V r~ A'm'* 'vjws2^ 5 > - s > - ITUTION t/>N0linilISN!^NVIN0SHllWSC/,S3 I H V« 8 HJL I B R AR 1 ES^SMITHSONIAN ^INSTITUTION ^NOIIOJ Jh©i 1NCP1QH THSON § mmy cq ^ o r' o _ j va a ii-* libraries Smithsonian institution NoiifuixsNi nvinqshiii/vs saiavaan libr/ z r~ 2 r ... z "* 2^^ m m W ^ g '^Oia^ m '^L ITUTION^ NOlinillSNl” NVIN0SH1IWS S3 I aVB 9 n~ LI B RAR I ES^SMITHSONIAN_INSTITUTlONwNOIin £ ► -r (/) z to * 2 CO 2 dvaan libraries Smithsonian institution NoiinmsNi nvinoshiiws S3iavaan_LiBR< — if) ~ tn — — or 1^1 < K _ O >2. _ - ___ _ - _l ~ 2 ITUTION2 N0IinilXSNI^NVIN0SHimS^S3 I a va a ll^L I B R AR I ES^SMITHSONIANJNSTITUTIQN^NOlin «<- (/) ZZ C/7 — — jvaan_ libraries Smithsonian institution NoiinmsNi nvinoshiiins^ss i avaa n z»-i br — ,v - — > itution NouniiiSNi NviNosHiiws^sa i a va a n_u Z \ Z ^ '’ ^ SRARIES^SMITHSONIAN INSTITUTION NOim ^ 4/5 L>J o ^ o ^ avaa n""* libraries smithsonian institution NouniiiSNi nvinoshiiws snavaan libr z •“ _ z &S. 33 > to N^oms^x rn ITUTION ^ N0llfUllSNI~NVIN0SHimS^S3 I H V H 9 H ~L I B R A R I E S ^SMITHSONIAN ~ INSTITUTION ^ NQI1I V™T VX E s : h AwffK'A z CO X JVH a n2 Li B RAR I ES^SMITHSONIAN INSTITUTION NOIini!lSNI_NVINOSHilWS S3 I a va a ll_ LI B R O XJ-vos*^ _ O z _j z VIITHSONIAN INSTITUTION NOlifllllSNI NVINOSHXMS S3iavaail LIBRARIES SMITHSONIAN INS z r~ z H 2 m ^ m x C/) _ {/> \ 2 /iNOSHims S3iavaan libraries Smithsonian institution NoiiruiiSNi nvinoshxiws s 3 1 SEoZ?' > ;■ ^ 5 XS^dcJx ^ * Z CO * 2 W *■ Z CO Z /iithsonian institution rtNoiiniusNi_NviNosHiiws rtS3 1 ava a n LI B RAR I ES SMITHSONIAN_INS’ O dCJX' ” q “ X^UusJX' o Xgx. DC^ “ '4fZ"y O /INOSHXMS S3iavaan LIBRARIESZSMITHSONIAN-,|NSTITUTIONZNOIXnXIXSNl'JNVINOSHXMS S3 I V z r- z r- z _ r- > z -1ITHS0NIAN INSTITUTION NOIXflXIXSNI NVINOSHXMS S3iavaail LIBRARIES SMITHSONIAN INS! ^ ^ z co z to z > w z -* 2 l/INOSHXWS^Sa I a va a nZLI B RAR I ESWSMITHSONIAN~INSTITUTION NOlinillSNI NVINOSHXIWS"’ S3 I ,0 - co — m ' ~ - w VIITHSONIAN INSTITUTION NOIXfUIXSNI NV1NOSHXIINS S3IHVaail LIBRARIES SMITHSONIAN INS' z r- z z r- m '±> NiOmstjx m \»dcx ^ m CO \ 2 CO _ CO VINOSHXIINS S3iavaan LIBRARIES SMITHSONIAN INSTITUTION NOIXnXIXSNI NVINOSHXmS S3 1 Z to 2” ... C/3 2 CO < ^ | i i | | | i VIITHSONIAN INSTITUTION NOIXOXIXSNI NVINOSHXIWS^SS I a Va 8 LI B R AR I ES^SMITHSONIAN INS 5 « — . CO ^ CO co xr^i9'wx ^ zTv^irx ^ 2 /? VOLUME 8 • PART 3 Palaeontology OCTOBER 1965 PUBLISHED BY THE PALAEONTOLOGICAL ASSOCIATION LONDON Price £3 THE PALAEONTOLOGICAL ASSOCIATION The Association was founded in 1957 to further the study of palaeontology. It holds meetings and demonstrations, and publishes the quarterly journal Palaeontology. Membership is open to individuals, institutions, libraries, &c., on payment of the appropriate annual subscription: Institute membership . . . . £5. 5s. (U.S. $15.50) Ordinary membership . . . . £3. 3s. (U.S. $9.50) Student membership . . . £2. 2s. (U.S. $6.50) There is no admission fee. Student members will be regarded as persons receiving full-time instruction at educational institutions recognized by the Council; on first applying for membership, they should obtain an application form from the Secretary or the Treasurer. Subscriptions are due each January, and should be sent to the Treasurer, Dr. C. Downie, Department of Geology, The University, Mappin Street, Sheffield 1, England. Palaeontology is devoted to the publication of papers (preferably illustrated) on all aspects of palaeontology and stratigraphical palaeontology. Four parts are published each year and are sent free to all members of the Association. Members who join for 1965 will receive Volume 8, Parts 1 to 4. All back numbers are still in print and may be ordered from B. H. Blackwell, Broad Street, Oxford, England, at the prices shown below (post free): Vol. 1 (for 1957-8) in 4 parts at £2 or U.S. $6.00 per part. Vol. 2 (for 1959) in 2 parts at £2 or U.S. $6.00 per part. Vol. 3 (for 1960) in 4 parts at £2 or U.S. $6.00 per part. Vol. 4 (for 1961) in 4 parts at £2 or U.S. $6.00 per part. Vol. 5 (for 1962) in 4 parts at £3 or U.S. $9.00 per part. Vol. 6 (for 1963) in 4 parts at £3 or U.S. $9.00 per part. Vol. 7 (for 1964) in 4 parts at £3 or U.S. $9.00 per part. A complete set, Volumes 1-7, consists of 26 parts and costs £64 or U.S. $192. Manuscripts on all aspects of palaeontology and stratigraphical palaeontology are invited. They should conform in style to those already published in this journal, and should be sent to Mr. N. F. Hughes, Department of Geology, Sedgwick Museum, Downing Street, Cambridge, England. A sheet of detailed instructions for authors will be supplied on request. © The Palaeontological Association , 1965 OSTRACODA FROM THE SUTTERBY MARL (U. APTIAN) OF SOUTH LINCOLNSHIRE by p. kaye and d. barker Abstract. The ostracod fauna of the Sutterby Marl at its type locality is described and contrasted with faunas of other British Lower Cretaceous horizons. Twenty-five species and subspecies have been found, of which two species and two subspecies are considered new. The basal member of what Swinnerton (1935) has called the Langton Series is a marl which forms a prominent springline in the area around Spilsby. This, the Sutterby Marl, can be seen in the edge of a field near Sutterby (Grid. Ref. TF 726391) where ploughing has cut into a bluff formed by the overlying Carstone. The field has many specimens of the belemnite Neohibolites ewaldi scattered over it. Whilst mapping the area in 1963, J. Newton-Smith, of Leicester University, dug a pit towards the bottom of the field and exposed a mottled yellow-brown marl containing many N. ewaldi. There was a line of phosphatic nodules towards the base. A stiff grey marl which underlay the nodules yielded no fossils, but the upper marls contained os- tracoda when washed down. Samples were also obtained from a pit, higher up the field, dug in April 1964 by a party consisting of Newton-Smith, Kaye, Barker, and others. These later samples have produced the richest Aptian fauna yet found in Britain which is described below. The original sample taken by Newton-Smith is equivalent to the lower of two samples collected in the later excavation. Twenty-five species and subspecies of ostracoda were recorded from the Sutterby Marl, of which two species and two subspecies are con- sidered new. Acknowledgements. The authors are most grateful to Mr. J. Newton-Smith for bringing the project to their notice and for his help in the field work. We are also grateful to certain members of Leicester and Reading University Geology Departments for help in the collecting of the samples. The photo- graphs used in the plates were taken by Mr. J. L. Watkins and text-fig. 1 was drawn by Mrs. J. Lees, both of Reading University Department to whom we are greatly indebted. Thanks are also due to Mrs. M. Kaye who typed the manuscript of this paper. SYSTEMATIC DESCRIPTIONS Order podocopida Suborder podocopina Superfamily cypridacea Family cyprididae Subfamily macrocypridinae Genus macrocypris Brady 1868 Macrocypris parva Kaye 1965a 1 965a Macrocypris parva Kaye, p. 75, pi. 5, figs. 1, 2. [Palaeontology, Vol. 8, Part 3, 1965, pp. 375-90, pis. 48-50.] B 6612 C C 376 PALAEONTOLOGY, VOLUME 8 Material. Three carapaces, BM Io 2712 from the Sutterby Marl, Sutterby, Lines. Remarks. The Sutterby findings extend the range of this species to the Upper Aptian. It has previously been recorded from the Hauterivian/Barremian of Speeton. Family incertae sedis Genus krausella Ulrich 1894 Krausella minuta Triebel 1936 Plate 48, figs. 12, 14 1936 Krausella minuta Triebel in Veen, p. 46, pi. 10, figs. 7-15. 1940 Krausella minuta Triebel; Bonnema, p. 115, pi. 3, figs. 32-34. ? 1946 Krausella minuta Triebel; Bold, p. 67, pi. 2, figs, la, b. Material. Five carapaces and a single left valve from the Sutterby Marl, Sutterby, Lines. BM Io 2694-6. Measurements. Length Height Carapace Io 2694 0-50 mm. 0-30 mm. Left valve Io 2695 0-45 mm. 0-28 mm. Remarks. This small, distinctive species was first described by Triebel from the Lower Cretaceous of Germany, and later by Bonnema (1940) from the Upper Chalk. The valves are small and smooth with the left valve being much larger than the right. Hingement appears to be by simple overlap rather than by a definite tooth arrangement. Family bairdiidae Genus pontocyprella Mandelstam 1956 Pontocyprella rara Kaye 1965a Plate 49, figs. 6-12 1965a Pontocyprella rara Kaye, p. 74, pi. 5, fig. 14. Material. Eighteen specimens from the upper sample and thirteen specimens from the lower sample BM Io 2662-9. Sutterby Marl, U. Aptian, Sutterby. Measurements. Length Height L.V. Io 2662 0-85 mm. 0-45 mm. R.V. Io 2665 0-85 mm. 0-40 mm. Remarks. Pontocyprella rara has only been found before as a few specimens in the basal Lower Aptian ( bodei zone) at Speeton, Yorkshire (Kaye 1965a). It is, however, one of the most abundant species in both of the Sutterby Marl samples. The Sutterby specimens are larger than the Speeton forms and the measurements of the latter approxi- mate to those of the penultimate instars at Sutterby. Full details of the adults of the species can now therefore be given. The major distinguishing features are the median position of the greatest height and greatest width, and the angular posterior end. The ventral margin is straight in the left valves and concave in the right valves. The pro- longation of the antero-dorsal margin so characteristic of the genus is well marked. Internally the most prominent features are the wide anterior and narrow posterior vestibules. Normal pore canals are small but rather abundant and well scattered over the lateral surface. The hinge consists of a long narrow bar in the right valves which KAYE AND BARKER: OSTRACODA FROM THE SUTTERBY MARL 377 fit into a long smooth groove in the left valves. Above the bar in the right valve there is a narrow marginal shelf. The muscle scars form a small rosette below the centre of the valve. They consist of four scars, two elongate anterior scars with two oval scars, one postero-dorsal, and the other posterior of them. Superfamily cytheracea Family cytherideidae Subfamily schulerideinae Genus schuleridea Swartz and Swain 1946 Schuleridea derooi Damotte and Grosdidier 1963 Plate 49, figs. 16, 19-21 1963 Schuleridea derooi Damotte and Grosdidier, p. 154, pi. 1, figs. 4a-/. Material. Thirty specimens from the Sutterby Marl, U. Aptian of Sutterby, BM Io 2673-7. Remarks. This species is abundant throughout the Sutterby Marl and has been recorded from the Lower and Upper Aptian of the Isle of Wight and the Lower Aptian of the Paris Basin. Genus dolocytheridea Triebel 1938 Dolocytheridea minuta Kaye 1963 Plate 48, figs. 15-17 1963c Dolocytheridea minuta Kaye, p. 34, pi. 1, figs. 4-5. 19656 Dolocytheridea minuta Kaye; Kaye, p. 37. Material. Six specimens from the lower sample, Sutterby Marl, Sutterby BM Io 2697-9. Remarks. This species though originally described from the Upper Hauterivian and Lower Barremian at Speeton has also been recorded from the Lower and Upper Aptian of the Isle of Wight and the Lower Aptian of the Paris Basin. It also occurs quite abundantly in the Gault Clay, Middle and Upper Albian, of southern England. Family cytheruridae Genus cytherura Sars 1866 Cytherura reticulosa (Chapman 1894) 1894 Cytheropteron reticulosum Chapman, p. 692, pi. 33, figs. 6a-c. 19646 Cytherura reticulosa (Chapman); Kaye, p. 318, pi. 55, figs. 7, 9. Remarks. This form occurs only rarely at Sutterby but is one of the most characteristic species of lower horizons in the Lincolnshire Lower Cretaceous. Genus dolocythere Mertens 1956 Dolocythere rara Mertens 1956 1956 Dolocythere rara Mertens, p. 192, pi. 10, figs. 33-37; pi. 13, figs. 91-93. 19646 Dolocythere rara Mertens; Kaye, p. 322, pi. 55, figs. 12, 14, 15. Remarks. This species occurs rarely in the upper sample of the Sutterby Marl. 378 PALAEONTOLOGY, VOLUME 8 Genus acrocythere Neale 1960 Acrocy there hauteriviana (Bartenstein) 1956 1956 Orthonotacythere hauteriviana Bartenstein, p. 532, pi. 3, figs. 80, 81. 1960 Acrocythere hauteriviana Bartenstein; Neale, p. 213, pi. 3, figs, la-b, pi. 4, figs. 10, 14. Remarks. A. hauteriviana has only been found rarely in the Sutterby Marl. It is extremely abundant at lower horizons in the ‘Boreal’ Lower Cretaceous of northern England. Genus eucytherura Muller 1894 Eucytherura ornata Kaye 1964a Plate 48, fig. 1 1 1964a Eucytherura ornata Kaye, p. 100, pi. 4, figs. 11-12. Material. Two valves BM Io 2692-3 from the lower sample, Sutterby Marl, Sutterby, Lines. Measurements. Length Height L.V. Io 2692 0-32 mm. 0T7 mm. R.V. Io 2693 0-32 mm. 017 mm. Remarks. This species was only represented before by a single valve from the Barre- mian at Speeton. Its highly distinctive ornament is not comparable to any other species of this, or related, genera. Genus cytheropteron Sars 1866 Subgenus cytheropteron Sars 1866 Cytheropteron (C.) cf. inaequivalve Bonnema 1941 Plate 48, figs. 8-10, 13 1941 Cytheropteron inaequivalve Bonnema, p. 27, pi. 6, figs. 24-28. EXPLANATION OF PLATE 48 All figs. X 66. Figs. 1-4. Cytheropteron ( Infracytheropteron ) lindumensis sp. nov. 1. L.V. (Holotype) lateral view, Io 2678. 2. R.V. (Paratype) lateral view, Io 2679. 3. R.V. (Paratype) lateral view, Io 2680. 4. L.V. (Paratype) lateral view, Io 2681. Figs. 5, 7. Cytheropteron ( Eocytheropteron ) nova reticulata ssp. nov. 5. R.V. (Paratype) lateral view, Io 2683. 7. L.V. (Holotype) lateral view, Io 2684. Fig. 6. Cytheropteron (C.) rugosa Kaye. 6. R.V. lateral view, Io 2686. Figs. 8-10, 13. Cytheropteron (C.) inaequivalve Bonnema. 8. R.V. lateral view, Io 2687. 9. L.V. lateral view, Io 2688. 10. L.V. lateral view, Io 2689. 13. Carapace dorsal view, Io 2690. Fig. 1 1 . Eucytherura ornata Kaye. 1 1 . L.V. lateral view, Io 2692. Figs. 12, 14. Krausella minuta Triebel. 12. Carapace from right, Io 2694. 14. L.V. internal view, Io 2695. Figs. 15-17. Dolocytheridea minuta Kaye. 15. L.V. lateral view, Io 2697. 16. R.V. lateral view, Io 2698. 17. L.V. internal view, Io 2699. Figs. 18-22. Orthonotacythere inversa tuberculata Kaye. 18. R.V. lateral view, Io 2701. 19. L.V. lateral view, Io 2702. 20. L.V. lateral view, Io 2703. 21. L.V. internal view, Io 2704. 22. R.V. lateral view, Io 2705. Figs. 23-25. IStillina cf. fluitans Bonnema. 23. L.V. lateral view, Io 2707. 24. L.V. lateral view, Io 2708. 25. R.V. lateral view, Io 2709. Palaeontology, Vol. 8 PLATE 48 % * I •ts^ • KAYE and BARKER, Lower Cretaceous Ostracoda KAYE AND BARKER: OSTRACODA FROM THE SUTTERBY MARL 379 Material. Eleven valves and one carapace BM Io 2687-91 from the Sutterby Marl, Sutterby, Lines. Measurements. Length Height L.V. Io 2688 0-37 mm. 0-23 mm. R.V. Io 2687 0-37 mm. 0-23 mm. Description. Valves small, elongate; dorsal margin arched in left valves but with weak cardinal angles in the right valves. Anterior margin broadly rounded, posterior margin angled at mid-height. A broad-based ventral alate expansion occurs directed posteriorly and tipped with a small spine. Lateral surface smooth. Duplicature moderately broad, crossed by few thick, straight radial pore canals. Normal pore canals not abundant, well scattered. Hinge crenulate, merodont. Remarks. The Sutterby specimens are almost identical to the Chalk form C. inaequivalve Bonnema (1941) differing in the broad base of the alae. This feature gives the alae a more triangular appearance when viewed dorsally. C. (C.) inaequivalve differs from C. v.-scriptum Veen (1936), C. nannisimum Damotte and Grosdidier (1963), C. reightonen- sis Kaye (1964a) and other Cretaceous species in the lack of surface ornament and the type of alae. Cytheropteron (C.) rugosa Kaye 19656 Plate 48, fig. 6 19656 Cytheropteron (C.) rugosa Kaye, p. 38, pi. 8, figs. 4-5. Material. One right valve BM Io 2686 from the upper sample, Sutterby Marl, U. Aptian, Lines. Remarks. This species is abundant in the Upper Aptian of the Isle of Wight and the Hauterivian and Barremian of Lincolnshire. The Sutterby specimen is identical in all its features. Subgenus eocytheropteron Alexander 1933 Cytheropteron {Eocytheropteron) nova Kaye 1964a ssp. reticulata ssp. nov. Plate 48, figs. 5, 7 Holotype. A left valve BM Io 2684 from the Sutterby Marl, U. Aptian, Sutterby, Lines. Paratypes. Six specimens BM Io 2683-5 from the same sample. Diagnosis. A subspecies of C. ( Eo .) nova Kaye with a row of prominent square reticula- tions along the crest of the alae. Measurements. Length Height L.V. Io 2684 (holotype) 0-62 mm. 0-37 mm. R.V. Io 2683 (paratype) 0-62 mm. 0-37 mm. Description. This subspecies is almost identical with C. {Eo.) nova s.str. from the Haute- rivian and Barremian at Speeton (Kaye 1964a) differing principally in having a prominent row of large square reticulations along the crest of the ventral alate expansion. The Sutterby subspecies is a little larger and has the alate expansion rather more drawn out and not quite as symmetrically rounded. They are almost certainly derived from C. {Eo.) nova s.s. 380 PALAEONTOLOGY, VOLUME 8 Subgenus infracytheropteron Kaye 1964 Cytheropteron {Infracytheropteron) exquisita Kaye 1964 1964a Cytheropteron ( Infracytheropteron ) exquisita Kaye, p. 105, pi. 5, figs. 9-10. Material. Two carapaces BM Io 2711 from the Sutterby Marl, Sutterby, Lines. Cytheropteron ( Infracytheropteron ) lindumensis sp. nov. Plate 48, figs. 1-4 Holotype. A left valve BM Io 2678 from the Upper Aptian, Sutterby Marl, Sutterby, Lines. Paratypes. Six adult valves and one juvenile BM Io 2679-81 from the same sample. Diagnosis. A small smooth species of Cytheropteron {Infracytheropteron) with a strongly arched dorsal margin in the right valve and an asymmetrical lateral alate expansion. Measurements. Length Height Holotype L.V. Io 2678 0-41 mm. 0-25 mm. Paratype R.V. Io 2679 0-42 mm. 0-25 mm. Description. Valves small, laterally compressed. Dorsal margin strongly arched in the right valves, weakly arched in the left valves. Greatest height at one-third length. Anterior margin broadly rounded, posterior margin angled at mid height. Lateral surface smooth, inflated with an alate expansion ventro-laterally. No median sulcus. Alate expansion low, asymmetrical, and weakly directed postero-ventrally. Ventral sur- face smooth. Duplicature fairly broad, crossed by few straight thick radial pore canals, six anteriorly, three posteriorly. Normal pore canals not abundant, concentrated along the crest of the ala. The hinge in the left valve consists of a broad marginal bar which fits into a prominent open-ended furrow in the right valve. Above the median furrow in the right valve is a strong curved marginal bar which fits above the bar of the left valve. The median bar of the left valve has terminal gaps to accommodate the margin of the right valve. Remarks. This species differs from the only other member of the subgenus, C. (/.) exquisita Kaye (1964a), in being larger and having a smooth lateral surface. The hinge, inflation, shape of the dorsal margin and relative inflation of the lateral surface above the alae are the easiest distinguishing features of the species from members of related sub- genera. Genus stillina Laurencich 1957 ? Stillina cf. fluitans (Bonnema) 1941 Plate 48, figs. 23-25 1941 Cytheropteron fluitans Bonnema, p. 27, pi. 6, figs. 29-36. Material. Eleven somewhat fragmentary valves BM Io 2707-9 from the Sutterby Marl, Sutterby, Lines. Measurements. Length Height L.V. Io 2707 0-42 mm. 0-22 mm. Description. This small highly distinctive species has previously only been recorded from KAYE AND BARKER: OSTRACODA FROM THE SUTTERBY MARL 381 the Upper Chalk but one of us (P.K.) has found it abundantly throughout the Gault Clay (M. and U. Albian) of southern England. The valves are very strongly compressed laterally and have a spine-like ventral ala. Posterior to the ala there is a prominent spine lying somewhat beyond the inflated area on the valve margin. It tends to be directed ventrally rather than ventro-laterally as in the case of the ala. The posterior is drawn out into a long upturned caudal process; the anterior margin is strongly denticulate. The duplicature is broad and crossed by few straight radial pore canals. There is in most Chalk and Albian specimens a prominent eye tubercle and keel-like ridge along the dorsal margin. These features tend to be absent in the Aptian specimens and in certain specimens from the higher horizons but may be a dimorphic characteristic. The hinge is merodont to weakly amphidont in the Aptian specimens but is more strongly amphi- dont in later forms. The anterior tooth in the right valve is knob-like whilst the median elements are not strongly divided. Though the external features are identical the hinge differs from that of true Stillina. As, however, the specimens do not fit into any other described genus they are left tentatively in that genus. Fuller description of the more abundant Albian material may finally resolve the difficulties. Genus orthonotacythere Alexander 1933 Orthonotacythere inversa tuber culata Kaye 1963 Plate 48, figs. 18-22 1963e Orthonotacythere inversa tuberculata Kaye, p. 436, pi. 61, figs. 11, 15, 16. Material. Fifteen specimens from the Sutterby Marl, U. Aptian, Sutterby. BM Io 2701-6. Remarks. This form is the youngest member of a morphological sequence of sub- species of Orthonotacythere inversa (Cornuel) 1846 found in the Speeton clay. It occurs in the Upper Barremian at Speeton and its range is extended by the Sutterby findings into the Upper Aptian. The dominant ornamental features of the subspecies are the pronounced ventral longitudinal ribbing and tuberculation, differing from the other sub- species of O. inversa in the lack of vertical costation on the lateral surface. O. inversa tuberculata differs from the species of Orthonotacythere found in the Aptian of the Isle of Wight such as O. atypica Kaye (19656) and O. catalaunica Damotte and Gros- didier (1963) in details of the ornament, particularly the costation. Orthonotacythere sp. B Plate 49, figs. 17, 18 Material. Six valves and fragments from the Sutterby Marl, Upper Aptian, at Sutterby, Lines. BM. Io 2670-2. Measurements. Length Height L.V. Io 2670 at least 0-70 mm. 0-45 mm. Description. A species of Orthonotacythere with a deep vertical median sulcus and a pronounced ventral longitudinal ridge. The valves are rather large and are devoid of reticulation. A weak swelling occurs in the antero-dorsal region, probably representing 382 PALAEONTOLOGY, VOLUME 8 the eye tubercle and associated tubercle but the rest of the lateral surface is smooth. The ventral longitudinal ridge is high and keel-like. It is ventrally arcuate and is thickened in places giving a somewhat fluted effect. It is not tuberculate. Two short, slightly oblique longitudinal ridges run below the major rib on the ventral undersurface. The internal features are identical to other members of the genus. Remarks. This species though represented by few generally broken specimens and therefore not completely described or named is undoubtedly distinct. It shows con- siderable similarities to O. inornata Kaye (19656) from the Upper Aptian of the Isle of Wight but differs in the keel-like nature of the ventral rib and the antero-dorsal tuber- culation. The lack of reticulation and poor tuberculation distinguish it from all other described species of the genus. Family bythocytheridae Genus monoceratina Roth 1928 Monoceratina tricuspidata (Jones and Hinde) 1890 Plate 48, fig. 13 1890 Cytheropteron cuspidatum tricuspidata Jones and Hinde, p. 38, pi. 3, figs. 6, 7. 1936 Monoceratina tricuspidata (Jones and Hinde); Veen, pp. 42, 43, pi. 2, figs. 4-11. 1940 Monoceratina tricuspidata (Jones and Hinde); Bonnema, p. 40, pi. 6, figs. 77-80. 1941 Monoceratina tricuspidata (Jones and Hinde); Triebel, p. 353. 1964c Monoceratina tricuspidata (Jones and Hinde); Kaye, p. 56, pi. 3, figs. 7, 8. Material. Eight specimens and fragments from the Sutterby Marl, U. Aptian, Sutterby, Lines. BM Io 2657-8. Measurements. Length Height R.V. Io 2657 0-62 mm. 0-27 mm. Remarks. This species has previously been recorded only from the Upper Chalk and its range is therefore considerably extended. One of us (P.K.) has, however, specimens of this species from the Cambridge Greensand (U. Albian) in his collections. The Sutterby specimens match the ornament and other features of the Chalk forms exactly. EXPLANATION OF PLATE 49 All figs. X 50. Figs. 1-5. Protocythere mertensi langtonensis ssp. nov. 1. L.V. (Holotype) lateral view, Io 2651. 2. Carapace (Paratype) dorsal view, Io 2652. 3. L.V. (Paratype) lateral view, Io 2653. 4. R.V. (Paratype) lateral view, Io 2654. 5. L.V. (Paratype) internal view, Io 2655. Figs. 6-12. Pontocyprella rara Kaye. 6. L.V. lateral view, Io 2662. 7. Carapace dorsal view, Io 2663. 8. L.V. lateral view, Io 2664. 9. R.V. lateral view, Io 2665. 10. R.V. lateral view, Io 2666. 11. R.V. internal view, Io 2667. 12. L.V. internal view, Io 2668. Fig. 13. Monoceratine tricuspidata (Jones and Hinde). 13. R.V. lateral view, Io 2657. Figs. 14, 15. Neocy there ( Physocythere ) cf. bordeti Damotte and Grosdidier. 14. L.V. lateral view, Io 2659. 15. R.V. lateral view, Io 2660. Figs. 16, 19-21. Schuleridea derooi Damotte and Grosdidier. 16. Female R.V. lateral view, Io 2673. 19. Male R.V. lateral view, Io 2674. 20. Female L.V. lateral view, Io 2675. 21. Male L.V. lateral view, Io 2676. Figs. 17, 18. Orthonotacythere sp. B. 17. R.V. lateral view, Io 2670. 18. R.V. lateral view, Io 2671. Palaeontology, Vol. 8 PLATE 49 r*. 4 . . W •' KAYE and BARKER, Lower Cretaceous Ostracoda KAYE AND BARKER: OSTRACODA FROM THE SUTTERBY MARL 383 Family progonocytheridae Subfamily progonocytherinae Genus neocythere Mertens 1956 Subgenus physocythere Kaye 1963a Neocythere {Physocythere) cf. bordeti (Damotte and Grosdidier) 1963 Plate 49, figs. 14, 15 1963 Centrocythere bordeti Damotte and Grosdidier, pp. 156-7, pi. 2, figs. 8 a-h. Material. Fifteen specimens from the Sutterby Marl, U. Aptian, Sutterby, Lines. BM Io 2659-61. Measurements. Length Height Male L.V. BM Io 2659 0-57 mm. 0-35 mm. Remarks. The Sutterby specimens agree well with Damotte and Grosdidier (1963) forms from the Lower Aptian of the Paris Basin in all details but hingement. The latter authors state that the species has a characteristic Centrocythere type hinge which from Mertens’s (1956) description of the type species should be amphidont with a high step- like anterior tooth and a divided posterior tooth separated by a weakly crenulate furrow in the right valves, a broad accommodation groove being present above the median element in the left valves. In the Sutterby specimens the hinge is merodont with a strongly crenulate median element in the left valve and a wide marginal shelf. The anterior and posterior teeth in the right valve are both strongly subdivided. On this basis the specimens fall better within the subgenus Physocythere than in Centrocythere. Subfamily protocytherinae Genus protocythere Triebel 1938 Protocythere derooi Oertli 1958 Plate 50, figs. 6, 8,9, 11 1958 Protocythere derooi Oertli, p. 1509, pi. 6, figs. 129-43. 19656 Protocythere derooi Oertli; Kaye, p. 44, pi. 6, fig. 10. Material. Six specimens from the lower sample, Sutterby Marl, U. Aptian, Sutterby, Lines. BM Io 2637-41. Measurements. Length Height Female L.V. Io 2633 0-67 mm. 0-39 mm. Female R.V. Io 2632 0-61 mm. 0-33 mm. Remarks. This species, first recorded from the Upper Aptian of SE. France, has also been found in the Upper Aptian of the Isle of Wight. Its most diagnostic feature is the cross-rib joining the median and dorsal longitudinal ribs posteriorly. Protocythere mertensi Kaye 1963 d ssp. langtonensis ssp. nov. Plate 49, figs. 1-5 Holotype. A left valve BM Io 2651 from the upper sample of the Sutterby Marl, Sutterby, Lines. Paratypes. Seven valves and one carapace BM Io 2652-6 from the same sample. 384 PALAEONTOLOGY, VOLUME 8 Diagnosis. A subspecies of Protocythere mertensi being much smaller in size and having a less well developed muscle node than the species sensu stricto. Measurements. Length Height L.V. Holotype Io 2651 0-75 mm. 0-42 mm. R.V. Paratype Io 2654 0-73 mm. 0-35 mm. Description. The specimens of this subspecies are closely similar to P. mertensi s.s. Kaye (1963d) but are smaller and differ in minor features of the ornament. The muscle node is poorly developed in P. mertensi langtonensis whilst the reticulate ornament runs uninterrupted across the crests of the longitudinal ribs. The ventral rib is less inflated and the ventral margin is longer and straighter than in P. mertensi s.s. The dorsal rib is also less convex and the eye tubercle less well developed than in the latter subspecies. Internally the major difference is the small number of radial pore canals (8-9 an- teriorly) in P. mertensi langtonensis compared with the large number (twenty anteriorly) characteristic of P. mertensi s.s. Remarks. This species is obviously very closely related to P. mertensi s.s. which occurs in the Lower Albian ( ewaldi Marl) at Speeton and is presumably ancestral to it. It is also closely similar to the Apto/Albian form P. gaultina Kaye (1963d) but lacks the characteristic anterior marginal ridge. From P. tricostata Triebel (1940) it differs in the smaller size and in the junction of the median and ventral ridges anteriorly. Family trachyleberididae Subfamily trachyleberidinae Genus cythereis Jones 1849 1940 non 1956 Cythereis bekumensis Triebel 1940 Plate 50, figs. 13-16 Cythereis bekumensis Triebel, p. 188, pi. 4, figs. 45-46, pi. 10, fig. 107. Cythereis aff. bekumensis Triebel; Deroo, p. 1518, pi. 4, fig. 58. Material. Six specimens BM Io 2646-50 from the lower sample, Sutterby Marl, Sutterby, Lines. Measurements. Male L.V. Io 2649 Female L.V. Io 2648 Female R.V. Io 2646 Length 0-95 mm. 0-87 mm. 0-87 mm. Height 0-50 mm. 0-50 mm. 0-46 mm. EXPLANATION OF PLATE 50 All figs, x 50. Figs. 1-5, 7. Cythereis sutterby ensis sp. nov. 1 . Male L.V. (Holotype) lateral view, Io 2630. 2. Male R.V. (Paratype) lateral view, Io 2631. 3. Female R.V. (Paratype) lateral view, Io 2632. 4. Female L.V. (Paratype) lateral view, Io 2633. 5. Male R.V. (Paratype) internal view, Io 2634. 7. Male L.V. (Paratype) internal view, Io 2635. Figs. 6, 8, 9, 11. Protocythere derooi Oertli. 6. L.V. lateral view, Io 2637. 8. L.V. lateral view, Io 2638. 9. R.V. lateral view, Io 2639. 1 1 . R.V. lateral view, Io 2640. Fig. 10. Cytherella ovata (Roemer). 10. R.V. lateral view, Io 2642. Fig. 12. Cytherelloidea cf. ovata Weber. 12. R.V. lateral view, Io 2644. Figs. 13-16. Cythereis bekumensis Triebel. 13. R.V. lateral view, Io 2646. 14. R.V. lateral view, Io 2647. 15. L.V. lateral view, Io 2648. 16. L.V. lateral view, Io 2649. Palaeontology, Vol. 8 PLATE 50 KAYE and BARKER, Lower Cretaceous Ostracoda KAYE AND BARKER: OSTRACODA FROM THE SUTTERBY MARL 385 Remarks. The major distinguishing features of this species are the strong lateral com- pression, the prominent muscle node and short weak median rib. The intercostal areas are weakly reticulate whilst the anterior marginal rib is well marked. The dorsal and ventral longitudinal ribs are keel-like, but postero-ventral inflation is very weak. C. bekumensis is similar to C. bartensteini Oertli (1958) but is larger and less inflated and not as strongly ornamented. It differs from C. geometrica s.s. Damotte and Gros- didier (1963) in being larger, reticulate, and having the ribs keel-like rather than rounded. Cythereis sutterbyensis sp. nov. Plate 50, figs. 1-5, 7 Holotype. A male left valve BM Io 2630 from the upper sample, Sutterby Marl, U. Aptian, Sutterby, Lines. Paratypes. Five valves and one carapace BM Io 2631-6 from the same sample. Diagnosis. A large species of Cythereis with heavily calcified valves. Median rib short, spined, separated from the prominent muscle node. Lateral surface weakly reticulate. Measurements. Holotype male L.V. Io 2630 Paratype female L.V. Io 2633 Paratype male R.V. Io 2634 Paratype female R.V. Io 2632 Length 1T2 mm. 105 mm. 1-07 mm. 100 mm. Height 0-60 mm. 0-64 mm. 0-59 mm. 0-59 mm. Description. Valves large, very strongly built. Dorsal and ventral margins straight con- verging posteriorly. Cardinal angles well marked, greatest height at a quarter length. Anterior margin broadly rounded ; posterior triangular, angled ventrally. Lateral surface moderately inflated and weakly reticulate. Dorsal and ventral ridges prominent and wrinkled. Median ridge very short with a series of spines on its crest, separated anteriorly from the high, ridged muscle node. Anterior and posterior marginal ridges well marked and bearing small tubercles. Eye tubercle prominent, anterior margin denticulate. Interior of valves characteristic of genus. Hinge strongly developed, terminal teeth in right valve, high with subdivisions on crest. Remarks. The most obvious characteristics of this species are its large size and strong build. It is closest in arrangement of the ornament to C. geometrica fittoni Kaye (19656) differing principally in the greater prominence of the ribs and reticulation. Suborder platycopina Family cytherellidae Genus cytherella Jones 1849 Cytherella ovata (Roemer) 1 840 Plate 50, fig. 10 1840 Cytherina ovata Roemer, p. 104, pi. 16, fig. 21. 1849 Cytherella ovata (Roemer); Jones, p. 28, pi. 7, figs. 24 a-g. 1890 Cytherella ovata (Roemer); Jones and Hinde, p. 44, pi. 3, figs. 48-54. 1956 Cytherella ovata (Roemer); Deroo, pp. 1508, 1523, pi. 1, figs. 4-6. 1958 Cytherella ovata (Roemer); Oertli, p. 1502, pi. 1, figs. 10-29. 