Genus LISSATRYPA Twenhofel, 1914 Lissatrypa sp. (Figs 28-31) MATERIAL. Sixteen articulated individuals, BB 93431 (Fig. 28), BB 93432 (Fig. 30), BB 93433 (Fig. 31) and BB 93434^6, and one pedicle valve, BB 93447 (Fig. 29). DIMENSIONS (in mm). length width thickness BB 93431, conjoined valves 9-8 9-3 6-3 BB 93439, conjoined valves 13-2 14-7 7-9 REMARKS. The internal structures have been delicately silicified in several of the specimens (Figs 30, 31), and the characteristic spiralia and brachiophores of Lissatrypa are clearly visible as set out in Copper (1973). Dr P. Copper has also seen the specimens and kindly confirmed the identi- fication. However, only the two measured specimens are completely clear of matrix, and although the shape and form of BB 93439 is similar to Lissatrypa atheroidea, the type species from Anticosti Island, Canada (Twenhofel 1914 : pi. 1), the smaller specimen (Fig. 28) is proportionately thicker and less sulcate, which precludes specific identification of the Iranian form at present. Lissatrypa ? sulcata from Gotland differs in possessing a distinctive narrow median depression in both valves (Bassett & Cocks 1974 : pi. 9). Superfamily SPIRIFERACEA King, 1846 Family CYRTIIDAE Fredericks, 1919 Genus EOSPIRIFER Schuchert, 1913 Eospirifer sp. (Figs 32-33) MATERIAL. Two pedicle valves, BB 93448 (Fig. 32) and BB 93449, and three brachial valves, BB 93450 (Fig. 33) and BB 93451-52. DIMENSIONS (in mm). length width thickness BB 93448, pedicle valve 24-7 29-9 9-9 BB 93451, brachial valve 21-1 30-7 8-7 REMARKS. These finely-ribbed Eospirifer specimens have a more pronounced sulcus than the type species, E. radiatus, which is known from the late Llandovery to the Ludlow of Europe, North America and Asia. There is no hint of lateral ribbing on any of the material available. Figs 32-41. Spiriferida from the Qarabil Limestone. (32, 33) Eospirifer sp. ; (32) BB 93448, pedicle valve viewed ventrally and laterally, x 1 ; (33) BB 93450, broken brachial valve showing the exterior and the interior, x 3. (34-39) Xerxespirifer iranicus gen. et sp. nov. ; (34) BB 93453, holotype, con- joined valves viewed dorsally, laterally and ventrally, x3; (35) BB 93454, conjoined valves, x3 - (35a) exterior viewed obliquely from anteroventrally, and (35b) interior with pedicle valve below, showing the distinctive muscle field, and a fragment of the brachial valve above; (36) BB 93464, pedicle valve exterior, x2; (37) BB 93463, brachial valve exterior, x2; (38) BB 93455, internal view of conjoined valves with brachial valve below and fragment of pedicle valve above, x 3 ; (39) BB 93462, brachial valve interior, with a small fragment of the pedicle valve attached to the posterior left of the hinge line, x 4-5. (40, 41 ) Plicoplasia sp. ; (40) BB 93469 exterior, interior and lateral views of pedicle valve, x5; (41) BB 93470, interior view of pedicle valve, x4-5. IRANIAN SILURIAN BRACHIOPODS 39 r*^?P^^ ' 32a 32b 33a 33b 34a 34b 34c 40a 40b 40c 40 L. R. M. COCKS Family AMBOCOELIIDAE George, 1931 Genus PLICOPLASIA Boucot, 1959 Plicoplasia sp. (Figs 40-41) MATERIAL. Forty-seven pedicle valves, including BB 93469 (Fig. 40), BB 93470 (Fig. 41) and BB 93471-98, many fragmentary, and two fragmentary brachial valves, BB 93499-500. DIMENSIONS (in mm). length width thickness BB 93472, pedicle valve 6-5 7-4 3-3 BB 93475, pedicle valve 6-3 7-7 3-4 REMARKS. This chiefly Devonian genus has only recently been described from beds of Silurian age (Johnson et al 1976 : 94), namely from the Ludlow of Nevada, U.S.A., the Wenlock of Arctic Canada and the Ludlow of Australia. The great numerical disproportion of pedicle over brachial valves in the Iranian sample suggests that these shells have been current-sorted and swept into the area of deposition from elsewhere; this is supported by the fact that nearly all the shells are broken anteriorly and laterally, where the valves are weaker. Family DELTHYRIDAE Waagen, 1883 Subfamily ACROSPIRIFERINAE Termier & Termier, 1949 Genus XERXESP1R1FER nov. TYPE SPECIES. Xerxespirifer imnicus sp. nov. (see below). DIAGNOSIS. Delthyrid like Howellella but with pinched-in pedicle muscle field, and with a median rib in the sulcus of the pedicle valve and a corresponding hollow in the brachial valve fold. REMARKS. In its characteristic median rib Xerxespirifer differs from Howellella in the same way as the kozlowskielline Holcospirifer differs from its relative Boucotinskia, and also as the eospiri- ferid Nikiforovaena differs from Striispirifer (for discussion see Bassett et al. 1976 : 620). The narrow width of the muscle field in the new genus is also a point of distinction from Howellella, as is also the more incurved beak. The presence of a small cardinal process is unusual in the Spiriferida. Xerxespirifer iranicus sp. nov. (Figs 34-39) DESCRIPTION. Exterior. Ventribiconvex, with valves elliptical in outline, maximum width anterior to hinge line. Pedicle valve beak incurved, with interarea small, curved and apsacline, small open delthyrium, housing a presumably functional pedicle posteriorly. Brachial valve with very small interarea and almost imperceptible umbo. Radial ornament plicate, with ribs rounded in cross-section, with from three to six ribs on each flank; a small but distinctive median rib on the pedicle valve sulcus and a complementary trough on the fold in the brachial valve. Concentric ornament of fine growth lines. Pedicle valve interior. Teeth small, but curved dorsoposteriorly (Fig. 35b), supported by small dental plates posteriorly only. There is no median septum. Muscle field well-impressed and pear-shaped, but narrow for the family, and extending less than half the valve length (Fig. 35b). Brachial valve interior. Cardinal process small but distinct (Fig. 39). Sockets rounded, bounded anteriorly by curved socket plates which coalesce laterally with the hinge line. There is no median septum. Brachiophore bases short and stubby, protruding ventrolaterally at about 80° to each other just anterior to the cardinal process, but separated from each other by a small cavity. The form of the spiralia is unknown. Adductor muscle scars obscure, but perhaps represented by a pair of suboval areas at about valve mid-length (seen broken off in Fig. 38). IRANIAN SILURIAN BRACHIOPODS 41 MATERIAL. Ten conjoined valves, including holotype BB 93453 (Fig. 34), BB 93454 (Fig. 35), BB 93455 (Fig. 38), BB 93456-61 and BB 93462 (Fig. 39), one other brachial valve, BB 93463 (Fig. 37), and 5 other pedicle valves, BB 93464 (Fig. 36) and BB 93465-8. DIMENSIONS (in mm). length width thickness BB 93453, conjoined valves, holotype 9-3 11-1 7-6 BB 93456, conjoined valves 13-2 16-1 10-8 BB 93459, conjoined valves c. 12 13-7 c. 9 REMARKS. The Iranian material described here as X. iranicus is the only material so far known which can be ascribed to Xerxespirifer. Whether or not the new genus is truly endemic to Iran is not yet known. X. iranicus is almost certainly the same form as that identified as Platystrophia sp. by Brice et al. (1973 : 183; pi. 21, fig. 13), but whose internal structures were then unknown. Acknowledgements I am most grateful to the members of the 1972 Oxford University expedition to Iran, especially R. Brand, J. Hollingdale and R. N. Mortimore, for presenting this material which they collected, as well as the supporting documentation. I also thank A. J. Boucot, M. G. Bassett and P. Copper for discussion, C. H. C. Brunton for reading the manuscript and P. C. Ensom for technical assistance. References Bancroft, B. B. 1949. Welsh Valentian brachiopods and the Strophomena antiquata group of fossil brachio- pods. (Ed. A. Lament.) 16 pp., 3 pis. Mexborough, privately printed. Barrande, J. 1879. Systeme silurien du centre de la Boheme. lere par tie: Recherches paleontologiques. 5, Classe des Mollusques. Ordre des Brachiopodes. 226 pp., 153 pis. Prague & Paris. Bassett, M. G. 1972-77. The articulate brachiopods from the Wenlock Series of the Welsh Borderland and South Wales. 2 : 27-78, pis 4-17 (1972); 3 : 79-122, pis 18-32 (1974); 4 : 123-176, pis 33-47 (1977). Palaeontogr. Soc. (Monogr.), London. & Cocks, L. R. M. 1974. A review of Silurian brachiopods from Gotland. Fossils Strata, Oslo, 3: 1-56, pis 1-11. & Holland, C. H. 1976. The affinities of two endemic Silurian brachiopods from the Dingle Peninsula, Ireland. Palaeontology, London, 19 : 615-625, pis 93-95. Boucot, A. J. 1959. Early Devonian Ambocoeliinae (Brachiopoda). /. Paleont., Menasha, 33 : 16-24, pis 1,2. 1960. In Boucot, A. J., Martinsson, A., Thorsteinsson, R., Walliser, O. H., Whittington, H. B. & Yochelson, E. A late Silurian fauna from the Sutherland River Formation, Devon Island, Canadian Arctic Archipelago. Bull. geol. Surv. Can., Ottawa, 65 : 1-51, pis 1-10. & Harper, C. W. 1968. Silurian to Lower Middle Devonian Chonetacea. /. Paleont., Tulsa, 42 : 143-176, pis 27-30. Brice, D., Lafuste, J., Lapparent, A. F. de, Pillet, J. & Yassini, I. 1973. Etude de deux gisements paleozoi'ques (Silurien et Devonien) de 1'Elbourz oriental (Iran). Annls Soc. geol. N., Lille, 93 : 177-218, pis 20-25. Chiang, K. K. 1972. Hesperorthis and two new Silurian species. /. Paleont., Menasha, 46 : 353-359, pi. 1. Cocks, L. R. M. 1967. Llandovery stropheodontids from the Welsh Borderland. Palaeontology, London, 10 : 245-265, pis 37-39. 1968. Some strophomenacean brachiopods from the British Lower Silurian. Bull. Br. Mus. not. Hist., London, (Geol.) 15 : 283-324, pis 1-14. 1970. Silurian brachiopods of the superfamily Plectambonitacea. Bull. Br. Mus. nat. Hist., London, (Geol.) 19: 139-203, pis 1-17. 1978. A review of British Lower Palaeozoic brachiopods, including a synoptic revision of Davidson's Monograph. Palaeontogr. Soc. (Monogr.), London. 256 pp. Copper, P. 1973. The type species ofLissatrypa (Silurian Brachiopoda). /. Paleont., Menasha, 47 : 70-76, pis 1, 2. Dalman, J. W. 1828. Uppstallning och Beskrivning af de i Sverige funne Terebratuliter. K. svenska VetenskAkad. Handl, Uppsala & Stockholm, 1827 : 85-155, pis 1-6. 42 L. R. M. COCKS Diirkoop, A. 1970. Brachiopoden aus dem Silur, Devon und Karbon in Afghanistan. Palaeontographica Stuttgart, A 134 : 153-225, pis 14-19. Foerste, A. F. 1914. Notes on the Lorraine Faunas of New York and the Province of Quebec. Bull. scient. Labs Denison Univ., Granville, Ohio, 17 : 247-340, pis 1-5. Hall, J. & Clarke, J. M. 1892. An introduction to the study of the genera of Palaeozoic Brachiopoda, Part I. Nat. Hist. N. Y., Albany, Palaeont. 8 (1). xvi + 367 pp., 41 pis. Havlicek, V. 1967. Brachiopoda of the suborder Strophomenidina in Czechoslovakia. Rozpr. ustfed. Ust. geol., Prague, 33 : 1-235, pis 1-52. Johnson, J. G., Boucot, A. J. & Murphy, M. A. 1976. Wenlockian and Ludlovian age brachiopods from the Roberts Mountain Formation of Central Nevada. Univ. Calif. Publs geol. Sci., Berkeley, 115 : 1-102, pis 1-55. King, W. 1850. A monograph of the Permian fossils of England. Palaeontogr. Soc. (Monogr.), London. xxxvii+258 pp., 29 pis. Ko/lowski, R. 1929. Les brachiopodes gothlandiens de la Podolie Polonaise. Palaeont. pol., Warsaw, 1 : 1-254, pis 1-12. Linnaeus, C. 1758. Systema Naturae .... 10th edition, 1. 824 pp. Stockholm. McLearn, F. H. 1924. Palaeontology of the Silurian Rocks of Arisaig, Nova Scotia. Mem. geol. Surv. Brch Canada, Ottawa, 137 : 1-180, pis 1-30. Menakova, G. N. 1964. [Brachiopods from the Lower Silurian deposits of the Zeravshano-Gissarsk mountain region.] Trudy Upravl. Geol. Sov. Minist. tadzhik. SSR, Moscow. 51 pp., 10 pis. [In Russian.] Nikiforova, O. I. 1937. Upper Silurian Brachiopoda of the central Asiatic part of the U.S.S.R. Monografii Paleont. SSSR, Leningrad, 35 (1) : 1-94, pis 1-14. [In Russian, English summary]. Sapelnikov, V. P. 1972. [Silurian Pentameracea from the eastern slopes of the central and northern Urals.] Trudy Inst. Geol. Geokhim. Akad. Nauk SSSR ural. nauch. Tsentr, Moscow, 98 : 1-295, pis 1-57. [In Russian.] - & Rukavishnikova, T. B. 1975. [Upper Ordovician, Silurian and lower Devonian pentamerids of Kazakhstan.] 227 pp., 43 pis. Moscow, Inst. Geol. Geokhim. Akad. Nauk SSR ural. nauch. Tsentr. [In Russian.] Schuchert, C. 1913. Brachiopoda. In Zittel, K. A. von (transl. & ed. Eastman, C. R.), Text-book of Palaeontology I (2nd ed.) : 355-420. London. - & Cooper, G. A. 1931. Synopsis of the brachiopod genera of the suborders Orthoidea and Penta- meroidea, with notes on the Telotremata. Am. J. Sci., New Haven, (5) 22 : 241-251. Shaler, N. S. 1865. List of the Brachiopoda from the Island of Anticosti sent by the Museum of Com- parative Zoology to different institutions in exchange for other specimens, with annotations. Bull. Mus. comp. Zool. Harv., Cambridge, Mass., 1 : 61-70. Sowerby, J. 1813. In Sowerby, J. & Sowerby, J. de C. 1812-46. The Mineral Conchology of Great Britain. 1 vols. London. Sowerby, J. de C. 1839. In Murchison, R. L, The Silurian System, xxxii+768 pp. London. Stocklin, J. 1974. Possible ancient continental margins in Iran. In Burk, C. A. & Drake, C. L. (eds), The geology of continental margins. 1009 pp. Berlin, Heidelberg and New York. Twenhofel, W. H. 1914. The Anticosti Island faunas. Mus. Bull. Can. geol. Surv., Ottawa, 19 : 1-35, pi. 1. Ulrich, E. O. & Cooper, G. A. 1942. New genera of Ordovician brachiopods. /. Paleont., Menasha, 16 : 620-626, pi. 90. Walmsley, V. G., Boucot, A. J. & Harper, C. W. 1969. Silurian and lower Devonian salopinid brachiopods. /. Paleont., Tulsa, 43 : 492-516, pis 71-80. Williams, A. et al. 1965. Brachiopoda. In Moore, R. C. (ed.), Treatise on Invertebrate Paleontology, H. xxxii+927 pp., illustr. Kansas. Wright, A. D. 1968. The brachiopod Dicoelosia biloba (Linnaeus) and related species. Ark. Zool., Stockholm, (2) 20 : 261-319, pis 1-7. Ziegler, A. M., Cocks, L. R. M. & Bambach, R. K. 1968. The composition and structure of lower Silurian marine communities. Lethaia, Oslo, 1 : 1-27. , Hansen, K. S., Johnson, M. E., Kelly, M. A., Scotese, C. R. & Van der Voo, R. 1977. Silurian continental distributions, paleogeography, climatology, and biogeography. Tectonophysics, Amsterdam, &c., 40: 13-51. Two new condylarths (Mammalia) from the early Eocene of southern England J. J. Hooker Department of Palaeontology, British Museum (Natural History), Cromwell Road, London SW7 5BD Synopsis Hyopsodus wardi sp. nov. is described as the first English species of the mainly North American genus Hyopsodus and is shown to be closely related to the earliest North American species H. loomisi. Lessness- ina packmani gen. et sp. nov. is described, extending the range of the anisonchine Periptychidae from exclusively the Palaeocene of North America into the early Eocene of Europe. Introduction Collecting over the past two decades in the early Eocene Blackheath Beds of Abbey Wood (London Borough of Bexley) has brought to light a number of mammalian specimens. Amongst these are several jaw fragments and isolated teeth representing two new species of condylarth. One belongs to Hyopsodus Leidy, the other to a new monotypic genus. The new Hyopsodus specimens have allowed this genus to be recognized in older collections from Abbey Wood and from the Suffolk Pebble Beds (early Eocene) of Kyson, Suffolk, where it had been misidentified. Most previous European records of Hyopsodus (Riitimeyer 1892, Teilhard de Chardin & Fraipont 1921) belong to other taxa (see Stehlin 1906 : 632-634; and below). Russell (1968) was the first to record the genus from Europe correctly; Dashzeveg (1977) has described an endemic species from Mongolia. A relatively small number of mammalian species have already been described from Abbey Wood (White 1931, Cooper 19320, b, Simons 1962, Van Valen 1965, KUhne 1969) but recent collecting by bulk sampling, sieving and acetic acid concentration has begun to provide glimpses of a much larger fauna. Remains are sparse and often fragmentary, but the taxa described here are among the best-represented so far. Stratigraphical details of various temporary sections dug in the Blackheath Beds of Abbey Wood appear in Epps & Priest (1933), Rundle (1970; contains numerous additional references), Ward & Cooper (1971) and Hooker (1975). Two lithostrati- graphic terms (Lessness Shell Bed and Abbey Wood Member) introduced by Cooper (1976 : 6) are used here but are included in the Blackheath Beds (informal usage) contrary to Cooper's interpretation. Material and abbreviations. Nearly all the specimens described here are in the Department of Palaeontology, British Museum (Natural History); their numbers are prefixed M. Those of other collections are prefixed as follows: University of California (UC); University of Michigan (UM); W. Morris private collection (MC). The initials of the collectors are indicated where relevant: Messrs S. A. Baldwin, A. G. Davis, F. J. Epps, W. George, P. R. Gurr, J. J. Hooker, R. G. Maynard, A. R. G. Packman and D. J. Ward. 1= maximum length; w= maximum width (wi and W2 being widths of trigonid and talonid respectively of lower teeth); L=left; R= right. Systematic descriptions Order CONDYLARTHRA Cope, 1881 Family HYOPSODONTIDAE Trouessart, 1879 Genus HYOPSODUS Leidy, 1870 TYPE-SPECIES. Hyopsodus paulus Leidy, 1870; Bridgerian, Middle Eocene, Wyoming, U.S.A. Bull. Br. Mus. not. Hist. (Geol.) 32 (1) 43-56. Issued 28 June 1979 43 44 J. J. HOOKER Hyopsodus wardi sp. nov. Figs 1-6; 8-13 HOLOTYPE. Right maxillary fragment with P4 alveoli and M1-2 (M 29762). PARATYPES. Twenty-four jaw fragments and isolated teeth listed in Table 1. DERIVATION OF NAME. Named after Mr D. J. Ward, who collected the holotype. TYPE HORIZON AND LOCALITY. Lessness Shell Bed within the lower sandy Abbey Wood Member of the Blackheath Beds, Sparnacian, early Eocene; the sandpit, Abbey Wood, London Borough of Bexley (National grid TQ 480786). RANGE. Apart from specimens from the type locality, two isolated teeth of Hyopsodus from the Suffolk Pebble Beds of Kyson, Suffolk (TM 269475), may represent this species and are referred to as H. cf. wardi. DIAGNOSIS. Upper molars: short and broad; ectocingula variably developed; M1-2 hypocones small and close to protocones; protostyles present; M3 triangular and without hypocone. Lower molars: short and broad; relatively transverse orientation of opposing buccal and lingual cusps ; paraconid prominent and may be isolated from metaconid on M i , small and close to metaconid on M.2', talonid notch narrow and with entoconulid; entoconid notch (sensu Gingerich & Simons 1977) very shallow. These characters distinguish H. wardi from all described species of Hyopsodus except H. loomisi McKenna, 1960; this is larger (Fig. 14; Table 1) but morphologically similar, except that it completely lacks upper molar ectocingula. Mean length of H. wardi M2 is 2-90 mm. DESCRIPTION AND COMPARISONS. (Dental terminology follows Van Valen (1966) and Szalay (1969) except that upper mesial and distal cingula are not separated into para- and precingula and meta- and postcingula respectively.) Size has been shown to be the only reliable character to differentiate contemporaneous species of North American Hyopsodus which are otherwise morphologically identical (Gingerich 1976). Morphology in the genus, however, changes with time. Dimensions of the few known upper teeth of H. wardi do not overlap with those of the holotype of H. loomisi, according to McKenna's (1960 : 106) measurements (see Fig. 14). Those of more extensive assemblages of both species would almost certainly overlap slightly. M 29640, the buccal half of a left M ' , assuming that the other three M!s are near the mean, must be close to the upper size limit for H. wardi. There is the slight possibility that M 29640 belongs to a larger species of Hyopsodus occurring alongside H. wardi, but more material is obviously needed to decide this. Apart from the measurements of the holotype, Guthrie (1967 : 38, text-fig. 27) has produced a histogram with measurements of two MIS of H. loomisi from Four Mile, Colorado (the type locality), which almost fit within the range of those of H. wardi MIS. However, H. 'miticulus' also from Four Mile (see measurements in McKenna, 1960 : table 6) may also refer to H. loomisi (personal communication, Dr P. D. Gingerich). If Guthrie's (1967) measurements of H. 'miticulus' from Four Mile are thus included with those of H. loomisi, the supposed bimodality not being obvious, the mean M2 length is increased. This mean measure conforms more closely to that expected from the upper tooth measurements of the holotype of H. loomisi, and is thus greater than that of H. wardi. Gingerich (1976 : text-figs 4-5) has provided statistical measurements of MI for a sequence of assemblages in the Big Horn Basin, Wyoming, identified as H. loomisi, which range fairly widely in size, but would encompass those of the holotype. The stratigraphically lowest of these (1976: text-fig. 5) would overlap considerably with H. wardi. Figs 1-6. Hyopsodus wardi sp. nov. from the Blackheath Beds of Abbey Wood; upper teeth. Fig. 1, LP2, M 32167. Fig. 2, RP3 (reversed), M 31997. Fig. 3, LP4, MC44. Fig. 4, holotype R maxilla with P4 alveoli and M1-2 (reversed), M 29762. Fig. 5, RM3 (reversed), M 29644. Fig. 6, LdP3, M 29745. 'a' suffixes are occlusal views; 'b' suffixes are buccal views. Fig. 7. Lessnessinapackmanigen. et sp. nov. from the Blackheath Beds of Abbey Wood; RM1, MC 17; a, buccal view; b, occlusal view. All specimens x6 and sprayed with ammonium chloride. NEW ENGLISH CONDYLARTHS 45 46 J. J. HOOKER P2 (M 32167) is almost identical in morphology to that of the holotype of H. loomisi (UC 44781 ; McKenna 1960 : 107, text-fig. 58). P3 is triangular in outline. M 29408 appears relatively slightly shorter than M 31997, even accepting the broken-off parastyle. The mesiolingual and distolingual margins are slightly con- cave in M 31997, straight in M 29408. The distal cingulum continues lingually beyond the junc- tion with the postprotocrista in M 29408, but stops at the junction in M 31997. The buccal margin is slightly convex in M 31997, straight in M 29408. M 31997 has a very slight lingual bulge on the postparacrista which suggests an incipient metacone, absent in M 29408. The mesial cingulum in both specimens projects lingually beyond the junction with the preprotocrista. In the holotype of H. loomisi the mesial cingulum and preprotocrista form a single crest. In outline shape, of the two Abbey Wood P3s M 29408 is more like the holotype of H. loomisi, whilst M 31997 is more like a referred specimen of H. loomisi (UM 63620) from Locality 383 of the Bighorn Basin. Thus at least one type of individual variation of H. loomisi is paralleled by H. wardi. Of the two P4s, MC 44 is complete and MC 65 is a protoconal fragment. Outline shape of MC 44 is very similar to that of the holotype of H. loomisi (UC 44781). Mesial and distal cingula on both have the same lingual extent as in UC 44781. The ectocingulum is complete in MC 44, missing in UC 44781. Presence of a small metacone on MC 44 is suggested by a lingual dentine bulge behind the paracone, unlike UC44781. There is a very small paraconule on MC 44 but not on MC 65. It is missing in McKenna's (1960) figure of UC 44781. The last two features appear to have only individual significance, being variable in different North American Hyopsodus species. The dP3 (M 29745) has a broken mesial edge. It is much smaller than the two P3s, more quadrate in outline, relatively lower-crowned, with a very small metacone and no metastyle. Deciduous premolars of Hyopsodus appear to be uncommon in North America. Gazin (1968 : 50; pi. 6, fig. 3) figured the 'best-preserved': dP2-4 of//, lepidus Matthew, 1909 from the Bridger Formation of the Green River Basin, Wyoming. Compared to dP3 in the figure, M 29745 was apparently less tapered mesially and the distobuccal corner is slightly acute instead of right-angled ; the figure is not detailed enough for further comparisons. The size ratios of dP3 to permanent teeth in //. wardi are similar to those of H. lepidus as shown by Gazin's illustrations (1968 : pi. 6, fig. 3 and pl. 7, fig. 3). On the upper molars, there is variation in development of the ectocingulum but on the available specimens it is never completely absent as it is on the holotype of H. loomisi. McKenna (1960 : 106) used this as a diagnostic character of the latter species. Possible variation in upper molar characters such as development of the hypocone and ectocingulum attributed to H. loomisi by Gazin (1962 : 62) refer to assemblages away from the type area of the species and they may not be referable to H. loomisi. Some of the problems have been discussed by Delson (1971 : 350-351). Most of the upper molars of H. wardi have an incipient mesostyle either in the form of a vertical ridge on the buccal wall or a denticle on the ectocingulum. The paraconule and metaconule cristae show much variation in development. There is variation in the buccal saliency of the paraconal region in M2 and M3. M3 varies more than M2 in size, proportions and degree of reduc- tion of the metacone. The distal cingulum on M3 is also variable, in M 29631 there being at its lingual end a minute denticle in the position of a hypocone. McKenna (1960) could find little significant to write about the lower teeth of H. loomisi and did not figure them. For comparison with H. wardi I have relied on referred material : a cast of UM 63621 from Locality 383, Bighorn Basin and figures in Gazin (1962 : pl. 9, figs 5-7) of speci- mens from Bitter Creek, Washakie Basin, Wyoming. The PSS and P4 from Abbey Wood are similar to these teeth in UM 63621 but the trigonids are slightly more open and the paraconids Figs 8-13. Hyopsodus wardi sp. nov. from the Blackheath Beds of Abbey Wood; lower teeth. Fig. 8, RP3, M 15126. Fig. 9, RP4, M 29757. Fig. 10, RM,, M 15123. Fig. 11, LM2 (reversed), M 29497. Fig. 12, fragment of R mandibular ramus with M3, M 15132. Fig. 13, fragment of R mandibular ramus with Mj_2, M 15146, in buccal view, 'a' suffixes are lingual views; 'b' suffixes are occlusal views; 'c' suffixes are buccal views. All specimens x 6 and sprayed with ammonium chloride. NEW ENGLISH CONDYLARTHS 47 48 J. J. HOOKER and metaconids smaller. There is a closer resemblance to one of the Bitter Creek specimens (Gazin 1962 : pi. 9, fig. 6). H. wardi lower molars are very similar to those of UM 63621 and the Bitter Creek specimens except that the crest joining the entoconid and hypoconulid is more prominent, causing these cusps to be less separated (i.e. very shallow entoconid notch). The MI paraconids seem a little less distinct in the Bitter Creek specimens than in H. wardi. Of the two H. wardi M i s, one (M 15123) has a distinct paraconid isolated from the metaconid; the other (M 15146), although at a similar wear stage, has a smaller paraconid fused to the metaconid. All the sufficiently well preserved lower molars have entoconulids (= entostylids) ; some have transversely elongated ectostylids unlike UM 63621. The M3 (M 15132) has a more distally tapering talonid than UM 63621 and, unlike the latter, a buccal cingulum round the protoconid. Only one specimen of H. wardi has MI and M2 associated. The isolated M 15123 on the one hand and M 29497 and M 29698 on the other can, however, be confidently identified as M i and M2 respectively. The former has a distinct paraconid and the talonid wider than the trigonid, whilst the latter two have very small paraconids close to the metaconids and talonids equal to or narrower than the trigonids. The other known preultimate molars are too worn or broken to be further identified. H. orientalis Dashzeveg 1977, from the early Eocene of Naran Bulak, Mongolia, has been diagnosed as having M2 significantly larger than M3 ; hypocone absent on M3 ; lower molars with well-developed paralophid and weak entoconid from which lingual hypoconulid is weakly separated. The size ratios between some of the Mls and M2s seem more significant than between M2 and M3, but not in every case (judged from Dashzeveg's figures and measurements). In addition, the M1-2 centrocristae are acutely buccally flexed, the M1-2 parastyles are prominent and, according to the figures, the lower cheek teeth appear higher-crowned than in other Hyopsodus species. The MI paraconid is stated to be present but appears much smaller than in H. wardi. Apart from the obvious differences H. orientalis is slightly smaller than H. wardi although the measurements overlap. Table 1. Length and width measurements (mm to nearest 0-05) of English Hyopsodus wardi sp. nov. (including H. cf. wardi.). Brackets indicate an estimate. Only a single width measurement is given for lower premolars Number Tooth 1. w. Number Tooth 1. w, w2 UPPER DENTITION LOWER DENTITION M 32167 (RGM) LP2 2-00 1 •45 M 15126 (FJE) RP3 2 •45 I- 75 M 29408 (JJH) RP3 - 2 •70 MC133 LP3 2 •45 1- 75 M 31997 (JJH) RP3 2-65 2 60 M 29757 (SAB) RP4 2 •75 2- 00 MC44 MC65 LP4 LP4 2-40 3 •40 M 15 146 (AGO) (RM, \RM2 3 3 •05 •15 2-30 2-60 (2-45) 2-60 M 29762 (DJW) ( RM1 2-65 3 •40 M 15123 (FJE) RMj 2 •85 2-20 2-30 (holotype) \ RM2 2-90 4 •00 M 29497 (WG) LM2 3 •50 (2-90) (2-90) M29631(ARGP) / LM2 \ LM3 2-80 2-30 4 3 •10 •35 M M 13766 (FJE) 13767 (FJE) RM1/2 RM1/2 3 3 •30 10 2-55 _ M 20214 (PRG) VLM' 2-70 3 •40 M 31881 (SAB) RM1/2 3 10 - - M 29640 (PRG) LM1 3-05 - M 15132(AGD) RM3 3 •40 2-55 2-10 M 29642 (PRG) LM1 2-75 3 •45 M 202 18 (PRG) LM2 3-15 4 •30 M 29643 (PRG) RM2 2-90 3 •90 M 29698* RM2 3 •25 2-65 2-40 M 29644 (PRG) RM3 2-15 3 •20 M 32144 (JJH) RM3 2-15 3 •20 M 29745 (SAB) LdP3 - 2 •05 M 29694* RM2 2-90 4 •45 *Hyopsodus cf. wardi; not paratypes. NEW ENGLISH CONDYLARTHS 49 The holotype right maxilla of H. wardi is truncated anteriorly at the level of the infraorbital foramen which forms part of the edge. It lies above the former position of the distal half of P3. The orbital floor is preserved with the opening of the maxillary canal above the junction of P4 (now missing) and M1. The base of the zygomatic arch is opposite M2. M 29631 is a left maxilla which shows the base of the zygomatic arch and part of the orbital floor. mm 2- mm w. Fig. 14. Scatter diagram of length versus width for upper cheek teeth of Hyopsodus wardi sp. nov. (hollow symbols) and of the holotype of H. loomisi McKenna (after McKenna 1960) (solid symbols). 0*=P2J n«=P3; O*=P4; AA=M1; VT=M2; ^ = M3. K indicates H. cf. wardi from Kyson. Brackets indicate that the width is estimated. DISCUSSION. It has been shown above that H. wardi is very similar to H. loomisi. If North American and English Hyopsodus evolved in parallel at the same rate, then the best transatlantic correlation of the Blackheath Beds of Abbey Wood would be with strata in the Rocky Mountains belonging to the Lower Graybullian land mammal age within the early Wasatchian (early Eocene). Savage (1971) pointed out the great similarities between the Sparnacian mammal faunas of north-west Europe and those of the early Wasatchian of Wyoming. He noted that c. 60 % of the genera found in Europe were identifiable in North America. This is even more than in the Pleistocene, when glaciations favoured circumpolar ranges, or in the immediately preceding Palaeocene. He suggested that the similarities reflect a sudden migration across the area. In a migration of this kind one would expect even the species to be identical, but this does not appear to be so. This may be because the basal Sparnacian fauna of the Conglomerat de Meudon is poorly known and the main Sparnacian faunas are from higher horizons. Gingerich (1977 : 63) has pointed out that the European adapid primate Pelycodus eppsi (C. F. Cooper) from Abbey Wood differs from all North American Pelycodus in that its incipient hypocone arises from the basal cingulum instead of from the postprotocingulum, the difference being the initiation of the subfamily characteristics of the Adapinae and Notharctinae respectively. It would seem therefore that by Abbey Wood times some faunal differentiation, if not also isola- tion of the two areas, had already taken place. This reduces the reliance one might otherwise place on evolutionary grade in correlating Sparnacian and early Wasatchian faunas. Conversely, any significant time separation of the two faunas would be expected to have produced individual specializations in H. wardi as they have in H. orientalis. None are apparent in the material at present available, tending to support the correlation suggested above. Hyopsodus is the second most abundantly represented element in the mammal fauna at Abbey Wood. It is surpassed only by Hyracotherium, which may reflect a collecting bias caused by past coarse-sieving methods. These were also the two most commonly found genera in the early Wasatchian mammal fauna of Powder River, Wyoming (Delson 1971 : table 1). 50 J. J. HOOKER The tooth identified tentatively by Teilhard de Chardin & Fraipont (1921) as Hyopsodus from the Eocene fissure filling of Vinalmont, Belgium, was dismissed without explanation by Gazin (1968 : 8). From the figure it appears definitely condylarthran, but differs from early Eocene Hyopsodus in the following ways. It is more robust; the talonid is higher relative to the trigonid; the paraconid is central instead of lingual ; the cristid obliqua trends only slightly mesiolingually and meets the trigonid at a high level; the talonid notch is narrow and shallow; it is much larger. Family PERIPTYCHIDAE Cope, 1882 Subfamily ANISONCHINAE Osborn & Earle, 1895 Genus LESSNESSINA nov. TYPE SPECIES. Lessnessina packmani sp. nov. DERIVATION OF NAME. After the type locality in the woods of Lessness Abbey. Feminine. DIAGNOSIS. Small anisonchine (P3-M3 length 11-0 mm; other measurements in Table 2). P3-M3 protocones large and high and with mesial and distal concavities. P1-2 relatively large. P3~4 similar to one another, short and broad. Upper molars with low buccal cusps and without ectocingulum. M1-2 with well-spaced paracones and metacones; prominent mesostyles; proto- cones approximately in line with and not buccal to small subequal protostyles and hypocones. M3 reduced. Lessnessina packmani sp. nov. (Figs 7, 15) HOLOTYPE. Left maxilla with alveoli for canine and P1 and complete P3-M3 (M 29632). PARATYPES (3). Right P4 (M 29756), right Mi (MC 17) and right M3 (M 29637). DOUBTFULLY REFERRED SPECIMEN. Protoconal fragment of a right or left P3 or P4 (M 34648). DERIVATION OF NAME. In memory of the late Mr A. R. G. Packman, who collected the excellent holotype. TYPE HORIZON AND LOCALITY. Lessness Shell Bed, Abbey Wood Member, Blackheath Beds, Sparnacian, early Eocene; the sandpit, Abbey Wood, London Borough of Bexley (TQ 480786). Known from no other horizon or locality. DIAGNOSIS. Only known species of the genus. DESCRIPTION. Dental formula. ^ -^- ^ ^4 . Only P2-M3 are materially known. Two, more mesial single-rooted teeth are represented by alveoli on the holotype. The more distal of these is not separated from either P2 or the more mesial alveolus by diastemata. It once contained a root which was slightly recurved and longer than either of those belonging to P2. The crown would have been broader than and almost as long as that of P2. The remnant distal half of the more mesial alveolus indicates a longer and slightly more recurved root than the first. These alveoli are considered to have contained the canine and P1, as there is no maxillary/ premaxillary suture to indicate that the mesial one is for an incisor, and no interalveolar diastema (i.e. occluding space for a lower canine) to indicate that the distal one is for the upper canine. P1 is normally small in anisonchines (Matthew 1937 : 145-160). Small size of this tooth in mammals is often a prelude to its loss and it could be argued that in a late anisonchine the distal alveolus is more likely to have housed the canine. However, as P2 in Lessnessina is somewhat enlarged, it seems more logical to postulate a foward continuation of this trend to explain a large P1. State of preservation of the holotype. The specimen is well-preserved but shows evidence of water transport. The natural edges (i.e. those not broken in collecting) approximately follow the sutural edges of the maxilla, but the convex bone surfaces are slightly worn and the buccal edges NEW ENGLISH CONDYLARTHS 51 Fig. 15. Lessnessina packmani gen. et sp. nov. from the Blackheath Beds of Abbey Wood; holotype L maxilla with C and P1 alveoli and P2-M3, M 29632; a, occlusal view; b, buccal view; c, lingual view. All views x 6 and sprayed with ammonium chloride. 52 J. J. HOOKER of the teeth, particularly P2, are somewhat abraded. Natural tooth wear is fairly advanced especially on M1, indicating an old individual. P4 is more worn than P3, which is more worn than P2. This could indicate either a mesially trending eruption sequence, perhaps in response to premolar enlargement in a mesial direction, or a relatively distal area for maximum mastication. Upper dentition. Some of the tooth surfaces are slightly rugose. P2 is long and narrow in outline, broader distally. It is dominated by the paracone. There are large distolingual and steep narrow mesial wear facets. It is separated from P3 by a short diastema. P3 is transversely elongated, with conical paracone and protocone. The parastyle is low but mesially salient. There are rather transverse pre- and postprotocristae, the latter being very weak. The long axis of the tooth is oblique so that the protocone is mesial to the paracone. P4 is almost identical to P3 but the parastyle is higher and extends more lingually; and the postprotocrista is stronger, joining the higher distal cingulum at a small cuspule. It is only slightly larger than P3 in the holotype. M1-2 are triangular in outline and transversely elongated. The lingual half shows a rather sudden narrowing where the large mesial and distal cingula terminate in small protostyle and hypocone. The mesostyle is linked to both paracone and metacone by a low, slightly buccally flexed centrocrista. Although on M1 of both the holotype and MC 17 the paracone and metacone are worn to equal heights, the smaller area of exposed metacone dentine shows that this cusp was lower than the paracone, as can be seen more easily on the other two molars. The most noticeable feature of the M1-2 and to a lesser extent of P3-4 and M3 is the shallow convex lingual slope of the protocone whose enamel dives basally below the rest of the crown until it merges without overhang into the massive lingual root. Wear on M1 of the holotype is deep enough to have fused the protocone, paraconule, meta- conule and the large mesial and distal cingula into one large area of dentine, with even a narrow dentinal isthmus extending from the paraconule region to the worn paracone tip. The holotype M2 is less worn and the worn mesial cingulum stands isolated. The paraconule is still fused to the protocone but oddly the metaconule is independent of the protocone but fused to the distal cingulum. From the wear pattern of the holotype M1 it is evident that this tooth also had its protocone and metaconule separated at an earlier wear stage. That this is a variable character is shown by MC 17 where the worn paraconule, protocone and metaconule are fused and indepen- dent of the cingula. The upper molar parastyles are strong but merge lingually with the mesial cingulum. Table 2. Length and width measurements (mm to nearest 0-05) of Lessnessina packmani gen. et sp. nov. Number Tooth 1. w. LP1 alveolus 2-00 1-30 LP2 2-50 1-65 LP3 2-10 2-75 M 29632 (ARGP) (holotype) LP4 2-20 3-20 LM1 2-45 3-30 LM2 2-60 3-90 LM3 2-05 2-95 M 29756 (SAB) RP4 2-00 2-60 MC17 RM1 2-20 2-95 M 29637 (PRO) RM3 1-70 2-10 The holotype M3 is rather worn and abraded and this tooth is best studied from the smaller paratype (M 29637). The tooth is triangular, tapering distally, with a terminal metacone. The paraconule and especially metaconule are poorly developed. Breakage on both specimens makes it difficult to decide whether or not there was a mesostyle. The distal cingulum curves at its lingual end up the side of the protocone. Lower dentition. Although unknown, this would probably have been similar to that of Hemithlaeus except that Mi_2 would have been relatively longer with more expanded talonids NEW ENGLISH CONDYLARTHS 53 bearing large hypoconids ; M3 relatively smaller with reduced hypocormlid but larger hypoconid; Pi_2 relatively large, ?2 a particularly tall tooth according to the long steep mesial wear facet of P2. Maxilla. The palatal process is preserved as far as the midline suture. Part of the lower half of the orbital floor remains and shows the maxillary canal beginning between P4 and M1 and emerging at the infraorbital foramen above the mesial half of P3. The lower orbital margin was formed by the now missing jugal, its contact with the zygomatic process of the maxilla becoming a groove anteriorly. The fact that a portion of orbital rim composed of maxilla occurs antero- dorsal to the jugomaxillary suture suggests that the lacrymal was confined to the orbit. The orbit appears to have extended anteriorly to above the junction of P3 or P4. A small elongated foramen occurs near the jugomaxillary suture at the level of P4. DISCUSSION OF RELATIONSHIPS. Lessnessina, the youngest anisonchine, extends the subfamily's upper stratigraphical limit from the Upper Palaeocene to the Lower Eocene and its geographical range from North America to Europe. Androconus Quinet, 1965, a supposed periptychid from the Oligocene of Hoogbutsel, Belgium, has been shown to be synonymous with the pantolestid insectivore Cryptopithecus Schlosser, 1890 (Heissig 1977). Because of this considerable range extension, the relationships of Lessnessina require some justification. It has the following unspecialized condylarth features. Bunodont cusps, in combina- tion with non-molariform distal premolars ; small upper molar hypocones ; upper cheek teeth with crowns passing basally into powerful roots without an overhanging lingual enamel edge; absence of M1-2 postprotocingulum. One or more of these distinguish it from superficially similar genera in the orders Insectivora such as Paschatherium and Primates such as Plesiadapis. From the primitive Artiodactyla such as Diacodexis, Lessnessina differs essentially in its reduced M3 and short transverse P3 as well as in some features which distinguish it also from hyopsodontid condylarths (see below). Accepting its condylarthran affinities, one sees that the general style of dentition of Lessnessina is found only in the Hyopsodontidae and the anisonchine Periptychidae. It has been compared Table 3. Comparative table of characters in Lessnessina, Anisonchinae (except Lessnessina) and Hyopsodontidae Character Lessnessina Anisonchinae Hyopsodontidae P3-M3 length 11 mm 16-27 mm 10-28 mm Nature of cusps Bunodont Bunodont Semibunodont, crests usually more prominent Short, broad, p3 outline Short, broad, resembling P4 in Triangular, unlike P4 resembling P4 Hemithlaeus and where known Anisonchus Upper molar buccal cusps Low Acute lingual tapering of upper molar outline with enamel extending basolingually Relative size of M1 metacone Relative size of M3 Facial extension of lacrymal Yes Smaller than paracone Reduced Probably none High Yes Smaller than paracone Nearly as large as M2 Little or none in Haploconus (fide Matthew 1937 : 159) Low No Equal to paracone Reduced Large in Hyopsodus 54 J. J. HOOKER with original specimens, casts or published figures of the genera contained in these groups, where the upper dentition is known. They comprise the Anisonchinae Hemithlaeus, Conacodon, Anison- chus and Haploconus (I follow Van Valen (1967) in referring Coriphagus to the Pentacodontidae), and the Hyopsodontidae Choeroclaenus, Mioclaenus, Dracoclaenus, Protoselene, Promioclaenus, Ellipsodon, Jepsenia, Litaletes, Hyopsodus, Haplomylus, Phenacodaptes, Apheliscus, Aletodon, Paratricuspiodon, Tricuspiodon, Louisina and Microhyus (I follow Van Valen (1967) in referring Paschatherium and Litolestes to the Adapisoricidae, to which I also tentatively add Haplaletes and Litomylus). Table 3 compares key characters of Lessnessina with the rest of the Anisonchinae and the Hyopsodontidae. Similarities with the Anisonchinae predominate, particularly the important features of P3 shape, the basolingual enamel extension of the upper molars and the ? non-facial lacrymal. Similarities to the Hyopsodontidae are restricted to rather more generalized characters. ANCESTRY. Among the known Palaeocene anisonchines, Lessnessina is most like the structurally primitive Hemithlaeus from the Puercan (early Palaeocene) of New Mexico. Hemithlaeus P3-4 are, however, somewhat inflated and the parastyles are reduced. Lessnessina P3-4 are thus more like P4 of the Puercan Conacodon cophater (Cope, 1884), but here P3 has lost its protocone. Fundamental morphological similarity is thus closest to a hypothetical latest Cretaceous or earliest Palaeocene common ancestor of Hemithlaeus and Conacodon, where the premolars have not become inflated, P3 has a protocone and the M1-3 hypocones have not enlarged. The evolution of Lessnessina from this stage would have involved : 1 . ? reduction in overall size 2. reduction of M3 3. enlargement of P1-2 4. enlargement of P3-M3 protocones 5. ? expansion of MI -2 talonids with enlargement of hypoconids (hypothetical) 6. relative lengthening of upper and probably lower molars 7. acquisition of upper molar mesostyles. This sequence of events is theoretical as there is no known genus which could be the immediate ancestor of Lessnessina. There is a gap of at least 10 million years between the latter and the probable time of its latest common ancestry with the other known anisonchines. The latter all appear to be too specialized to have given rise directly to Lessnessina. Since this paper was submitted I have received from the American Museum of Natural History, through the courtesy of Mr E. Manning, casts of lower dentitions of the hyopsodontid Oxyacodon. They are quite dissimilar in detail from those of other hyopsodontids, especially in the structure of the lower molars. The distinct median paraconids (especially Mj-s), high trigonids and low talonids, prominent entocristids extending to tips of entoconids and metaconids, and mesio- lingual cingula, all favour inclusion in the Anisonchinae. Oxyacodon lower molar talonids are relatively broader than in other anisonchines, but otherwise cusp pattern is rather similar to that of Hemithlaeus. This character is postulated above for the unknown Lessnessina lower molars. The reduced Oxyacodon Ps_4 talonids also conform to the morphology expected for Lessnessina Ps-4, from the latter's simple P3-4. For a related example, where uppers and lowers are known, compare the occlusal relationships of Phenacodaptes P|. It is possible, therefore, that Lessnessina is closer to Oxyacodon than to the other anisonchines. O. agapetillus (Cope) in particular extends the size range of anisonchines downwards almost to encompass Lessnessina. Acknowledgements I would like to thank the collectors and Dr D. E. Russell for making specimens available, Dr P. D. Gingerich for information and for providing casts of important specimens, staff of the Photo- graphic Unit of the British Museum (Natural History) for the photographic illustrations, Mr J. Brydon and Dr R. P. S. Jefferies for Russian translation, and Drs A. W. Gentry, W. R. Hamilton and K. A. Kermack for critically reading the manuscript. NEW ENGLISH CONDYLARTHS 55 References Cooper, C. F. 19320. The Genus Hyracotherium. A revision and description of new specimens found in England. Phil. Trans. R. Soc., London, B 221 : 431-448, pis 49-51. 19326. On some mammalian remains from the Lower Eocene of the London Clay. Ann. Mag. not. Hist., London, (10) 9 : 458-467, pis 11-12. Cooper, J. 1976. British Tertiary stratigraphical and rock terms, formal and informal, additional to Curry 1958, Lexique Stratigraphique International. Tertiary Res. Spec. Pap., London, 1. 37 pp., 1 tab. Dashzeveg, D. 1977. [On the discovery of Hyopsodus Leidy, 1870 (Mammalia, Condylarthra) in the Mongolian People's Republic. In Barsbold, R. & Vorobjeva, E. L. (eds), Mesozoic and Cenozoic faunas, floras and biostratigraphy of Mongolia.] Trudy sovmest. sov .-mongoV . paleont. Eksped., Moscow, 4 : 7-13, 3 text-figs.- [In Russian.] Delson, E. 1971. Fossil mammals of the early Wasatchian Powder River local fauna, Eocene of northeast Wyoming. Bull. Am. Mus. nat. Hist., New York, 146 : 304-364, 24 figs. Epps, F. J. & Priest, S. 1933. Field Meeting at Abbey Wood, Kent. Proc. Geol. Ass., London, 44 : 417-421, pi. 39. Gazin, C. L. 1962. A further study of the Lower Eocene mammalian faunas of southwestern Wyoming. Smithson. misc. Collns, Washington, 144 (1). vi+98 pp., 14 pis. 1968. A study of the Eocene condylarthran mammal Hyopsodus. Smithson. misc. Collns, Washington, 153(4). iv+ 90 pp., 13 pis. Gingerich, P. D. 1976. Paleontology and phylogeny: patterns of evolution at the species level in early Tertiary mammals. Am. J. Sci., New Haven, 276 : 1-28, 8 text-figs, 1 tab. 1977. New species of Eocene primates and the phylogeny of European Adapidae. Folia primatol, Basel, 28 : 60-80, 8 figs. & Simons, E. L. 1977. Systematics, phylogeny and evolution of early Eocene Adapidae (Mammalia, Primates) in North America. Contr. Mus. Paleont. Univ. Mich., Ann Arbor, 24 : 245-279, 1 pi., 13 text-figs, 7 tabs. Guthrie, D. A. 1967. The mammalian fauna of the Lysite Member, Wind River Formation (early Eocene) of Wyoming. Mem. sth. Calif. Acad. Sci., Los Angeles, 5 : 1-53, 36 figs. Heissig, K. 1977. Neues Material von Cryptopithecus (Mammalia, Pantolestidae) aus dem Mitteloligozan von 'Mohren 13' in Mittelfranken. Mitt, bayer. St. Paldont. Hist. Geol, Munich, 17:213-225, 24 text-figs. Hooker, J. J. 1975. Report of field meeting to Abbey Wood, London. Tertiary Times, Bromley, 2 : 159-162. Kiihne, W. G. 1969. A multituberculate from the Eocene of the London Basin. Proc. geol. Soc., London, 1658 : 199-202. McKenna, M. C. 1960. Fossil Mammalia from the early Wasatchian Four Mile Fauna, Eocene of north- west Colorado. Univ. Calif. Publs geol. Sci., Berkeley, 37 (1). 130 pp., 64 text-figs. Matthew, W. D. 1937. Paleocene faunas of the San Juan Basin, New Mexico. Trans. Am. phil. Soc., Philadelphia, (NS) 30. viii + 510 pp., 65 pis. Quinet, G. E. 1965. Un Condylarthre de Hoogbutsel. Bull. Inst. r. Sci. nat. Belg., Bruxelles, 41 (15) : 1-5, pi. 1. Rundlc, A. J. 1970. Report of field meeting to Abbey Wood, Kent. Tertiary Times, Bromley, 1 : 20-23. Russell, D. E. 1968. Succession en Europe, des faunes mammaliennes au debut du Tertiaire. In Colloque sur 1'Eocene. Mem. Bur. Rech. geol. minier., Paris, 58 : 291-296, 1 text-fig., 1 tab. Riitimeyer, L. 1892 (1891). Die Eocane Saugethier-Welt von Egerkingen. Gesammtdarstellung und dritter Nachtrag zu den 'Eocanen Saugethieren aus dem Gebiet des schweizerischen Jura (1862)'. Abh. schweiz. paldont. Ges., Basel, 18 : 1-153, pis 1-8. Savage, D. E. 1971. The Sparnacian- Wasatchian mammalian fauna, Early Eocene, of Europe and North America. Abh. hess. Landesamt. Bodenforsch., Wiesbaden, 60 : 154-158. Simons, E. L. 1962. A new Eocene primate genus, Cantius, and a revision of some allied European lemuroids. Bull. Br. Mus. nat. Hist., London, (Geol.) 7 : 1-36, 3 pis. Stehlin, H. G. 1906. Die Saugethiere des schweizerischen Eocaens. Critischer Catalog der Materialien. 4ter Teil: Dichobune - Mouillacitherium - Meniscodon - Oxacron. Abh. schweiz. paldont. Ges., Basel, 33 : 597-690, pi. 12, text-figs 65-98. Szalay, F. S. 1969. Mixodectidae, Microsyopidae, and the insectivore-primate transition. Bull. Am. Mus. nat. Hist., New York, 140 : 193-330, pis 17-57. Teilhard de Chardin, P. & Fraipont, C. 1921. Note sur la presence dans le Tertiaire inferieur de Belgique d'un Condylarthre appartenant au groupe des Hyopsodus. Bull. Acad. r. Belg. Cl. Sci., Bruxelles, (5) 7 : 357-360, 1 text-fig. 56 J. J. HOOKER Van Valen, L. 1965. Some European Proviverrini (Mammalia, Deltatheridia). Palaeontology, London, 8 : 638-665, 6 text-figs. - 1966. Deltatheridia, a new order of mammals. Bull. Am. Mus. not. Hist., New York, 132 : 1-126. pis 1-8. 1967. New Paleocene insectivores and insectivore classification. Bull. Am. Mus. not. Hist., New York, 135 : 217-284, pis 6-7. Ward, D. J. & Cooper, J. 1971. Report of Project Meeting to Abbey Wood, Kent. Tertiary Times, Bromley, 1 : 61-66. White, E. I. 1931. The vertebrate faunas of the English Eocene. Vol. 1. From the Thane t Sands to the Base- ment Bed of the London Clay. 123 pp., 1 pi., 162 text-figs, 4 tabs. London, British Museum (Natural History). Miocene sharks' teeth from Ecuador A. E. Longbottom Department of Palaeontology, British Museum (Natural History), Cromwell Road, London SW7 5BD Synopsis The first Miocene sharks from Ecuador are described. Representatives of nine genera are identified: Procarcharodon, Odontaspis, Hemipristis, Galeocerdo, Negaprion, Carcharhinus, Scoliodon, Isistius and Aetobatus. The shark fauna is comparable with those from the Miocene of other South American coun- tries, the Caribbean, North America and Europe. The first occurrence of Isistius triangulus (Probst) in the Miocene of the western hemisphere is noted. Introduction During the years 1974-75, geologists from the Institute of Geological Sciences (Overseas Division) and the British Museum (Natural History) collected from the Neogene of Ecuador. Among the collections obtained were many sharks' teeth, which represent their first record from the Ecua- dorian Miocene, and hence are of considerable interest. The teeth were collected by Drs C. R. Bristow and J. E. P. Whittaker. All but one of the teeth come from three localities CRB 123, a, b (Fig. 1), which occur along a 4 km stretch of coast just south of Bahia. One further tooth was collected from locality J 2a about 50 km north of Bahia. This tooth is included because of probable correlation with the beds from which the remainder were collected. The details of the localities are as follows. CRB 123 is a sea stack at Punta la Colorada, grid reference 602308, at 0° 37-6' S, 80° 27-6' W. CRB 123a is the cliff near the lighthouse at Punta Bellaca, grid reference 615330, at 0° 36-4' S, 80° 27' W. CRB 123b is the cliff at Punta la Gorda, grid reference 596298, at 0°38-2' S, 80° 28' W. Grid references refer to the Cartographia Censal CC-MIII-D3 (123) Bahia de Caraquez, 1 : 50 000 map. Locality J 2a is at 0° 12' S, 80° 20' W. A full description of the stratigraphy of this coastal area was published by Bristow (1976). He originally included the sharks' teeth from localities CRB 123, a, b in his faunal list for the Borbon Formation. Further work has shown, however, that the beds containing the teeth are transitional between the Onzole Formation and the overlying Borbon Formation. Here Borbon- type sandstones sometimes occur in Onzole-type blue silts and vice versa and it is difficult to refer the teeth to either of these Formations with certainty. Planktonic foraminifera from nearby sample CRB 124 (grid reference 609318) have dated these transitional beds as Zone N17 of Blow (1969), late Miocene (J. E. P. Whittaker, personal communication). Age-diagnostic species are Sphaeroidinellopsis paenedehiscens Blow, Globorotalia plesiotumida Blow & Banner and Globoro- talia humerosa Takayanagi & Saito. The sample J 2a is from a part of the so-called Jama Formation which is thought to be equivalent to the Borbon Formation and also late Miocene. The matrix is a medium-grained sandstone which has been recemented by weathering, especially at locality CRB 123, the sea stack from which the teeth were partly eroded out. This means that it is extremely hard and is difficult to remove without breaking the specimens. It tends to adhere more strongly to the roots, thus allowing only the crowns to become visible. Consequently many specimens consist of crown only. The teeth described are all now deposited in the collections of the British Museum (Natural History), London, and are referred to in this paper by the register number of that institution with a 'P' prefix. Faunal Description The fauna includes 9 genera, with 10 species, all in the subclass Selachii. Measurements quoted are the vertical height of the complete tooth, unless otherwise stated. Bull. Br. Mus. not. Hist. (Geol.) 32 (1) : 57-70. Issued 28 June 1979 57 58 A. LONGBOTTOM Bah id CRB123 Post-Daule Group rocks Borbon and Progreso formations Onzole and Subibaja formations Angostura Formation Pre-Daule Group rocks mm®& Fig. 1. Generalized geological and locality map of part of coastal Ecuador (from Bristow 1976); CRB 123 and J 2a are C. R. Bristow sample locality numbers. MIOCENE SHARKS' TEETH 59 Class CHONDRICHTHYES Subclass SELACHII Family ISURIDAE Genus PROC ARCH ARO DON Casier Procarcharodon megalodon (Agassiz) SYNONYMY. See Cappetta (1970). MATERIAL. Four incomplete specimens, P. 59278-81. LOCALITIES. Three from CRB 123, one from CRB 123b. DESCRIPTION (Fig. 2). The teeth are all fairly large (estimated height 8-9 cm) and typical of this species. Family ODONTASPIDAE Genus ODONTASPIS Agassiz Odontaspis acutissima Agassiz SYNONYMY. See Cappetta (1970). MATERIAL. Eleven teeth, P. 59265-74. LOCALITIES. Nine from CRB 123, two from CRB 123b. DESCRIPTION (Figs 3, 4). The teeth are typical of this species. Most of them lack a root and denticles, but they all have vertical striations on the inner face of the crown. Most appear to be anterior teeth. Family CARCHARHINIDAE Genus HEMIPRISTIS Agassiz Hemipristis serra Agassiz SYNONYMY. See Cappetta (1970). MATERIAL. Two specimens, P.59275-6. LOCALITY. Both from CRB 123. DESCRIPTION. One specimen consists of the root only and the other is the apex of a crown. Genus GALEOCERDO Miiller & Henle Galeocerdo aduncus Agassiz SYNONYMY. See Cappetta (1970). MATERIAL. One tooth, P.59277. LOCALITY. CRB 123. DESCRIPTION. This specimen is an almost complete tooth, still partly embedded in the matrix. The tooth is large, being 2-3 cm across the widest part at the base of the crown. The tip of the crown is missing so the height cannot be measured. Genus NEGAPRION Whitley Isolated teeth of this genus are difficult to distinguish from the lower teeth of some species of Carcharhinus. From descriptions of fossil and Recent (Bigelow & Schroeder 1948) examples of this genus some of the specimens in this collection are included in the genus Negaprion. All the specimens are here identified as N. eurybathrodon (Blake). 60 A. LONGBOTTOM I Fig. 2. Procarcharodon megalodon. P.59280, CRB 123, inner view, xO-7. Figs 3-4. Odontaspis acutissima, both CRB 123. (3) P.59267. (4) P.59266. (a) inner views; (b) outer views, x 2. Figs 5-7. Negaprion eurybathrodon. (5) P.59262, CRB 123, outer view, x2-2. (6) P.59263, CRB 123a, outer view, x2-9. (7) P.59257, CRB 123 : (a) inner view; (b) outer view; x2-4. MIOCENE SHARKS' TEETH 61 Negaprion eurybathrodon (Blake) SYNONYMY. See Cappetta (1970). MATERIAL. Nine teeth, P.59257-64. LOCALITIES. One from CRB 123a, seven from CRB 123, one from CRB 123b. DESCRIPTION (Figs 5-7). Some of the teeth are fairly complete, others incomplete or still partly embedded in matrix. Two of the larger teeth (heights 1 -7 and 2 cm) appear to be upper teeth by White's (1955) description. These have more triangular crowns than the lower teeth, and flatter outer faces. They have a small central depression in the base of the enamel of the outer face. The lateral extensions of enamel along the roots have wavy edges. These two teeth are slightly asymmetrical and are probably lateral in position, since they also have wide roots. The lower teeth are symmetrical, even those most lateral ones with wide roots. The lateral extensions of the enamel, when visible, show only slight crinkling along the edges in some specimens. The teeth have a convex inner face and a bulbous outer face to the crown, especially at the base of the enamel. It is this character, and the lack of serrations on the main part of the crown, that distinguishes these teeth from the lower teeth of the Carcharhinus species in this fauna. Genus CARCHARHINUS Blainville This genus contains many living species with a variety of tooth types. The upper teeth tend to have triangular crowns with fine regular serrations along the edges. The lower teeth have narrower upright crowns with or without serrations, depending on the species. The unserrated lower teeth of some species are difficult to distinguish from those of species in other genera (e.g. Negaprion, Hypopriori) and other anatomical details are used in identifying Recent species, since teeth alone are not reliable. Fossil teeth of the unserrated type can therefore only be arbitrarily attributed to a genus and species, as with the foregoing Negaprion. The same is true when trying to associate upper and lower teeth as one species. I have placed two of the following types of teeth into one species, C. egertoni, because of custom, and because they fit the accepted descriptions of this species in Leriche (1942). Carcharhinus egertoni (Agassiz) SYNONYMY. See Antunes & Jonet (1969). MATERIAL. Forty-nine upper teeth, P. 592 15-236, P. 59244-6. Six lower teeth, P. 59237-42. LOCALITIES. Upper teeth, 43 from CRB 123, two from CRB 123a, four from CRB 123b. Lower teeth, six from CRB 123b. DESCRIPTION (Figs 8-19). Most of the upper teeth lack roots or are embedded in the matrix. They have broadly triangular crowns on a wide root. The crowns all show strong regular serrations, which become slightly coarser towards the base. The heights of the crowns vary from 7-3 mm to 15 mm, and within the sample there is a continuous size range. Anterior and lateral teeth are present, with the crowns becoming more asymmetrical and the roots becoming wider in the lateral teeth. There are crenulations at the base of the enamel on the outer face of the larger teeth. Many teeth have a central, raised pad of enamel at the base of the outer face. In these teeth the roots are narrower and the serrations finer than those of Carcharhinus priscus. The crowns also tend to be more triangular. The central pad of enamel is not present in C. priscus. The majority of the teeth are larger than described examples of C. priscus. The lower teeth are much fewer and are restricted to one locality. They have slender crowns which are symmetrical. All the specimens collected have narrow roots. The outer faces are more convex than those of the upper teeth. The crowns bear serrations at the tip only and when viewed from the outer side the crowns expand at the serrated part. In most the cutting edge is only strongly developed at the tip and is not very clear on the lower half of the crown. This is not due to wear since the fine serrations are still visible at the tip. 62 A. LONGBOTTOM b 10a 12a Figs 8-12. Carcharhinus egertoni, upper teeth, all CRB 123. (8) P.59215. (9) P.59225. (10) P.59217. (11) P.59216. (12) P.59219. (a) inner views; (b) outer views. All approx. x2. These teeth fit the description of C. egertoni by Leriche (1942 : fig. 6). As White (1955) says, there does not seem to be much difference between this species and the Recent Carcharhinus longimanus (Poey), the lower teeth being very similar. The difference in the numbers of upper and lower teeth may be an effect of preservation or collection. However, there is a possibility that the upper teeth represent more than one species. Two of the smaller ones show some similarity to C. prisons in having narrower crowns and coarser serrations. MIOCENE SHARKS' TEETH 63 Three of the upper teeth (Figs 17-19) show some variation. The crowns lean distally. The distal edge is concave and the mesial edge convex, especially so about half-way up the crown. This pronounced convexity is not present in the other specimens. Leriche (1942 : pi. 8, fig. 4) and Eastman (1904 : fig. 1) figure teeth similar to these as C. egertoni. 13a 14a 15a Figs 13-16. Carcharhinus egertoni, lower teeth, all CRB 123b. (13) P.59237, x3. (14) P.59238, x3. (15)P.59239, x2-5. (16) P. 59240, x 3. (a) inner views; (b) outer views. Carcharhinus cf. priscus (Agassiz) SYNONYMY. See Cappetta (1970). MATERIAL. Seven ? lower teeth, P.59247-53. LOCALITIES. Six from CRB 123, one from CRB 123b. DESCRIPTION (Figs 23-24). The teeth all have slender crowns on wide bases. The crowns are upright or incline slightly distally. These teeth also have fine serrations along the whole edge of the crown. They are very similar to the following species but are much smaller. These teeth are like those of C. prisons shown in Cappetta (1970 : pi. 14, figs 1-20) and examples in the British Museum (Natural History) from Montpellier and Florida. However, they have more definite serrations and the roots are separated by a more acute angle. The crowns are also narrower. Carcharhinus sp. MATERIAL. Three ? lower teeth, P.59254-6. LOCALITIES. Two from CRB 123, one from J 2a. DESCRIPTION (Figs 20-22). The teeth have symmetrical crowns on wide bases. The crowns are narrower than those of the upper teeth of C. egertoni. The outer face is slightly convex, the inner 64 A. LONGBOTTOM more so. These teeth have fine regular serrations along the whole edge of the crown and along the basal lateral extensions of enamel, and so differ from the lower teeth of C. egtrtoni. The teeth are fairly large and measure 1 -6 cm, 1 -49 cm and 1 -35 cm in height. The teeth are similar to upper teeth of C. priscus, but differ in their larger size and finer crenu- lations. These crenulations do not become coarser towards the base as they do in C. priscus and C. egertoni. 