386 PALAEONTOLOGY, VOLUME 8 Material. Twenty-three specimens from the lower sample, Sutterby Marl, Sutterby, Lines. BM Io 2642-3. Measurements. Length Height R.V. Io 2642 0-90 mm. 0-57 mm. Remarks. This well-known Aptian /Albian species is particularly abundant in the lower sample at Sutterby but is rather rare in the upper one. The specimens correspond exactly with those found in the Albian both at Speeton and in southern England. Genus cytherelloidea Alexander 1929 Cytherelloidea cf. ovata Weber 1934 Plate 50, fig. 12 19636 Cytherelloidea cf. ovata Weber; Kaye, p. 116. Material. Two pre-adult specimens from the upper sample, Sutterby Marl. BM Io 2644-5. Remarks. These two juvenile specimens are undoubtedly identical to the forms described as C. cf. ovata from the basal Aptian at Speeton. They have a weaker dorsal rib, and a more arcuate median rib than true C. ovata Weber (1934) and also have the ventral rib poorly connected posteriorly. CONCLUSIONS The Sutterby Marl is quite rich in ostracoda, twenty-five species and subspecies being represented. Twenty of these species have been recorded in the Cretaceous elsewhere in the British Isles, two species and two subspecies being new, one species being left un- named. The Sutterby fauna is closely similar to the pre- Aptian ‘Boreal province’ faunas found at Speeton, Yorkshire and from Lincolnshire. Of the twenty known species in the Sutterby Marl, twelve are known from these ‘Boreal’ pre- Aptian deposits. Simi- larities to the top Barremian and basal Aptian faunas at Speeton are very close with characteristic species such as O. inversa tuberculata and Pontocyprella rara being par- ticularly well represented. A correlation between faunas in the ‘Boreal’ province is only to be expected but more striking is the dissimilarity between the fauna of the Sutterby Marl and the ewaldi Marl (L. Albian at Speeton). No species are common to these two units which have been regarded as equivalent by previous authors. Geographical con- siderations could perhaps account for this but the close similarities between the Sutterby and earlier Speeton faunas seems to discount it. Ecological differences are possible but again unlikely between two very similar clay horizons. A time as well as space difference seems the most likely explanation plus perhaps uplift along the Market Weighton axis in Aptian times (see Kaye 1964 d). Though having essentially a ‘ Boreal’ fauna the Sutterby Marl shows some similarities to the Aptian of southern England. Eight species recorded at Sutterby are also present in the U. Aptian of southern England; three of these being found in the Bargate Beds near Guildford and six in the Upper Aptian of the Isle of Wight. These similarities add weight to the theory of connexion of the ‘ Tethyan ’ and ‘ Boreal’ seas during the Aptian through the English Midlands (Kaye 19646, 19656). The three species common to the Sutterby Marl and Bargate Beds of Surrey, Cytherura reticulosa KAYE AND BARKER: OSTRACODA FROM THE SUTTERBY MARL 387 (Chapman), Dolocythere rara Mertens, and Acrocythere hautermana (Bartenstein) are all well-known ‘Boreal’ forms and seem to indicate a southward migration in Upper Aptian times. The Isle of Wight Aptian has six species in common with the Sutterby Marl. The three species found in the Lower Aptian are also found in the Upper Aptian in the island and there is therefore no evidence of a Lower Aptian connexion of the two seas. Three of the six U. Aptian species, common to both the Isle of Wight and Sutterby : i.e. Schuleridea derooi, Neocy there ( Ph .) bordeti, and Protocythere derooi, are ‘ Tethyan ’ forms well known from the Aptian of the Paris Basin and indicate northward migration whilst the remaining three species Dolocytheridea minuta, Dolocythere rara, and Cytheropteron (C.) rugosa are ‘Boreal’ species and indicate southward migration. Dolocytheridea minuta is somewhat anomalous for though being a characteristic Hauterivian and Barremian ‘Boreal’ form it is also known from the Lower Aptian of the Isle of Wight and the Paris Basin. Some migration route must therefore have been open to it in Lower Aptian times, presumably through S. Germany. Interchange of species does seem conclusive in the Upper Aptian but full mixing of the faunas is not likely, only a few forms having penetrated from one province to another. Ecological considerations and relative competition can perhaps account for this. The Sutterby fauna does not persist into the Gault seas, only five Sutterby species being found in the Albian. Five species are also common to the Sutterby Marl and the Chalk. Three of these species Cytheropteron (C.) inaequivalve, IStillina cf. fluitans and Monoceratina tricuspidata are characteristic Chalk forms though the latter two have been recently found by one of the authors (Kaye) in the Gault clay. This great extension of the range of these species is rather striking but the alate nature of all three may point to some ecological factor. Alate or ventrally expanded species are very common in the Sutterby Marl, with members of genera such as Cytheropteron, Monoceratina, Or thonotacy there, and Cythereis being particularly prominent. They also form a large proportion of Chalk ostracod fauna. Slight differences do occur between the two samples taken in the Sutterby Marl. The upper sample is richer in species and numbers than the lower one and in general most of the species found rarely in the lower one are more abundant in the upper one. The most distinctive differences between the two samples are the great abundance of Cytherella ovata in the lower sample when it is rare in the upper sample and also the absence of Cythereis bekumensis and Protocythere derooi from the upper sample. The latter two species seem to be replaced at the higher horizon by Cythereis sutterbyensis and Protocythere mertensi langtonensis which are absent from the lower sample. Correlation of the Sutterby Marl with other Cretaceous horizons in Britain is difficult and its exact zonal horizon is not clear. The bulk of the Marl is Upper Aptian whilst the phosphate nodule bed contains ammonites indicating many Lower Aptian horizons (Casey 1961, pp. 570-1). The Marls above contain ammonites of the genus Colombiceras, only known elsewhere in this country from the Aptian of Upware which is itself of ques- tionable horizon. It is likely that the basal Upper Aptian in southern England is the time equivalent of the Sutterby but the difference in lithology and difference in faunal pro- vince make exact correlation difficult. The Aptian in Yorkshire is very thin and the greatest similarities are between the Sutterby Marl and the pre- Aptian (plus basal Aptian bodei zone) at Speeton, rather than with the ewaldi Marl. The latter deposit has, however, only proved fossiliferous in its upper few feet where a lower Albian microfauna quite 388 PALAEONTOLOGY, VOLUME 8 SPECIES UPPER APTIAN SUTTERBY, LIMCS UPPER APTIAN SURREY LOWER APTIAN ISLE OF WIGHT UPPER APTIAN ISLE OF WIGHT PRE APTIAN N. ENGLAND ALBIAN ENGLAND POST ALBIAN BRITISH ISLES Cytherella ovata (Roemer) X X X Cytherelloidea cf. ovato Weber X X Macrocypris parva Kaye X X Pontocyprella rora Kaye X X Krausella minuta Triebel X X X Schuleridea derooi D.8G. X X X Dolocytheridea minuto Kaye X X X X X Cytherura reticulosa (Chapman) X X X Eucytherura ornata Koye X X Cytheropteron (C) inaequivalve Bonnema X X Cytheropteron (C) rugosa Kaye X X X C (Eocytheropteron) nova reticulata sp.nov. X C(lnfracytheropteron) exquisita Kaye X X C.(lnfracytheropteron) lindumensis spnov. X P Stillina cf. fluitans Bonnema X X X Dolocythere rara Mertens X X X X 0. inversa tuberculota Kaye X X Orthonotacythere sp. B. X Monoceratina tricuspidata (Jones 8 Hinde) X X X Neocythere (Ph.) cf. bordeti (DBG) X X X Acrocythere hauteriviana (Bartenstein) X X X Protocythere derooi Oertli X X Protocythere mertensi langtonensis sp.nov. X Cythereis bekumensis Triebel X X Cythereis sutterbyensis sp. nov. X Known distribution of species of Ostrocodo found in the Sutterby Marl TEXT-FIG. 1. KAYE AND BARKER: OSTRACODA FROM THE SUTTERBY MARL 389 unlike the Sutterby one is found. The Sutterby Marl is, therefore, either absent at Speeton or represented in the barren lower layer of the ewaldi Marl. Neohibolites ewaldi though giving its name to the lithological unit is never common at Speeton but is very abundant at Sutterby and it is by no means certain that the specimens found at the two localities are conspecific. The absence of the Carstone at Speeton and its prominence overlying the Sutterby Marl in Lincolnshire may indicate geographical separation of the two areas possibly due to uplift along the Market Weighton axis. The relationship of the ewaldi Marl to the Carstone is problematical, particularly as the Carstone itself is diachronous and it is likely that the upper fossiliferous portion of the ewaldi Marl is equivalent to part of the Carstone. The Carstone has, however, yielded a Middle Albian ostracod fauna at Melton in South Yorkshire and may be represented in part by the ‘Greensand Streak’ at Speeton and basal sands at West Heslerton (see Kaye 1964 d). The absence of the Gault and the thin Red Chalk in Lincolnshire may bear out this suggestion. The distribution and range of the ostracoda found at Sutterby are shown on text-fig. 1 . REFERENCES bartenstein, H. 1956. Zur Mikrofauna des englischen Hauterive. Senckenbergiana leth. 37, 509- 33, pi. 1-3. bonnema, j. h. 1940-1. Ostracoden aus der Kreide des Untergrundes der nordostlichen Niederlande. Natuurh. Maandbl 27, 91-95, 104-8, 115-18, 129-32, pi. 1-4, 28, 8-10, 21-24, 26-29, 40-43, 56-60, 70-72, pi. 5-7. bold, w. a. van den. 1946. Contribution to the study of Ostracoda with special reference to the Tertiary and Cretaceous microfauna of the Caribbean region. Diss. Univ. Utrecht ; J. H. De Bussy, Amsterdam; pp. 1-157, pi. 1-18. casey, r. 1961. The stratigraphical palaeontology of the Lower Greensand. Palaeontology, 3, 487 - 621, pi. 77-84. damotte, r. and grosdidier, e. 1963. Quelques ostracodes du Cretace Inferieur de la Champagne Humide. 2 — Aptien. Rev. Micropaleont. 6, 153-68, 3 pis. deroo, g. 1956. Etudes Critiques au sujet des Ostracodes marines de Cretace Inferieur et Moyen de la Champagne Humide et du Bas Boulonnais. Rev. Inst, front;. Petrole, 11, 1499-535, pi. 1-5. jones, t. r. 1849. A Monograph of the Entomostraca of the Cretaceous Formation of England. | Palaeontogr. Soc. [Monogr.], 40 pp., 7 pis. jones, t. r. and hinde, g. j. 1890. A supplementary Monograph of the Cretaceous Entomostraca of England and Ireland. Palaeontogr. Soc. [Monogr.], 77 pp., 4 pis. | kaye, p. 1963a. The ostracod genus Neocy there in the Speeton Clay. Palaeontology, 6, 274-81, pi. 41. 19636. Species of the Ostracod Family Cytherellidae from the British Lower Cretaceous. Senckenbergiana leth. 44, 109-25, pi. 18-20. 1963c. The interpretation of the Mesozoic Ostracod genera of the family Cytherideidae Sars I 1925. Rev. Micropaleont. 6, 23-40, 3 pis. ! 1963 d. Ostracoda of the subfamilies Protocytherinae and Trachyleberidinae from the British Lower Cretaceous. Paldont. Z. 37, 225-38, pi. 18, 19. I 1963c. The ostracod species Orthonotacythere inversa (Cornuel) and its allies in the Speeton Clay of Yorkshire. Palaeontology, 6, 430-9, pi. 61. jj 1964a. Ostracoda of the genera Eucytherura and Cytheropteron from the Speeton Clay, Geol. Mag. 101, 97-107, pi. 4, 5. 19646. Revision of the ostracoda from the Bargate Beds in Surrey. Palaeontology, 7, 317-30, pi. 54, 55. 1964c. Revision of British Marine Cretaceous Ostracoda with notes on additional forms. Bull. Brit. Mus. (Nat. Hist.) Geology, 10, 35-79, pi. 1-9. 390 PALAEONTOLOGY, VOLUME 8 kaye, p. 1964 d. Observations on the Speeton Clay (Lower Cretaceous). Geol. Mag. 101, 340-56, text- figs. 1-6. 1965a. Further ostracoda from the British Lower Cretaceous. Senckenbergiana leth. 46, 73-82, pi. 5. 19656. Ostracoda from the Aptian of the Isle of Wight. Paldont. Z. 39, 33-50, pis. 6-8. mertens, e. 1956. Zur Grenzziehung Alb/Cenoman in Nordwestdeutschland mit Hilfe von Ostra- coden. Geol. Jb. 72, 173-230, pi. 8-14. neale, j. w. 1960. Marine Lower Cretaceous Ostracoda from Yorkshire, England. Micropaleontology, 6, 203-24, pi. 1-4. oertli, H. j. 1958. Les Ostracodes de l’Aptien-Albien d’Apt. Rev. Inst, frang. Petrole, 13, 1499-537, pi. 1-9. roemer, f. A. 1840. Die Versteinerungen des Nordeutschen Kreidegebirges. 145 pp. 16 pis. Hannover. swtnnerton, h. h. 1935. The Rocks below the Red Chalk of Lincolnshire and their cephalopod faunas. Quart. J. geol. Soc. Lond. 91 , 1-46. triebel, e. 1940. Die Ostracoden der Deutschen Kreide. Ill Cytherideinae und Cytherinae aus der Unteren Kreide. Senckenbergiana leth. 22, 160-227, pi. 1-10. 1941. Zur Morphologie und Okologie der Fossilen Ostracoden. Senckenbergiana leth. 23, 294- 400, 15 pis. veen, j. e. 1936. Die Cytheridae der Maastrichter TufFkreide und des Kunrader Korallenkalkes von Sud-Limburg. Ill Die Gattungen Loxoconcha, Monoceratina, Paracytheridea, Xesteloberis, Cythe- ropteron und Cytherura. Natuurh. Maandbl. 25, 21-113, pi. 1-4. weber, h., 1934. Ostracoden aus dem Hauterive von Wenden am Mittelland-Kanal. Jber. niedersachs. geol. Ver. 26, 139-49, pi. 8-9. P. KAYE, Burmah Oil Exploration Co., 20 Esplanade, Scarborough, Yorks. d. barker, Paleoservices Ltd., 162 High St., Watford, Herts. Manuscript received 18 August 1964 UPPER JURASSIC AND LOWER CRETACEOUS MICROFOSSILS FROM THE HAUTES-ALPES by JUDITH TURNER Abstract. Common pelagic (Tethyan) microfossils from the Upper Jurassic and Berriasian rocks of La Faurie, Hautes-Alpes, France, are described and illustrated, and their stratigraphical distribution compared with that of similar assemblages elsewhere. The fossils include crinoids, tintinnids, radiolarians, and incertae sedis such as Globochaete, Cadosina, and Stomiosphaera. Lombardia Bronnimann is considered to be a synonym of Sacco- coma Agassiz. I n the department of the Hautes-Alpes, in south-eastern France, near the village of La Faurie, the Kimmeridgian, Tithonian and Berriasian stages are represented by fine- grained calcareous rocks varying from pure limestones to marls (text-fig. 1). Coccoliths may form a large part of the ground-mass of the rock. In the Kimmeridgian the pseudo- mutabilis Zone and the basal part of the beckeri Zone consist of dark marls up to 200 m. thick. Above are up to 100 m. of marly limestone and limestones, with occasional chert nodules, forming the main part of the beckeri Zone and of the Tithonian. Conglomerates, probably originating as the result of movement of unconsolidated sediments in mud- flows, are common in the Lower Tithonian, and are of extremely variable thickness ; they usually give rise to prominent topographic features. The Berriasian is formed of alter- nating marls and marly limestones. The beds are sporadically rich in ammonites, including aptychi, and belemnites and brachiopods may occur. The marls below the marly limestones of the beckeri Zone and the Lower Tithonian are particularly poor in macrofossils. Microfossils, however, are common in most beds, and various assemblages occur which can be used for correlation locally and which may eventually prove to be useful over a wide area of Southern Europe. References to the forms which occur are widely scattered and illustrations sometimes inadequate so that it has been thought useful to bring them together in one article. The commonest microfossils are : plates of small crinoids of the family Saccocomidae recovered from washings from marls of the Kimmeridgian and, seen in thin section only, tintinnids such as Calpionella and Tintinnopsella, radiolarians, and various incertae sedis such as Globochaete, Cadosina, and Stomiosphaera. The specimens figured are deposited in the Hunterian Museum, Glasgow. Phylum ECHINODERMATA Class CRINOIDEA Order roveacrinida Sieverts-Doreck 1953 Family saccocomidae d’Orbigny 1852 Genus saccocoma Agassiz 1835 [= Lombardia Bronnimann 1955] Description. Angular, usually bilaterally symmetrical shapes between 50/x and 150/z in length and formed of a single calcite crystal, occur in large numbers in thin sections of the Kimmeridgian and the lowermost Tithonian (PI. 51, fig. 2). There is now no doubt, [Palaeontology, Vol. 8, Part 3, 1965, pp. 391-6, pis. 51-52.] B 6612 D d TITHONIAN 392 PALAEONTOLOGY, VOLUME 8 since the work of Verniory (1954, p. 327), that these are random sections through the brachial plates of small, pelagic crinoids of the family Saccocomidae, although they have been described as various other echinoderm fragments, algae or sponge spicules (Lombard 1938, 1945; Bronnimann 1955). Various types of brachial plates are found in washings of the Kimmeridgian marls which correspond exactly to the description of Saccocoma given by Jaekel (1892). UPPER B MARLY LIMESTONE Z CO _ CZ> TO , c). It may be mentioned that the last mentioned species was at first included by its author (Adkins 1920) in Hamulina. Later it was separated by the same author (Adkins 1928) in a new genus, Worthoceras, of ‘uncertain position’. Both species are restricted to the Upper Albian of Texas, where Eoscaphites is unknown up to the present. The distinct trifid outline of L, the undivided element U, the general character of the quadrilobate suture-line (text- fig. 10), and the originally flat and not impressed dorsum (PI. 59, fig. 7a), make a hetero- morph ancestor of Worthoceras evident. The ptychoceratids especially have nearly a world-wide distribution in the Albian, where they are represented by unsculptured dwarfed forms with trifid lateral lobes. Of further importance for this presumed attri- bution and the separation from Eoscaphites is the fact that the phragmocone in early Worthoceras does not coincide with the spiral portion, as it does in Eoscaphites. Generally almost all of the straight shaft is septate, as visible in the reproduced type material (PI. 59, figs. 5-7). Moreover, no umbilical perforation is known from Worthoceras. In Europe, the centre of the scaphitid development, only one worthoceratid species is known: W. rochatianum (d’Orbigny) of Cenomanian and Turonian age (PL 60, figs. 4-6). This species is barely distinguishable from the Albian W. platydorsum (PI. 59, fig. 5). Even in regard to the sculpture, when it is preserved (PI. 60, figs. 5b, 6a), it agrees with the forms of the Texan Albian. The suture-lines were found to be in accordance also; likewise in W. rochatianum (text-fig. 10z/) they contain a distinctly trifid L and slightly enlarged saddles. The shaft is likewise septate (PI. 60, fig. 4b). The only noticeable distinction is the dorsal impression (PI. 60, fig. 4a), which is much more pronounced og B 6612 440 PALAEONTOLOGY, VOLUME 8 than in the preceding species. It can perhaps be regarded as a reminiscence of the ptychoceratid ancestor, where the smaller shaft generally remains strongly impressed in the dorsum of the larger one. text-fig. 10. Worthoceras. a-c, W. worthense (Adkins), a. The last suture-lines of the holotype, l! BEGA 20994; Upper Albian, Fort Worth, h — 3 mm. b. Last suture-lines, and c. Septal surface of I hypotype, BEGA 20995; same locality, h = 2 mm. d, W. rochatianum (d’Orbigny), external suture of Madagascan hypotype (ex Collignon 1929, text-fig. 36); Cenomanian, Diego-Suarez. x 8. e-g, W. vermiculum (Shumard). e. Last suture of neotype (Moreman 1942), BEGA 19827; Lower Turonian, Britton, h = 3 mm. f Last suture, and g, Septal surface of hypotypoid BEGA 35390; Lower Turonian, Dallas County, h = 3 mm. h, W. gibbosum Moreman, holotype, BEGA 19812; I Lower Turonian, Britton; external suture at h = 4-5 mm. Likewise the following species, the Lower Turonian W. vermiculum (Shumard) (here PI. 59, fig. 8; PI. 60, figs. 1, 2), is barely distinguishable from the forms described above. The spiral portion remains open whorled, the suture-line (text-figs. 10c, /) quadrilobate, the shell smooth and the final hook flat-sided and tender. In the septal J. WIEDMANN: ORIGIN, LIMITS, AND POSITION OF SCAPHITES 441 surface (text-fig. 10g) the simple configuration of the ptychoceratids is preserved. The differences concern the somewhat stouter final crozier, the shaft which is septate only in its lower part, the nearly bifid L, and especially the lateral mouth lappets perceptible only in this species. The attribution of W. gibbosum Moreman of the Texan Lower Turonian to the present genus seems more difficult. The holotype, here reproduced (Plate 60, fig. 3, text-fig. 10 h), has the general appearance of a true Scaphites, with a stout and inflated final hook which nearly touches the moderately involute spiral portion. The latter coincides with the septate whorls. Finally the suture, with its distinct bifid L, seems to belong to a true Scaphites. Therefore W. gibbosum was interpreted as a transitional form between Wortho- ceras and Scaphites minutus Moreman. As indicated above, the latter species is a true Scaphites indeed, belonging to a group of dwarfed (and lappeted) forms which are separated from the equalis main stock as a distinct subgenus Pteroscaphites. After study of W. gibbosum we come to the following conclusion, that this species in all its characters belongs to Worthoceras and cannot be connected with Sc. minutus nor with any other known Scaphites. If we compare the suture-line (text-fig. 1 1 h), the septal surface (PI. 60, fig. 3 a) and the ornamentation of the spiral coil (PI. 60, fig. 3b) with that of its presumed descendant Sc. minutus (PI. 58, figs. 5 d, e\ text-fig. 6b) or with the otoscaphitid type species (PI. 58, figs. 2-4; text-figs. 7, 8), then the fundamental differences become obvious. Unfortunately the internal suture-line is not exposed in the Texan forms, but it is possible to interpret it from the external portion and septal surface. As mentioned many times, all scaphitid micro- morphs are characterized through the position of the subdivided U on the umbilical seam and the importance of the median incision p. In W. gibbosum we find nothing comparable. Alone a small undivided U can be recognized, which remains in the internal suture, and the median incision of LjU does not increase in size. The initial coil of Sc. minutus is nearly closed and passes (as in all other Ptero- and Otoscaphites) through an inflated pachydiscoid stage. The initial whorls of W. gibbosum remain well exposed and completely smooth as in all other Worthoceras species. Finally no lateral lappets can be observed on the mouth-border, which, on the other hand, are well developed in Ptero- and Otoscaphites. It is therefore impossible to connect Worthoceras with any scaphitid member despite their striking homoeomorphy. This is less surprising if we regard the phylogenetic position of Worthoceras. Its very slow evolutionary trend, which we intentionally demonstrated here in all details, makes it evident that the genus represents a phylogenetic end form. Scaphites is able to compete with the contemporaneous ammonites through genetic regeneration and the development of a high number of lateral offshoots, whereas Worthoceras is evidently unable to regenerate and thus does not pass the Turonian/Coniacian boundary. Regard- ing these reasons it might be better to place Worthoceras into Ptychoceratinae, as re- cently defined by me (Wiedmann 19626). The measurements (in millimetres) of some Worthoceras here represented are : Worthoceras rochatianum D B d h b u MHNG Wi ‘Sc 710 20 4 6-8 2-8 (0-41) 2-5 (0-37) 2-6 (0-38) MHNG Wi ‘Sc ’/ll — — 70 3-0 (0-43) 2-5 (0-36) 3-0 (0-43) MHNG Wi ‘Sc ’/1 2 — — 70 3 0 (0-43) 2-7 (0-38) 2-8 (0-40) 442 PALAEONTOLOGY, VOLUME 8 Worthoceras vermiculum D B d h b u BEG A 19827 14 4-4 70 3 0 (0-43) 3 0 (0 43) 2-8 (0 40) BEGA 35390 15 ?40 6-5 2-8 (0-43) 2-7 (042) 2-7 (042) BEGA 35392 17-5 51 9-5 3-8 (040) ? 3-2 (0-34) Worthoceras gibbosum BEGA 19812 17-5 7-8 10 4-5 (045) 5-2 (0-52) 3 0 (0-30) text-fig. 1 1 . The trend of worthoceratid development. X 2. a, W. platydorsum (Scott), b, W. vermiculum (Shumard), c, W. gibbosum Moreman. text-fig. 12. The pre- sumed development of the labeceratids. xO-25. a, Labeceras, b, Ellipso- ceras, c, Myloceras. Finally we note the tendency to re-coil in Worthoceras (text-fig. 11), which seems to be a common trend in all Cretaceous heteromorphs (Cobban 1952, text-fig. 2; Casey 1960, text-fig. 5; Wiedmann 19626, text-fig. 35; and here text-figs. 12 and 14). J. WIEDMANN: ORIGIN, LIMITS, AND POSITION OF SCAPHITES 443 Labeceratidae. The systematic position of the family Labeceratidae, included in Scaphi- taceae by Spath, Wright and others, was discussed on a previous occasion (Wiedmann 19626, p. 98). This restricted group of Upper Albian forms, known only from the Southern hemisphere, was characterized by the following features : 1. Sculpture, especially in the primitive representatives, is hamitid; but it becomes rapidly anisoceratid through the development of umbilical and marginal tubercles. 2. Mode of uncoiling is hamitid/crioceratitid at first, the middle-aged whorls are some- times in contact, followed by a straight septate shaft and a scaphitoid final hook with lappeted mouth-border. These are the unique similarities with Scaphites. 3. Suture-line consists of only four elements; U remains simple and undivided, and is trifid like L; the saddles are normal sized (text-fig. 13). The lineage Labeceras-Ellipsoceras-Myloceras (text-fig. 12) represents, if correct, a further example for the general trend of many heteromorphs to re-acquire the original mode of coiling. This mode of uncoiling, the sculpture and especially the sutural characters are exactly the same as in some early Anisoceras like Metahamites or Idio- hamites. Hamitoides, included in Labeceratidae by Spath and Wright, and perhaps the origin of the stock, is in all its characters a true Anisoceras. Therefore Labeceratidae were reduced to subfamily rank and included in Anisoceratidae by the author (19626, p. 99). The reference of Labeceras to Hamitidae by Whitehouse (1926, p. 226) was found to be more correct than the attribution to the lytoceratid or scaphitid stock by Spath. Zuluscaphites. This genus was established by van Hoepen (1955) on a single ammonitid spiral portion of Middle Albian ( ?) age. Its lateral lobe is trifid and no scaphitid living- chamber is known. It is impossible to regard a form with these negative characters as the source of the always later appearing Scaphites. I believe that Zuluscaphites merely represents a lyelliceratid inner whorl (Wiedmann 1962u, p. 209). Other ammonitids with uncoiled body-chamber. True ammonitids with scaphitoid living- chamber are very common. They are represented in the Jurassic, with forms like the stephanoceratid genus Kheraites (text-fig. 1 3) or the oppeliid Proscaphites, as well as in the Cretaceous. Here especially in the acanthoceratids Vascoceras, Fallotites (text-fig. 13), Pseudotissotia and Texasia, the uncoiling of the body-chamber is a common feature. All these forms are evidently connected with the acanthoceratid stock, and cannot lead to Scaphites because of the presence of three or more umbilical lobes. The same difference can be stated for Binneyites which has recently been referred to Haplocerataceae by Wright in Casey (1961, p. 118). Already the external suture-line (text-fig. 13) is com- pletely different from the scaphitid one. No real relationship to Scaphites can be recognized in the above-mentioned lyto- ceratids, heteromorphs, and ammonitids of scaphitoid appearance. In the eoscaphitid species the presence of an umbilical perforation and the absence of a dorsal impression was demonstrated. Therefore it seems that we can find the true scaphitid ancestor among the uncoiled heteromorphs. This presumption is supported through the fact, that in all early scaphitids the suture-line consists only of four elements. Not only in Eoscaphites and early scaphitid ontogeny can the quadrilobate suture-line be stated, but it remains simplified throughout during the ontogeny of all scaphitids studied here. The presumed regeneration of three or four umbilical lobes in Scaphites cannot be accepted. Scaphites 444 PALAEONTOLOGY, VOLUME 8 is in its sutural characters a true heteromorph member. It should no longer be regarded as a degenerate offshoot of normal coiled ammonites ; it is a hamitid heteromorph with the repeatedly noted trend to secondary re-coiling. The similarity is striking if we compare the suture-line of E. circularis with that of contemporaneous hamitids, especially the Upper Albian Plesiohamites (‘ Lytohamites') similis (Casey). This form also agrees in sculpture and mode of uncoiling so completely, that the separation from Eoscaphites seems to me to be more than arbitrary. If the biogenetic rule is of any phylogenetic significance, then the slender initial whorls of Eoscaphites point to an also very slender hamitid species, of upper Middle Albian age, Hamites tenuis. This species is stratigraphically directly followed by the first Eoscaphites, at the base of the English Upper Albian. It might be added here, that the possibility of a hamitid source for Scaphites was briefly discussed but also abandoned by Spath (1942, p. 717), and therefore neglected thereafter. THE PRINCIPLES OF SCAPHITID DEVELOPMENT Having pursued the early scaphitid development step by step from the base of the Upper Albian, we have reached not only an idea of the principles of scaphitid develop- ment, but also the solution to some of the problems of interest here. It has been possible to demonstrate that there is no difficulty in connecting the ostensibly different suture types of Eoscaphites and Scaphites. Schindewolf doubted that the quadrilobate suture- line which exists throughout in Eoscaphites could be compatible with that of true scaphi- tid species from the uppermost Senonian. In the early ontogeny five lobes were stated to occur in these forms, and seven in the adult stage (Schindewolf 1961, text-fig. 58). As text-fig. 14 shows, the first eoscaphitid species from the basal Upper Albian and the latest Maastrichtian scaphitids are connected by continuous transitions in the sutures. We therefore have to conclude that also in the late Scaphites not more than four lobes (/, U, L, and E) are present. The presumably regenerated umbilical lobes are in reality the secondary saddle incisions of L/U, and therefore named pseudolobes (p ). This becomes underlined, when we consider the ontogenetic suture development (text-figs. 7-9, 15). Here the questionable elements always remain outside of the umbilical seam, EXPLANATION OF PLATE 60 Figs. 1, 2. Worthoceras vermiculum (Shumard). 1, Hypotype BEGA 35390; Lower Turonian (Eagle Ford group), Ellen Fork, Trinity River at St. Louis/Sta Fe-Railroad bridge, Dallas County (Texas, U.S.A.). The mouth-border seems unlappeted; the first half of the shaft is septate, a. Lateral, b, ventral view. 2, The neotype with preserved mouth-border; proposed by Moreman (1942, pi. 34, figs. 12, 13), BEGA 19827; Lower Turonian (Eagle Ford group), 2-7 miles S. Britton, E. Midlothian highway, Ellis County (Texas, U.S.A.). a. Frontal, b, ventral, c, lateral view. Fig. 3. Worthoceras gibbosum Moreman. Holotype BEGA 19812 (cf. Moreman 1942, pi. 34, figs. 7, 8); same age and locality, a , Frontal, b, lateral view of smooth phragmocone ; x 4 ; c, ventral, d, lateral, e, frontal view; x2. Figs. 4-6. Worthoceras rochatianum (d’Orbigny). 4, Hypotype MHNG Wi ‘Sc’/lO, nearly complete, shaft is partly septate; Upper (?) Turonian, Uchaux (France), a. Frontal view with distinct dorsal impression, b, lateral view (leg. Rochat). 5, Hypotype with weak ribs, MHNG Wi ‘Sc '/l 2 (leg. Rochat); same locality, a. Frontal, b, lateral view. 6, Hypotype MHNG Wi ‘Sc’/ll, with part of straight shaft ; (leg. Rochat) ; same locality, a. Frontal, b, lateral view. All x 3 unless otherwise stated. Palaeontology, Vol. 8 PLATE 60 WIEDMANN, Cretaceous scaphitids Macroscaphites/Costidiscus ( LYTOCERATACEAE) text-fig. 13. The scaphitoid homoeomorphs. Mocroscophites/Oostid/scus ( LYTOCERATACEAE) Jaubert/ceras (TETRAGONITACEAE) text-fig. 13. The scaphitoid homoeomorphs. 6 I: ORIGIN, LIMITS, AND POSITION OF SCAPHITES TURON/A N-MAA STR/CH T!A N Hoplosc. constrictus Sc. nigricollensis 447 J. WIEDMANN: ORIGIN, LIMITS, AND POSITION OF SCAPHITES where true umbilical lobes are generally produced. The suture development of the Cam- panian Sc. hippocrepis (Dekay) as given by Reeside (19276), can now be shown precisely in the manner illustrated here (text-fig. 15). Here the extreme lengthening of the external saddles, and the gradual increase of saddle incisions to lobe size, are the most obvious characters of the scaphitid suture development. Another important systematic feature is the element U, which becomes I U L E c text-fig. 15. The suture development of a late scaphitid species, Sc. (Sc.) hippocrepis (Dekay) (after Reeside 19276, pi. 15); Campanian, Big Horn County. divided by the umbilical seam at first (text-fig. 1 5 d). Thereafter the two different branches become more and more separated, and pass to the internal suture-line (text-figs. 15e, /). U remained fixed on the umbilical seam only in the micromorphs, as in some primitive Albian forms. In the adult stage of the Upper Cretaceous forms (text-fig. 15g) a greater number of umbilical lobes are feigned. Naturally we come to wrong conclusions, if we pay exclusive attention to them. However, if we trace the suture development onto- genetically (text-fig. 15) and phylogenetically (text-fig. 14), we can easily recognize that the so-called umbilical lobes are merely the incisions (px, pz . . .) of the lengthened saddle L\U. Also the species with a ‘ trifid ’ lateral lobe should not be regarded exclusively. We have seen that there are indeed forms (Sc. simplex and the members of the meriani group) in the immediate neighbourhood of the type species, in which a ‘ trifid ’ L becomes developed. If we consider, however, the phylogenetic and ontogenetic evolution of this element, then we prefer to use the term ‘asymmetric’. This asymmetry evidently de- pends on the retraction of the umbilical suture-line of some Upper Albian species, especially Sc. simplex. In the subsequent species, Sc. obliquus and equalis, the suture- line is less retracted, and the original bifidity rapidly becomes re-established. In Sc. equalis the lobe development of its ancestors becomes recapitulated. Bifid at first, the 448 PALAEONTOLOGY, VOLUME 8 lateral lobe becomes asymmetrical in early ontogeny and regains the primary bifid out- line thereafter. Only this ultimate change was observed by Pervinquiere (1907) and since then regarded as evidence for an ammonitid origin of all scaphitids. Likewise in some Senonian species a ‘ trifid ’ L was cited, but Cobban (1952) demonstrated that also in these forms an originally bifid lobe gradually becomes asymmetrical. On the contrary, the trifidity of the so-called ‘second lateral’ or ‘ C/2’ is the original outline of the mis- interpreted median incision of the saddle LIU. Corresponding to the suture development, the septal surface is always hamitid in early Eoscaphites (text-fig. lc). It rapidly receives its typical scaphitid shape (text-figs. Id, 2b, 6d, la, b ), characterized by its comparatively small lobes, elongated saddles and only one distinctly bipartite umbilical lobe on or near the umbilical seam. Also in the sculpture development a gradual change between Eoscaphites and the Scaphites type species can be observed. In E. circulars the sculpture remains simple and uniform (comparable with that of the contemporaneous Hamites). The branching of ribs, at first only occasionally observable, is a common feature in the first true Scaphites. Now the differentiation in primary and secondary ribs becomes more and more pro- nounced. While the primaries increase in strength, the point of bifurcation approaches the outer margin, and may be tuberculate or not. Likewise the degree of shell involution changes continually. In Eoscaphites the first whorls are loosely coiled and no dorsal impression can be recognized. The first whorls are extremely slender and circular in section, but they rapidly increase in size (about 400 per cent, per whorl) and involution. This change can be followed step by step in the eoscaphitid species, and there is no doubt that Eoscaphites had an uncoiled, hamitid ancestor. Already in the first true Scaphites (Sc. simplex) the maximum shell involution was reached. But an inversion of this tendency appears in the somewhat more open whorled Cenomanian species. The final hook which is highly elongated in Eoscaphites becomes more and more inflated and stouter in the true scaphitids. Finally, the mouth-border is merely separated from the initial coil. Here, one of the features of heteromorphous development becomes apparent, namely the tendency to abandon the loose coiling and to regain the normal ammonitid spire. Through all these gradual transitions the generic separation between Eoscaphites and Scaphites remains arbitrary. This separation, which we retain, may be based on the umbilical perforation of Eoscaphites, its extreme increase of whorl size, the absence of a dorsal impression and the undifferentiated elements U and L/C/. In pursuing the development of early scaphitids, we recognized, contrary to the present knowledge, an extreme splitting off into different lineages, especially in the Upper Albian and Cenomanian (text-fig. 16). Metascaphites and Otoscaphites can now be directly derived from an eoscaphitid origin, while the above separated meriani and similaris stocks originate from Sc. simplex, the first true Scaphites. All these forms merely need generic separation. Sc. minutus was erroneously regarded as a transitional form between Worthoceras and Otoscaphites by Wright. While the former is now recognized as a homoeomorph development of ptychoceratid origin, Otoscaphites is identified as a true scaphitid development, which parallels the equalis main stock, probably throughout the complete Upper Cretaceous. Text-fig. 16 represents a convincing example for Schindewolf’s typostrophism. The J. WIEDMANN: ORIGIN, LIMITS, AND POSITION OF SCAPHITES 449 text-fig. 16. The scaphitid relationships. WIEDMANN: ORIGIN, LIMITS, AND POSITION OF SCAPHITES 450 PALAEONTOLOGY, VOLUME 8 three stages of divergent evolutionary rate, which Schindewolf recognized in many different fossil groups and regarded as an essential evolutionary principle, are well represented here. In the primary stage of explosive development (typogenesis) nearly all diverging types are produced during a relatively short period (Upper Albian/Ceno- manian). It is followed by a long and stable period (Turonian/Santonian), where the types barely change (typostasis). Finally, in the Campanian and Maastrichtian, a new stage of explosive development can be observed, which is generally destructive in nature and announces the imminent extinction of the scaphitid stock (typolysis). The problem of parallel development in scaphitids has repeatedly been emphasized. The members of the similaris stock, the ptero- and otoscaphitids, are exclusively small- sized or dwarfed species, which almost exactly parallel the normal-sized members of the main stock. We have two solutions to this problem. The members of the similaris stock seem to represent atrophic equivalents of the normal-sized boreal scaphitids. They are restricted to the Southern areas and the Pacific border. The ptero- and otoscaphitid micromorphs, however, are generally found together with the normal-sized specimens. In the literature of the last century this fact was often interpreted as sexual dimorphism, and the micromorphs were not separated from the large-sized species. In the modern scaphitid systematics, as created by Wright, the micromorphs were treated as a distinct genus ( Pteroscaphites ) or subfamilies (Otoscaphitinae). I am unable to follow Wright in regarding the mouth lappets, which are common in these forms, as the basis for the proposed separation. It has been known for a long time, that lappeted and unlappeted forms are inseparably joined in many ammonitid stocks. Makowski (1963) made the fact, that smaller size and presence of (lateral) mouth lappets in ammonites characterize the male sex, highly possible. Naturally it is impossible to solve this problem absolutely. But all these arguments seem reason enough to re-integrate the micromorph series into the equalis main stock. Thus we come to the conclusion that the systematics of ‘Scaphitaceae’ may be rigorously simplified. Excluding Worthoceras and the Labeceratids as convergent develop- ments, the remaining true scaphitids may be restricted to one single family. If we regard the reduced number of four-lobe elements as a characteristic of all heteromorphs, these also may be grouped into one single superfamily, Ancylocerataceae, in which Scaphitidae, in the neighbourhood of the ancestral hamitids, can now easily be included. REFERENCES adkins, w. s. 1920. The Weno and Pawpaw formations of the Texas Comanchean. Univ. Texas Bull. 1856 (1918), 3-147, pi. 1-11. 1928. Handbook of Texas Cretaceous fossils. Univ. Texas Bull. 2838, 303 pp., 37 pis. — — and winton, w. m. 1 920. Paleontological correlation of the Fredericksburg and Washita forma- tions in North Texas. Univ. Texas Bull. 1945 (1919), 84 pp., 21 pis. alth, a. 1850. Geognostisch-palaontologische Beschreibung der nachsten Umgebung von Lemberg. Haidinger's Naturw. Abh. 3, 172-248, pi. 9-13. anderson, f. m. 1902. Cretaceous deposits of the Pacific Coast. Proc. Calif. Acad. Sci., ser. 3, Geol. 2, no. 1, 1-131, pi. 1-12. basse, E. 1952. Cephalopodes, Nautiloidea, Ammonoidea. In Traite de Paleontologie, Pivetau, ed., 2, 522-688, pi. 1-24. boule, m., lemoine, p., and thevenin, a. 1907. Cephalopodes cretaces des environs de Diego-Suarez. II. Ann. Paleont. 2, 1-56 (21-77), pi. 1-8 (8-15). J. WIEDMANN: ORIGIN, LIMITS, AND POSITION OF SCAPHITES 451 breistroffer, m. 1947. Sur les zones d’ Ammonites dans TAlbien de France et d’Angleterre. Trav. Lab. geol. Univ. Grenoble, 26, 1-88. brongniart, A. 1 822. In Brongniart and Cuvier, Description geologique des environs de Paris. Nouv. edit., viii+428 pp., 16 pis., Paris. brown, t. 1837-49. Illustrations of the Fossil Conchology of Great Britain and Ireland, viii+273 pp., 97 pis. London. casey, R. 1960, 1961. The Ammonoidea of the Lower Greensand. Palaeontogr. Soc. 113 (1959), xxxvi +44 pp., pi. 1-10, 115 (1961), pp. 119-216, pi. 26-35. cieslinskj, s. 1959. The Albian and Cenomanian in the Northern periphery of the Swiety Kryz Moun- tains. Inst. Geol. Prace, Warszawa, 28, 1-95, 8 pis. cobban, w. A. 1952. Scaphitoid cephalopods of the Colorado group. Geol. Surv. Prof. Paper 239, 42 pp., 21 pis. 1964. The late Cretaceous cephalopod Haresiceras Reeside and its possible origin. Geol. Surv. Prof. Paper 454 1, 19 pp., 3 pis. and scott, g. r. 1964. Multinodose scaphitid cephalopods from the lower part of the Pierre Shale and equivalent rocks in the conterminous United States. Geol. Surv. Prof. Paper 483-E, 1 3 pp., 4 pis. collignon, M. 1928-9. Les Cephalopodes du Cenomanien pyriteux de Diego-Suarez. Ann. Paleont. 17 (1928), 24 pp., 5 pis., 18 (1929), pp. 25-79, pi. 6-7. 1931. La faune du Cenomanien a fossiles pyriteux du Nord de Madagascar. Ann. Paleont. 20, 43-104, pi. 5-9. 1963. Atlas des fossiles caracteristiques de Madagascar (Ammonites). Fasc. X (Albien). Expl. roneot. Serv. Geol. Republ. Malgache, 184 pp., pi. 241-317. deshayes, g. p. 1831. Description de coquilles caracteristiques des terrains. 264 pp., 14 pis. Paris. diener, c. 1925. Ammonoidea neocretacea. Fossilium Catalogus 1-29, 244 pp. Berlin. dixon, f. 1850. The geology and fossils of the Tertiary and Cretaceous formations of Sussex, xvi+422 pp., 40 pis. London. drushtchic, v. v. 1962. Der Umfang der Ordnung Lytoceratida. Bjull. Mosk. obc. ispyt. prirody, otd. geol. 36, 132-3. (In Russian.) frech, f. 1915. Ober Scaphites. Centralbl. Miner, etc. 1915, 553-568. fritsch, a. 1872. Cephalopoden der bohmischen Kreideformation. 52 pp., 16 pis. Prague. GEINITZ, H. b. 1846. Grundriss der Versteinerungskunde. 813 pp., 26 pis. Dresden and Leipzig. 1871-5. Das Elbthalgebirge in Sachsen. 1. Teil: Der untere Quader. Palaeontographica 20 (1), 319 pp., 67 pis. (1875, 277-320). grossouvre, a. de. 1894. Recherches sur la Craie superieure. 2e partie: Paleontologie. Mem. serv. expl. Carte geol. France, 2 vol., 264 pp., 39 pis. gueranger, e. 1867. Album paleontologique du departement de la Sarthe. 20 pp., 25 pis. Le Mans. hauer, f. von. 1862. Uber die Petrefakten der Kreideformation des Bakonyer Waldes. Sitzungsber. Akad. Wiss. Wien, Math.-naturw. Kl. 44 (1861), 631-59, 3 pis. hebert, e. and munier-chalmas, e. 1875. Fossiles du bassin d’Uchaux. Ann. Sci. geol. 6, 113-32, pi. 4-6. herbich, F. 1888. Date paleontologice din Carpatii Romanestl. Anuar. biur. geol., Bucuresci, 3 (1885), 178-339, 29 pis. hoepen, e. c. n. van. 1955. New and little-known Ammonites from the Albian of Zululand. South Afr. J. Sci. 51, no. 12, 355-60. hyatt, a. 1900. Cephalopoda. In Zittel and Eastman, Textbook of Paleontology, 502-92. London and New York. jimbo, k. 1894. Beitrage zur Kenntnis der Fauna der Kreideformation von Hokkaido. Palaeont. Abh., Neue Serie, 2, no. 3, 47 pp., 9 pis. jukes-browne, a. j. 1875. On the relations of the Cambridge Gault and Greensand. Quart. J. Geol. Soc. 31, 256-316, pi. 14, 15. logan, w. n. 1899. Contributions to the Paleontology of the Upper Cretaceous series. Field Columbian Mus. 36, Geol. ser. 1, 205-16, pi. 22-26. luppov, n. p. and drushtchic, v. v. 1958. Ammonoidea (Ceratitida, Ammonitida), Endocochlia. Beilage: Coniconchia. Osnovy Paleont ologii, 359 pp., 70+7 pis. Moscow. (In Russian.) 452 PALAEONTOLOGY, VOLUME 8 makowski, h. 1963. Problem of sexual dimorphism in Ammonites. Palaeont. Polonica 12 (1962), 92 pp., 12+20 pis. mantell, G. A. 1822. The fossils of the South Downs, or illustrations of the geology of Sussex. 320 pp., 43 pis. London. matsumoto, t. 1963. A survey of fossils from Japan illustrated in classical Monographs, part X. K. Jimbo (1894): Beitrage zur Kenntnis der Fauna der Kreideformation von Hokkaido. 25 th Anniver- sary vol. Palaeont. Soc. Japan, 41-48, pi. 60-68. meek, f. b. 1876. A report on the invertebrate Cretaceous and Tertiary fossils of the Upper Missouri Country. Monogr. U.S. Geol. Surv. Terr. 9, 629 pp., 45 pis. moreman, w. l. 1 942. Paleontology of the Eagle Ford group of North and Central Texas. J. Paleont. 16, 192-220, pi. 31-34. neumayr, m. 1875. Die Ammoniten der Kreide und die Systematik der Ammonitiden. Z. deutsch. Geol. Ges. 27, 854-942. nowak, j. 1911. Untersuchungen liber die Cephalopoden der oberen Kreide in Polen. II. Teil. Die Skaphiten. Bull, intern. Acad. Sci. Cracovie, Cl. Sci. math. & nat., ser. B, (1911), 547-89, pi. 32, 33. 1915. Ober die bifiden Loben der oberkretazischen Ammoniten und ihre Bedeutung fur die Systematik. Ibid., 1915, 1-13. 1916. Zur Bedeutung von Scaphites fur die Gliederung der Oberkreide. Verb. geol. Reichsanst. 1916, 55-67. d’orbigny, a. 1840-2. Cephalopodes. In Paleontologie Frangaise, Terrain Cretace, I. 662 pp., 148 pis. (1842: pp. 431-662). Paris. 1850. Prodrome de Paleontologie stratigraphique universelle des Animaux mollusques et rayonnes II. 427 pp. Paris. parkinson, j. 1811. Organic remains of a former world. Vol. 3. xvi+479 pp., 22 pis. London. parona, c. f. and bonarelli, g. 1897. Fossili albiani d’Escragnolles, del Nizzardo e della Liguria occidentale. Palaeontogr. Italica 2, (1896), 53-112, pi. 10-14. passy, A. 1832. Description geologique du departement de la Seine-Inferieure. 371 pp., 20 pis. Rouen. pervinquiere, l. Etudes de Paleontologie Tunisienne. I. Cephalopodes des terrains secondaires. Dir. Gen. Trav. Pub/., Carte geol. Tunisie, 438 pp., 27 pis. Paris. 1910. Sur quelques Ammonites du Cretace Algerien. Mem. Soc. Geol. France 17, Mem. 42, 86 pp., 7 pis. pictet, f. j. 1848. Description des Mollusques fossiles qui se trouvent dans les Gres Verts des environs de Geneve. I. Cephalopodes. Mem. Soc. Phys. & Hist. Nat. Geneve 11, 257-412, pi. 1-15. and campiche, g. 1861-4. Description des fossiles du terrain cretace des environs de Sainte- Croix. II. Mater. Paleont. Suisse, 3e ser., 2, 752 pp., pi. 44-98 (1861 : 1-144 pp., pi. 44-57). quenstedt, f. A. 1846-9. Petrefactenkunde Deutschlands. 1. Abt., 1, Cephalopoden. 580 pp., 36 pis. (1847/8, 185-472). Tubingen. reeside, j. b. 1927a. The Scaphites, an Upper Cretaceous ammonite group. Geol. Surv. Prof Paper 150-B, 21-36, pi. 9-11. 19276. The Cephalopods of the Eagle Sandstone and related formations in the Western Interior of the United States. Geol. Surv. Prof. Paper 151, 40 pp., 45 pis. rollier, l. 1922. Phylogenie des Ammonoides. Eclog. geol. Helv. 17, 358-60, pi. 20-22. roman, f. 1938. Les Ammonites Jurassiques et Cretaces. Essai de genera. 554 pp., 53 pis. Paris. and mazeran, p. 1913. Monographic paleontologique de la faune du Turonien du bassin d’Uchaux et de ses dependances. Arch. Mus. Hist. Nat. Lyon 12, 137 pp., 35 text-figs., 1 1 pis. ruzhencev, v. E. 1960. Ammonoid classification problems. J. Paleont. 34, 609-19. schindewolf, o. H. 1950. Grundfragen der Palaontologie. 495 pp. Stuttgart. 1961. Studien zur Stammesgeschichte der Ammoniten. 1. Lieferung. Abh. Akad. Wiss. Literatur Mainz, Math.-naturw. Kl. 1960, no. 10, 109 pp., 2 pis. schluter, c. 1871-2. Cephalopoden der oberen deutschen Kreide 1. Palaeontographica 21, 120 pp. 35 pis. (1872, 25-120, pis. 9-35). smith, j. p. 1901. The larval coil of Baculites. Amer. Naturalist 35, 39-46, 2 pis. smith, w. D. 1905. The development of Scaphites. J. Geol. 13, 635-54. sornay, j. 1955. Ammonites nouvelles du Cretace de la region des Monts du Mellegue (Constantine). Publ. Serv. Carte geol. Algerie, Mem. Paleont. 18, 40 pp., 2 pis. J. WIEDMANN: ORIGIN, LIMITS, AND POSITION OF SCAPHITES 453 sowerby, j. 1812-23. The Mineral Conchology of Great Britain. 1-4 (pars), pi. 1-383, (1813 : pi. 10-44). London. sowerby, j. de c. 1823-46. The Mineral Conchology of Great Britain. 4-7, pi. 384-648. London. 1836. Descriptive notes respecting the shells figured in plates xi to xxiii. Appendix A in Fitton, On the strata below the Chalk . . . Transact. Geol. Soc. London, 2 ser., 4, 335-49. spath, l. F. 1922. On the Senonian ammonite fauna of Pondoland. Transact. R. Soc. South Afr. 10, 113-47, pi. 5-9. 1925. On Upper Albian ammonites from Portuguese East Africa, with an appendix on Upper Cretaceous ammonites from Matupoland. Ann. Transvaal Mus. 11, 179-200, pi. 28-37. 1933. The evolution of the Cephalopoda. Biol. Reviews 8, 418-62. 1934-42. A monograph of the Ammonoidea of the Gault. Palaeont. Soc. 86 (1932), 443-96, pi. 44-56. 90 (1936), 497-540, pi. 57, 58. 95 (1941), 609-68, pi. 65-72. 96 (1942), 669-720. stoliczka, f. 1863-6. The fossil Cephalopoda of the Cretaceous rocks of Southern India. Mem. Geol. Surv. India, Palaeontologia Indica, ser. 3, 41-216, pi. 26-94 (1865, 107-54, pi. 55-75). 1 868. Additional observations regarding the cephalopodous fauna of the South Indian Cretaceous deposits. Rec. Geol. Surv. India, 1, 32-37. sucic-protic, z. 1961. Stratigraphie et tectonique des montagnes de Golubac (Serbie orientale). Ann. geol. Penins. balkanique 28, 25-130, 9 pis. (In Serbian.) uhlig, v. 1883. Die Cephalopoden der Wernsdorfer Schichten. Denk-schr. Akad. fViss. Wien 46, 1 27— 290, pi. 1-32. whitehouse, f. w. 1926. The Cretaceous Ammonoidea of Eastern Australia. Mem. Queensland Mus. 8, 195-242, pi. 34-41. wiedmann, j. 1962a. Ammoniten aus der Vascogotischen Kreide (Nordspanien). I. Phylloceratina, Lytoceratina. Palaeontographica A 118, 119-237, pi. 8-14. 19626. Unterkreide- Ammoniten von Mallorca. 1. Lieferung: Lytoceratina, Aptychi. Abh. Akad. Wiss. Literatur Mainz Math.-naturw. Kl., 1962, no. 1, 148 pp., 10 pis. 1963. Entwicklungsprinzipien der Kreideammoniten. Paldont. Z. 37, 103-21, 1 pi. wright, c. w. 1952. A classification of the Cretaceous Ammonites. J. Paleont. 26, 213-22. 1953. Notes on Cretaceous Ammonites. I. Scaphitidae. Ann. & Mag. Nat. Hist., ser. 12, 6, 473-6. 1957. In Treatise of Invertebrate Paleontology, Part L, R. C. Moore (ed.), pp. L80-L490. New York. 1963. Cretaceous Ammonites from Bathurst Island, Northern Australia. Palaeontology 6, 597- 614, pi. 81-89. * and wright, e. v. 1951. A survey of the fossil Cephalopoda of the Chalk of Great Britain. Palaeont. Soc., pp. 1-41. yabe, h. 1909. Zur Stratigraphie und Palaontologie der oberen Kreide von Hokkaido und Sachalin. Z. deutsch. Geol. Ges. 61, Abh., 402-44. 1910. Die Scaphiten aus der Oberkreide von Hokkaido. Beitr. Pal. Geol. Osterr.-Ung. & Orients 23, 159-73, pi. 15. zittel, k. A. von. 1881-5. Handbuch der Palaontologie. 1. Abt. Palaozoologie. 2. Band. 893 pp. Miinchen and Leipzig. JOST WIEDMANN Geologisches Institut, Tubingen, Manuscript received 16 March 1964 Germany ISO RT HIS AND SALOPINA (BRACHIOPOD A) IN THE LUDLOVIAN OF THE WELSH BORDERLAND by VICTOR G. WALMSLEY Abstract. Examination of the enteletacean brachiopod members of the faunal assemblages now used in British Ludlovian stratigraphy (Holland, Lawson, and Walmsley 1963) and comparison with a large collection of Silurian and Lower Devonian brachiopods from many parts of the world, reveals that the species commonly determined as Dalmanella lunata (J. de C. Sowerby) and D. orbicularis (J. de C. Sowerby), have not yet been found, (though sometimes recorded), from areas outside Britain. However, one of the new species of Isorthis here described has a wide distribution in Britain and occurs in North America (Maine and New Brunswick) and South America (Venezuela). Salopina Boucot is emended and Salopina lunata is described, figured and distinguished from S. submedia (McLeam), which was figured as S. lunata by Boucot 1960, from New Bruns- wick. Isorthis Kozlowski is emended and Orthis orbicularis J. de C. Sowerby together with two new species (/. amplificata and I. clivosa), and two new subspecies (I. scuteformis scuteformis and I. scuteformis uskensis), from the Welsh Borderland and a new, related species (/. slitensis) from Gotland, are assigned to Isorthis and described and figured. In order that the refinements of stratigraphy and correlation recently achieved in the Ludlovian of the Welsh Borderland (Holland, Lawson, and Walmsley 1963), may have greater value in wider correlations, it is necessary that the faunas on which the work has largely been based, should be studied in detail and descriptions made available. Very little work has been published on British Upper Silurian brachiopods since Davidson’s Monograph of the last century. The writer had begun a study of Ludlovian enteletacean brachiopods based largely on his own collections from the Welsh Border- land when the opportunity to spend a year in the U.S.A. made possible the essential comparative studies. Enteletacean brachiopods of the following genera occur in the Ludlovian rocks of the Welsh Borderland : Isorthis, Salopina, Resserella (= Parmorthis), Dalejina (= Rhipi- domelloides), and Dicoelosia. The purpose of the present paper is to describe and figure the several species commonly determined as Dalmanella orbicularis (J. de C. Sowerby) and D. lunata (J. de C. Sowerby), important members of the successive faunal assemb- lages now recognized in the type Ludlovian (Holland, Lawson, and Walmsley 1959, Lawson 1960, Holland, Lawson, and Walmsley 1963). Further papers are in preparation (jointly with Dr. A. J. Boucot and Dr. C. W. Har- per) on the genera Isorthis, Salopina, Resserella, and Fascico Stella, based on the study of Dr. Boucot’s collection from the U.S.A., Canada, South America, Europe, and New Zealand. Examination of the enteletacean brachiopods in this collection, together with material kindly loaned by various museums and private individuals, has shown that, despite several references to the contrary, the species I. orbicularis and S. lunata have not yet been found outside the British area. In Britain, however, several isorthids, des- cribed here as new species and subspecies, have probably been confused as lD. orbicu- laris'. One of these new species (/. clivosa) has a wide distribution in Britain and occurs in North America (Maine and New Brunswick) and in South America (Venezuela). [Palaeontology, Vol. 8, Part 3, 1965, pp. 454-77, pis. 61-65.] V. G. WALMSLEY : ISORTHIS AND SALOPINA (BRACHIOPODA) 455 The species Orthis lunata and O. orbicularis were first described by J. de C. Sowerby in Murchison’s The Silurian System, 1839, p. 611 and figured on pi. 5, figs. 15 and 16 respectively. Sowerby commented on the difficulty of distinguishing these species ‘ unless it [the specimen] be very perfect, or show the impressions upon the cast of the charac- teristic ridges in the interior’. It was presumably this difficulty which caused Davidson (1869, p. 215) to put O. orbicularis in synonomy with O. lunata and his figs. 1-5 on pi. XXVIII show both species under the latter name, an unfortunate error which misled several later authors. Elies and Slater (1906), however, in their account of the rocks of the Ludlow district, recognized and listed both species, as did subsequent authors in the Welsh Borderland. Several of these faunal lists, moreover, also recorded ‘ Dalmanella spp. ’ from the Ludlovian — an indication that there were dalmanellid shells present, not easily assignable to either D. lunata or D. orbicularis. The genus Dalmanella was erected by Hall and Clarke (1892) and the complex nomen- clatorial history of this group has been recounted recently (Williams and Wright 1963, pp. 1-3). Since 1892, the species O. lunata and O. orbicularis have usually been assigned to Dalmanella, but in 1960 Boucot (in Boucot et al., p. 3) erected the genus Salopina with 0. lunata Sowerby as type species. Unfortunately the shell described and figured by him, in his pi. 1, figs. 6-12, is not O. lunata J. de C. Sowerby, but -S. submedia (McLearn) from the Jones Creek Formation ( ? Wenlockian-Ludlovian in age) of New Brunswick, Canada. This species has now been redescribed in a joint paper on the ‘Silurian Brachio- pods and Gastropods of Southern New Brunswick’ (Boucot, Johnson, Harper, and Walmsley — in press). Kozlowski 1929, p. 75, erected Isorthis as a subgenus of Dalmanella, and described 1. szajnochai, the type species, in considerable detail. Schuchert and Cooper 1932, pp. 149-50, redescribed Isorthis and raised it to generic rank. They assigned to it a number of American and European species but did not include D. orbicularis. Boucot (in Boucot et al., 1960, p. 5 and pi. I, figs. 13-20, pi. II, figs. 1-7) described and figured a small silicified isorthid from the Sutherland River Formation (late Silurian- early Devonian age) of Devon Island, Canadian Arctic Archipelago, as I. orbicularis (Sowerby) and recorded it (p. 6) as the first known occurrence of this species from North America. I have been able to study this material in Dr. Boucot’s collection and to determine that it is a new species quite distinct from I. orbicularis (J. de C. Sowerby), (Walmsley, Boucot, and Harper — in preparation). In the present paper the genus Salopina is emended and S. lunata (J. de C. Sowerby) is described and figured. O. orbicularis J. de C. Sowerby is assigned to Isorthis emended, together with two new species and two new subspecies from the Welsh Borderland and a closely related species from Gotland — all of which are here described and figured. MORPHOLOGICAL TERMINOLOGY The terminology used by Schuchert and Cooper (1932) and defined in their glossary, pp. 6-1 1 and pi. a, is largely adopted. However, further explanation is needed of a few terms found necessary in the following systematic descriptions : (a) Fulcral plate. This term was used by Schuchert and Cooper for a small concave plate connecting the brachiophore to the shell wall (see their pi. A, fig. 14), and consider- able taxonomic importance was attached to the presence or absence of such a plate. It is h h B 6612 456 PALAEONTOLOGY, VOLUME 8 clear from their definition (p. 8, pi. a, fig. 14), that this name was applied to a plate which formed the floor of the socket and beneath which there was a small cavity. The antero- lateral edge of such a plate (being also the edge of the socket), shows itself in the usual figures of the interior of the brachial valve as a short line connecting the brachiophore and the hinge line. In such views it is usually impossible to determine whether a cavity exists beneath the socket floor. A very similar, if not identical, aspect is produced by an alternative structure which is commonly present in enteletacean shells. Here, the socket is a depression in the upper surface of a solid block or pad of calcite, but in this case, there is no cavity beneath the socket floor. The antero-lateral edge of such a structure resembles that of a true fulcral plate. Since all morphologic stages can be seen between a true fulcral plate and a solid socket pad, it would appear that the leading edge of the socket may be formed either by for- ward extensions of the calcite forming the socket floor so as to overhang and eventually produce a lower cavity and a fulcral plate, or, possibly by resorption of the antero- lateral face of a solid socket pad which could produce a lower cavity and hence a fulcral plate. A third possibility exists. The socket pad condition may be a result of a true ful- cral plate becoming submerged in adventitious calcite which has filled in the lower cavity. It is not known which, if any, of these possible processes operated in the species here described, but from a purely descriptive point of view it is necessary to distinguish between the two common aspects. It is proposed therefore to restrict the use of the term fulcral plate to those situations where there is a lower cavity beneath the floor of the socket and to use the term socket pad where the floor of the socket is raised above the general surface of the valve but does not have a lower cavity beneath it. In their recent critical discussion of dalmanellid cardinalia, Williams and Wright (1963, pp. 4-1 1) also pointed out that transitional stages between fulcral plates and socket pads are to be expected. The useful term ‘socket pad’ is adopted from their paper. ( b ) Brachiophores. This term is used to designate the plate-like structures which de- velop forward from the margins of the notothyrium, forming the inner (medial) margins of the sockets. No separate ‘brachiophore processes’ or ‘brachiophore supports’, terms used by Schuchert and Cooper, are distinguished. ( c ) Rib branching. Although patterns of rib branching have been studied in the species here described, only minor differences occur and their taxonomic usefulness is outweighed by combinations of other morphological features. Nevertheless, the general pattern of branching in the medial region of the brachial valves of isorthids and salopi- nids has a character which immediately distinguishes these groups from Resserella [= Par mor this ], species of which have previously been confused with I. orbicularis. Davidson (1869, p. 218), for example, confused O. orbicularis (an isorthid) with O. elegantula (a resserellid) and found difficulty in convincing himself that O. canaliculata Lindstrom (again, an isorthid) was really different from O. basalis Dalman (a resserellid). In Isorthis and Salopina the branching of the costellae in the medial area of the brach- j ial valve is symmetrical about the shell’s plane of symmetry. Secondary and sometimes tertiary branches of the medial primary pair, appear on the medial side. In Resserella , however, an unusual asymmetrical arrangement occurs in which the successive branches appear alternately to ‘left’ and ‘right’ of the preceding costella. This occurs in all known species of Resserella and is responsible for the unusual appearance of the ribbing in the V. G. WALMSLEY: ISORTHIS AND SALOPINA (BRACHIOPODA) 457 sulcus, to which several authors have referred — notably Schuchert and Cooper 1932, p. 129, in discussion of Parmorthis. Text-fig. 1 shows diagrammatically the typical arrangement of medial costellae on the brachial valve of isorthids and salopinids. The abbreviations used in the following systematic descriptions are explained in this diagram. The 5 mm. length stage is chosen as arbitrary, but convenient for this group of brachio- pods. This figure also shows diagrammatically by contrast, the asymmetrical arrange- ment of the branching of medial costellae on the brachial valve of a resserellid shell. There are of course other generic distinctions, but the asymmetrical pattern of rib branching has not hitherto been recorded and is an easily observed external distinction. text-fig. 1 . Contrast in patterns of costellae bifurcation in lsorthis and Salopina (on left) and Resserella (on right), in the medial region of the brachial valve. Also, key to notation of costellae referred to in the text. Number of ribs. The abundance, strength, and angularity of costellae as well as their branching patterns can be important distinguishing characters and various authors have attempted to record these features. However, since the number of costellae reaching the anterior margin in a branching system depends to a large extent on the stage of growth of the shell, figures given for the number of costellae, or even the number per mm., along the anterior margin are not meaningful unless the size of shell is also quoted. Rarely in description of enteletacean brachiopods have figures concerning the frequency of costellae been related to size. Where the frequency of costellae is referred to in the following descriptions it is taken as the average number per mm. at the 5 mm. length growth stage. SYSTEMATIC DESCRIPTIONS Superfamily enteletacea Waagen 1884 [Nom. transl. Altskhova 1960 (ex Entelitinae Waagen 1884)] Family schizophoriidae Schuchert and LeVene, 1929 Subfamily schizophoriinae Schuchert and LeVene, 1929 Genus salopina Boucot, 1960, emended Type species. Orthis lunata J. de C. Sowerby in Murchison, 1839, p. 611, pi. 5, fig. 15. Diagnosis. Relatively small schizophoriinids having a gently convex brachial valve, a more convex pedicle valve, hollow costellae, a poorly developed median ridge in the 458 PALAEONTOLOGY, VOLUME 8 pedicle valve and an adductor muscle field in which transverse ridges may be oblique and the margins meet the bases of the brachiophores in a sharp flexure. Comparison. Salopina is distinguished from Schizophoria by having the pedicle valve always more convex than the brachial and by the absence of both a prominent median ridge in the pedicle valve and prominent antero-laterally directed transverse ridges in the adductor muscle field. From Isorthis, Salopina is distinguished by its hollow costellae, and by the absence both of a distinct median ridge in the pedicle valve and of a strongly impressed adductor muscle field. The flexure by which the lateral margins of the adductor muscle field merge into the bases of the brachiophores in Salopina is distinct from the smooth con- tinuity in the case of Isorthis. The genus Sphenophragmus Imbrie 1959 also has hollow costellae and a general form very close to Salopina but is distinguished by its delicate cardinalia with a cardinal pro- cess raised on a notothyrial platform and the absence of a pronounced median ridge in the brachial valve. In the pedicle valve, the dental lamellae are not medially concave and there is no raised median area in the delthyrial cavity. Description. Relatively small, subequally biconvex to almost plano-convex schizo- phoriinids with pedicle valve always having greater convexity. Brachial valve non-sulcate to gently sulcate. Outline variable from semicircular to subcircular or transversely ellipti- cal. Hinge line straight, usually equal to about two-thirds of the greatest width, which is near mid-length. Lateral commissures straight to slightly flexed, anterior commissure crenulate and either rectimarginate or unisulcate. Pedicle valve interarea longer than interarea of brachial valve, usually gently curved, apsacline. Brachial valve interarea plane, anacline. Delthyrium and notothyrium open, latter usually partially filled by the myophore of the cardinal process. Pedicle valve beak usually distinct but not