17a 18a Figs 17-19. Carcharhinus egertoni, upper teeth, all CRB 123. (17) P.59244, x2-9. (18) P.59245, x2-6. (19) P.59246, x 1-8. (a) inner views; (b) outer views. Genus SCOLIODON Miiller & Henle Scoliodon taxandriae Leriche SYNONYMY. See Antunes & Jonet (1969). MATERIAL. Four teeth, P.59285-8. LOCALITY. CRB 123b. DESCRIPTION. Two of the teeth (P.59285-6) have triangular crowns which are inclined distally. The mesial edge is a continuous concavity, the distal edge is more or less vertical and has a notch where it joins the distal denticle. The distal denticle has a slightly wavy edge. The inner and outer faces are convex, with the enamel on the outer face ending basally in a horizontal pad. A distinc- tive oblique nutritive groove on the inner face also produces a notch visible, from the outer side, between the roots. The two other teeth are about the same size as the previous two (5-5 mm and 3-7 mm high). They are similar to the above teeth except that the mesial edge is less concave and the outer 20o MIOCENE SHARKS' TEETH b 21 65 22a 23a 24a Figs 20-22. Carcharhinus sp. (20) P.59254, CRB 123; (a) inner view; (b) outer view: x2-9. (21) P.59255, CRB 123, outer view; x2. (22) P.59256, J 2a, (a) inner view; (b) outer view; x2-3. Figs 23-24. Carcharhinus priscus. (23) P.59247, CRB 123, x3-8. (24) P.59253, CRB 123b, x3-5. (a) inner views; (b) outer views. face is flat. Instead of a horizontal pad at the base of the outer face, there is a medial triangular depression in the enamel. Family SCYMNORHINIDAE Genus ISISTWS Gill Isistius triangulus (Probst) SYNONYMY. See Cappetta (1970). MATERIAL. Three lower teeth, P.59282-4. LOCALITIES. Two from CRB 123, one from CRB 123a. DESCRIPTION (Figs 25-26). Two of the specimens are still partly embedded in matrix. One of these shows half the root with an indication of the central foramen (Fig. 25). The other two 66 A. LONGBOTTOM specimens are crowns only. One specimen shows an indication of crenulations along the edge. All three crowns are triangular, very thin and typical of this species. They are all anterior teeth and fairly large for this species, with crown heights of 4-5 mm. This is the first definite record of Isistus triangulus in the Miocene of the western hemisphere. Casier (1958) described specimens of Isistius sp. from Barbados. In 1966 he redescribed them as Sphyraena kugleri Casier and figured more examples. However, his pi. 3, fig. 27 appears to be a specimen of Isistius triangulus and is thus included in Table 1 . Family MYLIOBATIDAE A few teeth of Aetobatus sp. and possibly Myliobatis sp. are present in the fauna but these are very worn or still embedded in the matrix and are not identifiable to species. Figs 25-26. Isistius triangulus, lower teeth. (25) P.59284, CRB 123a, inner view. (26) P.59282, CRB 1 23, outer view. Both x6-6. Discussion This is the first fauna of sharks to be described from the Miocene of Ecuador. The Miocene deposits along the coast of Ecuador are up to 5000 ft (1520 m) thick (Sheppard 1928) and much work has been done on the foraminifera and mollusca of these deposits. References to Miocene sharks from South America are sparse. Table 1 is a summary of the known geographical distri- bution of the species in this fauna. I know of no references to sharks' teeth from the Miocene of Colombia. Small faunas from Chile (Oliver-Schneider 1936, 1937), Peru (HofTstetter 1968) and Venezuela (Leriche 1938, Rodriguez 1968) contain little more than the cosmopolitan Procarcharo- don megalodon. The Peruvian fauna is thought to be uppermost Miocene like the one from Ecuador. Leriche (1938) says that Hemipristis serra and P. megalodon from Venezuela are also from the Upper Miocene. The other faunas are merely stated as being from the Miocene. Of the other South American countries, a large fauna is described from the Pirabas Formation of Brazil (Santos & Travassos 1960, Santos & Salgado 1971). This fauna is Lower Miocene in age, older than the Ecuador fauna, but five species are common to both faunas. The Brazilian fauna includes Scoliodon taxandriae, and this is the only other record of this species from the western hemisphere. Records of the Argentinian Miocene faunas are controversial. Sharks' teeth from the Patagon- ian and Parana formations have been described. Leriche (1907) described sharks from the MIOCENE SHARKS' TEETH 67 Ti- i i i I i /^^ ro -r T ~r ~r ' o7 ^» S + + + + + + o^ 2 N « P .2 CO l~ I* "l l" "T" "T" G rs] i-fm' _< + + + + + &> °^ U l~^ o « % "S S ~T ~r ~r ~r ~r ~r i § 'C s u ON + + + + + + — - 1 . II 00 S§ "^ § 0^ H " » "T" T T ' ~r T- '-i *~" o' ^H g^i^ tSo" " ^ n £ i,,. i g^^SiS.g^^^^ gf 2 8 -i ^?^"* "v ^5 | (^ »- cd^ o O.^5r^J ^nS ***> " _o 9 + i ^S^iS^SS^o*™ ^ ^ 1 s fS + + + s C "o o VO II ill i > ^-^ ON ctf o C ^ •+• + + + T + T + g r- "-1 E ^^ o g «0 + + + Si S> "o 2 ••§ g ^ ^ o ^ L^ -^. E~"J t^J ^H r^ /^*v — ^ c« O •* , , , fc^O^ rTfsfDo ^ ^J SB'S T-H * ' ' ** f^ QO^ ^C ^^ fl *"^ f^ ^^ •X"N ^^ „ 3 2 + + + + ™% ss o^'S'ssi^^ 1-a ro 1"H HH r^ .— nJ _J o *™H ^\ ^^ -•. ^— ^ W ^j "T^ 's~ ^ *^^ O '^H - — * G (S *""* + + I^H ctf *Tv IH "^ ^ cti ^\ 4) .^ (y Z 2 ^— - « ^ , ^ 'g S g d-S^ "3 '^ ^'S ^ § s o + d-or^c 1 Sci^'-'-Soctf'M 3 u |J O <— ' C So ^ J3 js C ^ T3 w> *5 o\ + + + + Illl fllllllll *-S § •a u oo + + + + + ^ Tt ^ W-! VO t-- 00 ON 0 -H d?j vo II C _o + + + + _|_ _|_ O |~, vo «" ON ON ov ON O M O, '^ ^ "^ C ^^ " ^* C 3 "O i/i N cS oo 2^ P S o £ f) ~r T 1 C3 > 3 as ON "O "B 'S "C ^ "^ CiJQ ^)" r— ^H P] (/I $ °° en ^ o? 0 <~ D, « S | .2 % S & -a • •§ ss 5 "S S« S o- S -S « ^s *» 6, w 5 a s I S? § § g &> S B .« •« .« ^ 1 Q •2 "§ "S "i 1 « 1 ^ a Q a "« ^ 53 ^ •§•§•§ :§ -S •§ |> r^ r^ r« 8 -| ^ < U O O to ^2 ^ , lP|t|i JljJ .S -ort5^S§^ ^«S3^ | SO£.S|»«^1 |odg liJ^I|||| Ili^i •Swol< >«M f H«dl 1 0 indicates specimens 2 This species is include 68 A. LONGBOTTOM Patagonian and thought it might contain a mixture of Miocene (with hums hastalis (Agassiz), Procarcharodon megalodon and Galeocerdo aduncus) and Oligocene species (with Odontaspis cuspidata Agassiz and hums desor-i (Agassiz)). Priem (1911) thought the fish and cetaceans indicated an Early Miocene age for the Patagonian, whereas the molluscs indicated an Eocene/ Cretaceous age. Ameghino (1889, 1906) described the Tertiaries of Argentina in great detail. He gave a list of sharks found in the Patagonian and said it was probably Early Eocene in age, despite the presence of Cretaceous species such as Corax rothi Ameghino, hums angustidens (Reuss) and Scapanorhynchus subulatus (Agassiz). Ameghino also gave a faunal list for the Parana group. This includes Carcharhinus egertoni, Procarcharodon megalodon, Galeocerdo aduncus and other typical Miocene species. He said they came from the marine Enterienne Formation of Parana, and concluded the formation was Upper Oligocene in age. Woodward (1900) questioned this and because of the presence of the above sharks thought the Parana was Upper Miocene or Lower Pliocene. More recently Pascual & Rivas (1971) have revised the vertebrate faunas of the Tertiary of Argentina. They give a shark faunal list, based on the works of Ameghino, for the marine facies of the Patagonian Formation and the 'Enterriense'. They arrive at a Lower Miocene age for the Patagonian, based on the cetacean and penguin faunas also found. Simpson (1972 : 6) agrees with this and says: 'Some marine beds confused with or even included in parts of the Patagonian Formation may be older, but it remains highly probable that the greater or typical part of the formation is not older than late Oligocene, more likely early Miocene, and that the same determination applies to the fossil penguins.' The shark fauna, as stated earlier, includes many Eocene and Cretaceous species. Pascual & Rivas (1971) give faunal lists for the 'Enterriense', 'Mesopotamiense' and 'Rionegrense' of the Parana. The shark determinations again are probably from Ameghino and include usual Miocene forms. The majority of other vertebrates listed, however, are terrestrial mammals and these indicate a Pliocene age for the Parana. Pascual & Rivas say that there seems to be a mixture of terrestrial and marine faunas and they have included all the genera cited in previous works whether terrestrial or marine. They say that the rarer forms seem to be older and possibly represent secondarily derived fossils. They do not say whether they consider the sharks to be derived. This does seem possible in view of their probable Miocene age. I have included P. megalodon, O. acutissima, C. egertoni, C. priscus, H. serra and G. aduncus in Table 1 as being probably from the Miocene of Argentina based on the above evidence. The information for the Caribbean region is mostly from Leriche (1938); other references are given in Table 1. Most of the faunas are merely described as Miocene in age. Exceptions to this are the specimens of N. eurybathrodon from Panama, which are from the Middle Miocene (White 1955). Casier (1958) describes H. serra, P. megalodon and C. egertoni from the Lower Miocene of Trinidad, and C. egertoni and H. serra from the Upper Miocene. The Bissex Hill Formation of Barbados is Lower Miocene and considered to be contemporaneous with the Calvert Formation of Maryland (Casier 1958, 1966). P. megalodon and H. serra from Mexico are also from the Lower Miocene (Leriche 1938, Kruckow 1959). The information for the North American faunas is taken from Leriche (1942) and Eastman (1904). The Maryland faunas have been studied by Eastman (1904), however the sharks' teeth all come from the Calvert Formation in the Lower Miocene. Aetobatus arcuatus is the only species recorded from the St Mary's Formation (Upper Miocene) by Eastman. Leriche (1942) says that a specimen of Odontaspis acutissima is also possibly from the Upper Miocene. The other American records in Leriche (1942) are Lower Miocene or undifferentiated Miocene. There are many more recent references to the Miocene of North America but it is not within the scope of this paper to review them all. The European distributions are taken from Cappetta (1970) and Antunes & Jonet (1969). Cappetta gives a good comparison of these faunas, including Lower, Middle and Upper Miocene. Table 2 (taken from Cappetta (1970) and Antunes & Jonet (1969)) shows that all the Ecuadorian species range into the Upper Miocene in Europe. The faunas from Montpellier and Portugal correspond closely to that of Ecuador. The Montpellier fauna does not include C. egertoni, N. eurybathrodon and the Carcharhinus sp., but the first two are present in Portugal. The Ecuadorian fauna is therefore typically Miocene and compares closely with those of MIOCENE SHARKS' TEETH 69 Table 2. Distribution of the species in the Oligocene and Miocene S Olig.— ^ ^ Mioc.' — L M U L M U Procarcharodon megalodon Odontaspis acutissima Hemipristis serra .... Galeocerdo aduncus. Negaprion eurybathrodon . Carcharhinus egertoni Carcharhinus priscus Scoliodon taxandriae Isistius triangulus .... Aetobatus arcuatuf .... T T T T f T Montpellier and Portugal, both of which have been very well studied. It differs from the Mont- pellier fauna in the absence of small sharks' teeth of several species. Further collecting in Ecuador is necessary for the study of the shark-tooth microfauna. Table 3 shows the bathymetric and climatic ranges of Recent members ot the genera present in the Ecuador collection. The molluscan and foraminiferal faunas of the Miocene of Ecuador both indicate deposition in warm, fairly shallow water. This interpretation is supported by the shark fauna, which has a predominance of supposed near-shore littoral forms. The Recent species are also tropical to subtropical in range. Isistius is the only pelagic form, but this is also neritic in habit, and could range into shallower water. This fauna from Ecuador is important for the confirmation of the presence of Scoliodon taxandriae and Isistius triangulus in the western hemisphere, and in the variety of Carcharhinus teeth present. Future collections from the Miocene deposits of Ecuador, and also Colombia, could prove useful in furthering the study of the genus Carcharhinus and in extending the known geographical range of other Miocene species. Table 3. Bathymetric and climatic ranges of the genera littoral pelag. bathy. trop. sub-trop. temp. Carcharodon + + Odontaspis x + Hemipristis + x Galeocerdo + + + x Negaprion x x Carcharhinus + + + Scoliodon x -t- -t- Isistius x + + Aetobatus + + X occurrence abundant + occurrence common = occurrence rare Acknowledgements I would like to thank Dr C. Patterson and Dr P. L. Forey, both of the British Museum (Natural History), for their constructive criticisms; Dr J. E. P. Whittaker and Dr C. R. Bristow for infor- mation about the sample localities; and Mr P. Richens of the British Museum (Natural History) photographic unit for the photographs. 70 A. LONGBOTTOM References Ameghino, F. 1889. Contribution al conocimiento de los mamiferos fosiles de la Republica Argentina. Adas Acad. nac. Cienc. Cordoba 6 : 1-1028. 1906. Les formations sedimentaires du Cretace superieur et du Tertiaire de Patagonie, avec un parallele entre leurs faunes mammalogiques et celles de 1'ancien continent. An. Mus. nac. Hist. not. B. Aires 15 (ser. 3a, 8) : 1-568. Antunes, M. Telles & Jonet, S. 1969. Requins de 1'Helvetien superieur et du Tortonien de Lisbonne. Revta Fac. Cienc. Univ. Lisb. C 16 (1) : 119-280. Bigelow, H. B. & Schroeder, W. C. 1948. Fishes of the Western North Atlantic. Part 1, Lancelets, cyclo- stomes and sharks. Mem. Sears Fdn Mar. Res., New Haven, 1 : 1-576. Blow, W. H. 1969. Late middle Eocene to Recent planktonic foraminiferal biostratigraphy. In Bronni- mann, P. & Renz, H. H. (eds), Proceedings First International Conference on Planktonic Microfossils, Geneva 1967, 1 : 199-422, pis 1-54. Leiden. Bristow, C. R. 1976. The Daule group, Ecuador (Middle Miocene - Pliocene). Newsl. Stratigr., Berlin & Stuttgart, 5 (2/3) : 190-200. Cappetta, H. 1970. Les selaciens du Miocene de la region de Montpellier. Palaeovertebrata, Montpellier, Mem. ext. : 1-139. Casier, E. 1958. Contribution a 1'etude des Poissons fossiles des Antilles. Schweiz. palaeont. Abh., Basel, 74 : 1-95. 1966. Sur la faune ichthyologique de la Formation de Bissex Hill et de la Serie oceanique, de 1'Ile de la Barbade, et sur 1'age de ces formations. Eclog. geol. Helv., Basel, 59 (1) : 493-515. Dartevelle, E. & Casier, E. 1943. Les Poissons fossiles du Bas-Congo et des regions voisines (premiere partie). Annls. Mus. r. Congo Beige, Brussels, A, ser. 3, 2 (1) : 1-200. Eastman, C. R. 1904. Pisces. In: Miocene (of Maryland). Systematic Palaeontology: 71-93. Maryland Geol. Survey. Hoffstetter, R. 1968. Un gisement de vertebres tertiaires a Sacaco (Sud-Perou), temoin neogene d'une migration de faunes australes au long de la cote occidentale sud-americaine. C. r. hebd. Seanc. Acad. Sci., Paris, D 267 (16) : 1273-1276. Kruckow, T. 1957. Die stratigraphische und palaogeographische Bedeutung der Miozanen Elasmo- branchier-Fauna von Baja California, Mexico. Neues Jb. Geol. Paldont. Mh., Stuttgart, 10 : 444-449. 1959. Das Miozan der Gulfkunstenebene von Mexico. Neues Jb. Geol. Paldont. Abh., Stuttgart, 109 (1) : 130-146. Leriche, M. 1907. Observations sur les Poissons du Patagonien recemment signales par M. Fl. Ameghino. Annls Soc. geol. N., Lille, 36 : 129-137. 1927. Les poissons de la Molasse suisse. Abh. schweiz. paldont. Ges., Basel, 46 : 1-26; 47 : 27-119. 1938. Contribution a 1'etude des poissons fossiles des pays riverains de la Mediterranee americaine (Venezuela, Trinite, Antilles, Mexique). Abh. schweiz. paldont. Ges., Basel, 61 : 1-42. 1942. Contribution a 1'etude des faunes ichthyologiques marines des terrains tertiaires de la plaine cotiere atlantique et du centre des Etats-Unis. Mem. Soc. geol. Fr., Paris, n.s. 20 (45) : 1-112. Oliver-Schneider, C. 1936. Comentarios sobre los peces fosiles de Chile. Revta chil. Hist, not., Santiago de Chili, 40 : 306-323. 1937. Sobre el verdadero nombre del selacio fosil 'Carcharias giganteus' Ph. Revta univ. Santiago, 32 (1) : 61-62. Pascual, R. & Rivas, O. E. O. 1971. Evolucion de las comunidades de los vertebrados de Terciario Argentino. Los aspectos paleozoogeograficos y paleoclimaticos relacionados. Ameghiniana, Buenos Aires, 8 (3-4) : 372-412. Priem, F. 1911. Poissons fossiles de la Republique Argentine. Bull. Soc. geol. Fr., Paris, (4a) 11 : 329-340. Rodriguez, S. E. 1968. Estratigrafia y Paleontologia del Mioceno en la Peninsula de Paraguana, Estado Falcon. Boln inf. Asoc. venezol. Geol. Min. Petrol., Caracas, 11 (5) : 127-152. Santos, R. da Silva & Travassos, H. 1960. Contribuicao a paleontologia do estado do Para. Peixes fosseis da Formacao Pirabas. Monografias Serv. Geol. Min. Brasil, Rio de Janiero, 16. viii + 35 pp. — & Salgado, M. S. 1971. Contribuicao a paleontologia do estado do Para. Novos restos de peixes da Formacao Pirabas. Bolm Mus. Para. Emilio Goeldi, Para, n.s. Geologia, 16 : 1-13. Sheppard, G. 1928. Notes on the Miocene of Ecuador. Bull. Am. Ass. Petrol. Geol., Chicago, 12 (6) : 671-673. Simpson, G. G. 1972. Conspectus of Patagonian fossil penguins. Am. Mus. Novit., New York, 2488 : 1-37. White, E. I. 1955. On Lamna eurybathrodon Blake. Ann. Mag. not. Hist., London, (12) 8 : 191-193. Woodward, A. S. 1900. On some fish-remains from the Parana Formation, Argentine Republic. Ann. Mag. not. Hist., London, (7) 6 : 1-7. A new fossil terrestrial isopod with implications for the East African Miocene land form S. F. Morris Department of Palaeontology, British Museum (Natural History), Cromwell Road, London SW7 5BD Synopsis A new species of fossil terrestrial isopod Eubelum rusingaense sp. nov. is described from the Lower Miocene of Rusinga Island, Lake Victoria, Kenya. It is suggested that the animal lived at about 2500 m within the montane forest litter. The invalid family group names Atracheata and Pseudotracheata (Vandel 1960) are replaced by Oniscacea and Porcellionoidea. Introduction The British-Kenya Miocene Expedition of 1947 (Leakey 1953, Leakey & Clark, 1955) collected a large quantity of fossil material from Rusinga Island, Lake Victoria, Kenya. A selection of arthropod fossils from this area has been figured, but without description, by Leakey (1953a : 445). Amongst the 'myriapod' specimens deposited in the British Museum (Natural History), London, were 46 specimens of a pseudotracheate terrestrial isopod of the family Eubelidae. This is the first recorded occurrence of the family as a fossil. This family, with the Armadillidae and Armadillidiidae, represent the most highly evolved group of conglobating forms within the Oniscoidea. Forms described from an older horizon, the Eocene-Oligocene Baltic Amber, all belong to the more primitive families Oniscidae, Porcellionidae and Trichoniscidae. The Armadillidiidae represented by Armadillidium molassicum (Heer, 1865) is found in the Miocene of Oeningen. The family Eubelidae is now essentially a family with a tropical distribution. The sole exception is Sphaerobathytropa ribauti Verhoeff, found in the central Pyrenees, but Vandel (1943) considered this species to represent a pre-glacial remnant fauna. Fossil terrestrial isopods are rarely found and the handful of species so far described hardly allows a phylogeny to be erected which is backed up by fossil evidence. Edney (1960), however, supposed that the marine isopods left the sea via the littoral in the middle Mesozoic, whilst Vandel (1943 : table 2) proposed that the terrestrial forms were developed from the marine Cirolanidae. The latter are well represented in the fossil record (Jurassic-Recent) if the genus Palaega is accepted within this family, as suggested by Rathbun (1935) and Imaizumi (1953). Considering the world-wide tropical distribution of the Eubelidae and especially their occurrence in the Americas, Africa and southern India, the latest date for the establishment of the family is Albian, since it was at about this time that the Guinea-Brazil areas became separated (Maack 1969). Edney (1954, 1968) and Warburg (1968) have dealt in detail with the morphological and be- havioural adaptations required to survive on land. Systematic description Order ISOPODA Latreille, 1817 Suborder ONISCOIDEA Latreille, 1803 Superfamily PORCELLIONOIDEA Verhoeff, 1918 [nom. transl. Morris herein, ex Porcellionidae Verhoeff, 1918] ( = Pseudotracheata Vandel, 1960). Vandel (1960 : 95) formalized the classification of the Oniscoidea. Unfortunately in the tribe Crinocheta the two family group names that he chose, Atracheata and Pseudotracheata, were not based on an included nominate genus and are therefore invalid names under the Rules Bull. Br. Mus. not. Hist. (Geol.) 32 (1) : 71-75. Issued 28 June 1979 71 72 S. F. MORRIS (Art. ll(e)). The replacement name for the Atracheata must be Oniscacea [nom. transl. Morris herein, ex Oniscia Rafinesque-Schmaltz, 1815 : 101]. The new names are used in exactly the same sense as Vandel's (1960) replaced names and with the same included families. Family EUBELIDAE Budde-Lund, 1899 Genus EUBELUM Budde-Lund, 1885 TYPE SPECIES. Eubelum lubricum Budde-Lund, 1885 by monotypy. RANGE. Lower Miocene to Recent. Eubelum rusingaense sp. nov. (Figs 1-11) DIAGNOSIS. Eubelum with three-jointed antennal flagellum. Eyes with 17-19 ommatidia. Profrons folded onto vertex. Triangular escutcheon present. Shallow groove on coxopodite ridge of first pereion segment. Schisma and sulcus arcuatus present. Telson not divisible into two parts. Uropod protopodite almost completely fills space between the fifth pleonite and telson. Distal part of telson converging posteriorly. LOCALITY AND HORIZON. Clark & Leakey (1951 : 9) locality R.la from between Hiwegi Hill and the east coast of Rusinga Island, Lake Victoria, Kenya. The beds are of Lower Hiwegi Series, Lower Miocene age. HOLOTYPE. BM(NH) In.61025 PARATYPES. In.61026-In.61070. DESCRIPTION. Cephalon with convex protrons pushed up over the vertex and extending as a ridge along the frontal line. The frons is prominent with shallow lateral antennal grooves. Antennary tubercle raised into ridge; clypeus convex. There is a prominent median spine of the frontal lamina. Labrum strongly convex, rounded, triangular with the ventral portion bearing two strong laterally-directed ridges. A triangular escutcheon is developed. The mandibles are quadrate. The anterior margin of the first pereion has a moderately sulcate margin and a straight pos- terior margin. The thickened coxopodite ridge of the epimeron of the first pereionite has a shallow ventrolateral groove. The epimeron is contiguous with the segment but the junction is marked by a deep sulcus arcuatus running parallel to and all the way along the lateral margin. A schisma is present and the internal lobe is the largest. The marginal edges of the second to fourth pereionites are rounded and thinned for their combined insertion into the schisma during enrollment. The fifth to seventh pereionite margins are squared off with a gentle lateral curvature. The pleonite segments are similarly squared off. The telson is not divisible into two parts, having arcuate lateral margins and a gently rounded posterior margin. The uropod protopodite fills the space between the telson and the fifth pleonite. The body is capable of complete or nearly complete conglobation. There are sparse minute pits on the thoracic segments especially along the posterior bands of the segments. The abdominal segments are smooth. A few random granules may be present on the anterior part of the telson. Figs 1-11. Eubelum rusingaense sp. nov. Lower Hiwegi Series, Lower Miocene of Rusinga Island, Kenya; all x5. 1, 2, anterior and left lateral views of holotype, In.61025. 3-11, paratypes. 3, ventral view of In. 6 1026, to show triarticulate flagellum of left second antenna. 4, ventral view of In. 61027, to show leg bases. 5, dorsolateral view of In. 61028, to show ommatidia of right eye. 6, left lateral view of In. 61029. 7, ventral view of In. 61030, to show clypeus and schisma. 8, ventral view of In. 61031, to show coxopodite ridge. 9, ventral view of In. 61032, showing imperfect preservation of legs. 10, anterolateral view of In. 61033, to show ommatidia of right eye. 11, anterolateral view of In. 61034, to show partial enrollment. Fig. 12. Eubelum ignavum Budde-Lund, 1899. Recent; 8000-9000 ft (2460-2775 m), Djem-Djem Forest, Ethiopia. Anterior view of BM(NH) Zool. Dept. no. 1949.2.2.235, x5. EAST AFRICAN ISOPOD 73 11 12 74 S. F. MORRIS DISCUSSION. The isopods are preserved in the round in a semi-transparent calcite. They occur with other, as yet undescribed, arthropods especially insects and myriapods. The nearest relations to the fossil species are Eubelum ignavum Budde-Lund, 1899 and E. dollfusii Budde-Lund, 1899. E. rusingaense differs from E. dollfusii by its continuous frontal line across the cephalon. E. ignavum, which has a continuous frontal line, differs because the posterior part of its profrons is not pushed up onto the vertex and its eye has only 14-15 ommatidia. The distal portion of the telson of both the Recent species has parallel sides and is relatively wider than the fossil species. E. rusingaense does have some similarities with the type species E. lubricum Budde-Lund from Zaire, Congo Republic and Cabinda (Angola) (Ferrara & Schmalfuss 1976) but the telson is not so convergent, the larger eyes have 23-24 ommatidia, the escutcheon is much less prominent but broader, and it differs in the shape of the supra-antennal line. Van Name (1920 : 85) compared E. asperius with E. ignavum, but Arcangeli (1950) made E. asperius the type species of a new subgenus Pseudethelum, based on the two-jointed flagellum of the second antenna. Ferrara & Schmalfuss (1976) went further and suggested that after a thorough revision of the genus the asperius group would warrant generic separation from Eubelum s.s. Enrollment. The specimens from Rusinga are found in many degrees of enrollment, from a flat to a relaxed arched condition ; six specimens show a nearly complete enrollment with only a slight gape between the frontal line and the pleotelson. The coxopodite ridge of the first pereionite is ventrally grooved but, although complete closure would seem possible, none of the specimens show contact between the groove and the squared-off margins of the epimera of the fifth to seventh pereionites and first to fifth pleonites. The face is flattened and the escutcheon is well developed. The method and order of enrollment appears to be as in Vandel (1943 : fig. 10). When straight, the specimens can protrude the cephalon by a hinged movement along the posterior margin of the cephalon. REMARKS. Although terrestrial isopods are not dependent on an aquatic environment, they are poorly adapted to resist desiccation and super-saturation. Their habits are governed by the need to avoid excesses, although they will move along a hydrocline towards the saturation point. Since their transpiration rate is greater than their normal uptake of water, the difference between the two rates has to be made up from moist substances taken up by mouth or by water taken directly through the body wall (Edney 1954). Their moisture requirements can be satisfied by a number of habitats, consequently isopods are not indicative of any particular one. E. rusingaense probably lived in a damp montane forest litter, but could have lived in moist open grassland, under stones or near a water margin. The genus Eubelum in East Africa is, today, confined to montane forest at fairly high altitudes, certainly above 2000 m, where there is a relatively high annual precipitation rate. The nearest modern relatives of E. rusingaense, E. ignavum and E. dollfusii, have so far only been found in the Ethiopian highlands between 7000 ft (c. 2150 m) and 9000 ft (c. 2775 m) in montane forest litter near water margins (Barnard 1940 : 365). Rusinga Island lay, before the Miocene, within the Kavirondo rift valley system at about 4000-4500 ft (1300-1 350m) (Shackleton 1951, Pulfrey 1960, Bishop & Trendall 1967). Andrews & Van Couvering (1975) deduced that the pre-uplift elevation of the sub-Miocene remnant surface was at about 1500 m. They further note that this is the lower limit of the montane forest. Their analysis of the fauna showed no clear predominance of any vegetational type, but indicated that large tracts of forest occurred in the immediate vicinity. If the distribution of the East African isopods was climatically controlled in the Miocene as it is now then the surface would have been at a somewhat higher altitude than has been previously suggested, i.e. within the montane forest belt at about 2500 m. Eubelum is not confined to the highlands of East and central Africa but occurs in the lowlands of West Africa. However, these largely belong to the asperius group, that is now thought to be generically separable from Eubelum s.s. Furthermore, Vandel (1945 : 226) considered that low- land species represented recent immigrants, whilst highland species were those that remained near the autochthon. This suggestion is supported by the close relationship between the new species and the Recent indigenous species. EAST AFRICAN ISOPOD 75 Acknowledgements I would like to thank Dr F. Ferrara, Florence, for help with the systematic position of the fossil within the Eubelidae. References Andrews, P. & Van Quivering, J. A. H. 1975. Palaeoenvironments in the East African Miocene. In Szalay, F. S. (ed.), Approaches to Primate Paleobiology. Contrib. Primatol., Basel, 5 : 62-103. Arcangeli, A. 1950. Isopodi terrestri. In: Exploration du Pare National Albert 2. Mission H. Damas (1935-1936), 15. 80 pp., 121 pis. Institut des Pares Nationaux du Congo Beige, Bruxelles.. Barnard, K. H. 1940. Entomological Expedition to Abyssinia, 1926-7 : Woodlice collected by Mr J. Omer-Cooper. Ann. Mag. nat. Hist., London, (11) 6 : 355-366, 8 figs. Bishop, W. W. & Trendall, A. F. 1967. Erosion-surfaces, tectonics and volcanic activity in Uganda. Q. Jlgeol. Soc. Land. 122 : 385-420, pis 13, 14. Budde-Lund, G. 1885. Crustacea Isopoda terrestria per Familias et Genera et species descripta. 320 pp. Havniae (Copenhagen). 1899. A revision of Crustacea Isopoda terrestria'' with additions and illustrations 1. Eubelum. 31 pp., 5 pis. Copenhagen. Clark, W. E. Le Gros & Leakey, L. S. B. 1951. The Miocene Hominoidea of East Africa. Fossil Mammals Afr., London, 1. 117 pp., 9 pis. Edney, E. B. 1954. Woodlice and the land habitat. Biol. Rev., Cambridge, 29 : 185-219. 1960. Terrestrial adaptations. In Waterman, T. H., Physiology of Crustacea, 1 : 367-393. New York. 1968. Transition from Water to Land in Isopod Crustacea. Am. Zooi, Utica, 8 : 309-326. Ferrara, F. & Schmalfuss, H. 1976. Terrestrial isopods from West Africa. Part 1 : family 'Eubelidae' Budde-Lund, 1899. Monitore zool. ital., Florence, (n.s.) Suppl. 7 : 1-114, 371 figs. Imaizumi, R. 1953. Note on Bathynomus sp. (Crustacea) from the Miocene of Japan. Short Pap. Inst. Geol. Paleont. Tohoku Univ., Sendai, 5 : 84-87. pi. 12. Leakey, L. S. B. 1953. [An exhibit of fossil insects from Miocene beds of Rusinga and Mwafangano islands in Lake Victoria, Kenya]. Proc. geol. Soc. Lond. 1498 : Ixxi. 1953o. Preserved in Three-Dimensional Form: Fossil Insects of 25,000,000 B.C. ///. Lond. News, March 21st, 1953 : 445. & Clark, W. E. Le Gros 1955. British-Kenya Miocene Expedition. Nature, Lond. 175 : 234. Maack, R. 1969. Kontinentaldrift und Geologic des siidatlantischen Ozeans. 164 pp. Berlin. Pulfrey, W. 1960. Shape of the sub-Miocene erosion bevel in Kenya. Bull. geol. Surv. Kenya, Nairobi, 3. 18 pp., map. Rafinesque-Schmalz, C. S. 1815. Analyse de la Nature. 224 pp. Palermo. Rathbun, M. J. 1935. Fossil Crustacea of the Atlantic and Gulf coastal plain. Spec. Pap. geol. Soc. Am., New York, 2. vii + 160 pp., 26 pis. Shackleton, R. M. 1951. A contribution to the geology of the Kavirondo rift valley. Q. J I geol. Soc. Lond. 106 : 345-389. Vandel, A. 1943. Essai sur 1'origine, 1'evolution et la classification des Oniscoidea (isopodes terrestres). Bull. biol. Fr. Belg., Paris, Suppl. 30. 136 pp., 77 figs. 1945. Le repartition geographique des Oniscoidea (Crustaces, Isopodes terrestres). Bull. biol. Fr. Belg., Paris, 79 : 221-272. 1960. Isopodes terrestres (Premiere Partie). Faune Fr., Paris, 64. 416 pp. Van Name, W. G. 1920. Isopods collected by the American Museum Congo expedition. Bull. Am. Mus. nat. Hist., New York, 43 : 41-108, 126 figs. Warburg, M. R. 1968. Behavioral Adaptations of Terrestrial Isopods. Am. Zool., Utica, 8 : 545-559. A re-evaluation of the fossil human calvaria from Singa, Sudan C. B. Stringer Subdepartment of Anthropology, British Museum (Natural History), Cromwell Road, London SW7 5BD Synopsis The first studies of the Singa calvaria concluded that it was an Upper Pleistocene 'proto-Bushman' fossil. More recently it has been suggested that it is a 'neanderthaloid' specimen or an anatomically modern specimen. This study concludes that the Singa calvaria is an unusual archaic Homo sapiens fossil, perhaps most closely related to the crania from Djebel Irhoud. Introduction The Singa calvaria (Fig. 1) was discovered by W. G. R. Bond in 1924, eroding from a caliche deposit within the 'Gezira clay' exposed in the west bank of the Blue Nile (Oakley, Campbell & Molleson 1977). The discovery was made in the Singa district of eastern Sudan, about 320 km south-east of Khartoum, and faunal and archaeological remains were also recovered, both at Fig. 1. Left lateral view of the Singa calvaria. BM(NH) Anthropology registration number M 1 5546. x 1. the Singa site and the related site of Abu Hugar, about 15km further south (upstream). The collection of faunal material contained some extinct species including an antilopine, a porcupine (Hystrix astasobae Bate 1951) and a bovine allocated by Bate to Homoioceras singae Bate 1949 (Bate 1951). The allocation of the bovine fossils from Singa and Abu Hugar to an extinct species of long-horned buffalo has recently been questioned, and it appears probable that the holotype skull from Singa belongs to a short-horned buffalo (Gentry & Gentry 1978). Nevertheless, Bate's conclusion that the fauna from Singa and Abu Hugar is of Upper Pleistocene age, and perhaps dates from an earlier phase of the Upper Pleistocene, still stands (Bate 1951). Bull. Br. Mus. not. Hist. (Geol.) 32 (1) : 77-83. Issued 28 June 1979 77 78 C. B. STRINGER Lacaille (1951) considered that the Singa and Abu Hugar implements had most affinity with the Middle Stone Age Proto-Stillbay industry of Lochard, Rhodesia, which would indicate on present dating evidence an earlier Upper Pleistocene age. However, Marks (1968) cautioned that the artefacts were nondescript, and although found in the same layer as the calvaria, were not necessarily of the same age. He agreed with Lacaille that some of the flakes might be 'Levallois' but remarked that Levallois flakes occurred in a variety of African Middle and Late Stone Age industries of widely-differing ages. He also commented that there were no grounds for comparison of the Singa artefacts with the Khormusan Levalloiso-Mousterian industry known from the Nile region. However, McBurney (1977) quoted a range of radiocarbon dates for the Khormusan from over 43 000 years BP (before present) to less than 20 000 years BP and suggested that the Singa calvaria might represent the population which produced what he considered to be 'the latest expression of the Mousterian so far known jn the world'. Whiteman submitted a crocodile tooth from Abu Hugar for radiocarbon dating, which produced an age of 17 300 ±2 000 years BP. Whiteman and Oakley considered that this late Pleistocene radiocarbon age might be due to contamination from the limestone in the Abu Hugar deposit or from the river Nile which regularly flooded its banks in the area (Whiteman 1971). The relatively low radiocarbon age is also incon- sistent with relative dating analyses which apparently confirmed the association of the Singa calvaria and the Abu Hugar fauna, but revealed that the bones contained only a negligible amount of collagen (Oakley et al. 1977) . Review of previous studies of the Singa calvaria The first description of the Singa calvaria was published by Woodward (1938), who concluded that the specimen was a brachycephalic variant of the 'Boskop type' of Homo sapiens known from the later Pleistocene of southern Africa. Modern Khoisan (Bushman) populations were regarded by Woodward as 'degenerate' survivors of this Boskop race. Wells (1951) compared the Singa calvaria with a wider range of Upper Pleistocene hominids, but similarly concluded that the Singa fossil was a north African variant of the proto-Bushman stock already known from sites such as Boskop and Matjes River. Briggs (1955) accepted its pre-Bushman affinities but also suggested that the Singa calvaria was comparable to the 'type C' male crania from the late Pleistocene/Holocene site of Afalou in Algeria. Anderson (1968) discounted any relationship between the Singa fossil and populations such as those represented at Afalou, and he emphasized that the latter groups were instead clearly linked to European Upper Palaeolithic and Mesolithic populations. A similar view was taken by Greene & Armelagos (1972), who compared the Singa calvaria with the Wadi Haifa Mesolithic crania from northern Sudan. They also saw little resem- blance between the Singa fossil and 'boskopoid' and 'bushmanoid' fossils from southern Africa. However, Wells (1972) reasserted that such a relationship existed, and suggested that the Singa fossil might be a paedomorphic descendant of the earlier Upper Pleistocene Omo (Kibish) population. Tobias, however, noted that the supposed 'bushmanoid' characters of the Singa specimen were restricted to the rear of the vault, whereas the anterior part of the calvaria was more archaic, even resembling the Broken Hill cranium in certain details (Tobias 1968). In a more extensive study, Brothwell (1974) also questioned the Khoisan features of the Singa fossil and suggested that its unusual cranial shape might be the result of a longitudinal growth defect affecting the parietal bones, producing a cranial shape which superficially resembled that of Bushman skulls. The cranial contours of the Singa specimen were compared to those of other Pleistocene crania, and the comparisons suggested that the Singa calvaria most resembled archaic (i.e. not anatomi- cally modern) Homo sapiens fossils such as the Gibraltar neanderthal cranium. Brothwell's data on cranial measurements were used in a multivariate principal component analysis conducted by Andrews, and the Singa specimen was placed as very distinct from various fossil and Recent Bushman crania as well as from a small sample of other Recent African crania (Brothwell 1974). Instead the analysis linked it with the Broken Hill cranium, but this was the only archaic Homo sapiens fossil included in the analysis from a range of relevant specimens. Brothwell concluded that the overall cranial morphology of the Singa calvaria lay closer to that of early Upper THE SINGA SKULL 79 Pleistocene crania from Europe, Asia and Africa than to late Pleistocene and Recent Khoisan specimens. He also suggested that the Singa fossil might represent a hybrid of 'neanderthaloid' and anatomically modern populations. However, Rightmire (1975) stated that in his view the Singa calvaria was not 'neanderthaloid' nor closely related to proto-Bushman or Bushman material from southern Africa. He agreed with Wells (1972) that it might be related to the Omo (Kibish) material, but concluded that it was probably a late Pleistocene example of robust but anatomically modern man. A new study of the Singa calvaria None of the studies of the Singa specimen have compared it in both metrical and morphological characteristics with a full range of relevant Upper Pleistocene crania as well as a range of more recent African material. In a multivariate study of Middle and Upper Pleistocene hominid crania using Mahalanobis' generalized distance statistic (D2) (Rao 1952 : 246), the Singa calvaria was compared with a wide range of other fossil crania (Stringer 1974). It was concluded that the specimen was not closely related metrically to modern Homo sapiens, but instead was more similar to the Homo erectus fossils from Ngandong (Java) and to archaic Homo sapiens crania such as the Broken Hill fossil and the European and south-west Asian Neanderthals. However, at that time it was not possible to use, in comparison, data from certain relevant fossil crania nor from large samples of Recent African populations (including Bushmen). Additionally, the multi- variate analyses used by Broth well (1974) and myself (Stringer 1974) had not distinguished factors of size from those of shape in assessing multivariate distances. The size and shape statistic of Penrose (1954) does allow such distinctions to be made whilst accurately reflecting true differences in size and shape (Corruccini 1973). This method of analysis is utilized here, modified by the use of logarithms of the raw cranial data rather than standardized data. This approach can be shown on theoretical and empirical grounds to give a shape distance which more truly reflects proportional change in shape, and it also avoids the problem of standardizing data from small or metrically divergent samples. The computation of the Penrose size and shape statistic in the case for P variables is based on the equation p- 2 # = & + £ 2 (dt - d}2 total distance size shape where di=Xi~ Xf', X( refers to log iih measure on the Singa calvaria and X- to log /th measure on another specimen or group. The total distance and size distance are calculated, and the shape distance is obtained by subtraction. Size distances which were negative before squaring are plotted separately in the results from those which were positive. Eleven cranial measurements (glabello-occipital length, maximum cranial breadth across parietals, bistephanic, biauricular and biasterionic breadths, frontal chord and subtense, parietal chord and subtense, and occipital chord and subtense) were taken on a range of Middle and Upper Pleistocene crania (samples generally as Stringer 1974). It was also possible to include for the first time comparative cranial data from the south-west Asian Upper Pleistocene sites of Qafzeh (where I have combined data from crania 6 and 9 with data from the comparable Skhul 5 specimen) and Shanidar (where I have combined data from Shanidar 1 with data from the comparable Amud 1 fossil to form an Asian Neanderthal group). Data from the robust Achegour 1 Mesolithic cranium were also included as this was representative of a late Pleistocene/early Holocene mid-Saharan population (Oakley et al. 1977). Large comparative samples of Recent African male and female crania were also included (sample size for each sex numbered between 34 and 58 individuals) using data collected by Howells (1973). However, it was not possible to use comparative data from the Sudanese Wadi Haifa Mesolithic crania (Greene & Armelagos 1972) since a different series of measurements were taken. The results of the size and shape analysis with the Singa calvaria as reference specimen are displayed in Fig. 2. It can be seen that the specimens and populations generally regarded as 80 C. B. STRINGER 'anatomically modern' are all very distinct in shape from the Singa fossil, although Bushman crania are closest to it of the Recent groups. Omo (Kibish) 1 and the Asian Neanderthal sample are also quite distant from the Singa specimen. However, the Homo erectus samples from Peking (Choukoutien) and Ngandong (Java), the archaic Homo sapiens crania from Djebel Irhoud, Saccopastore, Broken Hill and Omo (Kibish) 2, and the European Neanderthals, are all relatively close to the Singa calvaria in shape. Thus the Singa specimen apparently displays metrical fea- tures of Homo erectus and archaic Homo sapiens. •004 •002 N .Oml/2 As 4 .3 Neo. , ' B.H.t>0m2 .Pet .2 1 0-01 .0-02 0-03 0-04 . 0-05 Neo «| E 11 "•'•• SHAPE . -12 -002 So -•004 -.006 Sat 14 -•008 P.tc Fig. 2. Result of Penrose size and shape analysis using Singa calvaria as reference specimen. 1 - Qafzeh 6, 9 and Skhul 5; 2 - European Upper Palaeolithic group; 3 - North African Mesolithic group; 4 - AchegOur (Niger); 5, 6, male and female Egyptian; 7, 8, male and female Teita (Kenya); 9, 10, male and female Dogon (Mali); 11, 12, male and female Zulu; 13, 14, male and female Bushman. Si -Singa; D.I. - Djebel Irhoud 1; So -4 Ngandong (Solo) crania; Pek-4 Peking (Choukoutien) crania; Sac - Saccopastore 1 ; Nea - European Neanderthal group; B.H. - Broken Hill; Om-Omo (Kibish) crania, treated separately and as a group; Pet - Petralona ; As Nea- Amud 1 and Shanidar 1 ; I.E. - Iwo Eleru (Nigeria). Any attempt to classify the Singa fossil must take into account morphological, as well as metrical, characteristics. The shape analysis (Fig. 2) could not accurately reflect overall cranial morphology since the data employed are predominantly measured lengths of chords and sub- tenses in the sagittal plane and breadths of the lower part of the cranial vault. Shape resemblance between the Singa and Homo erectus specimens is a reflection of a short, relatively low cranial vault, broad at the base, a very short, flat parietal arch, and a protruding occipital bone. Differ- ences in shape are primarily caused by the larger occipital chord, parietal breadth and frontal dimensions of the Singa fossil. The much greater size of the Singa fossil is particularly marked in comparison with the Peking Homo erectus crania (Fig. 2). This is obviously related to the greater cranial volume of the Singa specimen. Previous estimates of the cranial capacity of the Singa fossil calculated from external measurements ranged from 1550 to 1600 ml (Wells 1951) and my own calculations confirm these figures. Using millet seed to measure the endocranial volume not occupied by matrix gave a value of approximately 1100ml, and examination of radiographs suggests that there is at least 300 ml of matrix remaining within the calvaria. It is hoped that the endocranial surface can eventually be cleaned to obtain a better estimate of endocranial volume, but it is clear that the cranial capacity of the Singa specimen lies well outside the range of 915- 1255 ml measured for eleven of the Peking and Ngandong crania (Day 1977). The Singa calvaria THE SINGA SKULL 81 also differs from Homo erectus crania in a whole range of morphological features. The supra- orbital torus in the Singa fossil is strongly developed laterally but does not form a thick continuous structure. The occipital bone, although projecting, is not strongly angulated, and the occipital plane is long compared to the nuchal plane, which is poorly defined. The occipital torus conforms to the Homo sapiens pattern, and those measurements of occipital and parietal thickness which can be taken fall within the range of values for Homo sapiens rather than for Homo erectus. Additionally the cranial vault displays none of the angularity and sagittal keeling characteristic of Homo erectus and Middle Pleistocene archaic Homo sapiens crania. In a recent study of Pleistocene hominids I investigated the use of angles and indices as a means of discriminating between Homo erectus, archaic Homo sapiens and anatomically modern Homo sapiens crania (Stringer 1978). Although there was often an overlap in ranges of values between the groups, it was usually possible to assign isolated specimens to one or other of these groups on the basis of their cranial angles and indices. Table 1 shows which angles and indices could Table 1. Classification of the Singa calvaria using ranges of variation in angles and indices for Pleistocene fossil hominid crania (position of Singa indicated by +) Z, u "60 60 0) 60 03 ffl Z PQ z £ £ O PQ Homo erectus (Peking and Ngandong) Archaic Homo sapiens (includes Petralona, Broken + Hill and Neanderthals) Anatomically modern Homo sapiens (includes + Qafzeh and Skhul) be determined for the Singa calvaria, and the group ranges within which the values for the Singa specimen fall (indicated by +). The nasio-frontal angle which measures the medial pro- jection of the upper face shows no discrimination in this case, but the other values appear to indicate a resemblance to archaic Homo sapiens crania. The only exception, the low value for the bregma angle, is related to the relatively short basion-nasion length of the Singa calvaria. How- ever, a relatively short basion-nasion length is also found in the Broken Hill cranium which has the lowest value for the bregma angle found in archaic Homo sapiens crania, and which also falls within the modern range of values for this angle. It seems certain that the Djebel Irhoud 1 fossil (which lacks the basion) would also have had a similarly low value for the bregma angle, sug- gesting that the Singa calvaria is in fact close to African archaic Homo sapiens crania in this feature. There are other characteristics in which the Singa calvaria resembles archaic rather than anatomically modern Homo sapiens. The temporal bone is relatively low and robust, the supra- orbital torus is broad and well-developed laterally and the interorbital area and malars are strongly built (Tobias 1968, Brothwell 1974 and my own observations). The mastoid processes are small and there is a well-developed occipitomastoid crest, as in many archaic Homo sapiens crania (Stewart 1964). But on the other hand the height of the occipital plane compared to the length of the nuchal plane is relatively large, as in anatomically modern Homo sapiens. 82 C. B. STRINGER Combining the metrical and morphological data given here it seems likely that Tobias (1968) and Brothwell (1974) are correct in questioning the proto-Bushman and modern affinities attri- buted to the Singa fossil. It is probably a relatively late archaic Homo sapiens specimen with an unusually short parietal arch which has affected the overall cranial shape (Brothwell 1974). Its affinities are more likely to lie with late archaic Homo sapiens fossils such as those from Omo (Kibish) and Djebel Irhoud rather than with earlier specimens such as the Broken Hill cranium. Whilst there seem to be no Upper Pleistocene crania which match the Singa calvaria closely in metrical characteristics it may be significant that the Djebel Irhoud 1 specimen is closest in overall Penrose (1954) distance (size + shape) to the Singa fossil. Both Djebel Irhoud 1 and the second specimen from the site are short broad crania with high frontal bones and relatively small brows compared with many other archaic Homo sapiens crania (Ennouchi 1968). Like the Singa fossil they also have relatively short deep infratemporal fossae, small mastoid processes and a large cranial capacity. It seems possible that the Singa calvaria in fact represents an unusual individual specimen drawn from a comparable archaic Homo sapiens population which lived in Sudan during the Upper Pleistocene. Acknowledgements I would like to thank the staff of the many institutes which allowed me to study fossil material in their care, and Professor W. W. Howells and Dr E. Trinkaus for providing comparative data used in this study. Dr M. Hills provided helpful discussion on the use of distance statistics in craniometric research. I would like to thank Mr R. Kruszynski for drawing Fig. 2 and the British Museum (Natural History) photographic unit for providing the photograph for Fig. 1. Funds towards the cost of travel for the collection of data were provided by the Medical Research Council, the Boise Fund, Oxford, and the British Museum (Natural History). References Anderson, J. E. 1968. Late Paleolithic skeletal remains from Nubia. In Wendorf, F., The prehistory of Nubia : 996-1040. Dallas, Fort Burgwin Research Center and Southern Methodist University Press. Bate, D. M. A. 1951. The mammals from Singa and Abu Hugar. In Arkell, A. J., Bate, D. M. A., Wells, L. H. & Lacaille, A. D., The Pleistocene fauna of two Blue Nile sites. Fossil Mammals Afr., London, 2 : 1-28. Briggs, L. C. 1955. The stone age races of northwest Africa. Bull. Am. Sch. prehist. Res., New Haven, 18 : 1-98. Brothwell, D. R. 1974. The Upper Pleistocene Singa skull: a problem in palaeontological interpretation. In Bernhard, W. & Kandler, A., Bevolkerungsbiologie : 534-545. Stuttgart. Corruccini, R. S. 1973. Size and shape in similarity coefficients based on metric characters. Am. J. phys. Anthrop., Washington, 38 : 743-753. Day, M. H. 1977. Guide to fossil man. xviii + 346 pp. London. Ennouchi, E. 1968. Le deuxieme crane de rhomme d'Irhoud. Annls Paleont., Paris, 54 : 117-128. Gentry, A. W. & Gentry, A. 1978. Fossil Bovidae (Mammalia) of Olduvai gorge, Tanzania. Part I. Bull. Br. Mus. nat. Hist., London, (Geol.) 29 : 289-446. Greene, D. L. & Armelagos, G. 1972. The Wadi Haifa Mesolithic population, xi + 146 pp., 23 pis. Amherst, University of Massachusetts. Howells, W. W. 1973. Cranial variation in Man. Pap. Peabody Mus., Cambridge, Mass., 67 : 1-298. Lacaille, A. D. 1951. The stone industry of Singa - Abu Hugar. In Arkell, A. J., Bate, D. M. A., Wells, L. H. & Lacaille, A. D., The Pleistocene fauna of two Blue Nile sites. Fossil Mammals Afr., London, 2 : 43-50. McBurney, C. B. M. 1977. Archaeology and the Homo sapiens sapiens problem in northern Africa (G. H. Kroon Memorial Lecture). 37 pp., 7 pis. Amsterdam, Netherlands Museum for Anthropology and Prehistory. Marks, A. E. 1968. The Khormusan: an Upper Pleistocene industry in Sudanese Nubia. In Wendorf, F., The prehistory of Nubia: 315-391. Dallas, Fort Burgwin Research Center and Southern Methodist University Press. Oakley, K. P., Campbell, B. G. & Molleson, T. I. 1977. Catalogue of fossil hominids. Part I. Africa (2nd ed.). xiii + 223 pp., 20 pis, 2 maps. London, British Museum (Natural History). THE SINGA SKULL 83 Penrose, L. S. 1954. Distance, size and shape. Ann. Eugen., London, 18 : 337-343. Rao, C. R. 1952. Advanced statistical methods in biometric research, xvii + 390 pp. New York. Rightmire, G. P. 1975. Problems in the study of later Pleistocene man in Africa. Am. Anthrop., Washington, 77 : 28-52. Stewart, T. D. 1964. A neglected primitive feature of the Swanscombe skull. In Ovey, C. D., The Swans- combe skull : 151-159. London, Royal Anthropological Institute. Stringer, C. B. 1974. A multivariate study of cranial variation in Middle and Upper Pleistocene human populations. Ph.D. Thesis, University of Bristol. - 1978. Some problems in Middle and Upper Pleistocene hominid relationships. In: Olivers, D. J. & Joysey, K. A. (eds), Recent advances in Primatology, 3 (Evolution) : 395-418. London. Tobias, P. V. 1968. Middle and early Upper Pleistocene members of the genus Homo in Africa. In Kurth, G., Evolution und Hominisation : 176-194. Stuttgart. Wells, L. H. 1951. The fossil human skull from Singa. In Arkell, A. J., Bate, D. M. A., Wells, L. H. &Lacaille, A. D., The Pleistocene fauna of two Blue Nile sites. Fossil Mammals Afr., London, 2 : 29-42. 1972. Late Stone Age and Middle Stone Age tool-makers. S. Afr. archaeol. Bull., Claremont, 27 : 5-9. Whiteman, A. J. 1971. The geology of the Sudan Republic, xiv + 290 pp., 22 pis. Oxford. Woodward, A. S. 1938. A fossil skull of an ancestral Bushman from the Anglo-Egyptian Sudan. Antiquity, Gloucester, 12 : 193-195. New species of Protorthoptera and Protodonata (Insecta) from the Upper Carboniferous of Britain, with a comment on the origin of wings P. E. S. Whalley Department of Entomology, British Museum (Natural History), Cromwell Road, London SW7 5BD Synopsis Narkeminopsis eddi gen. et sp. nov. (Protorthoptera) and Erasipteron bolsoveri sp. nov. (Protodonata) are described. The use of wings for functions other than flight in the earliest insects is discussed. Origin of wings New insect specimens from the Carboniferous of Britain are not common. The species of Proto- donata from Derbyshire described here is one of the oldest Carboniferous fossil insects found in Britain, the only earlier British fossil insect being the wingless Rhyniella from the Devonian. The two new species described show the complex wing venation typical of these early winged insects. With such specialized forms flying, we have to look further back in time for the origins of wings. Even if Rhyniella is a genuine fossil insect, and this is still disputed (Crowson 1970), it is unlikely to have been ancestral to the flying forms since the proponents of its insect affinities suggest that it is a collembolan. 1 Fig. 1. Narkeminopsis eddi gen. et sp. nov. Upper Carboniferous (Westphalian D), Radstock, near Bristol. Wing, 25-5 mm x9 mm. There are no insect fossils yet known for the critical period in insect evolution from the Upper Carboniferous back into the Devonian, when the wingless Rhyniella occurs; thus we can still only speculate on the factors which stimulated the production of wings. However, I believe there is evidence available from Recent insects which has been overlooked and which can be used to support a multiple-factor theory for the origin of wings. With the discovery of the circulation of the haemolymph in the wing veins of insects, first reported in 1934 by Yeager & Hendrickson, we can consider that wings may have evolved not Bull. Br. Mus. not. Hist. (Geol.) 32 (1) : 85-90. Issued 28 June 1979 85 86 P. E. S. WHALLEY only for flight, and perhaps gaseous exchange as suggested by Portier (1930), but also as heat- exchangers. Clench (1966) describes the use of wings in butterflies for heat exchange, and Gillett (1971) also suggests an alternative to their use in flight. If these facts are applied to the earliest insect, it is possible to consider that outgrowths of the body wall, which would have been developed to help increase the surface-area/volume ratio, could have been used as a sort of aerial gill and heat-exchanger. They would need some muscula- ture to move them for orientation, or even to allow more efficient heat or gaseous exchange. The outgrowths or 'proto wings' could then have been developed concurrently with flight, which became their most important function. It is not suggested that this is directly in line with the tracheal-gill theory (see Wigglesworth 1976) for which an aquatic stage is postulated, but merely that flight does not have to be considered as the only function of the early 'wing'. Fig. 2. Erasipteron bolsoveri sp. nov. Upper Carboniferous (Westphalian A), Derbyshire. Wing, 87 mm x!2 mm. Systematic descriptions Order PROTORTHOPTERA (including Protoblattoidea and Paraplecoptera) Carpenter (1966) considered there were few essential differences between species formerly placed in the Protoblattoidea and Paraplecoptera, and those in the Protorthoptera ; the latter name has priority. The species described below is placed in the Protorthoptera following Carpenter (1966), although the limits of the order still need to be defined. Family NARKEMIDAE Handlirsch, 1911 Professor Carpenter (in litt.) suggests that Narkeminopsis gen. nov. and Narkemina Martynov are sufficiently distinct from the poorly-preserved Narkema taeniatum Handlirsch, the type of the family Narkemidae, to be placed in a new family. N. taeniatum does show bandings of light and dark across the wing as in N. eddi sp. nov., but the dark bands are much narrower and there are CARBONIFEROUS INSECTS 87 differences in the origin of the median veins. From Narkemina angustatum Martynov (the type- species of Narkemina) N. eddi differs in the narrower costal area and fewer cross-veins, and has a longer basal part to the median vein (Carpenter, in litt.). The basal part of the wing of N. eddi (Fig. 3) clearly shows the origin of CuA and CuP. Pending further research, no new family is proposed for the new genus Narkeminopsis, which is provisionally placed in the Narkemidae (Upper Carboniferous). Genus NARKEMINOPSIS nov. TYPE-SPECIES. Narkeminopsis eddi sp. nov. DIAGNOSIS. Forewing. Costal margin smooth, gently curved. Sc with several unbranched oblique veins. Sc terminates on RI. Rs forks from R roughly halfway to termen. Costal margin narrower, cross-veins less numerous and basal part of median vein shorter than in Narkemina Martynov. Hindwing unknown. q CuP t CuA Fig. 3. Narkeminopsis eddi, wing venation, diagram. + convex veins; — concave veins. Narkeminopsis eddi sp. nov. (Figs 1,3) DIAGNOSIS. Forewing. 25 -5 mm x 9 mm. Wing impression (no counterpart) preserved upside down. Rs with several branches. Main wing cells distinctly spindle-shaped. MA and MP present, apical parts missing. CuA with several branches, CuP single to wing margin. Cross-veins promi- nent. Archedictyon present. Wing markings prominent as strong, dark, parallel cross-bands, each approximately 2 mm wide, at a slight angle across wing. HOLOTYPE. BM(NH) In.64531. Upper Carboniferous (Westphalien D); Radstock, nr Bristol, Avon; collected by E. Jarzembowski. DISCUSSION. The archedictyon on the wing is only faintly visible and can be seen between the branches of CuA and along part of the costal margin. The bands of darker colour are well defined and relatively smooth while the lighter areas are rougher, suggesting a more delicate (?less sclerotized) membrane which has virtually disappeared, leaving just the impressions of the wing veins. The transverse banding may well represent part of the original wing pattern. Order PROTODONATA (Meganisoptera) DIAGNOSIS. Dragonfly-like insects which lack the nodus, arculus and pterostigma of typical Odonata. Family ERASIPTERIDAE Carpenter, 1939 DIAGNOSIS (modified after Carpenter, 1939). CuA present, reduced. Anal-crossing vein between CuP and 1A. MP absent. Archedictyon present. Basal part of costa of forewing toothed. One included genus. 88 P. E. S. WHALLEY Genus ERASIPTERON Pruvost, 1933 TYPE-SPECIES. Erasipteron larischi Pruvost 1933, by monotypy. Upper Carboniferous (Namurian C), Czechoslovakia. DIAGNOSIS. Relatively few cross-veins. Sc roughly half length of forewing. Rs strongly branched in apical area of wing. Fig. 4. Erasipteron bolsoveri, reconstruction CuA Fig. 5. E. bolsoveri, basal third of wing, venation diagram. +convex veins; — concave veins. Erasipteron bolsoveri] sp. nov. (Figs 2, 4, 5) DIAGNOSIS. Forewing. 87 mm x 12 mm. No pattern visible. Concave and convex veins clear (Fig. 5). RI terminates near wing apex. Rs sharply curved down at apex. Rs and MA fork one- third from base. Anal veins curved in basal area. Anal cells simple, single row. Hindwing unknown (but see Discussion). HOLOTYPE. BM(NH) In.64532. Upper Carboniferous (Westphalien A); Derbyshire, Bolsover deep hard seam, at 586 m, approx. 15 cm above seam. Collected by M. Spencer. DISCUSSION. This species can be separated from E. larischi by the presence of cells in the area formed by the anal-crossing vein, and by the simple arrangement of cells in the basal part of the anal area. An incipient nodus is supposed to be present in E. larischi (Carpenter 1939; Kukalova 1964); no trace of a nodus can be seen in E. bolsoveri, but there is some damage to this part and the presence of a nodus cannot be entirely ruled out. The specimen from Derbyshire is the second species of the genus Erasipteron Pruvost, family Erasipteridae sensu Carpenter (1939). The type-species, Erasipteron larischi Pruvost, was described from a single hindwing from the Upper Carboniferous (Namurian C) of Czechoslovakia, and redescribed by Kukalova (1964); the new species is based on a forewing. The type specimen of the new species probably consists of two wings, although only one is immediately visible. The two wings lie exactly one above the other, separated by a matrix of 0-5-1-0 mm. These wings were broken in the basal third, the two parts lying at right-angles to t This name has already appeared a number of times in the popular press, but without formal diagnosis. CARBONIFEROUS INSECTS 89 one another (Fig. 2). The double nature of the fossil can be seen one-third from the apex, where some of the overlying wing has been removed. The lower wing matches exactly the size and vein positions of the upper wing. There is a trace of the underlying wing near the base, but this is not as clear as the apex. Two possibilities have to be considered. A single wing may have become separated into an upper and lower wing membrane and the space between filled up like a sac; the resulting 'double wing' would therefore be only the upper and lower membranes of one wing. Alternatively there are two wings which have been fossilized exactly together. From an examination of the specimen the latter explanation seems more probable; for example, the thickness of the matrix between suggests two wings rather than an upper and lower surface of one wing. It is difficult to tell from the amount of the lower wing revealed whether it is a fore or hind wing, but it is, in the apical part at least, an exact replica of the upper wing but with most of the convex-concave vein system flattened out. It is possible that the convex-concave system is identical with the upper wing. If this is correct then the lower wing is likely to be the hindwing from the same side as the upper wing (which I regard as a forewing), thus indicating that the insect was homoneurous, with a simpler anal area than in E. larischi. This would provide con- clusive evidence of the two species being generically distinct. Unfortunately this hypothesis is based on the condition of the convex-concave state of the lower wing, which is not clearly visible. With the information available I do not propose to describe a new genus but to redefine Erasipteron Pruvost. It is possible that there were differences between the forewings and hind- wings in species of Erasipteron, which would explain some of the discrepancies between E. larischi (hindwing) and E. bolsoveri (forewing), but data are not available to support or contradict this. The new species may well be homoneurous but the facts are inconclusive. E. bolsoveri is the most complete specimen of the larger, dragonfly-like Protodonata found in Britain, with an estimated wingspan of nearly 200 mm (8 inches), larger than any extant species. Acknowledgements I am grateful to Mr M. Spencer and Mr T. Judge, who presented the Protodonata to the Museum. The North Derbyshire Area of the National Coal Board provided the photograph, which was taken by Mr V. P. Gorswill. Professor F. M. Carpenter and Dr J. Kukalova-Peck provided help and advice on the interpretation of the specimens ; to both I offer my thanks. I am grateful to Dr M. Calver (Chief Palaeontologist, Institute of Geological Sciences) who offered help and advice. Mr C. J. Wood (I.G.S.) first drew my attention to the Protodonata. To my colleagues who have given useful advice I offer my thanks. Bibliography References marked* are not cited in the text, but are included as contributing to the bibliography of the subject. "Carpenter, F. M. 1931. Lower Permian insects of Kansas, Part 2. Am. J. Set., New Haven, 21 : 97-139. * 1933. Lower Permian insects of Kansas, Part 6. Proc. Am. Acad. Arts Sci., Boston, 68 : 411-503. 1939. Lower Permian insects of Kansas, Part 8. Proc. Am. Acad. Arts Sci., Boston, 73 : 29-70. * 1960. Studies on North American Carboniferous insects. Protodonata. Psyche, Camb. 67 : 98-110. 1966. Lower Permian insects of Kansas, Part 11. Protorthoptera and Protodonata. Psyche, Camb. 73 : 46-88. Clench, H. 1966. Behaviorial thermoregulation in Butterflies. Ecology, Brooklyn 47 : 1021-1034. Crowson, R. A. 1970. Classification and Biology, ix + 350 pp. London. * , Rolfe, W. D., Smart, J., Waterstone, C. D., Willey, E. C. & Wotton, R. J. 1967. Arthropoda: The Chelicerata, Pycnogonida, Palaeoisopus, Myriapoda and Insecta. In Harland, W. B. et al. (eds), The Fossil Record : 499-528. London. Gillett, J. D. 1971. Mosquitos. xiii+274 pp., 22 figs. London (World Naturalist series). Handlirsch, A. 1911. New Palaeozoic insects from the vicinity of Mazon Creek, Illinois. Am. J. Sci., New Haven, 31 : 323. 