over- hanging hinge line. Shell punctate, multicostellate with hollow costellae. Costellae increase by bifurcation and vary in degree of curvature as they radiate from the umbo. Costellae rounded to subangular. At 5 mm. length the number of costellae per mm. varies with species but is usually about three. The rate of appearance of secondary and tertiary branches varies with species. Interior of pedicle valve. The muscle field is commonly weakly impressed, confined to the posterior half and the median third of the valve and in some species is even more restricted. No pronounced median ridge is present but there is a very faintly raised median area separating the elongate diductor tracks which are bounded laterally by low, for- ward extensions of the medially concave dental lamellae. Teeth range in size with species and do not bear crural fossettes. Small lateral cavities external to the dental lamellae extend beneath the interarea. No pedicle callist present. Interior of brachial valve. A commonly weakly impressed adductor muscle field, one- third or less the width of the valve, extends up to three-quarters of the valve length. A wide median ridge, at least a quarter as wide as the muscle field, separates the adductor impressions which are bounded laterally by not very pronounced, slightly raised margins sub-parallel to the median ridge. These margins become weaker as they curve to meet the median ridge in front of the muscle field. Posteriorly they are directed laterally to the brachiophores, but by a sharp flexure merge into the brachiophore bases (see PI. 65, fig. 1). Faint transverse ridges directed normal to the median ridge or slightly antero- V. G. WALMSLEY: ISORTHIS AND SALOPINA (BRACHIOPODA) 459 laterally give a quadripartite aspect to the adductor field. The brachiophores vary in thickness and degree of divergence. Fulcral plates are present in most species, flooring triangular sockets beneath which are distinct, sometimes deep, lateral cavities. The cardinal process has a short, thin shaft and a slightly bulbous simple non-lobed myophore bearing crenulations on its posterior face. Crenulations of the anterior margin of the shell are wide, low, flat or slightly rounded and bear a median groove. Interspaces are narrow and rounded. Crenulations commonly extend over one-fifth of the shell length. Salopina lunata (J. de C. Sowerby) Plate 64, figs. 15-27; Plate 65, figs. 1-12, 28, 29 1839 Orthis lunata J. de C. Sowerby in Murchison, p. 61 1, pi. 5, fig. 15 (non fig. 16) and pi. 3, fig. 12 d. 1869 Orthis lunata Davidson, pp. 215-16, pi. 28, figs. 1, 2, 4 (non figs. 3, 3a), ? fig. 5. non 1 845 Orthis lunata Murchison, de Verneuil and Keyserling, Geol. de la Russie d’Europe, vol. 2, p. 189, pi. 13, figs. 6 a-d. non 1913 Dalmanella lunata Williams, Proc. U.S. Nat. Mus., 45, p. 337, pi. 3, figs. 1-5 and 8. non 1922 Orthis ( Dalmanella ) lunata Barrois, Pruvost and Dubois, Mem. Soc. geol. du Nord, 6, p. 77, pi. 11, figs. 4-12; p. 155, pi. 17, figs. 16-18. non 1924 Dalmanella lunata McLearn, p. 55, pi. 4, figs. 5-6. non 1960 Salopina lunata Boucot et al., p. 3, pi. 1, figs. 6-12. Diagnosis. Outline transversely elliptical to subquadrate, brachial valve sulcate with fine costellae which radiate with little curvature. Posterior costellae sub-parallel to hinge line in brachial valve. Brachiophores thin, long, pointed blades. Fulcral plates present. Comparison. The New Brunswick shell figured by Boucot (1960, pi. 1, figs. 6-12) as S. lunata is now recognized as S. submedia (McLearn) which differs from S. lunata especially in the nature of the costellae. These are coarser and develop a distinct curva- ture which results in the most posterior costellae meeting the hinge line at a fairly high angle. S. submedia , which is redescribed in ‘ Brachiopods and Gastropods of Southern New Brunswick’, Boucot, Johnson, Harper, and Walmsley (in press), also differs from S. lunata in its more circular outline and in the longer diductor impressions being bounded by less strongly curved dental lamellae. Description. Unequally biconvex, pedicle valve having greater convexity and brachial valve sulcate. Outline transversely elliptical to subquadrate. Hinge line straight, equal to about three-fifths of greatest width which occurs slightly anterior to mid-length. Length equals about three-quarters of the width. Anterior commissure, crenulate and faintly unisulcate, lateral commissures straight. Exterior of pedicle valve. Most convex in postero-median region, curvature decreasing toward lateral margins and slightly towards anterior margin. Outline subelliptical with anterior margin parallel to hinge line and lateral margins rounded, convex outwards. Cardinal angles obtuse and slightly less rounded than the antero-lateral margins. Beak projects one-tenth of the total length posterior to hinge line. Interarea triangular, with sharp lateral margins. Delthyrial margins subtending about 60°. Exterior of brachial valve. The shallow sulcus extends from umbonal region to anterior margin where it equals about a quarter of greatest width. Anteriorly diverging margins of sulcus poorly defined. Valve outline subelliptical with anterior margin parallel to 460 PALAEONTOLOGY, VOLUME 8 hinge line, lateral margins rounded, convex outwards. Cardinal angles obtuse, less rounded than anterolateral margins. Interarea plane, anacline, about half as long as area in pedicle valve. Rib branching pattern in median area shows P I (see text-fig. 1) quickly giving off secondary and second order branches towards midline. Secondaries give off tertiaries also medially, these being the most medial ribs on the valve. These do not arise together and points of branching can be one-fifth of the total length of valve apart. At 5 mm. length from the beak all these ribs have appeared, P II has also at this stage (5 mm.) produced secondary, tertiary and second order ribs so that increase in costellae is rapid. Interior of pedicle valve. The muscle field is confined to the posterior one-third of the valve and is also one-third as wide as greatest width. A very faintly raised median region is flat and up to a quarter as wide as the muscle field. In some specimens it narrows towards the anterior. This may represent the adductor muscle track but no distinct scars are impressed. Diductor tracks on either side are elongate, semi-elliptical with anteriorly convex anterior margins slightly raised above the valve floor. Laterally the diductor impressions are bounded by the dental lamellae which diverge outwards before converging again as walls of the delthyrial cavity. The teeth are small, triangular in plan EXPLANATION OF PLATE 61 Figs. 1-13. Isorthis amplificata sp. nov., Wenlock Limestone or basal Ludlovian siltstone, bank in field, 600 yards NW. of Porth-llong, Usk inlier, Britain (Nat. Grid Ref. 34829825). 1-2, Internal mould of pedicle valve (x 3), GSM 102206 and latex impression (x 3). Note deep crural fossettes. 3-4, Internal mould of pedicle valve (x3), GSM 102207, Paratype and latex impression (x3). Note vascular trunks. 5, Internal mould of brachial valve (x 3), GSM 102208. 6-7, Internal mould of brachial valve (x2), GSM 102209, and latex impression (x2). Note impressions of slightly crenulated sockets, thick brachiophores and their close proximity near the cardinal process. (See also figs. 20-21.) 8-9, Internal mould of brachial valve (x2), GSM 102210, and latex impression ( x 2). Note large brachiophores. 10-1 1 , Internal mould of brachial valve ( x 2), GSM 10221 1 , and latex impression (X2). 12-13, Internal mould of pedicle valve (x2), GSM 102212, and latex impression ( x 2). Figs. 14-16. Isorthis amplificata sp. nov., Wenlock Limestone or basal Ludlovian siltstone, old quarry at Cwm, Usk inlier, Britain (Nat. Grid Ref. 33320160). 14, Internal mould of pedicle valve (x2), GSM 102213. 15, Latex impression of same specimen (x2). 16. Internal mould of brachial valve (X 3), GSM 102214, Holotype. Figs. 17-19. Isorthis amplificata sp. nov., Wenlock Limestone or basal Ludlovian siltstone, bank in field, 600 yards NW. of Porth-llong, Usk inlier, Britain (Nat. Grid Ref. 34829825). 17, Internal mould of brachial valve (x2), GSM 102215. 18, Latex impression of same specimen (x2). 19, Oblique view of same latex impression showing divergence of brachiophores. Figs. 20-30. Isorthis amplificata sp. nov., Wenlock Limestone old quarry on west side of Lincoln Hill, Iron Bridge, Britain (Nat. Grid Ref. 670038). 20-24, Anterior, posterior, side, internal and external views of brachial valve (x3), GSM 102216, Paratype. 25, Interior of pedicle valve (x2), GSM 102217. 26-30, Posterior, anterior, side, pedicle valve and brachial valve (x2), GSM 102218, Paratype. Figs. 31-38. Isorthis clivosa sp. nov., 31-32, Mid Eltonian siltstone, trackside 300 yards south of Hafod-fawr, Llandovery area, Britain (Nat. Grid Ref. 81443096), internal mould of pedicle valve (x2), GSM 102219, and latex impression (x2). 33-34, Mid-late Eltonian, 6 yards upstream of mountain fence along Nant Cwm Clyd, 1,000 yards SE. of Cwm Clyd Farm, Llandovery area, Britain (Nat. Grid. Ref. 80152980), internal mould of brachial valve (x2), GSM 102220, and latex impres- sion ( x 2). 35-38, Mid Eltonian, 50 yards up R. Gwydderig from Bont Wen, on road A. 40. 35-36, External mould of pedicle valve (x2), GSM 102221, and latex impression (x2). 37-38, External mould of brachial valve ( x 2), GSM 102222, and latex impression ( x 2). Palaeontology , Vol. 8 PLATE 61 WALMSLEY, Isorthis V. G. WALMSLEY: ISORTHIS AND SALOPINA (BRACHIOPOD A) 461 and section and project normal to the hinge line. External to the teeth are lateral cavities roofed over by the interarea and bounded by the external face of the tooth and the floor of the valve. In large specimens these cavities may be obscured. Interior of brachial valve. The adductor muscle field extends about four-sevenths of the distance to the anterior margin and is slightly less than one-third as wide as the greatest shell width. A rounded, median ridge (about two-fifths width of muscle field) extends from the anterior margin of the muscle field posteriorly becoming somewhat carinate as it passes between the brachiophores. Flanking the median ridge are adductor impressions, narrower than the median ridge and bounded by very faintly raised margins which are sub-parallel to the median ridge, but slightly concave towards it. Posteriorly these margins, which are directed externally to the brachiophore bases, curve sharply medially and merge with the bases of the brachiophores. Anteriorly the margins fade towards the end of the median ridge. Anterior adductor impressions are smaller than the posterior pair and when slightly more impressed are clearly separated by faint trans- verse ridges. In many specimens the impressions are faint, the outer margins incomplete and the appearance is of divergent margins extending from immediately outside the brachiophores, fading about mid-length where the median ridge also may fade. The brachiophores are very long, thin, erect, pointed blades directed normal to the hinge line or only slightly divergent from each other. Their distal extremities are separated by a distance equal to one-fifth of greatest shell width. Their anterior edges are normal to the commissural plane. Proximally they are connected by short curved fulcral plates to the posterior edge of the valve floor along the hinge line. The fulcral plates floor very small sockets and roof over a much larger lateral cavity. The cardinal process consists of a thin blade-like shaft which arises from the crest of the posterior carinate portion of the median ridge between the brachiophores and terminates in a very slightly bulbous, non-lobed myophore. Anterior crenulations on the valve margin have a low rectangular cross-section and bear fine grooves along the mid-line of their internal faces. Interspaces are rounded. This type of anterior crenulation is very similar to that described as diagnostic of Rhipidomelloides. Type specimens. This species was erected and first described by J. de C. Sowerby in Murchison’s Silurian System, p. 611, and illustrated in pi. 3, fig. 12 d and pi. 5, fig. 15. Boucot (1960, p. 3) designated as ‘holotype’, ‘The specimen figured by Salter (op. cit.)’, referring to pi. 5, fig. 15. The figure was by J. de C. Sowerby, not Salter, and comprised not one but three specimens, an internal impression of a pedicle valve, an internal impression of a brachial valve and an external impression of a pedicle valve. None of the originals of these figures has been definitely identified but they may be related to GSM 51555 in the Geological Survey Museum, London. Specimen No. GSM 51555 is from the Geological Society Collection — Murchison Collection, Oaker Quarry (near Beecher) and the old label states ‘ PI. 5, fig. 1 5 ’. The impressions are all slightly distorted. Similar distortion is to be seen in specimens from Oaker Quarry (Nat. Grid reference 377816) about 5 miles east of Clun, but not in material from the Ludlow district or the inliers of Woolhope, May Hill, or Usk. Dr. Stubble- field (in correspondence) considers that this specimen is remotely eligible as the original of pi. 5, fig. 15, and is quite probably one of several syntypes. In order to clarify the situation, I designate the ringed specimen on GSM 51555 (an internal impression of a brachial valve) as lectotype and figure it Plate 65, fig. 28. It is arguable whether a specimen which is not certainly the original should be designated lectotype. However, the Geological Society Collection has been very carefully examined by the Geological Survey and it seems certain that no more likely specimen exists. To designate GSM 51555 as a neotype would hardly assign to this specimen its due significance. 462 PALAEONTOLOGY, VOLUME 8 Distribution. S. lunata ranges in the Ludlovian of Britain from the Lower Leintwardinian to the highest Whitcliffian where it is most abundant. In the inliers of Usk, May Hill, and Woolhope it has been recorded as present but very rare in the basal part of the Downtonian — immediately above the Ludlow Bone Bed. It should be noted that the original of Sowerby’s pi. 3, fig. 12 d, also of S. lunata, cannot be traced, but the tablet bearing specimens figured as 12a, b, c, e, and g is marked: ‘Devonian (Tilestones) Horeb Chapel, R. I. Murchison, F.G.S.’. In the Lower Leintwardinian, it is present at May Hill and Ludlow, rare at Usk and fairly common at Woolhope. In the Upper Leintwardinian it is fairly common at May Hill, Usk, Ludlow, and Builth, and common at Woolhope. In the Lower Whitcliffian, it is rare in the Leintwardine area, fairly common at Woolhope, Ludlow and Builth, and common at Usk and May Hill. In the Upper Whitcliffian, it is common at Usk, May Hill, Woolhope, Ludlow, the Leintwardine area, Builth, Knighton, and Kerry. It appears, therefore, to have migrated from the south-eastern part of the shelf area north-westwards to the basin area during the later part of Ludlovian time. (See also Holland and Lawson, 1963, p. 287.) Records of S. lunata from localities outside Britain have been checked and cannot be authenticated. Its sudden appearance in Britain in early Leintwardinian times raises the question of its ancestry. In the Hemse Group of Gotland (early Ludlovian age, probably mid Eltonian based on associated Pristiograptus nilssoni and Lobograptus scanicus), is a shell which is considered to be conspecific with D. conservatrix McLearn. The latter is a salopinid and is redescribed in ‘ Salopinid Brachiopods of the Silurian and Lower Devonian’ (Walmsley, Boucot, and Harper, in preparation). S. conservatrix is known from the Ross Brook Formation (late Llandovery age) of Nova Scotia. S. conservatrix seems to be the most likely ancestor of S. lunata. A species of Salopina from the Stonehouse Formation (Gedinnian age) of Arisaig, Nova Scotia, (Walmsley, Boucot and Harper, in preparation) is very close to S. [ Dalmanella ] missendenensis (Straw), from the Little Missenden Bore (1913) material. Unfortunately, there is too little bore material avail- able to establish specific identity. The age of the bore material was recorded, Straw and Woodward 1933, p. 139, as ‘either late Downtonian or post-Downtonian ’. It seems likely that both the Stonehouse and Little Missenden forms were derived from S. lunata. Subfamily isorthinae Schuchert and Cooper, 1931 Genus isorthis Kozlowski, 1929, emended Type species. I. szajnochai Kozlowski, 1929. Diagnosis. Planoconvex to equally biconvex isorthinids lacking hollow costellae, having a well-developed median ridge separating elongate diductor tracks in the pedicle valve, and a well-impressed adductor muscle field in which the transverse ridges are normal to the median ridge in the brachial valve. Cardinal process commonly simple, non-lobed especially in Silurian species, bilobed in some Devonian species. Brachiophores erect or postero-laterally inclined, never convergent on to the median ridge. Comparison. Isorthis is distinguished from Schizophoria by having the greater convexity always in the pedicle valve and by having transverse ridges in the adductor muscle field aligned normal to the median ridge whereas in Schizophoria they are oblique, directed antero-laterally from the median ridge and the brachial valve has the greater convexity. Schuchert and Cooper (1932, p. 149) suggested that Isorthis was probably derived from an early Silurian Schizophoria. No true Schizophoria is known, however, from the Silurian and it appears more likely that Schizophoria in the Devonian was in fact derived from Isorthis. Salopina is distinguished from Isorthis by having hollow costellae (see Boucot et al. 1 960, pi. 1, fig. 12) and by the distinctive flexure where the margins of the adductor muscle field meet the bases of the brachiophores (see PI. 65, fig. 1 and contrast with PI. 63, figs. 10 and 14). The median ridge in the pedicle valve is also less pronounced. V. G. WALMSLEY: ISORTHIS AND SALOPINA (BRACHIOPODA) 463 The distinction between Isorthis and both Dalmanella and Onniella as emended by Williams and Wright (1963, pp. 27, 28) lies mainly in the shorter more cordate pedicle muscle field of the latter genera and in the case of Dalmanella the convergence of the brachiophore bases on to the median ridge. Levenea is distinguished from Isorthis by its pentagonal delthyrial cavity and muscle field in the pedicle valve. Description. Biconvex, varying from almost planoconvex to subequally biconvex. Pedicle valve always having greater convexity. Brachial valve non-sulcate to sharply sulcate. Outline variable, sub-circular, transversely elliptical or shield shaped. Hinge line straight, usually between half and three-quarters as long as greatest shell width, which is near mid-length. Lateral commissures straight, anterior commissure recti-marginate and crenulate or unisulcate and crenulate. Pedicle valve interarea usually twice as long as brachial valve interarea, gently incurved, apsacline. Delthyrium triangular, open. Brachial valve interarea plane, anacline to orthocline. Notothyrium triangular, open, partially filled by cardinal process. Pedicle valve beak distinct but not overhanging hinge line, brachial valve beak slight. Shell multicostellate, costellae increasing by bifurcation. Pattern of bifurcation in medial region of brachial valve characterized by median pair of primary costellae branching medially very early. In some species these secondary costellae again branch medially to produce tertiary costellae. A second order of branching from the median primary may arise. The second pair of primary costellae produce secondaries and both primary and secondary costellae may again divide pro- ducing tertiaries and second order branches. At 5 mm. length the presence of tertiary and second order branching varies with species. All this branching is symmetrical as in Salopina, but quite different from the pattern developed in Resserella. The costellae are rounded, evenly spaced with rounded interspaces (striae). At 5 mm. length the number of costellae per mm. varies between three and five and is most commonly four. Interior of pedicle valve. The delthyrial cavity is usually fairly deep, but may have a delthyrial platform on which the median ridge and diductor tracks are raised above the general floor of the valve. In some species the anterior edge of the platform makes a steep descent. Dental lamellae vary in thickness and in their inclination relative to the plane of symmetry of the shell. A pedicle callist is present in some species. Teeth, triangular in plan and cross-section project normal to the hinge line, sometimes curving slightly posteriorly. Crural fossettes are usually present and vary in depth and position on the medial surface of the tooth. A median ridge extends from the delthyrial cavity, not reach- ing beyond the diductor impressions. The ridge varies in length, height, and width and may be parallel sided, narrow anteriorly, or widen anteriorly. In some species sub- elliptical adductor impressions occupy the posterior end of the ridge. In others no distinct adductor impressions are visible but faint striations along the median ridge may re- present an adductor track. Diductor tracks, separated by the median ridge, may be slightly or deeply impressed, narrow, elongate, sub-parallel or divergent anteriorly, rarely extending beyond mid-length. The diductor impressions have a smooth, non- flabellate outline and do not enclose the adductor impressions anteriorly. Pallial marking rarely visible. In some species vascular trunks extend forward from the anterior limit of the diductor tracks, diverging in line with these tracks, towards the anterior margin. 464 PALAEONTOLOGY, VOLUME 8 Crenulations of the anterior shell margin usually low and rounded but variable. Interior of brachial valve. The sub-circular to elongately elliptical adductor muscle field occupies the median third of the valve, usually extending slightly anterior to mid-length. A median ridge is always present varying from one-eighth to one-quarter as wide as the muscle field, usually constant in width, rarely tapering anteriorly. The degree of im- pression of the muscle scars varies, resulting in some species in a distinct transverse ‘ step ’ or ridge normal to the median ridge, separating the anterior and posterior pairs of impressions and giving a clearly quadripartite appearance, with the anterior pair either equal to or larger than the posterior pair. In other species there is no sign of such transverse marking. The muscle field is bounded by a raised margin of variable outline which anteriorly converges and fades towards the median ridge and posteriorly merges smoothly with the bases of the brachiophores. The brachiophores vary in angle of divergence, thickness, and length. In most species they are thin, flaring plates. Fulcral plates connect the lateral face of the brachiophore to the posterior margin of the valve and floor the dental sockets which vary in size. In some shells, socket pads are present instead of fulcral plates. The cardinal process EXPLANATION OF PLATE 62 Figs. 1-22. Isorthis clivosa sp. nov. 1-4, Mid Eltonian siltstone, trackside 300 yards SE. of Hafod-fawr, Llandovery area, Britain (Nat. Grid Ref. 81443096). 1-2, Internal mould of brachial valve (x2), GSM 102223, and latex impression (x 2). 3-4, Internal mould of pedicle valve (x 2), GSM 102224, and latex impression ( x 2). 5-6, Early Leintwardinian, roadside quarry, 1,080 yards ENE. of Mary Knoll House, Ludlow area, Britain (Nat. Grid Ref. 49107399), internal mould of pedicle valve (x2), GSM 102225, and latex impression (x2). 7-8, Early Bringewoodian, roadside on A. 40, 190 yards WNW. of Ynys-y-Bont, Llandovery area, Britain (Nat. Grid Ref. 83723225), internal mould of pedicle valve (x2), GSM 102226, and latex impression (x2). Note vascular trunks. 9-10, Early Bringewoodian, roadside, 1,000 yards east of Sluvad Farm, Usk inlier, Britain (Nat. Grid Ref. 32509920), internal mould of pedicle valve ( x 3), GSM 102227, and latex impression ( x 3). Note crural fossettes, thick dental lamellae, and wide median ridge. 1 1-12, Mid Eltonian, trackside, 300 yards south of Hafod-fawr, Llandovery area, Britain (Nat. Grid Ref. 81443096), internal mould, of brachial valve ( x 2), GSM 102228, Paratype. Note pear-shaped outline of adductor muscle field, and latex impression (x2). 13-14, Early Bringewoodian, roadside on A. 40, 190 yards WNW. of Ynys-y-Bont, Llandovery area, Britain (Nat. Grid. Ref. 83723225), internal mould of pedicle valve ( x 2), GSM 102229, Paratype, and latex impression ( x 2). Note abrupt end of median ridge. 15-18, Mid Eltonian, stream bank 600 yards SE. of Esgair llaethdu, Llandovery area, Britain (Nat. Grid Ref. 79022926). 15-16, Internal mould of pedicle valve (x2), GSM 102230, Paratype, and latex impression ( x 2). 17-18, Internal mould of pedicle valve ( x 3), GSM 102231 , and latex impression ( x 3). 19-20, Late Leintwardinian, roadside, 1,230 yards ENE. of Mary Knoll House, Ludlow area, Britain (Nat. Grid Ref. 49227407), internal mould of pedicle valve (x3), GSM 102232, Holotype, and latex impression (x 3). Note flat median ridge with abrupt anterior end and also the anterior crenulations. 21-22, Mid Eltonian, trackside 300 yards south of Hafod-fawr, Llandovery area, Britain (Nat. Grid Ref. 81443096), external mould of pedicle valve (x2), GSM 102233, and latex impression (x 2). Figs. 23-35. Isorthis slitensis sp. nov. 23-27, Slite Marl (Wenlockian), Cement plant quarry at Slite, Island of Gotland (Baltic), pedicle valve, side, brachial valve, anterior and posterior views ( x 2), GSM 102234, Paratype. 28-32, Slite Marl, canal ditch 800 m. SW. of Alby (north of Lergravsviken), Parish of Rute, Gotland, exterior, interior, anterior, posterior, and side views of brachial valve ( x 2), GSM 102235, Holotype. Note relatively long adductor muscle field and vascular marks. 33-35, Slite Marl, Cement plant quarry at Slite, Gotland. 33, Interior of pedicle valve ( x 2), GSM 102236, Paratype. 34, Interior of pedicle valve x (3), GSM 102237. Note relatively long diductor tracks and median ridge. 35, Interior of pedicle valve (x 3), GSM 102238. Note anterior crenulations. Palaeontology, Vol. 8 PLATE 62 WALMSLEY, Isorthis V. G. WALMSLEY : ISORTHIS AND SALOPINA (BR ACHIOPOD A) 465 arises as a thin laterally compressed or semi-cylindrical shaft from the median ridge between the brachiophores and expands slightly into a myophore which may be simple, non-lobed, and slightly bulbous, occupying the apical portion of the notothyrium, or may be distinctly bilobed as in several Devonian species. In well-preserved specimens the posterior face of the myophore often shows a series of chevron-shaped crenulations arranged with their apices pointing anteriorly. Geological Range. Silurian (Lower Llandoverian) to Devonian (Eifelian, ? Lower Givetian). Remarks. Kozlowski (1929, pp. 29 and 75) regarded the near equality of the convexity of the two valves and the simple cardinal process as being amongst the distinguishing characteristics of Isorthis. Schuchert and Cooper (1932, p. 149) described the lateral profile as ‘unequally to sub-equally biconvex’ and stated that the cardinal process was ‘small, bilobed, trilobed or multilobed’. In the twenty species of Isorthis which have been studied, the biconvexity varies from almost plano-convex to sub-equally convex. Silurian species generally have a simple cardinal process. Some Devonian species have a bilobed process. Isorthis orbicularis (J. de C. Sowerby) Plate 63, figs 1-15; Plate 64, figs. 1-14; Plate 65, figs. 30-31 1839 Orthis orbicularis J. de C. Sowerby in Murchison, p. 611, pi. 5, fig. 16 (non fig. 15). non 1924 Dalmanella orbicularis McLearn, p. 56, pi. 4, fig. 7. non 1942 Dalmanella orbicularis Dahmer, Senckenbergiana, 25, p. 116, figs. 14-16. non 1951 Dalmanella orbicularis Dahmer, pp. 91-94, pi. 7, fig. 1 ; pi. 9, figs. 20-21 ; pi. 10, fig. 6; pi. 11, fig. 22; pi. 12, fig. 10. non 1960 Isorthis orbicularis Boucot et al., p. 5, p. 1, figs. 13-20; pi. 2, figs. 1-7. Diagnosis. Unequally biconvex, sulcate, with circular to transversely elliptical outline. Median ridge of pedicle valve narrow, rounded, extending to centre of valve and separating sub-parallel diductor tracks which do not extend into anterior half of valve. Posteriorly, the median ridge separates a pair of short semi-elliptical adductor im- pressions. Brachial valve adductor muscle field well impressed, bounded by sub-circular raised margin, confined to posterior half of valve and distinctly quadripartite. Median ridge relatively wide and constant in width, posteriorly thickened between brachiophores. Muscle impressions sub-equal. Brachiophores short, divergent, and slightly flaring blades. Cardinal process non-lobed. Comparison. The distinctions between I. orbicularis and the other isorthids described in this paper are given under the respective species. I. orbicularis has been confused with Salopina lunata but these are now seen to represent different genera. I. orbicularis has also been confused with I. fornicatumcurvata Fuchs. The more elongate and less distinctly, less constantly quadripartite adductor muscle field of the later species, however, together with greater convexity of the brachial valve, clearly distinguish it. Description. Unequally biconvex to almost plano-convex, with pedicle valve at least twice as deep as brachial valve. Circular to transversely elliptical outline with well- 466 PALAEONTOLOGY, VOLUME 8 rounded cardinal angles. Anterior commissure crenulate and slightly sulcate, lateral commissure straight. Hinge line equal to two-thirds of greatest width, which is at mid- length. Width slightly greater than length, thickness almost half length. Exterior of pedicle valve. Evenly convex with distinct beak projecting one-sixth total length posterior to hinge line. Delthryium margins subtending about 60°. Lateral margins of the interarea smoothly rounded. Exterior of brachial valve. Slightly convex with shallow median sulcus widening from umbo to anterior margin where it equals about one-third of shell width. Lateral margins of sulcus not sharply defined. Outline, sub-circular to transversely elliptical with antero- lateral margins rather more rounded than postero-lateral margins. Interarea anacline. Notothyrium enclosing about 60°. The pattern of rib branching in the median area of the brachial valve consists of a median pair of P I, with P Is, P It and P I 2nd, and P II with P IIs. At 5 mm. length all these except sometimes P It, have appeared (see text-fig. 1). Interior of pedicle valve. A low, rounded, narrow median ridge extends from the delthyrial cavity to almost mid-length where it fades gradually. In the posterior part of the delthyrial cavity this ridge separates short subelliptical adductor scars, often well impressed. Anterior to the adductor impressions the sub-parallel diductor tracks are separated by the median ridge and end with rounded margins at mid-length. Laterally the diductor tracks are bounded by straight sub-parallel or slightly convergent ridges which extend forward from the dental lamellae, decreasing in height to mid-length. The teeth, which bear crural fossettes on their medial surfaces, are supported by dental lamellae which diverge slightly ventro-laterally before converging towards the floor of the delthyrial cavity. A small pedicle callist is present. Interior of brachial valve. The well-impressed adductor muscle field is confined to the posterior half of the valve, and is bounded by a circular to slightly elongate raised mar- gin, which merges posteriorly with the bases of the brachiophores and anteriorly fades as it curves towards the median ridge. A less well-developed pair of ridges runs antero- laterally from the junction of the anterior and posterior pairs of adductors. Four adductor impressions of sub-equal size are separated by a broad, low, rounded median ridge which is about one-quarter to one-third as wide as the muscle field and by narrower transverse ridges which are either parallel to the hinge line or directed slightly postero- laterally. This arrangement of ridges separating the muscle impressions gives a distinctive ‘cross’ pattern. Short, stubby, plate-like brachiophores flare antero-laterally. Antero- laterally directed sockets are supported by fulcral plates or, often in larger specimens, by socket pads. The cardinal process has a cylindrical shaft and a conical non-lobed myophore. Anterior crenulations are low, rounded without median grooves and are evenly spaced. Pallial marks are rarely seen but a pair of trunks running anteriorly from the adductor impressions divide before reaching the anterior margin. From the same origin a second pair run lateral to them. Variation in muscle field. Most variation occurs in the brachial muscle field where the highly symmetrical cross pattern may be extended to an elongate, roughly elliptical out- line with irregularly raised margins and transverse ridges which are obscure or faint, depending upon the degree of impression of the muscle scars. The pedicle valve shows some slight variation in the ridges bounding the diductor tracks. These ridges may be V. G. WALMSLEY : ISORTHIS AND SALOPINA (BR ACHIOPOD A) 467 straight and almost parallel, straight but slightly convergent anteriorly, or slightly curved (convex outwards). Type specimens. This species was first described by J. de C. Sowerby (in Murchison, Silurian System, p. 611). It was figured pi. 5, fig. 16 which is a multiple figure. The figured specimens are in the collection of the Geological Survey and Museum of Great Britain and are believed to be related to fig. 16 as follows : Geol. Soc. Colin. No. 6647 is taken to be the original of pi. 5, fig. 16, top left. Geol. Soc. Colin. No. 6648 is thought to be the original of fig. 16, bottom left, top right, and bottom right, but GSM 33239 may also have been used in drawing fig. 16, bottom right. Geol. Soc. Colin. No. 6647 is a poorly preserved internal mould of a pedicle valve. Geol. Soc. Colin. No. 6648 is a complete shell and GSM 33239 is a poorly preserved exterior of a brachial valve. In his brief description of O. orbicularis, J. de C. Sowerby ( Silurian System, p. 611) mentions many localities in the Welsh Borderlands, but does not state the horizon or localities from which his specimens figured in pi. 5, fig. 1 6 were obtained. Davidson (1869, pp. 215 and 218) included this species in O. lunata. McLearn (1924, p. 56) records D. orbicularis (J. de C. Sowerby) as occurring rarely in boulders in Stonehouse field at Arisaig, Nova Scotia — ‘probably from zone D of the Stonehouse Formation’. From his illustration pi. 4, fig. 7, it is not possible to say whether this is I. orbicularis but examination of a large collection of material from the Stonehouse Formation has failed to reveal a single specimen of I. orbicularis. I. cf. fornicatumcurvatci Fuchs, however, not referred to by McLearn is fairly abundant — and it seems likely that this record of I. orbicularis from Arisaig is mistaken. Dahmer (1951, p. 91) stated that the type was the specimen figured in Silurian System, pi. 5, fig. 16, but quoted two specimen numbers, 6647 and 6648, and referred to ‘ the figured ventral valve’. No. 6647 is a pedicle valve but 6648 is a complete shell. His selection of a type is thus not clear and in order to clarify the situation, Geol. Soc. Colin. No. 6647 is here selected as the lectotype. Geol. Soc. Colin. No. 6648 and GSM 33239 are figured syntypes. Specimens numbers 6647 and 6648 are refigured, pi. 65, figs. 30 and 3 1 . It should be noted, however, that the species from the Ebbe Sattel of Germany described and figured by Dahmer as D. orbicularis is a different species, I.fornicatumcurvata (Fuchs). Despite the fact that Dahmer put this latter species in synonomy with I. orbicularis, the two species are quite distinct, as detailed under Comparison, above p. 465. Boucot, et al. (1960, p. 5, pi. 1, figs. 13-20 and pi. 2, figs. 1-7) described a species from the Sutherland River Formation of Devon Island (Canadian Arctic Archipelago), as I. orbicularis and recorded the occurrence as the first known from North America. Examination of these silicified shells has revealed, however, that they differ in a number of important characteristics from I. orbicularis. In particular, the absence of a sulcus ; the lack of transverse ridges and the faintness of the median ridge in the brachial muscle field, which is thus not distinctly quadripartite and is only weakly impressed; the relatively shorter hinge line and the narrow double median ridge in the pedicle valve (see Boucot, pi. 2, fig. 3). Distribution. I. orbicularis is a common shell in the shelly facies of the Ludlovian of the Welsh Border- land. It ranges from the base of the Ludlovian (lowest Eltonian) to the lower part of the Whitcliffian but is most abundant in the lower Leintwardinian. In the basin facies of the Ludlovian in Central Wales, it has been recorded as common in the Tower siltstone group’ (see Lawson, 1960, p. 121), and rare or absent above the zone of Monograptus leintwardinensis. It has been recorded from several places outside Britain but examination of Boucot’s collections from Gotland, Germany, Podolia, New Brunswick, Nova Scotia, Quebec and Maine reveals that the isorthids in these areas are different species from I. orbicularis, several of them being new species which are being described in a separate paper. So far, no specimens of I. orbicularis (J. de C. Sowerby) have been seen by the author from localities outside the Welsh Borderland. Isorthis slitensis sp. nov. Plate 62, figs. 23-35. Diagnosis. Unequally biconvex, faintly sulcate, sub-circular outline with only slightly rounded cardinal angles. Median ridge of pedicle valve narrow, carinate to rounded. 