90 P. E. S. WHALLEY Kukalova, J. 1964. To the morphology of the oldest known Dragonfly, Erasipteron larischi Pruvost. Vest, fared. Ust. geol., Prague, 39 : 463-464. *Martynov, A. 1924. Sur 1'interpretation de la nevuration et de la tracheation des ailes des Odonata et des Agnathes. Ent. Obozr., Moscow, 18 : 147-174. Pruvost, P. 1933. Un Ancetre des Libellules dans le Terrain houiller de Tchecoslovaquie. Annls Soc. geol. N., Lille, 58: 149-155. Portier, P. 1930. Respiration pendant le vol chez les Lepidopteres. C. r. Seanc. Soc. Biol., Paris, 105 : 760-764. *Tillyard, R. 1928. The evolution of the order Odonata. Rec. Indian Mus., Calcutta, 30 : 151-172. Wigglesworth, V. B. 1976. The evolution of insect flight. In Rainey, R. C. (ed.), Insect Flight. Symp. R. ent. Soc. Lond. 7 : 255-269. *Wootton, R. J. 1976. The fossil record and insect flight. In Rainey, R. C. (ed.), Insect Flight. Symp. R. ent. Soc. Lond. 7 : 235-254. Yeager, J. F. & Hendrickson, G. O. 1934. Circulation of the blood in the wings and wing pads of the cockroach, Periplaneta americana Linn. Ann. ent. Soc. Am., Washington, 27 : 257-272. Monographs and other BM(NH) Publications related to Geology series Vol. 32 No. 1 Monographs Catalogue of fossil Hominids Pt. I.Africa. Ed. by K. P. Oakley, B. G. Campbell & T. I. Molleson 2nd edition 1977 A History of the classification of the phylum Brachiopoda H. M. Muir-Wood. 1955 (reprinted 1968) On the morphology and classification of the brachiopod suborder Chonetoidea. H. M. Muir-Wood 1962 Geology Bulletin Supplements No 6. Upper Jurassic rhynchonellid brachiopods from North- western Europe A. Childs 1969 No 11. Ordovician Brachiopoda from the Shelve District, Shropshire. A. Williams 1974 A complete list of titles in the Geology series is available Titles to be published in Volume 32 Miscellanea Palaeoenvironments and correlations of the Carboniferons rocks in west Fermanagh, Ireland. By C. H. C. Brunton & T. R. Mason. The Ordovician trilobite faunas of the Builth-Llandrindod Inlier, central Wales. Part III. By C. P. Hughes. The stratigraphy and brachiopods of the upper part of the type Caradoc of south Salop. By J. M. Hurst. The entire Geology series is now available Printed by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) Palaeoenvironments and correlations of tf Carboniferous rocks in west Fermanagh, Ireland C. H. C. Brunton & R. Mason Geology series Vol 32 No 2 30 August 1979 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique and ever-growing collections of the Museum, both by the scientific staif of the Museum and by specialists from elsewhere who make use of the Museum's resources. Many of the papers are works of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself, available separately, and individually priced. Volumes contain about 300 pages and are not necessarily completed within one calendar year. Subscriptions may be placed for one or more series. Subscriptions vary according to the contents of the Volume and are based on a forecast list of titles. As each Volume nears completion, subscribers are informed of the cost of the next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to : Publication Sales, British Museum (Natural History), Cromwell Road, London SW7 5BD, England. World List abbreviation: Bull Br. Mus. nat. Hist. (Geol.) Trustees of the British Museum (Natural History), 1979 ISSN 0007-1471 Geology series Vol 32 No 2 pp 91-108 British Museum (Natural History) Cromwell Road London SW7 5BD Issued 30 August 1979 Palaeoenvironments and correlations of the Carboniferous rocks in west Fermanagh, Ireland C. H. C. Brunton Department of Palaeontology, British Museum (Natural History), Cromwell Road, London SW7 5BD T. R. Mason Geology Department, University of Natal, Durban, Republic of South Africa Contents Synopsis ............. 91 Introduction ............. 91 Rock stratigraphical names .......... 93 Groups 93 Formations 95 Stratigraphical succession .......... 96 The Ballyshannon Group 96 Tournaisian . . . . . . . . . . .96 Visean 96 The Dartry Group 97 The Leitrim Group ........... 99 Correlations ............. 103 Discussion ............. 105 Acknowledgements . . . . . . . . . . .106 References ............. 106 Geological map of west Fermanagh ....... in rear pocket Synopsis The Carboniferous (Dinantian and low Namurian) rocks of west Fermanagh, south of Lower Lough Erne, are correlated with strata from surrounding areas in north-western Ireland and divided into forma- tions dated faunally and by the recognition of major sedimentary cycles. The proposed Ballyshannon Group consists of marine shales and limestones with two deltaic sandstone intercalations; the Dartry Group (new name) is entirely marine and includes a mud-mound 'reef phase; the Leitrim Group is of mixed marine, intertidal, lagoonal and deltaic sediments and includes sabkha deposits which may be correlated with those of eastern Canada. Introduction Since 1952 the Carboniferous geology of several areas of north-west Ireland has been described, but west Fermanagh has remained largely neglected since the original map (sheet 44, 1884) of the Geological Survey of Ireland and the Memoir (Symes & Wilkinson 1886) were published. The region investigated lies between Lower Lough Erne and the borders of counties Leitrim and Cavan (Fig. 1). Our purpose is to describe those features of the Carboniferous succession in the area which allow these rocks to be interpreted in terms of palaeoenvironments and permit their correlation with adjacent areas and with the Major Cycles of sedimentation described by Ramsbottom (1973). This paper results from work carried out by the authors indpendently from 1962 to 1974. Bull. Br. Mus. not. Hist. (Geol.) 32 (2) : 91-108 Issued 30 August 1979 91 92 C. H. C. BRUNTON & T. R. MASON GEOLOGY OF NORTH-WEST IRELAND Showing the position of Fermanagh and adjacent areas Upper Visean and Namurian shales & sandstones Tournaisian and Visean limestones S Londonderry Letterkenny |^x Intrusive Igneous Oswald 19,57 Donegal Ireland, political boundaries Padgef*>c\Lough Brandon 1968 . Fig. 1. Generalized geological map of north-western Ireland showing the area described herein together with the names of authors (with dates) who have described the surrounding regions. See also Figs 5 and 6. (After Geological Survey of Ireland, 1962.) CARBONIFEROUS OF WEST FERMANAGH 93 T. R. M. is generally responsible for the Leitrim Group and C. H. C. B. for the older strata. During this time we have described elements of the faunas: brachiopods by Brunton (1965, 19660, b, 1968), fish by Gardiner & Mason (1974), while bryozoans have been studied by Tavener- Smith (1965, 1973). A small syncline, preserving the youngest strata, has been described by Brandon (1972), and Wilson (1964) discussed some of the large Tertiary dykes in the area. Sheridan (1972) has presented a review of work on the Upper Old Red Sandstone and Lower Carboniferous rocks of north-west Ireland, in which he provided some useful information on a borehole sunk in west Fermanagh which penetrated to unexposed marine Tournaisian rocks. Lower to middle Visean strata mainly occupy areas of low ground below about 1 50 m and are not well exposed. The uplands to the south provide extensive dip and scarp exposures in the mid- to upper Visean limestones and sandstones, although the interbedded and overlying shales, which extend into the Namurian, are commonly limited to stream-section exposures. The exposures in west Fermanagh display the succession more completely than those of adjacent areas and four major cycles of transgression and regression from low Visean (Vlb) up to the low Namurian (El) are recognized. Most of the formation names are derived from the areas of Sligo, described by Oswald (1955), or Leitrim studied by Brandon (1972), but we introduce new stratigraphical Group names. After completion of this work the Geological Society of London published its report on 'A correlation of Dinantian rocks in the British Isles' (George et al. 1976). This proposed regional stage names corresponding to the six Major Cycles of Ramsbottom (1973), and we have added these names to Fig. 2. We have not emended our script since the stage names correspond to the cycles of sedimentation we discuss and because we disagree with the positioning of some of the cycle/stage boundaries suggested by George et al. for north-west Ireland. In general the cycles of sedimentation (Ramsbottom 1973) start with marine, commonly transgressive, sediments. The regional stages generally follow this concept, but in north-west Ireland George et al. (1976) place the start of some regional stages at the bases of sandstone sequences we believe to be of regressive, deltaic origin. Thus, there appears to be some inconsistency in the application of the stage boundaries and we feel it is inappropriate at present to define their positions stratigraphically in north-west Ireland; compare Fig. 2 with George et al. (1976: fig. 15). Rock stratigraphical names Groups The generalized succession in west Fermanagh is illustrated in Fig. 2. The rock units from the Ballyshannon Limestone to the Dergvone Shale are here considered as formations. These are divided into three Groups: the Ballyshannon Group, from the base of Ballyshannon Limestone up to the top of the Mullaghmore Sandstone; the Dartry Group, which finishes at the base of the Meenymore Formation, and the Leitrim Group, a unit named by Brandon (1972) for the Meenymore to Dergvone Shale Formations but without a defined upper limit. (In his thesis of 1968 Brandon assigned strata from the Glenade Sandstone up to the younger Lackagh Sandstone to the Leitrim Group, and all units below this to what he called the Sligo Group, a name which hitherto remained unpublished.) The Ballyshannon Group includes the oldest known Carboniferous rocks in north-west Ireland. These are commonly exposed as basal conglomerates of the Ballyshannon Limestone, at the base of the Bruckless Grits and Conglomerates in Co. Donegal, or as the Boyle Sandstone of the Carrick-on-Shannon area. The relatively abrupt change from the Mullaghmore Sandstone to the Benbulben Shale marks the base of the Dartry Group. Although there is no indication of unconformity between the two, the rocks of the Dartry Group are more consistently marine limestones than those of the Ballyshannon Group. Brandon (1972) mentions the Dartry Lime- stone Group, but without definition, and since we include formations other than the Dartry Limestone we omit Limestone from the Group name. Brandon's unpublished thesis name 'Sligo Group' could be used as a super-group to unite the Ballyshannon and Dartry Groups. 94 C. H. C. BRUNTON & T. R. MASON NAMURIAN 1000 scale '2 V3c V3b V3a V2b V2a Vlb Trans Fermanagh succession south of the Highland Boundary Fault zone showing correlation with major sedimentary cycles of Ramsbottom (1973) and stage names proposed by George et al. (1976) O " Q c •J o «£ o ft X) oo •P! — * 3 O 2 "o -* X u H IS CARBONIFEROUS OF WEST FERMANAGH 95 The junction between the Dartry Limestone Formation and the Meenymore Formation varies from apparent conformity to erosional disconformity. This junction marks the base of the Leitrim Group but the youngest formations are not exposed in west Fermanagh. Formations The formation names follow the rock unit names proposed by Oswald (1955) up to the Dartry Limestone, and by Brandon (1972) from there upwards, with the following exceptions. Basal Clastics is used informally to denote the unexposed mixed basal Carboniferous rocks in west Fermanagh, known only from rather scant bore-hole information. Similarly, the Dowra Sand- stone was named from Dowra no. 1 bore-hole, situated north-east of Lough Allen (Sheridan 1972). In this bore-hole the sandstone is 114m thick from a depth of about 1270 m. The type localities of the rock units described by Oswald (1955), and here accorded formation status, are as follows: 1. Ballyshannon Limestone - at Ballyshannon and the lower reaches of the river Erne, about 20 miles (32 km) north-west of Sligo. 2. Bundoran Shale -the foreshore and north side of Bundoran bay, about 16 miles (26km) north-west of Sligo. 3. Mullaghmore Sandstone - the headland of Mullaghmore, 14 miles (about 23 km) north of Sligo. 4. Benbulben Shale - exposures on the lower slopes of the mountain bearing this name (spelt Benbulbin on many maps), about 6 miles (10 km) north of Sligo. 5. Glencar Limestone - the north side of Glencar where the formation is well exposed on steep southern flanks of the Benbulben range. 6. Dartry Limestone - this takes its name from the Dartry mountains, between the Benbulben range and Lough Melvin to the north. The formation occurs in most of the highland areas of counties Sligo and Leitrim and is well exposed to the south of Glanade Lough, between the Benbulben and Dartry ranges. The limestone is typically dark to light grey, often crinoidal, and chert-bearing in contrast to the lime mud mound facies (reefs of previous authors) developed within this formation. The Meenymore Formation was described by Brandon (1972), who points out that it is equivalent to the oldest part of the Aghagrania Formation as described by West, Brandon & Smith (1968). In part, this formation is also equivalent to the Glenade Shales and Roscunnish Shales of Oswald (1955) and Caldwell (1959) respectively. The name is taken from Meenymore Townland, Glenboy, about 3 miles (5 km) ESE of Manor Hamilton, Co. Leitrim. The formation appears to be thin in this area and we suggest that the type locality, which has not yet been established, should be at the Aghagrania locality described by West, Brandon & Smith (1968) as the type locality of the Aghagrania Formation. The Aghagrania stream section lies about 1-5 miles (about 2-5 km) ENE of Drumshambo at the south end of Lough Allen, Co. Leitrim, where West, Brandon & Smith recorded 38-4 m of Meenymore laminated dolomite, limestones and shales. The Glenade Sandstone was named by Oswald (1955) for the sandstones capping parts of the Dartry mountains north of Glenade. However, in this region no younger formations are known so its upper boundary cannot be established. Brandon (1972) used the term for equivalent sand- stones in Co. Leitrim and estimated their thickness to be about 200 m in the Doagh Outlier region of west Fermanagh. We agree with this figure and believe our area provides a more complete sequence than exists in the Dartry hills. In Fermanagh the passage from the Meenymore to Glenade formations can be observed in sections of streams flowing north off the sandstone uplands, about 0-75 mile (1-2 km) south-west of Knockmore summit, 2-5 miles (4 km) WSW of Derry- gonnelly village. The top of the Glenade Formation is difficult to demonstrate owing to poor exposures at the base of the Bellavally Formation. However, between the sandstone hill of Trustia (993 ft (303 m) high and 5 miles (8 km) slightly south of west of Derrygonnelly) and Doagh school (about 1 mile (1-7 km) to the north) there are stream exposures of Bellavally shales and Doobally sandstone. Our interpretation (Mason 1978) differs slightly from that of Brandon (1972), who did not recognize the shales and Doobally sandstones of the Bellavally Formation. 96 C. H. C. BRUNTON & T. R. MASON Stratigraphical succession The Ballyshannon Group Tournaisian. Rocks of proven Tournaisian age are not exposed in the area but are known from the deep bore-hole at Big Dog, about 9 km WSW of Derrygonnelly, from a depth of over 1500 m. This bore, drilled by Marathon Petroleum (Ireland) Ltd, penetrated between 500 and 600 m of sandstones, shales, thin limestones and anhydrites, thought on macro- and micropalaeontological evidence to be of Upper Tournaisian (Tn3c) age by Sheridan, Hubbard & Oldroyd (1967), before being abandoned while still in marine strata. This great thickness of mainly arenaceous Carbon- iferous sediment contrasts with the few metres of basal .conglomerates and dolomites reported (Oswald 1955) only about 15 km to the north-west of Big Dog in the Ballyshannon region, at up to 50 m above sea-level. Here these basal Carboniferous rocks are of low Visean age (Sheridan 1972). Thus over about 15 km there is a fall in the base of the Carboniferous of nearly 2000 m and a corresponding increase in its age. In addition, the underlying strata in the Ballyshannon region are metamorphosed Moinian schists, whilst to the south-east, although not proven in the Big Dog bore, folded Lower Devonian and Old Red Sandstone rocks underlie the Carboniferous. Sheridan (1972) discussed the possibility of this apparently steep palaeoslope being the result of pene- contemporaneous fault movement but concluded that block faulting of an early Visean age resulted in the removal of the Tournaisian and Old Red Sandstones from the Ballyshannon horst region. It is becoming increasingly clear that some rocks in the British Isles much resembling the upper Old Red Sandstone (see Gayer et al. 1973) are really Tournaisian in age; these arenaceous transitional beds may be dated using spores, as has been done in Belgium. The lithologies of the lowest 600 m or so of the Big Dog bore indicate shallow marine and lagoonal conditions. Taken together, these two points lead to the conclusion that probably there was a long mid- and lower Tournaisian history of slow deposition in the area while the Old Red Sandstone lands gradually subsided. Periodically the lands were flooded with marine waters which led to the development of evaporites. Tournaisian subsidence to the south-east of the Highland Boundary Fault line was probably greater than that in the Ballyshannon region so that only thin, very condensed, erosional and regressive sediments older than Vlb remain on this horst. These, as yet poorly known, subsurface beds correlate with at least the upper part of Ramsbottom's (1973) First Major Cycle and may also include rocks within at least the lower part of his Cycle Two, which are of Via age on the Belgian scale. Visean. Major Cycle Three, starting probably at the base of Vlb (late C2 of the coral-brachiopod zonation), spread northwards into north-west Ireland as an important transgression and resulted in the deposition of the Ballyshannon Limestone. The regressive phase is represented by the Lower Bundoran Shale and the sporadically-developed Dowra Sandstone. The Ballyshannon Limestone strikes east-west from the type area into Co. Fermanagh, bordering the Moinian basement to the north, and in the region of Lower Lough Erne is repeated on the south shore by faulting. In our area this repeated section is exposed and extensively quarried from near Enniskillen in a north-westerly direction to Blaney. Being a relatively impervious and hard limestone it forms part of the south-west shore of the lake. Fine examples of Delepinea destinezi (Vaughan) have been described from the exposures on the north shore of the Lough (Pettigo Limestone of Simpson 1953) and date these rocks as from the C2S! Zone. The Dowra Sandstone was recorded from the Big Dog bore and, if correlated with the base of the Clonelly Sandstone of the Omagh basin (Simpson 1955), it represents a southerly extension of the conglomeratic and deltaic sandstones which spread from the north. Surface exposures of pre-upper Visean (V3) sandstones are uncommon in west Fermanagh and the separation of those attributed to the Dowra Sandstone and those correlated with the Mullaghmore (Derrygonnelly or Macnean) Sandstone is difficult. Both may contribute to the topographical ridge extending from near Enniskillen to Derrygonnelly, where the undoubtedly younger of the two is exposed, marking the upper regressive phase of Cycle Four. At Derrygonnelly this sandstone is well- sorted and contains about 50% calcite cement; it is between 100 and 200 m thick. In addition to CARBONIFEROUS OF WEST FERMANAGH 97 the microfauna reported by Sheridan (1972) a transported macro fauna of bryozoans, brachio- pods, coiled cephalopods and bivalves has been collected from temporary excavations which show the deposit to be fully marine. The environment in this area was probably a delta front, while coarser, less calcareous near-shore sands accumulated to the north. These sands, like the older Dowra Sandstone, were derived from the north and north-west, principally from the Moine and Dalradian areas of Co. Donegal. They extend from Donegal (the Mountcharles Sandstone), via the Mullaghmore type area on the Sligo coast, in a wide belt (known principally from bore-hole evidence) to the east, where they re-appear in the Dungannon region. The shales below and above the Mullaghmore Sandstone, the Upper Bundoran and Benbulben Shales respectively, are poorly exposed in west Fermanagh. They occupy low boggy ground, commonly covered by glacial drift and alluvium. The Dartry Group The shales and argillaceous limestones of the Benbulben Shale Formation represent the return to open marine conditions at the start of Major Cycle Five (V3b). This is an important cycle in north-western Ireland, its rocks are varied and in west Fermanagh increasingly well exposed. No junction between the Mullaghmore Sandstone (Derrygonnelly Sandstone locally) and Benbulben Shale can be seen in west Fermanagh and the shale is known only from small stream sections, principally in the Magho area, close to Lough Erne. It is a blue-grey muddy shale containing sparse thin limestones and bands of concretions, some of which contain a rich fauna of marine invertebrates. The gradational passage from the Benbulben Shale to the Glencar Limestone can be seen in the Magho area and from this point on a complete succession can be built up from a number of localities. The Glencar Limestone succession in Fermanagh is very similar to that in Co. Sligo, consisting of approximately equal proportions of interbedded argillaceous limestones and shales, as described by Schwarzacher (1964) from the type area. These rocks are commonly very fossilifer- ous, especially in their upper sections, containing abundant brachiopods, bryozoans, crinoids and (more locally) corals, and it is from the highest beds of this Formation that finely silicified brachio- pods and bryozoans have been described by Brunton (19660, b, 1968) and Tavener-Smith (1965, 1973). Palaeocurrent study indicates that clastic sediment was probably all derived from the north and north-west, with open shallow seas extending to the south, whilst in the Sligo area the Ox Mountains appear to have contributed sediment along with that from the Donegal region (Schwarzacher 1968). The full thickness of the Glencar Limestone is estimated at about 150 m, and towards the top the proportion of limestone to shale increases with thin beds of 'reefal' calcilutite or limestone conglomerates (Schwarzacher 1961) being found in areas where lime-mud mounds developed in the overlying Dartry Limestone. The Glencar Limestone to Dartry Limestone junction varies in nature between those areas of regularly-bedded Dartry Limestone, such as the Magho region in the north-west of our area, and those localities in which lime-mud mounds developed at an early date, such as the Killydrum region. (Following Wilson's (1975) synthesis of carbonate facies the term 'lime-mud mound' is used here for those facies commonly in the past called 'reefs'. For the sake of brevity the term 'reef is used on the Figures.) In the Magho area there is a gradational change over some metres of rock showing a reduction in the shale content and the development of relatively thick (5-20 cm) dark limestones containing nodular or tabular cherts. Unlike the Dartry Limestone of the type area, in Co. Sligo, thin shale partings persist almost throughout the formation in west Fermanagh. The junction in areas of lime-mud mounds is commonly abrupt, with the sole of the mud mound overlying crinoid-rich uppermost Glencar rocks. In other places the development of lime-mud mounds was delayed so that their bases are within the standard bedded Dartry Lime- stone facies. The overall picture during the time of deposition of the upper Glencar and Dartry limestones is of a reduction in the mud content and establishment of clearer-water shelf-margin 98 C. H. C. BRUNTON & T. R. MASON marine conditions. This period represents the most fully marine phase of Major Cycle Five, and the development of mud mounds indicates a shallowing of marine conditions, leading to the regressive phase and spread of deltaic Glenade Sandstone which terminated this major cycle. In west Fermanagh there are strong lateral facies changes in the Dartry Limestone, from regularly-bedded chert-bearing limestones to lime-mud mounds (Fig. 3) which influence the present upland regions, commonly producing a knoll-like topography. Faulting between Magho and Killydrum, associated with the Highland Boundary Fault zone, partially obscures this facies change, but it also provides fine sections through some mud mound complexes. GLENADE SANDSTONE MEENYMORE FORMATION DARTRY LIMESTONE SANDSTONE SHALES, SILTSTONES & MUDSTONES INCLUDING THIN LIMESTONES & EVAPORITES THIN-BEDDED, SHALEY, FOSSILIFEROUS LIMESTONE CARN LIMESTONE WELL BEDDED, CORAL-RICH LIMESTONE | | 'CORE'MICRITE THIN-BEDDED 'COVER' LIMESTONE 'BEDDED' MICRITE GLENCAR LIMESTONE 'REEF' FACIES DARK, CHERT-RICH BEDDED LIMESTONE Fig. 3. Diagrammatic representation of the principal facies associated with lime-mud mounds occurring in the highlands of west Fermanagh. Lateral and vertical dimensions are not to scale. The lime-mud mounds in this area are like those fully described by Schwarzacher (1961) from counties Sligo and Leitrim and resemble those from northern England (Black 1954, etc.) or, to some extent, the type Waulsort 'reefs' of Belgium, as well as the Waulsortian facies of south and central Ireland. Wilson (1975) widely discussed the various lithologies which have in the past been termed 'reefs', 'bioherms', etc. His general term for all such lithologies, 'carbonate buildup', was subdivided on compositional criteria to include his terms 'lime mud mounds' and 'linear mud accumulations', which were defined as lithologies in which 'lime mud matrix dominates other constituents such as organic boundstone and bioclastic debris. Such buildups are commonly perceived to accumulate both through hydrodynamic processes and in situ organic production.' It is this definition which most closely fits the facies previously termed 'reef complex' in the Dinantian of north-west Ireland. Within the nine facies belts recognized by Wilson (1975), belt five is mostly of organic reefs on platform margins, but includes non-organic reefs. Wilson distinguishes three types of facies. Type I is carbonate-mud and organic detritus accumulations; Type II is a margin slope, knoll reef, facies combining organic frame builders with lime-mud accumulations, and Type III is the facies of true frame-constructed organic reefs on platform rims. The Fermanagh carbonate buildup facies fall principally into Type I, but grade rarely into the Type II situation as the Dinantian sea became shallower. CARBONIFEROUS OF WEST FERMANAGH 99 The Fermanagh mud mound complexes include several lithologies, each of which may grade into the others laterally or vertically (Fig. 3). These include the following. 1. Massive core micrite in which bedding is absent, apart from occasional erosional planes. The rock is a light-coloured calcilutite containing a few cavities filled by sparry calcite. Fossils are rare. 2. Bedded micrite, which in hand specimen is virtually indistinguishable from the core micrite, contrasts in the field by being well bedded and commonly weathered to a rubbly surface. It forms the main mass of small complexes and grades from the tops and flanks of larger mud mounds having a true core, from which it dips at angles of up to about 40°. In these bedded micrites fossils are locally abundant, sometimes associated with the spar-filled cavities con- tributing to the rubbly weathered appearance. Elsewhere bryozoan fronds are preferentially aligned more or less parallel to the bedding planes (see Schwarzacher 1961). 3. Bedded cover limestones grade from the bedded micrite and, because the cavities are smaller and the individual beds thinner, their weathered surfaces erode to a fine rubbly appearance with individual pieces about 2 cm across. As the name suggests, these limestones are found above major mud mounds and are normally extremely rich (up to about 40%) in the debris of crinoids and blastoids. These organisms probably flourished in the shallow water at the tops of the mounds, locally in an environment more like the knoll reefs (Type II) of Wilson (1975). The limestone usually consists of broken shell debris producing a coarse lime sand. The relationships of adjacent mud mounds one to another and to their surrounding chert-rich bedded limestone is one of lateral as well as upward sedimentation through time (Fig. 3). In this way adjacent mounds coalesced or overlapped and the height, or time, to which any one complex developed varied. Thus some mounds are overlain by several metres of normally-bedded Dartry Limestone with chert, though lacking shale partings, deposited before the first shale member of the Meenymore Formation. In other places, flanking the highest mounds, a few metres of well- bedded clean limestones, some of which contain widespread colonies of Lithostrotion corals, were deposited prior to the return of the chert-bearing limestone facies. These coral-rich lime- stones, near the top of the Dartry Limestone formation, indicate a short period of shallow-water clear marine conditions before the deposition of the very argillaceous, but fossiliferous, limestones and shales locally known as Cam Limestones. These represent the start of the strongly regressive phase towards the end of the Fifth Major Cycle and locally they are the basal rocks of the Meenymore Formation. Elsewhere in west Fermanagh mud mound cover limestones are followed directly by Meenymore shales or by even younger Glenade Sandstone. The Leitrim Group The Leitrim Group includes all the formations overlying the Dartry Limestone, up to and including the Namurian Dergvone Shale. Of these formations only the Glenade Sandstone is well exposed. The best outcrops of the Meenymore Formation are found near the town of Garrison, where streams have cut through the peat- and drift-covered slopes above Lough Melvin. The Formations younger than the Glenade Sandstone are restricted to the area around the Doagh Syncline, which has been described by Brandon (1972). The lowest formation of the Leitrim Group is the Meenymore Formation, first described by Brandon (1972), with its type section in Co. Leitrim. This formation is found all over north-west Ireland (Brandon 1977) and its base is always sharply defined. It attains a maximum thickness of about 100 m in the Glen Syncline and is missing over some of the 'reef areas of the underlying Dartry Limestone. The formation may be divided into lower and upper parts by the main marker horizon, the prominent calcareous Quarry Sandstone Member. The lower part of the formation is mainly composed of shaly limestones and mudstone, while the sequence above the Quarry Sandstone Member is characterized by the presence of laminated carbonates. In both the upper and lower parts evaporites are found. As the Meenymore Formation is extremely variable in thickness and lithology, it is ideal for detailed sedimentological study, but this variation also makes it difficult to correlate the exposures. 1 00 C. H. C. BRUNTON & T. R. MASON A goniatite-nuculid fauna from Glennasheevar contains Bollandoceras indicative of the B2 subzone (V3b). This implies that the Glenade Sandstone above is of B2 or P]a age, since the Bellavally Formation contains a Plb fauna. The main lithologies of the Meenymore Formation may be summarized as follows. (a) Calcareous sandstones and siltstones. (b) Shaly and sandy limestones. (c) Laminated micrites, siltstones and dolomicrites (collectively known as laminated carbonates). (d) Limestones containing pseudomorphs after evaporite minerals. (e) Shales, mudstones and siltstones. Fissile, pyritic, micaceous and varying in colour. The most characteristic lithologies are the laminated carbonates, and their presence suggests that the formation was deposited in a sabkha-like environment (Gardiner & Mason 1974), a conclusion which was supported by the discovery of evaporites. The laminated carbonates can be subdivided into three main types. (a) Planar laminated carbonates; dolomite content variable, usually unfossiliferous. (b) Brecciated laminated carbonates; dedolomitized or with high dolomite content. Occasional fossil fragments. (c) Laminated carbonates; dolomitic, containing trace fossils. These rocks were deposited in a peritidal environment, as they all exhibit algal fabrics. The planar laminated carbonates consist of thin (5-15 cm) bands of flaggy carbonate, with a conspicuous lamination on the weathered surface. These laminae, which are parallel to bedding, are composed of very thin (0-1-1-0 mm thick), alternating layers of dolomite and calcite. It is likely that they were formed as described by Gebelein & Hoffman (1973), who claimed that there is sufficient magnesium complexed into organic algal sheath material to form dolomite during diagenesis. Scanning electron micrographs of the laminated rock types show the dolomite to be undoubtedly of secondary origin. Finely laminated lithologies like these are common in the geological column (Sander 1936; Sarin 1962; Laporte 1967; Hoffman 1970). The brecciated laminated carbonates were formed by subaerial desiccation of algal mats under arid conditions. The broken fragments are polygonal, averaging about 1 cm2 in area. The poly- gonal pieces are of a lighter buff colour than the greyish sediment-filled cracks separating them. Sections polished at right angles to bedding show the fragments of algal material to vary in thick- ness from 0-5 mm to 5-0 mm, whilst the sediment layers are 1-0 mm or less in thickness. In thin section these rocks show the following features. (a) Pseudomorphs of calcite after dolomite. (b) Palimpsest textures, with rhombic zones of ferric oxides remaining as ghosts within a new generation of calcite. (c) Possible dolomite inclusions within the calcite pseudomorphs (seen with the scanning electron microscope only). These features fulfil criteria established by Shearman et al. (1961) to define dedolomitized lithologies. The mechanism offered to explain this phenomenon utilizes the reversible reaction 2CaCO3 + MgSO4 ^ CaCO3MgCO3 + CaSO4, thus requiring a gypsiferous solution to effect dedolomitization. This could easily have been derived from the various evaporites within the Meenymore sequence, and the soluble magnesium sulphate produced by the reaction would be removed by rainfall. Gypsum was found at Killybeg crystallized along bedding planes and joint surfaces as a Recent deposit. Dedolomitization is essentially a near-surface or surface reaction. The remaining laminated carbonate contains tube-like trace fossils, which are often concen- trated on bedding surfaces. A detailed study of these (Mason in press) has led to the conclusion that they were the fossil remains of polychaete worm tubes. They were probably made by a small sedentary polychaete resembling the Recent species Sabella microphthalma which Neumann et al. (1970) describe as a common faunal element of subtidal algal mats of the Bahamas. These tubes exhibit a preferred orientation, and it is possible that they were sufficiently rigid to withstand CARBONIFEROUS OF WEST FERMANAGH 101 Abu Dhabi Sabkha Holocene Fermanagh Sabkha Carboniferous Supratidal facies Intertidal facies Subtidal facies Ground level Contorted laminated carbonate Sabkha facies with diagenetic nodular anhydrite Algal Mats---'' Lagoon al Facies Quartz pseudomorphing nodular sulphates Algal laminated carbonate Mudstone 10 cm Typical measured outcrop Meenacloyabane stream section • o Scale After Shearman 1966 Fig. 4. A typical sabkha cycle from Abu Dhabi and the Carboniferous analogue from a section in Co. Fermanagh. transportation and reworking by currents. The diagenetic history of this rock is one of progressive dolomitization. Many of the laminated carbonates are contorted and folded, suggesting that they were still plastic when deformed. Examination of the different types of distortion suggested kinship with some of the algal zones described by Kendall & Skipwith (1968) from the Persian Gulf. One algal limestone from Shroneallappa contains rounded colonies of Ortonella furcata which closely resemble the algal colonies of the cinder zone (Kendall & Skipwith 1968). If it is to be accepted that there is a close comparison between the west Fermanagh Meenymore Formation and the sedimentary regime of the present Persian Gulf, supporting evidence is required. In the Persian Gulf Shearman (1966) recognized a sabkha cycle of deposition comprising firstly sandy carbonate with sulphate nodules, then algal mats, and finally lagoonal facies. Similar cycles occur in the Meenymore Formation (Fig. 4), except that those of the Visean have been altered during diagenesis. A diagenetic history may be constructed to show the sequence of replacement. The main difference between the Carboniferous and Recent sediments is the Carboniferous lack of calcium sulphate, most of which has been replaced diagenetically by celestite, chalcedonic silica (lutecite) and megaquartz. Lens-shaped pseudomorphs after gypsum, and megaquartz replacement of anhydrite nodules, are also recorded. It appears that the depositional environment during Meenymore times resembled the coastal sabkhas of the Persian Gulf at the present time. 102 C. H. C. BRUNTON & T. R. MASON CARBONIFEROUS OF WEST FERMANAGH 1 03 The Glenade Sandstone is exposed as a series of prominent escarpments forming much of the high ground in the area. Sandstones are interbedded with thin shales whose presence may be inferred by the occurrence of shaly clasts in the basal parts of the sandstone outcrops. Some shales seen in a quarry face were pinched into lenses by compaction. Conglomeratic horizons are common near the base of the formation. They contain small (2 cm) shale clasts, well-rounded quartz pebbles (2-3 cm diameter) and coaly debris, thought to have formed in situ from drifted plant remains. Some horizons contain numerous unidentifiable plant fragments. Petrographically the Glenade Sandstone contains c. 95% quartz, 4% calcite or dolomite cement and 1 % feldspar. The occurrence of two generations of calcite, one of which is limpid and the other darker containing ferric oxide, suggests that an original dolomite cement has been dedolomitized. Rhombic crystals of iron-rich calcite resembling the rhombic ghosts of dolomite crystals occur in the brecciated laminated carbonates. In the type area Oswald (1955) described the Glenade Sandstone as containing up to 50% feldspar. The lower amount of feldspar in the Fermanagh outcrops probably reflects their greater distances from the inferred source areas to the north and north-west. All the Carboniferous sandstones in Fermanagh thin southwards, and each sandstone in the succession advanced further south than its predecessor. The thickest sandstones in the Leitrim area are of Namurian age (E2), suggesting that the shoreline was prograding and advanced south with time. The Glenade Sandstone thins from about 200 m in Fermanagh to about 30 m in Leitrim. Compaction in the Glenade Sandstone, in addition to pinching out the shaly horizons, also produces a characteristic 'wavy' bedding, with 'waves' up to an amplitude of 0-75 m and wave- length 4 m. Most other sedimentary structures are localized. Current bedding readings suggest a northerly source for the sediment. Large limestone clasts (500 ml) found in the basal part of the formation are thought to be eroded blocks of Dartry Limestone, as the basal Glenade Sandstone rests on mud mound lime- stone where the Meenymore Formation is attenuated or absent. These contacts are poorly exposed and can be deduced most easily from aerial photographs. The formations above the Glenade Sandstone are poorly exposed in Fermanagh and crop out only in the Doagh Syncline. This is the most northerly occurrence in Ireland of these strata, the nearest related succession being found in Leitrim. It is only possible to construct the succession up to the Dergvone Shale by mapping the various limestone marker horizons, and by collecting their distinctive faunas (Fig. 2). In west Fermanagh the succession is about twice as thick as that in Leitrim, which reflects the close proximity of Fermanagh to the source of the sediment on the old Carboniferous continent. Brandon (personal communication) has remarked on the apparent cyclicity of the Carraun Shale (Fig. 2), which is primarily composed of limestone-shale alternations, thus reinforcing Ramsbottom's (1973) views on the minor cyclicity of Major Cycle Six. Correlations Age determination and correlation based upon faunas in west Fermanagh is poor, except for Fig. 5. Generalized stratigraphical sections A-D, based on published work and our own studies, showing major facies changes and their correlations with the formation names derived from Oswald (1955) and Brandon (1972). From left to right the columns represent increasingly southerly positions and are inter- preted as being successively further from the Carboniferous landmass to the north. The Clogher-Slidve Beagh section (Paget 1952) may be atypical of that basin as a whole. Sheridan (1972) presented data from the Glenoo no. 1 well, drilled a few km south-west of Slieve Beagh, which when tabulated provides approximately 2260 m of sediment from near the base of the Dartry Limestone equivalent down to the top of the supposed Old Red Sandstone. This Glenoo succession includes about 250 m of sandstone, which can be correlated with the Mullaghmore Sandstone, and about 1900 m of limestones and thin shales of the Ballyshannon Limestone type. Below this about 1350m of mixed sandstones, shales, occasional limestones and anhydrites were penetrated which appear similar to those recorded from the base of the Big Dog well in Fermanagh, where they are marked as Basal Clastics in Fig. 2, p. 94. 104 C. H. C. BRUNTON & T. R. MASON CARBONIFEROUS OF WEST FERMANAGH 1 05 some of the younger strata. However, some faunas of significance have allowed us to correlate the formations in north-west Ireland (Figs 5, 6) and to suggest the zonal correlations we present in Fig. 2, p. 94. No stratigraphically important faunas have been described from the Bally- shannon Group, although where better exposed in County Sligo Oswald (1955) records Michelina megastoma (Phillips) (= M. grandis) and Delepinea destinezi (Vaughan) in the Ballyshannon Limestone itself. These are typical C2Si or Cycle Three fossils. D. destinezi is also known from the equivalent limestone exposures on the north shore of Lower Lough Erne near Kesh (Simpson 1953). In this region Simpson (1955) reported Davidsonina carbonaria (M'Coy) from sandstone (? Mullaghmore Sandstone) above the local Ballyshannon Limestone and this would correlate with the S2 or Cycle Four; however, this record has not been verified. Only in the uppermost Benbulben Shale are exposures good and fossils plentiful. Unfortun- ately, diagnostic macrofossils remain poorly known despite the excellence of silicified bryozoans and brachiopods described from the Glencar/Dartry Limestone junction. These faunas have been assigned by Tavener-Smith and Brunton to the Dx Zone, or Cycle Five. The characteristic brachiopods of this Cycle, Davidsonina septosa (Phillips) and Daviesiella llangollensis (Davidson), found in central and northern England, appear to be missing from north-west Ireland. This absence may result from palaeoenvironmental differences between the pale bioclastic limestones low in the D1 Zone, where they are normally found, and the contemporaneous muddy lime- stones of north-west Ireland. Of the abundant silicified brachiopod species collected from Co. Fermanagh, Plicatifera plicatilis (J. de C. Sowerby), Overtonia fimbrlata (J. de C. Sowerby), Krotovia spinulosa (J. Sowerby), Productina margaritacea (Phillips) and eomarginiferids such as Eomarginifera trispina Brunton are normally typical of D! limestones in Britain. Unfortunately, foraminifera are rare in most of these limestones and have not yet been studied in detail. However, small samples of Glencar and Dartry Limestones kindly examined by Professor R. Conil suggest to him (in lit. February 1975) a high V2b to V3a age for a low Glencar Limestone sample. A V3a age for this sample is equivalent to S2 Zone and would thus extend Cycle Four of Ramsbottom (1973) to the base of the Glencar Limestone. From the overlying Leitrim Group shales, good P age goniatites (including Goniatites striatus (J. Sowerby) and G.falcatus Roemer from the Bellavally Formation) have been collected, so the exact age of the Dartry Group remains somewhat equivocal, but it must broadly coincide with the whole of the Dt Zone or Cycle Five. The base of the Namurian is marked by the occurrence of Cravenoceras cf. leion Bisat and Eumorphoceras sp. in the Dergvone Shale from the Doagh outlier. In addition, Brandon (1972) records Eumorphoceras cf. rota Horn and E. cf. medusa Yates from these shales, which are the youngest Carboniferous rocks in west Fermanagh. Discussion The depositional environments of the west Fermanagh Lower Carboniferous strata may be reconstructed by studying the fauna and lithologies. The sedimentary regime varied from shelf carbonates and shales to deltaic sandstones and coastal sabkhas, and the recognition of Major Cycles Three to Six demonstrates the value of this approach to broad stratigraphic correlation. If this cyclicity is to be of more general help in correlation it is necessary to consider the gross aspects of various Carboniferous successions in the northern hemisphere. Fig. 6. Generalized stratigraphical sections E-H from areas as indicated in Fig. 5, based on published work, showing major facies changes and their correlations with the formation names derived from Oswald (1955)and Brandon (1972). From left to right the columns represent increasingly southerly posi- tions and are interpreted as being successively further from the Carboniferous landmass to the north, and follow on from A to D (Fig. 5). In the Donegal syncline (George & Oswald 1957) there is consider- able facies variation ; in the northern parts of the area great thicknesses (over 800 m) of grits and con- glomerates were recorded. This correlates with the thick pre-Clonelly Sandstone sequence of the Omagh syncline (Simpson 1955) shown in Fig. 5, A. The Highland Boundary Fault zone passes between sections F and G, affecting the pre-Carboniferous basement rocks as it does in west Fermanagh. 106 C. H. C. BRUNTON & T. R. MASON The map prepared by Smith et al. (1973) shows that the areas in which Lower Carboniferous sediments were laid down are now fragmented by continental drift, with some pieces now separ- ated by the North Atlantic Ocean. As eastern Canada and Ireland were formerly much closer, one would expect to find evidence of Visean evaporites and supratidal deposits in the Canadian succession. Schenk (1967), working on the Windsor Group of the Canadian Maritime Provinces, described lithologies which were 'almost identical' to the strand-line carbonates of the present Persian Gulf, Bahamas, Florida Bay and South Australia. He described two peritidal lithosomes, together forming the Macumber Formation, thought to be of Visean age, and this formation may be closely related to the Irish Meenymore Formation. Concerning the sequence underlying the Fermanagh Visean succession, Belt et al. (1967) correlated the 'cementstone' facies of eastern Canada, Northern Ireland (Cultra) and the midland valley of Scotland. Recent work on the Irish fauna suggests that the 'cementstones' are of Middle Tournaisian age (Brandon, personal communication). Samples from this succession contain calcitized and dolomitized enterolithic structures similar to those described as calcitized anhydrite by Shearman & Fuller (1969) from the Devonian of Canada. It seems that during the Tournaisian there was an evaporitic interlude, and the successions in eastern Canada, Ireland and Scotland are again remarkably similar. Carboniferous evaporites are also known from the Middle Carbon- iferous of Svalbard (Spitzbergen) (Holliday 1966), the Permo-Carboniferous of Bj0rn0ya, Svalbard (Folk & Siedlecka 1974) and the Lower Carboniferous of Alaska (Armstrong 1970); it is likely that other occurrences remain to be discovered and described in Russia and North America. Green (1961) concluded that the distribution of evaporites was limited to those areas falling between latitudes 45° N and 45° S. The British Isles, and probably Svalbard, would have been within these latitudes during Lower Carboniferous times (Smith et al. 1973), it is suggested that there is sufficient similarity between the British and Irish successions, and that of eastern Canada and possibly Svalbard, to permit them to be correlated on the basis of the over-all sequence. Acknowledgements Much of this work was completed whilst C. H. C. B. was in receipt of a DSIR grant and T. R. M. a post-graduate studentship from the Department of Education, Northern Ireland. Dr Allan Brandon of IGS provided one of us (T. R. M.) with much fruitful discussion, both in the laboratory and the field. We thank Dr C. G. Adams and Dr L. R. M. Cocks for their helpful comments on reading a draft of this paper and Angela Foster for assistance with the map. References Armstrong, A. K. 1970. Mississippian dolomites from the Lisburne Group, Killik River, Mount Bupto region, Brooks Range, Alaska. Bull. Am. Ass. Petrol. Geol., Chicago, 54 (2) : 265-274. Belt, E. S., Freshney, E. C. & Read, W. A. 1967. Sedimentology of Carboniferous cementstone facies, British Isles and eastern Canada. /. Geol., Chicago, 75 : 711-721. Black, W. W. 1954. Diagnostic characters of the Lower Carboniferous knoll-reefs in the north of England. Trans. Leeds geol. Ass. 6 : 262-297. Brandon, A. 1968. The geology of Carboniferous strata ( Visean - Namurian) in parts of counties Lei trim and Cavan, Irish Republic. Ph.D. Thesis, Univ. of Southampton (unpubl.). 1972. The Upper Visean and Namurian shales of the Doagh outlier, County Fermanagh, Northern Ireland. Ir. Nat. J., Belfast, 17 : 159-170. 1977. The Meenymore Formation - an extensive intertidal evaporitic formation in the Upper Visean (B2) of north-west Ireland. Rep. Inst. geol. Sci., London, 77/23. 14 pp. Brunton, C. H. C. 1965. The pedicle sheath of young productacean brachiopods. Palaeontology, London, 7 (4) : 703-704, pi. 109. 1966a. Silicified productoids from the Visean of County Fermanagh. Bull. Br. Mus. not. Hist., London, (Geol.) 12(5): 175-241, pis 1-19. 19666. Predation and shell damage in a Visean brachiopod fauna. Palaeontology, London, 9 (3) : 355-359, pi. 60. CARBONIFEROUS OF WEST FERMANAGH 107 1968. Silicified brachiopods from the Visean of County Fermanagh (II). Bull. Br. Mus. nat. Hist., London, (Geol.) 16 (1) : 1-70, pis 1-9. Caldwell, W. G. E. 1959. The Lower Carboniferous rocks of the Carrick-on-Shannon syncline. Q. Jl geol. Soc. Lond. 115 : 163-187. Dixon, O. A. 1972. Lower Carboniferous rocks between the Curlew and Ox Mountains, Northwestern Ireland. /. geol. Soc. Lond. 128 : 71-101. Folk, R. L. & Siedlecka, A. 1974. The 'schizohaline' environment: its sedimentary and diagenetic fabrics as exemplified by Late Palaeozoic rocks of Bear Island, Svalbard. Sedim. Geol., Amsterdam, 11 : 1-15. Gardiner, B. G. & Mason, T. R. 1974. On the occurrence of the palaeoniscid fish Elonichthys serratus in the Visean of County Fermanagh, with a note on the Enniskillen and Egerton collections. Proc. R. Ir. Acad., Dublin, 74 (B) : 31-36. Gayer, R. A., Allen, K. C, Bassett, M. G. & Edwards, D. 1973. The structure of the Taff Gorge area, Glamorgan, and the stratigraphy of the Old Red Sandstone-Carboniferous transition. Geol. J., Liverpool, 8 (2) : 345-374, pis 12-15. Gebelein, C. D. & Hoffman, P. 1973. Algal origin of dolomite laminations in stromatolite limestone. /. sedim. Petrol., Menasha, 43 : 603-613. George, T. N. & Oswald, D. H. 1957. The Carboniferous rocks of the Donegal syncline. Q. Jl geol. Soc. Lond. 113 : 137-179. et al. 1976. A correlation of Dinantian rocks in the British Isles. Spec. Rep. geol. Soc. Lond. 7 : 1-87. Green, R. 1961. Palaeoclimatic significance of evaporites. In Nairn, A. E. M. (ed.), Descriptive Palaeo- climatology : 61-88. New York. Hoffman, J. I. 1970. The nature of calcium-organic complexes in natural solutions and its implications on mineral equilibria fabstr.]. Diss. Abstr. int., Ann Arbor, 30 (11) : 5101B. Holliday, D. W. 1966. Nodular gypsum and anhydrite rocks in the Billefjorden region, Spitsbergen. Arbok norsk Polarinst., Oslo, 1965 : 65-73. Kendall, C. G. St C. & Skipwith, P. A. d'E. 1968. Recent algal mats of a Persian Gulf lagoon. /. sedim. Petrol., Menasha, 38 : 1040-1058. Laporte, L. F. 1967. Carbonate deposition near mean sea-level and resultant facies mosaic: Manlius Formation (Lower Devonian) of New York State. Bull. Am. Ass. Petrol. Geol., Chicago, 51 : 73-101. Mason, T. R. 1978. Revision of the stratigraphy of the Doagh Outlier, County Fermanagh, Northern Ireland. Ir. Nat. J., Belfast, 19 (5) : 162-163. 1979. Diagenesis of polychaete trace fossils from the Visean of Ireland. Lethaia, Oslo, 12. (In press). Neumann, A. C., Gebelein, C. D. & Scoffin, T. P. 1970. The composition, structure and erodability of subtidal algal mats, Abaco, Bahamas. /. sedim. Petrol., Menasha, 40 : 274-297. Oswald, D. H. 1955. The Carboniferous Rocks between the Ox Mountains and Donegal Bay. Q. Jl geol. Soc. Lond. Ill : 167-183. Paget, P. 1952. The geology of the Clogher-Slieve Beagh area, County Tyrone. Scient. Proc. R. Dubl. Soc. 26 : 63-83. 1953. The stratigraphy of Cuilcagh, Ireland. Geol. Mag., London, 90 : 17-26. Ramsbottom, W. H. C. 1973. Transgressions and regressions in the Dinantian: A new synthesis of British Dinantian stratigraphy. Proc. Yorks. geol. Soc., Leeds, 39 (4) : 567-607. Sander, B. 1936. Beitrage zur Kenntniss der Anlagerungsgefiige. (Rhythmische Kalke und Dolomite aus der Trias). Z. Krystallogr. Miner. Petrogr. B (Mineralog. petrogr. Mitt.), Leipzig, 48 : 27-209. Sarin, D. D. 1962. Cyclic sedimentation of primary dolomite and limestone. /. sedim. Petrol., Menasha, 32 : 451-471. Schenk, P. E. 1967. The Macumber Formation of the Maritime Provinces, Canada, a Mississippian analogue of Recent strand-line carbonates of the Persian Gulf. /. sedim. Petrol., Menasha, 37 : 365-376. Schwarzacher, W. 1961. Petrology and structure of some Lower Carboniferous reefs in north-west Ireland. Bull. Am. Ass. Petrol. Geol., Chicago, 45 : 1481-1503. 1964. An application of statistical time-series analysis of a limestone-shale sequence. /. Geol., Chicago, 72(2) : 195-213, pi. 1. 1968. [The vertical and lateral variation of a Carboniferous limestone area near Sligo (Ireland).] In:Voprosy matematicheskoi geologii [Topics in mathematical geology] : 158-167, figs. Leningrad, Akad. Nauk SSSR (Ordela Lenina Matematicheskil Inst.) [In Russian: Engl. transl. 1970, Consult. Bur. New York-London.] Shearman, D. J. 1966. Origin of marine evaporites by diagenesis. Trans. Instn Min. Metall., London, (B) 75 : 208-215. , Khouri, J. & Taha, S. 1961. On the replacement of dolomite by calcite in some Mesozoic limestones from the French Jura. Proc. Geol. Ass., London, 72 : 1-12. 108 C. H. C. BRUNTON & T. R. MASON & Fuller, J. G. 1969. Anhydrite diagenesis, calcitisation and organic laminites, Winnipegosis Formation, Middle Devonian, Saskatchewan. Bull. Can. Petrol. GeoL, Calgary, 17 : 496-525. Sheridan, D. J. R. 1972. Upper Old Red Sandstone and the Lower Carboniferous of the Slieve Beagh Syncline, and its setting in the N.W. Carboniferous basin, Ireland. Spec. Pap. geol. Surv. Ir., Dublin, 2. 129 pp. , Hubbard, W. F. & Oldroyd, R. W. L. 1967. A note on Tournaisian strata in Northern Ireland. Scient. Proc. R. Dubl. Soc. 3 : 33-37. Simpson, I. M. 1953. Daviesiella destinezi (Vaughan), a Lower Carboniferous index fossil in north-west Ireland. Geol. Mag., London, 90 : 193-200. 1955. The Lower Carboniferous stratigraphy of the Omagh Syncline, Northern Ireland. Q. Jl geol. Soc. Lond. 110 : 391-408. Symes, R. G. & Wilkinson, S. B. 1886. Explanatory Memoir . . . sheet 44 ... portions of Counties Fermanagh, Leitrim and Cavan. Mem. geol. Surv. Ireland, Dublin. 22 pp. Smith, A. D., Briden, J. C. & Drewry, G. E. 1973. Phanerozoic World Maps. In Hughes, N. F. (ed.), Organisms and Continents through time. Spec. Pap. Palaeont., London, 12 : 1-42. Tavener-Smith, R. 1965. A new fenestrate bryozoan from the Lower Carboniferous of County Fer- managh. Palaeontology, London, 8 : 478-491, pi. 66. 1973. Fenestrate bryozoa from the Visean of County Fermanagh, Ireland. Bull. Br. Mus. not. Hist., London, (Geol.) 23 (7) : 389-493, pis 1-26. West, I. M., Brandon, A. & Smith, M. 1968. A tidal flat evaporitic facies in the Visean of Ireland. /. sedim. Petrol, Menasha, 38 (4) : 1079-1093. Wilson, J. L. 1975. Carbonate Facies in Geologic History. 469 pp., 30 pis. Berlin, Heidelberg & New York. Wilson, R. L. 1964. The Tertiary Dykes of Magho Mountain, Co. Fermanagh. Ir. Nat. J., Belfast, 14 (11) : 254-257. Accepted for publication 24 February 1978 Titles to be published in Volume 32 Miscellanea Palaeoenvironments and correlations of the Carboniferous rocks in west Fermanagh, Ireland. By C. H. C. Brunton & T. R. Mason. The Ordovician trilobite faunas of the Builth-Llandrindod Inlier, central Wales. Part III. By C. P. Hughes. The stratigraphy and brachiopods of the upper part of the type Caradoc of south Salop. By J. M. Hurst. The entire Geology series is now available Printed by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) The Ordovician trilobite faunas of the Builtl Llandrindod Inlier, central Wales. Part III C. P. Hughes Geology series Vol 32 No 3 27 September 1979 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique and ever-growing collections of the Museum, both by the scientific staff of the Museum and by specialists from elsewhere who make use of the Museum's resources. Many of the papers are works of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself, available separately, and individually priced. Volumes contain about 300 pages and are not necessarily completed within one calendar year. Subscriptions may be placed for one or more series. Subscriptions vary according to the contents of the Volume and are based on a forecast list of titles. As each Volume nears completion, subscribers are informed of the cost of the next Volume and invited to renew their subscriptions. Orders and enquiries should be sent to: Publications Sales, British Museum (Natural History), Cromwell Road, London SW7 5BD, England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Geol.) Trustees of the British Museum (Natural History), 1979 ISSN 0007-1471 Geology series Vol 32 No 3 pp 109-181 British Museum (Natural History) Cromwell Road London SW7 5BD Issued 27 September 1979 The Ordovician trilobite faunas of the Builth- Llandrindod Inlier, central Wales. Part III C. P. Hughes Department of Geology, Sedgwick Museum, Downing Street, Cambridge CB2 3EQ Contents Synopsis . . . . . . . . . . . . .110 Terminology . . . . . . . . . . . . -110 Definition of measurements . . . . . . . . . .110 Stratigraphy 112 Systematic descriptions . . . . . . . . . . .113 Family Cyclopygidae Raymond . . . . . . . . .113 Genus Emmrichops Marek . . . . . . . . .113 Emmrichops ? extensus sp. nov. . . . . . . . -113 Genus Pricyclopyge Richter & Richter . . . . . . -114 Pricyclopyge wattisoni sp. nov. . . . . . . . .115 Cyclopygid gen. et sp. indet. . . . . . . . .116 Family Asaphidae Burmeister . . . . . . . . .117 Subfamily Asaphinae Burmeister . . . . . . . . .117 Genus Nobiliasaphus Pf ibyl & Vanek . . . . . . .117 Nobiliasaphus powysensis sp. nov. . . . . . . . .117 Genus Opsimasaphus Kielan . . . . . . . . .120 ? Opsimasaphus sp. indet. A . . . . . . . . . 121 ? Opsimasaphus sp. indet. B . . . . . . . . .121 Subfamily Ogygiocaridinae Raymond . . . . . . . .122 Genus Ogyginus Raymond . . . . . . . . .122 Ogyginus corndensis (Murchison) . . . . . . . .126 Ogyginus inter medius Elles . . . . . . . . .136 Ogyginus cf. intermedius Elles . . . . . . . .137 Ogyginus ? laticostatus (Salter) . . . . . . . .138 Genus Ogygiocarella Harrington & Leanza . . . . . .140 Ogygiocarella debuchii (Brongniart) . . . . . . .142 Ogygiocarella angustissima (Salter) . . . . . . 1 50 Ogygiocaridinid gen. et sp. indet. . . . . . . .153 Asaphid gen. et sp. indet. . . . . . . . . .153 Family Nileidae Angelin . . . . . . . . . .154 Genus Barrandia M'Coy . . . . . . . . .154 Barrandia cordai M'Coy . . . . . . . . .156 Barrandia cf. cordai M'Coy . . . . . . . . .158 Barrandia expansa sp. nov. . . . . . . . . .159 Barrandia ultima sp. nov. . . . . . . . . .160 Genus Homalopteon Salter . . . . . . . . .162 Homalopteon radians (M'Coy) . . . . . . . .164 Homalopteon murchisoni sp. nov. . . . . . . . .168 Family Odontopleuridae Burmeister ........ 171 Subfamily Odontopleurinae Burmeister . . . . . . .171 Genus Diacanthaspis Whittington . . . . . . . .171 Diacanthaspis ? sp. A . . . . . . . . . .171 Odontopleuridid gen. et sp. indet. . . . . . . . .172 Family Proetidae Salter. . . . . . . . . . .174 Genus Rorringtonia Whittard . . . . . . . . .174 Rorringtonia sp. . . . . . . . . . . .174 Acknowledgements . . . . . . . . . . . .175 References 175 Index 179 Bull. Br. Mus. not. Hist. (Geol.) 32 (3) : 109-181 Issued 27 September 1979 109 110 C. P. HUGHES Synopsis This paper, third in a series describing the Ordovician trilobite faunas of the Builth-Llandrindod inlier, describes representatives of the families Cyclopygidae, Asaphidae, Nileidae, Odontopleuridae and Proetidae. Emmrichops ? extensus, Pricyclopyge wattisoni, Nobiliasaphus powysensis, Barrandia expansa Barrandia ultima and Homalopteon murchisoni are described as new. A lectotype for Ogygiocarella angustissima (Salter) is selected. Terminology The terminology used is that of Harrington, Moore & Stubblefield (in Moore 1959) with, in addition, a few terms introduced since that date and two proposed here. The new terms intro- duced by Marek (1961) for the description of cyclopygids have not been adopted as the standard terms in use are believed adequate. Richter & Richter's (1940) 'greek letter' terminology for the path of the facial suture is used in the definition of some of the measurements taken on asaphids and nileids. The following new terms are proposed: PARATHORACIC SEGMENT — applied to the region of the transitory pygidium having the form of a thoracic segment but still fused into the pygidial exoskeleton (see Barrandia and Homa- lopteon). YOUNG HOLASPIS — applied to individuals having the full number of thoracic segments but not possessing all other holaspid characters. It is preferred to 'meraspis degree 8' (as applied to asaphids) of Whittard (1964:260) since the term meraspis excludes forms possessing the adult number of thoracic segments. The following terms not listed in Moore (1959) are also used: PARADOUBLURAL LINE — applied to the line on the dorsal surface of the exoskeleton directly above the inner margin of the doublure (Henningsmoen 1960 : 210). POSTERIOR INDENTATION — applied to the small indentation in the posterior margin of the transitory pygidium in asaphids (Henningsmoen 1960 : 230). As this structure is present in holaspides of certain asaphids (Henningsmoen 1960 : 220; pi. 1, figs 1, 2) this term is preferred to 'larval notch' (Osmolska 1962 : 57). Definition of measurements As in earlier parts of this study (Hughes 1969, 1971), the single orientation method of measure- ments proposed by Shaw (1957) has been followed. In those measurements involving distances between, or to, furrows, the measurements have been taken from the deepest (dorsoventrally) point in the furrow. All measurements are taken from internal moulds except where specifically stated to the contrary. Measurements shown as estimated indicate that measurement was not possible using both defined end points. In width measurements estimates were obtained by the doubling of measurements taken to the sagittal line from one defined end point; in length measurements an estimation was made of the position of one end point by projecting exsagittal features onto the sagittal line. Measurements given to two decimal places are rounded to the nearest 0-05 mm. Definitions of measurements (see Fig. 1): A maximum cephalic length - measured in sagittal line. B glabellar length - measured in sagittal line between anterior of glabella and posterior margin of occipital ring (asaphids only herein). C prepalpebral length - distance between anterior of cephalon and y of facial suture, as projected onto sagittal line (asaphids only herein). Ci palpebral length - distance between y and e of facial suture, as projected onto sagittal line (asaphids only herein). C2 postpalpebral length - distance between e of facial suture and posterior cephalic margin, as projected onto sagittal line. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 111 Fig. 1. Diagram showing measurements taken on cyclopygids, ogygids and nileids; also hypostomata and sutural terminology. 112 C. P. HUGHES C3 midpalpebral length - distance between 5 of facial suture and posterior cephalic margin, as pro- jected onto sagittal line. I maximum cephalic width - measured in a transverse direction (nileids only herein). Ii cephalic width - measured in a transverse direction along line of posterior cephalic margin (asaphids only herein). J posterior cranidial width - measured in a transverse direction along posterior cephalic margin. J! maximum anterior cranidial width - measured in a transverse direction. J2 midcranidial width - measured in a transverse direction between 8 of facial suture. K maximum glabellar width - measured in a transverse direction between axial furrows. KI posterior glabellar width - measured in a transverse direction between axial furrows, along posterior cephalic margin. K2 maximum anterior glabellar width - measured in a transverse direction (asaphids only herein). Q thoracic length - measured in an exsagittal line between anterior margin of anterior thoracic segment and posterior margin of posterior thoracic segment. Qi axial length of anterior thoracic segment - measured in sagittal line between anterior and posterior margins of anterior thoracic segment. RI anterior thoracic width - measured in a transverse direction between lateral extremities of anterior thoracic segment. R2 anterior thoracic axial width - measured in a transverse direction between axial furrows, along anterior margin of anterior thoracic segment. R3 posterior thoracic axial width - measured in a transverse direction, between axial furrows along posterior margin of posterior thoracic segment. R4 posterior thoracic width - measured in a transverse direction between lateral extremities of pos- terior thoracic segment. R5 measured in a transverse direction between inner doublural margin and axial furrow along anterior thoracic margin (nileids only herein). R6 measured in a transverse direction between inner doublural margin and axial furrow along posterior thoracic margin (nileids only herein). W maximum pygidial width - measured in a transverse direction. Wi intra-marginal pygidial width - measured in a transverse direction between distal extremities of anterior border furrows. X anterior pygidial axial width - measured in a transverse direction between axial furrows, along anterior margin. Xi posterior pygidial axial width - measured in a transverse direction between axial furrows across posterior of axis (asaphids only herein). Y pygidial axial length - measured in sagittal line between articulating furrow and posterior of axis. Z pygidial length - measured in sagittal line between articulating furrow and posterior pygidial margin. T.L. total length - measured in sagittal line between anterior cephalic margin and posterior pygidial margin. H! hypostomal length - measured in sagittal line between anterior and posterior hypostomal margins. H2 distance between anterior hypostomal margin and anterior of median lip as projected onto sagittal line (asaphids only herein). H3 distance between anterior hypostomal margin and anterior of maculae, as projected onto sagittal line (asaphids only herein). H4 distance between anterior hypostomal margin and posterior extremity of anterior wings, as projected onto sagittal line. H5 maximum hypostomal width - measured in a transverse direction. H6 posterior hypostomal width - measured in a transverse direction across the maximum width posterior of anterior wings. H7 distance between anterior hypostomal margin and posterior border furrow, measured in sagittal line (nileids only herein). H8 distance between anterior hypostomal margin and middle furrow, measured in sagittal line (nileids only herein). Stratigraphy In earlier parts of this study (Hughes 1969, 1971), the term Llandeilo Series was taken as essentially equivalent to the single assemblage zone of Glyptograptus teretiusculus, and beds of low Nema- BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 1 1 3 graptus gracilis Zone age were referred to basal Caradoc Series. Since then, however, Addison (in Williams et al. 1972 : 35-36) has demonstrated that assemblages indicative of N. gracilis Zone age are present in the middle and upper Llandeilo. In this work, therefore, the G. tere- tiusculus Zone is considered to be of lower Llandeilo age and the succeeding beds of basal N. gracilis Zone to be of middle and upper Llandeilo age, possibly with the upper horizons being of basal Caradoc age. Thus some beds are assigned an age herein which is different from the usage in my previous papers. The lithostratigraphical terminology of the region is badly in need of revision ; the only available names (e.g. Glyptograptus teretiusculus Shales), formalized in Williams et al. (1972), though unsatisfactory, are used herein, being the best terms available at present. Systematic descriptions Unless otherwise stated, all specimens are in the Department of Palaeontology of the British Museum (Natural History) (registered numbers without prefix, or with prefix I or It). Specimens in other collections are cited as follows: BU University of Birmingham GSM Geological Survey Museum (Institute of Geological Sciences, London) NMW National Museum of Wales, Cardiff OUM Museum, Oxford University SM Sedgwick Museum, Cambridge UCW University College, Aberystwyth Family CYCLOPYGIDAE Raymond, 1925 Genus EMMRICHOPS Marek, 1961 DIAGNOSIS. Cephalon about twice as wide as long. Eye relatively flat with much of the visual surface directed ventrally. Glabella with three pairs of transversely elongate muscle impressions. Rostral suture not arched downwards. Doublure relatively narrow. Thoracic axis wider than pleural regions; number of thoracic segments unknown. Pygidium unknown. TYPE SPECIES. Emmrichops planicephalus Marek, 1961. DISTRIBUTION. Dobrotiva Beds (lower Llandeilo) of Czechoslovakia (Marek 1961 : 52) and possibly Glyptograptus teretiusculus Shales (lower Llandeilo) of central Wales (herein). DISCUSSION. Marek (1961) considered the very wide cephalon sufficient to warrant the separation of this genus from Microparia Hawle & Corda, 1847, to which he believed it was allied owing to the similarity in cephalic musculature. Emmrichops ? extensus sp. nov. (Figs 3, 4, 6, 9) DIAGNOSIS. Cyclopygid with cranidium much wider than long. Pygidium with concave lateral margins and transversely elongate anterolateral angles. NAME. 'Dilated'. TYPE MATERIAL. Holotype, GSM 6142 (Fig. 4), internal mould of pygidium. Paratype, It. 2838 (Figs 3, 6, 9), external mould of pygidium and associated internal mould of damaged cranidium. DIMENSIONS (in mm). W X Y Z Holotype 24-7 7-5 6-5 8-4 It.2838 c. 14-0 4-0 3-7 5-8 For explanation of symbols see Fig. 1. TYPE LOCALITY AND HORIZON. The exact locality from which the holotype was collected is un- certain. The specimen is labelled 'Pen Cerrig, Builth' and it is possible that it is from the quarry at the south-western end of Pen-cerig Lake, although collecting at this locality has not yielded 114 C. P. HUGHES further specimens. The paratype is from exposures in the left bank of the stream section east of Bach-y-graig, 40 yd (37 m) upstream of the point where the footpath enters the wood at the western end of the section. If the holotype is from Pen-cerig, then it is from slightly younger beds than those from which the paratype was recovered, though both lie within the Glyptograptus teretiusculus Shales of lower Llandeilo age. DESCRIPTION. The single known cephalon is compressed, cracked and incomplete, but is clearly very much wider than long. It shows no obvious signs of glabellar furrows or nodes, though faint traces of possible muscle scars are present posteriorly. The preserved eye (right) is well developed; possibly not fused anteriorly with left. Lenses hexagonal except for inner median portion where they are modified to be approximately rectangular (Fig. 9). Thorax unknown. Pygidium nearly three times as wide as long, subtriangular with anterior margin gently convex forwards and lateral margins markedly concave (Figs 4, 6). Axis occupying nearly one-third of anterior width tapers gently to the posterior, occupies about four-fifths of pygidial length and terminates obtusely before reaching border furrow. Posterior of axis more clearly defined in holotype than in paratype pygidium where it merges into postaxial field. Anteriorly a single well-defined axial ring is developed with, in the holotype, faint traces of three further rings. Pleural field smooth and inflated. Wide, shallow border furrow present laterally and anteriorly, becoming less well defined posteriorly. Doublure narrow, underlies posterolateral border and bears several subparallel terrace lines. DISCUSSION. The species is here tentatively assigned to Emmrichops mainly on the basis of the very wide cephalon. Whilst it is recognized that the unusual outline of the holotype may be due to post-mortem deformation, the symmetry of the specimen together with the relatively small amount of cracking or buckling of the specimen argues against tectonic deformation. The lateral margins of the paratype pygidium are not well preserved but appear to be of the same general form as those of the holotype, though possibly with a slightly smaller anterolateral projection. Although the pygidium of Emmrichops has not been described, it is likely to be relatively wide and the pygidium described here is not obviously anomalous when compared to the cephalon of the type species. Of other cyclopygid genera in which the pygidium is known, Pricydopyge possibly exhibits the most similarity. However, the rather unusual outline of E. ? extensus distinguishes it from all known Pricydopyge species. Genus PRICYCLOPYGE R. & E. Richter, 1954 DIAGNOSIS. Cephalon larger than pygidium and somewhat wider than long. Glabella with four pairs of muscle scars. Librigena very small. Eye large, highly convex. Rostral suture, if developed, moderately vaulted upwards. Thorax of six segments. Pygidium subtriangular with distinct axis and border. TYPE SPECIES. Aeglina prisca Barrande, 1872. DISTRIBUTION. The genus is present throughout Europe in the lower Ordovician (Tremadoc - Llandeilo) and also the Arenig Series of Turkey (Dean 1973). The type species, originally described from the lower Llanvirn of Sarka, Bohemia, has since been recorded from the upper Llanvirn of Bulgaria (Spassow 1958: 18) and from the Arenig and lower Llanvirn Series of Britain (Whittard 1961 : 177). DISCUSSION. Aeglina prisca Barrande, 1872, designated type species by R. & E. Richter (1954), was considered by Marek (1961) to be conspecific with Pricydopyge binodosa (Murchison, 1859). However, it is here preferred to follow Whittard's later argument (1966:287) in favour of considering them as two separate species. The thoracic and pygidial morphology of Aspidaeglina Holub, 1911 is similar to that of Pricydopyge with the exception of the spinose thoracic pleural terminations in Aspidaeglina. The cephalon is at present very imperfectly known and further specimens may indicate it to be congeneric with Pricydopyge. Cydopyge Hawle & Corda, 1847 is distinguished from Pricydopyge BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 115 Fig. 2. Reconstruction of Pricyclopyge wattisoni sp. nov., c. x 3. by its semicircular pygidium, lack of a large pleural spine on the sixth pleural thoracic segment and by a different configuration of cephalic muscle scars. Pricyclopyge wattisoni sp. nov. (Figs 7, 8) DIAGNOSIS. Pricyclopyge with no median glabellar node and somewhat rounded outline to the triangular pygidium; glabellar margins ill-defined posteriorly. NAME. After the late Mr J. T. Wattison. TYPE MATERIAL. Holotype, BU 1913 (ex Wattison Collection) (Fig. 7), external mould of cranidium and anterior thoracic segment. Paratype, BU 1914 (ex Wattison Collection) (Fig. 8), external mould of pygidium. DIMENSIONS (in mm). A Holotype 9-5 R2 6-0 J K Qi Rt 12-0 10-8 1-6 12-0 W Wi X Z Paratype 13-0 11-0 4-0 7-4 For explanation of symbols see Fig. 1 . TYPE LOCALITY AND HORIZON. Both the type specimens are from the Glyptograptus teretiusculus Shales in the small quarry at the south-western end of Pen-cerig Lake. This is the only locality from which the species is known. DESCRIPTION. Cranidium elliptical, consisting almost entirely of large, gently convex, unfurrowed glabella. Cranidium slightly wider than long; posterior margin convex posteriorly. Axial furrow, marked by change of curvature between the flat fixigena and gently convex glabella, gently convex outwards medially; posteriorly indiscernible. Fixigena, present along three-quarters of glabella, very small, forming a narrow palpebral lobe lateral to glabella and expanding slightly 116 C. P. HUGHES into a small triangular area posterolaterally. Facial suture directed anteromedially at posterior margin, becoming parallel to axial furrow. Occipital furrow very shallow, delimiting a narrow (sag.) occipital ring. Posterior border furrow very shallow laterally, merging axially with occipital furrow. Librigena and eye unknown. Pygidium roundedly subtriangular with continuous border which is less pronounced pos- teriorly. Axis tapers posteriorly; delimited by axial furrow for only about half pygidial length. Articulating half ring and two axial rings visible, anterior one being the more prominent; pleural field smooth. DISCUSSION. Both the cranidium and the pygidium probably belong to the same species, if not the same individual, because of their close association on the same slab, compatible size, and the fact that they both appear to be best assigned to Pricydopyge. The species is distinct from all named species of Pricydopyge, although the pygidium described as 'cyclopygid E' by Whittard (1966: 289; pi. 49, figs 14, 15) may belong to this species. How- ever, the Shelve specimen is more rounded in outline and the axis is better defined. The presence of the median doublural spine in the Shelve specimen is of uncertain significance; the doublure is not known in the Builth specimen. P. binodosa (Murchison, 1859) differs in that it possesses a prominent median glabellar node and the cephalic and pygidial furrows are better developed. P. latifrons (Tjernvik, 1956), from the Ceratopyge Limestone of Sweden, is perhaps most similar to this new species, but the latter differs in having no well-developed median glabellar node nor any transverse terrace lines. P. obscura Marek, 1961 differs from P. wattisoni in having an elongate (rag.) narrow (tr.) glabella, while P. synophthalma (Kloucek, 1916) is distinguished by its much greater median transverse expansion of the glabella and better-defined axial furrow. P. (?) campestris Koroleva, 1967 is easily distinguished by the well-developed pygidial axis and the clearly-defined cephalic axial furrow. P. super ciliata Dean, 1973 from the Sobova Formation of Turkey is distinguished by its relatively long glabella, well-incised axial furrow and relatively wide fixigena. An internal mould of a thorax, It. 2837, is provisionally referred to P. wattisoni (Fig. 5) as it occurs with the type specimens of this species at Pen-cerig Lake. Cyclopygid gen. et sp. indet. (Fig. 13) FIGURED SPECIMEN. It.2839, internal mould of pygidium. Dimensions (in mm): W, 4-7. X, 1-4. Y (measurement taken to the posterior extremity of the axial furrow), 1-7. Z, 2-8. For explana- tion of symbols see Fig. 1. LOCALITY AND HORIZON. Stream section 15yd (14m) SW of the old quarry 350yd (320m) west of Maesgwynne, in Glyptograptus teretiusculus Shales. DESCRIPTION. Pygidium small, subsemicircular in outline. Axis occupies just under one-third of the width anteriorly and tapers to the rear. Axial furrow well developed anteriorly, but becomes progressively weaker posteriorly, finally dying out altogether at about three-fifths of the distance to posterior margin. Axis merges without interruption into postaxial field; no postaxial furrow present. One well-developed axial ring present. Pleural field gently convex and has one prominent furrowed rib, the furrow becoming more prominent distally. Anterior border narrow (sag.} separated by anterior border furrow. Doublure moderately wide, seemingly about half the length of the postaxial field in width, and bears a series of terrace lines parallel to the margin. DISCUSSION. This isolated pygidium of lower Llandeilo age is very like Microparia nudus Whittard, 1961 and Microparia (Degamelld) gigantea (Barrande, 1872) in outline, but differs from both in the nature of the furrowing of the axial and pleural regions. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 1 1 7 Family ASAPHIDAE Burmeister, 1843 Subfamily ASAPHINAE Burmeister, 1843 Genus NOB1LIASAPHUS Pribyl & Vanek, 1965 [= Pamirotchechites Balashova, 1966] DIAGNOSIS (emended from Pribyl & Vanek 1965 : 277-278). Like Opsimasaphus but differs in that glabella is well denned frontally, and laterally with preglabellar field up to two-fifths of cephalic length. Path of anterior branches of facial suture semicircular, intramarginal, sub- tending an angle of between 90° and 130° frontally. Glabella clearly furrowed with median tongue-like lobe divided by two transverse furrows, posteriorly deflected adaxially. Three small lateral glabellar lobes and two very faint circular muscle areas present lateral to median lobe. Thoracic segments terminate in short blunt spines. Pygidial outline parabolic with up to 17 axial rings. Axial ring furrows and surface sculpture, posteriorly deflected medially, giving central keel to axis, which may be pointed posteriorly. Pleural field with 14 to 16 ribs. Doublure narrow posteriorly, more concave than in Opsimasaphus. TYPE SPECIES. Asaphus nobilis Barrande, 1 846. DISTRIBUTION. The genus ranges from the Llanvirn to Ashgill Series. It is present in the Llanvirn - Caradoc (Soudleyan) of Czechoslovakia (Havlteek & Vanek 1966 : 41, 55; Kfiz & Pek 1972 : 165; Pribyl & Vanek 1965:278; Pribyl & Vanek 1968:192); Llanvirn and ? Llandeilo of Spain (Gil Cid 1972, 19720); middle and upper Llandeilo of Wales (herein); Llandeilo of the Pamir, U.S.S.R. (Balashova 1966, 1968); lower Ordovician of France ? (Racheboeuf 1970); Caradoc of Portugal (Thadeau 1947 : 220) and of Sardinia (Laufeld 1973); ? upper Ordovician of Afghanistan (Pillet & Lapparent 1969 : 326). DISCUSSION. On erecting Opsimasaphus, Kielan (1960 : 75-77) discussed 'Asaphus' nobilis, which had been placed previously in Pseudobasilicus by some (Reed 1930 : 308; Jaanusson 1953 : 445), and concluded that its generic attribution remained in doubt. Pribyl & Vanek (1965) erected Nobiliasaphus, type species Asaphus nobilis, as a subgenus of Opsimasaphus. Subsequently, Balashova (1966) erected Pamirotchechites, also with A. nobilis as type species. Balashova (1968) later assigned nobilis to Pseudobasilicus and Gil Cid (1972, 1972a) followed this. In 1971, how- ever, Balashova accepted Nobiliasaphus and gave it full generic status; this was supported by Kriz & Pek (1972). A full revision of this group of asaphids is beyond the scope of this paper, but the character- istic rearward kink in the pygidial axial ring furrows, so clearly portrayed by Barrande (1852 : pis 31, 32), in particular would seem to necessitate the rejection of close affinity between Pseudo- basilicus and 'Asaphus' nobilis. Further, this feature, and the better definition of the glabella, long preglabellar field and greater number of pygidial ribs, appear to be consistent differences between Nobiliasaphus and Opsimasaphus. It is clear, however, that Nobiliasaphus is closely related to Opsimasaphus, and it seems inappropriate to continue to place Nobiliasaphus in the Pseudoasaphidae of Balashova, 1969; both genera are thus included here in the Asaphidae. The relations of these genera to other asaphids are less clear and although they are placed here in the Asaphinae, both show some similarities to the ogygiocaridinid - niobinid group. The generic placing of specimens described as O. radiatus (Salter, 1866a) by Whittington (1966 : 71-78) and Ingham (1970 : 18-19) is not entirely certain due to indifferent preservation. Nobiliasaphus powysensis sp. nov. (Figs 10, 14-16, 20) DIAGNOSIS. Glabella occupies about three-fifths cephalic length. Anterior branches of facial suture widely divergent, becoming very nearly marginal; mid-point of eye about one-fifth of cephalic length from posterior margin. Pygidium subparabolic, axis slender, tapering posteriorly with at least 17 axial rings; anterior width about one-eighth of anterior pygidial width. Pleural field with 14-16 ribs. Border region smooth and moderately wide. NAME. From Powys, Wales. 118 C. P. HUGHES TYPE MATERIAL. Holotype: It. 13553 (ex University College, Aberystwyth, UCW 19451/2) (Figs 15, 16), internal and external moulds of entire specimen. Paratypes: It. 2924 (Fig. 20), internal and external moulds of disarticulated specimen consisting of cranidium, pygidium and some thoracic segments. 1. 2857a, internal mould of large pygidium. It. 13554, internal mould of pygidium. It.13555 (Fig. 10), internal and external moulds of pygidium. SM A44731 (Fig. 14), external mould of pygidium and part of thorax. TYPE LOCALITY AND HORIZON. The holotype is almost certainly from the Nemagraptus gracilis Shales, in the middle quarry, Llanfawr, Llandrindod (locality recorded in University of Wales, Aberystwyth, catalogue as ' "quarry near Hospital", Llandrindod Wells'). Paratypes It. 13554-5 are from the middle quarry, Llanfawr; It. 2924 is, and I.2857a and SM A44731 are most probably, from Gwern-yfed-fach quarry. DIMENSIONS (in mm). A B Q C2 J Jx J2 K Paratype It.2924 c. 22-5 13-5 c. 3-9 2-2 c. 30-0 c. 22-0 c. 16-5 8-0 RI R2 Qi w wx x xx Y z Holotype It.13553 c. 22-2 5-1 1-2 21-5 16-2 3-0 1-5 10-0 12-7 Paratypes It.2924 c. 42-0 5-5 2-3 c. 23-0 27-5 I.2857a c. 108-0 16-3 c. 7-5 c. 69-0 78-0 It.13554 14-2 3-0 1-5 9-9 11-8 It.13555 26-7 21-8 4-1 1-8 13-2 17-1* SMA44731 c. 26-4 c. 21-4 4-0 1-9 13-9 17-4* For explanation of symbols see Fig. 1. * - measured on external mould. DISTRIBUTION. Known only from the two localities yielding the type specimens. DESCRIPTION. Glabella relatively short, occupying about three-fifths of cephalic length and defined laterally by weak axial furrow, particularly anteriorly; frontally glabella merges into preglabellar Figs 3-4, 6, 9. Emmrichops ? extensus sp. nov., p. 113. Figs 3, 6, 9, Lower Llandeilo, left bank of stream section east of Bach-y-graig, 40 yd (37 m) upstream of the point where the footpath enters the wood at the western end of the section, SO 071610. Figs 3, 9, Paratype. Internal mould of cephalon, It. 2838. Fig. 3, x4; Fig. 9, x7. Fig. 6, Paratype. Latex impression from external mould of pygidium, It. 2838, x3-5. Fig. 4, Lower Llandeilo, ? small quarry at SW end of Pen-cerig Lake (see p. 113); SO 043541. Holotype. Internal mould of pygidium showing well-developed doublure, GSM 6142, x3. Fig. 5. ? Pricyclopyge wattisoni sp. nov., p. 116. Lower Llandeilo, locality as Figs 7-8. Internal mould of thofax, It.2837, x4. Figs 7-8. Pricyclopyge wattisoni sp. nov., p. 115. Lower Llandeilo, small quarry at SW end of Pen-cerig Lake, SO 043541. Fig. 7, Holotype. Latex impression from external mould of cranidium, BU 1913 (ex Wattison Coll.), x3. Fig. 8, Paratype. Latex impression from external mould of pygidium, BU 1914 (ex Wattison Coll.), x2-5. Figs 10, 14-16. Nobiliasaphus powysensis sp. nov., p. 117. Fig. 10, Middle-upper Llandeilo, middle quarry, Llanfawr, Llandrindod, SO 066617. Paratype. Internal mould of pygidium, It.13555, x 1-75. Figs 15-16, Middle-upper Llandeilo, ? middle quarry, Llanfawr, Llandrindod (see above), SO 066617. Holotype. Internal and external moulds of entire individual, It.13553 (ex University College of Wales, Aberystwyth, Collection, UCW 19451, 19452), x2. Fig. 14, Middle-upper Llandeilo, near Builth Road, Radnorshire (Powys), probably from Gwern-yfed-fach quarry, \ ml (805 m) SE of Builth Road station, SO 030526. Paratype. External mould of pygidium showing slight irregularity of rib develop- ment, SM A44731, x2. Fig. 11. ? Opsimasaphus sp. indet. B, p. 121. Lower Llandeilo, small quarry at SW end of Pen-cerig Lake, SO 043541. Mould of dorsal surface of hypostoma, It. 2840, x4. Fig. 12. Ogygiocaridinid gen. et sp. indet., p. 153. Lower Llandeilo, left bank of stream section east of Bach-y-graig, 40 yd (37 m) upstream of the point where the footpath enters the wood at the western end of the section, SO 071610. Internal mould of meraspid of degree 0, It. 2927, x20. Fig. 13. Cyclopygid gen. et sp. indet., p. 116. Lower Llandeilo, stream section 15 yd (14 m) SW of the old quarry 350yd (320m) west ot Maesgwynne, SO 059566. Internal mould of pygidium, It. 2839, x8. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 119 120 C. P. HUGHES field with only a very shallow preglabellar furrow. A slight, though distinct, constriction of the glabella developed approximately opposite (tr.) posterior of palpebral lobe. Anterior to this, glabella expands slightly, attaining its maximum width just anterior of palpebral lobe; obtusely rounded frontally. Lateral glabellar furrows poorly preserved, but basal pair prominent, declined steeply posteromedially commencing just posterior to mid-point of palpebral lobe, and de- limiting a posteriorly-tapering, wedge-shaped median lobe, with small median glabellar node at its posterior extremity. A further very weak pair of furrows may be developed just anterior to palpebral lobe and declined posteromedially at about 45° to sagittal line. Occipital furrow continuous; occipital ring distinct, with posterior margin slightly convex posteriorly. Anterior branches of facial suture widely divergent, subtending an angle of about 90° at the sagittal line, and intramarginal frontally. Frontal area large and nearly flat; low median ridge may extend across the frontal area (Figs 15, 16, 20), but may be an artefact of preservation. Palpebral lobe semicircular and more or less flat. Eye large, crescentic; no traces of lenses known. Immediately to the posterior, facial suture is more or less transversely directed and just in front of the posterior border furrow; laterally it is deflected to cut the posterior margin. Posterior border furrow well developed and straight. Posterior border convex and only slightly narrower (exsag.} than occipital ring. Librigena smooth, very gently convex, with wide concave border. Genal spine prominent, extending back to fifth thoracic segment. Cephalic doublure slightly wider than border; median suture present frontally. Thorax approximately rectangular in outline, though it tapers slightly to the posterior. Axial furrows deep, slightly scalloped and converging slightly towards the posterior. Deep apodemal pit developed in the axial furrow at posterior of each segment. Posterior band on the axial rings, anteriorly convex. Pleural furrow commences at inner anterior corner of pleura and directed slightly obliquely to anterior margin, dying out before it reaches the bluntly-pointed, posteriorly directed pleural termination. Doublure probably underlies about one-third of pleural width. Pygidium subparabolic with slender axis, occupying from one-sixth to one-eighth of pygidial width anteriorly and tapering uniformly to the rear where it is slightly less than half its anterior width. At least 17 axial rings, possibly as many as 20, are developed, with an apodemal pit, most clearly developed anteriorly, present in the axial furrow at posterior of each ring. Pleural field flat, bearing 14-16 well-defined ribs which bear weak furrows. Distally the ribs are deflected posteriorly, the point of deflection occurring at about two-thirds the way along the anterior ribs, moving progressively to about the mid-point on the shorter, more posteriorly placed ribs. SM A44731 shows slight irregularity in rib development on left pleural field. Border region wide with ribs only just extending onto it. Doublure bears terrace lines oblique to the inner margin; inner edge appears to be scalloped. DISCUSSION. This new species is clearly distinguished from all known species of Nobiliasaphus by the much higher number of pygidial ribs. There are also differences in outline of the pygidium and minor differences in cephalic proportions (see Kfiz & Pek 1972, 1974). Genus OPSIMASAPHUS Kielan, 1960 DIAGNOSIS. See Kielan 1960 : 75. TYPE SPECIES. Opsimasaphus jaanussoni Kielan, 1960. DISTRIBUTION. The genus ranges from the ? lower Llandeilo to Ashgill Series. It is present in the lower Llandeilo of Wales (?) (herein); the Caradoc of Ireland (Brenchley et al. 1967, M. Romano personal communication 1973) and northern Poland (Modlinski 1967); the Ashgill of Bohemia, Poland and Sweden (Kielan 1960), Britain (Ingham 1970) and Kazakhstan (Nikitin et al. 1968) and possibly of Quebec (Lesperance 1968); ? the lower Ordovician of France (Racheboeuf 1970). DISCUSSION. With the relatively recent acceptance of Nobiliasaphus as a separate genus many of the reported occurrences of Opsimasaphus must remain in some doubt until this whole group of asaphids is revised. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 121 ? Opsimasaphus sp. indet. A (Figs 18-19) FIGURED MATERIAL. It. 2925 (Fig. 19), internal mould of incomplete cranidium. Dimensions (in mm): A, 25-5. B, 18-6. J, 39-0. J1? 28-3. J2, 22-5. K2, 14-0. For explanation of symbols see Fig. 1. It. 2926 (Fig. 18), fragment of external mould of cranidium. LOCALITY AND HORIZON. Both figured specimens are from the Glyptograptus teretiusculus Shales in the old quarry 570 yd (521 m) north of Wye Cottage. DESCRIPTION. The single internal mould of an incomplete cranidium shows some similarities to Nobiliasaphus powysensis. However, the glabella is relatively longer, occupying nearly three- quarters of the cephalic length, and is relatively wider and much more clearly defined anteriorly. The palpebral lobe bears a distinct furrow. The external mould reveals that at least the frontal part of the glabella bears a sculpture of fine raised ridges lying subparallel to the preglabellar furrow. DISCUSSION. These two specimens are clearly of the Opsimasaphus - Nobiliasaphus type. They are here designated ? Opsimasaphus mainly because of the relatively long glabella which is well defined frontally, both of which are features characteristic of O. jaanussoni. However, more definite generic assignment must await the discovery of further specimens. ? Opsimasaphus sp. indet. B. (Fig. 11) FIGURED SPECIMEN. It. 2840, mould of dorsal surface of hypostoma. DIMENSIONS. See Fig. 17. Fig. 17. Diagram showing the dimensions of ? Opsimasaphus sp. indet. B. LOCALITY AND HORIZON. From the uppermost part of the Glyptograptus teretiusculus Shales in the quarry at the south-western end of Pen-cerig Lake. DESCRIPTION. Middle body subcircular, gently convex and not differentiated into anterior and posterior lobes. It is separated from border regions by a clearly-developed border furrow which 122 C. P. HUGHES is deepest posterolaterally. Lateral border widens posteriorly to reach its maximum width just behind the posterior of the middle body. Posterior border is about two-thirds as long (exsag.) as middle body, and deeply notched medially. Prominent circular maculae present, situated at base of the projections of the posterior border. DISCUSSION. This specimen is clearly similar to the hypostoma of Opsimasaphus. The placing of isolated hypostomata is, however, difficult and it is possible that it could belong to other asaphid genera having hypostomata with deeply notched posterior borders. Opsimasaphus is preferred to Nobiliasaphus as an assignment, because other possible ? Opsimasaphus specimens are known from beds of similar age within the Builth region. The hypostoma of Basilicus Salter, 1 849 shows strong similarities, but the present specimen lacks the anterior wings of that genus. If the placing of either ? Opsimasaphus sp. indet. A or B is correct, the occurrence in the lower Llandeilo is of interest as being the earliest record of the genus. Subfamily OGYGIOCARIDINAE Raymond, 1937 Genus O G YGINUS Raymond, 1912 DIAGNOSIS. The diagnosis of Whittard (1964 : 245-246) is followed here. TYPE SPECIES. Asaphus corndensis Murchison, 1839. DISTRIBUTION. The genus is apparently confined to Wales, Salop (Shropshire) and Brittany. Records of the genus from both North and South America (Kayser 1876 : 24; pi. 2, figs 3, 4; Kobayashi 1937 : 495; pi. 5, fig. 9; Rusconi 1950 : 84; Ross 1957 : 494; pi. 42, fig. 7) are very tentative; that of Kayser probably belongs to Ogygitella australis Harrington & Leanza (1957 : 142-143) and those of Ross and Kobayashi may be bathyurids and not asaphids. The specimen Figs 18-19. ? Opsimasaphus sp. indet. A, p. 121. Lower Llandeilo, old quarry 570yd (521 m) north of Wye Cottage, SO 047538. Fig. 18, external mould of fragment of cranidium showing surface sculpture on glabella, It.2926, x2. Fig. 19, internal mould of incomplete cranidium, It.2925, x2. Fig. 20. Nobiliasaphus powysensis sp. nov., p. 117. Middle-upper Llandeilo, quarry at Gwern-yfed-fach, i ml (805 m) SE of Builth Road station, SO 030526. Paratype. Internal mould of cranidium with associated external moulds of thoracic segments and pygidium, It. 2924, x 1-5. Figs 21-33. Ogyginus corndensis (Murchison), p. 126. Figs 21, 23, 26, 32-33, Lower Llanvirn, small quarries 600 yd (549 m) east of Upper Gilwern, SO 092582. Fig. 21, internal mould of small meraspid transitory pygidium of unknown degree, It.2849, x 20. Fig. 23, internal mould of transitory pygidium of unknown degree with two pleural ribs and probably two axial rings, It. 2875, x 20. Fig. 26, internal mould of entire meraspid of degree 4, showing the single pair of well-developed lateral glabellar furrows and continuous occipital furrow, It. 2873, x 10. Fig. 32, internal mould of meraspid of degree 5, showing single pair of lateral glabellar furrows and form of occipital furrow, It.2876, x 10. Fig. 33, internal mould of meraspid probably of degree 6, It. 2866, x 10. Figs 22, 31, Lower Llanvirn, eastern end of cliff section on left bank of Camnant Brook, midway between The Court and Pen-dre, SO 088567. Fig. 22, latex impression from external mould of meraspid transitory pygidium of unknown degree, showing trace of one furrow on the left pleural field and a posterior indentation, It. 2845, x 20. Fig. 31, internal mould of meraspid of degree 5, with continuous occipital furrow, It. 2846, x 10. Figs 24, 28-29, Lower Llandeilo, stream section 15 yd (14 m) SW of the old quarry, 350 yd (320 m) west of Maesgwynne, SO 059566. Fig. 24, latex impression from external mould of meraspid transitory pygidium, showing three pairs of ribs and axial rings and also a prominent posterior indentation, It.2880, x 20. Fig. 28, latex impression from external mould of meraspid transitory pygidium, with two pairs of pleural ribs, It. 2863, x 20. Fig. 29, latex impression from external mould of transitory pygidium of unknown degree, It.2864, x 20. Fig. 25, Lower Llandeilo, left bank of stream section east of Bach-y-graig, 65 yd (59 m) upstream of the point where the footpath enters the wood at the western end of the section, SO 072610. External mould of meraspid transitory pygidium of unknown degree, showing two ribs on the left pleural field and three on the right; note also the clear axial furrowing and posterior indentation, It.2881, x20. Figs 27, 30, Lower Llandeilo, left bank of stream section east of Bach-y-graig, 40 yd (37 m) upstream of the point where the footpath enters the wood at the western end of the section, SO 071610. Internal mould of meraspid of degree 4, showing details of the eye, It.2877. Fig. 27, x25; Fig. 30, x 10. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 123 124 C. P. HUGHES referred to Ogyginus sp. aff. corndensis from the lower Ordovician of South Korea (Kobayashi 1934 : 553; pi. 4, fig. 17) has since been placed in Birmaniles (Kobayashi 1950 : 527). The genus is known to range with certainty from the lower Llanvirn to the upper part of the lower Llandeilo Series, the occurrence in north-west France being of ? Arenig age (Henry 1971 : 66). The genus may also be represented in the Arenig (? D. hirundo Zone) of Whitesands Bay, Dyfed (Pem- brokeshire) (see Whittard 1964 : 246). f.s. C D Fig. 34. Diagram showing the position of the anterior branch of the facial suture in Ogyginus corn- densis (Murchison). A, It.2844 (Fig. 48); B, 59188; C, 59186 (Fig. 52); D, It.2865 (Fig. 58), showing anterior outline of hypostoma. f.s. - anterior branch of facial suture; h.s. - hypostomal suture. DISCUSSION. Whittard (1964: 245; pi. 42, fig. 3) believed the facial suture of Ogyginus to be marginal and not intramarginal as had been thought previously, though as advocated elsewhere (Hughes 19710: 179) his usage of the terms 'niobiform' and 'isoteliform' to describe the path of the suture anteriorly should be discontinued. However, I suggested (Hughes 1972) that Figs 35-43. Ogyginus corndensis (Murchison), p. 126. Figs 35, 37, 39-43, Lower Llanvirn, small quarries 600 yd (549 m) east of Upper Gilwern, SO 092582. Fig. 35, internal mould of meraspid of degree 6, showing slight frontal expansion of glabella and discontinuous occipital furrow, It. 2869, x 10. Fig. 37, latex impression from external mould of meraspid of degree 7, showing differentiation of occipital ring into an anterior and posterior region, It.2871, x 10. Fig. 39, internal mould of meraspid of degree 7, with continuous, though medially shallowing, occipital furrow, It. 2855, x 10. Fig. 40, internal mould of librigena with eye, It. 2870, x3. Fig. 41, internal mould of smallest known certain holaspis, It.2847, x 5. Fig. 42, internal mould of slightly larger holaspis, It. 2854, x 5. Fig. 43, internal mould of still larger holaspis, It. 2867, x5. Fig. 36, Lower Llandeilo, stream section at Wellfield Lodge immediately above where the stream is piped under the road, SO 044528. Internal mould of meraspid of degree 7, with continuous occipital furrow and genal spine extending back to fourth thoracic seg- ment, It. 2862, x 10. Fig. 38, Lower Llandeilo, left bank of stream section east of Bach-y-graig, 40yd (37 m) upstream of the point where the footpath enters the wood at the western end of the section, SO 071610. Internal mould of slightly disarticulated possible holaspis which has retained a continuous occipital furrow; note relatively large eye lenses, It. 2878, x 10. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 125 126 C. P. HUGHES Whittard was mistaken and that the suture was intramarginal as Salter (1866 : 130) originally described. Proof of this is difficult since the anterior portion of the cephalon is rarely well pre- served, presumably because the cuticle was thin. However, It. 2844 (Fig. 48) shows the anterior branch of the facial suture on the right side to be intramarginal, at least for the greater part of its course. It is believed that Whittard's interpretation (1964:250-251) of specimens 59188 and 59186 (see Figs 34, 52) is incorrect and that the position of the suture in these specimens is as Salter originally described. Against this it must be pointed out that specimens GSM 102135-6 (Whittard 1964 : pi. 42, figs 3, 4) indicate that the suture must be very nearly marginal in some individuals. Even allowing for the reinterpretation of the course of the facial suture, this genus and Ogygio- carella Harrington & Leanza, 1957 are clearly closely related. Differences which support the continued separation of the genera include the general lack of anterior glabellar expansion and the more pronounced furrowing in Ogygiocarella, together with the lack of spindle-shaped pleural furrows and the zetoidal axial furrow, and the greater number of pygidial ribs and axial rings. Hoekaspis Kobayashi, 1937, Megalaspidella Kobayashi, 1937 and Ogygiocaris Angelin, 1854 resemble Ogyginus in many ways, as does Ogygitella Harrington & Leanza, 1957, though this latter genus is easily distinguished by its deeply notched hypostoma. Ogyginus corndensis (Murchison, 1839) (Figs 2 1-6 1,66) 1839 Asaphus corndensis Murchison : 663. 1940 Ogyginus corndensis (Murchison) mut. intermedius Elles (pars) : 406, 428-429, 432; pi. 31, fig. 11 only. 1964 Ogyginus corndensis (Murchison); Whittard : 250-254; pi. 42, figs 2-8; pi. 43, fig. 1 (includes full synonymy). 1964 Ogyginus intermedius Elles; Whittard (pars) : pi. 41, fig. 2 only. 1964 Ogyginus corndensis (Murchison) var. septenarius Whittard : 254; pi. 43, figs 2-4. 1964 Ogyginus corndensis (Murchison) var. novenarius Whittard : 254; pi. 43, figs 5-1 1. 1966 Ogyginus corndensis (Murchison); Whittard : 302. 1966 Ogyginus corndensis (Murchison) var. septenarius Whittard; Whittard : 302. 1972 Ogyginus corndensis (Murchison); Hughes : 13-15; fig. 6. DIAGNOSIS. Ogyginus with eye situated at or slightly behind the mid-length. Thoracic axial furrows zetoidal except in meraspis. Pygidium with between seven and nine smooth ribs, approxi- mately ten axial rings. HOLOTYPE. GSM, Geol. Soc. Coll. GSb 4094 (Fig. 49). Dimensions: T.L. c. 66.0 mm; I c. 50-0 mm. For explanation of symbols see Fig. 1. DISTRIBUTION. The species is known from the upper Llanvirn and lower Llandeilo of the Shelve area (see Discussion, p. 135). In the Builth region the species first occurs in the upper part of the lower Llanvirn, and is present until the upper part of the lower Llandeilo. The species is at present unknown outside these two regions. DESCRIPTION. As this species was fully described by Whittard (1964 : 250-255) a full redescrip- tion is unnecessary, except for the hypostoma, of which complete examples are here described for the first time. Comment is made only on features where study of the Builth material adds to, or differs from, Whittard's description. Complete individuals are ovate, having a length : width ratio of about 3 : 2. Although data are only available for a limited number of complete individuals (see Fig. 44) of a sample from east of Upper Gilwern, these show no evidence of distinct wide and narrow forms. Further, the length- width data for 96 pygidia from the same locality fail to show such forms, although as expected some variation is present (see Fig. 56). Some cephala are slightly more angular anteriorly than others (Figs 48, 52). Three pairs of short, weak lateral glabellar furrows are developed. The anterior pair, rarely preserved, is situated BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 127 60- 50- 4O- 30- 20- 2- HOLASPIDES- Pygidial widtli(W) in mm. MERASPIDES Total length (TL) in mm. 20 3O 40 5O 60 80 Fig. 44. Graph showing differing regression lines for meraspides and holaspides of Ogyginus corndensis (Murchison). The two dashed lines represent total length : width ratios of 6 : 5 and 6 : 4, and illustrate that larger specimens are relatively wider but that distinct 'wide' and 'narrow' forms are not present. opposite anterior of palpebral lobe and directed slightly anteromedially, and extends about half-way to sagittal line. In common with the other lateral glabellar furrows, abaxial end barely reaches axial furrow (Figs 54, 55). Median pair consists of shorter, transversely-directed furrows situated approximately opposite mid-point of palpebral lobe, extending about half-way to sagittal line and failing to reach axial furrow. Like anterior pair they are rarely preserved. Basal pair most strongly developed and situated opposite posterior of palpebral lobe, being directed posteromedially at between 30° and 40°. They extend slightly further towards sagittal line than do the other two pairs (Figs 54, 55). Lateral occipital lobe virtually isolated by medial convergence of lateral occipital furrow and posterior band furrow (Fig. 55). Immediately anterior to posterior band is a small median glabellar node. On well-preserved internal moulds a low median ridge extends anterior to glabellar node terminating on frontal lobe opposite point where glabella is widest (Figs 55, 61). Posterior border uniform in width and relatively wide, being deflected only very slightly at the fulcrum. Eye length in young holaspides about one-third of cephalic length; in large specimens eye relatively smaller, the ratio being reduced to about one-fifth. Visual surface approximately crescentic but widens anteriorly (Fig. 40). Eye with up to two thousand hexagonal lenses, situated close to axial furrow, with mid-point between two-fifths and half of cephalic length from posterior margin. Posterior branch of facial suture, immediately posterior to palpebral lobe, makes angle 128 C. P. HUGHES of about 60° with sagittal line. Anterior branch follows a strongly arcuate path from anterior of palpebral lobe, being convex outwards, but is probably entirely dorsal anteriorly (Fig. 48) (see Discussion, p. 124). Hypostoma, excluding large anterior wings, slightly longer than wide. Arcuate middle furrow separates oval middle body into distinct anterior and posterior lobes. In all but the largest speci- men (Fig. 58), anterior lobe oval, long axis directed sagittally, and separated from anterior wing by distinct though shallow anterior border furrow (Fig. 59). In It. 2865, total length 15-5 mm, no border furrow is developed and anterior lobe merges anterolaterally with anterior wing. Posterior lobe crescentic and much shorter (sag.), occupying only about one-fifth total length. Prominent maculae developed at anterolateral corner of posterior lobe and characterized by their smooth surface. Posterior margin projects backwards sagittally to form median lip; posterior border furrow wide and shallow. Lateral border well developed, but shoulder not prominent. Lateral notch occurs at just over two-fifths the length from anterior. Both dorsal and ventral surfaces of hypostoma, with exception of maculae, bear widely-spaced terrace lines approximately concentric about mid-point of anterior border, but which become progressively more bowed sagittally posteriorly (Fig. 58). The median ridge on posterior half of the anterior lobe in It. 2860 (Fig. 60) is not thought to be an original feature as other moulds of the ventral surface fail to show any such ridge. Anterior margin of hypostoma abuts directly onto cephalic doublure, which widens to form a process flanking anterolateral margin of the anterior wing (Fig. 59). Median suture connects anterior branches of facial suture to hypostomal suture, all being functional. Posterior margin of hypostoma extends for only a short distance behind posterior of eye (Figs 59, 60); fixigena and hypostoma have been displaced slightly to the rear in the holotype (Fig. 49). Thorax rectangular; about twice as wide as long. Axis occupies about one-quarter of total width and tapers slightly to the rear in small specimens, but becomes nearly parallel-sided in larger individuals (Fig. 46). In very large specimens tapering is again apparent but restricted to posterior half of axis. Pygidium becomes relatively slightly longer with increase in length. Only anterior five or six axial rings well marked. Articulating half-ring furrows generally well defined, becoming shallower medially. Axis occupies up to nine-tenths of pygidial length in small specimens (Fig. 47), but in large specimens only about three-quarters (Fig. 52). Pleural field gently convex (tr.) with eight or nine smooth, unfurrowed, gently convex ribs. Due to post-mortem deformation faint furrows may be developed, particularly on anterior ribs. Irregular or asymmetrical develop- ment of ribs is very rare, being known on only four specimens (e.g. Figs 50, 51, 53). Anterior border deflected slightly to posterior laterally with large articulating facets bearing Figs 45-55. Ogyginus corndensis (Murchison), p. 126. Figs 45-47, 50-54, Lower Llanvirn, small quarries 600 yd (549 m) east of Upper Gilwern, SO 092582. Fig. 45, internal mould of small holaspis showing general adult form, It. 2859, x 5. Fig. 46, internal mould of thorax and pygidium with nine pleural ribs, It. 2848, x 1. Fig. 47, internal mould of small pygidium with only seven pleural ribs; note relatively long axis, It. 2858, x6. Fig. 50, external mould of pygidial fragment, showing irregular rib develop- ment, It. 2857, x2. Fig. 51, external mould of pygidium, showing incomplete separation of the fifth and sixth pleural ribs on the left pleural field (right side of specimen), It. 2872, x2. Fig. 52, internal mould of nearly complete individual with intramarginal facial suture (see p. 124) and relatively short pygidial axis, 59186, xf. Fig. 53, internal mould of pygidium showing incomplete development of furrow separating sixth and seventh pleural ribs, It. 2861, x2. Fig. 54, latex impression from external mould showing glabellar furrows, It. 2850, x2. Fig. 48, Lower Llandeilo, left bank of stream section east of Bach-y-graig, 80 yd (73 m) upstream of the point where the footpath enters the wood at the western end of the section, SO 072610. Specimen with exoskeleton preserved showing intramarginal facial suture, It. 2844, Y. 2. Fig. 49, Weston Beds, near Middleton, Salop (Shropshire). Holotype. Internal mould of damaged specimen; note weak development of eighth pleural rib. GSM Geol. Soc. Coll. GSb.4094, Y- 1. Fig. 55, Lower Llanvirn, left bank of stream SW of Gilwern, about 100 yd (91 m) from its source, SO 080581. Internal mould of cranidium showing details of occipital ring and node and median glabellar ridge, It.2842, x4. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 129 130 C. P. HUGHES [583] [SE] Pygidial length(Z) •% • * Maximum pygidial width (W) in mm o 4 £ 5 zb eb So £ Fig. 56. Plot of length and width of 96 pygidia of Ogyginus corndensis (Murchison) from the small quarries 600 yd (549 m) east of Upper Gilwern. + - two specimens. terrace lines; lateral and posterior borders well developed. Doublure relatively wide, being somewhat wider than borders. In small specimens it is generally poorly preserved, but is thought to be relatively narrower than in large specimens where it may be up to one-quarter of pygidial length (Fig. 52). Posteriorly 12-16 relatively widely spaced terrace lines are developed. These lie parallel to the margin on inner half, but are slightly oblique on outer parts. The obliqueness tends to increase towards anterolateral corners and extra lines may be intercalated (Fig. 52). Posteriorly, terrace lines on inner portion of doublure appear to reflect the posterior indentation found in the early meraspid stage. ONTOGENY. An almost complete suite of meraspid stages is known, principally from the small quarries east of Upper Gilwern. Figs 57-61, 66. Ogyginus corndensis (Murchison), p. 126. Figs 57-61, Lower Llanvirn, small quarries 600 yd (549 m) east of Upper Gilwern, SO 092582. Fig. 57, internal mould of typical pygidium and part of thorax, It. 2874, x3. Fig. 58, mould of dorsal surface of hypostoma, It. 2865, x3-5. Fig. 59, external mould of librigena with mould of dorsal surface of hypostoma and part of cephalic doublure, It. 2853, x 3. Fig. 60, mould of ventral surface of hypostoma, with internal mould of librigenae showing median suture, It. 2860, x2. Fig. 61, internal mould of cranidium, showing median pair of lateral glabellar furrows, It.2851, x4. Fig. 66, Lower Llanvirn, left bank of stream SW of Gilwern, about 100 yd (91 m) from its source, SO 080581. Enrolled specimen, It.2843, x4. Figs 62-64. Ogyginus intermedius Elles, p. 1 36. Figs 62, 64, Lower Llanvirn, left bank of upper reaches of Camnant Brook 210 yd (192 m) S 13° W of the fence crossing near the stream source, SO 088575. Fig. 62, internal mould of librigenae with mould of ventral surface of hypostoma, It. 2883, x 1. Fig. 64, internal mould of pygidium with ribs furrowed along entire length, It. 2882, xl. Fig. 63, Lower Llanvirn, left bank of upper reaches of Camnant Brook 240 yd (219 m) S 13° W of the fence crossing near the stream source, SO 088575. Internal mould of pygidium with nine ribs furrowed at distal end, It.2885, x3. Fig. 65. ? Ogyginus intermedius Elles, p. 137. Lower Llanvirn, left bank of Howey Brook, 30 yd (27 m) below the cliff section i ml (805 m) ESE of Carregwiber, SO 089582. Internal mould of meraspid transitory pygidium, It. 2889, xlO. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 131 -/ 132 C. P. HUGHES Degree uncertain. A single isolated globose transitory pygidium, 0-85 mm wide and 0-45 mm long, is roughly semicircular in outline with well-developed posterior indentation (Fig. 22). Axis well defined and tapers posteriorly. Only one pleural rib and at least one axial ring are developed, in contrast to the two ribs developed on transitory pygidium of the degree 0 ogygio- caridinid meraspis described below (p. 153). A further specimen (It. 2849) of similar dimensions (Fig. 21) has no ribs or axial rings preserved. The degree to which these specimens belong cannot be determined, though their size suggests that they may be degree 2 or 1, or possibly even degree 0. Several further isolated transitory pygidia are known which are slightly larger than these two specimens but still smaller than the smallest known transitory pygidium of degree 4 meraspis. These further specimens, whose width ranges from 1-2 to 1-5 mm, are very similar to the smaller ones except that two or three pleural ribs are developed (Figs 23, 24), with up to three axial rings. In all these specimens the ribs occupy, approximately, anterior half of pleural field and become less well developed posteriorly. These transitory pygidia are thought to belong to either degree 2 or 3. Degree 4. Two complete specimens have lengths of 2-9 mm and 3-2 mm, the cephalon account- ing for 1-2 mm and 1-5 mm respectively. Cephalon moderately convex (tr.), subsemicircular in outline, being very slightly less than twice as wide as long. Glabella well differentiated, but shows very little, if any, widening of anterior lobe. Single pair of short, clearly incised, lateral glabellar furrows reaches half-way to sagittal line at one-quarter the way along glabella, as measured from the occipital furrow. Occipital ring swollen and bowed backwards sagittally and clearly differentiated by the occipital furrow. Eye well developed, occupies about one-third of cephalic length and has relatively few large lenses (Figs 27. 30). Anterior branch of facial suture not seen, but posterior branch appears to follow a similar course to that of holaspid form. Genal region moderately convex with a pronounced, straight posterior border furrow; posterior border narrow medially, becoming slightly wider distally. The four thoracic segments, although preservation is not good, appear to be essentially similar to those of holaspides. Axis strongly convex (tr.), bounded laterally by well-developed axial furrows. Pleural region appears to possess the typical spindle-shaped pleural furrows, with the fulcrum occurring at about three-quarters of the length out from the axial furrow. Transitory pygidia approximately semicircular, 2 mm wide and strongly convex (tr.) with well-defined convex axial region. Anterior portion of the axis tapers gently to the rear, whilst posterior portion is parallel-sided. Up to nine axial rings developed with at least four ribs on strongly convex pleural field. Outer region of pleural field steeply declined, reaching maximum posteriorly where it is nearly vertical. Border relatively wide laterally and flat, being about one- tenth of pygidial length in width, becoming narrower posteriorly where a shallow terminal incurvation is developed medially. Degree 5. Two complete specimens, It. 2846 and It. 2876 (Figs 31, 32), are 3-3 and 3-9 mm long with cephala 1-5 and 1-7 mm respectively. Cephalon roughly semicircular, only a little over one and a half times as wide as long. Glabella moderately convex (//-.) with no frontal expansion. The single pair of lateral glabellar furrows differs from that of degree 4 in being directed postero- medially rather than transversely. Occipital ring swollen, clearly differentiated by deep occipital furrow which is transversely directed for a short distance laterally before bending to lie parallel to posterior margin of occipital ring (Fig. 32). This curved central portion of the furrow is con- siderably shallower than lateral parts, thus anticipating the condition found in degree 6 which lacks central portion of the furrow. Eye well developed, occupies about one-third of cephalic length. Anterior branch of facial suture not known; course of posterior branch similar to that of adult. The five thoracic segments are typically those of Ogyginus, except that axial furrow is more or less straight and shows little sign of the zetoidal shape it so characteristically assumes in the adult form. Transitory pygidia semicircular, 1-0 mm and 1-2 mm long and 2-1 mm and 2-3 mm wide respectively. Axis shows typical adult outline and has up to seven axial rings. Pleural field moderately convex with six or possibly seven ribs. Border region flat with maximum width equal BUILTH-LLANDRINDOD ORDOVICIAN TR1LOBITES 133 to one-ninth of pygidial length, becoming narrower and less well defined posteriorly. Preserva- tion is such that it is impossible to determine any posterior indentation. Degree 6. Only two specimens (It. 2869, Fig. 35 and SM A51177) certainly of degree 6 are known; they are about 5-4 mm and 4-2 mm in length, with cephala 2-2 mm and 1-7 mm long respectively. One further specimen (It. 2866, Fig. 33) may be of degree 6, but transitory pygidium has overridden the thorax slightly so that no definite count can be made of thoracic segments. Length of this specimen is also about 5-4 mm. Cephalon roughly semicircular in outline, though only just over one and a half times as wide as long. Although the anterior is poorly preserved, the well-defined glabella clearly expands frontally. Single pair of lateral glabellar furrows directed posteromedially as in degree 5. As in earlier degrees, occipital ring swollen and bowed backwards sagittally but occipital furrow no longer continuous, being represented by two short, transversely-directed laterally-placed furrows. Eye well developed and occupies about one-third of cephalic length. Course of anterior branch of facial suture not discernible; that of posterior branch as in adult. Posterior border furrow clearly developed and straight; posterior border narrow medially, widening laterally. Thoracic segments have the same general form as the adult. Fulcrum, however, situated three- fifths way along pleura as measured from axial furrow, and axial furrow has not attained typical zetoidal shape. Transitory pygidium only very slightly less than twice as wide as long. Axis clearly differen- tiated and same shape as in adults, with at least seven and possibly ten axial rings. Pleural region convex, being steeply declined posteriorly. Seven or eight ribs appear to be present. Border region varies in width between one-eighth and one-ninth of pygidial length and is nearly horizontal laterally, becoming narrower and less well defined posteriorly. Degree 7. Three complete specimens known (It.2855, Fig. 39; It. 2871, Fig. 37; It. 2862, Fig. 36). Preservation of It.2855 is superior to that of the other two and the following is based mainly on this specimen. Total length 5-5 mm (5-7 mm in It. 2871; 4-7 mm in It. 2862) with cephalon 2-2 mm in both It.2855 and It. 2871 (1-9 mm in It. 2862). Cephalon semicircular with glabella expanding frontally (see It. 2871). Single pair of lateral glabellar furrows directed transversely. Eye occupies approximately one-third of cephalic length. Occipital furrow in It.2855 entire and shows very little, if any, shallowing medially. It. 2871, however, exhibits differentiation of occipital ring into anterolateral lobes and a posterior band as is found in the adult form. In It. 2862 occipital furrow still appears continuous medially as in It.2855. Course of both anterior and posterior branches of facial suture appears to be similar to that in holaspides. In It. 2862 a short genal spine is present, reaching back as far as fourth thoracic segment. Hypostoma, although not well preserved, appears to be similar to that of adult (It.2855, Fig. 39). Thorax is essentially like that of adult form except that axial furrow has not attained the typical zetoidal shape. Transitory pygidium showing seven, possibly eight, pleural ribs in It.2855, but only five or six in It. 2871. Axial furrowing not well preserved, but there appear to be at least six axial rings. Border region horizontal laterally, in marked contrast to the steep outer portion of the pleural region, and about one-eighth of pygidial length in breadth. Posteriorly border region becomes narrower and less well defined. Young holaspides. The smallest specimen, total length 6-8 mm, is not well preserved (Fig. 41), but its length is about 1-2 times greater than that known for any meraspis. Occipital ring still somewhat swollen although posterior band clearly present. A possible young holaspis (It. 2878, Fig. 38) still retains a deep, continuous furrow. Eye slightly larger relative to the length of the cephalon than in meraspides, and set well to the posterior. Eye lenses are few and relatively large (Fig. 38). Short genal spine present. Thorax closely resembles that of larger individuals, except that axial furrow does not show the typical zetoidal shape. Pygidium relatively wide, just over two and a half times as wide as long. Posteriorly axial rings and pleural ribs poorly preserved, but only seven ribs appear to be developed. Pygidium less strongly convex and with less narrowing of the border region posteriorly than in meraspides. No posterior indentation developed. 134 C. P. HUGHES A slightly larger specimen (It. 2854, Fig. 42), length 7-1 mm, shows eight ribs present on the pleural field. DISCUSSION OF DEVELOPMENT. Table 1 shows that specimen size, as measured by its total length, is not by itself indicative of a particular meraspid degree; this may in part at least be a reflection of varying amounts of tectonic deformation. If, however, specimens from the single locality east of Upper Gilwern are considered alone, then there is a progressive increase in total length, although at times very small, with increase in meraspid degree. There is no constant growth factor between each degree (see Table 2); this may be due to the paucity of data or to there being more than one instar in some or all degrees. Table 1. Measurements and rib counts on meraspid and young holaspid specimens of Ogyginus corn- densis (Murchison). For explanation of symbols see Fig. 1. * - specimen from the small quarries 600yd (549 m) east of Upper Gilwern. All measurements in mm. Specimen T.L. Number A C Q C2 W Z Pleural ribs Axial rings Degree Left Right It.2845 It.2849* It.2875* It.2880 It.2863 - - - - - 0-85 0-85 1-2 1-3 c. 1-5 0-45 1 0-50 ? 0-70 2 0-85 3 c. 0-95 2 7 7 2 3 2 1+? 7 2+ ? 3+ ? 3+ ? 7 7 7 7 7 It.2881 - - - - - c. 1-5 0-80 2 3 3 7 It.2873* 2-9 1-2 - - - 2-0 1-0 4+ ? 4+ ? 6, ?7 4 It.2877 3-2 1-5 - 0-55 - 2-3 1-0 4+ ? 5+ ? 9 4 It.2846 3-3 1-5 - - - 2-1 1-0 4 + 6, ?7 7+ ? 5 It.2876* 3-9 1-7 0-8 0-6 0-3 2-3 1-2 5, ?6 5, ?6 5, ?6 5 SM A51177 4-2 c. 1-7 0-6 0-7 - 3-1 1-3 7 7 ?9 6 It.2866* c. 5-4 c. 2-3 - 0-8 - 3-2 1-4 7 6 + 8 ?6 It.2869* c. 5-4 c. 2-2 0-95 0-75 0-5 3-2 1-7 7 8 7, ?10 6 It.2862 4-7 1-9 - - - 3-1 1-4 7, ?8 7, ?8 4 + 7 It.2855* 5-5 2-2 1-05 0-75 0-4 3-4 1-6 7, ?8 7, ?8 6+ ? 7 It.2871* 5-7 2-2 - - - 3-7 1-7 5, ?6 5. ?6 7 7 It.2878 c. 5-1 c. 2-0 - 0-8 - 3-3 1-4 7 ?8 8 + ? holaspis It.2847* 6-8 c. 2-9 c. 1-4 c. 1-0 c. 0-5 4-5 1-9 ?7 ?7 7 holaspis It.2854* 7-1 2-8 - - - 4-0 1-8 8 8 6 + holaspis Table 2. Mean lengths (T.L.) and growth factors between meraspid degrees of Ogyginus corndensis (Murchison) from the small quarries 600 yd (549 m) east of Upper Gilwern. 8* taken as the smallest two holaspides known which are assumed to belong to the same instar. All measurements in mm. n - number of specimens. Degree Mean T.L. n Growth factor 4 5 6 7 8* 2-9 3-9 c. 5-4 5-6 6-95 1 1 2 2 2 1-33 c. 1-38 c. 1-04 1-24 The most interesting morphological feature in the ontogeny of the species is that of the development of the occipital ring and associated features. In general their development shows a progressive change towards the condition found in the adult form, with the deep continuous occipital furrow of degree 4 being reduced to short, transversely-directed furrows by degree 6, with the posterior band appearing in degree 7. However, there are some apparent exceptions, e.g. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 135 It. 2855 of degree 7 (Fig. 39) clearly shows a deep continuous occipital furrow with little shallow- ing medially; a similar condition is found in a ? young holaspis (Fig. 38). The earliest stage in which the differentiation of the occipital ring into anterolateral lobes and a posterior band occurs is degree 7. A latex impression from an external mould (Fig. 37) clearly shows this, although an internal mould of the same degree shows no such differentiation. In earlier degrees there are no such differences apparent between internal and external moulds. Whilst it is considered that specimens It. 2845-6 from the lower Llanvirn belong to O. corn- densis, the possibility remains that they may represent meraspides of Ogyginus intermedius Elles (see p. 136) which may be present at the same locality. The specimens are here assigned to O. corndensis on account of their similarity to other mersapides of this species, a holaspis of which is known from the locality. The lack of a distal furrow on the anterior rib may not pre- clude it from O. intermedius, as the ontogeny of the latter species is virtually unknown at present. It may be that early, if not all, meraspides of these two species are indistinguishable. BIOMETRICAL DATA. Abundant well-preserved material has yielded considerable data which are summarized in Tables 3 & 4. Table 3. Bivariate statistics for holaspid specimens of Ogyginus corndensis (Murchison) from the small quarries 600yd (549 m) east of Upper Gilwern. All measurements in mm. For explanation of symbols see Fig. 1. x: y var. x var. y var. a A d 12-99 21-15 2-66 0-28 ( )-93 0-93 0-57 0-0085 _ _ 7 A C3 12-99 21-15 5-60 4-10 ] •00 1-00 1-02 0-0013 0-44 0-0002 7 B J 19-80 159-78 11-80 56-78 ] •00 1-00 1-00 0-0006 0-60 0-0002 12 B J: 10-28 42-52 9-91 53-41 •00 1-00 1-13 0-0005 - - 18 B K2 9-86 35-40 7-09 21-99 •00 1-00 1-08 0-0003 - - 30 J Ji 20-79 177-87 12-01 84-81 ( )-99 1-00 1-16 0-0009 - - 9 K2 K! 7-03 21-37 5-84 14-90 •00 1-00 1-01 0-0003 0-84 0-0002 28 Ri Q 16-50 73-73 8-04 15-47 •00 1-00 0-95 0-0007 0-46 0-0002 8 Rn I 1^2 16-50 73-73 4-29 4-14 •00 1-00 0-92 0-0006 - - 8 Rn l K.4 17-14 82-16 17-63 85-42 •00 1-00 0-99 0-0008 1-02 0-0008 7 R2 R3 4-35 4-77 4-10 5-05 •00 1-00 1-08 0-0005 - - 17 R< RS 19-70 206-39 4-83 14-58 •00 1-00 1-07 0-0002 - - 13 W Z 13-66 61-90 7-09 19-03 ( )-99 1-00 1-06 0-0001 - - 88 W X 13-60 62-53 3-03 3-18 ( )-99 0-99 1-01 0-0001 0-23 0-00001 94 W, Y 11-09 39-61 6-07 13-32 ( )-99 0-99 1-05 0-0001 - - 107 Z Y 7-48 23-85 6-63 17-49 •00 1-00 0-97 0-00002 - - 123 Y X 6-38 14-11 3-07 3-16 ( )-99 0-99 0-98 0-0001 0-47 0-00003 131 Table 4. Measurements of hypostomata of Ogyginus corndensis (Murchison) from the small quarries 660yd (549 m) east of Upper Gilwern. First three specimens are moulds of the dorsal surface, the last two, moulds of the ventral surface. All measurements in mm. For explanation of symbols see Fig. 1. H2 H: H4 NMW 68.376.G208 5-3 4-6 3-7 2-4 _ 4-2 It.2853 9-0 8-1 6-5 4-0 c. 10-2 6-0 It.2865 15-5 13-0 11-3 7-0 18-5 13-0 lt.2860 10-2 8-9 7-0 5-0 10-7 6-0 NMW68.376.G209 10-6 9-5 7-5 4-5 - - DISCUSSION. Although the holotype is from the Weston Beds, near Middleton, Salop (Shrop- shire), Salter (1866) based his description and figures mainly on specimens from Builth where the species is very abundant. 