468 PALAEONTOLOGY, VOLUME 8 extending into anterior half of valve and separating wide parallel diductor tracks. Brachial valve adductor muscle field moderately well impressed, bounded by circular raised margin, extending into anterior half of valve, distinctly quadripartite, muscle impressions of equal size. Median ridge of constant width, thickened posteriorly between brachiophores. Brachiophores moderately thick, straight, divergent blades. Comparison. This species is very close to I. orbicularis, but is distinguished by having the following: (i) less circular outline with less rounded cardinal angles, (ii) sulcus less defined, (iii) the valves of almost equal length, (iv) both pedicle and brachial valve muscle fields relatively longer, extending into the anterior half of the shell. Description. Unequally biconvex to almost plano-convex with pedicle valve at least twice as deep as the brachial valve. Circular to sub-quadrate outline with cardinal angles only slightly rounded and antero-lateral margins well rounded. Anterior commissure crenu- late, slightly sulcate, lateral commissures straight. Hinge line equal to two-thirds of greatest width which is at mid-length. Width slightly greater than length, thickness slightly more than half of the length. Exterior of pedicle valve. Evenly convex with distinct beak which barely projects posterior to the brachial valve. Lateral margins of the interarea smoothly rounded. Margins of delthyrium enclosing 60°. Exterior of brachial valve. Slightly convex with faint median sulcus widening anteriorly. Outline somewhat shield-shaped with cardinal angles only slightly rounded. Interarea anacline. Notothyrium enclosing about 70°, partially filled by the cardinal process which protrudes slightly. The branching of the costellae in the median area of the brachial valve results in P I and P II having P Is and P IIs at 5 mm. length. Tertiary and second order branches appear only at or after this stage. Interior of pedicle valve. A median ridge, carinate in small forms, becoming rounded in larger shells, extends from the delthyrial cavity to about mid-length of the valve where it ends in a gradual slope. In one large shell the median ridge bears a very faint median groove on its posterior two-thirds. No separate adductor impressions are seen. Parallel diductor tracks which extend beyond the anterior end of the median ridge and are separated by it, terminate imperceptibly just anterior to mid-length. Laterally the diduc- tor tracks are defined by low ridges extending forward from the bases of the dental EXPLANATION OF PLATE 63 Figs. 1-15. Isorthis orbicularis (J. de C. Sowerby). 1-4, Early Leintwardinian, south end of old quarry, behind Darran Farm, Usk inlier, Britain (Nat. Grid Ref. 32759795). 1 , Internal mould of pedicle valve ( x 3), GSM 102239. Note impressions of adductor and diductor muscles. 2, Latex impression (x 3) of same specimen. Note diductor tracks confined to posterior half of valve. 3, Enlarged view ( x 5) showing adductor and diductor impressions and subparallel dental lamellae. 4, Side view of specimen in fig. 1, showing absence of delthyrial platform. 5-8, Early Bringewoodian, roadside quarry at Porth-llong, Usk inlier, Britain (Nat. Grid Ref. 35209780), posterior view (x 3), internal mould of pedicle valve ( x 2), GSM 102240, latex impression ( x 2) and enlarged ( x 5). Note adductor scars in relation to the narrow median ridge. 9-1 5, Early Leintwardinian, south end of old quarry behind Darran Farm, Usk inlier, Britain (Nat. Grid Ref. 32759795). 9-10, Internal mould of brachial valve ( x 2), GSM 102241 , and latex impression ( x 2). 1 1-12, External mould of brachial valve ( x 2), GSM 102242, and latex impression (x2). 13-15, Internal mould of brachial valve (x2), GSM 102243, latex impression (x 2) and enlarged view of mould (x 5). Note adductor muscle field con- fined to posterior half of valve and also vascular trunks. Palaeontology , Vol. 8 PLATE 63 WALMSLEY, Isorthis V. G. WALMSLEY: ISORTHIS AND SALOPINA (BRACHIOPODA) 469 lamellae. The ridges terminate opposite the anterior end of the median ridge. The teeth are strong, have a rounded lateral profile, and deeply incised crural fossettes on the anterior edges of their medial faces. Dental lamellae diverge slightly outwards before converging slightly towards the floor of the delthyrial cavity. A small pedicle callist is present. Interior of brachial valve. Well-impressed adductor muscle scars form a distinct quadri- partite pattern surrounded by a circular to slightly elongate raised margin, the lateral portions of which fuse with the brachiophore bases. Anteriorly the raised margin decreases as it curves in towards the median ridge. The median ridge, which is broad and low, equals about one-quarter of the width of the muscle field. Transverse ridges caused by the deeper impressions of the anterior pair of muscle scars are parallel to the hinge line. The muscle scars are of equal size. The brachiophores diverge from the inner margins of the notothyrium, with a slight flare, at about 74° from each other. They are fairly thick stubby plates with a carinate posterior edge and bluntly rounded extremities. The postero-lateral faces of the brachio- phores merge into fulcral plates which floor the sockets or into socket pads. The triangular sockets widen antero-laterally and are slightly roofed over at the posterior end by the interarea. The cardinal process, which arises from a posterior thickened and raised portion of the median ridge (notothyrial platform), has a short, thick cylindrical shaft and a conical myophore which largely fills the notothyrium from which it protrudes slightly. This projecting posterior portion of the myophore bears five or six fine crenula- tions arranged in a chevron pattern with apices directed anteriorly. Along the anterior margin of the valve, the crenulations are low and rounded with narrower interspaces. Holotype. GSM 102235, brachial valve (figured PI. 62, figs. 28-32). Slite Marl (Wenlockian). Canal ditch 800 m. south-west of Alby (north of Lergravsviken), parish of Rute, Gotland. Paratypes. GSM 102234, whole shell (figured PI. 62, figs. 23-27); GSM 102236, pedicle valve (figured PI. 62, fig. 33); GSM 102237, pedicle valve (figured PI. 62, fig. 34); and GSM 102238, pedicle valve (figured PI. 62, fig. 35). Slite Marl (Wenlockian). Cement plant Quarry at Slite, Gotland. Distribution. This species has only been seen in material collected by Boucot from the Slite Marl of Gotland. Additional material was kindly provided by Dr. Hede from the same beds but from a different locality. It is not present in collections available from other horizons in Gotland, nor is it present in collections from other areas. The age of the Slite Marl is given (Regnell and Hede, 1960, p. 49) as Wenlockian. According to Boucot (verbal communication) it is probably — but not certainly — Upper Wenlockian. The close similarity between I. slitensis of Wenlockian age and I. orbicularis of Ludlovian age suggests that the latter may have been derived from the former. It is also the reason for including a Gotland shell in this paper as comparisons are thereby made easier. Isorthis amplificata sp. nov Plate 61, figs. 1-30 1869 Orthis crassa Lindstrom; Davidson, p. 213, pi. 27, figs. 18 and 19 a (non fig. 17). Diagnosis. Unequally biconvex, faintly sulcate with circular to transversely elliptical outline Median ridge of pedicle valve low, flat, moderately wide for about one-third of the valve length then tapering gently ; separating diductor tracks which do not reach mid- length. Brachial valve adductor muscle field well impressed, bounded by circular raised mar- gin extending well into anterior half of valve; distinctly quadripartite. Median ridge 470 PALAEONTOLOGY, VOLUME 8 narrows between larger anterior pair of impressions, and posteriorly thickens between the brachiophores. Brachiophores distinctive, very thick, straight, erect, only slightly divergent; their posterior edges separated only by the width of the cardinal process. Sockets large and deep; teeth massive. Comparison. This species is fairly close to I. orbicularis but is distinguished by the following: (i) Unusually thick non-flaring brachiophores which almost meet at their bases and which have a posterior edge aligned normal to the commissure plane, whereas in I. orbi- cularis the posterior edge is only slightly oblique to this plane and the brachiophores are thinner and flare outwards as they diverge. (ii) More massive teeth with deeper crural fossettes and larger sockets. (iii) The wider median ridge and absence of distinct adductor scars in the pedicle valve. (iv) Pear-shaped outline of the muscle field in the brachial valve. There is an amplification, especially of teeth and brachiophores, to which the name refers. Description. Unequally biconvex, with pedicle valve twice as deep as brachial valve. Outline sub-circular to transversely elliptical with rounded cardinal angles. Anterior commissure crenulate and gently sulcate, lateral commissure straight. Hinge line slightly greater than half the greatest width which is slightly posterior to mid-length. Width slightly greater than length, thickness almost half the length. Exterior of pedicle valve. Evenly convex with well-developed beak which overhangs the pedicle interarea and is almost in contact with the beak of the brachial valve. The pedicle beak projects one-tenth of total length beyond the hinge line. Lateral margins of interarea sharp. Exterior of brachial valve. Slightly convex with a very shallow sulcus widening from the umbo to the anterior margin. Lateral margins of sulcus indefinite. Outline trans- versely elliptical. Cardinal angles rounded. Interarea orthocline. Notothyrium having margins which diverge only slightly, partly filled by myophore of the cardinal process which protrudes slightly. The pattern of branching of the costellae in the median area of the brachial valve isPI + PIs + PIt + PI 2nd and P II + P IIs, all of which have been produced at the 5 mm. stage. Interior of pedicle valve. A low, rounded median ridge extends from the delthyrial cavity to the centre of the valve where it ends in a gentle slope. The posterior extremity of the delthyrial cavity is flat. Adductor scars are not seen. Parallel, well impressed, narrow diductor tracks separated by the median ridge, extend to the anterior end of the ridge where they terminate with rounded margins. Low extensions of the dental lamellae bound the diductor tracks laterally. The thick vertical dental lamellae support massive teeth, whose anterior extremities are ridge-like, directed antero-laterally. The medial faces of the teeth bear deeply cut crural fossettes which are straight and inclined antero- ventrally. The posterior faces of the teeth bear accessory sockets directed antero- laterally. No pedicle callist seen. Interior of brachial valve. Well impressed, pear-shaped adductor muscle field defined by raised margins which posteriorly fuse with the anterior edges of the brachiophore bases and anteriorly decrease towards the median ridge. In one specimen, short, low V. G. WALMSLEY : ISORTHIS AND SALOPINA (BRACHIOPODA) 471 oblique ridges extend outwards from the central part of each margin in an antero-lateral direction (see PI. 61, fig. 18). Muscle field equals almost two-thirds shell length and over one-third greatest width. The median ridge is narrow and rounded between anterior pair of impressions but becomes wider and flatter between posterior pair. The anterior pair is larger and sub- elliptical, posterior pair trigonal and not always clearly separated. The anterior pair well impressed. When the posterior pair is also well impressed, the transverse ridge, normal to the median ridge, emphasizes the quadripartite character. The brachiophores are very thick, straight, erect, and proximally are separated pos- teriorly only by the width of the cardinal process. Distally they diverge slightly, their bases merging with the raised margin of the muscle field. In some specimens fulcral plates are visible, joining the postero-lateral surface of the brachiophore to the posterior edge of the valve and flooring the sockets which are partially excavated beneath the interarea. Sockets are large and widely divergent and sometimes crenulated. The cardinal process arises as a short shaft between the posterior edges of the brachiophores and expands to a bulbous non-lobed myophore which partially protrudes from the noto- thyrium. Crenulations of the anterior shell margin are low and evenly rounded with rounded interspaces. Holotype. GSM 102214, internal mould of brachial valve (figured PI. 61, fig. 16). Siltstone immediately above Wenlock Limestone (? latest Wenlockian/earliest Ludlovian). Old Quarry, at Cwm, Usk inlier, Britain (Nat. Grid Ref. 33320160). Paratypes. GSM 102207, internal mould of pedicle valve (figured PI. 61, fig. 3). Siltstone immediately above Wenlock Limestone (? latest Wenlockian/earliest Ludlovian). Bank in field, 600 yards north-west of Porth-llong, Usk inlier, Britain (Nat. Grid Ref. 34829825). GSM 102216, brachial valve (figured PI. 61, figs. 20-24) and GSM 102218, whole shell (figured PI. 61, figs. 26-30). Wenlock Limestone (Wenlockian). Old Quarry on west side of Lincoln Hill, Iron Bridge, Britain (Nat. Grid Ref. 670038). Distribution. This species occurs commonly in the four to six feet of buff and rust-coloured decalcified siltstones immediately overlying the Wenlock limestone in the Usk inlier. In these beds the shells have been dissolved and very clear internal and external moulds were collected, the majority being moulds of the brachial valve. Boucot collected shells, here considered to be conspecific, from the Wenlock lime- stone at Iron Bridge, Shrewsbury (Grid Ref. 670038 Old Quarry on west side of Lincoln Hill). So far, these are the only known occurrences of this species, the age of which is Wenlockian and possibly earliest Ludlovian. (See Walmsley, 1959, pp. 487 and 506-7 for discussion of the age of the beds at Usk). Isorthis clivosa sp. nov. Plate 61, figs. 31-38; Plate 62, figs. 1-22 Diagnosis. Unequally biconvex, nonsulcate to faintly sulcate with circular to transversely elliptical outline. Median ridge of pedicle valve wide, ending in broad slope and separating narrow deeply impressed diductor tracks. Well marked vascular grooves diverge from anterior ends of diductor tracts toward anterior margin. Brachial valve has distinctly quadripartite adductor muscle field bounded by circular raised margin not extending into anterior half of valve. Median ridge narrows between larger anterior pair of impressions. Dental lamellae thick with flattened anterior edges. Brachiophores, divergent thin blades. 472 PALAEONTOLOGY, VOLUME 8 Comparison. This species is distinct from all other isorthids described here because of the pedicle valve’s wide median ridge with its broad anterior slope, to which the name , refers. In this respect it resembles I. ar curia (Hall and Clarke) but differs from it in having a distinctly quadripartite muscle field in the brachial valve. In the combination of these two characters it resembles I. szajnochai Kozlowski, the type species of Isorthis j but may be easily distinguished by its less convex brachial valve, the narrowing of the brachial valve median ridge anteriorly, the much lower pedicle valve median ridge, the shorter more curved pedicle valve interarea and the thinner brachiophores. It is possible 1 that I. szajnochai was derived from I. clivosa, or a form close to it. I. neocrassa (Niki- forova) from the lower, (?) middle, Llandovery of the Siberian Platform appears to be very close to I. clivosa (see Nikiforova and Andreeva 1961, pp. 113-16 and pi. 18, figs. 1-21) but there are apparent differences in the muscle fields and cardinalia. I. neo- crassa is possibly ancestral to I. clivosa. Description. Unequally biconvex with pedicle valve having greater convexity. Outline sub-circular to transversely elliptical. Cardinal angles rounded. Anterior commissure crenulate and rectimarginate to faintly sulcate. Lateral commissure straight. Hinge line equal to half greatest width of shell which is at mid-length. Width equal to or slightly greater than length. EXPLANATION OF PLATE 64 Figs. 1-14. Isorthis orbicularis (J. de C. Sowerby). 1-2, Early Bringewoodian siltstone, roadside quarry at Porth-llong, Usk inlier, Britain (Nat. Grid Ref. 35209780), external mould of brachial valve (x2), GSM 102244, and latex impression (x2). Note sulcus and costellae. 3-5, Early Bringe- woodian, bank in yard at Coed-y-paen Farm, Usk inlier, Britain (Nat. Grid Ref. 33479825), internal mould of brachial valve ( X 3), GSM 102245, latex impression (x3) and enlarged posterior view ( x 5) of impression showing flaring brachiophores and small myophore. 6-7, Early Leintwardinian, south end of old quarry behind Darran Farm, Usk inlier, Britain (Nat. Grid Ref. 32759795), external mould of brachial valve ( x 2), GSM 102246, and latex impression ( x 2). 8-9, Early Leint- wardinian, roadside 1,000 yards west of Dinham Bridge, Ludlow area, Britain (Nat. Grid Ref. 49817441), internal mould of brachial valve (x2), GSM 102247, and latex impression (x2). 10, Early Bringewoodian, roadside quarry at Porth-llong, Usk inlier, Britain (Nat. Grid Ref. 35209780), internal mould of pedicle valve ( x 2), GSM 102248. Note transverse form and slightly curved dental lamellae. 11-14, Early Leintwardinian, south end of old quarry behind Darran Farm, Usk inlier, Britain (Nat. Grid Ref. 32759795). 1 1-12, Internal mould of brachial valve of large gerontic speci- men ( x 2), GSM 102249, and latex impression ( x 2). Note vascular trunks. 1 3-14, External mould of pedicle valve (x2), GSM 102250, and latex impression (x2). Figs. 15-27. Salopina lunata (J. de C. Sowerby). 15-19, Early Leintwardinian siltstone, south end of old quarry behind Darran Farm, Usk inlier, Britain (Nat. Grid Ref. 32759795). 15-16, Internal mould of brachial valve ( x 3), GSM 102251, and latex impression ( x 3). 17, Side view of specimen in fig. 18 (x3). 18-19, Internal mould of pedicle valve (x3), GSM 102252, and latex impression ( x 3). 20-23, Late WhitclifAan, 700 yards north of Mary Knoll House, Ludlow area, Britain (Nat. Grid Ref. 48217429). 20-21, External mould of brachial valve (x2), GSM 102253, and latex impression (x2). Note posterior costellae, subparallel to hinge line. 22-23, External mould of pedicle valve (x2), GSM 102254, and latex impression (x2). 24-25, Whitcliffian, trackside 220 yards NNW. of Upper Berthlwyd Farm, Usk inlier, Britain (Nat. Grid Ref. 37470600), internal mould of brachial valve (x 3), GSM 102255, and latex impression (x 3). Note wide median ridge and anterior crenulations. 26-27, Early Leintwardinian, south end of old quarry behind Darran Farm, Usk inlier, Britain (Nat. Grid Ref. 32759795), internal mould of pedicle valve ( x 3), GSM 102256, and latex impression (x 3). Note slightly raised median area and medially concave dental lamellae. Palaeontology, Vol. 8 PLATE 64 t’V* MM WALMSLEY, Isorthis and Salopina V. G. WALMSLEY: ISORTHIS AND SALOPINA (BRACHIOPODA) 473 Exterior of pedicle valve. Evenly convex with beak projecting about one-seventh shell length posterior to hinge line. Margins of delthyrium subtending about 50°. Lateral margins of interarea sharp. Exterior of brachial valve. Convex, with very faint wide median sulcus in some specimens. Outline transversely subelliptical, interarea one-third as long as in pedicle valve, anacline. Interior of pedicle valve. A wide, long median ridge arising in the posterior part of the delthyrial cavity, extends forward to the centre of the valve where it descends in a broad slope to the valve floor. In the posterior half of the delthyrial cavity, faint, narrow adductor tracks flank the median ridge which anteriorly separates narrow, deeply im- pressed, slightly divergent diductor tracks extending to the end of the median ridge. Beyond this point vascular grooves continue in the line of the diductor tracks, diverging and fading towards the anterior margin. Laterally the diductor tracks are defined by forward extensions of the vertical dental lamellae which are fairly thick and have flattened anterior edges. In some specimens they extend well beyond the end of the median ridge, lateral to the vascular grooves. Teeth moderately large, with blunt extremities, laterally directed accessory sockets and shallow crural fossettes in the com- missural plane. A small pedicle callist is present. Interior of brachial valve. A well-impressed brachial adductor muscle field confined to the central third of the posterior half of the valve, is bounded by sub-circular raised margins which posteriorly fuse with the bases of the brachiophores and anteriorly decrease towards the median ridge. The median ridge is rounded and narrow — equal to one-seventh of the width of muscle field — between the larger anterior pair of im- pressions and wider between the posterior pair. The anterior pair is more deeply impressed than the posterior impressions, resulting in a marked step across the muscle field normal to the median ridge, and a quadripartite aspect. The brachiophores are thin blades with vertical anterior edges, diverging at 60°. They bound the moderately large sockets which are raised on socket pads. The cardinal pro- cess arises with a very short shaft from the posteriorly thickened median ridge between the brachiophores. The myophore is bulbous and non-lobed. Holotype. GSM 102232, internal mould of pedicle valve (figured PI. 62, fig. 19). Late Leintwardinian. Roadside exposure 1,230 yards east-north-east of Mary Knoll House, Ludlow area (Nat. Grid Ref. 49227407). Paratypes. GSM 102228, internal mould of brachial valve (figured PL 62, fig. 11). Mid Eltonian. Trackside, 300 yards south of Hafod-Fawr, Llandovery area (Nat. Grid Ref. 8 1 443096). GSM 102229, internal mould of pedicle valve (figured PI. 62, fig. 13). Early Bringewoodian. Roadside on A. 40, 190 yards west-north-west of Ynys-y-Bont, Llandovery area (Nat. Grid. Ref. 83723225). GSM 102230, internal mould of pedicle valve (figured PI. 62, fig. 15). Mid Eltonian. Stream bank 600 yards south-east of Esgair llaethdu, Llandovery area (Nat. Grid Ref. 79022926). Distribution. I. clivosa associated with I. orbicularis occurs fairly commonly in material presented to the writer by Dr. J. Price from the Black Cock Beds (= middle and upper Eltonian) and the lower Cwm Clyd Beds (= lower Bringewoodian) of the Llandovery area in Wales. Material collected by Dr. Boucot from the Rhumney River section, Cardiff, S. Wales (Nat. Grid Ref. 200794) also contains both species. In the writer’s collections from the Welsh Borderland I. clivosa occurs extremely rarely in Upper Forest Beds (lower Bringewoodian) of Usk and the Lower and Upper Leintwardine Beds of the Ludlow anticline, in all cases associated with abundant I. orbicularis. In a small collection from the Lower Leintwardine Beds of Oreleton, presented by Dr. Lawson, two specimens of I. clivosa were found. Poorly located material from the Ludlovian of the Denbigh Moors (N. Wales) also contained two specimens of I. clivosa. In the British Ludlovian, therefore, the known range is from middle 474 PALAEONTOLOGY, VOLUME 8 Eltonian to upper Leintwardinian and this species seems to have been more common in the basin facies, rarely moving into the shelf area. The writer has identified this species, however, in collections made by Boucot from the upper Wenlockian of Golden Grove, near Llandeilo (Wales), and from the Canaston Beds (Upper Landovery) of Pembrokeshire (Wales). It also occurs in New Brunswick, Canada, in beds believed to be of Upper Llandovery age and also in beds of Ludlovian age. (Boucot, Johnson, Harper, and Walmsley — in press). Collections from various localities in Maine — determined by Boucot as Upper Llandovery (C4) — Lower Wenlock and Lower Ludlow in age, and a collection of Upper Llandovery (C3) — Ludlow age from the Merida Andes of Venezuela also contain 7. clivosa. An account of these occurrences is in preparation. The total known range to date, therefore, is from Upper Llandovery to Upper Leintwardinian. Isorthis scuteformis, sp. nov. Plate 65, figs. 13-27 Diagnosis. Unequally biconvex isorthids, non-sulcate to faintly sulcate, with shield- shaped outline and lacking a delthyrial platform in the pedicle valve. EXPLANATION OF PLATE 65 Figs. 1-12. Salopina lunata (J. de C. Sowerby). 1-3, Late Whitcliffian siltstone, 700 yards north of Mary Knoll House, Ludlow area, Britain (Nat. Grid Ref. 48217429). 2-3, Internal mould of brachial valve (x2), GSM 102257, and latex impression (X2). 1, is enlarged view of latter (x5) to show flexure where the muscle field margin joins the base of the brachiophore. 4-5, Late Whit- cliffian, trackside 100 yards north of Brook Farm, Usk inlier, Britain (Nat. Grid Ref. 32329768), internal mould of pedicle valve (x2), GSM 102258, and latex impression (x2). 6, Early Leint- wardinian, trackside 200 yards NNE. of Hilla Farm, Uskin lier, Britain (Nat. Grid Ref. 37770688), internal moulds of brachial and pedicle valves ( x 2), GSM 102259, showing common preservation. 7-8, Late Leintwardinian, small quarry 400 yards NW. of Walnut Tree Farm, Usk inlier, Britain (Nat. Grid Ref. 33059680), external mould of brachial valve ( x 3), GSM 102260, and latex impression (x3). 9-12, Late Whitcliffian 700 yards north of Mary Knoll House, Usk inlier, Britain (Nat. Grid Ref. 48217429). 9-10, Internal mould of brachial valve ( x 3), GSM 102261, and latex impres- sion ( x 3). 1 1-12, External mould of pedicle valve ( x 2), GSM 102262, and latex impression ( x 2). Figs. 13-22. Isorthis scuteformis scuteformis subsp. nov. Early Eltonian mudstone, railway cutting, eastern entrance to Ledbury Tunnel, Malvern area, Britain (Nat. Grid Ref. 386724). 13-17, Pedicle valve, brachial valve, posterior, anterior, and side views (x2), GSM 102263, Paratype. Note shield-shaped outline. 18-19, Interior of pedicle valve (x2), GSM 102264, Paratype, and oblique view of same specimen ( x 2) to show teeth. 20-22, Posterior, side, and interior views ( x 3), GSM 1 02265, Holotype. Note constant width of median ridge and continuous, fairly straight margins of adductor muscle field. Figs. 23-27. Isorthis scuteformis uskensis subsp. nov. 23-24, Eltonian siltstone, 150 yards SE of Pres- coed Farm, Usk inlier, Britain (Nat. Grid Ref. 34959962), internal mould of brachial valve ( x 2), GSM 102266, Holotype, and latex impression (x2). Note larger anterior pair of adductor im- pressions, disjunct curved margins of muscle field, and anteriorly narrowed median ridge. 25-27, Late Eltonian, trackside 1 ,000 yards north of Llangibby Castle, Usk inlier, Britain (Nat. Grid Ref. 36509825). 25, Internal mould of brachial valve (x3), GSM 102267, Paratype. 26-27, Internal mould of pedicle valve (x3), GSM 102268, Paratype, and latex impression (x3). Note shield- shaped outline, pedicle callist, and short median ridge, small pointed teeth and anterior crenulations. Figs. 28-29. Salopina lunata (J. de C. Sowerby), GSM 51555. 28, Internal mould of brachial valve (x 5), Lectotype. Same specimen shown top left (left of round spot) (x 1£) in order to define the position of lectotype on the slab GSM 51555. Note, specimens on this slab all somewhat distorted (see p. 461). Museum label on this specimen states ‘Oaker Quarry’. Figs. 30-31. Isorthis orbicularis (J. de C. Sowerby). Horizons and localities not known. 30, Internal mould of pedicle valve (x 2), Lectotype. GSC 6647 (see p. 467). 31, Brachial valve of whole shell (x2). GSC 6648. Palaeontology, Vol. 8 PLATE 65 WALMSLEY, Isorthis and Salopina V. G. WALMSLEY: ISORTHIS AND SALOPINA (BRACHIOPODA) 475 Comparison. I. scuteformis is distinguished from I. orbicularis by its shield-shaped outline and general absence of a sulcus in the brachial valve which is relatively more elongate. I. arcuaria (Hall and Clarke) also has a shield-shaped outline and a non-sulcate brachial valve in which the adductor muscle field is not distinctly quadripartite but the pedicle valves of these two species are quite distinct in that I. arcuaria has a wide median ridge raised on a delthyrial platform which ends in a steep anterior slope. Description. Unequally biconvex, with the pedicle valve having greater convexity. Distinctly shield-shaped outline with relatively straight lateral margins, only slightly rounded cardinal angles and antero-lateral margins which curve smoothly round to the anterior margin. Anterior commissure crenulate and rectimarginate, lateral commissures straight. Hinge line half as long as greatest width of shell, which is at mid-length. Thickness equal to half length. Exterior of pedicle valve. Convex, slightly arched along mid-line. Beak distinct, pro- jecting one-tenth total length beyond hinge line. Interarea with sharp lateral margins. Exterior of brachial valve. Gently convex, non-sulcate or faintly sulcate. Outline distinctly shield-shaped. Interarea, anacline. Notothyrium with only slightly divergent margins. Interior of pedicle valve. A relatively short, low, rounded median ridge extends from the delthyrial cavity to mid-length, where it fades gradually. It separates well-impressed sub-parallel elongate adductor tracks. Sub-parallel, low, forward extensions of the short dental lamellae bound the diductor tracks laterally and end opposite the anterior limit of the median ridge. Interior of brachial valve. The well-impressed adductor muscle field occupies the median third of the posterior half of the valve and is bounded by raised margins which posteriorly fuse with the bases of the brachiophores and anteriorly terminate opposite the end of the median ridge. The brachiophores are short, thin, flaring plates. Sockets are raised on socket pads. The cardinal process has a bulbous, non-lobed myophore which occupies the noto- thyrium. Isorthis scuteformis scuteformis, subsp. nov. Plate 65, figs. 13-22 This sub-species of I. scuteformis is characterized by the following: Pedicle valve: incurved beak, comparatively massive blunt teeth, which lack crural fossettes, and the absence of a pedicle callist. Brachial valve: adductor muscle field not distinctly quadri- partite, bounded by continuous, almost straight, lateral margins. The median ridge is uniformly wide. The brachiophores, at the hinge line, are separated only by the width of the cardinal process which has a very short shaft. The sockets are short. Holotype. GSM 102265, brachial valve (figured PI. 65, figs. 20-22). Early Eltonian. Railway cutting at eastern entrance to Ledbury Tunnel, Malvern area (Nat. Grid Ref. 386724). Paratypes. GSM 102263, whole shell (figured PI. 65, figs. 13-17) and GSM 102264, pedicle valve (figured PI. 65, figs. 18-19). Early Eltonian. Railway cutting at eastern entrance to Ledbury Tunnel, Malvern area (Nat. Grif Ref. 386724). Distribution. This sub-species is present in material collected by Boucot from the Ludlovian (Eltonian) of the Malvern area, Britain (railway cutting near eastern entrance of Ledbury railway tunnel. Grid 476 PALAEONTOLOGY, VOLUME 8 Ref. 386724). It has not yet been seen in equivalent beds of the Ludlow area or in the other Welsh Borderland Eltonian localities, nor from other Ludlovian areas. Isorthis scuteformis uskensis, subsp. nov. Plate 65, figs. 23-27 This sub-species differs from I. scuteformis scuteformis by the following: Pedicle valve : beak is not incurved and the small, very pointed teeth bear deep crural fossettes. A pedicle callist is present. Brachial valve: adductor muscle field is distinctly quadri- partite with the anterior pair of impressions considerably larger than the posterior pair. The margins of the muscle field are curved and disjunct. The median ridge narrows sharply where it passes between the anterior pair of impressions. The brachiophores are more widely separated from each other at the hinge line, the cardinal process has a long slender shaft and the sockets are deep. Holotype. GSM 102266, internal mould of brachial valve (figured PI. 65, fig. 23). Basal Lower Forest Beds (early Eltonian). 