136 C. P. HUGHES Although the forms occurring in the Builth and Shelve regions clearly belong to a single species, there is an interesting difference in the stratigraphical distribution of the various forms based on the number of pygidial ribs present. Whittard (1964 : 254) proposed the two varieties O. corndensis var. septenarius and O. corndensis var. novenarius to accommodate forms with seven and nine ribs. In the Builth region, unlike Salop (Shropshire), all three forms are found together in the Llanvirn, although the vast majority of specimens have eight ribs. In the lower Llandeilo, however, eight- and nine-rib forms occur in about equal numbers, with seven-rib forms occurring only rarely. In the Shelve region, however, eight-rib forms are restricted to the Llanvirn. Seven-rib forms, which as in the Builth region are rare, occur with eight-rib forms in the Llanvirn excluding the topmost beds (Betton Beds). The nine-rib form occurs with eight- rib forms in the Betton Beds, but is the only form occurring in the Llandeilo. Although the proportion of the various forms within an assemblage has a similar stratigraphical significance in both regions, it is felt that there is little justification for formal recognition of three sub- species. It seems likely the species was evolving under a selection pressure favouring an increased number of ribs, which differed in the two areas. Whittard (1964 : 253) indicated the distinctions between O. corndensis and O. grandis Whittard, 1964, O. porcatus Whittard, 1964 and O. intermedius Elles, 1940. To these may be added O. armoricanus (Tromelin & Lebesconte, 1876), redescribed by Henry (1971 : 66-68), which is distinguished by its short genal spine, lack of well-developed occipital ring and lack of occipital node. Two possible examples of enrolled specimens are known but both are flattened and may be the result of the disarticulated parts of an exuvia being preserved on top of one another. Ogyginus intermedius Elles, 1940 (Figs 62-64, 68, 71-73, 75) 1940 Ogyginus corndensis (Murchison) mut. intermedius Elles (pars) : 395, 397-398, 428-429; pi. 31, fig. 10, non fig. 11. 1964 Ogyginus intermedius Elles; Whittard (pars) : 246-248; pi. 41, figs 1, 3, 4, non fig. 2. 1966 Ogyginus intermedius Elles; Whittard : 302, 305. DIAGNOSIS. Ogyginus with relatively small eye situated well to the anterior; glabella expanding onto anterior border. Nine or ten ribs on pleural field with short pleural furrow on distal portion of anterior ribs. HOLOTYPE. SM A10087 (Fig. 73), internal mould of complete specimen. Dimensions (in mm): T.L., 61-7. A, 20-3. C, c. 11-2. Q, c. 3-0. C2, 6-7. C3, c. 8-2. Il7 c. 41-1. K2, 13-5. Q, 18-9. Qlt 2-5. R1? 38-2. R3, 8-0. R4, 38-4. W, 38-0. X, 7-8. Y, 18-7. Z, 22-5. For explanation of symbols see Fig. 1. TYPE LOCALITY AND HORIZON. Didymograptus bifidus Shales south-east of Elusendy, Golden Grove, near Llandeilo. DISTRIBUTION. Elles (1940 : 429) originally recorded the species from the lower Llanvirn of the Llandeilo and Builth regions together with rare occurrences in the upper Llanvirn and Llandeilo of Builth. Whittard (1964 : 248) found the species to be restricted to the lower Llanvirn in the Shelve region, and noted that it was also present in the bifidus fauna from west of Llan Mill, Dyfed (Pembrokeshire). Specimens identified as O. intermedius from the upper Llanvirn and lower Llandeilo of the Builth region by Elles are here placed in O. corndensis, so that the species is now restricted to the lower Llanvirn throughout the Anglo- Welsh region. In the Builth district the species is most common in the beds exposed in the upper reaches of the Camnant Brook, where it is more abundant than O. corndensis. The only other localities within the area where it is certainly present are the 'cliff' section, Howey Brook, Hendy Bank and Frank's Bridge. Elles (1940 : 398) also records it from east of Bwlchyfedwen, but no speci- mens have been traced to confirm this. DESCRIPTION. Comment is made only on features in which the Builth material differs from, or adds to, Whittard's redescription (1964: 246-248). Cephalon known in the Builth region from BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 137 very few rather poor fragmentary specimens. Three pairs of lateral glabellar furrows appear to be present together with a discontinuous occipital furrow (Fig. 68), although it is uncertain which depressions are true lateral glabellar furrows and which are due to compression. Anterior and median pairs situated approximately opposite small palpebral lobe, with anterior pair directed slightly anteromedially and median pair more or less transverse. Posterior pair situated about half-way between median pair and occipital furrow and directed posteromedially. Lateral glabellar furrows extend no more than about three-quarters of the way to sagittal line (Fig. 68). Thorax known only from a few poorly-preserved specimens. However, some larger specimens have an axial furrow of moderately well-developed zetoidal form, whilst others show the sup- posedly more characteristic scalloping (Fig. 68). Pygidium with axis tapering over its anterior half and well defined by deep axial furrow. Eleven or twelve axial rings may be developed, together with a short terminal piece. Ring furrows continuous though considerably shallower medially. Nine or ten ribs may be developed (Figs 63, 71), with short, shallow furrows present on the distal portions of up to the first six ribs (Fig. 63). The furrows along the entire length of the ribs in It. 2882 (Fig. 64) are thought to be the result of compression. It. 2886 shows a rib which apparently bifurcates on the right pleural field (Fig. 75). Doublure relatively narrow, in width only about one-eighth of pygidial length and bearing terrace lines. A single slightly distorted isolate pygidium is known which may represent an immature example of this species (Fig. 65). It shows at least eight axial rings. Although they are poorly defined posteriorly it appears that only eight ribs are present, the anterior one having a short furrow present distally. The presence of the furrow on the anterior rib suggests that this may belong to O. intermedius rather than O. corndensis. The development of only eight ribs suggests that the specimen may be of a late stage meraspis rather than a young holaspis. In size it is clearly comparable to meraspides of O. corndensis of degree 7. DISCUSSION. About 30 specimens are known from Builth and they show more variation in the number of pygidial axial rings and pleural ribs than those (not more than 15) known from the Shelve area (Whittard 1964 : 246-248; 1966 : 302). In the light of the variation in the number of pygidial ribs in O. corndensis it is not surprising to find a similar variation in O. intermedius. The moderately well-developed zetoidal aspect of the axial furrow on the thorax of some large specimens may reflect inherent variation, as in O. corndensis. Whittard (1964:246, 248) considered Ogygiocaris henningsmoeni conspecific with O. inter- medius. Struve's (1962) description and illustrations were of rather incomplete material and a detailed comparison is not possible; however, the glabellar furrowing and apparently larger, more posteriorly placed eye in O. henningsmoeni suggest that the two forms are distinct. The species is distinguished from all other Ogyginus species by the small eye situated well forward, and a glabella encroaching onto the anterior border. Isolate pygidia with nine ribs are very difficult to distinguish from nine-rib forms of O. corndensis, though the short furrow on the distal portion of the anterior ribs in O. intermedius may be utilized if compression is not too great. For a fuller comparison with related species see the discussion of O. corndensis (p. 135) and Whittard (1964 : 253). Ogyginus cf. intermedius Elles, 1940 (Fig. 177) FIGURED SPECIMEN. It. 2890, internal mould of pygidium. LOCALITY AND HORIZON. Didymograptus bifidus Shales, left bank of the stream section in the upper reaches of the Camnant Brook, 270 yd (247 m) below the fence crossing the stream above the ravine. DESCRIPTION. This pygidium is very similar to pygidia here attributed to O. intermedius, except that the doublure is considerably wider, the width being slightly over one-quarter of the total pygidial length. 138 C. P. HUGHES DISCUSSION. It has been shown that in the closely-related species O. corndensis the relative width of the doublure is dependent on the size of the individual and increases with increase in size. Since this specimen is no larger than specimens of O. intermedius from the same locality, the wider doublure would seem to indicate a significant difference between this specimen and O. intermedius. Ogyginus 1 laticostatus (Salter, 1866) (Fig. 109) 1851 Isotelus (Basilicus) ? laticostatus (Green sp.); Sedgwick & M'Coy (pars) : 170; pi. IE, fig. 18a only (non laticostatus of Green, 1832). 1852 Isotelus laticostatus (Green sp.); Sedgwick & M'Coy : 366. 1866 Asaphus (Basilicus) laticostatus Sedgwick & M'Coy; Salter : 158-159; pi. 18, fig. 6. 1931 Ogyginus ? laticostatus (McCoy); Reed : 462. 1953 'Asaphus' laticostatus M'Coy; Jaanusson : 445. HOLOTYPE. SM A16693, internal mould of large pygidium. Holotype by monotypy. LOCALITY AND HORIZON. This specimen was originally cited as having been collected from Maen Goran, Llangollen (Sedgwick & M'Coy 1851: 170). Subsequently (Salter 1866: 159; 1866fl : 311-312) stated that it was from Maen Goran, Builth. No Maen Goran has been traced in either district, but there is a house named Maen Cowyn at Llanelwedd just north of Builth. This may be a modern spelling of Maen Goran (see note by G. L. Elles in the Sedgwick Museum Catalogue under the entry for the holotype); it might be noted, however, that Maen Cowyn was in use as early as 1894 (Woods 1894 : 572). Thus it is thought possible that the specimen came from near Maen Cowyn, possibly from the small track-side quarry, about 270 yd (247 m) west of Maen Cowyn, in ashy beds possibly of upper Llanvirn age. DESCRIPTION. Cephalon and thorax unknown. Pygidium, though damaged, is roughly semicircular in outline. Axis slightly under one-quarter of pygidial width anteriorly, tapering posteriorly and defined by deep axial furrow. Exact number of axial rings uncertain but at least ten developed. Ring furrows continuous though very much shallower medially than laterally. At least nine, and possibly ten, flat-topped pleural ribs present. Doublure relatively narrow, being about one-seventh of the pygidial length in width. It appears to have a simple arcuate inner margin and bears terrace lines. DISCUSSION. If the location and age of this specimen are correct it represents the only occurrence of a trilobite from the upper Llanvirn in the southernmost part of the Inlier. Figs 67, 69-70, 74. Ogygiocarella debuchii (Brongniart), p. 142. Fig. 67, Lower Llandeilo, old quarry 350yd (320 m) west of Maesgwynne, SO 059566. Internal mould of pygidium, showing doublure and form of axial rings, It.2900, x 1-5. Figs 69-70, Lower Llandeilo, old quarry 570yd (521 m) north of Wye Cottage, SO 047538. Fig. 69, internal mould of pygidium showing furrowing of axis, It.2904, x2. Fig. 70, external mould of librigena with mould of dorsal surface of doublure and part of hypo- stoma, It.2911, xl-5. Fig. 74, Lower Llandeilo, stream section 15yd (14m) SW of the old quarry 350 yd (320 m) west of Maesgwynne, SO 059566. Internal mould of damaged cranidium showing sculpture of fine ridges on glabella and occipital ring, It. 2899, x2-5. Figs 68, 71-73, 75. Ogyginus intermedius Elles, p. 136. Figs 68, 75, Lower Llanvirn, right bank of upper reaches of Camnant Brook 175yd (160m) S 13° W of the fence crossing near the stream source, SO 088575. Fig. 68, internal mould of cranidium and part of thorax, showing glabellar furrows and scalloped nature of thoracic axial furrow, It. 2887, x 1. Fig. 75, internal mould of pygidium showing irregular rib development, It. 2886, xl. Fig. 71, Lower Llanvirn, right bank of upper reaches of Camnant Brook 220yd (201 m) S 13° W of the fence crossing near the stream source, SO 088575. Internal mould of small pygidium with an incipient tenth pleural rib developed, It. 2884, x7-5. Fig. 72, ? Lower Llanvirn, Frank's Bridge, Builth. Internal mould of complete large individual, SM A45484, xO-5. Fig. 73, Lower Llanvirn, SE of Elusendy, Golden Grove, Llandeilo. Holotype. Internal mould of complete specimen, SM A10087, x 1. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 139 140 C. P. HUGHES Salter (1866: 158-159; 18660:311-312) commented on M'Coy's inexplicable reference of this specimen to Isotelus (Basilicus ?) laticostatus Green, 1832, which Salter believed was in fact a phacopid, and indicated that the specimen clearly represented a distinct form with genuine asaphid affinities. Reed (1931 : 462) considered that the annulation of the axis, the relative width of the doublure and the nature of the pleural ribs indicated close affinity to Ogyginus, a sugges- tion also given some support by Jaanusson (1953 : 445-446). The specimen is very similar to large pygidia of O. intermedius and a tentative generic assignment to Ogyginus would seem the most plausible proposition. In view of the relatively poor preservation and the uncertainties regarding the age and locality of this specimen it is proposed that the species name should be retained but restricted to the holotype. Genus OGYG1OCARELLA Harrington & Leanza, 1957 DIAGNOSIS. Outline oval, isopygous; cephalon and pygidium subsemicircular. Glabella parallel- sided posteriorly, expanding slightly frontally; four pairs of lateral glabellar furrows; eye situated in posterior half of cephalon; anterior branch of facial suture marginal frontally. Hypostoma with entire posterior margin. Thorax with scalloped axial furrow. Pygidium generally with 1 1 or more ribs and about 14 axial rings; inner edge of doublure sinuous. TYPE SPECIES. Asaphus debuchii (Brongniart, 1822). By original designation of Harrington & Leanza (1957 : 160-161), where it is misquoted as 'Asaphus debuchianus (Brongniart, 1822)' (see Jaanusson in Moore 1959 : O352; Jaanusson 1956; Whittard 1964 : 255-256; Hughes 1972)1. DISTRIBUTION. Genus characteristic of the Anglo-Welsh region, particularly south Wales and Salop (Shropshire); only a single specimen is known from north Wales (MacGregor 1963 : 792-793). Also present in the Lower Llanvirn of Peru (Newell & Tafur 1944: pi. 92, figs 2-4; Hughes, Rickards & Williams in press). DISCUSSION. Despite the ^interpretation of the course of the facial suture in Ogyginus (p. 124) the differences between Ogygiocarella and Ogyginus are few. However, the difference in facial suture pattern together with the greater expansion of the frontal glabellar lobe, the zetoidal axial furrow, the non-sinuate inner margin of the pygidial doublure and the generally lower number of pleural ribs in Ogyginus are considered to be sufficient to retain them as separate genera. Ogygiocarella also shows similarities to Ogygiocaris Angelin, 1854, but it may be distinguished by its marginal facial suture anteriorly, narrower cephalic doublure and lack of any posterior indentation in holaspides. 1 For suppression of Asaphus debuchianus Brongniart in Desmarest, 1817, non 1822, see Opin. Decl. int. Commn zool. Norn., London, 18 (4) : 241-256 (1958). Figs 76-85. Ogygiocarella debuchii (Brongniart), p. 142. Figs 76-77, 82, Lower Llandeilo, stream section 15yd (14m) SW of the old quarry 350yd (320m) west of Maesgwynne, SO 059566. Fig. 76, latex impression from external mould of the smallest known transitory pygidium, It. 2897, x20. Fig. 77, internal mould of transitory pygidium showing at least three ribs, It. 2896, x 20. Fig. 82, latex impression from external mould of meraspid of degree 4, It.2898, x 15. Fig. 78, Lower Llandeilo, 160 yd (146 m) SE of Tre Coed, on left bank of stream, SO 054552. Internal mould of transitory pygidium with eight ribs, It. 2892, x 10. Figs 79, 81, 83-84, Lower Llandeilo, left bank of stream section east of Bach-y- graig, 45 yd (41 m) upstream of the point where the footpath enters the wood at the western end of the section, SO 071610. Figs 79, 83, internal mould of longitudinally compressed meraspid of degree 7, It.2913. Fig. 79, x 15; Fig. 83, x 10. Fig. 81, details of eye, It.2894, x 15. Fig. 84, internal mould of transversely compressed meraspid of degree 7, It.2893, x 10. Fig. 80, Lower Llandeilo, old quarry 570 yd (521 m) north of Wye Cottage, SO 047538. Internal mould of librigena and mould of ventral surface of doublure, It.2902, xl-5. Fig. 85, Lower Llandeilo, stream section at Wellfield Lodge immediately above where the stream is piped under the road, SO 044528. Internal mould of young holaspis, It. 2891, x5. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 141 '85 142 C. P. HUGHES Ogygiocarella debuchii (Brongniart, 1822) (Figs 67, 69-70, 74, 76-86, 88-89, 91-92, 97) 1843 Ogygia Buchii (Brongniart); Goldfuss : 555. 1846 Asaphus Buchii Brongniart; De la Beche : 31. 1940 Ogygiocaris buchi (Brongniart) Elles (pars) : 406-412. 1964 Ogygiocarella debuchii (Brongniart); Whittard : 255-261 ; pi. 44, figs 1-1 1 ; pi. 45, figs 1-8 (includes fuller synonymy). 1966 Ogygiocarella debuchii (Brongniart) ; Whittard : 302. 1970 Ogygiocarella debuchii (Brongniart); Toghill : 122. 1972 Ogygiocarella debuchii (Brongniart); Hughes : 7-17. DIAGNOSIS. Ogygiocarella generally having eleven pygidial ribs, rarely ten or twelve. LECTOTYPE. Whittard (1964 : 261) selected as lectotype the specimen on which Brongniart (1822) based his plate 2, fig. 2A. This was unsatisfactory since the whereabouts of this specimen was, and is still, unknown. Nevertheless, it is proposed to accept Whittard's designation in order to preserve current usage of the name O. debuchii (see also Whittard 1964 : 256; Hughes 1972). DISTRIBUTION. The species is common throughout much of south and central Wales and Salop (Shropshire), although some differences in stratigraphical range occur in different areas. The species first occurs near the top of the upper Llanvirn of the Builth district. De la Beche (1846 : 31) reported the species from the Grey Feldspar Sands at Tan-y-graig at a horizon somewhat lower than this, but no specimens have been located in substantiation. It persists until just below the top of the lower Llandeilo. In Salop it does not appear until the upper part of the lower Llandeilo (upper part of the Meadowtown Beds) and continues into the upper Llandeilo (Rorrington Beds), while in the Llandeilo region the species appears to be restricted to the Llandeilo. Numer- ous occurrences of the species have also been recorded from south Wales, but until the trilobites of this region are revised there is some doubt as to whether they can all be referred to O. debuchii, though it is certainly present in the Hendre Shales at Mydrim (Toghill 1970 : 122). DESCRIPTION. The following descriptive notes supplement Whittard's (1964:257-261) descrip- tion. Complete individuals are oval in outline being about one and a half times longer than wide. Some variation occurs in length/width ratio, but there is no evidence of the existence of distinct wide and narrow forms (Fig. 101) as claimed by Whittard (1964 : 257). Cephalon approximately two and a half times as wide as long. In uncompressed specimens, unfurrowed frontal portion of the glabella moderately convex (//•.) with steep convex frontal face (cf. 59207, Fig. 86; SM A49721). Total length of glabella and occipital ring about one and three-quarters that of maximum glabellar width. Palpebral lobe situated entirely in posterior half of cephalon, mid-point being only about one-third of cephalic length from posterior margin. Eye moderately large, with all lenses approximately equal in size and hexagonal in outline. A relatively small specimen (eye length 3-0 mm) has visual surface with approximately 1000 Figs 86, 88-89, 91-92. Ogygiocarella debuchii (Brongniart), above. Fig. 86, unknown loc. in the Builth district. External mould of specimen lacking librigenae, 59207, x 1. Fig. 88, ? Llandeilo, loc. unknown but probably from the Llandeilo district. Internal mould of complete specimen showing well-developed nodes on the thoracic and pygidial axis, GSM 12880, x 1. Figs 89, 92, Lower Llandeilo, old quarry 570 yd (521 m) north of Wye Cottage, SO 047538. Fig. 89, internal mould of pygidium with twelve ribs, It.2908, xl-5. Fig. 92, details of eye lenses, It.2910, x 10. Fig. 91, Lower Llandeilo, left bank of stream section east of Bach-y-graig, 20 yd (18 m) upstream of the point where the footpath enters the wood at the western end of the section, SO 072610. Internal mould of cephalon and part of thorax prepared to reveal mould of ventral surface of hypostoma, It. 2895, x 2. Figs 87, 90. Ogygiocarella angustissima (Salter), p. 150. Fig. 87, Middle-upper Llandeilo, quarry at Gwern-yfed-fach, i ml (805 m) SE of Builth Road station, SO 030526. Internal mould of small pygidium with a slight posterior indentation, It. 2923, x7. Fig. 90, Middle-upper Llandeilo, middle quarry, Llanfawr, Llandrindod, SO 066617. Mould of dorsal surface of hypostoma, It. 2917, x3. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 143 Wii i 92 144 C. P. HUGHES 90- 80- 70- 60- 50^ 40- 30- 20- 10- Number of specimens 94i 10-fi 13 13+i Number of pygidial ribs Fig. 93. Histogram of the number of pygidial ribs developed in Ogygiocarella debuchii (Brongniart) (left peak) and Ogygiocarella angustissima (Salter) (right peak). lenses of about 0-033 mm diameter (Fig. 81). A larger specimen (eye length 8-5 mm) has approxi- mately 4000 lenses of 0-055 mm diameter (Fig. 92). Visual surface approximately crescentic in. outline, appearing to be widest anteriorly in dorsal view owing to the steeper inclination of the visual surface posteriorly. External surface of frontal glabellar lobe bears a Bertillon pattern of fine, closely-spaced ridges (Figs 74, 86). Similar ridges are also present medially on posterior of glabella and occipital Figs 94-96, 98-100. Ogygiocarella angustissima (Salter), p. 150. Figs 94-95, Middle-upper Llandeilo, loc. uncertain (see p. 151). Fig. 94, Lectotype. Internal mould, 59198, vl. Fig. 95, Paralectotype. Internal mould showing glabellar furrowing, 59199a, / \-5. Figs 96, 100, Middle-upper Llandeilo, middle quarry, Llanfawr, Llandrindod, SO 066617. Fig. 96, part of cephalic doublure, showing possible panderian node near posterior margin, It. 2919, x3. Fig. 100, external mould of pygidial fragment showing surface sculpture, It. 2914, /3. Fig. 98, ? Middle-upper Llandeilo, loc. unknown but in the Builth district. Internal mould of virtually complete specimen having fourteen pygidial ribs, OUM B.123, xl (figured as Asaphus debuchii by Brongniart 1822: pi. 2, fig. 2c). Fig. 99, Middle-upper Llandeilo, ? quarry at Gwern-yfed-fach, i ml (805 m) SE of Builth Road station, SO 030526. Internal mould of large pygidium and posterior part of thorax, showing sinuate inner edge of pygidial doublure, I.2857b, xf Fig. 97. Ogygiocarella debuchii (Brongniart), p. 142. Lower Llandeilo, old quarry 570yd (521 m) north of Wye Cottage, SO 047538. Internal mould of relatively narrow pygidium, It. 2906, x 1. BUILTH-LLANDR1NDOD ORDOVICIAN TRILOBITES 145 'j**&ssis&£K,. V 100 146 C. P. HUGHES 20- 10- Cephalic length (A) in mm. Glabellar length (B) in mm. Posterior cranidial width (J) in mm. 10 20 30 40 50 60 Fig. 101. Plots showing similarities between the cephalic dimensions of Ogygiocarella debuchii (Brongniart) ( • O) and Ogygiocarella angustissima (Salter) ( + A)- Points enclosed by a circle are approximate. ring (Fig. 74); here, however, they are not so markedly convex forwards. Remaining portions of dorsal cephalic exoskeleton smooth. Doublure moderately wide, though narrower anteriorly owing to location of hypostoma, and bearing fine terrace lines. Thorax tapering slightly posteriorly. Axial furrows also converge slightly posteriorly. Nodes on axial rings occasionally developed in the Builth material; the only specimen known with these fully developed is GSM 12880, probably from Llandeilo (Fig. 88). No panderian openings or protuberances developed. Pygidial axis infundibular, occupying slightly over one-sixth of pygidial width anteriorly and with up to 13 axial rings (rather than the 14 mentioned by Whittard), plus terminal piece. Generally, however, only between seven and nine axial rings are clearly differentiated, with little trace of furrowing on posterior of axis. Posteriorly apodemal pits in axial furrow become progressively less pronounced, disappearing at about the eighth or ninth axial ring. Typically eleven well-developed ribs present (see Table 11, p. 150). Posterior band of ribs is more convex (exsag.) than anterior one and distal to paradoublural line; ribs are deflected slightly more to the rear, this becoming progressively less posteriorly. Externally posterior bands bear raised lines lying oblique to interpleural furrows; distally these lines become subparallel to the margin Figs 102-108, 110. Ogygiocarella angustissima (Salter), p. 150. Figs 102-103, 105-107, Middle-upper Llandeilo, middle quarry, Llanfawr, Llandrindod, SO 066617. Fig. 102, external mould of pygidium showing apodemal pits, It. 2921, x 2. Fig. 103, latex impression from external mould showing sculpture, form of occipital ring, fixigenae and palpebral lobes, It. 2918, x2. Fig. 105, internal mould of eye showing traces of eye lenses, It. 2920, x 10. Fig. 106, internal mould of damaged librigena showing sculpture and median suture, It. 291 5, x 3. Fig. 107, internal mould of part of thoracic segment, showing form and sculpture of the thoracic doublure, It. 2916, x2. Figs 104, 110, Middle-upper Llandeilo, quarry at Gwern-yfed-fach, i ml (805m) SE of Builth Road station, SO 030526. Fig. 104, internal mould of librigena showing sculpture of cephalic doublure, It. 2922, x7. Fig. 110, external mould showing slight irregularity in the eighth rib on the left pleural field, 59206, xO-5. Fig. 108, Middle- upper Llandeilo, ? Harper's quarry, NE of Wellfield, SO 038532. Internal mould of nearly complete specimen showing general form and position of the hypostoma, GSM 12918, x 1. Fig. 109. Ogyginus ? laticostatus (Salter), p. 138. ? Upper Llanvirn, 'Maen Goran, Builth'; ? small quarry about 270 yd (247 m) west of Maen Cowyn, SO 054525 (see p. 138). Internal mould of pygidium, SMA16693, xl. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 147 ' /" •''' 1 148 C. P. HUGHES and extend onto the distal extremities of anterior bands (Fig. 86). Strong terrace lines developed on doublure subparallel to margins, becoming bunched together in post-axial region (Fig. 86). ONTOGENY. Although only a few meraspid specimens are known from the Builth region, they show some interesting differences from the ontogenetic series described by Whittard (1964: 259-260) from the Shelve region. The Builth material consists of one entire specimen of degree 4 and two of degree 7, together with a few isolated transitory pygidia of uncertain degree. Isolated transitory pygidia. Smallest known specimen, It. 2897 (Fig. 76) is 1-3 mm wide and 0-8 mm long, roughly semicircular, globose, with a prominent posterior indentation. At least two ribs are present on anterior part of pleural field. Axial region approximately parallel-sided and weakly furrowed. Posterior and lateral borders well developed and flat. It. 2896 (Fig. 77), width and length 1-6 mm and 0-8 mm respectively, is very similar but has at least three ribs developed. It. 2892 (Fig. 78), 3-0 mm wide and 1-7 mm long, has eight ribs, the anterior ones having well-developed furrows along their entire length (//-.). The development of eight ribs would suggest, by comparison with Whittard's ontogenetic series, degree 3 or 4, but the well- developed pleural furrows are more characteristic of his degree 7 specimens. Degree 4. Single known specimen, It. 2898 (Fig. 82) is about 4-0 mm long. Cephalon not well preserved axially, but occipital ring appears simple, bowed posteriorly with continuous occipital furrow and no lateral occipital lobes. Basal pair of lateral glabellar furrows well developed. Eye large, almost half the cephalic length, with relatively few (up to 100) large lenses. Thorax slightly disarticulated, but appears to consist of only four segments, similar to those of holaspides but with relatively deeper pleural furrows. Transitory pygidium semielliptical, nearly three times as wide as long. Seven unfurrowed ribs developed. Axis with at least six axial rings and extending close to indented posterior margin. Degree 7. Two entire specimens are 6-5 mm and 8-2 mm long (see Table 5). Although basically similar to holaspides, the cephalon differs in various ways. Occipital ring simple with continuous furrow and no lateral occipital lobes. Glabella progressively expands slightly anteriorly and has only three pairs of lateral glabellar furrows, the anterior pair of the adult being absent. Eye still relatively large, being only just under half the cephalic length. Posterior border furrow dies out distally approximately in line (exsag.) with the fulcrum. As in holaspides, genal spine relatively short, reaching only to fourth thoracic segment. The seven thoracic segments are as in holaspides. Transitory pygidium semielliptical, with 9-11 ribs which may be shallowly furrowed along their entire length (tr.) (Fig. 83). Ribs deflected slightly posteriorly on reaching paradoublural line. Axis infundibular with ten axial rings and a terminal piece. No posterior indentation developed. Apart from being slightly larger, specimen It. 2893 (Fig. 84) is similar. Table 5. Measurements and rib counts on meraspid specimens of Ogygiocarella debuchii (Brongniart). All measurements in mm. For explanation of symbols see Fig. 1. Specimen T.L. AW Z Ribs Degree It.2897 _ 1-3 0-8 2+1 ? It.2896 - - 1-6 0-8 3 ? It.2892 — — 3-0 1-7 8 ? It.2898 c. 4-0 c. 2-0 3-0 c. 1-1 7 4 It.2913 6-5 2-9 5-2 1-9 9 7 lt.2893 8-2 3-3 4-7 2-2 7 + 7 DISCUSSION OF ONTOGENY. The series described by Whittard shows that transitory pygidia of degree 1 probably had six ribs and suggests that degree 0 transitory pygidia possessed fewer. The present study, however, has revealed meraspid transitory pygidia with only two or three ribs developed. Whether these are of degree 0 or not cannot be ascertained. Their size and the BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 149 number of ribs developed would suggest that they are earlier than degree 4, but more likely to be of degree 2 or 3 than 0 or 1. The present study also shows that variations in the number of ribs developed at a particular degree do occur. A striking difference between the Builth and Shelve meraspides is that of size, the Builth specimens being generally much smaller. Table 6 gives the comparative data for specimens from the two areas. As in the holaspid form, variations in the overall length/width ratios are known in the meraspides (Figs 83, 84; Whittard 1964 : pi. 45, figs 6, 7), although much of this variation results from post-mortem deformation. Table 6. Comparative data for meraspid specimens of Ogygiocarella debuchii (Brongniart) from the Builth and Shelve regions. Column headed 'Locality' gives an indication of which specimens are from the same locality as others. All measurements in mm. * indicates a further incipient rib developed. For explanation of symbols see Fig. 1. Builth Specimen Degree T.L. A Z Ribs Locality It.2897 7 _ _ 0-8 2+ ? A It.2896 7 - - 0-8 3 A It.2898 4 c. 4-0 c. 2-0 c. 1-1 7 A It.2892 7 - - 1-7 8 B It.2913 7 6-5 2-9 1-9 9 C It. 2893 7 8-2 3-3 2-2 7 + C Shelve GSM 102160 1 _ _ 1-3 6 1 GSM 102163 2 3-3 - 1-5 7* 2 GSM 102164 3 c. 4-0 - 1-6 7 + 2 GSM 102165 4 6-4 3-0 2-2 9 3 GSM 102166 5 5-7 2-6 1-8 10 4 GSM 102167 7 14-5 - 2-8 10 3 BIOMETRICAL DATA. Despite the common occurrence of the species in the Builth district, rela- tively few measurements are possible owing to the damaged state of much of the available material. Pygidia have provided most of the available data; insufficient cephalic and thoracic data are available to allow any formal analysis. Measurements are given in certain cases to give some quantitative idea of the dimensions and variation of some features. For ease of com- parison, various dimensions of O. debuchii and O. angustissima (p. 146) are plotted on Fig. 101. In the cases where bivariate analyses have been possible, they have been computed using the total measurable sample from the Builth inlier rather than the sample from any specific locality. Table 7. Ratios of the maximum cephalic length (A), thoracic length (Q), maximum pygidial width (W) and pygidial length (Z) to the total length (T.L.) for Ogygiocarella debuchii (Brongniart). Specimen T.L. A/T.L. Q/T.L. W/T.L. Z/T.L. NMW68.376.G210 26-4 0-33 0-31 0-67 0-36 NMW68.376.G211 33-5 0-33 0-32 0-67 0-35 NMW68.376.G212 c. 35-5 c. 0-30 c. 0-31 c. 0-71 c. 0-39 NMW68.376.G213 c. 36-0 c. 0-32 c. 0-32 c. 0-67 c. 0-36 NMW68.376.G214 38-4 0-33 0-29 c. 0-71 0-38 150 C. P. HUGHES Table 8. Cephalic measurements for Ogygiocarella debuchii (Brongniart). All measurements in mm. For explanation of symbols see Fig. 1. Specimen A B C3 J Ji K NMW68.376.G215 10-7 9-5 - c. 20-0 _ NMW68.376.G211 11-0 10-0 - - - 5-6 NMW68.376.G214 12-5 11-4 - 24-0 c. 6-0 NMW68.376.G216 12-6 c. 10-5 4-4 c. 21-0 5-5 NMW68.376.G217 17-5 16-2 - 29-0 16-8 9.4 NMW68.376.G218 c. 22-3 - - 44-0 _ NMW68.376.G219 23-0 21-3 - - - 12-4 NMW 68.376.G220 - c. 11.0 - - c. 6-0 Table 9. Bivariate statistics for the thorax and pygidium of Ogygiocarella debuchii (Brongniart). All measurements in mm. For explanation of symbols see Fig. 1. X y X var. x y var. y r re a var. a a var. a n Q Rz 15-08 41-29 7-21 10-68 0-95 0-95 1-06 0-0156 0-51 0-0036 9 R» R, 7-87 12-79 6-72 8-94 1-00 1-00 0-98 0-0005 0-84 0-0004 6 W Z 30-08 299-07 18-28 131-04 0-97 0-98 1-08 0-0025 0-66 0-0008 29 W X 25-92 253-52 4-61 9-33 0-99 0-99 1-07 0-0007 - - 24 Z Y 14-52 83-82 12-07 54-16 1-00 1-00 0-97 0-0001 - - 38 z X 14-55 96-65 4-35 8-47 0-97 0-97 0-99 0-0015 0-30 0-0001 39 Table 10. Thoracic measurements for Ogygiocarella debuchii (Brongniart). All measurements in mm. For explanation of symbols see Fig. 1. Specimen RI Q Specimen R! Q NMW 68 376.G216 c. 23-0 9-8 NMW 68, 376.G221 c. 33-0 14-0 NMW 68, .376.G211 c. 24-0 10-8 It.2903 c. 44-0 23-2 NMW 68, 376.G215 c. 24-5 10-9 NMW 68, 376.G217 c. 53-6 23-8 NMW 68, 376.G213 25-4 11-4 NMW 68. 376.G222 c. 55-0 23-2 NMW 68, 376.G214 31-2 11-2 NMW 68, 376.G223 c. 61-0 24-0 Table 11. Frequency distribution of ribs on the pleural fields of Ogygiocarella debuchii (Brongniart). The suffix 'i' indicates the development of a further incipient rib. See also Fig. 93, p. 144). Number of ribs 9i 10 lOi 11 Hi 12 Number of specimens 0 1 7 82 1 2 DISCUSSION. Under a variety of names, this is one of the most commonly cited species in British Ordovician palaeontological literature, being first figured by Lhuyd in 1698. The validity and authorship of the species has been the subject of much debate, the two most recent contributions being by Whittard (1964) and Hughes (1972). Ogygiocarella angustissima (Salter, 1 865-6) (Figs 87, 90, 94-96, 98-100, 102-108, 110) 1 846 Ogygia Buchii (Brongniart) ; Burmeister (pars) : 59-61 . 1865 Ogygia angustissima Salter : pi. 14, figs 8-9 (figures only). 1 866 Ogygia angustissima Salter : 1 29 (text). 1940 Ogygia buchi (Brongniart); Elles (pars) : 408, 411, 412, 414-419, 421- 433. 1964 Ogygiocarella debuchii (Brongniart) var. angustissima (Salter) Whittard : 261-262; pi. 45, fig. 9; pi. 26, fig. 1. Includes earlier synonymy. 1966 Ogygiocarella debuchii (Brongniart) var. angustissima (Salter); Whittard : 302. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 1 5 1 DIAGNOSIS. Ogygiocarella with 12-14 ribs developed on pleural field, majority of individuals having 13. LECTOTYPE (herein selected): 59198 (Fig. 94), internal mould of nearly complete individual. Dimensions (in mm): T.L., 33-5. A, 11-2. B, 10-0. C, 5-3. Clt 2-1. C2, 3-8. J, c. 19-8. Klf c. 4-3. RL c. 22-0. R2, 3-7. R3, 3-1. R4, c. 21-0. Q, 10-8. Qlf 1-2. W, 19-7. X, 2-9. Y, 11-5. Z, 12-0. For explanation of symbols see Fig. 1. Both pleural fields have thirteen ribs and at least nine axial rings developed. Paralectotypes : 59199a (Fig. 95), 59199b, internal moulds of nearly complete specimens, 59199a lacking librigenae. TYPE LOCALITY AND HORIZON. Specimen 59198, here selected as lectotype, is labelled as coming from 'Llandeilo Flags, Gilwern' and was recorded as such by Salter in his original description (1866 : 129). However, the beds exposed at and around Gilwern are of uppermost lower Llanvirn age, whereas this species is characteristic of the highest lower Llandeilo to upper Llandeilo rocks of the Builth area. The lithology of the lectotype slab is unlike any found at Gilwern, but is similar to that of the Nemagraptus gracilis Shales of middle-upper Llandeilo age, and it seems likely that the specimen is either from the middle quarry, Llanfawr, Llandrindod, or possibly from Gwern-yfed-fach quarry, near Builth Road. While it is accepted that the designa- tion of a lectotype from uncertain locality leaves much to be desired, a similar problem over locality exists for the other syntypes (59199a, b); they are listed as from Gwern-fydd, but their lithology strongly suggests that they are from Harper's quarry, a quarter mile (402 m) north-west of Wellfield. Specimen 59198 has, therefore, been selected as lectotype as it is the most perfect of the three syntypes. DISTRIBUTION. The species appears to be restricted to the high lower Llandeilo and middle to upper Llandeilo. Until the faunas of south Wales have been restudied little can be said about its occurrence south of Builth except that it has recently been recorded from the topmost Llandeilo Flags at Pant-yr-hendre and Llan quarries near Mydrim, Dyfed (Carmarthenshire) (Toghill 1970: 122). DESCRIPTION. Complete individuals oval in outline, about one and a half times as long as wide, but with some variation (see Table 12). The cephalon is identical to that of O. debuchii except that the palpebral lobe may be slightly more anteriorly placed in larger specimens of O. angustissima, but insufficient data are available at present to confirm this quantitatively. Eye appears to be similar to that of O. debuchii; one specimen (eye length 6-0 mm) has c. 2500 lenses (Fig. 105). Hypostoma apparently like that of O. debuchii. Pygidium, although basically similar to that of O. debuchii, differs in some features. Outline tends to be parabolic, slightly longer relative to its width than in O. debuchii, except in small specimens (width less than 27 mm). Infundibular axis occupies about one-seventh of pygidial width anteriorly. Between 12 and 14 ribs may be developed on the pleural field, with over three- quarters of the specimens having 13 ribs (see Table 15 and Fig. 93). At least 14 axial rings may be developed, but generally only the anterior seven or eight are clearly discernible (as in O. debuchii). A comparison of the respective bivariate statistics for pygidial length and anterior axial width reveals that the pygidial axis of O. angustissima is narrower, with respect to pygidial length, than in O. debuchii. As in other asaphids, irregularities in the rib development are rare, only a single example being known in O. angustissima (Fig. 1 10). In all other features the pygidium appears to be identical to that of O. debuchii, with the exception that some small holaspid speci- mens show a slight posterior indentation (Fig. 87). BIOMETRICAL DATA. Slightly more data are available for O. angustissima than for O. debuchii, and they are presented in a similar manner, some slight modifications being made to suit the extra data available. 152 C. P. HUGHES Table 12. Ratios of the maximum cephalic length (A), thoracic length (Q), maximum pygidial width (W) and pygidial length (Z) to the total length (T.L.) for Ogygiocarella angustissima (Salter). Specimen T.L. A/T.L. Q/T.L. W/T.L. Z/T.L. NMW68.376.G224 c. 29-0 c. 0-34 _ c. 0-62 c . 0-34 BM 59198 33-5 0-33 0-32 0-58 0-36 SM A44685 38-6 0-32 0-31 - 0-39 BM 1.8076 48-1 0-30 0-31 0-69 0-39 SM A44686 50-5 0-30 0-31 0-70 0-39 NMW68.376.G225 54-0 0-30 0-32 0-61 0-38 BM 1.59215 60-4 0-29 0-31 0-67 0-40 BM 1.4381 65-0 0-32 - 0-78 0-39 BMI.1330 70-5 0-33 0-30 c. 0-63 0-37 GSM 12918 71-5 0-29 0-31 0-68 0-41 SM A44696 71-8 c. 0-28 0-32 - 0-39 BM 59211 83-4 0-30 0-31" 0-69 0-39 BM 59202 88-0 0-31 0-30 0-65 0-40 OUMB123 96-8 0-29 0-29 0-65 0-40 SM A44714 101-9 0-30 0-29 0-64 0-40 Table 13. Cephalic measurements for Ogygiocarella angustissima (Salter). All measurements in mm. For explanation of symbols see Fig. 1. Specimen A B C3 J Jt K2 NMW68.376.G226 10-4 c. 9-5 c. 4-1 _ _ _ BM 59198 11-2 10-0 3-8 c. 19-8 - - SM A44685 12-4 - 4-6 - - _ NMW 68.376.G227 13-0 11-9 c. 5-4 - - - SM A44730 13-6 - - 27-0 12-3 - SM A44686 15-0 - - - 14-2 - NMW68.376.G228 17-3 - - c. 33-6 _ - BM 59215 17-4 - - 37-1 - - NMW 68.376.G229 20-7 - - c. 44-2 21-5 - BM 1.4381 20-7 - 7-9 44-2 - - NMW 68.376.G230 21-6 - 8-9 c. 41-0 18-4 - BMI.1330 23-3 21-7 - c. 39-5 - _ NMW68.376.G231 23-5 20-0 c. 6-5 46-0 - - BM 59211 25-2 23-0 c. 10-8 50-0 - 15-7 BM 59202 27-0 24-6 12-4 47-0 - c. 14-0 BM 59203 29-0 26-5 - 61-3 - - OUM B123 29-6 27-0 - c. 54-4 - _ BM 59211 30-0 27-0 - - - - SM A44714 30-3 26-8 12-4 c. 58-5 c. 28-5 16-8 SM A44723 34-0 30-5 14-1 68-6 29-0 19-2 DISCUSSION. Apart from the number of pygidial ribs, the differences between O. angustissima and O. debuchii are slight. If samples of each form in the Builth region are considered, there is a clear distinction based on the pygidial rib counts, although both species include morpholo- gically intermediate specimens. There is also a clear stratigraphical separation between the two species in the Builth region. In the Shelve Inlier the situation is not quite so straightforward, as the two forms have an overlapping stratigraphical range with both occurring in the upper- most lower Llandeilo. This suggests that the fossil record has preserved details of a stock under- going speciation, with concomitant blurring of sharp distinctions between the two species. BU1LTH-LLANDRINDOD ORDOVICIAN TRILOBITES 153 Table 14. Bivariate statistics for Ogygiocarella angustissima (Salter). All measurements in mm. For explanation of symbols see Fig. 1. X y X var. x y var. y r re a var. a a var. a n A B 25-10 51-48 22-64 41-89 1-00 1-00 1-00 0-0007 0-90 0-0006 11 Ri Q 49-34 435-01 21-29 74-70 0-98 0-98 0-96 0-0055 0-41 0-0011 8 Q R2 20-23 48-74 8-53 11-55 0-97 0-97 1-14 0-0044 0-49 0-0008 19 R* Ra 8-48 10-96 7-32 7-87 0-99 0-99 0-98 0-0009 0-85 0-0007 20 R4 Q 54-80 175-52 23-71 27-80 0-96 0-96 0-92 0-0065 0-40 0-0012 11 W Z 46-31 349-97 27-51 115-22 0-93 0-94 0-97 0-0039 0-57 0-0015 32 W X 47-18 351-83 7-88 10-26 0-89 0-90 1-02 0-0056 0-17 0-0002 34 z Y 26-89 111-61 22-36 70-82 1-00 1-00 0-96 0-0002 0-80 0-0001 42 z X 26-54 112-58 7-25 9-61 0-95 0-96 1-06 0-0018 0-29 0-0001 54 Table 15. Frequency distribution of ribs on the pleural fields of Ogygiocarella angustissima (Salter). The suffix T indicates the development of a further incipient rib. See also Fig. 93. Number of ribs 12 12i 13 13i 14 Number of specimens 9 3 77 2 5 Ogygiocaridinid gen. et sp. indet. (Fig. 12) FIGURED SPECIMEN. It. 2927, internal and external moulds of complete meraspid of degree 0. LOCALITY AND HORIZON. Left bank of stream section east of Bach-y-graig, 40 yd (37 m) upstream of the point where the footpath enters the wood at the western end of the section; in Glypto- graptus teretiusculus Shales. DESCRIPTION. Exoskeleton apparently complete, oval in outline, being 1-4 mm long and 1-05 mm wide. Cephalon semicircular with no genal spines. Glabella parallel-sided; axial furrow deep. Glabellar furrows consist of two notch-like indentations on either side of glabella situated at one-third and two-thirds the way along the glabella. Occipital ring bowed backwards sagittally, occipital furrow continuous but deepest laterally. Posterior border and border furrow well defined. Genal region convex, with no eye. Facial suture may, however, be submarginal with the librigena, and with it the eye, missing. Transitory pygidium semicircular and slightly smaller than cephalon, being 0-55 mm long. Axis well defined, occupying one-fifth of pygidial width anteriorly and tapering posteriorly to become parallel-sided over the posterior half. Two ribs occupy anterior half of the gently convex pleural fields. Number of axial rings developed cannot be determined. DISCUSSION. The taxonomic placing of isolated occurrences of very early growth stages is difficult. Among the genera known in association with this specimen is Barrandia, but although the early ontogenetic stages of this genus are not known, the general characters of both the cephalon and transitory pygidium of this specimen are more akin to the ogygiocaridinids, and it would seem most likely to belong to either Ogyginus corndensis (Murchison) or Ogygiocarella debuchii (Brongniart), both species being known from the same locality. The apparent lack of eyes is noteworthy, as relatively large, prominent eyes are present in degree 4 meraspides of both O. debuchii and O. corndensis. However, Evitt (1961) found that no visual surface is discernible in the protaspides of some asaphids, although its future position is indicated by the course of the facial suture. Asaphid gen. et sp. indet. (Fig. 112) FIGURED SPECIMEN. It. 2928, internal mould of glabella and part of left fixigena. Dimensions (in mm): B, 11-6. Cl5 c. 3-8. C2, 3-5. K2, 8-0. For explanation of symbols see Fig. 1. 154 C. P. HUGHES LOCALITY AND HORIZON. Gwern-yfed-fach quarry, \ ml (805 m) south-east of Builth Road Station; Nemagraptus gracilis Shales. DESCRIPTION. Glabella well defined, constricted at about one-fifth of its length, expanding slightly to just anterior of the palpebral lobe, obtusely rounded anteriorly. Three pairs of lateral glabellar furrows present. Posterior pair situated opposite the narrowest part of glabella and develops pits adaxially. Median pair moderately deep, curved posteromedially, situated at about half-way along glabella, and distally fail to reach axial furrow. Anterior pair situated opposite anterior of palpebral lobe, short, shallow and directed slightly anteromedially. Occipital furrow continuous, though central portion shallower and bowed posteriorly. Very little of fixigena preserved, but apparently postericr portion extends for some way laterally posterior to eye. Both posterior border and border furrow moderately well developed. DISCUSSION. The affinities of this specimen are unknown, although it bears some similarity to Nobiliasaphus powysensis sp. nov. (see p. 117). However, the glabella furrowing, occipital ring and position of the eye serve to distinguish it from that species. Specimen 1.1331, labelled as 'Ogygia . . . from Gwernyfod', is probably closely related to this form. Family NILEIDAE Angelin, 1854 Genus BARRANDIA M'Coy, 1849 DIAGNOSIS. Exoskeleton oval, depressed and subisopygous. Glabella smooth, expanded frontally; axial furrow shallow; fixigena narrow (tr.) posteriorly. Eye well developed; genal spine short. Thorax of eight segments; doublure extending beneath more than half pleural width. Pygidium subsemicircular to subelliptical with prominent, weakly furrowed, axis and generally smooth pleural field. TYPE SPECIES. Barrandia cordai M'Coy, 1849. DISTRIBUTION. The genus first occurs in the Lower Llanvirn of south Wales, Salop (Shropshire) and Pont Seiont, north Wales. In the Builth area it is restricted to the lower Llandeilo, a horizon at which it is unknown elsewhere. The record of B. homfrayi Hicks from the Lower Llanvirn at Builth (Elles 1940 : 395) is considered to be based on a misidentification. The genus has been recorded also from the mid-Ordovician of North America (Hintze 1953; Webb 1956). How- ever, it is thought likely that the specimen referred to by Hintze belongs to Aponileus (see Hu 1963) and that those mentioned by Webb also belong to that or a related genus. DISCUSSION. The assignment of Barrandia to the Nileidae rather than the Asaphidae has been subject to some debate (Reed 1931; Prantl & Pfibyl 1949; Jaanusson in Moore 1959). New Figs 111, 114-119. Barrandia cordai M'Coy, p. 156. Fig. Ill, Llandeilo, Builth, exact loc. unknown (see p. 156). Holotype. Internal mould of damaged specimen, SM A15626, x2-5. Figs 114-119, Lower Llandeilo, left bank of Dulas Brook, 150yd (137 m) SW of the old quarry 350 yd (320 m) west of Maesgwynne, SO 058564. Figs 114, 117, internal and external moulds of pygidium showing furrow separating parathoracic segment, It. 2933, x5. Fig. 115, internal mould of pygidium, It. 2929, x6. Fig. 116, internal mould of cranidium, It.2930, x6. Fig. 118, internal mould of cranidium showing axial furrows fading out anteriorly, It.2931, x6. Fig. 119, internal mould of pygidium with para thoracic segment, It.2932, x 6. Fig. 112. Asaphid gen. et sp. indet., p. 153. Middle-upper Llandeilo, quarry at Gwern-yfed-fach, ^ ml (805 m) SE of Builth Road station, SO 030538. Internal mould of glabella and part of left fixigena, It.2928, x3. Figs 113, 120-124. Barrandia cf. cordai M'Coy, p. 158. Figs 113, 120-122, 124, Lower Llandeilo, stream section 15yd (14m) SW of the old quarry 350yd (320m) west of Maesgwynne, SO 059566. Figs 113, 124, internal mould of entire specimen, It. 2934. Fig. 113, enlargement of eye, x 10; Fig. 124, x4. Fig. 120, external mould of transitory pygidium, showing weakly-developed ribs on the pleural fields, It. 2939, x8. Figs 121, 122, internal and external moulds of transitory pygidium, showing traces of two ribs, It.2938, x6. Fig. 123, Lower Llandeilo, old quarry 350yd (320m) west of Maesgwynne, SO 059566. Internal mould showing cephalic doublure, It.2936, x 3. BUILTH-LLANDRINDOD ORDOVICIAN TR1LOBITES 155 123 124 156 C. P. HUGHES information regarding the nileids has recently become available (Fortey 1975) and it is believed that the general form of the exoskeleton, in particular that of the hypostoma (despite the presence of a furrow passing across the median body), cephalic doublure and eye, suggests that the assignment to the Nileidae is correct. The eye-like 'nodular areas' described by Whittard (19610 : 221) are clearly seen in Barrandia and Homalopteon from the Builth area (Figs 113, 149), and have the concavo-convex outline characteristic of many trilobite compound eyes. Eye lenses in nileids are minute and commonly can only be seen with difficulty in well-preserved material (Fortey 1975). Despite the lack of visible lens structures in Barrandia and Homalopteon it is difficult to imagine these 'nodular areas' as anything but eyes. The distinction between this genus and Homalopteon Salter is considered under the latter genus (p. 164). Owing to the common crushing and distortion of specimens, and also damage to, or complete loss of, the dorsal surface over the doublure, specific determinations are generally difficult. This applies particularly to Barrandia with its very wide doublure, although owing to the thin nature of the exoskeleton details of axial rings and pleural ribs are commonly distin- guishable on both internal and external moulds; data from both have been included in the following specific descriptions. Systematic difficulties, especially assessments of variation, are further increased by the rareness of most species. Out of a total of fewer than 30 specimens from the Stapeley Shales, Whittard recognized five species based essentially on differences in the furrowing of the pygidial axis and pleural region. Some specific separations have been made in the present study on similar criteria, but it is possible that in both cases the differences may prove to be of an intraspecific nature. Barrandia cordai M'Coy, 1849 (Figs 111, 114-119, 134) 1849 Barrandia cordai M'Coy; 409-410, text-fig. 1931 Barrandia cordai M'Coy; Reed : 443, 467 (gives earlier synonymy). 1940 Barrandia cordai M'Coy; Elles (pars) : 412. 1949 Barrandia cordai M'Coy; Prantl & Pfibyl : 2, 4, 5, 14. 196k/ Barrandia cordai M'Coy; Whittard : 221-223, 226-227; pi. 33, fig. 8. DIAGNOSIS. Barrandia with subelliptical pygidium, having narrow conical axis with first two axial rings distinct; third and fourth indistinct; pleural field smooth apart from well-developed border furrow. HOLOTYPE. SM A 15626 (Fig. Ill), internal mould of damaged entire individual. Dimensions (in mm): T.L., c. 25-0. A, c. 8-8. I, c. 17-0. J,, c. 9-0. Q, c. 8-5. Ql5 1-4. R:, 15-5. R2, 5-1. R4, 13-2. R5, 2-3. R6, 2-0. W, 12-6. X, 3-2. Z, 7-7. Dimensions of hypostoma It. 2948 (in mm): H1? 8-2. H4, 2-9. H5, 7-7. H6, c. 8-3. H7, 7-0. H8, 5-8. For explanation of symbols see Fig. 1. TYPE LOCALITY AND HORIZON. The locality for the holotype was given by M'Coy (1849 : 410) as 'Black Wenlock shale of Builth'. Salter (1866: 143) quotes it as from Tenkerrig, near Builth, Radnorshire'. Elles (1940:412,421) recorded the species from various localities which she believed to be of Llandeilo and basal Caradoc age; Whittard (196 la : 227), however, considered the species to be most likely of basal Caradoc age (basal N. gracilis Zone). Addison (in W'lliams et al. 1972 : 35-36) has shown that the basal portion of the N. gracilis Zone is of middle and upper Llandeilo age. The holotype is preserved in a dark greyish-blue shale which compares with both lower and middle to upper Llandeilo rocks of the Builth sequence. The present study indicates that Homalopteon radians (M'Coy) is the only nileid occurring at Pen-cerrig and that forms attributable to B. cordai are known only from the exposures in the left bank of the Dulas Brook, 150 yd (137 m) south-west of the old quarry 350 yd (320 m) west of Maesgwynne, and from the old quarry in the right bank of the stream 570yd (521 m) north of Wye Cottage. Both these exposures are in Glyptograptus teretiusculus Shales of lower Llandeilo age. DESCRIPTION. The holotype is the only complete specimen known. Its length is about one and a half times its maximum width. Cephalon roughly semicircular, about one-third of the total length of the specimen. Glabella BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 157 characteristically devoid of furrows, the two longitudinal furrows on the holotype being due to distortion, which, owing to the thin exoskeleton, is relatively common and may show a degree of symmetry in some specimens. Posteriorly the glabella occupies slightly over three-fifths of the cranidial width, anteriorly expanding in width to form the whole of the anterior part of the cranidium. Axial furrows shallow and subparallel posteriorly, becoming divergent and shal- lower anteriorly and fading out altogether before reaching lateral margins (Fig. 118). Fixigena a narrow flat band alongside the posterior two-thirds of the glabella, merging anteriorly with the expanded frontal region of the glabella. No discrete palpebral lobe developed and relatively little posterior expansion of the fixigena. Facial suture gently sigmoidal, being curved outwards in its anterior and posterior portion, becoming marginal anteriorly. Librigena subtriangular with a short genal spine. No median suture developed, and the librigenae form a single unit with the ventral doublure. The eye is not certainly known. The so-called eye previously described is thought to be an artefact resulting from imperfect preservation. Cephalic doublure poorly known, but anteriorly appears to have projections on the inner margin presumably surrounding the anterior wings of the hypostoma. Doublure bears a series of parallel terrace lines, at least 12 being present anteromedially in the holotype. Hypostoma subrectangular with small anterior wing extending for about one-third of the hypostomal length. Lateral margins subparallel though slightly convex outwards. Posterior wing small, short (sag.); posterior margin protruded into a broad mesial tongue-like extension. Anterior lobe subcircular, much larger than posterior lobe, and bounded by a well-developed middle and lateral furrow; posterior lobe crescentic. Posterior border furrow well developed mesially, though shallower than middle furrow with which it merges anterolaterally. Maculae absent. Surface bears fine raised lines generally conforming in shape to that of the middle furrow; on the posterior border, however, they become modified to be subparallel to the posterior margin. Eight thoracic segments are present as in other Barrandia spp., not seven as shown in Moore (1959 : O357, fig. 267.9). Axial rings simple, bowed gently posteriorly and with a well-developed articulating furrow. Anteriorly the axis occupies about two-fifths of thoracic width and tapers slightly to the rear. Axial furrow well incised and slightly zetoidal. Pleurae falcate, each having a strong diagonal furrow commencing at the inner anterior corner and fading out about half- way across the pleural region. Wide doublure underlies nearly two-thirds of pleural region and bears numerous outwardly concave terrace lines. Pygidium subelliptical with anterior margin convex forwards. Axis relatively narrow and cone- shaped, occupying slightly under one-quarter of the anterior width and extending for about two-thirds of the pygidial length. In the holotype the axis is poorly preserved, but nevertheless shows a faint indication of one axial ring. It. 2932 (Fig. 119) shows two distinct, and possibly up to four faintly-defined, axial rings and a terminal piece. This specimen, however, is small, and the furrowing may become lost in the larger individuals. Anterior border well developed, becoming shallower distally and failing to reach the lateral margins. Remainder of pleural field thought to be smooth, although owing to the common loss of the dorsal exoskeleton in this region it is difficult to be certain. Wide doublure underlies almost entire pleural region and Table 16. Measurements for Barrandia cordai M'Coy. All measurements in mm. For explanation of symbols see Fig. 1 . Specimen A J Ki W X Y Z It.2929 _ _ _ 8-0 1-9 2-5 4-3 It.2930 4-3 c. 4-7 2-8 - - - - It. 2931 5-6 5-7 3-6 - - - - It.2932 - - - 7-7 2-0 2-1 3-7 It. 2933 - - - 7-9 2-2 2-1 c.4-2 NMW68.376.G232 c. 5-7 - 3-5 - - - - NMW68.376.G233 6-5 c. 7-0 4-3 - - - - 158 C. P. HUGHES bears fine terrace lines subparallel to the margins. Posteromesially about 15 terrace lines are present. ONTOGENY. Two small pygidia similar to those of holaspides are known in which a 'thoracic' segment appears to be fused along the anterior margin (Figs 1 14, 117, 1 19). It. 2933 (Fig. 1 17) does not show any furrow between the 'thoracic' segment and the rest of the pygidium on the external surface, but a clear furrow is impressed on the internal mould (Fig. 1 14), indicating a slight thickening of the exoskeleton along the line of junction. It is thought that both these specimens represent late meraspides with a parathoracic segment developed prior to release into the thorax. Fortey (1975 : 63) has shown that the release of the final segment in Nileidae may occur at a comparatively late stage in ontogeny. A meraspis of B. ultima sp. nov. (Fig. 130) shows two segments developed in the anterior part of a relatively large transitory pygidium. Unfortunately nothing more is known of the ontogenies of Barrandia species. DISCUSSION. Previous descriptions of this species have been based on the holotype, which, although a relatively complete specimen, has certain crucial regions either missing or poorly preserved. Elles believed the species to be fairly widespread in all post-Llanvirn beds of the Builth region but did not give any redescription. The present study has shown that the species is rare in the Builth inlier. B. cordai is distinguished from B. cf. cordai by its narrower pygidial axis and from B. expansa sp. nov. (p. 159) by the different outline of the pygidial axis. B. ultima sp. nov. (p. 160) has a better-developed cephalic axial furrow, relatively wider (tr.) fixigena and a greater number of axial rings on the pygidium. B. homfrayi Hicks, 1875 differs in having a relatively wide furrowed pygidial axis and a relatively wide cranidium. B. tasgarensis Whittard, 196 la, which is very similar to though narrower than B. homfrayi, is distinguished by the lack of furrows on the pygidial axis, while B. bianularis Whittard, 196 la, is readily distinguished by its relatively wide, stumpy pygidial axis with two clearly-defined axial rings. B. parabolica Whittard, 196 la is in many ways similar to B. cordai, but the marked parabolic outline of the cephalon and pygidium appears to distinguish Whittard's species. The hypostoma is similar to that described by Whit- tard (196 la : 223-224) in B. homfrayi but the anterior wing is more anteriorly placed and the posterior lobe narrower (tr.) in B. cordai. Barrandia cf. cordai M'Coy (Figs. 113, 120-126) DIMENSIONS (in mm): It. 2934: T.L., c. 23-7. A, c. 7-5. I, c. 18-0. K]5 c. 5-6. Q, c. 8-8. Ql5 1-3. R19 c. 16-0. R2, 5-3. W X Y Z It.2934 13-0 3-6 4-0 7-4 It.2935 10-5 3-0 3-2 5-5 It.2937 6-2 1-8 2-0 3-2 It.2938 6-7 1-9 2-2 It.2939 c. 6-0 1-5 c. 2-0 3-3 For explanation of symbols see Fig. 1. LOCALITIES AND HORIZON. It. 2935-6 are from the old quarry 350 yd (320 m) west of Maesgwynne, and It.2934, 2937-9 are from the stream section 15 yd (14 m) south-west of this quarry; both are in Glyptograptus teretiusculus Shales of lower Llandeilo age. DESCRIPTION. It. 2934-5 are very like B. cordai, except that the pygidial axis is somewhat broader and has a slightly expanded anterior portion similar to that of B. expansa sp. nov. It.2934 also shows one eye moderately well preserved, situated at about the mid-length of the cephalon, with a clear concavo-convex outline (Figs 113, 124). No lens structures are preserved. It. 2936 is the only other specimen of Barrandia known from the old quarry west of Maes- gwynne and is included here. It shows the mesial portion of the cephalic doublure with about BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 1 59 16 subparallel terrace lines anteriorly, and the projection on the inner margin that abuts against the anterior wing of the hypostoma (Fig. 123). ONTOGENY. Three smaller Barrandia pygidia from the stream section appear to have two well- developed axial rings and two pleural ribs. The ontogeny of Barrandia species suggests that these probably represent meraspides, rather than a small species characterized by a ribbed pleural field. Although an impression of the dorsal exoskeleton is absent in the pleural regions in both internal moulds and not well preserved in the external mould, traces of the two ribs are present axially (Figs 120-122), which, together with the two anterior axial rings, may make up para- thoracic segments. The lateral extent of these ribs is not known; traces on the doublure in It. 2937 suggest they continue to near the lateral margin, whereas in It. 2939 they appear to die out rapidly with only the deeper anterior border furrow encroaching onto the wide border region. Wide doublure underlies most of pleural region, and bears terrace lines subparallel to pygidial margins. DISCUSSION. The slightly wider axis in this form may be an expression of intraspecific variation present in B. cordai, but until more material is available it is best considered as a form closely related to B. cordai. If the small pygidia with two ribs are correctly assigned to the meraspid period they support the other evidence suggesting that the late retention of parathoracic segments is a characteristic of all Barrandia species. Barrandia expansa sp. nov. (Figs 127-129) DIAGNOSIS. Barrandia with marked expansion of anterior portion of pygidial axis; pleural field with well-developed anterior border furrow; axis with two axial rings and large terminal piece. NAME. 'Spread out'. TYPE MATERIAL. Holotype It. 2940 (Fig. 128), internal mould of pygidium. Paratypes, It. 2941 (Fig. 127), external mould of thorax; It. 2942 (Fig. 129), internal mould of pygidium. DIMENSIONS (in mm): Qx Rx R2 R4 W X Y Z Holotype c. 14-5 4-3 c. 4-0 7-3 It.2941 1-2 11-0 4-1 10-5 It.2942 c. 14-0 c. 4-0 c. 4-6 For explanation of symbols see Fig. 1. LOCALITIES AND HORIZON. The holotype and It.2941 are from the left bank of the stream section east of Bach-y-graig, 45 yd (41 m) upstream of the point where the footpath enters the wood at the western end of the section. It.2942 is from the left bank of the stream section, 120yd (109 m) south-east of Tre Coed. Both are in Glyptograptus teretiusculus Shales of lower Llandeilo age. The species is known with certainty only at these two localities. DESCRIPTION. Cephalon unknown. Thoracic axial rings simple, with terrace lines on external surface. Pleurae falcate, with pleural furrow extending obliquely from the anterolateral corner towards the posterolateral corner, dying out just over one-third of the distance across the pleura at a point thought to coincide with the inner margin of the doublure. Dorsal surface bearing a series of fine, gently sigmoidal, raised lines in an anterolateral-posteromedian direction. Pygidial axis slightly over half pygidial length, anteriorly occupying about three-tenths of greatest pygidial width. Axis funnel-shaped, with axial furrows converging posteriorly in the anterior half, becoming parallel-sided in posterior portion. Two well-developed axial rings present in the anterior tapering portion; posterior portion without furrows. Axis bears fine, transversely- directed, raised lines. Pleural field with well-developed anterior border furrow. Wide doublure underlies most of pleural field, and has terrace lines subparallel to pygidial margins. DISCUSSION. Only the three figured specimens are known. Of other Barrandia, B. expansa appears to be most like B. bianularis Whittard, 196 la, but it is readily distinguished from this by its infundibular axis. 160 C. P. HUGHES Barrandia ultima sp. nov. (Figs 130-133, 135-136, 140) 1940 Barrandia (Homalopteon) radians (M'Coy); Elles (pars) : 412. DIAGNOSIS. Barrandia having relatively well developed cephalic axial furrow reaching lateral cranidial margin. Pygidial axis with probably four axial rings and terminal piece; pleural field smooth apart from anterior border furrow. NAME. 'Last'. TYPE MATERIAL. Holotype, It. 2945 (Fig. 136), internal mould of entire, but damaged, specimen. Paratypes, It. 2943 (Fig. 135), internal mould of pygidium with parathoracic segment. It. 2944 (Fig. 131), internal mould of cranidium. BU 455 (ex Chamberlain Collection 147/14) (Fig. 140), internal mould of damaged entire specimen. DIMENSIONS (in mm). Holotype: T.L., c. 15-4. A, c. 5-8. I, c. 13-4. KL, 3-9. Q, 5-2. Qx, 0-8. R]5 c. 12-0. R2, c. 3-5. W X Y Z Holotype c. 9-0 2-4 2-5 4-4 It.2943 8-3 1-7 2-0 3-4 BU455 9-7 2-5 3-0 5-0 For explanation of symbols see Fig. 1. LOCALITY AND HORIZON. All type specimens are from the stream section 400 yd (365 m) north- east of Gorse, north-east of Llandrindod; Glyptograptus teretiusculus Shales of Lower Llandeilo age. Restricted to the type locality and to beds a little higher in the succession exposed about 20 yd (18 m) downstream from it. DESCRIPTION. Similar to B. cordai and a comparative description only is required. Complete individuals tend to be slightly wider than B. cordai having a length : width ratio of about 12:9 against 12 : 8 for B. cordai, but how flattening may have altered the original proportions is not known. Cephalon differs from B. cordai in that the axial furrow is more strongly developed anteriorly and extends to the lateral cranidial margin, to cut it immediately anterior of 5, and the fixigena is slightly wider (tr.). Traces of the hypostoma are seen in It. 2945 (Fig. 136), and it appears to conform to the general Barrandia pattern with large subcircular anterior lobe and a short (sag.) crescentic posterior lobe. Posterior margin protruded into a median lip and surface of median body bearing a series of subconcentric raised lines. Figs 125-126. Barrandia cf. cordai M'Coy, p. 158. Lower Llandeilo, old quarry 350yd (320 m) west of Maesgwynne, SO 059566. Internal and external moulds of pygidium showing wide, slightly forwardly expanding axis and transverse sculpture on dorsal surface, It. 2935, x 6. Figs 127-129. Barrandia expansa sp. nov., p. 159. Figs 127-128. Lower Llandeilo, left bank of stream section east of Bach-y-graig, 45 yd (41 m) upstream of the point where the footpath enters the wood at the western end of the section, SO 071610. Fig. 127, Paratype. External mould of thorax, It. 2941, x5. Fig. 128, Holotype. Internal mould of pygidium showing frontal expansion of axis, It. 2940, x4. Fig. 129, Lower Llandeilo, 120yd (109 m) SE of Tre Coed on left bank of stream, SO 054552. Paratype. Internal mould of pygidium, It. 2942, x3-5. Figs 130-133, 135. Barrandia ultima sp. nov., above. Lower Llandeilo, stream section 400 yd (365 m) NE of Gorse, SO 072617. Fig. 130, internal mould of meraspid ? of degree 6, BU 456 (ex Chamber- lain Coll. 147/14), x6. Fig. 131, Paratype. Internal mould of cranidium, showing axial furrows extend- ing to lateral margins and relatively wide fixigena, It. 2944, >:6. Fig. 132, internal mould of pygidium with parathoracic segment and furrowed axis, It. 2947, x6. Fig. 133, internal mould of meraspid of degree 5 or 6, It.2946, x6. Fig. 135, Paratype. Internal mould of pygidium with parathoracic seg- ment, It.2943, x6. Fig. 134. Barrandia cordai M'Coy, p. 156. Lower Llandeilo, stream section 15 yd (14 m) SW of the old quarry 350yd (320m) west of Maesgwynne, SO 059566. Internal mould of hypostoma, It. 2948, x4. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 161 135 162 C. P. HUGHES Thorax appears to be identical to B. cordai. Pygidium differs from B. cordai in that four axial rings and a short terminal piece are almost certainly present (Fig. 135) and the doublure is narrower. Paratype pygidium It. 2943, with its parathoracic segment, is only a little smaller than other pygidia with no such segment and is another example of late retention of the eighth thoracic segment in the pygidium (p. 158). ONTOGENY. Two late meraspides are known, but determination of their precise degree is not possible. BU 456 is most likely a degree 6 and It. 2946 may be degree 5 or 6. Cephalon, as far as can be determined, very similar to that of holaspides. BU 456 shows the eye but no trace of lens structures. Genal spine (Fig. 130) possibly relatively larger than in holaspides but this may be due to distortion. BU 456 also shows parts of six thoracic segments, which appear to be freely articulating, and two parathoracic segments. It. 2946 has the anterior region very poorly preserved and it is more difficult to assess where the front of the transitory pygidium lies. It is believed that there are also two parathoracic segments present. In outline it is similar to the holaspis pygidium, but it is relatively longer with a larger axis. DISCUSSION. This species is distinguished from other Barrandia species by the relatively narrow pygidial doublure. The species is the youngest known Barrandia, from beds near the top of the Glyptograptus teretiusculus Shales above the level at which Ogyginus corndensis dies out. Other species previously recorded from approximately this level or higher (Elles 1940) are here placed in Homalopteon. Genus HOMALOPTEON Salter, 1866 DIAGNOSIS. Exoskeleton oval, depressed and subisopygous. Glabella expanded frontally; four pairs of very shallow lateral glabellar furrows; axial furrow shallow particularly anteriorly; fixigena narrow expanding posteriorly. Eye well developed; librigenae fused; genal spine short. Eight thoracic segments, doublure extending under slightly less than half pleural width. Pygidium subsemicircular; axial rings prominent; up to three ribs on gently convex pleural field. TYPE SPECIES. Ogygia portlockii Salter, 1849. DISTRIBUTION. The type species is from the Raheen Shales (Caradoc) of Co. Waterford, Ireland. In central Wales the genus is present in the uppermost lower, and middle to upper Llandeilo; it was also recorded by Salter (1867 : 179) from Abereiddy Bay, probably from beds of lower Llandeilo age. DISCUSSION. The genus was separated at subgeneric level from Barrandia by Salter (1 866 : 137-1 38) on characters which are summarized below. He placed both in the Asaphidae, but Reed (1931 : 468), having given full generic status to Homalopteon, assigned them to the Nileidae on the basis of the absence of the median suture. In Moore (1959 : O352) Jaanusson tentatively placed Homalopteon back in the Asaphidae. Homalopteon is certainly very closely allied to Barrandia and is thus a nileid. Figs 136, 140. Barrandia ultima sp. nov., p. 160. Lower Llandeilo, stream section 400 yd (365 m) NE of Gorse, SO 072617. Fig. 136, Holotype. Internal mould showing traces of hypostoma and cephalic doublure, It. 2945, x4. Fig. 140, Paratype. Internal mould of nearly complete specimen, BU 455 (ex Chamberlain Coll. 147/14), x4. Figs 137-139, 141-144. Homalopteon radians (M'Coy), p. 164. Figs 137-138, 141-142, Lower Llandeilo, small quarry at SW end of Pen-cerig Lake, SO 043541. Fig. 137, internal mould of cranidium showing traces of glabellar furrows, It. 2960, x5. Fig. 138, internal mould showing cephalic axial furrows, It. 2950, x5. Fig. 141, external mould of thorax and pygidium showing surface sculpture, It. 2949, x2. Fig. 142, internal mould of larger pygidium showing four axial rings and a terminal piece, It. 2961 , x 2. Figs 143-144, Lower Llandeilo, 'Pen-cerig, Builth'; probably from small quarry at SW end of Pen-cerig Lake, SO 043541. Fig. 143, Lectotype. Internal mould of pygidium, SM A16691, x4. Fig. 144, Para- lectotype. External mould of pygidium, SM A16692, x4. Fig. 139, Lower Llandeilo, Builth, exact loc. unknown. Internal mould of complete specimen, NMW 98.433, x2. BUILTH-LLANDRINDOD ORDOVICIAN TRILOBITES 163 144 164 C. P. HUGHES In retaining Homalopteon as a separate genus, some modification to Sailer's original definition is required (Table 17). Table 17. Summary of important differences between Barrandia and Homalopteon, as envisaged by Salter (1866) and the present author. „ Barrandia Homalopteon Barrandia Homalopteon Glabella Unfurrowed Furrowed Unfurrowed Four furrows Axial furrow Incomplete Complete — — Fixigena — — Narrow posteriorly; no posterior border Expanded posteriorly; narrow posterior border Eye Subcentral Anterior — — Pleura Fulcrum near axis Fulcrum remote — — Doublure — — >i pleural width 0) CO o u -o c * « w £ -C c u. U o terna t •ne st on 0) -D O X T3 C O CO -Q -D C 0) U> co *• CO *• T3 3'-- c o u 3 _, <1^ 3 5 _, O 0> E 3 ^ - CO c O c c o c o C o c o o o o 0) t) O) c • — c o c mm o o c o _o .c 0 3 c 0 M - CO I U 3 • o k/N- x J ^ 0) C x 0) C X — o A — -° c *~ 0 * Q. 4, O i- w 4, C O a, c o *- •— t^ \ 41 ""/L -Q -Q a -° ° ° *. """ 0) to o D Ixft7 -Q £ E -§ £ 0 «, 0 0) "> O -^ 0- c "•> *Vi E < 0 j E 4> E -D -o e U1 O •— 4) * *• O 4) g i_ •- C ,_ ^ Q. c £ s O O O T. 0 U 0 1 4 D 01 J 0 J 44033 u o 4 D y U 0 1 4 D W J 0 -j a 1 1 ! A 6 u o i X a u a q 3 < -1 "- I CO U. Ov Cf— o , o o CO » »> -° .; ± o u o V4-OV o o> o D) ,/, U » i •> " « O *" *- ^ u c °', * to •" 41 C • 3 4, 0 0> D ~° C "D 0) V o ^_ "D — o UCK c — • O — c 05 a* * CO O) 00 E — O oa 4) * -o ~° CCN c o 0 -= c ca o T3 1— * I ^ * -1 z ,_ c 0 0 OQ'" 00 u u 5 jaddn a|PP! W J 3 MOT SOVId 311IAONO1 D o « T3 O O) "" O 0) 9 -C Ol 01 O ^- 4> (A 0 0 41 U *> 0 D * * "3 * f c *- • 1 * >o a = -D *- u "D * = - m D "u "* "" «) Q. * " *cx Q. C 41 U 4) < CD o c ° 0) u. 1 •" E 3 ° C Q.1- "~ U ^ 1/5 CO 5 ! < _• U co dnoj£) u o 4 D y d n o a Q a 1 1 i A 6 u o -| •do 1 1 D M 4 D H 3 0.0 1 => X ^ •-"3 o c o *~ => o -*= C CO 0 = ° o -o U u 0) < to oo X 1 0) O) 0) C £ — O) s — = 1 *• ^l 0 U CO 0) M x , ,_• X Qj CD __ Q» *" , -^ o^ -1 o c — C O _* o ~ ° « -o o- o t- ° CO U ~* u- I 0) u, 00 0) c *- OJ .ij - c -o M 'E -5 ,,-D ™ -o o> o> I-D-C ^ «J^ 2! 5 c CO ^-Q^ o> CO " \ X 01 ^ "O 0) — O n~I II I! 4) /8UO|SpUD$ DOpDJDQj S >j D O y D|Dg pUD O|l8pUD|-| c o -ceo D . «i -D 35 X C ^ 4) o> -^ *-' c E = 0 c--"^ "D !c 0 0 -C 0 = •--•^^o)^: ^ o "« :E o CO CO 3NOisaNvs ooavavo 188 J. M. HURST He stated that the Onny Valley presented the best section and as can be seen from Fig. 2 an attempted subdivision of the strata was made. However this unit was far more comprehensive than Murchison probably intended, as subsequent researchers showed it to contain rocks Pre-Cambrian to Silurian in age. Later Ramsay & Aveline (1848) and Forbes (1848) still retained the Upper Llandovery rocks of the Longmynd area in the Caradoc. Not until the work of Sedgwick (1852) was there realisation of the inherent mixture in the Caradoc. He noted the trilo- bites Ampyx and Trinudeus near Cheney Longville and differentiated the 'Horderley' and 'May Hill' (now Upper Llandovery) Caradoc faunas, suggesting that the latter were younger. Ramsay (1853) revised his earlier interpretation (Ramsay & Aveline 1848), connecting the Caradoc Sandstones of the Longmynd with the Wenlock Shales. This small paper heralded the major changes of the following year, in which Salter & Aveline (1854) published their classic results. They divided the revised Caradoc Sandstone into five parts, from youngest to oldest: 1. Trinudeus Shales 2. Thin-bedded flags of greenish sandstone 3. Thick-bedded flags of Horderley 4. Hoar Edge Grits 5. Shales of Harnage and Shineton. The first three units are the ones dealt with here (Fig. 2). Callaway (1877) differentiated the Caradoc Harnage Shales from the Tremadoc Shineton Shales, placing the former above the Hoar Edge Grits. He also proposed the name Chatwall Sandstone to replace the 'thick-bedded flags of Horderley' of Salter & Aveline. In 1884 La Touche introduced the names Onny Shales, Cheney Longville Flags and Horderley Sandstone corres- ponding respectively to units 1 to 3 of Salter & Aveline (Fig. 2). From this it is apparent that the Chatwall Sandstone and Horderley Sandstone are synonymous. Lapworth & Watts (1894) subsequently erected Acton Scott Beds for strata contained between the Onny Shales above and the Cheney Longville Flags below (Fig. 2). Cobbold (1901) recognized the Alternata Limestone for the first time but placed it in with the Chatwall or Soudley Sandstone. With the coining of the term Soudley Sandstone, there were three names for a single unit, i.e. Chatwall, Horderley and Soudley Sandstones. Lapworth (1916) proposed the name Caradoc Series and divided the strata into a number of 'Groups' (Fig. 2). These 'Groups' were further subdivided, but no information regarding litholo- gies, faunas or type localities was offered. Nevertheless, Lapworth was the first person to recognize the heterogeneous nature of the Cheney Longville Flags by subdividing the unit into three. Bancroft subdivided the type Caradoc Series into units characterized by the faunas, mainly of brachiopods and trilobites (Fig. 2). He first named the Soudleyan, Longvillian and Marshbrook- ian Stages in the predominantly sandstone units in the middle of the Series (1929a). Later the same year (1929&) he listed six stages for the whole Caradoc, in ascending order Girvanian, Harnagian, Soudleyan, Longvillian, Marshbrookian and Actonian (Fig. 1). At this point the youngest stage, the Onnian, was not introduced by name, although the subdivision was incorporated in his table. Bancroft also privately published a short paper in 1933, showing the vertical extent and lithological and faunal composition of the Stages Costonian to Onnian. The Costonian was not defined and was simply introduced as a replacement for the Girvanian (Fig. 1). A paper published posthumously (Bancroft 1945) added more information regarding the Stages, but unfortunately they still remained ill-defined in terms of vertical extent. Further, many of the subdivisions were defined before the species of brachiopods and trilobites upon which they depended had been formally erected. These inconsistencies have marred Bancroft's work and made it difficult to follow. Nevertheless, as Dean (1958) indicates, Bancroft revolutionized stratigraphical research in the Ordovician. Not only did he collect faunas in bulk from single horizons, but he also tried to introduce statistical methods into his palaeontological work (Bancroft 1928#), and his methods far surpass many of today. Dean (1958, 19600, 1964) reviewed and revised the Stages and Zones of Bancroft based on a thorough redescription of the trilobite faunas (Dean 1960-63) and also published extensive faunal lists, but did not formally revise the lithostratigraphy (Fig. 2). He suggested that the name Acton Scott Beds sensu stricto should be reserved for the yellow calcareous sandstones of the CARADOC STRATIGRAPHY AND BRACHIOPODS 189 C H RONOSTRATIGRAPHY UTMOST R ATIGRAPH Y BIOSTRAT GRAPHY Series Stage irachiopods Trilobites O O Q < at < O i_ 0) Q. a D O n n i a n Onny Shale Formation O n n i e 1 1 a Sericoidea O n n i a A c t o n i a n Acton Scott Fm. W i stanstow/ — ->^H e n 1 e y Member J Member O n n i e 1 1 a Cryptothyrij PI a t y 1 i c h a s Chasmops Ragdon Member O n n i e 1 1 a Hedstroemina Cheney Lo n g ville Fm. Crosspipes Member Broeggerolithus Marshbrookian D a 1 m a n e 1 1 a Woo 1 s t o n i a n G 1 y n b o r o Member K j a e r i n a Bancroftina H e t e r or t h i s Alternate Limestone Formation i_ _i L o n gvillia n Morderley Sandstone Formation Fig. 3 Revised stratigraphical classification of the type upper Caradoc Series. Acton Scott district, whilst he referred to the time-equivalent division elsewhere simply as 'grey mudstones and limestones' (Dean 1958). The Geological Survey of Great Britain (Greig et al. 1968) essentially followed the previous nomenclatures, although they amalgamated all the facies of the Actonian into an Acton Scott Group, and did not subdivide the Cheney Longville Flags. Lithostratigraphy : Cheney Longville to Soudley The lithostratigraphical units recognized here are essentially those of Bancroft (19290, b, 1933) and Dean (1958, 1964). However, it has proved desirable to revise some of the lithostratigraphic boundaries, as well as to formalize the nomenclature and to define boundaries precisely (Fig. 3). The stratigraphy of the Soudley to Cheney Longville area will be described first, followed by a short account of the Cardington - Chatwall district (Fig. 1). a. Horderley Sandstone Formation This formation immediately underlies the Alternata Limestone Formation from Soudley (SO 477918) to Cheney Longville (SO 415852). It has a maximum thickness of 180 m, in the Onny Valley (SO 4 15858), and consists of laminated and cross-bedded medium-grained sandstones (Fig. 16, p. 213). The basal beds are some 70m thick and commence with laminated medium- grained sandstones which pass gradationally from the underlying Glenburrell Beds (Dean 1964). Interbedded shale units are common and may reach 10 cm in thickness. The middle part of the sequence is well exposed in the Long Lane Quarries (SO 413842) and the large Onny Valley Quar- ries (SO 415858), and consists of 50 to 60 m of massive, low angle cross-laminated or parallel- laminated medium-grained sandstone (Figs 19, 21). Shelly coquinas are common in the laminated sandstones, often reaching a thickness of 50 cm. The uppermost 30 or 40 m of this formation is only developed in the Onny Valley, and consists of parallel to very low-angle cross-laminated medium-grained sandstone. Interbedded thin bioturbated fine sandstones (c. 5 cm) occur in the upper 5 m and form a transition into the overlying Alternata Limestone Formation. 190 J. M. HURST O A . v E c o A i i 5_ 4-1 II *• -* x \7 o> V^ • -o E •- *.' •- m 0 ,, o 1 eo u_ simple tube burrow low angle cross lamination comminuted shell debris ripple drift cross lamination parallel lamination c o q u i n a complete bioturbation mottled bioturbation Fig. 4 Sedimentological log of the boundary between the Alternata Limestone Formation and Cheney Longville Formation at Soudley Quarry, SO 477918. CARADOC STRATIGRAPHY AND BRACHIOPODS 191 To define a lower lithological boundary to this formation, or to subdivide it, is outside the scope of this paper. However, an upper boundary is defined since it is also the base of the Alternata Limestone Formation. Some 10 m west of a footbridge across the Onny River (SO 418856) the contact between the Horderley Sandstone Formation and the Alternata Limestone Formation is exposed. The top of the Horderley Sandstone Formation is defined at the base of the first thick shelly limestone influx (coquina) containing Heterorthis alternate/, Trematis punctata and occasional phosphatic nodules and fossils (see Fig. 12, p. 202). In this section the initial coquina is some 50 cm thick, and occurs on both sides of the river bed. This section is also described below as the standard section for the base of the Woolstonian Stage (Fig. 12). North of the Onny Valley in Marshbrook Railway Cutting (SO 440905) a faulted outcrop of the Horderley Sandstone Formation and Alternata Limestone Formation is seen. The contact is not now accessible (see Hurst 1979), but from the available exposure the Horderley Sandstone Formation does not appear to be as complete as in the Onny Valley. It consists of approximately 15 m of laminated medium-grained purple or greenish sandstone with thin shelly lenses along the laminae and with occasional pebble stringers 1 cm thick. These sediments are most similar to the middle or upper part of the Onny Valley sequence and there does not appear to be a gradation into sediments typical of the upper Horderley Sandstone Formation of the Onny Valley or into the overlying Alternata Limestone Formation. At Soudley Quarry (SO 4779 1 8) parallel-laminated medium-grained purple and green sandstones with occasional thin shelly layers (Fig. 16) are overlain by the Alternata Limestone Formation (Hurst 1979). This is presumably one of the localities that prompted Cobbold (1901) to refer to a 'Chatwall or Soudley Sandstone'. As Dean (1960a) has demonstrated and will be discussed later, the Chatwall Sandstone is a lithostratigraphic unit distinct from the Soudley Sandstone. However, there is no lithological difference between the Horderley Sandstone Formation and Soudley Sandstone. Thus the term Soudley Sandstone is abandoned in favour of the Horderley Sandstone Formation, which has priority. b. Alternata Limestone Formation This stratum is named after the brachiopod Heterorthis alternata which occurs abundantly in shelly limestone lenses (Fig. 24). Cobbold's name (1901) is retained because the index species Heterorthis alternata is the type-species of the genus and the Alternata Limestone Formation includes the type-locality for the species. The Alternata Limestone Formation consists of a series of interbedded bioturbated silty shales and sandy silts (c. 10cm), laminated and graded fine sandstones often with a basal shell lag (10 to 50 cm thick), and thick shelly limestone bands (Fig. 22). The latter can be 50 cm thick, and consist of packed coarse shell material set in a matrix of sandy silt. The shells show no signs of grading or preferred orientation, but the beds appear crudely laminated owing to the alignment of large flat shells of Heterorthis alternata and Kjaerina bipartita (Fig. 24). As Dean (1958, 1964) pointed out, this formation is a useful mapping horizon at the base of the Cheney Longville Formation. Nowhere does it exceed 30 m in thickness, and its maximum development is at Woolston (SO 423873). Between the roadside exposure west of Cheney Long- ville (SO 416852) and the Onny Valley (SO 418856 and SO 418857) it maintains a thickness of approximately 20 to 25 m. At Marshbrook (SO 440905) only some 15 m are exposed but its true thickness is probably greater than this. Only 2 m of Alternata Limestone Formation occurs at Soudley Quarry (SO 4779 18). This small thickness, and the fact that phosphatic nodules and phosphatized fossils also occur, suggests a condensed sequence (Fig. 16). Between Soudley and Cheney Longville it is only the basal metre of the succession which contains phosphatized frag- ments. An analysis of the facies changes at this horizon, with detailed sections, are given by Hurst (1979). Discussion of section correlations of the uppermost Horderley Sandstone Forma- tion and Alternata Limestone Formation are given later under Chronostratigraphy, p. 210. Soudley Quarry (SO 477918) is the only place where the contact of the Alternata Limestone Formation and the Cheney Longville Formation is seen. This is not the ideal locality to define the junction of these two formations. However, in the Marshbrook (SO 440905), Woolston (SO 192 J. M. HURST Fig. 5 Examples of sedimentary logs of the Glynboro Member of the Cheney Longville Formation, from (A) the stream section at Soudley SO 478916 and (B) the old river cliff locality in the Onny Valley at SO 422890. Symbols explained in Fig. 4. CARADOC STRATIGRAPHY AND BRACHIOPODS 193 423873) and Onny Valley (SO 418857) sections there is a gap, of 5 to 10m stratigraphically, between the highest undoubted Alternata Limestone Formation exposure and lowest strata of the Cheney Longville Formation. As can be seen from Fig. 4, p. 190, the basal Cheney Longville Formation is readily distinguishable from the Alternata Limestone Formation. The top of the latter is taken to coincide with the top of the last thick shelly limestone coquina with abundant Heterorthis alternata. Above this the Cheney Longville Formation is characterized by the absence of shelly lenses, and the sediment is also slightly coarser. c. Cheney Longville Formation (Figs 17, 18, 20) The Cheney Longville Formation is an homogeneous sequence of sands, silts and sandy silts, some 180 to 200 m thick and weathering greenish-yellow. It was originally named by La Touche (1884), Fig. 2. Two members are recognized. Glynboro Member. The basal member, named after the cottage of Glynboro (SO 418859) in the Onny Valley, corresponds precisely to the Lower Cheney Longville Flags of Dean (1958). It comprises approximately 120 m of buff yellowish to olive green weathering siltstones and sand- stones. Sandstone beds 10 to 50cm thick are interbedded with 20cm bioturbated sandy silt units (Fig. 17). The sandstone beds are graded and have planar erosive bases and occasionally a persistent basal coquina which gradationally passes up into parallel or very low-angle cross- lamination. The bioturbated horizons account for approximately 10% of the succession in the lower half of the Member and up to 25% in the upper part. The fauna is best in the coquinas, but occurs as comminuted debris throughout. A very sparse fauna occurs in the bioturbated horizons and is the same species as in the shell beds. An example of a sedimentary facies log of this member is shown in Fig. 5 (see also Hurst 1979). The base of the Glynboro Member and the Cheney Longville Formation is defined at Soudley Quarry (Fig. 4). Very few good sections exist in the lowest part of this Member, the most extensive being in the stream section (SO 478916) south-west of Soudley Quarry (see Hurst 1979). Poor sporadic exposures occur in the lane west of Cheney Longville and in the Onny Valley. The higher beds are better represented. An almost continuous laneside exposure due west of Cheney Longville (SO 418851) includes the transition into the overlying Crosspipes Member. Further limited exposures occur at Woolston, and in an old cart track at Whittingslow (SO 436892) there is another continous exposure into the Crosspipes Member. An extensive section is present in an old river cliff in the Onny Valley (SO 422890). Crosspipes Member. This member, named after Crosspipes Farm (SO 429892), comprises a maximum of 60m and corresponds to the Upper Cheney Longville Flags of Dean (1958). It consists of thin laminated and graded very fine sandstones (c. 10-15 cm) interbedded with thick (up to 50 cm) bioturbated sandy and silty shales (Figs 18, 20). The tops of the fine sandstone beds are penetrated by simple tube burrows and grade into the overlying bioturbated units. Fauna is abundant in the bioturbated units and the same species occur concentrated as lensing shell lags at the base of the laminated sandstones (Fig. 25). Some lenses are up to 15 cm thick and are persistent. Occasionally in the bioturbated silty shales thin lenses of small fragments of volcanic glass, pumice and orthoclase feldspar crystals can be traced across the outcrop. The thickness of the laminated fine sandstones decreases quickly up through the member until the top 10 or 15m consists primarily of bioturbated and partially bioturbated sandy silts and sandy shales. Shell concentrations up to 5 cm thick become common in the top 10 m of strata, but they are not associated with laminated graded sand units (Fig. 18). Fig. 6 gives examples of sedimentary logs from this member. The Crosspipes Member (lowest part) is well exposed at the motte and bailey at Cheney Longville (SO 419849), in an extensive river section in Marshwood (SO 442891) and a river section near Ragleth Hill (SO 451908). The base of the Crosspipes Member, and thus the top of the underlying Glynboro Member, is taken in a fairly continuous roadside exposure at Cheney Longville (SO 418851 to SO 419849), and coincides with the stage boundary between the Woolstonian and the Marshbrookian (Fig. 1 3, 194 J. M. HURST Fig. 6 Examples of sedimentary logs of the Crosspipes Member of the Cheney Longville Formation, from (A) the Minton road section at SO 439900 and (B) the stream section in Marshwood at SO 422891. Symbols explained in Fig. 4. p. 203). The lithological boundary between the two members is gradational over several metres. The base of the Crosspipes Member is taken at the increase in thickness of bioturbated units intercalated with thin laminated sandstone beds. This change is also accompanied by a decrease in sediment grain size from predominantly sands and sandy silts to silts and silty shales (Fig. 20). The fauna is also more abundant in the Crosspipes Member. d. Acton Scott Formation In the Onny Valley above the Cheney Longville Flags and below the Onny Shale Formation (Fig. 3) the sequence consists of two basic facies, a siltstone below and a calcareous siltstone above. However, to the north, around Acton Scott, the siltstones are succeeded by hard calcareous sandstones. Each of these three divisions is defined here as a member. Ragdon Member. Named after the village of Ragdon (SO 458915), this comprises approximately 30 m of mottled (intensely bioturbated) siltstones, weathering buff to green. Occasionally burrow- disrupted current laminae occur, and non-bioturbated primary bedding accounts for about 2% of the succession. Isolated laminated siltstone beds with a basal layer of shell debris occur rarely. The rich and diverse fauna is distributed randomly throughout the sequence, although paper-thin layers of bedding-parallel shell fragments and convex-upward shells (not associated with laminated silts) occur at a frequency of about one layer per 5 m (Fig. 26). This member weathers easily to form low-lying land, and as a result is poorly exposed, mainly in overgrown stream sections. It is not exposed south of the Onny Valley; a stratigraphically low but poor exposure under a tree in an abandoned meander of the river is the most southerly out- crop (SO 423854). Two small outcrops in the middle part of the member occur on the north side of the river in a roadside bank (SO 424855) and in an overgrown valley known as Dandy Hollow 195 Section at A-B 92 ip 4m I jp 13ml a p 5 m 35B °P 5m I « 46 300 m Limy Bed £ x i_ E c -1 c 0 O • — c o o E o M Alterna 0 u_ ~o 0 dstone c C c o 0 o CO ™ o — X E '> a> L. a> 0 c <0-05) from P. attenuata from the 226 J. M. HURST Meadowtown Beds but not from the other samples of P. attenuata from the Rorrington Beds and the Upper Llanvirn Betton Beds. P. lockleyi and P. attenuata further differ in that the former has a deeper brachial and pedicle valve. Also, the pedicle trough of the new species is significantly longer and the ornament consists solely of fine concentric growth lines, there being no indication of fila as in P. attenuata. Subfamily GLOSSELLINAE Cooper, 1956 Genus PSEUDOLINGULA Mickwitz, 1909 Pseudolingula sp. Fig. 56 Four fragmentary specimens from the Crosspipes Member at Marshwood Quarry (loc. 27) and one from the Acton Scott Formation (loc. 44) are the sole representatives of Pseudolingula in the Upper Caradoc. A relatively complete articulated specimen (BB 72255) was 14-5 mm long and 10-mm wide with a biconvex parallel-sided shell. Shell surface appears to have been smooth. Family PATERULIDAE Cooper, 1956 Genus PATERULA Barrande, 1879 Paterula cf. subcircularis Cooper Figs 57-60 1956 Paterula subcircularis Cooper : 239-240; pi. 24, figs 7-10. DESCRIPTION. Shallow biconvex subcircular Paterula with a brachial valve 95% as wide as long and a pedicle valve 93 % as wide as long; shell surface ornamented by very fine concentric growth lines; width of shallow pedicle notch 5% of the length of the valve and situated immediately posterior to the ventral beak which is submarginal and located at 12% of the length of the valve; submarginal dorsal beak, located at 1 1 % of the length of the brachial valve. Well-defined limbus 7 % (range 4 % to 10 % for 7 valves) and 8 % (range 5 % to 10 % for 9 valves) as wide as the pedicle and brachial valves are long, respectively. Pedicle valve with extremely faint short median ridge extending anteriorly from pedicle notch; two lines diverge anteriorly from the beak. HORIZON AND LOCALITIES. All specimens from the basal Onny Shale Formation; loc. 49. Repre- sentative dimensions of pedicle valves: BB 72257, length 1-3 mm, width 1-3 mm; BB 72258, length 2-1 mm, width 2-0 mm. Brachial valves: BB 72259, length 2-0 mm, width 2-0 mm; BB 72260, length 1-1 mm, width 1-1 mm. Paterula sp. Fig. 61 Several internal brachial moulds of a paterulid (e.g. BB 72261, length 1-8 mm, width 1-4 mm) have been recovered from the Acton Scott Formation at loc. 44. Two valves were on average 74% as wide as long with a limbus 10% as wide as the valves are long, and a submarginal dorsal beak located at 14% of the length of the valve. Internal features unknown. DISCUSSION. The Caradoc Paterula belong to two distinct species. The sample from the Onny Shale Formation was used to derive the statistics in Tables 5-7. The species is comparable with P. subcircularis, described by Cooper (1956 : 239) from the Springtown Shale of Oklahoma, in the near-circular valve shape and shallow biconvexity. However, the Caradoc specimens differ from the large P. subcircularis in their small size and in having a shallow pedicle notch as opposed to a deep one. These minor morphological differences do not justify the erection of a new species. No useful purpose is served in comparing the scarce Paterula sp. with described species. CARADOC STRATIGRAPHY AND BRACHIOPODS 227 Genus ELLIPTOGLOSSA Cooper, 1956 Elliptoglossa sp. Fig. 63 A single articulated specimen (BB 72263) from the Acton Scott Formation at Chuney Pool (loc. 43) is provisionally assigned to Elliptoglossa. The shallow biconvex shell was 4-8 mm long and 4-2 mm wide, with a subelliptical outline and a shell surface ornamented by very fine growth lines. Genus L1NGULOPS Hall, 1872 Lingulops sp. Fig. 62 Complementary moulds of a brachial valve (BB 72264) from a temporary exposure in the Cross- pipes Member at loc. 24 are the sole representatives of the genus in the Upper Caradoc. The valve, which had a subelliptical outline, was 3-5 mm long, 2-2 mm wide and 0-5 mm deep, with a submarginal beak located 1 mm from the posterior margin. Muscle field anteriorly impressed. Shell surface ornamented by very fine growth lines. Family CRANIOPSIDAE Williams, 1963 Genus PARACRANIOPS Williams, 1963 Paracraniops doyleae sp. nov. Figs 64-69 DESCRIPTION. Oval, subequi valve Paracraniops with a dorsal (?) valve 81 % as wide as long and 17% as deep as long; transverse profile evenly convex, longitudinal profile eccentrically convex with a submarginal apex slightly overhanging a truncated posterior margin with a small groove; anteriorly flattened. Ventral (?) valve 82% as wide as long and 14% as deep as long; transverse and longitudinal profiles, and apex and posterior margin characteristics as for dorsal (?) valve. Valve edges with small flange about 7 % length of valve. Shell surface ornamented by lamellose growth lines which become widely spaced anteromedially. Ventral (?) interior with elongate, semicircular, slightly elevated platform, with well-defined periphery, extending anteriorly from the valve apex for 60 % the length of the valve, and containing a pair of posterior adductor scars diverging widely at over 100° from the valve apex, bounded anterolaterally by a pair of elongate oval anterior adductor scars; centre of platform accommo- dating elongate peardrop-shaped oblique internal scar. Dorsal (?) interior with two very poorly defined ridges extending from the valve apex for approximately three-quarters of valve length to the edge of the poorly-defined anterior adductor scar. Posterior adductor and oblique internal scars very poorly developed but appearing to diverge widely from valve apex. NAME. After Ms M. C. Doyle. DIMENSIONS. length width Internal mould of pedicle valve, BB 72278, holotype 4-5 4-0 Internal mould of pedicle valve, BB 72279 4-3 3-5 Internal mould of pedicle valve, BB 72280 4-0 3-3 Internal mould of brachial valve, BB 72281 4-5 4-0 Internal mould of brachial valve, BB 72282 3-5 3-0 External mould of brachial valve, BB 72283 4-5 3-9 HORIZON AND LOCALITIES. Type material from loc. 44. Also recorded from Iocs 11, 12, 15, 19, 24, 26, 30, 32, 34, 35, 43, 45, 46, 47 and 48. 228 J. M. HURST 73 74 Figs 64-77 Paracraniops doyleae sp. nov., internal moulds of pedicle (?) valves: 65 = holotype BB 72278, x6-2, 67 = BB 72279, x6, 66 = BB 72280, x6; internal moulds of brachial valves: 64 - BB 72281, x6, 68 = BB 72282, x6, 69 = BB 72283, x5. Trematis punctata (J. de C. Sowerby) : 70 = brachial valve BB 54739, x3, 71 = pedicle valve BB 72256, x 1-5. Schizocrania salopiensis Williams, brachial valves: 72 = BB 24955, x 3-5, 73 = BB 24953, x3, 74 = BB 30077, x 3-5, 75 = BB 72265, x 3-5, 76 = BB 72266, x 2-5, 77 = BB 72267, x3-5. DISCUSSION. Paracraniops occurs sporadically, but occasionally locally very abundantly, through- out the Cheney Longville Formation and the Acton Scott Formation. Only material from the Acton Scott Formation was well enough preserved for statistical analysis, the results of which are presented in Tables 8-12. The Paracraniops from the Cheney Longville Formation is considered conspecific with P. doyleae. The Salopian Paracraniops is closely related to P. macella Williams, which is known from the older Longvillian (Gelli-grin Ashes) rocks of Bala, north Wales (Williams 1963 : 346-349). They principally differ in that the former is more transverse and has a brachial valve deeper than the pedicle valve. Further, in the ventral (?) interior of P. doyleae the posterior adductor scars are obtusely splayed and in the dorsal (?) interior the ridges more widely splayed. CARADOC STRATIGRAPHY AND BRACHIOPODS 229 Order ACROTRETIDA Kuhn, 1949 Suborder ACROTRETIDINA Rowell, 1949 Superfamily DISCINACEA Gray, 1840 Family TREMATIDAE Schuchert, 1893 Genus TREMA TIS Sharpe, 1848 Trematis punctata (J. de C. Sowerby) Figs 70, 71 1839 Orbicula granulata J. de C. Sowerby /// Murchison : 636. 1839 Orbicula punctata J. de C. Sowerby in Murchison : pi. 20, fig. 5. 1866 Discina (Trematis) punctata (J. de C. Sowerby) Davidson : 69; pi. 6, fig. 9. DESCRIPTION. Subcircular, gently convex brachial valve (e.g. BB 54739, length 18-1 mm, width 19-2 mm), about one-sixth as deep as long with maximum depth posteriorly; small dorsal umbo but no pseudointerarea observed on specimens available. Ornament of small hexagonal pits arranged consistently in a hexagonal megapattern. Dorsal interior posteriorly thickened with shallow broad median ridge separating two very shallow oval muscle scars. Pedicle valve evenly convex, depressed posteriorly, with eccentric beak and open triangular pedicle notch. HORIZON AND LOCALITIES. Occurs at Iocs 1, 2, 3, 4, 5, 6, 7 and 8. Lectotype (selected Cocks 1978, not figured here) GSM 6878, from Cheney Longville Formation west of Church Preen. DISCUSSION. Wright (1963) analysed the pit arrangement of Trematis. T. punctata has hexagonally arranged pits, and is similar to T. millepunctata Hall and T. craigensis Reed. Genus SCHIZOCRANIA Hall & Whitfield, 1875 Schizocrania hewardi sp. nov. Figs 78-83 DESCRIPTION. Schizocrania with subcircular brachial valve, 97% as long as wide and 23% as deep as long, transverse and asymmetric profiles evenly convex; umbo posteriorly placed over- hanging straight truncated posterior margin with a strong groove; shell surface ornamented by faint persistent growth lines and straight radial capillae. Dorsal interior with small elongately oval, widely divergent, posterior adductor scars, weakly impressed and extending anteriorly from the posterior margin for 23% of the length of the valve; anterior adductor not impressed; pedicle valve unknown. NAME. After Dr A. P. Heward. DIMENSIONS. length width Exterior of brachial valve, BB 72268, holotype 1-2 1-3 Exterior of brachial valve, BB 72269 1-0 1-2 Exterior of brachial valve, BB 72270 1-1 1-2 Exterior of brachial valve, BB 7227 1 1-1 1-2 HORIZON AND LOCALITY. All the material from the upper part of the Onny Shale Formation, from the Onny river cliff; loc. 51. Schizocrania salopiensis Williams Figs 72-77 1974 Schizocrania salopiensis Williams : 44-46; pi. 6, figs 22-26. DESCRIPTION. Schizocrania with subcircular to subelliptical or subtriangular brachial valve, 95 % as long as wide and 29 % as deep as long, evenly convex in both transverse and longitudinal profile; submarginal umbo not overhanging smoothly rounded posterior part of valve; shell 230 J. M. HURST surface ornamented by very faint impersistent growth lines and radial capillae, which branch dichotomously and curve towards the anterior and lateral margins. Dorsal interior with large slightly divergent subcircular to oval posterior adductor scars, weakly impressed and extending anteriorly from the posterior margin for 29% of the length of the valve; small circular anterior adductor scars situated near the centre of the valve and separated by a very shallow median ridge; pedicle valve unknown. HORIZON AND LOCALITIES. BB 72265-7 are from loc. 44; BB 24953, BB 24955 and BB 30077 from loc. 27. Also recorded at Iocs 36 and 45. Dimensions of brachial valves ranged from BB 72265, with length 4-6 mm and width 5-5 mm, to BB 24953, length 12-6 mm and width 10-4 mm. DISCUSSION. Schizocrania brachial valves are fairly common in the Upper Caradoc rocks, which have provided two samples used to derive the statistics given in Tables 13-15. Data on shell shape, depth and length of posterior adductor scars show that S. hewardi differs from S. salopiensis Williams (1974 : 46) in valve shape (P< 0-01) and length of the posterior adductor scars (P< 0-05). They do not differ in shell depth. Further, S. hewardi has an incurved posteriorly-placed umbo along a straight truncated valve margin with a strong groove, whilst S. salopiensis has a rounded posterior profile without an overhanging umbo. The posterior adductor scars of S. hewardi are also more widely splayed. S. hewardi differs from S. salopiensis Williams, from the Upper Llandeilo Rorrington Beds of the Shelve district, in terms of valve shape (P< 0-001), depth (/><0-001) and length of posterior adductor scars (/><0-05). From the younger sample of S. salopiensis from the Lower Caradoc Spy Wood Grit S. hewardi differs in valve shape (P< 0-001) and length of posterior adductor scars (P<0-05). There is no difference in valve depth of the two species. Such differences are considered sufficiently important to warrant the erection of a new species, even though no pedicle valves of either species have yet been recovered. The Schizocrania from the Crosspipes Member and Acton Scott Formation is similar to S. salopiensis in respect of its rounded rather than truncated posterior margin and submarginal umbo. However, it differs significantly (/><0-05) from both the Spy Wood Grit and Rorrington Beds samples of S. salopiensis (see Williams 1974 : 46) in shell shape as it grew longer relative to width. This difference is not considered sufficiently important to warrant the erection of a new species. Family DISCINIDAE Gray, 1840 Subfamily ORBICULOIDEINAE Schuchert & Le Vene, 1929 Genus ORBICULOIDEA d'Orbigny, 1847 Orbiculoidea ovata sp. nov. Figs 84-88 DESCRIPTION. Biconical, subcircular to suboval Orbiculoidea with a brachial valve on average 92% as wide as long (range 76% to 99% for 5 valves), 27% as deep as long (range 27% to 31 % for 5 valves); transverse profile evenly convex, longitudinal profile eccentrically convex with a beak located on average 37% (range 29% to 41 % for 5 valves) of the length of the brachial valve; pedicle valve on average 86 % as wide as long (values for 2 valves 80 % and 92 %), 23-5 % as deep as long (values for 2 valves 23% and 24%); transverse profile evenly convex, longitudinal profile eccentrically convex with a conical beak located 39 % (value for 2 valves 39 % each) of the length of the pedicle valve, which is flattened anteriorly; surface of both valves ornamented solely by coarse well-rounded concentric growth lines. Figs 78-83 Schizocrania hewardi sp. nov., all brachial valves: 79 and 81 = holotype, BB 72268, viewed from above and obliquely, x,75; 78 and 80 = BB 72271, viewed half obliquely and from side, x75 and x 100 respectively; 82 = BB 72269, partly exfoliated valve viewed from above, x 100; 83 = oblique enlargement of ornament on BB 72271, x400. Figs 78-83 are all stereoscan photomicrographs with approximate magnifications. 231 232 J. M. HURST Figs 84-92 Orbiculoidea ovata sp. nov., brachial valves: BB 72273, x2-7, 86 - BB 24949, x2; pedicle valves: x 2-7. Orbiculoidea sp. : 89 = brachial valve BB 24948, valve BB 24950, x8-5. Schizotreta sp. 2, brachial valves: x3. 84 = BB 72274, x 2-5, 85 - holotype 87 = BB 72275, x3, 88 = BB 31260, x3-3. Schizotreta sp. 1 : 92 = pedicle 90 - BB 72276, x3-5, 91 = BB 72277, Dorsal interior with very indistinct muscle impressions anterior of apex and from which two indistinct pallial tracks extend anterolaterally. Ventral interior smooth and with pedicle groove extending posteriorly from the umbo for 31 % (values for 2 valves 31 % each) of the length of the valve. DIMENSIONS. Exterior of brachial valve, BB 72273, holotype Exterior of brachial valve, BB 72274 Exterior of pedicle valve, BB 72275 Partially exfoliated pedicle valve, BB 31260 Internal mould of brachial valve, BB 24949 length 9-8 8-7 6-5 11-5 11-5 width 9-2 8-6 5-2 10-0 11-0 HORIZON AND LOCALITIES. BB 72273-4 from type locality, loc. 44. Paratypes BB 24949 from loc. 47, BB 72275 from loc. 48 and BB 31260 from well excavation at Plaish. Also recorded from Iocs 35 and 43. Orbiculoidea sp. Fig. 89 There is a partially exfoliated brachial valve in the BM(NH) collection (BB 24948) from calcareous sediments of the Acton Scott Formation at Batch Gutter in the Onny Valley (SO 4245 8538). The valve, which was 6-0 mm long, 5-7 mm wide and 1-6 mm deep, was subcircular in outline with the beak situated 2-2 mm forward of the posterior margin. The transverse profile was evenly convex and the external surface ornamented by two sets of fila, one very fine and on average 2 mm apart and the other intercalated and very impersistent. Orbiculoidea occurs sparsely in Upper Caradoc rocks of east Salop. Most specimens belong to the new species O. ovata, which differs from the present Orbiculoidea sp. only in details of the ornament. Genus SCHIZOTRETA Kutorga, 1848 Schizotreta sp. 1 Fig. 92 An internal mould of a pedicle valve in the BM(NH) collection (BB 24950) is the sole representa- tive of Schizotreta sp. 1. It is from the Onny Shale Formation of the Onny river cliff, loc. 51. The CARADOC STRATIGRAPHY AND BRACHIOPODS 233 valve, which was approximately 3-1 mm long, 2-2 mm wide and 0-6 mm deep, was oval in outline with the low depressed beak situated 1 -0 mm forward of the truncated posterior margin. The transverse profile was evenly convex and the external surface ornamented by coarse fila. Schizotreta sp. 2 Figs 90, 91 Two partially exfoliated brachial valves (BB 72276-7) from loc. 44 of the Acton Scott Formation are the only representatives of Schizotreta sp.-2 from the east Salop Upper Caradoc. BB 72276, 5-5 mm long, 5-3 mm wide and 1-1 mm deep, is suboval in outline with a rounded anterior margin, parallel lateral margins and a truncated posterior with a slightly overhanging marginal beak; its internal features are unknown. The shell surface is ornamented by very coarse concentric ridges. BB 72277a, b, a fragmentary specimen, is 7 mm wide and over 7-5 mm long and 1-5 mm deep. DISCUSSION. It is considered unlikely that Schizotreta sp. 1 and sp. 2 are conspecific, because of their difference in ornament. In the development of strong coarse concentric ridges, submarginal beak and truncated posterior margin Schizotreta sp. 2 closely resembles the older S. corrugata Cooper from the Porterfield Pratt Ferry Formation of Alabama (Cooper 1956 : 277). Class ARTICULATA Huxley, 1869 Order ORTHIDA Schuchert & Cooper, 1932 Suborder ORTHIDINA Schuchert & Cooper, 1932 Superfamily ORTHACEA Woodward, 1852 Family ORTHIDAE Woodward, 1852 Subfamily PRODUCTORTHINAE Schuchert & Cooper, 1931 Genus NICOLELLA Reed, 1917 Nicolella actoniae (J. de C. Sowerby) 1839 Orthis actoniae J. de C. Sowerby in Murchison : 639; pi. 20, fig. 16 (left); non pi. 20, fig. 16 (right). 1963 Nicolella actoniae (J. de C. Sowerby); Williams : 352-356; pi. 1, figs 15-19. cf. 1974 Nicolella cf. actoniae (J. de C. Sowerby); Williams : 57-58; pi. 9, figs 1-6. cf. 1977 Nicolella cf. actoniae (J. de C. Sowerby); Mitchell : 31-32; pi. 3, figs 14-20. Williams (1963) has redescribed and figured N. actoniae, and his work will not be duplicated here. He records the species definitely from loc. 45, where in my collections it is rare, and possibly from loc. 47, where I have not found it. It also occurs at loc. 43 and abundantly at loc. 44. All these localities are in the Acton Scott Formation. Family DOLERORTHIDAE Opik, 1934 Genus DOLERORTHIS Schuchert & Cooper, 1931 Dolerorthis virgata (J. de C. Sowerby) Figs 93-105 1839 Orthis virgata J. de C. Sowerby in Murchison : 639; pi. 20, fig. 15. 1869 Orthis calligramma var. virgata J. de C. Sowerby; Davidson : 240 pars; pi. 35, figs 23, 24, non figs 1-22. 1869 Orthis virgata J. de C. Sowerby; Davidson : 240; pi. 37, fig. 2. 1958 Dolerorthis duftonensis (Reed); Dean : pi. 25, fig. 1. 1978 Dolerorthis virgata (J. de C. Sowerby); Cocks : 45. DESCRIPTION. Subcircular to subquadrate, unequally bioconvex Dolerorthis with rectangular cardinal margins; brachial valve 76% as long as wide, evenly convex, non-sulcate and 14% as deep as long; dorsal interarea anacline, notothyrium open; pedicle valve 84% as long as wide, 234 J. M. HURST evenly convex and 24% as deep as long; long apscaline interarea, delthyrium open; ornamen- tation costellate commonly with two ribs per mm, 10 mm anteromedially of the dorsal umbo, cancellated by fine concentric lamellae. Ventral muscle scar subpentagonal, extending anteriorly for 35% of the length of the valve and 30% as wide as the valve is long; lanceolate adductor scar not surrounded by diductor tracks; teeth strong, supported by short subparallel dental lamellae extending anteriorly of umbo for 17% of the valve length. Dorsal interior with simple ridge-like cardinal process; brachiophores simple, stout, widely divergent with their bases extending anteriorly of the dorsal umbo for 16% of the valve length and splaying laterally for 33% of the valve length; dorsal adductor field quadripartite with posterior scars larger than anterior and extending anteriorly for 50% of the length of the valve. MATERIAL. Lectotype (selected Cocks 1978: 45) GSM Geol. Soc. Coll. 6904 (Fig. 95), length 17-9 mm, width 22-2 mm, from calcareous sandstones at Acton Scott, possibly Iocs 47 or 48. Other material from loc. 12, coll. O18 (e.g. Figs 93, 102, BB 72395, length 19-8 mm and width 25-0 mm, and Fig. 94, BB 72396, length 14-5 mm and width 19-0 mm), O112 (e.g. Fig. 100, BB 72397, length 24 mm and width 26-0 mm, and Fig. 105, BB 72398, length 9-8 mm and width 13-8 mm) and O19 (e.g. Fig. 97, BB 72399 and Fig. 104, BB 72400); also loc. 27, coll. M20 (e.g. Fig. 98, BB 72401, Fig. 103, BB 72402, Fig. 99, BB 72403 and Fig. 96, BB 72404); also loc. 40, coll O103 (e.g. Fig. 101, BB 72405) and Iocs 10, 11, 17, 32, 41 and 48. DISCUSSION. Dolerorthis occurs sparingly in the Upper Caradoc, but the small samples have been combined to obtain the statistics in Tables 16-25. The sample from the Alternata Limestone and Glynboro Member does not differ in terms of the known bivariate characteristics from the Crosspipes Member and Acton Scott Formation sample, although a larger collection may show the later sample to consist of slightly more coarsely-ribbed individuals. However, such a minor difference is not taxonomically important, so all Upper Caradoc Dolerorthis are included in D. virgata. D. virgata differs from D. duftonensis prolixa Williams in a number of characters, namely in having a relatively longer brachial valve (P<0-01), a slower increase in depth of the brachial valve (/><0-02), relatively longer dorsal cardinalia (/><0-01) and a slower increase in length of the dorsal muscle field(P< 0-01). D. virgata does not differ in any of its statistical data from those given by Williams (1963 : 60-61) for D. duftonensis. However, there may be significant differences between the two as the former species appears to have a shallower pedicle valve and a longer ventral muscle field. D. duftonensis could be a synonym of D. virgata, but until more is known about the variation of D. duftonensis the two species are kept separate. Family DINORTHIDAE Schuchert & Cooper, 1931 Genus D1NORTH1S Hall & Clarke, 1892 Dinar this sp. Figs 106, 111 DESCRIPTION. Subquadrate, dorsibiconvex Dinorthis with a brachial valve 76% as long as wide and 26% as deep as long; pedicle valve weakly convex with slightly raised umbonal region, and Figs 93-111 Dolerorthis virgata (J. de C. Sowerby), internal moulds of pedicle valves: 93 = BB 72395a, xl-6, 94 = BB 72396, x2-6, 95 = lectotype, GSM 6904, x2, 96 = BB 72404, >:!•!, 97 = BB 72399, x 1-2, 99 = BB 72403, x 1-1 ; external mould of pedicle valve: 102 = BB 72395b, x2-5; internal moulds of brachial valves: 98 = BB 72401, x 1-3, 100 = BB 72397, x 1-5, 101 = BB 72405, Y.2, 103 = BB 72402, X 1-2, 104 - BB 72400, x2, 105 - BB 72398, x2. Dinorthis sp., internal mould of pedicle valve: 106 = BB 72407, :< 1-5; internal mould of brachial valve: 111 = BB 72406, xO-8. Platystrophia sp. 1, internal mould of pedicle valve: 107 = BB 72394, xl; internal moulds of brachial valves: 108 = BB 72368, x2, 109 = BB 72367, xl-2. Platystrophia sp. 2, internal mould of brachial valve: 110 = BB 72369, x3-5. CARADOC STRATIGRAPHY AND BRACHIOPODS 235 107 236 J. M. HURST 73% as long as wide; ventral interarea apscaline, dorsal interarea orthocline; radial ornamenta- tion of strong costae; ventral muscle scar elongately subrectangular, 40% as long as the valve and 28 % as wide as the valve is long; teeth blunt and supported by short dental lamellae extending anteriorly for 15% of the valve length; cardinal process differentiated into short stout shaft and expanded myophore; brachiophores stout, unsupported and extending forward of dorsal umbo for 15% the length of the valve and splaying laterally for 29% the length of the valve; dorsal adductor field suboval, weakly impressed and extending anterior of dorsal umbo for about half the length of the valve. HORIZON AND LOCALITIES. BB 72406, brachial valve (length 28-7 mm, width 38-0 mm) from loc. 29, BB 72407, pedicle valve (length 15-4 mm, width 21-2 mm) from loc. 13. Also recorded from Iocs 15, 18, 30 and 32. Family PLECTORTHIDAE Schuchert & Le Vene, 1929 Subfamily PLATYSTROPHIINAE Schuchert & Le Vene, 1929 Genus PLATYSTROPHIA King, 1850 Platystrophia sp. 1 Figs 107-109 The only specimens, from the Crosspipes Member, are an incomplete complementary external and internal mould of a brachial valve (BB 72367; loc. 24, colln WFT2), a complete internal mould (BB 72368; loc. 24, colln WFT1) and an internal mould of a pedicle valve (BB 72394; loc. 27, colln M6). The brachial valves were 68 % as long as wide and 33 % as deep as long with a well-defined slightly rounded fold about 34% as wide as the valve. The radial ornamentation consisted of four costae on the fold and five on each lateral slope, with a wavelength of 1 mm, 5mm anteromedially of the dorsal umbo. Cardinal process ridge-like; short, thick, slightly divergent brachiophores bound wide notothyrium, extending anteriorly for 17% of the length of the valve and splaying laterally for 34% of the valve length; dorsal adductor scars strongly impressed, quadripartite, extending anteriorly for 49% the length of the valve on either side of a low thin median ridge. The pedicle valve was 25 mm wide, 20-2 mm long and 6 mm deep, with an elongately oval, subtriangular muscle field 32% as long as the valve and 18% as wide as the valve is long; small thick pedicle callist; teeth stout, supported by short dental plates extending anteriorly for 16% the length of the valve. Platystrophia sp. 2 Fig. 110 The only specimen, from the Acton Scott Formation, is an internal mould of a brachial valve (BB 72369 ; loc. 43, colln CP3). The valve, which was 4-7 mm long, 6-7 mm wide and 1 -7 mm deep, Figs 112-144 Rhactorthis grandis sp. nov., internal moulds of pedicle valves: 113 = BB 72349, x2, 114=8872351, x3-7, 115 = 8872353, x3, 116 = 8872352, x3, 117 = 8872357, x4, 118 = BB72354a, x3-3, 120 = BB 72350, x4; partially conjoined valves, holotype: 112 = BB 72348, x 4; external mould of pedicle valve: 1 19 = BB 72354b, x 3-3; internal and external moulds of brachial valves: 121 = BB 72359a, x2-5, 122 = BB 72359b, x2-5, 123 = BB 72362, X2-3, 124 = BB72361b, x3, 125 = BB72361a, x3, 126 = BB 72358, x3-2, 127 = BB 72360, x2-7, 128 = BB 72356, x5, 129 = BB 72355, x5-3, 130 = BB 72363, x3. Rhactorthis actoniaesp. nov., holotype, internal mould of pedicle valve: 131 = BB 72364, x 1-5; internal and external moulds of brachial valves: 132 = BB 72365, x3-3, 133 = BB 72366a, x2, 134 = BB 72366b, x2-2. Rhactorthis cf. crassa Williams, internal and external moulds of pedicle valves: 135 = BB 72381, X3-2, 136 = BB 72380, x2, 137 = BB 72382, x6-5, 138 = BB 72379a, x4-5, 139 = BB 72379b, X4-5; internal moulds of brachial valves: 140 = BB 72377, x5, 141 = BB 72376, x2, 142 = BB 72378, x 5-5. Gelidorthis sp., internal mould of pedicle valve: 143 = BB 72373, x 5. Plectorthid gen. et sp. indet., internal mould of pedicle valve: 144 = BB 72374, x3. CARADOC STRATIGRAPHY AND BRACHIOPODS 237 144 238 J. M. HURST had a poorly-developed rounded sulcus 2-5 mm wide. Radial ornamentation of three costae on the fold and seven on each lateral slope. Cardinal process ridge-like, brachiophore bases sub- parallel and extending anteriorly for 13% the length of the valve; fulcral plates present. Subfamily RHACTORTHINAE Williams, 1963 Genus RHACTORTHIS Williams, 1963 TYPE SPECIES. Rhactorthis crassa Williams, 1963 by original designation. Rhactorthis actoniae sp. nov. Figs 131-134 DESCRIPTION. Subcircular unequally biconvex Rhactorthis with sulcate brachial valve, 80% as wide as long and 20 % as deep as long; evenly convex pedicle valve 83 % as wide as long and 25 % as deep as long, pedicle interarea slightly curved apscaline, brachial interarea straight anacline; delthyrium and notothyrium open; radial ornamentation commonly of five costellae (measured on 4 valves) per mm, 2 mm anteromedially of the dorsal umbo, commonly arising by dichotomous branching; disruptive growth lines occasionally present. Ventral muscle scar subpentagonal and extending anteriorly for 30% the length of the valve and 33% as wide as the valve is long; diductor scar does not surround median adductor scar; teeth small, supported by receding dental lamellae and extending anteriorly for 18 % the length of the valve. Dorsal interior with very shallow median ridge, at posterior end of which arises continuous shaft of cardinal process which is crenulated posteriorly, fills notothyrium and projects slightly beyond hinge line; brachiophores very short, divergent and extending anteriorly for 17% the length of the valve, with delicate convergent bases which curve laterally parallel to hinge line for 34% the length of the valve; adductor scars very weakly impressed, elongately oval and extending anteriorly for 44% the length of the valve. NAME. Referring to the locality near Acton Scott. DIMENSIONS. length width Internal mould of pedicle valve, BB 72364, holotype 6-5 12-0 Internal mould of brachial valve, BB 72365 5-1 6-0 External and internal moulds of brachial valve, BB 72366a, b 9-6 10-7 HORIZON AND LOCALITIES. Type material from loc. 47, colln CF3, the only locality from which the species has been recovered. DISCUSSION. See under R. grandis, p. 239. Rhactorthis cf. crassa Williams Figs 135-142 cf. 1963 Rhactorthis crassa Williams : 372-375; pi. 4, figs 1-6. DESCRIPTION. Subcircular to subquadrate, moderately biconvex Rhactorthis with weakly sulcate brachial valve 82 % as wide as long and 25 % as deep as long, lateral slopes moderately strong, evenly convex with shallow rounded sulcus originating at umbo; pedicle valve moderately convex, 82% as wide as long and 32% as deep as long; ventral interarea apscaline, dorsal interarea anacline; delthyrium and notothyrium open; radial ornamentation of five to eight low angular costellae per mm, 2 mm anteromedially of the dorsal umbo, arising by dichotomous branching; internal branching common, externals not observed; concentric lamellae common, but no major disruption by growth lines. Ventral muscle scar elongately subpentagonal and extending anteriorly for 32% of the valve length and 28 % as wide as the valve is long; adductor scar not surrounded anteriorly by diductor CARADOC STRATIGRAPHY AND BRACHIOPODS 239 impressions; teeth small and supported by short divergent dental lamellae extending anteriorly for 20% of the valve length; dental lamellae continued along sides of muscle field as shell ridges and meet anteromedially to surround muscle field completely. Dorsal interior with wide rounded median ridge; cardinal process shaft broad and continuous with posterior end of median ridge; myophore crenulated, completely filling notothyrium and extending slightly beyond hinge line; subparallel to divergent short, stout brachiophores extending anteriorly for 17% the length of the valve, with convergent bases which sometimes curve laterally for 43% of the length of the valve; sockets small and deep; adductor scars very weakly impressed, situated on either side of median ridge and extending anteriorly for 45% of the valve length. MATERIAL AND LOCALITIES. Loc. 43, colln CP4 (e.g. Fig. 141, BB 72376, length 8-5 mm, width 9-6 mm), colln CP1 (e.g. Fig. 136, BB 72380, length 9-0 mm, width 10-0 mm and Fig. 142, BB 72378) and CP2 (e.g. Fig. 135, BB 72381). Loc. 45, colln AS4 (e.g. Fig. 140, BB 72377, length 3-2 mm, width 4-0 mm and Figs 138-9, BB 72379, length 4-5 mm, width 5-2 mm) and colln AS6 (e.g. Fig. 137, BB 72382, length 2-6 mm and width 3-1 mm.) Rhactorthis grandis sp. nov. Figs 112-130 DESCRIPTION. Subcircular, strongly biconvex Rhactorthis with strongly sulcate brachial valve 78% as wide as long and 23 % as deep as long, lateral slopes strong, evenly convex and with deep, persistent, rounded sulcus originating at umbo; pedicle valve strongly convex, sometimes strongly carinate, 84% as wide as long and 33 % as deep as long; apscaline ventral interarea over one-fifth as long as the pedicle valve; anacline dorsal interarea over one-eighth as long as dorsal valve; delthyrium and notothyrium open ; radial ornamentation commonly of five or six costellae per mm, 2 mm anteromedially of the dorsal ubmo, arising by dichotomous branching, externals rarely developed, internals common ; growth lines lamellose with occasional large disruptive growth lines. Ventral muscle scar rounded subpentagonal and extending anteriorly for 31 % the length of the valve and 34% as wide as the valve is long; wide adductor scar not surrounded anteriorly by diductor impressions; teeth small and supported by very short, delicate dental lamellae extending anteriorly for 19% the length of the valve. Dorsal interior with well-developed ridge posteriorly; wide ridge-like cardinal process contin- uous with median ridge and posteriorly crenulated, filling notothyrium, projecting slightly beyond hinge line; divergent, short brachiophores extending anteriorly for 16% the length of the brachial valve and with convergent bases which curve laterally away from median plane, for 37 % the length of the valve, parallel with hinge line to define small deep sockets; adductor scars very weakly impressed, situated on either side of median ridge, immediately anterior of the brachio- phores and extending anteriorly for 48 % thelength of the valve. DIMENSIONS. length width BB 72348, holotype, pedicle valve (Fig. 112) 5-0 5-8 holotype, brachial valve 4-3 5-8 BB 72349, pedicle internal mould (Fig. 113) 5-8 8-2 BB 72350, pedicle internal mould (Fig. 120) 3-2 4-5 BB 72352, pedicle internal mould (Fig. 116) 6-5 6-6 BB 72355, brachial external and internal mould (Fig. 129) 3-8 4-5 BB 72363, brachial external and internal mould (Fig. 130) 5-9 7-0 BB 72359, brachial external and internal mould (Fig. 121) 7-6 8-6 LOCALITIES. BB 72348-53 and BB 72359-62 from loc. 25, colln WFT7; BB 72354 from loc. 25, colln WFT9; BB 72355-8 and BB 72363 from loc. 28, colln Mil. Also recorded from Iocs 22 and 33. DISCUSSION. Four samples of Rhactorthis have been assessed statistically, and the data set out in Tables 26-35. Rhactorthis from Iocs 25 (A in tables) and 28 (B in tables) in the O. reuschi Zone of the Crosspipes Member do not differ in their bivariate characteristics and consequently are treated as representing a single new species, R. grandis. 