150 yards south-east of Prescoed Farm, Usk inlier (Nat. Grid Ref. 34959962). Paratypes. GSM 102267, internal mould of brachial valve (figured PI. 65, fig. 25) and GSM 102268, internal mould of pedicle valve (figured PI. 65, fig. 26). Lower Forest Beds (late Eltonian). Trackside, 1,000 yards north of Llangibby Castle, Usk inlier (Nat. Grid Ref. 36509825). Distribution. So far, this subspecies is known only from the Lower Forest Beds (Eltonian) of the Usk inlier. It is possible that it occurs elsewhere in the Eltonian but has been determined as ‘ Dalmanella ’ orbicularis. Acknowledgements. It is a great pleasure to record my gratitude to the Division of Geological Sciences of the California Institute of Technology for the award of a Research Fellowship in Palaeontology (1962/3) and to the Council of the University College, Swansea, for allowing me leave of absence for a year. This work was carried out during part of the tenure of this Fellowship. To the Fulbright Commission for a travel grant to the U.S.A., and to the Royal Society and the Geological Society for awards towards the cost of field work in Maine, New Brunswick, Quebec, and Nova Scotia, I am ex- tremely grateful. I owe Dr. A. J. Boucot a great deal for his vigorous and stimulating discussions and guidance both in the field in the northern Appalachians and in the laboratory, as well as for the generous and wholehearted way in which he made his magnificent collection available to me. Dr. J. E. Hede of Lund University very kindly supplied additional material from Gotland. Dr. J. Price, Dr. J. D. Lawson, Mr. J. H. McD. Whitaker, and Dr. E. V. Tucker kindly supplied material from their mapping areas in the Welsh Borderland. I must also record my indebtedness to Dr. G. A. Cooper of the National Museum, Washington, Professor H. B. Whittington of the Museum of Comparative Zoology, Harvard, Dr. K. Waage of the Peabody Museum, Dr. W. Struve of the Natur Museum, Senckenberg, Frankfurt, Dr. B. S. Norford, Geological Survey of Canada, and Dr. T. W. Amsden of the Oklahoma Geological Survey for the loan of material in their care, and to Dr. C. J. Stubblefield, Director of the Geological Survey of Great Britain, for the loan of Sowerby’s figured specimens, for very helpful discussions on the types of I. orbicularis and S. lunata, and for having drawn my attention to the Little Missenden Borehole material. All the specimens figured in this paper have now been deposited in the Geological Survey Museum, London, and bear GSM numbers. REFERENCES boucot, a. j. et al. 1960. A late Silurian Fauna from the Sutherland River Formation, Devon Island, Canadian Arctic Archipelago. Bull. geol. Surv. Can. 65. , Johnson, j., harper, c., and walmsley, v. G., (in press), Silurian Brachiopods and Gastropods of Southern New Brunswick. Bull. geol. Surv. Can. V. G. WALMSLEY : ISORTHIS AND SALOPINA (BRACHIOPODA) 477 dahmer, g. 1951. Die fauna der nach-ordovizischen Glieder der verse-Schichten mit Ausschluss der Trilobiten, Crinoiden U. Anthozoen, Palaeontographica, 101, A (1-4). Davidson, t. 1864-71. A Monograph of the British fossil Brachiopoda, vol. 3. Palaeontogr. Soc. [Monogr.]. elles, g. l. and slater, i. l. 1906. The Highest Silurian rocks of the Ludlow District. Quart. J. geol. Soc. Lond. 62, 195-222, pi. 22. hall, j. and clarke, j. m. 1892. An Introduction to the study of Brachiopoda, Part 1 ,45th Ann. Rep. New York State Mus., pp. 449-616. Holland, c. h. and lawson, j. d. 1963. Facies patterns in the Ludlovian of Wales and the Welsh Borderland. Lpool Manchr geol. J. 3. , lawson, j. d. and walmsley, v. g. 1959. A revised classification of the Ludlovian succession at Ludlow. Nature, Lond., 184, 1037-9. , , . 1963. The Silurian Rocks of the Ludlow District, Shropshire. Bull. Brit. Mus. (Nat. Hist.) Geology, 8, 93-171. imbrie, j. 1959. Brachiopods of the Traverse Group (Devonian) of Michigan. Pt. 1. Bull. Amer. Mus. nat. Hist. 116, New York. kozlowski, r. 1929. Les Brachiopodes Gothlandiens de la Podolie Polonaise. Palaeont. polon. 1. lawson, j. d. 1960. The Succession of shelly faunas in the British Ludlovian. Rept. Int. geol. Congr. 21st Sess., Norden. 7, 114-25. mclearn, f. h. 1924. Palaeontology of the Silurian rocks of Arisaig, Nova Scotia. Mem. geol. Surv. Canada, 137. murchison, r. i. 1839. The Silurian System. London. Nikiforova, o. and andreeva, o. n. 1961. Ordovician and Silurian stratigraphy of the Siberian Plat- form and its palaeontological basis (Brachiopoda). Trudy Geol. Inst. (Vsegei) 56. (In Russian). regnell, g. and hede, j. e. 1960. The Lower Palaeozoic of Scania — The Silurian of Gotland. Guide to Excursions A 22 and C 17. Int. Geol. Congr. 21st Sess., Norden. schuchert, c. and cooper, g. a. 1932. Brachiopod genera of the suborders Orthoidea and Penta- meroidea; Peabody Mus. Nat. Hist., Mem. 4. straw, s. h. and smith woodward, a. 1933. The fauna of the Palaeozoic Rocks of the Little Missen- den Boring. Summ. Prog. geol. Surv. G. B., 1932, Pt. II. walmsley, v. g. 1959. The geology of the Usk inlier (Monmouthshire). Quart. J. geol. Soc. Lond. 114, 483-521. williams, a., and wright, a. d. 1963. The Classification of the ‘ Orthis testudinaria Dalman’ Group of Brachiopods. J. Paleont. 37, 1-32. Manuscript received 10 January 1964 V. G. WALMSLEY, Department of Geology, University College of Swansea. A NEW FENESTRATE BRYOZOAN FROM THE LOWER CARBONIFEROUS OF COUNTY FERMANAGH by R. TAVENER-SMITH Abstract. Collections of bryozoa made recently in south Fermanagh include numbers of small but complete Fenestella- like colonies. These show considerable morphological variation, but careful examination indicates that they all belong to a single homogeneous population. The sample is described, and it is considered that the distinctive pattern of branching shown by the colonies and the presence in each of a strongly developed proximal spine places them in a new genus, for which the name Ptilofenestella is proposed. A collection of bryozoa has been made from thin beds of muddy limestone on the south side of Carrick Lough (text-fig. 1), County Fermanagh, Northern Ireland. The strata occur not far below the local top of the Carboniferous Limestone and are probably of Dx age, corresponding stratigraphically with the lower part of the Dartry Limestone (Oswald 1955) of the Leitrim-Sligo area. They are highly fossiliferous, and because the original fossil material has been to some extent replaced by silica, treatment of blocks with dilute hydrochloric acid has yielded a considerable fauna. This is characterized by the presence of large numbers of finely preserved bryo- zoan fragments of varying size and shape. Brachiopods, lamellibranchs, gastropods, trilobites, and sponges are also present but constitute only minor elements of the fauna. Representatives of the first three groups are usually of small size, serving still further to emphasize the predominantly bryozoan nature of the assemblage. Among the bryozoa, [Palaeontology, Vol. 8, Part 3, 1965, pp. 478-91, pi. 66.] R. TAVENER-SMITH: A NEW FENESTRATE BRYOZOAN 479 fenestrate cryptostomes are most numerous and a preliminary examination of these showed the common presence of a distinctively shaped form that is the subject of this paper. The colonies in question are quite small. Though varying considerably in size, they do not exceed a centimetre in length or 6 mm. in diameter. The zoarial shape is very characteristic, being hemispherical proximally but cylindrical distally, at which extremity it is open. The reticulate meshwork of branches and dissepiments forming the colonial skeleton thus presents the appearance of an elongate basket open at one end. The frame- work is strong, and this, coupled with its small size, probably accounts for the preserva- tion of a large number of specimens. Zoaria of this sort are associated in the collections with fenestellid fragments of many kinds, and it was at first thought that they represent the broken-off proximal extremities of larger funnel-shaped colonies of Fenestella. Closer examination showed that this is not so, because the ends of branches forming the zoarial baskets are usually not frac- tured, but are quite undamaged (PI. 66, fig. 11). The specimens are therefore complete colonies, but the possibility remains that they are immature forms that were killed off at an early stage in development and segregated by current action. This is unlikely for several reasons. First, fossil material from the locality concerned is in general poorly sorted and shows no sign of having been carried far. Delicate structures like long, spiny outgrowths are commonly preserved without damage, for example. Again, if the basket- shaped colonies are all young forms the associated fauna would be expected to yield fragments with a comparable meshwork and appearance representing broken parts of larger, adult expansions. None satisfying this requirement was found in the course of examining more than 5,000 fenestellid fragments from the same locality and horizon. Finally, the arrangement of branches in the distal part of some of the longer zoaria suggests that these are in a late stage of development. A branch thus situated may in some cases be seen to terminate abruptly while the pair on either side of it grow more closely together beyond, so that the normal dissepimental length is soon re-established. When two or three branches fail in this way at several points round the periphery of the cylinder the diameter is reduced so that it begins to close in distally (PI. 66, fig. 4). Branch failure of this kind, though by no means universal, is not uncommon among the larger specimens and similar occurrences in other stocks have been attributed by Elias and Condra (1957, p. 56) to a decline in the vigour of growth during the later, or gerontastic phase of colonial development. For these reasons it seems likely that the specimens are complete, fully developed zoaria which, by reason of their compact shape and structure, have been preserved without damage in considerable numbers. EXAMINATION OF SPECIMENS External features. Although all the colonies are basically alike in that they have a basket- like shape, they are quite variable in several respects, as even a cursory examination shows. The most obvious differences relate to the length of the basket (from 3 to 10 mm.), its diameter (1| to 6 mm.), and the number of branches composing it (13-34). There is also considerable variation in the texture of the meshwork of branches and dissepiments forming the basket. Some colonies, not always the smallest, have a fine mesh in which the usual fenestrule size is about 0-22 x 0-57 mm., while in others, not always the largest, 480 PALAEONTOLOGY, VOLUME 8 the mesh is much coarser and fenestrules may be up to 0-44x0-88 mm. Less noticeable differences affect the spacing of zooecial apertures and carinal nodes, the size of aper- tures, and width of branches. In view of these differences an examination was carried out to determine the limits of morphological variation in the assemblage, and to ascertain whether more than one taxonomic group was represented. It seemed reasonable to examine at the outset the grosser differences of zoarial size and structure, with the exception of the length of colonies which is a function of age and therefore of no immediate interest. The range of zoarial diameter was determined by measuring this feature at a standard distance of 3 mm. from the growth origin (to eliminate possible effects due to age differences) in a hundred colonies chosen to be fairly representative of the collection. The resultant data were used to construct the histogram of text-fig. 2a. Next the diameter of seventy-five colonies was plotted against the number of branches composing them (text-fig. 2b), all readings being taken at the same distance from the proximal end as before. The two graphs demonstrate the exist- ence of a continuous series in each case, and it is evident that differences in these re- spects noticed in the collection are those between end members of the series which, when viewed together, may differ considerably from one another. The second graph also shows that a fairly constant proportional relationship exists between the number of branches in a colony and its diameter. This explains why the meshwork of large and small colonies is often (though not always) of similar texture. In order to examine the nature and extent of structural variability in greater detail a sample of fifty-five of the best preserved colonies was chosen so as to present as far as possible an adequate cross-section of the collection. Each of these was then measured for twelve variables, between ten and twenty readings being taken in every case, depending on the size and state of preservation of the specimen. Nearly 9,000 measurements were made, all being taken from the mature, cylindrical parts of zoaria, the short proximal ends being avoided because of the difficulty of making accurate measurements there due to the strong curvature of the meshwork. The dimen- sions recorded were as follows : (a) Those regarded as standard in systematic work on fenestrate bryozoans, namely the number of branches in 10 mm. measured transversely to the branch axes ; the number of fenestrules in the same distance measured parallel to the branch axes ; the number of zooecial apertures in one row per 5 mm., and the number of carinal nodes, also in 5 mm. The procedure used in recording these data was that recommended by Condra and Elias (1944, pp. 54-57). ( b ) In some circumstances continuous variables are more useful than discontinuous ones, and measurements were therefore made of fenestrule length, fenestrule width (each measured between the mid-points of opposite sides), inter-apertural distance, and internodal distance (the distance between adjacent centres in both cases — see text-fig. 3). These were the continuous variables most nearly equivalent to the discontinuous ones mentioned above. (c) Branch width and apertural diameter. Measurements of the first were made at right angles to the branch axis midway along the length of a fenestrule, so as to obtain the normal width, unaffected by branch-dissepiment junctions or branch division. In these colonies there are no notable differences in branch width immediately before and after branch division, and so such measurements were not taken. In recording branch Dtamater of colony. Diamctar of colony. Numbar of branches in 10 mm. D Number of tenestrules in 10mm. E. 10 15 20 25 Number of apertures in 5mm. F 20. 15. 10 . 5. 0. 45 Number of nodes in 5mm. G. 015 0-25 0 35mm. Branch width. H. 2° 0-05 0-1 0-15mm. Apertural diameter. h-m-. Internodal space Fenestrule length. I. 30 40 50 eo 30-5 160 <0-53 c. 70-80 SU 12109 > 26 0 13-5 <0-52 64 SU 12110 > 23 0 110 <0-48 64 All the specimens are worn, so the measured values of Dc are less than the true values, and calcu- lated values of Dt/Dc are inflated accordingly. This is shown here and in subsequent tables by the use of the signs > and <. There are over sixty septa. The long major septa leave an axial space only 3 mm. across, into which project the counter and cardinal septa. These generally intertwine at the axis, and frequently one or both extend nearly to the opposite side of the axial space. The remaining major septa frequently are pinnate about the counter-cardinal plane. The minor septa, about half as long as the major, terminate at the margin of the tabularium. In the dissepimentarium both orders of septa are straight and fusiform, the major septa being more strongly dilated than the minor. Rare irregular carinae may occur out- side the zone of dilatation. The ends of the counter and cardinal septa are generally slightly dilated. The trabeculae are very slender, arranged in wide disphylloid fans. The zone of divergence of the fans is near the periphery; outside this zone, the trabeculae are moderately inclined outwards ; inside the zone they are increasingly axially inclined, until D. L. STRUSZ: DISPHY LLIDAE AND PHACELLOPHYLL1DAE 527 they are about 70-80° from the vertical at the margin of the tabularium. In the tabularium they curve upwards once more. The biseriate tabularium consists of a very wide axial zone in which the tabular floors are domed, surrounded by a narrow trench-like periaxial zone. The axial zone is com- posed of numerous small, fairly globose tabellae, while the periaxial zone is made up of a series of flat or sagging tabellae, which tends to be vertically discontinuous. Consequently, the outermost axial tabellae often interleave with the innermost dis- sepiments, the only difference then being in the direction of inclination. The dissepimentarium generally equals \ R, and is composed of 9-14 series of small globose dissepiments. These diverge quite markedly from a zone near the periphery, so dividing the dissepimentarium into three zones. The outer zone is of 2-3 series of dis- sepiments which are moderately to strongly inclined outwards; the middle zone is of 1 or 2 series of globose horizontal plates ; and the inner zone is of 6-9 series of strongly axially inclined plates. Comparison. M. bellense is close to the type species in size and number of septa (50-60 in M. annamiticum), and also in having a wide dissepimentarium. The major differences are in the arrangement of the tabellae, and in the strongly fusiform septa in M. bellense. Another point of difference is in the marked elongation of the counter and cardinal septa — a character lacking in all described species assigned to Mansuyphyllum, but which is found in a number of Garra disphyllid species. Known localities. Be- 10 (common), Cr-4. Mansuyphyllum parvu/um sp. nov. Plate 72, figs. 4—6; Plate 73, fig. 1 ; text-figs. 3, la, b Holotype. SU 14224 (PI. 72, fig. 6, PI. 73, fig. 1), loc. Ct-40. Other material figured: SU 14223 (loc. Ct-40), 18151 (loc. CAT/255). Derivation of name. Latin parvulus, very small. Diagnosis. Diminutive Mansuyphyllum with numerous fusiform septa; calix deep, with everted rim. Description. Solitary turbinate to trochoid corallites, whose maximum diameter is about 10 mm. The epitheca is moderately rugate, and bears shallow, irregular septal grooves. There are also irregular lateral talons. The calix is wide and deep, with a flat or slightly domed floor, and a rather wide everted rim; the major septa project above the floor and walls of the calix. Dimensions in mm. Specimen Loc. Dc Dt Dt/Dc n SU 12296 Cr-36 b > 60 40 <0-6 52 ,, 8-8 5-5? 0-6? > 32* SU 14223 Ct-40 5-2 3 0 0-58 38 SU 14224 ,, 8-8 c. 4-5 c. 0-5 48 ,, 101 > 38* SU 18152 CAT/255 7-5 4-5 0-6 52 Calical section. 528 PALAEONTOLOGY, VOLUME 8 Adult corallites have about 48-52 septa. Serial transverse peels of SU 18151 show that the septa are inserted quite rapidly: at Dc = 5-9 mm., n = 46; at 8 mm., n = 50, and at 9-5 mm. (the base of the calice) n = 50 still. The first section is from about half-way between the apece of the corallite and the base of the calix. The septa are strongly fusi- form, with the major the more strongly dilated; they attenuate rapidly within the tabularium. The major septa are long, more than § R ; one, probably the counter sep- tum, extends into the axial space (which is 1-5-2 mm. across), and may reach its opposite side. The other septa may be slightly pinnate about the counter-cardinal plane. The minor septa equal R, ending at the margin of the tabularium. Towards the periphery there are poorly developed irregular carinae, while in at least one transverse section there are short discontinuities in the septa about the margin of the tabularium. text-fig. 3. Mansuyphyllum parvulum sp. nov. Longitudinal sections, x5. a, Holotype SU 14224 (loc. Ct-40), traced from photograph of celluloid ‘peel’ (PI. 72, fig. 6). b, SU 18151 (loc. CAT/255), traced from photograph (PI. 72, fig. 5). The slender trabeculae are curved, arranged in disphylloid fans. Those at the periphery are directed only slightly outwards, while those at the margin of the tabularium are directed axially at about 30° from the vertical. The tabularium is irregularly biseriate ; the wide axial series consists of flatly domed complete and incomplete tabulae, while the narrow periaxial series consists of small, flat to inclined tabellae. The outer margins of the axial series are often supplemented by globose vesicular tabellae. There are three to four series of dissepiments (more in the extensions of dissepimental tissue into the larger talons). These are irregular in size, generally strongly globose. The innermost series is moderately to strongly axially inclined, and the others are horizontal to slightly peripherally inclined. The inner one or two series are generally strongly invested with fibrous septal tissue. Comparison. M. parvulum differs from all previously described species in its small size. It differs from the type species in having a slightly reflexed calical rim. Very similar in size and structure is M. catombalense sp. nov., described hereunder. For detailed comparison, see p. 530. Remarks. This species at first seems close to Kunthia Schluter 1885, with its deep calix and fusiform septa (see Stumm 1949, pi. 12, figs. 22-23). However, in that genus the calix reaches almost to the apex of the corallite, which is not the case in M. parvulum. As D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPH YLLIDEA 529 the type, K. crateriformis, is poorly known, further comparison must await its re- examination. Known localities. Cr-36 b, Ct-40 (type), P-13, CAT/255. Mansuyphyllum catombalense sp. nov. Plate 73, figs. 2 a-c\ text-figs. 4, Id Holotype. SU 14155 (PI. 73, figs. 2 a-c); paratype. SU 14156, both from loc. Ct-18. Derivation of name. From the parish of Catombal, in which occurs the type locality. Diagnosis. Small Mansuyphyllum with narrow dissepimentarium; minor septa often discontinuous; counter septum extends across narrow axial space. Description. This species is known only from thin sections, and so the external features are poorly known. Corallites are ceratoid, with marked growth irregularities, and talons are developed for lateral attachment. Adult corallites near the mode are 7 or 8 mm. in diameter; the maximum known is 10 mm. The calix has a narrow rounded rim, a steeply sloping wall, and a wide floor containing a broad, low, flat-topped axial boss. Dimensions in mm. Specimen Dc Dt Dt/Dc n Li Li Axial space SU 14155 7-3 3-4 0-47 48 0-9 R 0-6 R 0-8 9-6 40 0-42 44 0-8 R 0-5 R 1-0 7-5 3-9 0-52 48 0-9 R 0-4 R 1-0 SU 14156 70 3-4 0-49 40 0-9 R 0-6 R 0-9 There are 40-48 septa in adults. The major are long (08-0-9 R), leaving an axial space about 1 -0 mm. wide. The counter septum extends across this space, almost meeting the cardinal septum; it may be slightly dilated, or turned aside, but there is no definite columella. The counter-lateral septa are a little withdrawn, and the remaining septa are pinnately arranged about the counter-cardinal plane. The minor septa equal about \ R , and terminate just inside the tabularium. Septal dilatation is fusiform, and moderate to very strong. In a zone about 0-5 mm. wide around the tabularium the septa are fre- quently dilated so as to come into contact ; outside this zone they may, rarely, retain this strong dilatation to the periphery, but are normally only moderately dilated. In the dissepimentarium the septa may be straight and smooth, or zigzag, or irregularly carinate. In the tabularium the septa are usually straight, and only rarely bear poorly developed carinae. The minor septa are generally discontinuous in the outer part of the dissepimentarium, and may be so throughout their length, at times to the extent of occurring as a series of detached trabeculae. In only one section are there peripherally discontinuous major septa, and these are in a portion of the corallite immediately below a sudden growth constriction. The trabeculae are monacanthine and thick — up to 0-5 mm. in dilated portions of the septa ; there is only one radial series in each septum. At the periphery they are vertical or slightly axially inclined ; this inclination from the vertical increases steadily inwards, so that the individual trabeculae are curved. The inclination is about 60° from vertical at the margin of the tabularium. Within the tabularium the inclination appears to decrease, the trabeculae being generally directed sharply upwards. 530 PALAEONTOLOGY, VOLUME 8 The wide tabularium is biseriate. The axial series consists of wide, flat-topped domes. The periaxial series consists of peripherally inclined tabellae. The margins of the axial series are in places supplemented by small globose tabellae. A plot of Dt against Dc shows that this ratio remains constant during growth (text-fig. 4). The growth equation is: Dt = 0-5 Dc-\- 0-4. The narrow dissepimentarium, about J R, is made up of one to five series of small globose dissepiments; these are horizontal at the periphery, increasing to steeply inclined or vertical at the tabularium. text-fig. 4. Plots of n against Dc (above) and Dt against Dc (below) for Mansuyphyllum catombalense sp. nov. and subsp. nov ? Comparison. A very similar species is M. parvulum sp. nov. Both are small, with fusiform septa and an elongate counter septum. They have similar biseriate tabularia, and not very dissimilar dissepimentaria. The differences are: (1) Mansuyphyllum parvulum is turbinate to trochoid; Mansuyphyllum catombalense is ceratoid. (2) Mansuyphyllum parvulum has an axial space of about 1 -5-2-0 mm. In Mansuy- phyllum catombalense it is TO mm. or less. (3) The minor septa in Mansuyphyllum parvulum are very rarely discontinuous; in Mansuyphyllum catombalense they are generally so, as also at times are the major septa. (4) Unlike Mansuyphyllum parvulum, in Mansuyphyllum catombalense the septa may remain strongly dilated outwards from the zone of maximum dilatation. (5) In Mansuyphyllum parvulum, DtjDc is about 0-5-0-6, in Mansuyphyllum catom- balense about 0-4-0-5. D. L. STRUSZ: DISPHYLLIDAE AND PH ACELLOPHYLLIDEA 531 It is nevertheless clear that the two species are quite closely related. M. parvulum is known from calcarenites or silty calcarenites, while M. catombalense is known only from two localities in a crinoid-coral biostrome. It is therefore highly likely that the two species have become differentiated by adaptation to the two distinct environments. See also p. 540. Known localities. Ct-18 (type), Ct-28. Mansuyphyllum catombalense subsp. nov? Plate 72, figs. 8 a, b ; text-fig. If Material. Transverse and longitudinal sections, SU 13268 from loc. Cr-100. Description. The only known corallite is worn, but is apparently solitary and ceratoid. It is at least 1 5 mm. in diameter. Calix and epitheca are unknown. There are fifty-four septa, showing a marked fusiform dilatation at the margin of the dissepimentarium; the major are considerably more dilated than the minor. The pinnate arrangement of the septa is more marked than in M. catombalense s.s., and both the counter and cardinal septa extend into the oval axial space, almost meeting at its centre. The trabeculae do not appear to be arranged in a disphylloid fan system. They are rather wavy, and axially directed at about 60° from the vertical. The tabularium is not well known, but appears to be biseriate, with a periaxial series of globose tabellae supple- menting an axial series of flat to domed tabulae. The dissepimentarium is wide, com- posed of at least nine series of small, globose, strongly inclined dissepiments. Comparison. This specimen differs from M. catombalense s.s. essentially in its greater size and number of septa, and in having a relatively wider dissepimentarium with con- siderably more series of dissepiments. The fusiform dilatation of the septa around the tabularium is also more abrupt. Remarks. The limited material does not warrant naming, but it seems likely that this is a subspecies of M. catombalense which has differentiated in response to a change of environment. The latter species has been found only in a crinoidal biostrome, while SU 13268 is from a large coral biostrome near Wellington, which appears to have been a rather quieter environment, associated with deposits of pellet calcarenites and algal limestones. See also text-fig. 7. Mansuyphyllum sp. A Plate 72, figs. 9 a, b ; text-fig. 7c Material. SU 20101 (PI. 72, figs. 9a, b ), loc. Be-10; SU 12250, loc. Cr-4. Description. As the only material consists of thin sections, the precise external features are unknown. There appear to be some transverse growth irregularities, but septal grooves are only intermittently developed. The calix is unknown. Adult corallites are about 9 mm. in diameter. Dimensions in mm. Specimen Dc Dt Dt/Dc n U SU 20101 8-6 4-7 0-55 50 c. R -1 R 9y 5-4 2-9 0-54 36 SU 12250 9-6 4-8 0-50 50 9 532 PALAEONTOLOGY, VOLUME 8 The epitheca is thin, and is lined internally by lamellar tissue no more than 0-2 mm. thick. In adults, n = 40-50. The major septa are long, leaving an axial space 1 -Ox 0-6 mm. ; their ends are pinnately arranged about the counter-cardinal plane. The counter septum crosses the axial space to unite with the end of the cardinal septum. The counter- lateral septa are a little withdrawn, and in the smallest of the available transverse sec- tions the minor septa between them and the counter septum are nearly as long as the major septa. The other minor septa are about \ R, barely entering the tabularium. Septal dilatation is very strongly fusiform, and at the margin of the tabularium spreads over the innermost dissepimental surfaces in the very narrow interseptal loculi, to form a stereozone up to 1 mm. wide. In the dissepimentarium the thinner parts of the septa develop irregular carinae. The trabeculae are arranged in typical disphylloid half-fans. Vertical or slightly inwardly directed at the periphery, their inclination increases axially until within the tabularium they are inclined inwards at about 60° from the vertical. The tabularium is a little over \ Dc; only one fragmentary longitudinal section is available, and in this the tabulae are apparently incomplete, globose, and possibly arranged in two series, the axial ones being flatter than those near the dissepimentarium. The narrow dis- sepimentarium is composed of some four or five series of small, highly globose vesicles. These are horizontal at the periphery, becoming very steeply inclined at the margin of the tabularium. The dissepiments also decrease in size axially. Comparison. These two specimens appear to be about half-way between M. parvulum and M. catombalense. For a detailed comparison see table, p. 541. MansuyphyUum ? sp. B Plate 73, fig. 3 Material. SU 20102 (PI. 73, fig. 3), loc. Cr-106; SU 17117, loc. Gn-10. Description. The two known specimens are small, trochoid corallites; the epitheca is without septal grooves, but bears strong growth wrinkles. The calix is unknown. EXPLANATION OF PLATE 72 Figs. 1-3. MansuyphyUum bellense sp. nov. 1 a, b, Holotype SU 11295; transverse (a) and longitudinal (b) sections; celluloid ‘peels’, x2. 2, Topotype SU 12110; transverse section, showing elongate cardinal and counter septa, x 2. 3 a, b, Topotype SU 20099; transverse {a) and longitudinal (b) sec- tions, x 2 (see also text-fig. le). All from loc. Be-10. Figs. 4-6. MansuyphyUum parvulum sp. nov. 4 a, b. Topotype SU 14223, loc. Ct-40; longitudinal sec- tion (a) showing lateral talons, and transverse ( b ) section, x4 (see also text-fig. la). 5, SU 18151, loc. CAT/255 ; longitudinal section of fragmentary corallite, x 4 (see also text-fig. 3b). 