240 J. M. HURST R. actoniae differs from R. grandis from loc. 25 in having a slower increase in depth of the pedicle valve (/><0-01) and in a relatively narrower ventral muscle field (/*<0-05). R. actoniae differs from R. grandis from loc. 28 in having a relatively narrower ventral muscle field (P< 0-01) and a slower increase in length of the dorsal muscle field (P< 0-05). Further, the pedicle valves of R. actoniae are moderately evenly convex whereas those of R. grandis are often carinate, resembling R. melmerbiensis (Reed) from the Dufton shales (Reed 1910 : pi. 23), but not as deep (Williams 1963 : 375). The two species further differ in the configuration of the dorsal cardinalia, with R. melmerbiensis apparently possessing shorter brachiophores whose bases do not splay to run parallel to the hinge line. R. crassa Williams, the type species of the genus, from the Gelli-grin Calcareous Ashes (Long- villian), differs from R. grandis in having relatively wider dorsal cardinalia (P< 0-001 with loc. 25 and P< 0-01 with loc. 28 R. grandis). Further, R. grandis has a carinate pedicle valve. R. grandis has fine radial ornamentation with counts of 5, 6 and 7 ribs per mm, 2 mm anteromedially of the dorsal umbones of 8, 8 and 1 specimens respectively, figures which are comparable to the costae counts derived from the published photographs of R. crassa (Williams 1963 : pi. 4) but not consistent with the species description (see Williams 1963 : 374). The brachial valve of R. grandis has 3a arising earlier than 1 a and earlier than 2a in 4 out of 5 and 5 out of 5 specimens respectively ; these proportions are significantly different from those in R. crassa (P<0-2 and P<0-02 respectively). R. cf. crassa differs from R. grandis in the faster relative increase in length of the dorsal cardina- lia (P< 0-05-0-02) and length of the adductor scar (/><0-05), relatively shallow pedicle valve (/*<0-05) and relatively longer pedicle muscle field (/><0-02). From R. actoniae, R. cf. crassa differs in having a relatively longer brachial valve (/><0-05) and relatively shorter dorsal muscle field (/><0-05). Further, the ribbing of R. cf. crassa is fine, having 5, 6, 7 and 8 costellae per mm, 2 mm anteromedially of the dorsal umbo on 2, 1, 1 and 1 shells respectively. This distribution is significantly finer than in R. grandis (/*<0-03), and in R. actoniae (/><0-05) which has 4 costellae at the appropriate distance from the dorsal umbo on 4 shells. R. cf. crassa differs from R. crassa only in having a significantly (/*<0-02) faster increase in length of the ventral muscle field, This difference is not sufficient to warrant the erection of a new species, so the specimens of Rhactorthis from the Actonian siltstones of Iocs 43 and 45 are referred to R. cf. crassa. Subfamily PLECTORTHINAE Schuchert & Le Vene, 1929 Genus GEL1DORTHIS Havlicek, 1971 Gelidorthis sp. Fig. 143 A complete internal and external mould of a very small pedicle valve (BB 72373) from loc. 48, colln ASH I , is provisionally assigned to Gelidorthis. The valve, which was 2-3 mm long and 3-2 mm Figs 145-180 Skenidioides cf. costatus Cooper, internal moulds of pedicle valves: 145 = BB 72388, x2-5, 146 = BB 72391, x4-6, 147 = BB 72390, x5-4, 148 = BB 72393, x4, 149 = BB 72389, x3, 150 = BB 72392, x4-3; internal and external moulds of brachial valves: 151 = BB 72384a, X3-5, 152 = BB72384b, x3-5, 153 = BB 72383a, x4, 154 = BB 72383, x4, 155 = BB 72386, x5, 156 = BB72385b, x5-7, 157 = BB72385a, x5-7, 158 = BB72387b, x3-4, 159 = BB72387a, x3-4. ? Draboviasp., internal mould of brachial valve: 160 = BB 72372, x 1-7. Destombesium sp., internal mould of brachial valve: 161 = BB 72375, x2. Dalmanella multiplicata multiplicata (Bancroft), internal and external moulds of brachial valves: 162 = lectotype BB 9379, xl-5, 163 = BB72308a, xl-4, 164 = BB 72308b, xl-5, 165 = BB 72296b, xl-8, 166 = BB 72296a, x2, 167 - BB 72297, x2, 168 = BB 72295a, x2-l, 169 = BB 72295b, x2, 170 = BB 72309, x2, 178 - BB 72304, xl-3, 179 = BB 72305a, x2, 180= BB 72305b, xl-3; internal moulds of pedicle valves: 171 = BB 72292, x2, 172=6872293, xl-8, 173 = 6872307, x3, 174 = BB 72294, x2, 175 = BB 72306, x2-7, 176 = BB 72303, xl-8, 177 = 88 72302, xl-3. CARADOC STRATIGRAPHY AND BRACHIOPODS 241 145 147 148 150 146 m - 151 153 W FT- 155 156 160 161 154 167 168 169 m Sj^^to msjjir 177 180 242 J. M. HURST wide, was evenly convex and very shallow. Radial ornamentation of relatively coarse angular ribs; obscure ventral muscle field; dental plates very short and subparallel. Plectorthid gen. et sp. indet. Fig. 144 A complete pedicle valve (BB 72374) from loc. 44, colln OA32, is a plectorthid. The subcircular valve was 6-8 mm long, 10-0 mm wide and 2-3 mm deep. The external surface was ornamented by branching hollow costellae. The teeth were small and supported by short, widely-divergent dental lamellae extending anteriorly for 12% of the length of the valve; muscle scar rounded and sub- pentagonal, extending anteriorly for 21 % of the length of the valve and 30% as wide as the valve is long. Family SKENIDIIDAE Kozbwski, 1929 Genus SKENIDIOIDES Schuchert & Cooper, 1931 Skenidioides cf. costatus Cooper Figs 145-159 cf. 1956 Skenidioides cf. costatus Cooper : 493-494; pi. 97, figs 33-48. cf. 1962 Skenidioides cf. costatus Cooper; Williams : 126-127; pi. 1 1, figs 24-27, 52. cf. 1963 Skenidioides cf. costatus Cooper; Williams : 375-377 ; pi. 4, figs 7-14. cf. 1974 Skenidioides cf. costatus Cooper; Williams: 82-83; pi. 13, figs 14-16; pi. 14, figs 1-3. DESCRIPTION. Ventribiconvex, oval Skenidioides with a pedicle valve 74% as long as wide and 26% as deep as long; brachial valve very gently convex, 70% as long as wide and with a sharp median sulcus originating at umbo; radial ornamentation costellate with 7 or 8 rounded costae occupying each lateral slope and between 3 and 5 costae per mm, 2 mm anteromedially of the dorsal umbo; internal branching common, external absent. Ventral interior with free spondylium 26% as deep as the length of the valve. Dorsal interior with thin ridge-like shaft of cardinal process continuous posteriorly with thin high median septum which is 81 % as long as the valve; brachiophore bases converging onto dorsal septum and extending anteriorly for 37 % of the valve length and 52 % as wide as the valve is long; adductor muscle field impressed about median septum and extending anteriorly for nearly three-quarters of the valve length. MATERIAL AND LOCALITIES. Loc. 27, colln M2 (e.g. Fig. 147, BB 72390, length 3-8 mm, width 5-0 mm). Loc. 43, colln CP2 (e.g. Fig. 145, BB 72388, length 6-2 mm, width 8-0 mm and Fig. 149, BB 72389). Loc. 44, colln OA29 (e.g. Figs 151-2, BB 72384, length 3-9 mm, width 5-8 mm, and Figs 156-7, BB 72385, length 2-5 mm, width 3-5 mm). Loc. 45, colln AS4 (e.g. Figs 153-4, BB 72383, length 4-0 mm, width 5-2 mm) and colln OA30 (e.g. Fig. 155, BB 72386). Loc. 48, colln ASH1 (e.g. Fig. 148, BB 72393, and Fig. 150, BB 72392, length 3-5 mm, width 5-4 mm) and colln ASH2 (e.g. Figs 158-9, BB 72387, and Fig. 146, BB 72391). DISCUSSION. Skenidioides occurs sporadically throughout the Actonian. Two small samples, one from calcareous sandstones (loc. 48) and the other from siltstones (Iocs 43-5), have been statisti- cally analysed (see Tables 36-42). They do not differ from each other in bivariate characteristics, or from 51. cf. costatus (see Williams 1963) found in the Gelli-grin Calcareous Ashes of north Wales. The sample from the siltstone has similar ribbing to S. cf. costatus from the Gelli-grin Calcareous Ashes but it may be slightly finer than in the Skenidioides from the calcareous sand- stones, in which one dorsal valve had only 3 ribs at the fixed distance from the umbo as opposed to 4 or 5. However, this possible difference is not considered important enough to warrant taxonomic recognition and so all Skenidioides from the Actonian are recorded as S. cf. costatus. CARADOC STRATIGRAPHY AND BRACHIOPODS 243 Superfamily ENTELETACEA Waagen, 1884 Family SCHIZOPHORHDAE Schuchert & Le Vene, 1929 Subfamily DRABOVIINAE Havlicek, 1950 Genus DESTOMBESIUM Havlicek, 1970 Destombesium sp. Fig. 161 An internal mould of a brachial valve (BB 72375) from loc. 25, colln WFT8 is provisionally assigned to Destombesium. The subcircular valve was 6-3 mm long and 7 mm wide, with a well- developed median sulcus, and a short, planar, anacline interarea. Valve interior with very thin undifferentiated cardinal process filling narrow notothyrium; brachiophores thin, subparallel and extending anteriorly for 16 % of the valve length and splayed laterally for 25 % of the valve length; brachiophore bases convergent; fulcral plates present; adductor muscle scars not impressed. Although the single specimen cannot be statistically compared with any described species of the genus, it is similar to D. zagoraensis from the Ashgill of Morocco (see Havlicek 1971 : pi. 15). Genus DRABOVIA Havlicek, 1950 ? Drabovia sp. Fig. 160 The internal mould (BB 72372) of a brachial valve about 8 mm long and 10 mm wide from the Acton Scott Beds (loc. 46) is questionably assigned to Drabovia. The valve is subcircular in outline and about one-quarter as deep as long and with a shallow median sulcus. Interarea short, slightly curved, anacline, cardinal process with small, bulbous, crenulated myophore and long thin shaft; brachiophores short with convergent bases and about one-fifth as long as the valve and splaying laterally for one-quarter the valve width, sockets defined by fulcral plates; muscle scar unknown. Family DALMANELLIDAE Schuchert, 1913 Genus DALMANELLA Hall & Clarke, 1892 Dalmanella multiplicata multiplicata (Bancroft) Figs 162-185 1928a Wattsella multiplicata Bancroft : 58-59; pi. 2, figs 1 1-14. DESCRIPTION. Typically rounded ventribiconvex Dalmanella, with an evenly, weakly convex brachial valve 80% as long as wide and with a very shallow median sulcus; pedicle valve 85 % as long as wide and 26% as deep as the valve is long; delthyrium and notothyrium open, pedicle callist inconspicuous, ventral interarea curved apscaline, longer than anacline dorsal interarea; radial ornamentation of costellae commonly 4 to 6 per mm, 2 mm anteromedially of the dorsal umbo; rib branching complex with both external and internal secondaries freely developed except on the first primary which only has internals. Ventral muscle field bilobed, with diductor scars extending anteriorly for 32% of the length of the pedicle valve but not surrounding the median adductor muscle field; diductor scars 32% as wide as the pedicle valve is long; teeth small, with shallow crural fossettes variably developed, and supported by dental plates, extending anteriorly for 21 % of the length of the pedicle valve, along the sides of the diductor muscle area as a pair of strong, slightly convergent ridges. Dorsal interior with cardinal process consisting of thin linear shaft with small, expanded, rounded myophore which is crenulated; brachiophores short, strong, divergent, with parallel to subparallel bases flanking well-developed notothyrial platform and extending anteriorly for 20% of the length of the brachial valve and splaying laterally for 31 % of the valve length; elongately suboval adductor scars poorly differentiated, often obscured by ribbing, situated on either side of low median ridge and extending anteriorly for 50% of the length of the brachial valve. 244 J. M. HURST MATERIAL AND LOCALITIES. Lectotype (selected Cocks 1978 : 62) Fig. 162, BB 9379, length 16-0 mm, width 20-7 mm; also Figs 165-9 and 171, 172, 174, BB 72292-7, all from loc. 18, new material colln x. Figs 176-180, BB 72302-5, from loc. 19, colln CL3. Figs 181-5, BB 723'lO-3, from loc. 27, colln Ml. Figs 163-4, BB 72308, length 13-9 mm, width 16-0 mm; Fig. 170, BB 72309, length 5-8 mm, width 6-3 mm; Fig. 173, BB 72307, length 3-8 mm, width 4-5 mm and Fig. 175, BB 72306, length 5-8 mm, width 7-0 mm; all from loc. 15, colln W8. Also recorded from Iocs 10-14, 16, 17, 21-3, 26 and 32, all in the Cheney Longville Formation. DISCUSSION. See p. 248. Dalmanella multiplicata (Bancroft) prima subsp. nov. Figs 186-190 DESCRIPTION. Like Dalmanella multiplicata but with a brachial valve 75% as long as wide; divergent brachiophores with subparallel brachiophore bases 22% as long as the brachial valve and splaying laterally for 36% the length of the valve; dorsal muscle field 53% as long as the brachial valve; pedicle valve 86% as long as wide and 30% as deep as long; dental plates extend anteriorly for 21 % of the length of the pedicle valve; ventral muscle field 28% as long as the pedicle valve and 34% as wide as the valve is long; radial ornamentation of costellae commonly 5 per mm, 2 mm anteromedially of the dorsal umbo. DIMENSIONS. length width Internal mould of pedicle valve, BB 72298, Fig. 190, holotype 3-6 4-3 Internal mould of pedicle valve, BB 72299, Fig. 189 3-5 4-2 Internal mould of brachial valve, BB 72300, Fig. 188 1-9 3-0 External and internal mould of brachial valve, BB 72301a b, Figs 186-7 3-1 4-8 HORIZON AND LOCALITIES. Type material from loc. 3 (colln O60). Also recorded from Iocs 1, 2, 4-8, all in the Alternata Limestone Formation. DISCUSSION. See p. 248. Dalmanella unguis unguis (J. de C. Sowerby) Figs 200-207 1839 Terebratula unguis J. de C. Sowerby in Murchison : 640; pi. 21, fig. 13. 1871 Orthis unguis (J. de C. Sowerby) Davidson : 257; pi. 37, figs 16-22. 1945 Wattsella unguis (J. de C. Sowerby) Bancroft : 196-197; pi. 23, figs 8-11; pi. 24, figs 6-7. 1958 Dalmanella unguis (J. de C. Sowerby) Dean : pi. 25, fig. 4. Figs 181-218 Dalmanella multiplicata multiplicata (Bancroft), internal moulds of pedicle valves: 181 = BB 72313, x2-5, 182 = BB 72312, x2-3; internal and external moulds of brachial valves: 183 = BB72310a, x2-4, 184 = BB72310b, x2-2, 185 = BB 72311, x2-5. Dalmanella multi- plicataprima subsp. nov., internal and external moulds of brachial valves: 186 = BB72301b, >'4-8, 187 = BB72301a, x3-8, 188 = BB 72300, x4-3; internal moulds of pedicle valves: 189 = BB 72299, x4-5, 190 = holotype BB 72298, x4-8. Dalmanella wattsi (Bancroft), internal moulds of pedicle valves: 191 = BB 72316, xl-5, 192 = BB 72315, xl-5, 193 = BB 72314, x 1 ; internal and external moulds of brachial valves: 194 = BB72317b, xl-2, 195 = BB72317a, xl-2, 196 = BB72318a, xl-4, 197 = BB72318b, xl-1, 198 = BB 72336a, xl-1, 199 = BB 72336b, xl-3. Dalmanella unguis unguis (J. de C. Sowerby), internal and external moulds of brachial valves: 200 = BB 72322, x2, 201 = BB 72321, x2, 202 = BB 72323a, xl-6, 203 = BB 72323b, xl-4; internal moulds of pedicle valves: 204 = BB72319, xl-3, 205 = BB 72320, x 2-3, 206 = BB 72324, xl-3, 207 = BB 72325, x 1-9. Dalmanella unguis ultima subsp. nov., internal moulds of pedicle valves: 208 = BB 72334, x3-5, 209 = BB 72328, x2-4, 210 = BB 72327, x 3, 217 = holotype BB 72326, x4, 218 = BB 72335, x2; internal and external moulds of brachial valves: 211 = BB 72332, x 5-2, 212 = BB 72331, x 3-2, 213 = BB 72329, x 2, 214 = BB 72330, x2, 215 = BB 72333a, x 1-7, 216 = BB72333b, xl-7. CARADOC STRATIGRAPHY AND BRACHIOPODS 245 186 195 200 205 182 f \ 183 184 187 188 189 196 198 201 202 185 190 199 203 204 206 208 209 210 207 217 218 211 212 213 214 215 216 246 J. M. HURST DESCRIPTION. Large ventribiconvex Dalmanella with evenly convex brachial valve 81 % as long as wide and with an obscure median sulcus; pedicle valve 87% as long as wide and 32% as deep as the valve is long; delthyrium and notothyrium open, pedicle callist obscure, ventral interarea moderately high, curved apscaline, longer than anacline dorsal interarea; radial ornamentation of very coarse plicae, commonly 2 or 3 ribs per mm, 2 mm anteromedially of the dorsal umbo; branching fairly simple with externals rarely developed. Ventral muscle field bilobed with diductor muscles extending anteriorly for 39% of the length of the pedicle valve and appearing not to surround the median adductor muscle field; diductor muscle field 30% as wide as the pedicle valve is long; teeth small, stubby, with variably developed crural fossettes and supported by strong dental plates extending anteriorly for 22 % the length of the pedicle valve and extending forward as strong convergent ridges bounding the diductor scars. Dorsal interior with cardinal process consisting of a short, moderately thick linear shaft with expanded, rounded, crenulated single or bilobed myophore; brachiophores short, strong, divergent, with divergent bases flanking poorly-developed notothyrial platform and extending anteriorly for 20% of the length of the brachial valve and splaying laterally for 29 % of the valve length; weakly impressed, elongately oval adductor scars poorly differentiated and separated posteriorly by a very low rounded median ridge and extending anteriorly for 49 % of the length of the brachial valve; interior of valve often strongly ribbed and obscuring dorsal adductor scar. MATERIAL AND LOCALITIES. Lectotype (selected Cocks 1978 : 63) GSM Geol. Soc. Colin 6864, from Cheney Longville Flags, 'Horderley' (exact locality unknown). Recorded here from the Crosspipes Member, loc. 27, colln M5 (e.g. Fig. 201, BB 72321, length 7-2 mm, width 8-8 mm; Fig. 204, BB 72319, length 13-2 mm, width 14-9 mm; Fig. 205, BB 72320, length 8-9 mm, width 9-4 mm) and loc. 24, colln WFT1 (e.g. Fig. 200, BB 72322; Figs 202-3, BB 72323; Figs 206-7, BB 72324-5). Also from Iocs 2 1 , 29, 30 and 33. DISCUSSION. See p. 248. Dalmanella unguis (J. de C. Sowerby) ultima subsp. nov. Figs 208-218 DESCRIPTION. Like Dalmanella unguis but with a brachial valve 78% as long as wide; divergent brachiophores with slightly divergent brachiophore bases extending anteriorly for 33 % of the length of the brachial valve and splaying laterally for 22% of length of valve; dorsal muscle field 54% as long as the brachial valve; pedicle valve 81 % as long as wide and 26% as deep as long; dental plates extend anteriorly for 23% of the length of the pedicle valve; ventral muscle scar 35 % as long as the pedicle valve and 40% as wide as the valve is long; radial ornamentation of costellae commonly 4 or 5 per mm, 2 mm anteromedially of the dorsal umbo. DIMENSIONS. length width Internal mould of pedicle valve, BB 72326, holotype 3-1 3-9 Internal mould of pedicle valve, BB 72327 3-7 4-2 Internal mould of pedicle valve, BB 72328 3-5 5-0 Internal mould of brachial valve, BB 72329 5-8 6-6 Internal mould of brachial valve, BB 72330 5-8 6-5 Internal mould of brachial valve, BB 72331 3-0 3-7 Internal mould of brachial valve, BB 72332 1-7 2-2 External and internal moulds of brachial valve, BB72333a, b 5-4 7-0 Internal mould of pedicle valve, BB 72334 2-8 3-6 Internal mould of pedicle valve, BB 72335 5-3 6-8 HORIZON AND LOCALITIES. BB 72326-31 from loc. 47, colln CF1. BB 72332-5 from loc. 27, colln M8. Also recorded from ioc. 21. • DISCUSSION. See p. 248. CARADOC STRATIGRAPHY AND BRACHIOPODS 247 ONNY VALLEY MARSHBROOK SPECIES Z 4 < Acton _CF1 Z 0 Scott ^/\A/vA/\/^ 1— Fm. A A /\A/\* A/ u V V D. unguis ultima Z /O89 — M26 *: o Crosspipes —088 — M9 0 oc CO I CO Member O87 — WFT1 /v\ o ~M6 D. unguis — M3 D~ unguis Iwattsi — Ml 5 W5 — CL3 W6 — W8 Z D. multiplicata multiplicata Z o Glynboro —017 CO — X 6 Member o D. multiplicata prima WFT14 Alternate Limestone — O60 Fig. 219 The stratigraphical location of the Dalmanella samples used in the bivariate analysis in the Onny Valley and Marshbrook sections, and the range of the Dalmanella species and subspecies. Dalmanella watt si (Bancroft) Figs 191-199 1928o Wattsella wattsi Bancroft : 58; pi. 1, figs 1-5. 1945 Wattsella wattsi Bancroft : 195-196; pi. 23, fig. 7; pi. 24, figs 2-5. 1958 Dalmanella wattsi (Bancroft) Dean : pi. 25, figs 2-3. 1963 Dalmanella wattsi (Bancroft); Williams & Wright : pi. 1, figs 2, 6, 9; pi. 2, fig. 3. DESCRIPTION. Large ventribiconvex Dalmanella with an evenly, feebly convex brachial valve 82% as long as wide and with a shallow, often obscure median sulcus; pedicle valve 85% as long as wide and 23% as deep as the valve is long; delthyrium and notothyrium open, pedicle callist inconspicuous, ventral interarea moderately high, curved apscaline, longer than anacline dorsal interarea; radial ornamentation of coarse costellae, 3 or 4 per mm, 2 mm anteromedially of the dorsal umbo; branching complex, both externals and internals freely developed except on the first primary, which only has internals. Ventral muscle field bilobed, with diductor muscles extending anteriorly for 38% of the length of the pedicle valve but not surrounding the median elongately oval adductor muscle field; 248 J. M. HURST diductor muscle field 28 % as wide as the pedicle valve is long; teeth small with variably developed shallow crural fossettes, and supported by strong dental plates extending anteriorly for 18 % of the length of the pedicle valve and extending forward as strong, slightly convergent ridges bounding the diductor scars. Dorsal interior with cardinal process consisting of strong, rounded shaft with expanded single or bilobed crenulated myophore extending slightly beyond the hinge; brachiophores very short, strongly divergent, with parallel bases flanking very well developed notothyrial platform and extending anteriorly for 18 % of the length of the brachial valve, and splaying laterally for 30% of the valve length; weakly-impressed elongately oval adductor scars poorly differentiated, situated on either side of low, rounded median ridge and extending anteriorly for 45 % of the length of the brachial valve. MATERIAL AND LOCALITIES. Lectotype (selected Cocks 1978 : 63) BB 73833, from loc. 27, from which came colln M3, including Fig. 191, BB 72316, length 15-5 mm, width 19-4 mm; Fig. 192, BB 72315, length 17-0 mm, width 19-0 mm; Fig 193, BB 72314, length 16-8 mm, width 22-0 mm; Figs 194-5, BB 72317, length 14-0 mm, width 17-0 mm; Figs 196-7, BB 72318, length 13-2 mm, width 15-2 mm; Figs 198-9, BB 72336, length 14-5 mm, width 18-4 mm. Other specimens from loc. 31, also in the Crosspipes Member. DISCUSSION. Dalmanella occurs sporadically throughout the Alternata Limestone Formation and Glynboro Member, but is particularly abundant in the uppermost part of the Glynboro Member and the Crosspipes Member. The calcareous sandstone Henley Member within the Acton Scott Formation at Acton Scott also yields a Dalmanella. Fig. 219 shows the stratigraphical positions, in two separate sections in the Caradoc district, of the samples used to derive the statistics given in Tables 43-53. The overwhelming abundance of Dalmanella populations precludes a bed-by-bed bivariate analysis. The sampling interval chosen for statistical analysis of the populations was primarily dictated by outcrop availability, but where sedimentary facies changes occurred, a close sampling interval was employed. A qualitative assessment of Dalmanella populations was made between each measured collection. Only one measurable population of Dalmanella multiplicata prima subsp. nov. (colln O60, loc. 63) was available from the Alternata Limestone, and this differed significantly from the stratigraphically younger colln WFT14 (loc. 14), in having a brachial valve which increased in width faster (P< 0-01) and in its longer ventral muscle scar (P< 0-05). Further, the O60 population differs significantly from the topotype D. multiplicata multiplicata (Bancroft) population (colln x; loc. 18), in having a brachial valve which increased in width faster (P< 0-01) and in faster growth in relative depth of the pedicle valve (/><0-02). In other respects, including ribbing (Table 53), the new subspecies closely resembles D. multiplicata multiplicata. In view of these morphological differences only a new subspecies has been recognized, so as to emphasize the close relationship with D. multiplicata multiplicata. The new subspecies appears to be limited to the Alternata Limestone. The upper Glynboro Member and the lower part of the Crosspipes Member (up to and including the Dalmanella wattsi Zone of Bancroft 19290, £,1933) contain abundant Dalmanella populations with much morphological variation. The D. multiplicata multiplicata topotype population (colln x) differs significantly from the W8 population in having relatively narrower dorsal cardina- lia (/*<0-05) and faster increase in relative length of the ventral muscle field. The W8 population does not differ from the W6 population with respect to any character, and in turn the W6 popula- tion does not differ from the W5 population nor from the Ml population (see Fig. 219). Dalmanella population CL3 does not differ in any character from population W5, and only differs from the older population O17 in having relatively wider dorsal cardinalia (/><0-05). The O17 population does not differ in any character from the WFT14 population, which in turn differs solely from the topotype population of D. multiplicata multiplicata in having relatively wider dorsal cardinalia (P< 0-05). In view of the non-directional but persistent morphological variation expressed by Dalmanella populations in the Glynboro and lower Crosspipes Members (to the top of the D. wattsi Zone of Bancroft 1929a, b, 1933), in both sections, it is proposed to include all the populations present in CARADOC STRATIGRAPHY AND BRACHIOPODS 249 the one subspecies, D. multiplicata multiplicata (Bancroft). This is supported by the limited data on ribbing (Table 53), which show that there is no significant difference in ribbing densities of the D. multiplicata topotype population and the younger population (Ml) of the same species. Bancroft (1928#, 1945) gave the type locality for Dalmanella wattsi, in the old quarry in Marsh- wood (loc. 27), in the Heterorthina praeculta beds, 3 in (76 mm) below the base of the 'Wattselld1 unguis beds. The topotype population (collection M3) differs significantly from the stratigraphi- cally older M 1 population on four counts, the relatively wider brachial valve (P< 0-05), the slower increase in length of the dorsal cardinalia (/*<0-05), the relatively longer pedicle valve (P<0-01), the relatively wider pedicle valve (/><0-05), the slower growth in relative depth of the pedicle valve (P< 0-05) and the wider ventral muscle scar (P< 0-05). D. unguis (colls M5 and M6 pooled) has significantly coarser ribbing (P< 0-05). From the topotype population of D. multiplicata multiplicata, D. unguis differs in having significantly coarser ribbing (P< 0-001), faster increase in length of the dorsal cardinalia (P< 0-02) and slower growth in relative depth of the pedicle valve (P< 0-001). The Dalmanella populations present in the upper part (D. unguis and O. reuschi Zones of Bancroft 19290, b) of the Crosspipes Member also display a great deal of morphological varia- bility. Population M5, which occurs stratigraphically above the topotype D. unguis population, differs in having relatively longer dorsal cardinalia (P< 0-01) and faster growth in relative depth of the pedicle valve. There is no ribbing difference. There is no significant bivariate character difference between Dalmanella populations M5-M8, M8-M9 and M9-M26. Population WFTI differs from the topotype of D. unguis in having relatively shorter dorsal cardinalia (/)<0-01) and in the faster increase in length of the dorsal adductor muscle field. Further, the WFTI population does not differ, as regards bivariate charac- ters, from the stratigraphically younger O87-89 population. However, this latter population differs from the youngest population in the Marshbrook section (M26) in the faster increase in relative width of the brachial valve (P< 0-01). Thus the Dalmanella populations of the Upper Cheney Longville Formation show no consis- tent morphological variation. However, the populations in the upper part of the D. unguis Zone, in both sections, show a consistent difference from the older populations, in having finer ribs (Table 53). The M6, M5 and WFTI populations do not differ significantly as regards coarseness of ribbing. However, the younger populations of M8-M9 (pooled data) differ from M6, M5 and WFTI (all at P< 0-001), whilst O87-89 differs from the same collections (/»< 0-01). This consistent morphological difference is afforded taxonomic recognition, with the erection of the subspecies D. unguis ultima to cover forms with a modal rib count of 4 or 5 per mm, at 2 mm anteromedially from the dorsal umbo, as opposed to the 2 or 3 of D. unguis unguis. In calcareous bioturbated sandstones of the stratigraphically younger Acton Scott Formation D. unguis ultima briefly reappears (loc. 47). The population CF1 does not differ from the previous populations of the subspecies, in either bivariate statistics or ribbing. Dean (1958 : 205) reported sympatry between D. wattsi and D. unguis, at least in the Marsh- brook area. This is contrary to the findings here. Large collections of basal D. unguis Zone specimens show a wide range of continuous morphological variation especially with respect to 250 J. M. HURST ribbing, which is the easiest qualitative character by which to distinguish the two species. Pre- sumably Dean would place the finer-ribbed specimens in D. wattsi, but this would involve splitting a continuous morphological series into two categories and would be unjustified. Genus BANCROFTINA Sinclair, 1946 1933 Raymondella Bancroft : 3 (nomen nudurri). 1938 Raymondella Bancroft; Whittington : 249 (non Reed 1935 : 8). 1945 Raymondella Bancroft : 197. 1946 Bancroftina Sinclair : 295. DIAGNOSIS (emended). Subcircular to subquadrate ventribiconvex shells with shallow sulcate brachial valve ; radial ornamentation finely costellate with first order costellae branching internally, more commonly than externally, in the first four sectors. Ventral interarea moderately long, curved and apscaline with open delthyrium; dorsal interarea short, weekly curved, anacline, notothyrium open and commonly filled with cardinal process; shell punctate. Ventral interior with massive teeth supported by well-developed subparallel to divergent dental lamellae; muscle field subtriangular to bilobed in outline with submedian diductor lobes not enclosing median adductor scar; pedicle callist well developed; pallial sinus pattern unknown. Dorsal interior with stout bilobed or trilobed cardinal process; brachiophores very stout and short, widely divergent with their bases greatly divergent relative to their tops and often sub- parallel to hinge line; fulcral plates rarely developed; ancillary struts often strong, fused to well- developed median ridge; dorsal adductor scar elongately oval and bipartite or quadripartite. TYPE SPECIES. Raymondella typa Whittington, by original definition of Sinclair (1946 : 295). DISCUSSION. Bancroft (1933 : 3) initially proposed the name Raymondella, with R. typa as genotype, for a group of middle Caradoc dalmanellid brachiopods from the upper Horderley Sandstone, Alternata Limestone and lower Longville Flags (all Longvillian) of the type Caradoc. The diagnosis was inadequate and most subsequent workers (see Williams & Wright 1963 : 27) have treated this as a nomen nudum. However, Whittington (1938 : 249) fully described Raymondella typa, based on composite material including specimens of his own from silty ashes of the Berwyn Dome and the material Bancroft had at his disposal. Whittington attributed the genus and the species to Bancroft. Raymondella was further described by Bancroft (1945) but as the authorship of the genus has to be attributed to the first full description of Whittington (1938), the name was by that time occupied (Reed 1935). Thus Sinclair (1946, 1949) proposed the name Bancroftina to cover the Caradoc dalmanellid and designated Raymondella typa Whittington as the type species. Bancroftina hemtti sp. nov. Figs 247-250 DESCRIPTION. Large subcircular, unequally biconvex Bancroftina, with shallow sulcate brachial valve 79 % as long as wide and less than one-tenth as deep as long, and pedicle valve 80% as long as wide and 28% as deep as long; radial ornamentation moderately coarsely costellate with a modal count of 4 costellae per mm, 2 mm anterior of the dorsal umbo, external branching earlier Figs 220-246 Bancroftina whittingtoni sp. nov., internal moulds of pedicle valves: 220 = BU 241, xl-8, 221 = BB 73681, x2, 222 = BB 72469, xl-5, 223 = BB 73677, xl-7, 224 = BB 73678, xl-9, 225 = BB 72467, x2, 226 = BB 72468, xl-8; internal and external moulds of brachial valves: 227 = holotype BB 72475, x 1-8, 228 = BB 72476, x 1-9, 229 = BB 72472, x 1-9, 230 = BB 72473, xl-9, 231 = BB 72471, xl-7, 232 - BB 72474, xl-7, 233 = BB 72470, xl-9. Bancroftina typa (Whittington), internal moulds of pedicle valves: 234 = BB 72480, x2, 235 = BB 72481, x2, 244 = BB 72479, x 1-3, 245 = BB 72486, x 1-1, 246 = BB 72487, x 1-6; internal and external moulds of brachial valves: 236 = BB 72483, x2, 237 - BB 72482, x2, 238 = lectotype BU 242, xl-8, 239 = lectotype of robusta BB 10302, x2-5, 240 = BB 72484, xl-8, 241 = BB 72477, xl-9, 242 = BB 72478, xl-5, 243 = BB 72485, xl-7. CARADOC STRATIGRAPHY AND BRACHIOPODS 251 220 222 223 224 221 225 226 230 234 235 239 227 231 236 240 244 229 228 232 233 237 238 242 241 243 —~rf\^ /T *'j .0' \ 245 252 J. M. HURST than internal slightly more common in sectors I to IV; dental lamellae very strong, weakly diver- gent and extending anteriorly for 17% of the length of the valve, crural fossettes absent, ventral muscle scar extending anteriorly for 35% of the length of the valve and 30% as wide as the valve is long; heavily calcified cardinalia with divergent brachiophore bases 14% the length of the brachial valve and 27% as wide as the valve is long; ancillary struts massive, ankylosed to median ridge, bipartite dorsal adductor scar extending anteriorly for 42% the length of the valve. NAME. After Dr R. A. Hewitt. DIMENSIONS. length width Internal mould of brachial valve, BB 72488, holotype 19-3 23-0 Internal mould of pedicle valve, BB 72489 16-7 18-6 Internal mould of brachial valve, BB 72490 16-7 20-1 Internal mould of pedicle valve, BB 72491 18-3 23-6 HORIZON AND LOCALITIES. All figured specimens from loc. 32; holotype BB 72488, colln LM4, remainder colln LM6. Also recorded from Iocs 26 and 27. DISCUSSION. See p. 253. Bancroftina typa (Whittington) Figs 234-246 1933 Raymondella typa Bancroft : 3; nomen nudum. 1938 Raymondella typa Bancroft; Whittington : 249 pars; pi. 10, fig. 13, non figs 12, 14. 1945 Raymondella typa Bancroft; Bancroft : 198-201 ; pi. 25, figs 1-10. 1945 Raymondella robusta Bancroft : 201 ; pi. 25, figs 1 1, 12; pi. 26, figs 1-3. 1946 Bancroftina typa (Bancroft) Sinclair : 295. 1959 Bancroftina typa (Whittington); Cave & Dean : 294; pi. 53, figs 5, 6. 1963 Bancroftina typa (Whittington); Williams & Wright : 7; pi. 1, figs 12, 15; pi. 2, fig. 6. 1963 Bancroftina robusta (Bancroft) Williams & Wright : pi. 1, fig. 13. DESCRIPTION. Subcircular to subquadrate, unequally biconvex Bancroftina, with shallow sulcate brachial valve 79 % as long as wide and less than one-eighth as deep as long, and pedicle valve 87% as long as wide and 30% as deep as long; radial ornamentation finely costellate with a modal count of 5 costellae per mm, 2 mm anterior of dorsal umbo, external branching very rarely occurring earlier than internal in sectors I to IV; dental lamellae strongly developed, divergent and extending anteriorly for 21 % of the length of the valve, with moderately strong crural fossettes ; ventral muscle scar extending anteriorly for 35 % of the length of the valve and 32% as wide as the valve is long; strong cardinalia with widely divergent brachiophore bases 19 % the length of the brachial valve and 34% as wide as the valve is long; bipartite dorsal adductor scar extending anteriorly for 50 % the length of the valve. MATERIAL AND LOCALITIES. Lectotype (Fig. 238) BU 242 (selected Cave & Dean 1959 : 295), length 10-7 mm, width 14-7 mm, from small quarry in upper Horderley Sandstone at New House, grid ref. SO 418859, from which also came BB 72480 (Fig. 234), length 10-1 mm, width 1 1-7 mm; BB 72481 (Fig. 235), length 9-7 mm, width 10-4 mm; BB 72482 (Fig. 237), length 9-5 mm, width 10-6 mm; BB 72483 (Fig. 236), length 9-5 mm, width 13-2 mm. From loc. 14, colln WFT19, e.g. BB 72477 (Fig. 241), length 12-4 mm, width 15-2 mm; BB 72478 (Fig. 242), length 9-7 mm, width 13-0 mm; BB 72479 (Fig. 244), length 13-8 mm, width 16-4 mm. From loc. 19, e.g. BB 10302 (Fig. 239), lectotype of robusta, length 13-6 mm, width 17-6 mm. From loc. 17, colln S9, e.g. BB 72484-7 (Figs 240, 243, 245-6). Also from Iocs 1-8, 10 and 16. Bancroftina whittingtoni sp. nov. Figs 220-233 1938 Raymondella typa Bancroft; Whittington : 249 pars; pi. 10, figs 12, 14, non fig. 13. DESCRIPTION. Semicircular to subquadrate, unequally biconvex Bancroftina, with shallow sulcate CARADOC STRATIGRAPHY AND BRACHIOPODS 253 brachial valve 77% as long as wide and less than one-eighth as deep as long, and pedicle valve 81 % as long as wide and 34% as deep as long; radial ornamentation very finely costellate with modal count of 7 costellae per mm, 2 mm anterior of dorsal umbo; in sectors I to IV internal branching occurs earlier than external branching; dental lamellae strongly developed, sub- parallel to slightly divergent, and extending anteriorly for 20% of the length of the valve, ventral muscle scar extending anteriorly for 32% of the length of the valve and 31 % as wide as the valve is long; strong cardinalia with widely divergent brachiophore bases extending anteriorly for 20% the length of the brachial valve and 32% as wide as the valve is long; dorsal adductor scar extend- ing anteriorly for 53 % the length of the valve. NAME. After Professor H. B. Whittington. DIMENSIONS. length width Internal mould of brachial valve, BB 72475, holotype 12-3 13-8 Internal mould of pedicle valve, BB 73678 10-5 13-2 Internal mould of pedicle valve, BU 241 13-2 15-2 Internal mould of pedicle valve, BB 73677 10-8 12-5 Internal mould of pedicle valve, BB 72467 9-5 10-6 Internal mould of pedicle valve, BB 72468 7-5 8-0 Internal mould of pedicle valve, BB 72469 12-6 16-0 External mould of brachial valve, BB 72470 9-9 12-2 Internal mould of brachial valve, BB 72471 14-3 17-2 Internal mould of brachial valve, BB 72472 7-5 10-7 Internal mould of brachial valve, BB 72473 12-9 15-2 Internal mould of brachial valve, BB 72474 12-7 14-8 HORIZON AND LOCALITIES. This species is known only from the type beds in the Berwyn Hills (Whittington 1938). The sample used in this study comes from a disused quarry in the Bryngwyn Beds, 360 m north-east of the spot height in the ancient camp at Bryngwyn Hill, some 4-8 km south-west of Llansantffraid; SJ 188184. DISCUSSION. Bancroftina is very abundant in the upper Horderley Sandstone, Alternata Limestone and lower Cheney Longville Formation. Bancroft (1945) recognized Bancroftina typa (Whitting- ton) from the Horderley Sandstone and erected a new species B. robusta, from the lower Cheney Longville Formation. Previously, Bancroft (1933 : table adjacent to p. 4) had referred to an un- described species of Bancroftina, B. gigantea (a nomen nudum), as characterizing 165 ft (50 m) of strata, in east Shropshire (Salop), above his Kjaerina bipartita Zone (= Alternata Limestone). The type locality of B. robusta falls within his zone of B. gigantea, suggesting that he was in fact referring to the same species. The statistical data derived in Tables 54-64 are based on collections which Whittington described as B. typa from the Berwyn Hills, B. typa from the Horderley Sandstone and B. robusta. As Bancroft (1945 : 201) emphasized the importance of size in the species classification of this genus, populations with the largest and smallest individuals and a population from the Alternata Limestone and upper Cheney Longville Formation have also been statistically analysed. As regards bivariate statistics, Bancroftina from the Horderley Sandstone, Alternata Limestone and Glynboro Member, including B. typa and B. robusta topotypes, do not differ significantly. Bancroftina from the Berwyn Hills, north Powys, differs consistently from the above populations in having a slower relative increase in width of the dorsal cardinalia (P< 0-05-0-001) and relatively longer dorsal cardinalia (P< 0-05-0-01) and so a new species, B. whittingtoni, is described here. B. hewitti does not differ significantly from the older populations of Bancroftina from Salop but does differ significantly from the Berwyn Hills Bancroftina in having relatively shorter dental lamellae (P< 0-05). B. whittingtoni from the Berwyn Hills has significantly finer ribbing (see Table 64) than B. hewitti (P< 0-001) and the remaining Bancroftina populations (/)<0-01). B. hewitti differs signi- ficantly (P< 0-01) from the other Salop populations in having the coarsest costellae. 254 J. M. HURST _The earlier insertion of eight costeljae relative to eight others has been analysed (see Table 63). lb)la, 2b)2a, 3ala)2a, 3a"l)3bl and 4b)4b are constantly associated throughout the seven popula- tions of the three species. 2a!)2b is very variable, forming no consistent patterns. B. hewitti differs significantly from the other species in the earlier insertion of 3a relative to 3aTa -(P< 0-001) and 3a relative to 3c (P< 0-001). These two rib patterns are constantly associated in B. typa and B. whittingtoni. To summarize, there are three species of Bancroftina in the Caradoc of Wales and the Welsh Borderland : 1 . B. whittingtoni from the Berwyn Hills, which differs from B. typa in the configuration of the dorsal cardinalia and in having finer ribbing. From B. hewitti it differs in its finer ribbing and dental lamellae. 2. B. typa, which includes populations from the Horderley Sandstone originally assigned by Bancroft to B. robusta. 3. B. hewitti, which is characterized by its coarse ribbing and the costellae insertions 3c)3a and 3ala)3a. Further, the species has very heavily calcified dorsal cardinalia with secondary calcite deposition around the brachiophores forming large blocks. Its ancillary struts are very strong and in the pedicle valve it does not possess crural fossettes as do B. whittingtoni and B. typa. Genus CR YP TO THYRIS Bancroft, 1945 DIAGNOSIS (emended). Subcircular, unequally biconvex dalmanellid with a deeply convex pedicle valve and sulcate brachial valve; radial ornamentation multicostellate; ventral interarea moder- ately long, curved, apscaline with open delthyrium and prominent umbo; dorsal interarea short and curved, anacline, notothyrium filled by cardinal process; shell punctate. Ventral interior with massive teeth supported by strong subparallel dental plates with strong crural fossettes on their medial faces, pedicle callist well developed, muscle field subtriangular to cordate in outline, submedian diductor lobes extending beyond dental lamellae and not surround- ing anterior edge of median adductor field. Dorsal interior with stout, undifferentiated, bilobed and anteriorly crenulated cardinal process, brachiophores with massive triangular pads and with their bases slightly divergent relative to their tops, fulcral plates rarely developed; dorsal adductor scar elongate and impressed, quadri- partite about strong median ridge. DISCUSSION. Bancroft (1928«, 1945) and more recently Kemezys (1968) noted that Cryptothyris had rhipidomelloid ribbing. However, in internal features this genus belongs unequivocally in the Dalmanellidae, an assignment made in 1965 by Williams et al. TYPE SPECIES. Resserella paracyclica Bancroft, 1928. Figs 247-275 Bancroftina hewitti sp. nov., internal moulds of pedicle valves : 247 = BB 72491, x 1 -2, 248 = BB 72489, xl-2; internal moulds of brachial valves: 249 = holotype BB 72488, xl-4, 250 — BB 72490, x 1-2. Cryptothyris paracyclica (Bancroft), internal moulds of brachial valves: 251 - lectotype BB 24172, x2-5, 252 = BB 72502, xl-9, 253 = BB 73564, x2-4, 254 - BB 24184, xl-7, 255 = BB 24180, xl-6, 256 = BB 72499, x2-2, 257 = BB 72498, x2; external surface of brachial valve: 258 = BB 72500, X2-5; internal moulds of pedicle valves: 260 = BB 24220, x2,261 = BB 24221, x2,262 - BB 72501, xl-5,263 = BB24216, x2,264 = BB 24222, x 1-1, 265 = BB 72496, x 1-6, 266 = BB 72497, xl-6; external surface of pedicle valve and lecto- type of synonymized C. cyclica: 259 = BB 24218, x2-5. Onniella a velinei Bancroft, internal mould of brachial valve: 267 = lectotype, BB 10326, x2. Onniella reuschi Bancroft, internal and external moulds of brachial valves: 268 = BB 72414, xl-7, 269 = BB 72415, xl-8, 270 = BB 72412, xl-5, 271 = BB 72413, xl-4, 272 = BB 72422, x2, 273 - BB 72420, x2-5, 274 = BB 10248 (lectotype of synonymized O. aspasia), x3, 275 = BB 72423, x 1-9. CARADOC STRATIGRAPHY AND BRACHIOPODS 255 261 262 265 266 263 * V 272 273 274 275 256 J. M. HURST Cryptothyris paracyclica (Bancroft) Figs 25 1-266 19280 Ressere lla paracyclica Bancroft : 56; pi. 1, figs 6-9. 19280 Resserella cyclica Bancroft : 56; pi. 1, fig. 10. 1945 Cryptothyris paracyclica (Bancroft) Bancroft : 223. 1958 Cryptothyris paracyclica (Bancroft); Dean : pi. 25, fig. 7. 1963 Cryptothyris paracyclica (Bancroft); Williams & Wright : pi. 2, figs 9, 14, 15. DESCRIPTION (emended). Subcircular, unequally biconvex Cryptothyris, with evenly convex brachial valve 80% as long as wide, and pedicle valve 82% as long as wide and 30% as deep as long; radial ornamentation finely costellate with modal count of 4 costellae per mm, 2 mm anteromedially of the dorsal umbo, median costellae commonly developed; subparallel dental lamellae 20 % as long as pedicle valve, ventral muscle scar extending anteriorly for 36 % the length of the valve and 32% as wide as the valve is long; cardinalia massive with very slightly divergent brachiophore bases extending anteriorly for 19% the length of the valve and splaying laterally for 31 % of the valve length; dorsal adductor scar extending anteriorly for 54% of the length of the valve. MATERIAL AND LOCALITIES. Lectotype (Fig. 251) BB 24172 (selected Cocks 1978 : 64), length 9-2 mm, width 11-3 mm; lectotype of cyclica (Fig. 259) BB 24218, length 9-2 mm, width 9-6 mm; also Fig. 253, BB 73564, length 7-1 mm, width 9-4 mm; Fig. 254, BB 24184, length 10-0 mm, width 11-7 mm; Fig. 255, BB 24180, length 9-7 mm, width 11-2 mm; Fig. 257, BB 72498, length 11-1 mm, width 13-0 mm; Fig. 258, BB 72500, length 5-6 mm, width 6-9 mm; Fig. 260, BB 24220, length 8-5 mm, width 9-6 mm; Fig. 261, BB 24221, length 8-9 mm, width 11 -Omm; Fig. 263, BB 24216, length 12-0 mm, width 14-3 mm; Fig. 264, BB 24222, length 10-9 mm, width 12-0 mm; all from loc. 39. Also loc. 44, colln OA1 (e.g. Figs 256, 265-6, BB 72499, BB 72496-7), loc. 43, colln CP2 (e.g. Fig. 252, BB 72502, and Fig. 262, BB 72501). Also recorded from Iocs 40-2, 45 and 48, all from the Acton Scott Formation. DISCUSSION. Bancroft (19280, 1945) erected Resserella paracyclica and R. cyclica, which he later transferred to Cryptothyris, from the same type locality in the Acton Scott Formation of the Onny Valley, and distinguished between them by the width of the hinge line. At the type locality Cryptothyris is very rare, but from a study of the small amount of Bancroft material in the BM(NH) it is apparent that, as regards hinge width, there is continuous variation between the two species and thus C. cyclica is treated as being synonymous with C. paracyclica. At loc. 44 in the Acton Scott Formation of Marshbrook a small collection of C. paracyclica has been statistically analysed to give some idea of the variation of the species involved (see Tables 65-73). Genus ONNIELLA Bancroft, 1928 Onniella broeggeri Bancroft Figs 316-335 1928a Onniella broggeri Bancroft : 56-57; pi. 2, figs 1-5. 1945 Onniella broggeri Bancroft; Bancroft : 222-223; pi. 30, figs 16-17. Figs 276-305 Onniella reuschi Bancroft, internal moulds of brachial valves: 276 = BB 72425, x2, 277 = BB 72424, x2-4, 278 = BB 72431, xl-5, 279 = BB 72432, xl-7; internal moulds of pedicle valves: 280 = BB 72417, xl-8, 281 = BB 10239 (lectotype of synonymized O. grandis), x2, 282 = BB 72411, x2-4, 283 = BB 72408, xl-7, 284 = BB 72410, xl-8, 285 = BB72418, x2, 286 = BB 72421, x2, 287 = BB 72419, x2-2, 288 - BB 72428, xl-7, 289 - BB 72427, x2-5, 290 = BB 72426, x3, 291 = BB 72430, xl-7, 292 = BB 72429, x2-9, 293 = BB 72409, x 1-7. Onniella depressa Bancroft, conjoined valves: 294 = brachial view of BB 10254 (lectotype of synonymized O. sinuata), x 1-5, 295 = pedicle view of BB 10254, x 1-5, 296 = brachial view of lectotype BB 10251, xl-5, 297 = pedicle view of lectotype BB 10251, xl-5; internal moulds of pedicle valves, 298 = BB 72435, x4, 299 = BB 72466, x4, 300 - BB 72434, x2-5, 301 = BB 72444, X2-5, 302 = BB 72433, x2-5, 303 = BB 73679, x2-5, 304 - BB 72443, x2-5, 305 » BB 72441, x2-5. CARADOC STRATIGRAPHY AND BRACHIOPODS 257 297 298 299 301 ~*i^^W ^^Nfe I302 304 305 258 J. M. HURST 1945 Onniella i neons tans Bancroft : 221-222; pi. 31, figs 1-3. 1958 Onniella inconstans Bancroft; Dean : pi. 25, figs 8-9. 1958 Onniella broeggeri Bancroft; Dean : pi. 25, figs 10-11. 1965 Onniella broeggeri Bancroft; Cave : pi. 12, figs F, I, L, O, P. DESCRIPTION (emended). Subquadrate, unequally biconvex Onniella, with evenly convex, sulcate brachial valve 76% as long as wide, and pedicle valve 82% as long as wide and 28% as deep as long; shell with right-angular cardinal extremities; dorsal interarea short, curved, anacline, notothyrium open; ventral interarea moderately long, curved apscaline, delthyrium open; radial ornamentation commonly 6 costellae per mm, 2 mm anterior of dorsal umbo, branching freely developed. Ventral interior with massive stout teeth supported by subparallel to parallel dental lamellae, which extend anteriorly for 22% of the length of the valve, and have moderately strong to very strong crural fossettes on their inner faces; dental plates continued along sides of diductor muscle field as a pair of strong ridges; adult ventral muscle field angular subpentagonal and extending anteriorly for 34% of length of the pedicle valve, maximum width 30% of the valve length; diductor scars elongately oval and produced as short lobes in front of median lanceolate adductors, but not surrounding them. Dorsal interior with massively calcified cardinalia 21 % as long as the brachial valve and 34% as wide as the valve length; brachiophores widely divergent and with weakly divergent bases which are often masked by thick secondary shell deposition to form thick triangular pads anky- losed to median ridge by thick ancillary struts in the form of a straight bar; inner faces of brachio- phores with pits; cardinal process with thick shaft continuous with posterior end of median ridge, myophore bilobed and posteriorly crenulated, filling notothyrium and projecting slightly beyond hinge line; adductor muscle field subquadrate, raised on shallow platform and with smaller posterior elements, divided by median ridge and extending anteriorly for 52 % of length of the brachial valve. MATERIAL AND LOCALITIES. Lectotype (Figs 316-7) BB 24077 (selected Cocks 1978 : 66), length 6-8 mm, width 8-6 mm, from loc. 50; also (colln O35) Fig. 320, BB 72458, length 4-1 mm, width 4-8 mm; Fig. 321, BB 72457, length 9-9 mm, width 12-8 mm; Fig. 323, BB 72456, est. length 10-5 mm, width 15-0 mm; Fig. 324, BB 72455, length 7-4 mm, width 8-9 mm; Fig. 328, BB 72459, length 8-4 mm, width 10-4 mm; Fig. 329, BB 72463, length 5-2 mm, width 6-8 mm; Fig. 330, BB 72462, length 1 1-4 mm, width 13-8 mm; Fig. 335, BB 72460, length 9-1 mm, width 12-6 mm. Lectotype of inconstans (Figs 318-9) BB 10259, length 11-1 mm, width 12-3 mm, from loc. 49, also (colln O98) Fig. 322, BB 72448; Fig. 325, BB 72450; Fig. 326, BB 72447; Fig. 327, BB 72449; Figs 331-4, BB 72451-4, also in the Onny Shale Formation. DISCUSSION. See p. 261. Onniella depressa Bancroft Figs 294-3 15 1945 Onniella depressa Bancroft : 218-221 ; pi. 30, figs 10-13. 1945 Onniella sinuata Bancroft : 222; pi. 30, figs 14-15. DESCRIPTION (emended). Subcircular to subquadrate unequally biconvex Onniella, with evenly Figs 306-331 Onniella depressa Bancroft, internal and external moulds of brachial valves: 306 = BB 72442, x2-5, 307 = BB 72438, x2-5, 308 = BB 72465, x3, 309 = BB 72440, x3, 310 = BB 72446, x3, 311 = BB 72464, x2-5, 312 = BB 72436, x2-5, 313 = BB 72439, x2-5, 314 = BB 72445, x2-5, 315 = BB 72437, x2-5. Onniella broeggeri Bancroft, conjoined valves: 316-7 = pedicle and brachial views respectively of lectotype BB 24077, x 2-5, 318-9 = pedicle and brachial views respectively of BB 10259 (lectotype of synonymized O. inconstans), x 1-5; internal moulds of brachial valves: 320 - BB 72458, x5, 321 - BB 72457, x2-5, 322 - BB 72448, x2-5, 323 = BB 72456, x2-5, 324 = BB 72455, x3, 325 = BB 72450, x2-5, 326 = BB 72447, x2-5, 327 = BB 72449, x 2-5; internal moulds of pedicle valves: 328 = BB 72459, x 2-5, 329 = BB 72463, x4, 330 = BB 72462, x2-5, 331 - BB 72454, x2-5. 259 309 315 316 328 329 313 * 317 IP 318 310 314 319 330 331 260 J. M. HURST convex, deeply sulcate brachial valve 79 % as long as wide, and a pedicle valve 83 % as long as wide and 34% as deep as long; dorsal interarea short, slightly curved, anacline, notothyrium open ; ventral interarea short, curved, apscaline, delthyrium open ; radial ornamentation commonly 6 costellae per mm, 2 mm anterior of dorsal umbo, branching freely developed. Ventral interior with massive teeth supported by subparallel dental lamellae which extend anteriorly for 23% of the length of the valve and have weak or moderately strong crural fossettes on their inner faces; adult ventral muscle field angular cordate to subpentagonal and extending anteriorly for 35 % the length of the pedicle valve, maximum width 32 % of the valve length ; elongate oval diductor lobes do not surround median lanceolate adductor scar. Dorsal interior with massive cardinalia 18% as long as the brachial valve and 34% as wide as the valve length; brachiophores slightly divergent with weakly divergent bases, commonly expanded by secondary shell deposition to form thick triangular pads ankylosed to median ridge by ancillary struts; inner faces of brachiophore with pits; cardinal process with thick shaft posteriorly bilobed and crenulated, filling notothyrium and projecting slightly beyond hinge line; adductor scar subquadrate with smaller posterior elements, divided by median ridge and extending anteriorly for 48 % the length of the brachial valve. MATERIAL AND LOCALITIES. Lectotype (Figs 296-7) BB 10251 (selected Cocks 1978 : 66), length 12-7 mm, width 14-8 mm, from loc. 37, also (colln O24) Fig. 298, BB 72435, length 5-2 mm, width 6-8 mm; Fig. 300, BB 72434, length 7-8 mm, width 10-0 mm; Fig. 302, BB 72433, length 13-5 mm, width 16-8 mm; Fig. 307, BB 72438, length 7-1 mm, width 9-4 mm; Fig. 312, BB 72436, length 10-6 mm, est. width 11-0 mm; Fig. 315, BB 72437. Lectotype for sinuata Figs 294-5, BB 10254, length 10-7 mm, width 11-8 mm, from loc. 40. From loc. 41, colln O104, Fig. 310, BB 72446, Fig. 314, BB 72445; also colln O105, Fig. 301, BB 72444, Figs 304-6, BB 72441-3, Fig. 309, BB 72440, Fig. 313, BB 72439. From loc. 44, colln OA29, Fig. 299, BB 72466, Fig. 308, BB 72465, Fig. 311, BB 72464, also colln OA30, Fig. 303, BB 73679, length 11 -6 mm, width 14-3 mm. Also recorded from Iocs 36, 38 and 39, all in the Acton Scott Formation. DISCUSSION. See p. 261. Onniella reuschi Bancroft Figs 268-293 1928o Onniella reuschi Bancroft : 57-58; pi. 2, figs 10-13. 1945 Onniella reuschi Bancroft; Bancroft : 215-217; pi. 29, figs 1-6; pi. 30, figs 1-3. 1945 Onniella grandis Bancroft : 217; pi. 29, figs 7, 8, 14; pi. 30, figs 4-6. 1945 Onniella aspasia Bancroft : 217-218; pi. 29, figs 10-12; pi. 30, figs 7-9. 1958 Onniella reuschi Bancroft; Dean : pi. 25, figs 5-6. 1963 Onniella grandis Bancroft; Williams & Wright : pi. 1, figs 1, 5, 8; pi. 2, fig. 5. DESCRIPTION (emended). Large, subcircular unequally biconvex Onniella, with evenly convex weakly sulcate brachial valve 74% as long as wide and about one-ninth as deep as long, and a pedicle valve 80% as long as wide and 22% as deep as long; dorsal interarea short, straight, anacline, notothyrium open; ventral interarea short, curved, apscaline, delthyrium open; radial ornamentation commonly 7 costellae per mm, 2 mm anterior of dorsal umbo, internal branching predominantly developed in sectors I to IV. Ventral interior with strong, massive teeth supported by divergent dental plates which extend anteriorly for 19% of the length of the valve and have very weak crural fossettes on their inner faces; adult ventral muscle field rounded cordate and extending anteriorly for 34% the length of the pedicle valve, maximum width 31 % of the valve length; oval diductor lobes do not surround median lanceolate adductor scar. Dorsal interior with massive subparallel brachiophores which extend anteriorly for 17% of the valve length and splay laterally for 31 % of the valve length; brachiophore secondarily thickened and with divergent bases; ancillary struts strong; cardinal process consisting of thick shaft posteriorly continuous with median ridge and with massive, ovoid, posteriorly crenulated, bilobed myophore, completely filling notothyrium and projecting slightly posterior of hinge line; thick CARADOC STRATIGRAPHY AND BRACHIOPODS 261 median ridge divides poorly-impressed adductor scars which extend anteriorly for 47% of the length of the valve. MATERIAL AND LOCALITIES. Lectotype BB 10236 (selected Cocks 1978 : 66), length 8-2 mm, width 10-9 mm, from loc. 25, as is Fig. 283, BB 72408, length 12-7 mm, width 17-6 mm (colln WFT7). Also loc. 28, collns M 10 and Mil, including Figs 268-271, BB 72412-5, Figs 282, 284, 293, BB 72409-11. Loc. 46, Fig. 281, BB 10239 (lectotype of grandis, length 13-8 mm, width 17-0 mm) also (colln AS5) Figs 272, 273, 280, 285-7, BB 72417-22. Loc. 35, Fig. 274, BB 10248 (lectotype of aspasia, length 13-8 mm, width 17-0 mm) and colln DH1, Figs 275-7, 288-290, BB 72423-8. Loc. 47, colln CF4, Figs 278-9, 291-2, BB 72429-32. Also recorded from Iocs 22, 33-5, 43-5. DISCUSSION. Onniella is abundant in the uppermost Crosspipes Member, Acton Scott Formation and Onny Shale Formation. Bancroft (1928a : 56-57) erected two species from the Upper Caradoc, O. reuschi and O. broeggeri, the latter of which he designated 'genotype'. He subsequently (1945 : 211-222) erected a series of seven species (five of them new) which, arranged in strati- graphical order, are O. reuschi, O. grandis, O. aspasia, O. depressa, O. sinuata, O. inconstans and O. broeggeri. He considered there to be two phylogenetic groups (1945 : 215), firstly the group of O. grandis forms, which encompassed O. reuschi, O. grandis, O. aspasia, O. depressa and O. inconstans. Bancroft claimed this group exhibited multiple progressive trends which were con- tinuous and unidirectional. O. aspasia and O. inconstans were considered to be slightly divergent from the main lines of descent. The second group was that of O. broeggeri, which also includes O. sinuata. Distinction between the seven species and the two species groups he drew in terms of shell size (an unreliable feature owing to transport and other environmental controls), impression of muscle scars, ribbing densities and relative branching of internal and external costellae. As pointed out by Williams & Wright (1963) and Williams (1963), in Bancroft's study of Howellites he did not subject his data to tests of significance, so that chosen characters, particularly rib patterns, used for specific diagnosis may not have been as sufficiently different as he believed and thus so reliable taxonomically. Bancroft also erected some species of Onniella on only a limited amount of material and often with little knowledge of the internal structures and variation. For example, he erected O. sinuata on a small number of conjoined valves and only one dorsal interior; appar- ently he had no ventral interiors. For the present study large collections of topotypic material were obtained, as well as accessory populations (e.g. colln M 1 1 corresponds to the cotype material used by Bancroft in describing O. reuschi, colln CF4 is an Actonian population of O. reuschi and colln O33 is stratigraphically the highest population of any Onniella in the Caradoc area and is assigned to O. broeggeri}. The statistical data given in Tables 74-82 are based on these collections. As regards bivariate statistics, species were arranged in stratigraphical order and each popula- tion tested against those adjacent. The topotype population (WFT7), cotype population (Mil) and colln CF4 do not differ from each other in bivariate or other characters and are treated as the species O. reuschi. Also population O33 does not differ from topotype O. broeggeri. 0. reuschi (WFT7) differs from O. grandis solely in having a slower increase in width of the dorsal cardinalia (P < 0-05), yet O. grandis and O. aspasia, O. reuschi (WFT7) and O. aspasia do not differ as regards bivariate characteristics. O. reuschi differs from O. depressa in having a relatively narrower brachial valve (P < 0-05) and a slower increase in length of the dorsal cardinalia (P < 0-05). However, O. reuschi and O. depressa do not differ from O. inconstans, which itself does not differ from O. reuschi (WFT7) or O. broeggeri. Further, O. broeggeri does not differ from the topotype or cotype populations of O. reuschi, or from the Actonian population CF4. In summary, bivariate statistics show that the seven 'species' of Onniella are very similar mor- phologically and do not segregate into neat species patterns. Thus it is misleading for Bancroft (1945) to distinguish 1. O. reuschi from O. grandis in terms of the greater size and larger muscle field of the latter 2. O. aspasia from O. grandis in terms of the delicate nature of the crural plates and smaller dorsal muscle area of the former 262 J. M. HURST 3. O. depressa from the remaining species by its relatively long shell and rounded contour 4. O. broeggeri by its small size, and 5. O. inconstans by the characteristic form of the valves. Although the length/width relationship of the adult ventral muscle field (in valves above 4 mm in length) does not change relative proportions from species to species, the actual configuration of the scar does show changes of shape. There are two basic ventral muscle field configurations, here referred to as 'LT- and 'V'-shaped (Fig. 352). Table 83 lists the distribution of pedicle valves with the various muscle field shapes, from the topotype populations. Chi-squared (x2) tests show there to be three morphological groups in which the component species do not differ significantly: 1 . O. reuschi, O. grandis and O. aspasia 2. O. depressa and O. sinuata 3. O. inconstans and O. broeggeri All three species of Group 1 have predominantly U-shaped muscle fields and differ significantly (/><0-05) from both species of Group 2, which have approximately equal U- and V-shaped muscle fields. They also differ from both species of Group 3 (P< 0-01), which have a predominantly V-shaped muscle field. Both species of Group 2 differ from both Group 3 species (P<0-05). Strength of the crural fossettes varies amongst the seven 'species'. On specimens above 4 mm in length these have been categorized as absent, weak or strong. The distributions for the topotype populations are shown in Table 84. Chi-squared (x2) tests again define the same three groups, with the component species not significantly different. All species of Group 1 differ from Group 2 and 3 species (P< 0-01 in each case) in having weak or absent crural fossettes. Group 2 species, with strong or weak crural fossettes, differ significantly from Group 3 species (/><0-01), which have predominantly strong crural fossettes. As noted by Bancroft (1945), the impression of the dorsal muscle field of Onniella varies. Table 85 lists the distribution of brachial valves with strongly or weakly impressed dorsal adductor scar. The following two groups are evident, in which the component species do not differ signi- ficantly: 1 . O. reuschi, O. grandis, O. aspasia, O. depressa and O. sinuata 2. O. inconstans and O. broeggeri Group 1 species, with a weakly-impressed dorsal adductor scar, differ significantly from Group 2 species (P<0-01), which have a strongly-impressed adductor scar. Bancroft mainly defined Onniella species on their ribbing patterns. Consequently, costellae densities as well as costellae insertions have been analysed (see Tables 86, 87). For costellae densities data were grouped into 5/6 and 7/8 costellae and tested for significance. O. reuschi, O. aspasia and O. grandis do not differ significantly. The former two have slightly finer ribbing and differ significantly (P<0-05) from the slightly coarser-ribbed O. depressa and O. sinuata, which themselves are similar and also do not differ significantly from O. inconstans and O. broeggeri. O. grandis does not differ significantly from any species. In conclusion, although there is a general trend for finer-ribbed species to occur in the top of the Cheney Longville Formation (i.e. O. reuschi) and lower Acton Scott Formation (i.e. O. grandis and O. aspasia), and for coarser- ribbed species to occur in the upper Acton Scott Formation (i.e. O. depressa and O. sinuata) and Onny Shale Formation (i.e. O. inconstans and O. broeggeri), much of the variation is not significant. Figs 332-351 Onniella broeggeri Bancroft, internal moulds of pedicle valves : 332 = BB 7245 1 , x 2-5, 333 = BB 72452, x2-5, 334 = BB 72453, x2-5, 335 = BB 72460, x2-5. Horderleyel/a cf. plicata Bancroft, internal moulds of pedicle valves: 336 = BB 72492, x2-5, 337 = BB 72493, xl-2; internal and external moulds of brachial valves: 338 = BB 72494, x 1-4, 339 = BB 72495a, X 1-2, 340 = BB 72495b, X 0-9. Reuschella bilobata (J. de C. Sowerby), internal moulds of pedicle valves: 341 = BB9269 (lectotype of synonymized R. semiglobata), xl-5, 342 = BB 72508, xl, 343 = BB 72509, X 0-9, 344 = BB 72507, xO-9; internal and external moulds of brachial valves: 345 = BB 72505, > 1-1, 346 = BB 72506, xl, 347 = BB 72504b, xl-3, 348 = BB 72504a, x2, 349 = BB 72503, xO-9. Internal mould of pedicle valve of Triplesia sp.: 350 = BB 72370, x3. Internal mould of pedicle valve of Bicuspinasp.: 351 = BB 72371, x3. CARADOC STRATIGRAPHY AND BRACHIOPODS 263 332 335 334 333 336 337 338 339 340 344 346 347 345 348 349 350 351 264 J. M. HURST diduc tor muscle u Fig. 352 adduc tor muscle Diagrammatic representation of the U- and V-shaped pedicle muscle fields of Onniella. The earlier insertion of eight costellae relative to eight others has been analysed (Table 87); the data for O. broeggeri are from Williams (1963 : 408). lTJ)l a are constantly fairly evenly associated throughout all seven 'species'. With regard to the earlier insertion of 2b)2a, both O. depressa and O. sinuata are similar and differ significantly (/*<0-05) from the remaining species in the earlier insertion of 2a. O. inconstans and O. broeggeri have similar insertion rates for 2aT)2b and differ significantly (/><0-05) from the remaining species in the earlier insertion of 2b. Rib patterns 3aTa)3a and 3aTa)2a are constant throughout all seven species, with the earlier insertion of 3a and 2a respectively. 3a is also regularly inserted earlier than 3c in all seven species. Regarding 3al)2bl, O. depressa and O. sinuata have similar insertion rates, but in having the earlier insertion of 3al differ significantly from the remaining species (P< 0-05-0-01). With the insertion of 4b)4b O. inconstans and O. broeggeri are similar, yet significantly differ (P< 0-001) from the remaining species in the earlier insertion of 4o. In summary, O. reuschi, O. grandis and O. aspasia are conspecific, as are O. depressajO. sinuata arid O. inconstansjO. broeggeri. Consequently, three species are recognized, O. reuschi, O. depressa and O. broeggeri. They differ in the configuration of the ventral muscle field and the strength of the crural fossettes. Further, O. broeggeri can be distinguished by its more heavily impressed quadripartite dorsal adductor muscle field, the significantly earlier insertion of both 2b relative to 2aT and 4b relative to 4b. O. depressa can be further distinguished by the significantly earlier insertion of 2a relative to 2b and 3al relative to 2bT. O. reuschi tends to have the finest ribbing although the distribution within a population is not significantly different from O. depressa and O. broeggeri. Family HARKNESSELLIDAE Bancroft, 1928 Genus HORDERLEYELLA Bancroft, 1928 Horderleyella cf. plicata Bancroft Figs 336-340 19286 Horderleyella plicata Bancroft : 186-187; pi. 1, figs 15-18. Figs 353-378 Heterorthina notata (Barrande), internal moulds of pedicle valves: 353 = BB 72342, x 1-6, 354 = BB 72341, x 1-6, 355 = BB 72346, x 1-6, 356 = BB 73683, x 1-8, 357 = BB 73684, xl-8, 358 = BB 72345, x2-l, 359 - BB 72343, xl-6, 360 = BB 72344, xl-8; internal and external moulds of brachial valves: 361 = BB 72340, x 1-7, 362 = BB 72347, x2-2, 363 = BB 72337, x 1-5, 364 - BB 72338, x 1-2, 365 = BB 72339, x 1-1, 366 = BB 73682, x 1-3. Heterorthina praeculta Bancroft, internal moulds of pedicle valves: 367 = BB 72284, x 1-3, 368 = BB 72285, xl-5, 369 = BB 72286, xl, 370 = BB 72287, xl-2; internal and external moulds of brachial valves: 371 - BB 72288a, xl-6, 372 = BB 72288b, xl-2, 373 = 72291a, xl-2, 374 = BB72291b, xl-2, 375 = BB 72289b, xl-3, 376 = BB 72289a, xl-8, 377 = BB 72290b, xl-1, 378 = BB72290a, xl-3. CARADOC STRATIGRAPHY AND BRACHIOPODS 265 353 357 361 365 370 375 354 355 356 FT*. 358 359 360 m- 362 j m 363 364 366 367 368 369 372 373 374 378 266 J. M. HURST 1945 H order leyella plicata Bancroft : 236-237; pi. 31, figs 4-7; pi. 32, figs 1-4. 1974 Horderleyella cf. plicata Bancroft; Williams : 102-103; pi. 16, figs 17, 19, 20; pi. 17, fig. 1. DESCRIPTION. Ventribiconvex Horderleyella with right-angled cardinal margins, brachial valve approximately seven-tenths as long as wide and one-fifth as deep as long, pedicle valve three- quarters as long as wide and up to one-third as deep as long, and varying from weakly convex carinate forms to strongly convex non-carinate forms, ventral interarea long, slightly curved apscaline, delthyrium open, dorsal interarea short flat anacline, notothyrium open; radial ornamentation fascicostellate and ornamented by 4 or 5 fascicostellae per mm, 2 mm antero- medially of the dorsal umbo. Ventral interior with small teeth and weak crural fossettes, supported by short divergent dental lamellae which extend anteriorly for one-seventh the valve length and extend forward as fairly strong ridges peripheral to the muscle field and curving back to the mid-line; ventral muscle scar bilobed, one-third as long as the valve and one-fifth as wide as the valve is long, and consisting of a narrow, elongately oval median adductor impression not enclosed by triangular diductor tracks. Dorsal interior with cardinal process consisting of short, slender shaft expanding anteriorly for one-sixth the valve length and splaying laterally for one-quarter of the length of the valve, fulcral plates defining sockets; muscle field obscure. DIMENSIONS. length width Internal mould of pedicle valve, BB 72492 7-3 (est.) 8-9 (est.) Internal mould of pedicle valve, BB 72493 14-2 16-8 Internal mould of brachial valve, BB 72494 10-5 (est.) 14-2 Internal and external mould of brachial valve, BB72495a, b 13-0 17-6 HORIZON AND LOCALITIES. BB 72492 from colln H, loc. 31. BB 72493, BB 72494 and BB 72495,a b from colln Mil, loc. 28. DISCUSSION. A small number of moulds recovered from the Crosspipes Member (Marshbrookian) appear to be conspecific with H. plicata Bancroft from the Costonian and Harnagian of the type Caradoc. H. plicata may possibly be slightly coarser ribbed, but that aside this stock appears to have undergone little morphological change through the period Costonian-Marshbrookian. Genus REUSCHELLA Bancroft, 1928 TYPE SPECIES. Reuschella semiglobata Bancroft by original designation of Bancroft (19286 : 180), treated here as a junior subjective synonym of Orthis bilobata J. de C. Sowerby. Reuschella bilobata (J. de C. Sowerby) Figs 341-349 1839 Orthis bilobata J. de C. Sowerby in Murchison : 640; pi. 19, fig. 7. 1869 Orthis vespertilio J. de C. Sowerby; Davidson : 236 pars; pi. 30, figs 11, 17, 18, non figs 12-16. \928b Reuschella semiglobata Bancroft : 188-190; pi. 2, figs 9-12. 1945 Reuschella bilobata (J. de C. Sowerby) Bancroft : 239. DESCRIPTION. Strongly dorsibiconvex subquadrate Reuschella with rounded cardinal margins, brachial valve approximately three-quarters as long as wide and one-third as deep as long with a very shallow sinus, pedicle valve slightly concave in cross-section parallel to hinge and approxi- mately three-quarters as long as wide and with a mean depth relative to length of one-eighth at the sharp median carina, lateral folds undulate; ventral interarea long, curved apscaline, delthyrium open, dorsal interarea short, flat anacline and with open notothyrium; radial ornamentation medium to coarsely fascicostellate. Ventral interior with massive teeth and supported by divergent to subparallel dental plates which extend anteriorly for one-quarter the valve length and extend forward as strong ridges peripheral to muscle field; ventral muscle scar bilobed, two-fifths as long as valve and one-third CARADOC STRATIGRAPHY AND BRACHIOPODS 267 as wide as valve length, and consisting of a narrow lanceolate median adductor impression not enclosed by elongate triangular diductor tracks. Dorsal interior with cardinal process consisting of massive rounded shaft and crenulated, ridged bulbous myophore completely filling notothyrium, notothyrial platform massive and indented anteriorly by moderately deep pits of adductor muscle scar, brachiophores divergent, extending anteriorly for one-seventh of the valve length and splaying laterally for one-third of the valve length, sockets deep and crenulated, muscle scar obscure. MATERIAL AND LOCALITIES. Lectotype (selected Cocks 1978 : 76) GSM Geol. Soc. Colin GSd 5527, from Acton Scott Formation, 'Acton Scott' (exact locality unknown). Lectotype of semiglobata (Fig. 341) BB9269, length 25-7 mm, width 31-0 mm, from loc. 42. Loc. 24, colln WFT1, e.g. Fig. 349, BB 72503, length 24-4 mm, width 29-5 mm; also colln WFT2, e.g. Figs 347-8, BB 72504, length 11 -4 mm, width 15-0 mm. Loc. 1, colln CL7, e.g. Fig. 342, BB 72508, length 25-1 mm, width 33-3 mm. Loc. 44, colln OA25, e.g. Fig. 345, BB 72505, length 23-8 mm, width 36-2 mm. BB 72509 (Fig. 343), BB 72507 (Fig. 344) and BB 72506 (Fig. 346) are with no precise locality labels, but are probably from loc. 47. Known from Iocs 1-48. DISCUSSION. Reuschella occurs throughout the Alternata Limestone to Acton Scott Formation but never in any abundance. J. de C. Sowerby in Murchison (1839) described Orthis bilobata from some part of this sequence. Bancroft (1928&), when he erected Reuschella, commented on the fact that O. bilobata belonged to Reuschella, but because it was not precisely located and known only by the interior of the brachial valve he ignored it. Instead, he erected R. semiglobata, saying it differed from R. bilobata in having small wings, a more conspicuous sinus in the brachial valve, absence of well-defined dorsal muscle scars and in having a greater longitudinal extension of the crural plates. New collections, and examination of Bancroft's type material, show R. semiglobata to be a variable stock within which is encompassed the single figured specimen of R. bilobata. Thus R. semiglobata is treated as a synonym of R. bilobata. Further, all post-Lower Longvillian Reuschella specimens are assigned to R. bilobata. Bancroft (19286) described R. horderleyensis from pre-Upper Longvillian strata, in the type Caradoc. He did, however, note that in terms of shell shape some R. horderleyensis specimens are comparable to R. bilobata. However, until more is known of the variability of R. bilobata the two species are best kept separate. Family HETERORTHIDAE Schuchert & Cooper, 1931 Genus HETERORTHIS Hall & Clarke, 1892 Heterorthis alternata (J. de C. Sowerby) 1839 Orthis alternata J. de C. Sowerby in Murchison : 638; pi. 19, fig. 6. 1871 Orthis alternata J. de C. Sowerby; Davidson : 264 pars; pi. 31, figs 7-8, non figs 1-6. 1963 Heterorthis alternata (J. de C. Sowerby) Williams : 418; pi. 9, figs 1-6, 8, 9. Williams (1963) has redescribed and figured this species, and his work will not be duplicated here. The species occurs at Iocs 1-8 inclusive and 44. Further, Dean (1958) reported a thin shell bed of H. alternata in the lowest Acton Scott Formation of the Onny Valley, but this is not now exposed, and Harper (1978) another thin shell bed in the Onny Shale Formation. Genus HETERORTHINA Hall & Clarke, 1892 1928a Heterorthina Bancroft : 59. 1960 Elsaella Alichova : 192. 1975 Elsaella Alichova; Hints : 78. DIAGNOSIS. Transverse, very gently and unequally biconvex heterorthids with more deeply convex pedicle valve and shallow, persistently sulcate brachial valve; ventral interarea short, 268 J. M. HURST curved, apscaline with open delthyrium; dorsal interarea short, anacline; notothyrium com- pletely filled by cardinal process; shell punctate; radial ornamentation multicostellate, of fine to very fine costellae with both external and internal branches freely developed in all sectors; costellae curving in posterior regions of shell to intersect hinge line. Ventral interior with small teeth supported by short, blunt, divergent dental lamellae; pedicle callist well developed; muscle field large, elongately pentagonal, occasionally with semiflabellate diductor tracks, anteriorly not enclosing lanceolate adductor scar (very rarely anterior lobes of convergent diductor scars meet medially to surround adductor scars completely.) Dorsal interior with strong cardinal process, anterior shaft with cleft, myophore expanded into large ovoid structure which early in ontogeny is bilobed but which with maturity develops up to four small septa dividing myophore into separate lobes; posterior end of myophore crenulated and extending well beyond hinge line; brachiophores greatly divergent, with divergent bases, fulcral plates absent; dorsal adductor scars elongate with approximately equal posterior and anterior elements. TYPE SPECIES. Heterorthina praeculta Bancroft, by original designation of Bancroft (1928# : 59). DISCUSSION. Bancroft (1928« : 59-60) erected this genus; whilst recognizing it was distinct from Heterorthis alternata (Sowerby) he noted that it shared many characteristics. The assignment of Heterorthina to a suprageneric category was formerly unsettled. Williams & Wright (1963) placed it in the Dalmanellidae. Later Harper, Boucot & Walmsley (1969) question- ably assigned the genus to the subfamily Heterorthinae Schuchert & Cooper, of the Rhipidomel- lidae Schuchert, 1913. It is now clear from the thorough work of Havlicek (1970) that Heterorthina should unquestionably be assigned to the Heterorthidae of Schuchert & Cooper, 1 93 1 . The diverg- ing brachiophore bases and lobation of the cardinal process myophore into a bilobed, trilobed or quadrilobed structure suggest strong affinities with Tafilaltia Havlicek, 1970 and Svobodaina Havlicek, 1950. The configuration of the pedicle valve muscle field of Heterorthina also suggests such affinities. Such a line of derivation involved increasing complexity of the cardinal process myophore and lobation of the pedicle diductor muscle field. It is interesting to note that young specimens of the pedicle valves of Svobodaina inclyta (Barrande), figured by Havlicek (1970 : pi. 5), are remarkably similar, especially in diductor muscle scar configuration, to adult Heterorthina. Further, Havlicek (1970 : 23) notes that, rarely, the anterior lobes of the adductor scars of Svobodaina touch, to surround the adductor tracks completely, a feature also occasionally seen in Heterorthina praeculta. Williams & Wright (1963) placed the genus Elsaella Alichova in synonymy with Heterorthina. Recently, Hints (1975) has brought Elsaella out of synonymy, believing the central position of the pedicle adductor scar, trilobed cardinal process and possibly the rhipidomellid type of ribbing warrant a separate genus. The trilobed cardinal process is common in Heterorthina and the central position of the pedicle adductor field is occasionally attained in H. praeculta through the medial expansion of the diductor tracks (see also Howe 1965 : text-fig. 6). The only apparent difference between the two genera is the coarser ribbing of Elsaella. These small differences do not warrant the retention of the separate genus Elsaella. Heterorthina morgatensis Melou 1975 Heterorthina morgatensis Melou : 192-203; pi. 20, figs 1-9; pi. 21, figs 1-12; pi. 22, figs 1-5. DESCRIPTION. Small, transverse, unequally biconvex Heterorthina, with weakly convex brachial valve 68 % as long as wide; pedicle valve 71 % as long as wide; delthyrium and notothyrium open, ventral interarea weakly curved apscaline; dorsal interarea short anacline; radial ornamentation of fine costellae 4 or 5 per mm, 5 mm anteromedially of dorsal umbo; costellae in posterior region curving to intersect hinge line. Ventral muscle field strongly bilobed, flabellate, extending anteriorly for 61 % of the length of the pedicle valve, but not surrounding the median adductor field; diductor scars 58% as wide as the pedicle valve is long. CARADOC STRATIGRAPHY AND BRACHIOPODS 269 Dorsal interior with broad, shallow median ridge, at posterior end of which arises the moderately thick cleft shaft of the cardinal process, which expands into bilobed myophore filling most of the notothyrium; brachiophores divergent, short, strong, with broad divergent bases extending forward for 21 % of the length of the brachial valve and splaying laterally for 26% of the valve length; notothyrial platform poorly developed; sockets narrow and well-defined, without fulcral plates; adductor scars elongately oval, impressed quadripartite and extending anteriorly for 50 % the length of the valve. DISCUSSION. See p. 271. Heterorthina notata (Barrande) Figs 353-366 1879 Orthis notata Barrande : pi. 66, fig. II (non pi. 127, fig. VIII). 1932 Heterorthis barrandei Schuchert & Cooper : 137. 1950 Heterorthina notata (Barrande) Havlicek : 105-106; pi. VIII, figs 1, 2, 8, 10 and 13. 1975 Heterorthina notata (Barrande); Melou : pi. 22, figs 8-9. 1977 Heterorthina notata notata (Barrande); Havlicek : 116; pi. 20, figs 18-24. DESCRIPTION. Transverse, unequally biconvex Heterorthina, with evenly weakly convex, shallow sulcate brachial valve 76% as long as wide; pedicle valve 80% as long as wide and 18 % as deep as the valve is long; delthyrium and notothyrium open, pedicle callist large, ventral interarea weakly curved apscaline, dorsal interarea short anacline and with prominent growth lines; radial ornamentation of fine costellae commonly 4 per mm, 5 mm anteromedially of the dorsal umbo; costellae curve to intersect hinge line. Ventral muscle field bilobed with diductor scars, showing well-developed growth increments, extending anteriorly for 42% of the length of the pedicle valve but not surrounding the elongately oval median adductor field; diductor scars 36% as wide as the valve is long; teeth very small, supported by short receding dental plates lacking crural fossettes and extending anteriorly for 15% of the length of the pedicle valve; shell ridge extensions of the dental plates bounding the lateral and anterior parts of the diductor scars. Dorsal interior with broad median ridge, at posterior end of which arises very broad shaft of cardinal process; myophore extremely large, filling notothyrium, bilobed in individuals below 9 mm in length and trilobed or quadrilobed by septa in more mature specimens, crenulated posteriorly and extending well beyond hinge line; brachiophores divergent, short, very strong and with broad divergent bases, extending forward for 15% of the length of the brachial valve and splaying laterally for 26% of the valve length; notothyrial platform very well developed; well- defined narrow sockets lacking crural fossettes; adductor scars elongately suboval, impressed particularly posteriorly on either side of broad median ridge and extending anteriorly for 48% the length of the valve. DIMENSIONS. length width Internal mould of brachial valve, BB 72337 14-0 17-0 Internal mould of brachial valve, BB 72338 15-7 20-8 Internal mould of brachial valve, BB 72339 18-0 23-0 Internal mould of brachial valve, BB 72340 16-2 20-0 Internal mould of brachial valve, BB 72347 9-1 12-1 External mould of brachial valve, BB 73682 11-4 13-0 Internal mould of pedicle valve, BB 72341 15-1 19-8 Internal mould of pedicle valve, BB 72342 15-6 21-4 Internal mould of pedicle valve, BB 72343 14-6 17-2 Internal mould of pedicle valve, BB 72344 17-5 20-8 Internal mould of pedicle valve, BB 72345 12-3 15-9 Internal mould of pedicle valve, BB 72346 13-4 18-0 Internal mould of pedicle valve, BB 73683 14-1 16-4 Internal mould of pedicle valve, BB 73684 8-7 1 1 -8 HORIZON AND LOCALITIES. All figured material is from the type locality in a temporary trench along a railway line at Palmovka Hill, Praha - Liben, Czechoslovakia; Beroun Series, Zahofany Formation. 270 j. M. HURST DISCUSSION. See p. 271. Heterorthina praeculta Bancroft Figs 367-379 1928« Heterorthina praeculta Bancroft : 59-60; pi. 2, figs 14-21. 1963 Heterorthina praeculta Bancroft; Williams & Wright : 28; pi. 1, fig. 19. 1965 Heterorthina praeculta Bancroft; Wright in Williams et al.: H335. 1967 Heterorthina macfarlani Neuman : A10-A12; pi. 2, figs 1-19. 1975 Heterorthina praeculta Bancroft; Melou : 202; pi. 22, figs 6-7. 1975 Heterorthina kerfornei Melou : 203-207; pi. 23, figs 1-14; pi. 24, figs 1-13. DESCRIPTION. Transverse, unequally biconvex Heterorthina with an evenly weakly convex to flat, weakly sulcate brachial valve 76% as long as wide; pedicle valve 80% as long as wide and 24% as deep as the valve is long; delthyrium and notothyrium open, pedicle callist small, ventral intera°rea slightly curved apscaline, dorsal interarea short anacline; radial ornamentation of very fine costellae commonly 5 or 6 per mm, 5 mm anteromedially of the dorsal umbo ; costellae in posterior region curve to intersect hinge line. Fig. 379 Range of variation in pedicle musculature of Heterorthina praeculta. Note variation from V to U shapes from 1 to 4. Ventral muscle field bilobed with diductor scars extending anteriorly for 43% of the length of the pedicle valve but not surrounding the elongately oval median adductor scar (rarely adductor scar is surrounded), from which they are separated by shallow depressions (expressed as ridges in mould); diductor scars 39% as wide as the pedicle valve is long; teeth small, supported by short receding dental plates, extending anteriorly for 14% of the length of the pedicle valve and with strong crural fossettes; shallow ridges bound the diductor scars anteriorly. Dorsal interior with broad median ridge, at posterior end of which arises broad, deeply cleft shaft of cardinal process; myophore large, bilobed or trilobed and crenulated posteriorly ex- tending beyond hinge line; brachiophores short strong, divergent, with broad divergent bases, extending forward for 17% of the length of the brachial valve and splaying laterally for 31 % of the length of the valve; sockets narrow, well-defined, without fulcral plates; adductor scars sub- oval, impressed on either side of median ridge and extending anteriorly for 47% the length of the valve. MATERIAL AND LOCALITIES. Lectotype BB 9152 (selected Cocks 1978 : 74) from loc. 31, from which also comes all the material figured here, e.g. Fig. 367, BB 72284, length 17-0 mm, width 20-0 mm; Fig. 368, BB 72285, length 9-0 mm, width 10-0 mm; Fig. 369, BB 72286, length 17-0 mm, width 19-8 mm; Fig. 370, BB 72287, length 16-5 mm, width 21-0 mm; Figs 371-2, BB 72288, length 15-3 mm, width 18-6 mm; Figs 373-4, BB 72291, length 10-6 mm, width 13-6 mm; Figs 375-6, BB 72289, length 8-3 mm, width 12-2 mm; Figs 377-8, BB 72290, est. length 17-0 mm, width 20-5 mm. Also recorded from Iocs 1 1, 29 and 32. Outside Britain the species is known from the Grier Limestone Member of the Lexington Limestone (Soudleyan-Longvillian) of North America, and the Schistes de Postolonnec (Caradoc) of Brittany, France. 271 DISCUSSION. A topotypic sample of H.praeculta has been statistically compared with H. macfarlani (data given by Neuman 1967 : Al 1 and Melou 1975 : 201-202) from the Grier Limestone Mem- ber of the Lexington Limestone, U.S.A. (Soudleyan - Longvillian ?), H. kerfornei (data given by Melou 1975 : 205-206) from the Schistes de Postolonnec (Caradoc), France, and H. morgatensis (data given by Melou 1975 : 197-198), also from the Schistes de Postolonnec (Tables 88-96). The comparisons show that H. macfarlani and H. kerfornei do not differ significantly from H. praeculta. 1 cm Fig. 380 Range of variation in pedicle musculature of Heterorthina kerfornei. Note variation from V to U shapes from 1 to 8. Melou (1975 : 206) claims that H. kerfornei differs significantly from H. praeculta in terms of size, form of the ventral muscle field and morphology of the cardinal process. Absolute size as a taxonomic character is unreliable and the relative dimensions of the ventral muscle fields do not differ significantly. Further, the overall diductor muscle field variation expressed by a population of//, praeculta (Fig. 379) is contained within the same variation limits as H. kerfornei (Fig. 380). From the figured specimens of H. kerfornei populations appear to contain a predominance of individuals which have massive cardinal process shafts filling the notothyrial cavity. In //. praeculta the cardinal process shaft is more variable, with individuals having either massive or non-massive ones, and thus H. kerfornei populations appear to be end members of the variation of topotype H.praeculta. There are therefore no consistent significant morphological differences between H.praeculta, H. macfarlani and H. kerfornei, consequently the relationships between the three species are most conveniently expressed by assuming that they in fact represent a single species, H. praeculta. H. morgatensis differs from H. praeculta in the significantly (P< 0-005) faster increase in relative length of the dorsal cardinalia. The dimensions of the ventral muscle field do not differ, yet the diductor tracks of H. morgatensis are frequently lobate whereas those of H. praeculta are not. Further, the cardinal process of the French species is very strongly bilobate, with a deep groove separating each lobe. From the figured specimens (Melou 1975 : pi. 21) H. morgatensis has 4 (LPB 2798) or 5 (LPB 2800) costellae per mm, 5 mm anteromedially of the dorsal umbo, and so appears more coarsely ribbed than H.praeculta. A topotypic sample of H. notata from near Prague differs significantly from H. praeculta in the slower relative width of the dorsal cardinalia (P< 0-001), the coarser ribbing (see Table 97; /*<0-01) and in lacking crural fossettes (Table 98; /*<0-01). The pedicle valve diductor tracks show a wide range of variation (Fig. 381) from simple pentagonal scars to exaggerated bilobed tracks, approaching those of Svobodaina. H. notata differs significantly from H. morgatensis in having a slower relative increase in dorsal cardinalia width (P< 0-05), and in having shorter dorsal cardinalia (P< 0-05). The ribbing of the two species approximates, yet the structure of the cardinal process is different. 272 J. M. HURST One of the earliest-described species referred to Heterorthina, H.fairmontensis (Foerste, 1909 : 322) of Maysville age (Onnian - Pusgillian), appears very similar to H. praeculta. Neuman (1967) considered the species to be very similar to H. macfarlani. Until more is known about the variation expressed by H.fairmontensis the two species are kept separate. The configuration of the pedicle valve muscle field (i.e. the complete surrounding of the adduc- tor scar by the diductor tracks), the coarse ribbing and the apparent lack of external branches in the first four sectors of ribbing in the brachial valve indicate that the type species of Elsaella is probably a distinct species of Heterorthina. Fig. 381 Range of variation in pedicle musculature of Heterorthina notata. Note variation from V to U shapes from 1 to 7. Genus MARIONITES Cooper & Muir-Wood, 1951 DIAGNOSIS (emended). Subcircular heterorthiids with deeply and evenly convex brachial valve with narrow median sulcus, ventral valve weakly resupinate and carinate in transverse profile; radial ornamentation multicostellate with costellae directed posteriorly along hinge line, branching freely developed ; ventral interarea short, straight orthocline to very weakly anacline, notothyrium open and filled by massive cardinal process. Ventral interior with massive teeth supported by thick, divergent, receding dental lamellae, pedicle callist well developed, muscle field flabellate surrounded by sinuous ridges, with submedian diductor lobes extending beyond, but not completely surrounding, median lanceolate adductor scar. Dorsal interior with massive, incipiently bilobed cardinal process, brachiophores greatly divergent, brachiophore bases weakly to moderately divergent, notothyrial platform massive, ankylosed to brachiophore bases and extending anteriorly into broad median ridge, weak fulcral plates occasionally developed; dorsal adductor scars elongate with posterior elements deeply inserted into anterior region of notothyrial platform. TYPE SPECIES. Marionella typa Bancroft, by original designation of Cooper & Muir-Wood (1951). DISCUSSION. The taxonomic position of Marionites has been unstable. Bancroft (19286) erected the genus Marionella and placed it in a new subfamily, the Harknessellinae, along with the genera Harknessella Reed, 1917, Reuschella Bancroft, 1928, Smeathenella Bancroft, 1928 and Horder- leyella Bancroft, 1928. He thought that the genus was morphologically intermediate between Harknessella and the Rhipidomellidae. Subsequently, Schuchert & Cooper (1932) relegated Marionella to a subgenus of Dinorthis, a position with wh ich Bancroft apparently agreed (Schuchert & Cooper 1932 : 96). They concluded that despite the lack of information regarding the punctate or impunctate nature of the shell, the internal features suggest dinorthiid affinities. The brachiopod name Marionella, being a junior homonym, was replaced by Marionites (Cooper & Muir-Wood 1951). Williams et al. (1965) placed Marionites in the family Plaesiomyidae Schuchert, 1913, as a subgenus of Multicostella Schuchert & Cooper, 1931. The nature of the multicostellate ribbing of Marionites (i.e. posteriorly-directed costellae along the hinge line) clearly suggests this genus CARADOC STRATIGRAPHY AND BRACHIOPODS 273 383 384 382 385 386 387 388 IB(^fc--^—^fc—^gj§(BI 392 393 394 395 Figs 382-395 Marionites typus (Bancroft), internal moulds of pedicle valves: 382 = lectotype BB9126, x2, 383 = BB72529, xl-5, 384 - BB 72532, x2, 385 = BB 72530, xl-3, 386 = BB 72531, xl-3; internal and external casts of pedicle valves: 387 = BB 9126, x 1-3, 388 = BB 24626, xl-3; internal and external moulds of brachial valves: 389 = BB 72537, xl-6, 390 = BB 24627, xl-7, 391 = BB 72534b, xl-9, 392 = BB 72533, x2-2, 393 = 6872536, xl-2, 394 = BB 24628, xl-2; internal cast of brachial valve: 395 = BB 9124, x 1-1. belongs in the Heterorthidae of Schuchert & Cooper 1931 (see Havlicek 1970 : 8-9; 1977 : 119). However, as the available material is preserved as internal and external moulds in sandstone, it is not known beyond doubt whether the shell was punctate, although it appears to be very finely so. Further, the nature of the flabellate ventral muscle scar, the massive elliptical cardinal process and brachiophore configuration are all suggestive of heterorthid affinities. None the less, in having a strongly convex brachial valve and resupinate and carinate pedicle valve Marionites resembles some members of the Harknessellidae. The similarity between Marionites and some genera of the Plaesiomyidae is a clear case of homoeomorphy. Havlicek's (1977) Marionites 274 J. M. HURST resupinatus appears to fall within the variation of M. typus, but until this can be adequately demonstrated they are left as separate species. Marionites typus (Bancroft) Figs 382-395 19286 Marionella typa Bancroft : 190-191 ; pi. 2, figs 13-16. 1932 Marionella typa Bancroft; Schuchert & Cooper : pi. 8, figs 16-18. 1951 Marionites typa (Bancroft) Cooper & Muir-Wood : 195. 1965 Marionites typa (Bancroft); Williams et al. : H321, fig. 202, 4a-d. DESCRIPTION. Subcircular Marionites with evenly convex, weakly sulcate brachial valve 12% as long as wide and 28% as deep as long, and resupinate pedicle valve 75% as long as wide; radial ornamentation finely multicostellate, branching freely developed; dental lamellae strong, diver- gent and extending anteriorly for 15 % the length of the pedicle valve, ventral muscle scar extend- ing anteriorly for 49% the length of the valve and 38% as wide as the valve is long; massive cardinalia with divergent brachiophore bases extending anteriorly for 18% of the length of the brachial valve and 28% wide as the valve is long; dorsal adductor scar extending forward of umbo for 45 % the length of the valve. MATERIAL AND LOCALITIES. Lectotype (Figs 382, 387) BB 9126 (selected Cocks 1978 : 51), length 18-0 mm, width 23-5 mm, from loc. 1 (= Bancroft loc. N6). From the same locality came BB 72529 (Fig. 383), length 16-2 mm, width 21-1 mm, BB 72530 (Fig. 385), length 16-8 mm, width 22-3 mm; BB 72531 (Fig. 386), length 15-9 mm, width 24-3 mm; BB 72532 (Fig. 384), length 134mm, width 16-7 mm; BB 72533 (Fig. 392), length 12-0 mm, width 15-3 mm; BB 72534 (Figs 390-1), length 13-5 mm, width 19-1 mm; BB 72535-6 (Figs 393-4); BB9124 (Fig. 395); BB 24626 (Fig. 388). Also from loc. 7, colln S3, e.g. BB 72537 (Fig. 389), length 15-6 mm, width 22-1 mm. Also recorded from Iocs 5 and 6, all in the Alternata Limestone Formation. Suborder TRIPLESIIDINA Moore, 1952 Superfamily TRIPLECIACEA Schuchert, 1913 Family TRIPLECIIDAE Schuchert, 1913 Genus TR1PLESIA Hall, 1859 Triplesia sp. Fig. 350 A distorted internal mould of a pedicle valve (BB 72370) from loc. 44, colln OA13 of the Acton Scott Formation is the sole representative of Triplesia in the Upper Caradoc rocks. The valve was approximately 4-5 mm long, 4-8 mm wide and 1 • 1 mm deep. Its outline was elongately subcircular, with rounded cardinal angles and a deep, slightly angular sulcus. Umbonal area crushed, but the dental lamellae appear short and widely splayed. Genus B1CUSPINA Havlicek, 1950 Bicuspina sp. Fig. 351 A distorted internal mould of a pedicle valve (BB 72371) from loc. 43, colln CP4 is the sole representative of Bicuspina in the Upper Caradoc. The valve was approximately 6-5 mm long, 10-6 mm wide and 1-7 mm deep. Its outline was subquadrate, with a straight, wide hinge line and with a rounded sulcus. Dental plates thin and divergent; pedicle tube obscured. CARADOC STRATIGRAPHY AND BRACHIOPODS 275 404 406 410 Figs 396-411 Leptestiina oeplki (Whittington), internal moulds of pedicle valves: 396 = BB 72510, x2-5, 397 = BB 72510, x3-0, 398 = BB 72511, x3, 400 = BB 72509, x3-2, 401 = BB 72512, x3-4; external mould of brachial valve: 402 = BB 72513, x3-2. Leptestiina sp., internal moulds of pedicle valves: 399 = BB 72514, x3, 403 = BB 72515, x3-7, 407 = BB 72516, x3; internal mould of brachial valve: 408 = BB 72517, x3-3. Eoplectodonta cf. rhombica (M'Coy), internal moulds of pedicle valves: 404 = BB 72519, x2, 405 = BB 72518, xl-8, 409 = BB 72521, xl-5, 410 = BB 72520, x2, 411 = BB 72519, x 1-4; external mould of brachial valve: 406 = BB 72522, xl-8. Order STROPHOMENID A Opik, 1934 Suborder STROPHOMENIDINA Opik, 1934 Superfamily PLECTAMBONITACEA Jones, 1928 Family LEPTELLINIDAE Ulrich & Cooper, 1936 Subfamily LEPTESTIININAE Havlicek, 1961 Genus LEPTESTIINA Havlicek, 1952 Leptestiina oepiki (Whittington) Figs 396-398, 400-402 1938 Sampo oepiki Whittington : 255; pi. 10, figs 15-16; pi. 11, fig. 10. 1963 Leptestiina oepiki (Whittington) Williams : 428-430; pi. 10, figs 15, 16, 19-21. aff.1977 Leptestiina aff. oepiki (Whittington); Mitchell : 76; pi. 15, figs 1-7. DESCRIPTION. Elongate, semicircular Leptestiina with evenly concavo-convex longitudinal profile, with pedicle valve about 60 % as long as wide and 40 % as deep as long ; ventral interarea apscaline, pseudodeltidium small and apical; ornamentation not known. Ventral interior with very short blunt teeth, supported by short receding dental plates; ventral muscle field subpentagonal in outline, extending anteriorly for about 30% of the length of the pedicle valve and about 75 % as long as wide, adductor field lanceolate, deeply impressed pos- teriorly and separated from adductor scars by ridges, diductor scars almost meeting anteriorly 276 J. M. HURST to surround adductors, whole muscle field surrounded by shell ridges which fuse posteriorly with dental lamellae. Brachial interior not known. DIMENSIONS. length width Internal mould of pedicle valve, BB 72509 6-5 10-9 Internal mould of pedicle valve, BB 72510 8-4 12-6 Internal mould of pedicle valve, BB 7251 1 7-2 13-5 Internal mould of pedicle valve, BB 725 12 7-2 10-8 External mould of brachial valve, BB 72513 10-7 HORIZON AND LOCALITIES. All figured material from loc. 44, colln OA1. Also recorded from Iocs 46 and 48. Leptestiina sp. Figs 399, 403, 407, 408 DESCRIPTION. Subcircular Leptestiina with evenly concavo-convex horizontal profile, with pedicle valve about 60% as long as wide and 35% as deep as long, ventral interarea apscaline, pseudo- deltidium small and apical, dorsal interarea short, hypercline; ornament unknown. Ventral interior with short blunt teeth supported by receding dental lamellae; ventral muscle field pentagonal in outline, extending anteriorly for about 30% of the length of the valve and 75 % as long as wide, adductor field lanceolate, weakly impressed, diductor scars almost meeting anteriorly to surround diductor scar. Dorsal interior with simple cardinal process consisting of anteriorly bifurcating, thin shaft connecting with notothyrial platform, socket ridges widely disposed subparallel to hinge defining a pair of oval sockets, lophophore platform 60% as long as wide, weakly defined posteriorly but apparently connecting with distal ends of socket ridges, medially incised anteriorly and orna- mented by radiating ridges. DIMENSIONS. length width Internal mould of pedicle valve, BB 72514 2-4 5-2 Internal mould of pedicle valve, BB 72515 4-0 8-6 Internal mould of pedicle valve, BB 72516 5-7 9-8 HORIZON AND LOCALITIES. All figured material from loc. 47, colln CF1. DISCUSSION. The few moulds of Leptestiina from the Acton Scott Formation show that there are two stocks present, which differ from each other in terms of outline, valve depth and ventral muscle field configuration. However, it is difficult to assess the importance of these differences until more is known about the variability of the stocks. Figs 412-444 Sowerbyella sericea (J. de C. Sowerby), internal moulds of pedicle valves: 412 = BB 73590, x2, 413 = BB 73591, xl-8, 414 = BB 73592, x2, 415 - BB 73593, xl-9, 416 = BB 73602, xl-9, 418 - BB 73601, xl-6, 419 = BB 73604, x2, 420 = BB 73603, x2, 421 = BB 73600, x2; internal moulds of brachial valves: 417 = BB 73607, xl-8, 423 = BB 73608, x 1-9, 424 = BB 73606a, xl-8, 426 - BB 73597, x 1-8, 427 = BB 73598, x2, 428 = BB 73595a, xl-4, 431 = BB 73609, xl-8, 432 = BB 73594, xl-7; external moulds of brachial valves: 422 = BB 73599, xl-8, 425 = BB 73606b, xl-9, 429 = BB 73595b, xl-6, 430 = BB 73596, xl-6. Chonetoidea radiatula (Barrande), internal moulds of pedicle valves: 433 = BB 73568, x3, 434 = BB 73565, x3-5, 435 = BB 73566, x3, 436 - BB 73569, x3-7, 437 = BB 73567, x3-6; external moulds of brachial valves: 438 = BB 73576, x3-6, 439 - BB 73577, x3-2; internal moulds of brachial valves: 440 = BB 73575, x2-7, 441 = BB 73573, x3-8, 442 - BB 73574, x3,443 = BB 73572, x 3-2, 444 = BB 73571, x3-5. CARADOC STRATIGRAPHY AND BRACHIOPODS 277 412 422 426 431 436 441 413 414 415 418 423 424 427 428 429 432 433 434 :, 442 443 416 421 425 430 435 ' ••di / p 536 539 540 294 J. M. HURST Family CHRISTIANIIDAE Williams, 1953 Genus CHRISTIANIA Hall & Clarke, 1892 Christiania hollii (Davidson) Figs 542-557 1871 Strophotnena Hollii Davidson : 303; pi. 42, figs 18-19. 1958 ' Rafinesquina1 holli (Davidson) Dean : pi. 25, fig. 12. 1978 Christiania hollii (Davidson) Cocks : 123, 200. DESCRIPTION. Semicircular to subquadrate concavo-convex Christiania with pedicle valve 75% as long as wide and 30% as deep as long and occasionally with sharp-sided sulcus, nature of delthy- rium unclear but with apical foramen. Ornament of raised, sharp-sided, concentric ridges. Ventral interior with impressed oval node containing muscle field and extending forward for about 50% of the valve length, elongate oval adductor track situated posteriorly and surrounded anteriorly by diductor track, dental plates very short and divergent. Dorsal interior with thin, widely divergent socket ridges which lie parallel to hinge line, anterior to which is anchor-shaped protegular node. Outer pair of longitudinal septa separated and dis- placed slightly anterior of socket ridges, inner septa short and diverging at about 40°, nature of cardinal process unknown. DIMENSIONS. length width Internal mould of pedicle valve, BB 73610 1-4 1-8 Internal mould of pedicle valve, BB 7361 1 1-6 2-4 Internal mould of pedicle valve, BB 73612 1-6 2-2 Internal mould of pedicle valve, BB 73613 1-7 2-8 Internal mould of pedicle valve, BB 73614 1-9 2-6 Internal mould of brachial valve, BB 73620 1-7 2-5 External mould of brachial valve, BB 73623 1-6 2-1 External mould of brachial valve, BB 73624 1-4 2-3 External mould of pedicle valve, BB 73625 1-5 2-1 HORIZON AND LOCALITIES. Recorded from Iocs 50 and 51 . All Davidson's type specimens are on a small block collected by Morris from loc. 51. DISCUSSION. This peculiar brachiopod can be allocated to the genus Christiania on the basis of the brachial valve septa (Cocks 1978). C. hollii resembles the younger C. port lock i Mitchell from the Killey Bridge Formation of Pomeroy, Ireland (Mitchell 1977). The only obvious difference between the two species is that the young growth stages of C. portlocki have much finer concentric ornament. As no larger specimens of C. hollii exist it is difficult to compare the species any further. Family LEPTAENIDAE Hall & Clarke, 1894 Genus LEPTAENA Dalman, 1828 Leptaena salopiensis Williams 1963 Leptaena salopiensis Williams : 461-462; pi. 15, figs 7, 8, 10-16. REMARKS. This species, described and figured by Williams (1963), is known from Iocs 35, 42-5 and 47 of the Acton Scott Formation. CARADOC STRATIGRAPHY AND BRACHIOPODS 295 552 556 557 Figs 542-557 Christiania hollii (Davidson), internal moulds of pedicle valves: 542 = BB73612, x8-4, 543 = BB 73613, x 10, 544 = BB 73616, x8-3, 545 = BB 73614, x8-5, 546 = BB 73610, x9, 547 = BB 73615, x 8, 549 = BB 73611, :<9; internal moulds of brachial valves: 548 = BB 73622, x8-5, 550 = BB 73619, x 10, 551 - BB 73620, xll-6, 555 = BB 73618, x9-5, 557 = BB 73621, x9-2, 556 - BB 73617, x9; external moulds of brachial valves: 552 = BB 73623, x 13-8, 554 = BB 73624, x8-3; external mould of pedicle valve: 553 = BB 73625, x9. Order SPIRIFERIDA Waagen, 1883 Suborder ATRYPIDINA Moore, 1952 Superfamily ATRYPACEA Gill, 1871 Family ATRYPIDAE Gill, 1871 Subfamily ZYGOSPIRINAE Waagen, 1883 Genus ZYGOSPIRA Hall, 1862 ? Zygospira sp. Figs 5 16-521 DESCRIPTION. Small biconvex Zygospira, with a brachial valve 95% as long as wide and 20% as deep as- long, and a pedicle valve 30% as deep as long; dorsal fold and ventral sulcus weakly developed, 35% as wide as the valve, rostrate with highly curved ventral umbo, radial ornamenta- tion of rounded costae with up to 2 on the dorsal fold and 4 to 7 on the lateral slopes. Ventral interior with very small teeth extending anteriorly for about 20 % the length of the valve, muscle field unknown. Dorsal interior with small, medially cleft triangular hinge plate; muscle scars not impressed. DIMENSIONS. Internal mould of brachial valve, BB 72528 Internal mould of brachial valve, BB 72525 Internal mould of brachial valve, BB 72524 Internal mould of pedicle valve, BB 72523 length 3-5 4-0 3-9 3-3 width 3-5 4-0 4-3 HORIZON AND LOCALITIES. BB 72528 from loc. 43, colln CP4. BB 72523, BB 72526 and BB 72525 from loc. 44, collns OA10, OA13 and OA14 respectively. BB 72524 from loc. 35, colln DH1 and BB 72527 from loc. 45, colln AS3. Also recorded from Iocs 42 and 46. 296 J. M. HURST DISCUSSION. Occasional disarticulated valves from the Acton Scott Formation are provisionally assigned to Zygospira on the basis of their hinge plate and valve configuration. However, the sulcal development is not reminiscent of Zygospira. Acknowledgements This work was initiated during the tenure of a N.E.R.C. Postdoctoral fellowship at the University of Birmingham, and I am indebted to Dr Alwyn Williams, F.R.S., for suggesting the project and for his encouragement at all times. Professor W. T. Dean kindly read the stratigraphical part of the manuscript, and Professor A. D. Wright and Dr L. R. M. Cocks the systematic part, and I am grateful for all their helpful comments. Dr J. S. Peel kindly identified the gastropods, and Dr N. J. Morris and Dr P. W. Bretsky verified the bivalve identifications; Mr S. F. Morris provided the trilobite specimens for photography. Drs V. Havlicek and A. W. A. Rushton kindly sent compara- tive material, and I am indebted to R. Watkins, R. H. Hewitt, M. G. Lockley, J. Owen and C. Wilcox for assistance in the field and laboratory, and to Clare Hurst for much help. The manu- script is published by permission of the Director of the Geological Survey of Greenland. References Alichova, T. N. 1960. [Order Orthida.] In Orlov, U. A. (ed.), Osnovy paleontologii, 7 (Bryozoans, Brachio- pods) : 183-197. Moscow, Akad. Nauk SSSR. Bancroft, B. B. 1928a. On the notational representation of the rib-system in Orthacea. Mem. Proc. Manchr lit. phi I. Soc. 72 : 53-90, pis 1-3. - 19286. The Harknessellinae. Mem. Proc. Manchr lit. phil. Soc. 72 : 173-196, pis 1-2. 1929a. Some new genera and species of Strophomenacea from the Upper Ordovician of Shropshire. Mem. Proc. Manchr lit. phil. Soc. 73 : 33-65, pis 1-2. - 19296. Some new species of Cryptolithus (s.l.) from the Upper Ordovician. Mem. Proc. Manchr lit. phil. Soc. 73 : 67-98, pis 1-2. 1933. Correlation tables of the stages Costonian to Onnian in England and Wales. 4 pp., 3 tables. Blakeney, Glos. (Privately printed.) 1945. The brachiopod zonal indices of the stages Costonian to Onnian in Britain. /. Paleont., Menasha, 19 : 181-252, pis 22-38. 1949. Upper Ordovician trilobites of zonal value in south-east Shropshire (ed. Lament, A.). Proc. Roy. Soc., London, (B) 136 : 291-315, pis 9-11. Barrande, J. 1 879. Classe des Mollusques. Ordre des Brachiopodes. Systeme Silurien du centre de la Boheme, 5. 226 pp., 153 pis. Prague & Paris. Bassett, D. A., Whittington, H. B. & Williams, A. 1966. The stratigraphy of the Bala district, Merioneth- shire. Q. Jlgeol. Soc. Lond. 122 : 219-271. Bergstrom, S. M. 1964. Remarks on some Ordovician conodont faunas from Wales. Ada Univ. lund. (II) 1964(3) : 1-67. Brenchley, P. J. 1969. The relationship between Caradocian volcanicity and sedimentation in North Wales. In Wood, A. (ed.), The Pre-Cambrian and Lower Palaeozoic rocks of Wales : 181-202. Cardiff. Bretsky, P. W. 1970. Upper Ordovician ecology of the central Appalachians. Bull. Peabody Mus. nat. Hist., New Haven, 34. 150 pp., 44 pis. Callaway, C. 1877. On a new area of Upper Cambrian rocks in South Shropshire, with a description of a new fauna. Q. Jlgeol. Soc. Lond. 33 : 652-671, pi. 1. Cave, R. 1965. The Nod Glas sediments of Caradoc age in North Wales. Geol. J., Liverpool, 4 : 279-298, pi. 12. & Dean, W. T. 1959. Four British Ordovician species of Dalmanelloid brachiopod. Palaeontology, London, 1 : 292-297, pi. 53. Cobbold, E. S. 1901. Geology. In Campbell-Hyslop, C. W. (ed.), Church Stretton, 1 : i-ii, 1-115, 5 pis. Shrewsbury. Cocks, L. R. M. 1970. The Silurian brachiopods of the superfamily Plectambonitacea. Bull. Br. Mus. nat. Hist., London, (Geol.) 19 : 139-203, pis 1-17. 1978. A review of British Lower Palaeozoic brachiopods, including a synoptic revision of Davidson's Monograph. Palaeontogr. Soc. (Monogr.), London. 256 pp. CARADOC STRATIGRAPHY AND BRACHIOPODS 297 Cooper, G. A. 1956. Chazyan and related brachiopods. Smithson. misc. Collns, Washington, 127 : 1-1245, pis 1-269. - & Muir-Wood, H. M. 1951. Brachiopod homonyms./. Wash. Acad. Sci., Washington, 41 : 195-196. Cooper, M. R. 1977. Eustacy during the Cretaceous: its implications and importance. Palaeogeogr. Palaeoclimat. Palaeoecol., Amsterdam, 22 : 1-60. Davidson, T. 1866-71. British Silurian Brachiopoda, 3. Palaeontogr. Soc. (Monogr.), London. 397 pp., 50 pis. Dean, W. T. 1958. The faunal succession in the Caradoc Series of south Shropshire. Bull. Br. Mas. nat. Hist., London, (Geol.) 3 : 191-231, pis 24-26. - 1960o. The Ordovician rocks of the Chatwall District, Shropshire. Geol. Mag., London, 97 : 163- 171. - 1960-63. The Ordovician trilobite faunas of south Shropshire. I. Bull. Br. Mus. nat. Hist., London, (Geol.) 4 : 71-143, pis 11-19 (19606). II. he. cit., 5 : 311-358, pis 49-55 (1961). III. loc. cit., 7 : 215- 254, pis 37-46 (1963o). IV. loc.cit.,9: 1-18, pis 1-2(19636). - 1960c. The use of shelly faunas in a comparison of the Caradoc Series in England, Wales and parts of Scandinavia. 21 st Int. geol. Congr., Copenhagen, 7 : 82-87. - 1964. The geology of the Ordovician and adjacent strata in the southern Caradoc district of Shrop- shire. Bull. Br. Mus. nat. Hist., London, (Geol.) 9 : 257-296. Foerste, A. F. 1909. Preliminary notes on Cincinnatian and Lexington fossils. Bull, sclent. Labs Denison Univ., Granville, 14 : 289-334, pis 7-11. Forbes, E. 1848. Memorandum respecting some fossiliferous localities alluded to in Prof. Ramsay and Mr Aveline's paper, as noted on the spot. Q. Jlgeol. Soc. Land. 4 : 297-299. Greig, D. C., Wright, J. E., Hains, B. A. & Mitchell, G. H. 1968. Geology of the country around Church Stretton, Craven Arms, Wenlock Edge and Brown Clee. (Sheet memoir 166.) xiv + 330pp. London (Inst. geol. Sci.). Harper, C. W. & Boucot, A. J. 1978. The Stropheodontacea. Part I : Leptostrophiidae, E' stropheodontidae and Strophonellidae. Palaeontographica, Stuttgart, A 161 : 55-118, pis 1-18. — , - - & Walmsley, V. G. 1969. The rhipidomellid brachiopod subfamilies Heterorthinae and Platyorthinae (New). J. Paleont., Tulsa, 43 : 74-92, pis 15-17. Harper, D. A. T. 1978. The occurrence of the Ordovician brachiopod Heterorthis alternata (J. de C. Sowerby) in the topmost Onnian of the type Caradoc area. Geol. Mag., London, 115 : 301-304. Havlicek, V. 1950. The Ordovician Brachiopoda from Bohemia. Rozpr. ustred. Ust. geol., Prague, 13 : 1-135, pis 1-13. - 1952. O ordovickych zastupcich celedi Plectambonitidae (Brachiopoda). Sb. ustred. Ust. geol., Prague, 19 : 397-428, pis 13-15. - 1967. Brachiopoda of the suborder Strophomenidina in Czechoslovakia. Rozpr. ustred. Ust. geol., Prague, 33. 235 pp., 52 pis. - 1970. Heterorthidae (Brachiopoda) in the Mediterranean province. Sb. geol. Ved. Praha (P) 12 : 7- 39, pis 1-11. 1971. Brachiopodes de 1'Ordovicien du Maroc. Notes Mem. Serv. Mines Carte geol. Maroc, Rabat, 230 : 1-135, pis 1-26. - 1977. Brachiopods of the Order Orthida in Czechoslovakia. Rozpr. ustred. Ust. geol., Prague, 44 : 1-327, pis 1-56. Hints, L. 1975. [Enteletacean brachiopods from the Baltic Ordovician.] 117 pp., 23 pis. Tallinn, Estonian Akad. Nauk. (In Russian, English resume.) Holland, C. H., Lawson, J. D. & Walmsley, V. G. 1963. The Silurian rocks of the Ludlow district, Shrop- shire. Bull. Br. Mus. nat. Hist., London, (Geol.) 8 : 93-171. Holtedahl, O. 1916. Strophomenidae of the Kristiania Region. Skr. VidenskSelsk. Christiania 12. 118 pp., 16 pis. Howe, H. J. 1965. Dalmanellidae from the Montoya Group (Ordovician) of Trans-Pecos Texas. /. Paleont., Menasha, 39 : 235-247, pi. 36. Hurst, J. M. 1978. A phenetic strategy model for dalmanellid brachiopods. Palaeontology, London, 21 : 535-554, pis 57-59. - 1979. The environment of deposition of the Caradoc Alternata Limestone and contiguous deposits of south Shropshire. Geol. J., Liverpool, 14 (1) : 15-40. & Hewitt, R. A. 1977. On tabular problematica from the type Caradoc (Ordovician) of England. NeuesJb. Geol. Paldont. Abh., Stuttgart, 153 : 147-169. Jones, O. T. 1928. Plectambonites and some allied genera. Mem. geol. Surv. Gr. Br. Palaeont., London, 1(5) : 367-527, pis 21-25. 298 J. M. HURST Kemezys, K. J. 1968. Arrangements of costellae, setae and vascula in enteletacean brachiopods. /. Paleont., Tulsa, 42 : 88-93. Lapworth, C. 1916. (In 'Palaeontological Work'.) Mem. geol. Surv. Summ. Progr., London, 1915 : 36-38. - & Watts, W. W. 1894. The Geology of South Shropshire. Proc. Geol. Ass., London, 13 : 297-355. La Touche, J. D. 1884. A handbook of the Geology of Shropshire. 91 pp., 22 pis. London & Shrewsbury. Lindstrom, M. 1953. On the Lower Chasmops beds in the Fagelsang district (Scania). Geol. For. Stockh. Forh.75 : 125-148,1 pi. M'Coy, F. 1852. In Sedgwick, A. & M'Coy, F. A systematic description of the British Palaeozoic fossils in the Geological Museum of the University of Cambridge. 661 pp., 27 pis. London & Cambridge. Melou, M. 1975. Le genre Heterorthina (Brachiopoda, Orthida) dans la formation des Schistes de Posto- lonnec (Ordovicien), Finistere, France. Geobios, Lyons, 8 : 191-208, pis 20-24. Mitchell, W. 1. 1977. The Ordovician Brachiopoda from Pomeroy, Co. Tyrone. Palaeontogr. Soc. (Monogr.), London. 138 pp., 28 pis. Murchison, R. I. 1839. The Silurian System . . . (&c.). xxxii + 768 pp., 36 pis., London. Neuman, R. B. 1967. Some silicified Middle Ordovician brachiopods from Kentucky. Prof. Pap. U.S. geol. Surv., Washington, 583-A : 1-14, pis 1-2. Pocock, R. W. & Whitehead, T. H. 1948. British Regional Geology. The Welsh Borderland (2nd ed.). iv + 83 pp., 11 pis. London (Geol. Survey U.K.). Ramsay, A. C. 1853. On the physical structure and succession of some of the Lower Palaeozoic rocks of North Wales and part of Shropshire. Q. Jlgeol. Soc. Lond. 9 : 161-176. & Aveline, W. T. 1848. Sketch of the structure of north and south Wales. Q. Jl geol. Soc. Lond. 4 : 294-297. Reed, F. R. C. 1910. New fossils from the Dufton Shales. Geol. Mag., London (dec. 5) 7 : 294-329, pis 23-24. 1935. The Lower Palaeozoic trilobites of Girvan. Supplement no. 3. Palaeontogr. Soc. (Monogr.), London. 64 pp., 4 pis. 1952. Revision of certain Ordovician fossils from County Tyrone. Proc. R. Ir. Acad., Dublin, (B) 55 : 29-136, pis 1-5. Salmon, E. S. 1942. Mohawkian Rafinesquinae. /. Paleont., Menasha, 16 : 564-603, pis 85-87. Salter, J. W. & Aveline, W. T. 1854. On the Caradoc Sandstone of Shropshire. Q. Jl geol. Soc. Lond. 10 : 62-73. Appendix: Salter, J. W. Description of some of the fossils, loc. cit. : 73-75. Schiener, E. J. 1970. Sedimentology and petrography of three tuff horizons in the Caradocian sequence of the Bala area (North Wales). Geol. J., Liverpool, 7 : 25-46. Schuchert, C. & Cooper, G. A. 1932. Brachiopod genera of the suborders Orthoidea and Pentameroidea. Mem. Peabody Mus. Yale, New Haven, 4 : 1-270, pis A, 1-29. Sedgwick, A. 1852. On the classification and nomenclature of the Lower Palaeozoic rocks of England and Wales. Q. Jlgeol. Soc. Lond. 8 : 136-168. Sinclair, G. W. 1946. Bancroftina, a new brachiopod name. /. Paleont., Menasha, 20 : 295. 1949. Raymondella, Bancroftina, Reedaspis. J. Paleont., Menasha, 23 : 438^439. Sowerby, J. de C. 1839. See Murchison. Spjeldnaes, N. 1957. The Middle Ordovician of the Oslo Region. 8. Brachiopods of the Suborder Stropho- menida. Norsk geol. Tiddskr., Bergen, 37 : 1-214, pis 1-14. Watkins, R. 1979. Benthic community organization in the Ludlow Series of the Welsh Borderland. Bull. Brit. Mus. not. Hist., London, (Geol.) 31 (3): 175-280. Whittington, H. B. 1938. New Caradocian brachiopods from the Berwyn Hills, North Wales. Ann. Mag. not. ///*/., London, (11) 2 : 241-259, pis 10-11. & Williams, A. 1955. The fauna of the Derfel Limestone of the Arenig District, North Wales. Phil. Trans. R. Soc., London, (B) 238 : 397-430, pis 38-40. Williams, A. 1953. The geology of the Llandeilo district, Carmarthenshire. Q. Jl geol. Soc. Lond. 108 : 177-208. - 1962. The Barr and Lower Ardmillan Series (Caradoc) of the Girvan District, south-west Ayrshire, with descriptions of the Brachiopoda. Mem. geol. Soc. Lond. 3 : 1-267, pis 1-20. 1963. The Caradocian brachiopod faunas of the Bala district, Merionethshire. Bull. Br. Mus. not. Hist., London, (Geol.) 8 : 327-471, pis 1-16. 1969o. Ordovician of the British Isles. In Kay, M. (ed.), North Atlantic - geology and continental drift : a symposium. Mem. Am. Ass. Petrol. Geol., Tulsa, 12 : 236-264. 19696. Ordovician faunal provinces with reference to brachiopod distribution. In Wood, A. (ed.), The Pre-Cambrian and Lower Palaeozoic rocks of Wales : 1 1 7-1 54. Cardiff. 1972. Distribution of brachiopod assemblages in relation to Ordovician palaeogeography. Spec. Pap. Palaeont., London, 12 : 241-269. CARADOC STRATIGRAPHY AND BRACHIOPODS 299 - 1974. Ordovician Brachiopoda from the Shelve district, Shropshire. Bull. Br. Mus. nat. Hist., London, (Geol.) Suppl. 11 : 1-163, pis 1-28. — et al. 1965. Brachiopoda. In Moore, R. C. (ed.), Treatise on Invertebrate Paleontology, H. 927 pp., illustr. Kansas. Strachan, I., Bassett, D. A., Dean, W. T., Ingham, J. K., Wright, A. D. & Whittington, H. B. 1972. A correlation of Ordovician rocks in the British Isles. Spec. Rep. geol. Soc. Lond. 3 : 1-74. — & Wright, A. D. 1963. The classification of the 'Orthis testudinaria Dalman' group of brachiopods. /. Paleont., Menasha, 37 : 1-32, pis 1-2. Wright, A. D. 1963. The fauna of the Portrane Limestone, I. The inarticulate brachiopods. Bull. Br. Mus. nat. Hist., London, (Geol.) 8 : 221-254, pis 1-4. Index New taxonomic names and the page numbers of the principal references are in bold type. An asterisk (*) denotes a figure. acknowledgements 296 Acrotretida, Acrotretidina 229-33 Acton Calcareous Beds 187* Acton Group 187*, 200 Acton Scott 185, 189, 194-5, 197*, 197-8, 204, 206, 218, 223-4, 238, 248, 267 Beds, Actonian 185, 186*, 187*, 188, 189*, 195*, 197, 200-1, 204*, 211-2, 218-9, 225, 228, 240, 242-3, 247 Formation, Group 185, 187*, 189*, 194-8, 195*, 196*, 200, 203, 207-10, 214, 215*, 218-9, 222-3, 225-7, 230, 232-4, 236, 247-9, 256, 260-2, 267, 274, 276, 278, 290, 294, 296 Aegiromeninae 280-1 Alabama 233 Alternata Limestone 187*, 188, 199, 205*, 247-8, 250, 253, 286 Formation 185, 186*, 187*, 189, 189*, 190*, 191, 193, 198, 200, 202*, 203*, 205*, 206-10, 212, 213*, 214, 215*, 216, 218, 221, 234, 244, 248, 267, 274, 278; see Heterorthis alternata ? Ambonychia radiata 208 Ampyx 188 Ampyxella edgelli 204-5, 209, 212, 217* Anglo-Welsh area 219 Annelida 207 Archinacella sp. 203, 209 ? Arjamannia sp. 209 Articulata 233-96 ash falls 219-20 Ashgill 219, 243 Atrypacea, Atrypidae, Atrypidina 295-6 'Avicula' orbicular is 201, 203, 208, 21 1 Bala 187*, 219-20, 228,280 Bancroft, localities of 221-3 notation of 221 zones of 201, 206 Bancroft ina 189, 201, 207, 250, 253-4 gigantea 253 hemtti 1 85, 20 1 , 207, 211, 250, 252, 253-4, 255 * robust a 251, 252-3 typa (& Zone) 200-1, 207, 210, 214, 216, 250, 251*, 252, 253-4 whittingtoni 185, 251*, 252-4 Batch Gutter 196*, 232 bellerophontaceans 203, 209 Beroun Series 269 Berwyn Dome 250 Hills 253 Bicuspina sp. 207, 263*, 274 biostratigraphy 201*, 200-6, 207-10 Birmingham University Museum 220 Birrels Wood Flags 187* Bishop's Castle railway 196*, 221 Bivalvia 201-4, 208-9; see palaeotoxodonts indet. 209, 212 bivariate statistics 220, 223, 261 brachiopods (biostratigraphy) 200, 206-8, 219 faunas 220, 223-96; see under species British Museum (Natural History) 220, 223 Brittany 270 Broeggerolithus 189 globiceps 200, 206 longiceps 200-1, 206, 209-11 transiens 202-3, 206, 209, 211, 217* Brongniartella bisulcata 200, 202, 206, 209-11, 215*, 217*, 218 minor 200, 210 cf. parva 206 Bryngwyn Beds, Hill 253 bryozoa 208; see button, prasoporid, ramose, stick Burrell's Coppice 212, 222 button bryozoans 212 calcareous mudstones 187* Calyptaulax actonensis 203-5, 209, 217* Caradoc (name) 185, 186*, 187*, 188, 189* Sandstone 187*, 188 stages 206 transgression 218 Cardington 189, 198, 200 Carinaropsis sp. 203, 209 300 J. M. HURST Cephalopoda 209 Chasmops 189, 218 extensus 200-3, 205, 209, 212, 217* Chatwall 185, 186*, 189, 198-9, 199*, 205*, 216, 218, 221 Flags 198-9, 205* Group 187* Limestone Formation 185, 199*, 199, 200, 205*, 206, 214, 216, 218 Sandstone 187*, 188, 191, 198, 205* Formation 185, 199*, 199, 206, 214, 216, 218 Chelmick Flags & Shales 187* Cheney Longville 186*, 188-9, 191, 193, 200, 203*, 206, 214, 221-2, 286 Flags 185, 187*, 188-9, 193-4, 198, 205*, 246; see Longville Flags &c. Formation 185, 187*, 189*, 190*, 191, 192*, 193-4, 194*, 195*, 195, 196*, 200, 202*, 203*, 204*, 206-12, 215*, 218, 221-2, 228-9, 244, 249, 253, 262, 278, 286, 290 Chonetes radiatulus 280 Chonetoidea 281 radiatula 203, 207, 21 1, 277*, 280, 281 stoermeri 28 1 ? sp. 2 280 Christ iania hollii 205, 207, 294, 295* portlocki 294 Christianiidae 294 chronostratigraphy 206-7, 210-12 Chuney Pool 195*, 195, 223, 227 Church Farm 197* Church Preen 229 Clathrospira sp. 204, 209 condensed sequences 219 conodonts 220 corals, solitary 210 Corner, The 222 correlation outside Caradoc 219-20 problems within area 212, 214, 216, 218-19 Coston 185 Costonian, Coston Beds 186*, 188, 266 Craniopsidae 227-8 crinoid ossicles 210-11 Cross Fell Inlier 219 Crosspipes Farm 193 Crosspipes Member 185, 187*, 189*, 193-4, 194*, 195*, 195, 196*, 201, 203*, 204*, 206-12, 213*, 215*, 222, 226-7, 230, 234, 236, 239, 247-9, 261,266 Crug Limestone 220 Cryptolithus cobboldi, gracilis, super bus 201 Cryptothyris 189, 254 paracyclica 203-4, 206-7, 212, 218, 254, 255*, 256 Cuneamya cf. miamiensis 203, 208 Cwms 199 Cycloconcha sp. 208 ? cycloconchid 208, 212 Cymbularia sp. 203, 205, 209, 212 Cyrtoceras sp. 203, 209 Cyrtolites sp. 209 Czechoslovakia 269 Dalmanella 189, 211, 248; stratigraphical ranges of species 247* indica Zone 214 lepta (& Zone) 206, 214 multiplicata multiplicata 201-2, 207, 211, 241*, 243-4, 245*, 248-9 prima 185, 207, 244, 245*, 248 unguis ultima 185, 206-7, 245*, 246, 249 unguis 196, 202, 207, 21 1, 244, 245*, 246, 249 Zone, beds 2 11, 249, 282 wattsi (& Zone) 196, 202, 207, 215*, 245*, 247-50, 281-2 Dalmanellidae 243-64, 268 Dandy Hollow 194, 222 Decoroproetus sp. 209, 211 deposit data 220, 223 Destombesium zagoraensis 243 sp. 207, 241*, 243 Dinorthidae 234-6 Dinorthis 272 sp. 207, 234, 235*, 236 Discina (Trematis) punctata 229 Discinacea 229-33 Discinidae 230-3 Dolerorthidae 233-4 Dolerorthis duftonensis 233-4 prolixa 234 virgata 201, 206-7, 212, 233-4, 235* Drabovia ? sp. 207, 241*, 243 Draboviinae 243 Elliptoglossa sp. 207, 224*, 227 Elsaella 267-8, 272 Encrinurus sp. 209 Enteletacea 243-74 Eobronteus sp. 205, 209 Eoplectodonta cf. rhombica 204, 207, 212, 275*, 278, 280 faunal list 207-10 Flexicalymene car ac tad 202-3, 209, 211, 215*, 217* cobboldi 210 (Reacalymene) horderleyensis 200 laticeps 203-4, 210, 212, 217* onniensis 205, 210 salteri 1 \ 0 Gastropoda 203-5, 209-10 indet. spp. 209, 211-12 Gelidorthis sp. 237*, 240, 242 Gelli-grin (Calcareous) Ashes 228, 240, 242, 280 Geological Survey Museum 220 Girvanian 188; see Costonian CARADOC STRATIGRAPHY AND BRACHIOPODS 301 Glenburrel Beds 186*, 189 Glossellinae 226 Glynboro 193 Membdr 185, 187*, 189*, 192*, 193, 200-1 203*, 205*, 207-10, 212, 213*, 221-2 234 247-8, 253, 286 Glyptocrinus Flags 198 graptolites 219 Gravicalymene 2 1 8 inflata 210 cf.praecox 206, 210, 212 Gretton 200, 290 Grier Limestone Member 270 Harknessella 272 Harknessellidae 264-7, 273 Harknessellinae 272 Harnage Shales 188 Harnagian 186*, 188, 266 Hatton 195, 197 Hedstroemina 189, 202, 284, 287-8 fragilis 202-3, 206-7, 211, 287, 288, 289*, 290 parva 288, 289*, 290 robusta 201, 288, 289*, 290 Henley 197 Member 185, 187*, 189*, 197*, 197-8, 200, 206, 214, 215*, 218-9, 248 Heterorthidae 267-74 Heterorthina 267-8, 271-2 fairmontensis 272 kerfornei21Q-\, 271* macfarlani 270-2 morgatensis 268-9, 271 notata 265*, 269, 272* pmeculta 187, 201-2, 207, 265*, 268, 270-2, 270* beds 249, 282 superzone 202 Heterorthinae 268 Heterorthis 189 alternata 187, 191, 193, 199-200, 205-7, 210-1 1, 214, 215*, 218, 267, 268; see also Alternata Limestone barrandei 269 historical review 185-9 Hoar Edge Grits 188 Holtedahlina 290 Hope Bowdler 286 Horderley 188, 246, 286 Flags, Sandstone 187*, 188 Sandstone Formation 185, 186*, 187*, 189*, 189-91, 190*, 199-200, 202*, 203*, 205*, 206-7, 210, 212, 213*, 214, 216, 218, 250, 252-4 Horderleyella 272 cf. plicata 202, 207, 263*, 264, 266 Hough, The 195, 222-3 Howellites 261 Hyolithes sp. 210 lllaenus (Parillaenus) cf. fallax 204-5, 210 sp. 210 Inarticulata 223-33 Killey Bridge Formation 294 Kjaerina 189, 201, 207, 282, 284, 286-8 bipartite (& Zone) 191, 200-1, 207, 210, 215* 253, 283*, 284, 286-7 geniculata 187, 201, 220, 284, 286-7 aff. jonesi 206 typa 201, 207, 211, 284, 285*, 286-7 Kjerulfina 290, 292 broeggeri (broggeri) 292 limbata 291-2 polycyma 201-3, 206-7, 211, 289*, 291-2, 293* trigonal is 201-2, 207, 211, 290-1, 292, 293* Khucekia (Phacopidina) apiculata 200, 210, 217* Kokenospira 204 sp. 203, 209, 212 lectotypes, selection of 220 Leptaena 218 salopiensis 202-3, 206-7, 294 sericea, sericea rhombica 278 Leptaenidae 294 Leptellinidae 275-6 Leptestiina oepiki 204, 275*, 275-6 sp. 206, 275*, 276 Leptestiininae 275-6 Lexington Limestone (N. America) 270-1 Lingulacea 223-8 Lingulella sp. 207, 222*, 225 Lingulellinae 225-6 Lingulida 223-8 Lingulops sp. 207, 224*, 227 Liospira sp. 203, 209 lithostratigraphy 189-200, 201* Llandeilo 187* Llandovery 188, 198 Llansantffraid 253 localities 221-3 Lonchodomas pennatus 204-5, 210, 212, 217* Long Lane 1 89 Longmynd 188 Longville Flags 187*, 201, 207, 210-11, 250 Group 187* Lane 210; see Cheney Longville Longvillia grandis 281 Longvillian 185, 186*, 188, 189*, 198, 201, 202*, 205*, 206-7, 210, 218-20, 228, 240, 250, 267, 280 Lophospira cf. gyrogonia 206 sp. 206, 209 Lytospira sp. 204-5, 209 Marionella 272 typa 272, 274 302 J. M. HURST Marionites 272-4 resupinatus 273-4 typus (typa) 207, 272, 273*, 274 Marshbrook 191, 204, 205*, 212, 214, 218, 221-2, 247, 249 Marshbrookian 185, 186*, 188, 189*, 193, 195*, 201, 203*, 204*, 210, 211, 212, 215*, 219-20, 247, 266, 287 Marshwood (Marsh Wood) 193, 194*, 195, 204, 211,215*, 222, 281-2 May Hill 188 Maysville 272 measurements 220-1, 221* Minton 194*, 222 ? Modiolodon sp. 208 Modiolopsidae 208 Modiolopsis cf. modiolaris 208, 210 sp. A 203, 208, 211 sp. B 203, 205, 208 sp. 208 Monoplacophora 203, 209 Morocco 243 Multicostella 272 muscle field configurations 262, 264, 270-2 New House 202*, 221,252 Nicolella actoniae 203-4, 208, 218, 220, 233 Nuculites planulatus 202-5, 208, 211-12, 215* sp. 208 Oakwood 195*, 195, 197, 223 Obolidae 223-6 Obolinae 223-5 Obolus 223-5 salopiensis 185, 203, 208, 212, 222*, 223-4 subditivus 224 sp. 206, 222*, 225 Oklahoma 226 Onnia 189, 205, 212 cobboldi (& Zone) 201, 205, 210, 212, 217* gracilis (& Zone) 201, 205, 210 superba (& Zone) 201, 205, 210, 217* Onnian, Onnia Beds 185, 186*, 187*, 188, 189*, 198, 201, 205-6, 212, 219 Onnicalymene, see Flexicalymene Onniella 189, 205, 211, 261-2, 264* aspasia 255*, 260-2, 264 avelinei 255* broeggeri, broggeri 205, 208, 212, 256, 258, 261-2, 263*, 264 depressa 204-5, 208, 212, 257*, 258, 259*, 260, 261-2, 264 grandis 201, 257*, 260-2, 264 inconstans 258, 259*, 261-2, 264 reuschi (& Zone) 201-4, 206, 208, 211, 215*, 239, 249, 255*, 257*, 260-2, 264 sinuata 257*, 258, 260-2, 264 sp. 205 Onnioides laevis 207 Onny Shales 187*, 188, 201, 223-5 Shale Formation 185, 187*, 189*, 194, 196*, 198, 205-10, 218-9, 223, 226, 229, 232, 258, 261-2, 267, 281 Valley, River 185, 186*, 188-9, 191, 192*, 193-5, 196*, 197-200, 202*, 203, 205*, 206-7, 210-12, 213*, 214, 216, 218-9, 221-3, 229, 232, 247, 256, 267, 278 Orbicula granulata, punctata 229 Orbiculoidea ovata 185, 208, 230, 232, 232* salopiensis 206 sp. 208, 232, 232* Orbiculoideinae 230-3 Orthacea 233-42 Orthida 233-75 Orthidae 233 Orthidina 233-74 Orthis actoniae 233 alternata 267 bilobata 266-7 calligramma virgata 230 grandis 281 notata 269 unguis 244 vespertilio 266 virgata 233 'Orthoceras"1 pic turn 203, 209 cf. subundulatum 203, 209, 212 Orthotetinae 290 orthocones 203 Oslo 281 Ostracoda 209, 211-12 Otarion sp. 206, 210-11 Palaeoglossa attenuata 225-6 lockleyi 185, 208, 211, 224*, 225-6 Palaeoneilo sp. 208 Palaeonucula sp. 205, 208, 212 palaeotaxodonts 205 Palmovka Hill, Praha-Liben 269 ? Parabasilicus powisii 206 Paracraniops doyleae 185, 201, 206, 208, 211, 227-8, 228* macella 228 Parillaenus, see Illaenus Paterula subcircularis 226 cf. subcircularis 208, 212, 224*, 226 sp. 208, 224*, 226 Paterulidae 226-7 Phacopidina, see Kloucekia Phragmolites sp. 209 phyllocariids 210 Plaesiomyidae 272-3 Plaish 232 Platystrophia sp. 1 208, 235*, 236 sp. 2 208, 235*, 236,238 Platystrophinae 236-8 Platylichas 189, 218 laxatus 204-5, 210, 217* CARADOC STRATIGRAPHY AND BRACHIOPODS 303 Plectambonitacea 275-81 Plectorthidae 236-42 gen. et sp. indet. 208, 237*, 242 Plectorthinae 240-2 pleurotomariaceans, indet. 209 Pomeroy, Ireland 294 Porterfield Pratt Ferry Formation 233 Poterioceras sp. 203, 209 Praearca sp. 209, 212 Praeleda sp. 203, 209 prasoporid bryozoan 202, 206, 208, 210-11, 215* Primaspis 218 caractaci 203, 206, 210-11, 217* Primitia sp. 209 Productorthinae 233 Pseudolingula sp. 208, 224*, 226 Pseudosphaeroxochus sp. 205, 210 Quinny Brook 222 Rafinesquina 287-8, 290 holli 294 robusta 288 Rafinesquininae 282-92 Ragdon 194, 195* Member 185, 187*, 189*, 194-5, 195*, 196*, 197, 203, 207-10, 215*, 219 Ragleth Hill 193, 195*, 195, 197, 211, 213* ramose bryozoan 202, 204, 206, 208, 210-12 Raymondella 250 gigantea 201 robusta 252 typa 250, 252 Reacalymene, see Flexicalymene Remopleurella burmeisteri 205, 210, 217* Remopleurides lot us onniensis 204, 210, 212 Resserella cyclica 256 paracyclica 201, 254, 256 Reuschella 266, 272 bilobata 202-3, 206, 208, 211, 218, 263*, 266-7 horderleyensis 267 semiglobata 263*, 266-7 Rhactorthinae 238-40 Rhactorthis 238, 239 actoniae 185, 206, 215*, 237*, 238, 240 crassa 238, 240 cf. crassa 203, 208, 215*, 237*, 238-9, 240 graiulis 185, 202, 208, 211, 237*, 238, 239-40 melmerbiensis 240 Rhipidomellidae 268, 272 Rock Garden 197* Rose Villa 204, 223 Rostricellula sp. 206 Rushbury 186* Sampo oepiki 275 Schistes de Postolonnec 270-1 Schizocrania hewardi 1 85, 205, 208, 229, 230, 23 1 * salopiensis 208, 21 1, 228*, 229-30 Schizophoridae 243 Schizotreta corrugata 233 spp. 1, 2 208, 232*, 232-3 scolecodonts 207 Sericoidea 189 abdita2%\ homolensis 205, 208, 212, 279*, 280-1 re strict a 281 sp. 205 Shelve district 225, 230 Shineton Shales 188 Similodonta sp. 202-5, 209, 211-12, 215* Sinuites sp. 205-6, 209 Skenidiidae 242 Skenidioides cf. costatus 203, 206, 208, 241 *, 242 Smeathenella 212 Smeathen Wood Beds 186* Soudley 189, 190*, 191, 192*, 193, 199-200, 205*, 212, 213*, 214, 215*, 216, 218, 221-2, 278 Soudleyan 185, 186*, 188, 205*, 214, 216, 218 Soudley Sandstone 188, 191 Sower by el la 205 rhombica 278 sericea 200-2, 206, 208, 210-11, 215*, 218, 220, 277*, 278 soudleyensis 206 Sowerbyellidae 278-81 Sowerbyellinae 278-80 Spiriferida 295-6 Springtown Shale 226 stage definitions, see chronostratigraphy statistical methods 220-1 stick bryozoan 202, 206, 208 Stretton Hills 185, 186*, 198-9 Strophomena 290 bipartita 284 broggeri 292 grandis 201-2, 206, 208, 211, 215*, 281-2, 283* Hollii 294 Strophomenacea 281-94 Strophomenida 275-94 Strophomenidae 281-92 Strophomenidina 275-94 Strophomeninae 281-2, 290 Strophonellidae 292 Svobodaina 268, 271 inclyta 268 swell lags 213*, 215* Swiss Cottage 195* systematic methods 220-1 Tafilaltia 268 Tallinnella scripta 209 Temnodiscus sp. 203, 205, 209, 212 Tentaculites anglicus 201-2, 206-7, 211, 215* Terebratula unguis 244 Ticklerton 197 Tremadoc 188 Trematidae 229-30 304 J. M. HURST Trematis craigensis, millepunctata 229 punctata 189, 200, 206, 208, 210, 214, 228*, 229 Tretaspis ceriodes favus 204-5, 210 Triarthrus cf. linnarssoni 205, 210 Trilobita 200-1, 203-6, 209-10, 219; see under species Trinucleus 188 Trinucleus Shales, Beds 187*, 188 Tripleciacea, Tripleciidae 274 Triplesia sp. 208, 263*, 274 Triplesiidina 274 trochiform gastropods, indet. 209 Trochonema sp. 209 Upper Barn 290 Wattsella multiplicata 243 unguis 187, 201, 244; beds 249 wattsi 187, 201, 247 Wenlock Shales 188 Welsh Basin 219-20 Whittingslow 193, 221-2 Wilderness, The 199 Wistanstow 197, 290 Member 185, 187*, 189*, 195*, 196*, 197, 198, 203*, 207-10, 218-9 Woolston 186*, 191, 193, 206, 221-2 Woolstonian 185, 189*, 191, 193, 202*, 203*, 205*, 206-7, 210-11, 218-20, 247 Wrekin 185 Zahofany Formation 269 Zygospira ? sp. 203, 208, 215*, 289*, 295-6 Zygospirinae 295-6 British Museum (Natural History) Publications related to Geology series Vol. 32 No. 4 Monographs A history of the classification of the phylum Brachiopoda. H. M. Muir-Wood 1955 (reprinted 1968), 124 pp, 12 text figures, 8vo, £7-15 On the morphology and classification of the brachiopod suborder Chonetoidea. H. M. Muir-Wood 1962, v + 132 pp, 16 plates, 24 text figures, 4to, £5-50 Geology Bulletin Supplements Upper Jurassic rhynchonellid brachiopods from North-western Europe. A. Childs Bull. British Museum (Natural History) Geol. Suppl. No 6, 1969, 119 pp, 12 plates, 40 text figures, 4to paper, £7-00 Ordoyician Brachiopoda from the Shelve District, Shropshire. A. Williams Bull. British Museum (Natural History) Geol. Suppl. No 11, 1974, 163 pp, 28 plates, 11 text figures, 110 tables, 4to paper, £13-00 Titles to be published in Volume 32 Miscellanea Palaeoenvironments and correlations of the Carboniferous rocks in west Fermanagh, Ireland. By C. H. C. Brunton & T. R. Mason. The Ordovician trilobite faunas of the Builth-Llandrindod Inlier, central Wales. Part III. By C. P. Hughes. The stratigraphy and brachiopods of the upper part of the type Caradoc of south Salop. By J. M. Hurst. The entire Geology series is now available as a set Printed by Henry Ling Ltd, Dorchester