6, Holotype SU 14224, loc. Ct-40; longitudinal section — celluloid ‘peel’ — <4 (see also text-fig. 3d). Fig. 7. Zelolasma gemmiforme (Etheridge fil.). SU 5278, loc. MM-10, figured Hill (1942c, pi. 6, fig. 6) ; transverse and longitudinal sections of corallites growing on the surface of a large solitary rugose coral (not shown) ; x 4 (see also text-fig. 5). Photograph courtesy Prof. D. Hill. Figs. 8a, b. MansuyphyUum catombalense subsp. nov? SU 13268, loc. Cr-100; transverse (a) and longitudinal ( b ) sections of fragmentary corallite, x 4 (see also text-fig. If). Figs. 9a, b. MansuyphyUum sp. A. SU 20101, loc. Be-10; transverse (a) and longitudinal ( b ) sections, ■ 4 (see also text-fig. 7c). Palaeontology, Vol. 8 PLATE 72 STRUSZ, Australian Devonian Rugosa D. L. STRUSZ: DISPHYLLIDAE AND PH ACELLOPHYLLID AE 533 Dimensions in mm. Specimen Dc Dt Dt/Dc n Ti U SU 20102 60 20 0-33 36 1 R I R SU 17117 7-5 3 0 0-40 38 i R t R The septa are strongly fusiform, the major little longer than the minor, and extending axially about §-§ R. The minor septa end at the margin of the tabularium. In one speci- men the septal dilatation is so strong that it forms a stereozone encompassing all but the outer 0-8 mm. of the dissepimentarium. In both specimens, there are numerous yardarm and zigzag carinae in the outer part of the dissepimentarium. The trabeculae are almost parallel, slightly axially directed, the inclination from the vertical being only a little greater at the tabularium than at the periphery. The tabularium is J-§ Dc across ; in the one available longitudinal section it is composed of irregular, distant, complete, flat to domed tabulae. In the same section, there are about four series of small, highly globose dissepiments, moderately inclined towards the axis. Comparison. Mansuyphyllum sp. B differs from M. parvulum in the different trabecular structure of its fewer septa, and in having a uniseriate tabularium. Remarks. This species is doubtfully included in Mansuyphyllum, because it does not have the biseriate tabularium characteristic of the type species. The carinate septa suggest Tipheophyllum, but that genus also has a biseriate tabularium, divided at the ends of the major septa. Further material is required before the precise affinities of this species are known, and until then it is best left un-named. Mansuyphyllum sp. C Plate 73, fig. 6 Material. SU 17117 (PI. 73, fig. 6), 17118, loc. Gn-10. Description. The external shape is unknown, both specimens being worn. The calix has a sharp rim, steep walls, and a wide flat floor. Dc is at least 3 cm. Dimensions in mm. Specimen Dc Dt Dt/Dc SU 17117 > 29 13-5 <0-47 SU 17118 > 15 6-5 <0-43 The major septa equal about § R; they are radial, and straight or slightly wavy. The minor septa, only about £ R, terminate within the dissepimentarium. The septa are fusi- form, moderately dilated, the major more than the minor; they are thin in the tabu- larium. The zone of maximum dilatation is about the middle of the dissepimentarium. The trabeculae form disphylloid fans, with the zone of divergence in the middle of the dissepimentarium. Peripherally they are slightly outwardly directed, and towards the tabularium they are increasingly axially directed, so that at the margin of the tabularium they are inclined at 60° from the vertical. The tabularium is wide — at least \ Dc. It is biseriate. The wide axial series consists of complete and incomplete tabulae, forming flat-topped domes. The narrower periaxial series consists of numerous horizontal and slightly outwardly inclined globose to sagging tabellae. Many of the tabulae are quite thickly coated with fibrous sclerenchyme. 534 PALAEONTOLOGY, VOLUME 8 The dissepimentarium contains nine or ten series of dissepiments. The outermost two to four series are small, globose, and vary from slightly outwardly inclined at the peri- phery, to horizontal in the zone of maximum septal dilatation, where they are often coated by extensions of the septal tissue. Axially from here, the dissepiments are increasingly inwardly inclined and elongate, until at the margin of the tabularium they are nearly vertical. Comparison. These specimens differ from the type species in their slightly greater size, in a wider tabularium, and in the more dilated septa. From M. bellense they differ in having a relatively wider tabularium. From other Australian species of the same size they differ in the form of the tabularium. Further material is required, but it may be possible that these specimens are closely related to M. bellense. Genus zelolasma Pedder 1964 1964 Zelolasma Pedder, p. 364. Type Species : Diphyphyllum gemmiformis Etheridge fil. 1902, pp. 253-5, pi. 37, fig. 1; pi. 39, figs. 1, 2; pi. 40, fig. 1. Diagnosis. Phaceloid to subcerioid disphyllid with frequent multiple peripheral budding. Septa subequal, generally thin, smooth peripherally, but wavy, slightly carinate, and sometimes slightly dilated, axially. Trabeculae in half-fans. Narrow dissepimentarium of a few series of small globose plates. Tabularium dominantly uniseriate, of gently sagging tabulae. Remarks. Of the other phaceloid disphyllids so far described, Disphyllum has a biseriate tabularium (in the type species — see Hill 1939a, p. 225), Cylindrophyllum has well- developed yardarm carinae, and Acinophyllum McLaren has unequal septa, peripherally carinate, and prominent connecting processes. The new genus Paradisphyllum, described below, has a considerably more complex structure. See also p. 535. Zelolasma gemmiforme (Etheridge fil. 1902) Plate 72, fig. 7 ; text-fig. 5 For complete synonymy, see Pedder 1964, p. 365. EXPLANATION OF PLATE 73 Figs. 1 a, b. Mansuyphyllum parvulum sp. nov. Holotype SU 14224, loc. Ct-40; (a) transverse section through calice (see also text-fig. 7b), ( b ) oblique longitudinal section; x 4. Figs. 2 a-c. Mansuyphyllum catombalense sp. nov. Holotype SU 14155, loc. Ct-18; (a, b ) transverse sections of two corallites, (c) longitudinal section; note septal discontinuity in b, c; x4 (see also text-fig. Id). 2 a, b, photographs by Mr. A. G. Smith, University of Queensland. Figs. 3 a, b. Mansuyphyllum ? sp. B. SU 20102, loc. Cr-106; transverse (a) and longitudinal (b) sections of corallite surrounded by a stromatoporoid; x 4. Figs. 4, 5. Disphyllum sp. cf. gregorii (Etheridge fil.). 4a, b, SU 16143, loc. E-16; transverse (a) and longitudinal ( b ) sections, x 2. 5, SU 16153, loc. E-21 ; transverse section through base of calix, x 2. Fig. 6. Mansuyphyllum sp. C. SU 17117, loc. Gn-10; oblique longitudinal section, X2. Fig. 7. Phillipsastrea oculoides Hill. Holotype SU 5281, loc. MM-10; figured Hill (1942c, pi. 6, fig. 9); transverse section, x 2. Photograph courtesy Prof. D. Hill. Palaeontology, Vol. 8 PLATE 73 STRUSZ, Australian Devonian Rugosa D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLIDAE 535 Holotype. AM F 5171: Cavan Bluff Limestone, Taemas bridge, north bank of Murrumbidgee R., parish Warroo, near Yass, N.S.W. See also Pedder 1964, p. 365. Material figured. SU 5278, loc. MM-10; figured Hill (1942c, pi. 6, fig. 6). Diagnosis. As for the genus. Remarks. Pedder has noted that the figures of maximum Dc and n given by both Etheridge Jr. and Hill are too low, and has given corrected values of Dc= 13-0 mm., and n = 50. The specimen figured herein and by Hill (1942c) from the Garra Formation agrees very closely with the type material. Pedder, in his synonymy, queries inclusion of this specimen in the species, but I do not consider this doubt to be justified. As no further Garra material is available, no additions to existing descriptions are possible. Known range. The species is so far known from text-fig. 5. Portion of a small colony of Zelolasma gemmiforme (Etheridge fil.) ,, ^ •• no, , A , ... n ■ adhering to the surface of another rugosan; he Couviman ? of the Murrumbidgee River area traced f‘om photograph (P1. 72, fig. 7). SU (Siegenian?, according to Pedder, p. 365), and from a locality (MM-10) in the Garra Formation probably towards the top, and most likely of early Couvinian or late Emsian age. 5278, figured Hill (1942c, pi. 6, fig. 6). X4. Genus disphyllum de Fromentel 1861 1861 Disphyllum de Fromentel, p. 302 ( fide Lang, Smith, and Thomas 1940, p. 53). 1893 Cannophyllum E. J. Chapman, p. 45 (obj. syn., Stumm 1949, p. 33). 1935 Disphyllum de Fromentel; Lang and Smith, p. 545 ( partim ). This paper contains a very full synonymy to 1935 ; however, the authors took a broad view of the genus, which has not been substantiated by recent workers. 1956 Disphyllum de Fromentel; Hill, p. F280, in Moore. (See for further synonymy, post-1935.) Type species : Cyathophyllum caespitosum Goldfuss 1826, p. 60 ( partim ) = Disphyllum goldfussi (Geinitz 1846); subsequent designation Lang and Smith 1934, p. 80. Diagnosis. Fasciculate Disphyllidae with smooth septa, whose maximum dilatation is peripheral; tabularium biseriate, the periaxial tabellae axially inclined; several series of globose dissepiments; trabeculae parallel or in half-fans. Discussion. As shown by Lang and Smith (1934, 1935), and also by Smith (1945, p. 21), the holotype of D. goldfussi has smooth septa which are moderately dilated in the dis- sepimentarium (particularly at the periphery) and thin in the tabularium. The tabularium is biseriate, and the dissepimentarium consists of several series of dissepiments. Other fasciculate disphyllids are Cylindrophyllum Simpson 1900, Acinophyllum McLaren 1959, Zelolasma Pedder 1964, and Paradisphyllum sp. nov. The first is characterized by strongly developed yardarm carinae in both dissepimentarium and tabularium, and by a fairly nar- row dissepimentarium. Acinophyllum was erected to include those American disphylloid species previously placed in Synaptophyllum Simpson 1900, which McLaren has shown to be a stauriid. The type species is characterized by a narrow dissepimentarium, n n B 6612 536 PALAEONTOLOGY, VOLUME 8 a uniseriate tabularium, septa which are weakly dilated peripherally and zigzag cari- nate in the dissepimentarium, and numerous connecting processes. Zelolasma differs from Acinophyllum principally in lacking the connecting processes, and in having sub- equal septa which are only carinate (and sometimes slightly dilated) axially. All three genera have trabeculae either subparallel or arranged in half-fans. Paradisphyllum has a more complex structure, with strong disphylloid trabecular fans, a wide dissepimen- tarium, and a biseriate tabularium. The simplicity of the tabularium in Acinophyllum is one of the diagnostic features stressed by McLaren. However, the taxonomic value of this feature is debatable. Thus Sudetia Rozkowska 1960, has as type a species in which both uniseriate and biseriate tabularia occur. Those specimens with a uniseriate tabularium resemble very strongly Acinophyllum. This raises the question of the validity of both genera. Sudetia was erected for a ‘descendant form’ of Peneckiella minor kunthi ( vide Rozkowska 1960, p. 33), in which the horseshoe dissepiments have become obsolete. However, the dissepimental types figured for Sudetia lateseptata, the type, are not even ‘ peneckielloid ’ (see p. 556). Moreover, I find it difficult to see how S. lateseptata can be a descendant of Peneckiella minor kunthi (Dames), as the former species is known only from the one locality, where it is intergrown with its supposed predecessor. It is likely that Sudetia Rozkowska 1960, is a junior synonym of Acinophyllum McLaren 1959. Disphyllum gregorii (Etheridge fil. 1 892) 1892 Campophyllum gregorii Etheridge fil. in Jack and Etheridge Jr., p. 60, pi. 3, figs. 15-18, Locality — Reid Gap, N. Qld. ; Givetian. 1895a Campophyllum gregorii Etheridge fil. p. 522, pi. 40, fig. 2. Locality — the same. 1942a Disphyllum gregorii (Etheridge); Hill, p. 247, pi. 8, figs. 1-4. Locality — the same, also Fanning R. and Burdekin Downs areas ; Givetian. non 1912 Campophyllum gregorii Etheridge; Chapman, p. 219, pi. 34, figs. 3-5, which is Brevi- phyllum recessum (Hill 1940). Lectotype. By subsequent designation Hill 1942a, p. 247. GSQ F 1655; figured Etheridge Jr., pi. 3, fig. 15, in Jack and Etheridge Jr. 1892. Type locality. Regan’s Limestone Quarry, probably portion 397v, parish Magenta ; Reid Gap, near Townsville, N. Queensland. Reid River Limestone, Givetian (Hill 1942a). Diagnosis. ‘ Disphyllum with ceratoid to cylindrical corallites with about 30 septa of each order; typically the major septa reach about half way to the axis, while the minor septa are less than half as long; there is typically one or two series of small, very globose dissepiments, and the septa are dilated so that they extend laterally over the upper sur- faces of the dissepiments; typically the tabulae are complete and horizontal, supple- mented at the margins by smaller plates; variability is great; the number of series of dissepiments may increase, the septa may become long, and sometimes curved about a small axial space, when the tabulae become incomplete on concave floors ; the dilatation of the septa varies in amount and position.’ (Hill 1942a, p. 248.) Remarks. Points of difference of D. gregorii from Breviphyllum recessum not brought out in the descriptions of Hill (1940, 1942a) are: firstly, in B. recessum the septa are thin, relatively short, and in transverse section exhibit discontinuities within the tabularium. D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLIDAE 537 In D. gregorii the septa are usually dilated at some point in their length, and are never discontinuous. The trabeculae in D. gregorii are thick monacanths, subparallel, and directed axially at a high angle from the vertical ; trabeculae are not visible in the holo- type of B. recessum. Finally, the epitheca of B. recessum is marked by deep, narrow septal grooves, separated by wide flat interseptal ‘ridges’. The septal grooves in D. gregorii , when developed, are very shallow and wide, essentially the intersection of two neighbour- ing slightly rounded to flat interseptal ridges. Disphyllum sp. cf. gregorii (Etheridge fil. 1892) Plate 73, figs. 4, 5 Material. Several fragments from Iocs. Cr-12, E-16, and E-21. Figured specimens. SU 16143 (loc. E-16), 16153 (loc. E-21). Description. The available fragments vary somewhat in structure, some sections being very like some figured sections of B. recessum (Hill 19406, e.g. pi. 9, fig. 7). However, the Garra specimens all have about sixty septa, with Dc = 13-16 mm., contrasting with n — 44 and Dc = 10 mm. for normal B. recessum. They also have the type of septal groove found in D. gregorii, and continuous septa. Dimensions in mm. Specimen Locality Dc Dt Dt/Dc n A l2 SU 16143 E-16 15 9-5 0-63 c. 70 c. i R c. \ R SU 16145 ,, > 11-5 9-5 <0-83 64 2 R iR SU 16153 E-21 c. 14 8 0 c. 0-57 66 c. R c. iR The specimens agree most closely with Hill’s description of specimens from the Reid Gap area, in that all have about five or six series of dissepiments, and highly variable septal length and dilatation. Generally the septa dilate at the margin of the tabularium, where the dilatation may spread over the dissepiments, and then they either remain at a constant moderate dilatation throughout the dissepimentarium, or show successive moderate wedge-wise dilatations over the surfaces of successive series of dissepiments. However, this may vary even in the one transverse section, with the septa on one side remaining attenuate throughout. Similarly the tabularium is variable; in some cases the tabulae are complete, flat with down-turned edges, supplemented by periaxial inclined tabellae; in others, the tabulae are incomplete and sagging, and the periaxial tabellae are both more numerous and less regular in size. Unfortunately none of the sections is sufficiently well preserved to show clearly the trabecular structure. There is a suggestion that the trabeculae are thin, and axially directed at a large angle from the vertical (as in D. gregorii ), but I have not seen the thick monacanths so characteristic of the Queensland species. Genus paradisphyllum gen. nov. Type species. Paradisphyllum harundinetum sp. nov. Diagnosis. Fasciculate disphyllids with septal trabeculae arranged in disphylloid fans. Septa fusiform, carinate outside the zone of greatest dilatation; the counter septum 538 PALAEONTOLOGY, VOLUME 8 may be elongate. Minor septa, and sometimes major septa, discontinuous near peri- phery. Tabularium as in Disphyllum. Remarks. Paradisphyllum differs from other fasciculate disphyllids essentially in the arrangement of the trabeculae in marked disphylloid fans (see p. 523), and so in having a calix with a strongly reflexed rim. The type species is further distinguished by having discontinuous septa. I would tentatively include in Paradisphyllum the Victorian species Disphyllum cognatum Philip 1962 (p. 177, pi. 24, figs. 5, 10, text-fig. 4b), from the Gedinnian Cooper’s Creek Formation. This species has a similar trabecular arrangement to that of P. harundinetum ; the only essential differences are that apparently the major septa are of uniform length, and that the septa are continuous. I consider that the trabecular structure is of far greater significance than the septal discontinuities. Another possible species is D. [ Synaptophyllum ] densum Smith 1945 (p. 22, pi. 12, figs. 3 a-c). As described and figured by Smith, this species has all the diagnostic features of Paradisphyllum. However, McLaren (1959, p. 30) has noted that horseshoe dis- sepiments are present in one longitudinal section of Smith’s type material. All the other longitudinal sections show trabecular fans, and so McLaren has tentatively assigned the species to Phacellophyllum. Further study is needed on this matter. Range. Paradisphyllum is definitely known from the Emsian or early Couvinian of the Garra Formation, and is probably also represented in the Gedinnian of Victoria. It may also be present in the Frasnian of Canada. Paradisphyllum harundinetum gen. et sp. nov. Plate 74, figs. 1, 3; text-figs. 6, 7 g, h Holotype. SU 13236 (PI. 74, fig. 1), loc. Cr-100. Paratype SU 20100 (PI. 74, fig. 3). Derivation of name. Latin harundinetum, a thicket of reeds ; refers to the appearance of the trabeculae in longitudinal section. Diagnosis. Dendroid to subcerioid Paradisphyllum about 7 mm. in diameter, with both orders of septa frequently discontinuous peripherally; with very irregular dissepiments, and crowded tabulae. Description. The corallum is small, generally dendroid, and usually in the form of a low dome. In crowded, subcerioid parts, individual corallites are bounded by slightly curved to irregular walls ; otherwise they may be separated by up to 1 cm., when they are ceratoid to cylindrical. The calix has a wide reflexed rim surrounding a shallow cup-shaped axial depression. Budding is lateral. Dc of adults = 5-11 mm., those near the mode being 6-7-5 mm. EXPLANATION OF PLATE 74 Figs. 1, 3. Paradisphyllum harundinetum gen. et sp. nov. 1, Holotype SU 13236, loc. Cr-100; x 4 (see also text-fig. 7/z); specimen collected by Dr. J. R. Conolly. 3, Paratype SU 20100, loc. Cr-100; longitudinal section, x 4 (see also text-fig. Ig). Figs. 2a, b. Hexagonaria approximans cribellum subsp. nov. Holotype SU 13259, loc. Cr-100; trans- verse (a) and longitudinal (b) sections, x 2 ; specimen collected by Dr. J. R. Conolly. Fig. 4. Billingsastraea aperta (Hill). SU 13261, loc. Cr-100; transverse section of corallum with rela- tively strong septal dilatation, x2. Palaeontology, Vol. 8 PLATE 74 STRUSZ, Australian Devonian Rugosa D. L. STRUSZ: DISPHYLL1DAE AND PH ACELLOPH YLL1D AE 539 Dimensions in mm. Specimen ( max. SU 13236 mean (min. No. of readings fmax. mean Dc Dt DtjDc n 11-2 4-2 0-78 44 6-6 2-9 0-46 36 1-2 0-9 0-33 6 33 33 33 33 8-0 3-6 0-51 34 6-3 2-8 045 U J jL'O ‘-tJ 4-4 1-9 0-39 32 6 6 6 2 (.min. No. of readings 32 The thin epitheca shows strong transverse growth irregularities, but no septal grooves ; it is lined by a thin (0-2 mm.) fibrous stereozone, from which project the septa. For adults, n is generally 38-42. The major septa are long, straight in the wide dissepi- mentarium, and straight or slightly curved in the tabularium. They are unequally with- drawn from the axis, leaving a space 1-1-5 mm. across, into which the counter septum, and sometimes the cardinal septum, projects. The septa often show a moderate pinnate arrangement about the counter-cardinal plane, in which case the counter-lateral septa may be slightly withdrawn; however, the septa may be completely radial. L2 is about \ to § R; the minor septa end just inside the tabularium. The septa are fusiform, and may be so strongly dilated as to form a stereozone up to 1 mm. wide near the inner margin of the dissepimentarium. However, the degree of dilatation is highly variable, even from one side of a corallite to the other, and in some juveniles it may be entirely absent. Towards the periphery the septa generally become attenuate, and bear irregular zigzag carinae. In this zone the minor septa are generally more or less discontinuous; less commonly the major septa also break up into dis- continuous fragments. The monacanthine trabeculae are arranged in broad disphylloid fans, closely resem- bling those which characterize the family Phacellophyllidae. The zone of divergence of the trabeculae corresponds exactly with the zone of maximum septal dilatation, at about the mid-radius, and about two-thirds of the way from the periphery to the tabularium. The biseriate tabularium is about Dc across, and consists of numerous closely crowded tabellae. The axial series is of globose plates, arranged in broad domes, and interfingering marginally with the periaxial tabellae, which are flat or gently sagging. The dissepiments are small, globose, in five to eight series. Their inclination varies from vertical at the inner margin of the dissepimentarium, through horizontal in the zone of maximum septal dilatation, to moderately outwardly inclined outside this zone. In this outer zone the dissepiments may become lonsdaleoid. Variation. As only one of the colonies so far found is relatively large and the total material is limited there are not enough data for a detailed study of intra-specific variation. No strong correlation between n and Dc could be found (text-fig. 6a), but a plot of Dt against Dc (text-fig. 6b) was clear enough to give a growth equation of Dt = 0-35 Dc+ 0-5. Comparison. P. harundinetum is apparently one of a group of closely similar disphyllid corals occurring in the Garra Formation, and is probably derived from one of these by acquisition of a colonial habit. The group is discussed below. Known localities. This species is only known from the biostrome at loc. Cr-100. 540 PALAEONTOLOGY, VOLUME 8 Discussion of Phytogeny. During examination of sections of a large number of Garra disphyllids, it became apparent that many of them had several fairly distinctive features in common, suggesting that they may be in fact closely related. To check this, their salient features were compared in a chart (p. 531), and plots of available data for n\Dc and Dt/Dc were made. These show that the various species within this group of Garra disphyllids are indeed probably closely related, and a study of their stratigraphic distribution, in so far as it is known, further suggests a phylogenetic plexus, of which the progenitor is probably the small solitary coral M. parvulum (known from beds closer to the base of the formation than those containing the other species). 10 (mm) Dt (mm) 5rt x SU 13235 ® SU 13236 ASU 20100 10 (mm) text-fig. 6. Plots of (a) n against Dc and ( b ) Dt against Dc, for three specimens of Paradisphyllum harundinetum gen. et sp. nov. The features common to the species of this postulated plexus, and which distinguish them from other Garra disphyllids, are : (1) Strongly fusiform septa, peripherally weakly to strongly carinate. (2) Trabeculae arranged in half-fans or full disphylloid fans. (3) Long major septa, frequently pinnate about the counter-cardinal plane. (4) Counter septum elongate, generally reaching the axis, and in some extending completely across the axial space. The differences between the various species are summarized in the table, p. 541. Note that for M. catombalense and P. harundinetum, the data for Dt/Dc, while rather limited, are sufficiently closely correlated to allow the determination of approximate growth equations (see also text-figs. 4, 6). The inferred relationships of the species are shown in text-fig. 7. The most nearly certain is the lineage M. parvulum — M. sp. A — M. catombalense, while the least certain is M. bellense. The latter shares with the others most of the features distinctive of the group, but is considerably larger. As intermediate specimens have not been found, the relationship of M. bellense to the remaining species is as yet uncertain. D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPH YLLID AE 541 Table comparing importance features of Paradisphyllum harundinetum gen. et sp. nov., and those species of MansuyphyUum which are con- sidered to be phylogenetically related. . GENUS c i e s bellense Me parvulum n suyphyl/u, sp. A m catombatense catombalense subsD. nov? Paradisphyllum harundinetum Corallum : trochoid ? turbinate-trochoid ceratoid caratoid ? dendroid-subcerioid | known range : Dc < N? of readings: ( normal adults : 20+ — 30-5+ mm/ 5-2 — 10* 1 mm / 5-4- 9 6mmy^ 1-6 - 8-7 mm // 15+ mm 1*2 - 11-2 mm 25 - 30 mm \ 9 -10 mm 9 mm \ 7- 8 mm ? \ 6 -7-5 mm ^ Growth curve, Dt = ? ? ? 0-5 Dc ♦ 0-4 ? 0-35 Dc ♦ 0-5 Diameter, axial space : c. 3 mm c. 2 mm 10 x 0-6 mm c. 1mm 1-7 x 1-2 mm 10 - 1-5 mm SEPTA: f known range : n j N! of readings: ( normal adults : C.50-C.76 // 32+ - 52 36-50 28-48 6-44 c. 50 CS0? 0 42 ^ 5/ x c 40 'X Li > 0-9 R c. 0-7 R « to. 9 R » 0-8 R 0-8 - 0-9 R l2 > 0-5 R 0-3- 0-5 R < c. 0-5 R > >0-3R 0-5- 0-7 R f major : continuous ? may be peripherally continuity j _ carinae in diss’ahum : common common? numerous Septal pattern: pinnate? frequently sometimes, slightly ( ■ ■■ alw ays > strongly frequently, moderately K < ' elongate ■ » variable length KL ( i equal to metasepta ) ( slightly withdrawn ) withdrawn may be slightly withdrawn c elongate t qual to metasepta .. .. .A elongate sometimes elongate Series of dissepiments: 9-14 3-4 — i 1 - 5 > 9 5 -8 Width of diss-arium: c. 0-5 R / m r. N. Table comparing importance features of Paradisphyllum harundinetum gen. et sp. nov., and those species of Mansuyphyllum which are con- sidered to be phylogenetically related. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLIDAE 542 PALAEONTOLOGY, VOLUME 8 CAT . 255 text-fig. 7. Inferred phylogenetic relationships between species of Mansuyphyllum , and Paradisphyl- lum harundinetum ; all x 1 -5, traced from photographs (see PI. 72-74). Localities from which the various species are known are shown in their probable relative stratigraphic positions; where known with reasonable accuracy, these are in heavy lettering, a. Transverse and longitudinal sections, SU 14223 (loc. Ct-40). b. Transverse section through calix, holotype SU 14224 (loc. Ct— 40). c, Transverse and longitudinal sections, SU 20101. d, Transverse and longitudinal sections, holotype SU 14155. e. Transverse and longitudinal sections, paratype SU 20099. /, Transverse and longitudinal sections, SU 13268. g. Longitudinal section, paratype SU 20100. h, Transverse section, holotype SU 13236. D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPH YLLID AE 543 Genus hexagonaria Giirich 1896 1896 Hexagonaria Giirich, p. 171 ( fide Lang, Smith, and Thomas 1940, p. 69). 1949 Hexagonaria Giirich, Stumm, p. 33. See for list of synonyms. 1954 Hexagonaria, Moenke, p. 452. See for a full discussion of the genus (taxonomy and morphology). Type species. By subsequent designation Lang, Smith, and Thomas 1940, p. 69, Cyathophyllum hexagonum Goldfuss 1826, p. 61, pi. 19, figs. 5 e,f pi. 20, figs, la, 6; non pi. 19, figs. 5 a-d. Diagnosis. ‘Cerioid corals with individual corallites separated by polygonal walls. Calices usually with an axial pit and a peripheral platform. Septa radially arranged, of two orders, major extend into tabularium while minor are confined to dissepimentarium. They are lightly or heavily carinate, rarely dilated. No modification of protosepta is visible. Dissepimentarium is wide and composed of many rows of horizontal or inclined dissepiments. Tabularium is relatively narrow and composed of closely set, complete or incomplete tabulae, that are horizontally disposed.’ (Stumm 1949, p. 33.) Discussion. A list of synonyms may be found in Stumm (1949). It should be noted also that in many cases Soshkina has included in her genus Megaphyllum species which are Hexagonaria, as well as some, including the type, which are considered by Stumm (1949) and others to be Disphyllum. In her discussion of Megaphyllum (1954, p. 37), Soshkina referred critically to Stumm’s illustration of H. hexagona (1948, pi. 6, figs. 1, 2), stating that fig. 1 is of the holotype, and fig. 2 of a different species. Yet according to Stumm, both are sections of the same specimen, a hypotype. Still referring to this, Soshkina also stated : ‘ . . . [Stumm] cannot distinguish the species of the genus Phillipsastraea from the massive colonies belonging to the group of species “ Cyathophyllum hexagonum Goldf.”, in other words belonging to genus Megaphyllum Soshk., . . . ’ This, with other statements on the same page, apparently indicates that Soshkina considered the type species of Hexagonaria to be a species of her genus Megaphyllum. At the same time, she placed Prismatophyllum Simpson 1900, in synonymy with Megaphyllum. In both cases, there is clear contravention of the rules of nomenclatural priority (Stoll et al. 1961 : Art. 23). This apparent confusion of type species and of priority has unfortunately caused con- siderable divergence of usage in the work of authors from Europe, America, and Australia on the one hand, and from Russia on the other. Megaphyllum bulvankerae Soshkina 1954 (p. 38) and M. columellare Soshkina 1954 (p. 40) are species of Hexagonaria. Hexagonaria approximans (Chapman 1914) 1914 Cyathophyllum approximans Chapman, p. 304, pi. 47, figs. 5, 6. East Gippsland, Victoria : Early Devonian. 1939a Prismatophyllum approximans (Chapman); Hill, p. 234. 19546 Hexagonaria approximans (Chapman); Hill, p. 107, pi. 6, figs. 3a, b. 1954 b Hexagonaria aff. approximans (Chapman); Hill, p. 108, pi. 6, figs. A a, b. Waratah Bay, Victoria: Bell Point Limestone, ‘possibly Couvinian’. 1962 Hexagonaria approximans (Chapman); Philip, p. 177, pi. 24, figs. 4, 8, 9. Tyers R., Gippsland, Victoria: Cooper’s Creek Formation, early Gedinnian. Holotype, NM 1247 : specimen figured Chapman 1914, pi. 47, figs. 5, 6. Quoted and figured as holotype by Hill 19546, p. 107, pi. 6, figs. 3a, 6. Type locality given by Chapman as ‘Cooper’s Creek, behind 544 PALAEONTOLOGY, VOLUME 8 Chinaman’s Garden’; probably from the early Gedinnian Cooper’s Creek Formation of Philip 1962, p. 127. Chapman also listed specimen no. 746, but this is not quoted by either Hill or Philip. Diagnosis. Large Hexagonaria with thirty-two to forty-two long, slightly fusiform septa, with carinae well developed near periphery; trabeculae arranged in half-fans or in broad disphylloid fans ; tabularium narrow, biseriate. Remarks. This species has been well described by Philip (1962) who, from the type material and a large collection from the Tyers River area, was able also to assess varia- tion, which he found to be considerable. One point to note is that according to Philip the trabeculae are directed inwards at all times, whereas according to Hill (19456, p. 107) . . there may be an area of divergence . . . near the inner margin of the dissepi- mentarium’. This is not clear in her figures of the holotype, but is quite definite in her H. aff. approximans (pi. vi, fig. 46), which Philip placed in synonymy with H. approxi- mans. Known range. The holotype comes from Cooper’s Creek, probably from the same horizon as the Cooper’s Creek Formation of Philip (1962), some miles to the south. Philip considered this horizon to be probably Gedinnian in age. The species is also known from the Bell Point Limestone of Waratah Bay, Victoria; Hill (19546) considered this horizon to be \ . . possibly Couvinian’. Philip (19606) placed it between the Cooper’s Creek-Loyola faunas and the Buchan fauna, probably Siegenian. Hexagonaria approximans cribellum subsp. nov. Plate 74, figs. 2 a, b\ Plate 75, figs. 1 a, b; text-figs. 8, 9 Holotype, SU 13259 (PI. 74, figs. 2a, b), loc. Cr-100. Other material figured: SU 13260 (loc. Cr-100). Derivation of name. Latin cribellum, a small sieve ; in reference to the complex carinae. Diagnosis. H. approximans with numerous strongly fusiform septa, and tabulae with upturned margins. Description. All known coralla are small, and apparently in the shape of low domes. The calix has a narrow rim, either flat or slightly everted, and rather steep sides sur- rounding a relatively shallow axial pit whose floor is gently domed. Dimensions in mm. (representative corallites). Specimen Dc Dt Dt/Dc n Li U SU 13259 4-9 1-2 0-24 40 0-8 R 0-5-0-7 R 12-7 4-4 0-35 52 0-7 R 0-6 R 13-3 50 0-38 56 0-8 R 0-5 R 22-1 60 0-27 54 0-8 R 0-7 R SU 13260 >30 5-5 <018 64 c. 0-9 R c. 0-8 R In adult corallites, Dc is about 2 cm. Each corallite is bounded by a very thin epitheca, lined by a fibrous stereozone 0-2-0-5 mm. thick. For adults, n = 54-64. The major septa are long, leaving an axial space of 3-4 mm. ; the minor septa are little shorter, and end at the margin of the tabularium. Attenuate in the tabularium, the septa are strongly fusiform in the dissepimentarium, with the zone of maximum dilatation being R in from the periphery. Peripherally the septa are generally thin, but bear prominent, D. L. STRUSZ: DISPHYLLIDAE AND PH ACELLOPHYLLID AE 545 flanged zigzag carinae, and occasionally break up into naotic segments consisting of these complex carinae without the intervening lamellar portions of the septa. The thin trabeculae are arranged in strongly asymmetrical disphylloid fans, the zone of divergence approximately corresponding to the zone of maximum septal dilatation. At the periphery the trabeculae are vertical, or more generally are inclined outwards GARRA Fm. 70° n ( • SU 132 59 ( ♦ SU 13260 10 VICTORIA 20 A Waratah Bay o Tyers Quarry 30 DC >4Q (mm) H. approximans cribet/um 40 • AO . ® a ° "* H. approximans o o A text-fig. 8. Plots of n (above) and Dt/Dc (below) against Dc for Hexagonaria approximans (Chapman) and H. approximans cribellum subsp. nov. Data for H. approximans s.s. obtained from published figures: Hill (19546 — Waratah Bay) and Philip (1962 — Tyers Quarry). at a small angle from the vertical. Inward from the zone of divergence they are in- creasingly curved, and at the margin of the tabularium they are generally inclined at a very high angle from the vertical, and may be horizontal. Within the tabularium their inclination becomes increasingly steep axially. A plot of Dt against Dc shows a fairly strong correlation (text-fig. 8) ; for the holotype, the corresponding growth equation is approximately: Dt = 0-33 Dc. The tabular floors are flat or gently domed, with sharply upturned margins. The tabulae are crowded, complete or more usually incomplete, and slightly sagging to slightly domed. Those which abut against the dissepimentarium are moderately to sharply upturned marginally. Rarely, the tabulae are marginally supplemented by slightly globose tabellae, inclined towards the axis. The wide dissepimentarium consists of up to eighteen series of small, highly globose 546 PALAEONTOLOGY, VOLUME 8 dissepiments, which are horizontal or outwardly inclined peripherally, horizontal until about the fourth series, then steeply and evenly inclined towards the tabularium until the last two to four series, which are vertical and elongate. Towards the periphery, the dissepiments in each interseptal loculus are frequently imperfectly arranged in three series: one series of transverse, cylindroidal plates, and two lateral series of convex, text-fig. 9. Dissepimental structure, Hexagonaria approximans cribellum subsp. nov. a. Transverse section; b, tangential longitudinal section; SU 13260 (loc. Cr-100), x5 (cf. PI. 75, figs. 1 a, b). c. Schematic representation, xlO; 1 — tangential longitudinal section; 2 — radial longitudinal section (part of septum removed to show dissepiments); 3 — transverse section, ep — epitheca; st — peripheral stereozone : tr — transverse (cylindroidal) dissepiments ; / — lateral dissepiments. incomplete plates, resting partly on the sides of the septa, and partly on the cylindroidal plates or on each other (text-fig. 9). Comparison. This subspecies differs from the original species principally in having con- siderably more septa — average about sixty as against about thirty-eight. This also shows up in plots of n against Dc for the two forms (text-fig. 8). In addition, the septa are in general more strongly dilated, and the trabeculae are arranged in more strongly de- veloped fans. There seem to be more series of dissepiments, and the tabular floors tend to be sagging rather than domed. In other respects, including the high degree of vari- ability (particularly in size), the new subspecies is quite close to H. approximans s.s. Remarks. In view of the definite differences from the Victorian species, and in the interests D. L. STRUSZ: DISPHYLLIDAE AND PH ACELLOPH YLLID AE 547 of stratigraphic precision, I feel it is desirable to separate the Garra form at the sub- species level. H. approximans cribeJlum is, however, quite probably a descendant of the Victorian species. Known localities. Loc. Cr-100. Genus billings astrae a Grabau 1917 1917 Billingsastraea Grabau, p. 957 ( fide Lang, Smith, and Thomas 1940, p. 27). 1958 Billingsastraea Grabau; Schouppe, p. 253. Contains an exhaustive synonymy. Type species. Phillipsastrea vemeuili Milne-Edwards and Haime 1851, p. 447, pi. 10, fig. 5 ( fide Lang, Smith, and Thomas 1940, p. 27). Diagnosis. Massive Disphyllidae with walls between corallites absent, or represented by fibrous sclerenchyme, but no epitheca. Septa may be confluent, abutting, or peripherally discontinuous. Discussion. In an extensive study of the ‘ Phillipsastrea ’ group of Rugosa, Schouppe (1958) has shown that Phillipsastrea sensu stricto possesses a zone of horseshoe dissepi- ments, and so is a senior synonym oiP achy phy Hum Milne-Edwards and Haime. Schouppe placed the species without horseshoe dissepiments in Billingsastraea. Until certain nomen- clatural problems raised by this revision are resolved by the International Commission (see p. 524), I have accepted this re-assignment. The following Australian species lack series of horseshoe dissepiments, and should, following Schouppe’s revision, be placed in Billingsastraea : Phillip sastraea aperta Hill 19426; P. callosa Hill 19426; P. carinata Hill 1942; P. delicatula Hill 1936; P. linearis Hill 19426 (P. walli Etheridge of Chapman 1914, non Etheridge Jr. 1892); P. maculosa Hill 19426; P. speciosa Chapman 1914. P. cur rani Etheridge fil. 1892, poses a problem which can only be solved by further study. Some specimens from the type locality (Limekilns, north of Bathurst, N.S.W.) lack horseshoes, while others have perfectly developed series of horseshoes. It is possible that two species are present. P. oculoidesYiiW 1942c, possesses definite small horseshoe dissepiments, and so remains in Phillipsastrea d’Orbigny 1849, sensu Schouppe 1958. Phillipsastrea sp. Hill 1954a, p. 14, pi. 3, fig. 2, is probably a Billingsastraea, but no longitudinal section is available. Billingsastraea aperta (Hill 1942) Plate 74, fig. 4; Plate 75, figs. 2, 3 19426 Phillipsastraea aperta Hill, p. 154 (non? pi. 2, figs, la, b). 1942c Phillipsastraea aperta Hill; Hill, p. 183, pi. 6, figs, la, b. Holotype : SU 7289 (PI. 75, figs. 3 a, b), loc. Cr-1 13. Other material figured : SU 13261, 20104, both loc. Cr-100. Diagnosis. Astraeoid Billingsastraea with widely spaced and only slightly dilated septa, with numerous small globose dissepiments, and with elongate tabellae arranged on slightly domed tabular floors. (After Hill 19426, p. 154.) Description. The description given by Hill (19426) is very brief, and so the species is redescribed below. 548 PALAEONTOLOGY, VOLUME 8 The corallum is partly astraeoid, partly thamnastraeoid. The calicular pits are rather deep, with steep walls, gently domed floors, and rounded everted rims. The pits are about 6 mm. in diameter, while Ts (see p. 522) is about 13 mm. Dt n n/Dt Ts Mean 5-2 38* 7-3 13-7 Max. 60 46 7-9 17-9 Min. 3-8 30 5-3 101 No. of readings 9 8 8 24 * Mean n to the nearest even number. Data from 2 coralla, loc. Cr-100. The major septa are long, leaving an axial space 1-2 mm. across, while the minor septa end just inside the tabularium. The axial space is elongate in the counter-cardinal plane; the cardinal septum projects a short distance into the space, but the counter septum extends right across it, its end either uniting with the cardinal septum, or turned abruptly aside. The remaining septa show a slight pinnate arrangement. The septa are fusiform; in a zone 2 -5-3 -5 mm. wide at the inner margin of the dissepimentarium they are usually as wide as the interseptal loculi. Away from this zone the dilatation decreases rapidly, the septa being attenuate peripherally and axially. The slender trabeculae are arranged in well-developed asymmetrical disphylloid fans, whose zone of divergence corresponds exactly with the zone of maximum septal dilatation. Dt = 3-5 mm. ; the tabularia are composed of numerous flat or gently domed, com- plete and incomplete tabulae, arranged biserially. In the broad axial zone the tabulae are supplemented by moderately convex tabellae. The dissepimentarium is composed of numerous small globose dissepiments, horizon- tal in the zone of maximum septal dilatation, steeply axially inclined inwards from this zone, and moderately peripherally inclined outside the zone. Midway between tabularia the dissepiments are again horizontal. Comparison. B. aperta very closely resembles Pseudoacervularia roemeri (Verneuil and Haime 1850) of Rozkowska (1953), from the Frasnian of the Holy Cross Mts., Poland. This resemblance is most noticeable in the arrangement of the dissepiments, the trabe- cular fans, and the degree of variation of septal dilatation. Remarks. SU 6199, from the Loomberah Limestone of Tamworth, N.S.W., placed in B. aperta by Hill (1942c, p. 154, pi. 2, figs, la, b ), differs from the holotype, and from other specimens from the type horizon, in that its septa are slightly carinate in the EXPLANATION OF PLATE 75 Figs. 1 a, b. Hexagonaria approximans cribellum subsp. nov. Topotype SU 13260, loc. Cr-100; trans- verse (o) and longitudinal (b) sections, x 2; note large diameter. (See also text-fig. 9 a, b.) Figs. 2, 3. Billingsastraea aperta (Hill). 2, SU 20104, loc. Cr-100; longitudinal section, X2. 3a, b, Holotype SU 7287, loc. Cr-1 13, figured Hill (1942c, pi. 6, figs. 7 a, b); longitudinal (a) and transverse (b) sections, x 2, photographs by courtesy of Prof. D. Hill. Fig. 4. Peneckiella mesa { Hill). Topotype SU 20103, loc. Gn-20; transverse section, x4. Specimen collected E. M. Basnett. Palaeontology, Vol. 8 PLATE 75 STRUSZ, Australian Devonian Rugosa D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLIDAE 549 tabularium, and are less strongly dilated in the dissepimentarium. The Tamworth form may not be conspecific with the Wellington specimens. Known localities. Cr-100, -113 (both in the same biostrome); P-43 (probably the same horizon as the above). Billingsastraea speciosa (Chapman 1914) Plate 76; text-figs. 10-14 1914 Phillipsastraea speciosa Chapman, p. 306, pi. 49, figs. 10, 11 ; pi. 50, figs. 12-14. Loyola, Victoria; Gedinnian. 1939a Phillipsastraea speciosa Chapman; Hill, p. 237, pi. 16, figs. 1, 2 only. 1942c Phillipsastraea speciosa Chapman; Hill, p. 183, pi. 6, figs. 8a, b. Wellington, N.S.W. ; Garra Formation, Emsian ? 1942c Phillipsastraea sp. Hill, p. 186 (not figured). Wellington, N.S.W. ; Garra Formation, Emsian ? 1962 Phillipsastrea speciosa Chapman; Philip, p. 176, pi. 24, fig. 6. Tyers R., Victoria; Cooper’s Creek Formation, Gedinnian. Holotype. GSV 2487 (Ferguson Colin.), and thin sections NM 1387, 1388 cut therefrom. Griffith’s Quarry, Loyola, near Mansfield, Victoria. The locality was considered by Philip (1960a) to be an equivalent of the Cooper’s Creek Formation of the Walhalla basin (south of Mansfield), and therefore early Gedinnian in age. Diagnosis. Astraeoid and thamnsastraeoid, with moderately to strongly fusiform septa, of which the major may reach the axis, or may withdraw almost to the dissepimentarium ; tabulae strongly concave. Description. A large number of specimens have been collected from the Garra Formation ; as the species is known from very few specimens at the several Victorian localities, I con- sider it worth while to describe fully the Garra material. The coralla are either lamellar, growing on a flat surface of sand, &c., or are broad, low, mushroom-shaped expansions. The largest corallum collected is about 16 cm. across and 6 cm. deep. Only one specimen has an unworn upper surface exposed : on it, the calices are shallow, saucer-shaped, with only slight axial depressions. Adjacent corallites meet in sharp, slightly raised rims. Thin sections show that individual coralla may be wholly thamnastraeoid, wholly astraeoid, or more usually both. Within one corallum the septa of adjacent corallites may follow several patterns: (1) the septa may terminate within the dissepimentarium, leaving a small gap between their ends ; (2) they may meet at a more or less abrupt angle, their ends being coincident or alternate; or (3) they may be completely confluent. Moreover, in a fair proportion of coralla, there may develop between some corallites an outer pseudotheca (Rozkowska 1953, see particularly p. 52; this structure is essen- tially the trabecular wall of Flower 1961, p. 26), formed by the union of the bifurcating ends of septa. This simulates the wall between ceriod corallites (e.g. Hexagonaria ), but may be clearly distinguished from that structure in lacking the median dark line (axial plate of Flower 1961) which indicates the presence of an epitheca around the individual corallites. The outer pseudotheca is in fact constructed of the same trabeculae as are the septa. When formed at all in B. speciosa, it is usually very incomplete ; consequently in thin section it is generally not possible to measure Dc. A more objective measure is Ts, 550 PALAEONTOLOGY, VOLUME 8 the distance between the axes of corallites whose septa abut or are confluent. The data are summarized in the following table of dimensions, and in somewhat more detail under ‘Variation’. Dimensions in mm. Dt n Ts Mean 1-76 23-9* 24f 5-12 Minimum 0-9 16 1-5 Maximum 2-8 32 13-8 Lowest meant 1-26 18-7* 18f 3-26 Highest meanf 2-37 27-8* 28f 7-49 No. of readings 374 374 1,050 * Calculated mean; f mean to nearest even number; $ means for individual coralla. Readings taken from 12 coralla, from localities Cr-94, -100, -113, P-26, -43. Most coralla show a fairly small range of n, but this range differs markedly from one corallum to another (see ‘Variation’). Most septa are greatly dilated, forming a wide stereozone around the tabularium. Outside this stereozone the degree of dilatation varies greatly. Generally the septa are fusiform, but in some coralla they remain dilated throughout the dissepimentarium. This dilatation is often confined to more or less clearly differentiated horizontal layers, which are separated vertically by layers in which the septal dilatation is much reduced. In the dilated layers, the dissepiments may be almost completely suppressed. In the regions of reduced dilatation the septa vary from strongly fusiform, with the stereozone around the tabularium, to only weakly fusiform and almost attenuate, so that there is no stereozone. In the dissepimentarium the septa may be smooth, but are more often irregularly carinate, the carinae being generally of the zigzag type. The carinae are usually only poorly developed, but on occasions may be very strong; the latter occurs mostly when the septa are fairly thin. Some septa split into two or three parallel thin strands in the dissepimentarium. Finally, there often occur septa which for a short distance become fragmented into discrete trabeculae. Minor septa barely enter the tabularium. The major septa are more variable; in some cases they extend unequally to the axis, either as continuous plates or as discrete trabe- culae ; in other cases they withdraw, sometimes to become equal in length to the minor septa. One septum, probably the counter septum, generally extends to the axis, where it is frequently enlarged, forming a rudimentary columella. The slender trabeculae are arranged in strong asymmetrical disphylloid fans. The zone of divergence corresponds with the centre of the septal stereozone. In the tabularium the trabeculae are axially inclined at about 45° from the vertical. The tabularium is narrow. In the coralla measured, Dt varied from 0-9-2-8 mm., but a plot of all Dt showed two peaks at 1-4 mm. and 2-0 mm. (see ‘Variation’), and most coralla correspond to one or other of these. The tabulae are moderately sagging, with EXPLANATION OF PLATE 76 Figs. 1 a-c. Billingsastraea speciosa (Chapman). SU 1 3206, loc. Cr-100; (a) transverse section showing a region of reduced septal dilatation ; ( b ) transverse section of a region of maximum septal dilatation, showing two corallites with shortened major septa; (c) longitudinal section; all x4. Specimen col- lected by Dr. J. R. Conolly. Palaeontology, Vol. 8 PLATE 76 STRUSZ, Australian Devonian Rugosa D. L. STRUSZ: D1SPHYLLIDAE AND PHACELLOPHYLLID AE 551 rather strongly upturned margins which are supplemented by strongly inclined elongate tabellae. The dissepimentarium is composed of small globose plates. In a narrow zone around each tabularium, there are two or three series which are very strongly inclined and rather elongate. Outside these, in the zone of maximum septal dilatation, there are one to three series of more or less horizontal dissepiments ; beyond these the dissepiments are gently outwardly inclined until midway between tabularia when they are again horizontal. Ontogeny. The earliest discernible stage in the formation of a corallite occurs generally at the point of intersection of three corallites. Careful inspection of the septa reveals a very small region in which there are six radially arranged plates ; some of these may text-fig. 10. Billingsastraea speciosa (Chapman) ; transverse section x 5, showing several buds. Traced from photograph of celluloid ‘peel’. be the ‘peripheral’ ends of septa, others may be carina-like offshoots of septa (text-fig. 10). The next observed stage is a corallite of about T2 mm. diameter, with some fourteen septa still not divisible into major and minor. At this stage, the tabularium is distinguish- able, Dt = 0-6 mm. Beyond this point, offshoots are not distinguishable from adult corallites. Apparently, therefore, the growth of a bud and the rate of septal insertion are both very rapid. This type of increase is the exact equivalent of the extra-tentacular budding of scleractinians (see Wells, p. F 348, and fig. 250, in Moore 1956). Variation. The three measurable variables in B. speciosa are Dt, n (measured at the margins of the tabularia), and Ts. Only for n and Dt can bivariate analysis be employed using individual readings ; for comparison with Ts, mean values of coralla must be com- pared. It has been found that all three variables show considerable variation from one corallum to another, but relatively little within the one corallum. Ts (text-fig. 11): this shows the widest range for data from one corallum. Frequency histograms show a broad spread more often than a sharp peak. This spread is even more pronounced in a plot of all data from twelve coralla: the range is from 1-5 mm. to 13-8 mm. overall. Using class intervals of 0-5 mm., the mode is 4-2 mm., and 68 per cent, of the data fall within the interval of classes 2-7-7 -2 mm. n (text-fig. 12): frequency histograms generally show a pronounced peak and a fairly small spread. However, the modes for the 12 coralla measured vary from 18 to 28, 552 PALAEONTOLOGY, VOLUME 8 27 4.2 7.2 text-fig. 1 1 . Billingsastraea speciosa (Chapman) ; frequency histograms (a-h) and percentage frequency curve ( i ) for Ts. a, loc. Cr-94; b-d, loc. Cr-100; e, loc. Ge-3; f-h, loc. P-43. i. Total of Ts from twelve coralla, localities Cr-94, -100, P-26, -43, Ge-3. text-fig. 12. Billingsastraea speciosa (Chapman) -, percentage frequency curve for n: 374 readings from twelve coralla, as text-fig. 1 1 . with a complete range between. A frequency curve for total n is also a fairly tight curve, with a peak at n = 24, and overall range 16-32. Dt (text-fig. 13): this provides the most interesting frequency distribution. Individual histograms vary considerably both in position and spread. However, a curve for total Dt is distinctly bimodal, modes 1-4 mm. and 2-0 mm., with a sharp low at 1-7 mm. Examination of the histograms in the light of this curve shows that only three of the twelve coralla have modes near 1 -7 mm., and of these two also show a very wide spread D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLID AE 553 of data. Most coralla fall clearly to one or other side of 1-7 mm., and in no one of these coralla do more than 4 per cent, of the measurements fall on the other side of 1 -7 mm. from the mode for that corallum. The plot of n against Dt (text-fig. 14) for all data reflects this dichotomy in Dt, but individual plots give a very wide spread in all directions. The correlation, while not close, text-fig. 13. Billingsastraea speciosa (Chapman) ; frequency histograms ( a-l ) and percentage frequency curve (m) for Dt. is enough to suggest that correlation would be obtained for individual coralla by measur- ing a large number of corallites. The dichotomy in Dt might suggest the possibility of dividing the species into two formae on this basis ; unfortunately this is not possible, as several coralla cannot easily be placed on one side or the other of the dividing line. Comparison. The Garra specimens, apart from the greater degree of variation shown by the much larger sample available, differ from the Victorian specimens principally in the high proportion of tabularia in which there is an elongate, columella-like counter ( ?) septum. The layered effect of the septal dilatation in the Garra coralla is visible to a lesser degree in the sections of the holotype. The relations of B. speciosa to other species have been noted by Hill (1939). Known localities. Locs. Cr-94, -100, -111,-113; Ge-3 ; P-26, -43 ; BR1 /177. All except 554 PALAEONTOLOGY, VOLUME 8 Ge-3 are apparently from the same biostrome, cropping out on either side of a major syncline in the Catombal Group, near Wellington. Known range. B. speciosa is known from the early Gedinnian of Loyola and the Tyers R. in Victoria, and the Emsian or early Couvinian of the Garra Formation. text-fig. 14. Billingsastraea speciosa (Chapman); correlation between n and Dt for twelve coralla, as text-fig. 11. Density plot contoured at 1, 2, 3, 4, 5 per cent, of total n. Family phacellophyllidae Wedekind 1922 Diagnosis. Solitary or colonial Rugosa with a well-developed dissepimentarium which includes a vertical series of horseshoe-shaped dissepiments ; in this series the horseshoe dissepiments may be replaced by peneckielloid dissepiments. Septa of two orders, com- posed of one or more radial series of monacanthine trabeculae which are arranged in fans whose zone of divergence coincides with the series of horseshoe dissepiments : the septa are frequently fusiform, and may be carinate. The tabularium may be simple, or composed of tabellae arranged in two or more series. Remarks. The structure of this family has been discussed in detail on pp. 523-4. Peneckiel- loid dissepiments, a variant form of horseshoe dissepiments, are discussed on p. 555. D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLIDAE 555 Genus peneckiella Soshkina 1939 1939 Peneckiella Soshkina, p. 23. 1949 Synaptophyllum Simpson; Stumm p. 37, non Simpson, 1900. (Subj. synonym, McLaren 1959, pp. 16, 22 q.v. for further discussion on the confusion of these two genera). 1953 Thamnophyllum Penecke 1894 ( partim ); Rozkowska, p. 14. 1958 Peneckiella Soshkina; Schouppe, p. 299. Contains an extensive synonymy. 1959 Peneckiella Soshkina; McLaren, p. 22. 1960 Peneckiella Soshkina; Rozkowska, p. 29. 1961 Peneckiella Soshkina; Lenz, p. 505. Type species. Diphyphyllum minus F. A. Romer 1855, p. 29, pi. 6, figs. 12 a-e. Soshkina 1939, p. 23, incorrectly called the species Peneckiella minima. Minus is the neuter form of minor (m., f.), the com- parative of parvus-, minimus is the superlative of this word. The type species is therefore Peneckiella minor (Romer). Diagnosis. Fasciculate or cerioid phacellophyllid with septa smooth or weakly carinate, dilated peripherally, or fusiform; with a peripheral series of horseshoe and peneckielloid dissepiments, supplemented by inner, and occasional outer, accessory dissepiments; tabulae generally complete. Discussion. This genus has been fully discussed recently by Fliigel (1956), Schouppe (1958), and McLaren (1959). Fliigel described topotypic material, and examined the holotype of Diphyphyllum minus Romer. Unfortunately, he only figured a transverse section, and a diagrammatic longitudinal section. Schouppe described and figured topotypic material. His figures show distinct trabecular fans, a series of horseshoe dissepiments, and occasional dissepiments on either side of this series. These dissepi- ments are here termed inner and outer accessory dissepiments. McLaren (1959) discussed the relationship of several American species to the type species of Peneckiella , and concluded that they differed from it in lacking horseshoe dissepiments. For the American species he erected the disphyllid genus Acinophyllum, which is characterized by carinate septa and only one or two series of normal dis- sepiments. Rozkowska (1960) described a subspecies of the type species, from the Frasnian of Poland, and, after examining topotypic material from Grund, concluded that the dis- sepimental structure of P. minor (s.l.) is variable. She distinguished four types of dis- sepiment, which are also recognized herein (text-fig. 15) : (i) horseshoe dissepiments sensu stricto\ (ii) ‘horizontal’ dissepiments, usually occurring as what are here termed outer accessory dissepiments; (iii) ‘peneckielloid’ dissepiments; and (iv) a rare variant of (iii), ‘sigmoidal’ dissepiments. She found all four types in the one corallite, but more usually only one or two types occur, particularly (i) and (iii) together. The ‘horizontal’ dissepiments somewhat resemble those in Thamnophyllum, but are far less uniformly developed. The ‘peneckielloid’ dissepiments consist of an inner half the same as for horseshoe dissepiments, and an outer half which slopes fairly steeply, straight down to the periphery. The ‘sigmoidal’ type is a variant of this, closely resembling a combination of horseshoe and horizontal dissepiments, particularly when the sigmoidal outer part touches the surface of a preceding dissepiment before proceeding to the periphery. As well as occurring in P. minor kunthi (Dames) of Rozkowska (1960, see figs. 22, 25, 27-29), this variability in dissepimental form can be seen in Schouppe’s (1958) figures of P. minor (Romer), and also in the specimens of P. mesa (Hill) described below. 556 PALAEONTOLOGY, VOLUME 8 Peneckiella mesa (Hill 1942) Plate 75, fig. 4; Plate 77, figs. 1, 3; Plate 78, fig. 1; text-figs. 16-21 1940a Disphyllum praecox ? Hill p. 399 (in text). Wellington, N.S.W. : ‘Devonian?’ 1942c Disphyllum mesa Hill p. 185, pi. 5, figs. 4, 5. Wellington; Garra Formation, Emsian? ?1954a Peneckiella teicherti Hill p. 25, pi. 2, figs. 29 a, b. West Kimberley Ra., W. Aust. (Fitzroy Basin); ‘ Atrypa beds’, Frasnian. non 1961 Peneckiella teicherti Hill?; Lenz p. 505, pi. 1, figs. 1, 2. Lower Mackenzie Valley, Canada ; Ramparts Limestone, Givetian. Holotype. SU 5276 (PI. 77, fig. 1), loc. Gn-20. Figured Hill 1942c, pi. 5, fig. 4. Other material figured: SU 17125, 20103, both loc. Gn-20. text-fig. 15. Dissepimental types found in phacellophyllid species (all diagrammatic, about x2 to x4). a. Horseshoe dissepiments ; b, peneckielloid dissepiments ; c, sigmoidal dissepiments ; d, horseshoe and horizontal dissepiments (as in Thamnophyllum Penecke); e, horseshoe and accessory dissepiments. ep — epitheca; hs — horseshoe series; p — peneckielloid series; si — sigmoidal series; ho — horizontal series; oad — outer accessory dissepiments; iad — inner accessory dissepiments; d — dissepimentarium ; t — tabularium. Diagnosis. Fasciculate Peneckiella with moderately carinate or smooth, thin or fusiform septa, high mesa-shaped tabulae, and dominantly peneckielloid dissepiments. Description. Much additional material has been collected, and so the species is fully re-described. The corallum is usually phaceloid. The corallites have a thin epitheca, with shallow septal grooves and only slight growth wrinkling. Dc is usually about 3-5 mm., but may be over 8 mm. The calix is deep, with vertical walls; the rim is narrow and slightly everted; the floor of the calix consists of a wide mesa-shaped axial boss, surrounded by a deep, narrow concentric trench. There are generally 30-40 septa (see table, p. 557) ; EXPLANATION OF PLATE 77 Figs. 1, 3. Peneckiella mesa (Hill). 1, Holotype SU 5276, loc. Gn-20, figured Hill (1942c, pi. 5, fig. 4); transverse section, x4. 3, Topotype SU 17125, loc. Gn-20; oblique and longitudinal sections — note calix and dissepiments in lowermost corallite (see also text-fig. 166); x 4. Figs. 2 a-c. Peneckiella teicherti Hill. Holotype, Univ. W. Aust. 33,515, figured Hill (1955a, pi. 2, fig. 29); Givetian or Frasnian, W. Kimberleys. a, longitudinal and b, c, transverse sections; x4, photographs by courtesy of Prof. D. Hill. Palaeontology, Vol. 8 PLATE 77 STRUSZ, Australian Devonian Rugosa D. L. STRUSZ: DISPHYLLIDAE AND PH ACELLOPH YLLIDAE 557 the major are long, unequal, leaving an axial space of c. % Dc; L2 is about J Lu the minor septa normally extended a little way into the tabularium. The septa are weakly to strongly dilated, generally fusiform, with the zone of maximum dilatation a little outside the margin of the tabularium. This dilatation frequently spreads over the surfaces of the innermost dissepiments, and forms an inner stereozone. The septa may also dilate peripherally. Within the dissepimentarium, where the dilatation is slight or moderate, irregular zigzag carinae frequently develop. In the tabularium the septa are thin, and straight or wavy. Specimen SU 5276* SU 17125f SU 17126f SU 13145$ Dc (mean) (mm.) 4-1 3-1 3-0 40 Dc (max.) 6-4 5-6 4-7 8-4 Dc (min.) 2-4 1-4 1-8 1-9 No. measurements 20 47 25 70 Dt (mean) (mm.) 2-5 1-9 1-8 2-3 Dt (max.) 3-9 3-3 2-8 3-2 Dt (min.) 1-5 10 1-2 1-3 No. measurements 20 47 25 58 Dt/Dc 0 61 0-61 0-59 0-56 n (mean) 32-5§ 314 30-8 37-1 32|| 32 30 38 n (max.) 40 37 38 46 n (min.) 23 22 26 28 No. measurements 20 33 18 58 n/Dc (mean) (mm-1) 0-83 M2 Ml 0-98 n/Dc (max.) 108 1-72 1-44 1-69 n/Dc (min.) 0-48 0-63 0-83 0-60 No. measurements 20 33 18 58 Z,2 (mean) 0-39 7? 0-38 R 0-46 7? L2 (max.) 0-51 7? 0-48 7? 0-59 7? L2 (min.) 0-21 R 0-24 7? 0-25 7? No. measurements 35 25 70 * holotype, loc. Gn-20; f topotypes, loc. Gn-20; :t homeotype, loc. Cr-54; § calculated mean; || mean to nearest even number of septa. The septa consist of slender, straight, or slightly curved trabeculae. These are arranged in phacellophylloid fans symmetrical about the axis of the series of horseshoe and peneckielloid dissepiments. The zone of divergence of the fans corresponds to the zone of greatest septal dilatation. The tabulae are generally complete, rarely incomplete. They are mesa-shaped; the axial parts are flat or gently sagging, while near the dissepimentarium they are turned abruptly down. The periaxial zone is narrow and deep ; in it the tabulae, augmented by occasional concave tabellae, are gutter-shaped, with flat or concave floors, high vertical inner walls, and low, upward-slanting outer walls which abut onto the dissepiments. Rare incomplete axial tabulae rest, on all sides, on the edges of the preceding ‘mesa’. Dt is c. 0-6 Dc, and the axial zone of the tabularium is about | Dt. The narrow dissepimentarium consists mainly of a single series of peneckielloid dissepiments, augmented by small horseshoe dissepiments, which in some corallites are 558 PALAEONTOLOGY, VOLUME 8 the dominant element, and by inner and outer accessory dissepiments. Rarely, the peneckielloid and horseshoe dissepiments fail, and the dissepimentarium is composed of several series of globose dissepiments whose inclination diverges from the mid- radius of the dissepimentarium. Ontogeny. Budding is parricidal. Two or three buds appear on top of the calical rim, and rapidly expand, sometimes to come into contact, while growth of the parent corallite ceases completely. Septa and dissepiments are added very rapidly, so that juvenile corallites soon assume the proportions of adults. text-fig. 16. Longitudinal sections, 10, of: a, Peneckiella minor kunthi (Dames), after Rozkowska (1960, fig. 27); b, P. mesa (Hill), SU 17125, traced from photograph (PI. 77, fig. 3). Variation (text-figs. 17-21). In a series of papers on Devonian phacellophyllids, Rozkow- ska (1953, 1956, 1957, 1960) has made a detailed study of many species, and has dis- tinguished them partly on a morphological basis, and partly statistically. Her statistical approach used the variation of n and Dt relative to Dc. To graph the data, she divided those for Dc into classes of 0-5 mm., and found the arithmetic mean of the data for n or Dt in each such class. For comparative purposes, similar data were obtained from a number of colonies collected from three localities in the Garra Formation, including the type. Scatter diagrams and mean curves for the individual colonies showed a greater spread than that of curves for separate species as figured by Rozkowska. As a further test, curves for n :Dc were constructed for one corallum, using class sizes of 0-3 mm. and 0-5 mm., and in the latter case class limits coinciding with those of Rozkowska (1960) for P. minor kunthi, and removed from these limits by half a class interval. These curves showed that the choice of class size and position has a very strong effect on the form of the resulting curve. Consequently the data and curves provided by Rozkowska cannot safely be used for comparison with other species, and doubt is thrown on Rozkowska’s application of these methods. Variation of n\Dc for P. mesa is wide; this is well shown in cumulative frequency curves of the ratio n/Dc, which vary strongly in both position and shape. However, scatter diagrams of Dt : Dc for the coralla studied show fairly strong concentrations, to which closely similar isometric growth curves, of the approximate form Dt = 0-6 Dc, D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLIDAE 559 R6zkowska: >s toJ 1 0 2 Dc text-fig. 17. Mean curves for n against Dc, Peneckiella mesa and P. minor kunthi (from Rozkowska 1960, fig. 26). 50 1" • 1 corallite . 2 coratlites . 3 20- . 4 . * 5 » Dct 0 2 4 6 mm. text-fig. 18. Scatter diagrams for n against Dc, Peneckiella mesa\ 138 measurements from 5 coralla, Iocs. Cr-54, Gn-20. 560 PALAEONTOLOGY, VOLUME 8 text-fig. 19. Cumulative frequency curves for n/Dc, Peneckiella mesa (five coralla) and P. minor kunthi (calculated from Rozkowska 1960). text-fig. 20. Scatter diagram and estimated mean curve for Dt against Dc, Peneckiella mesa', 158 measurements from five coralla, Iocs. Cr-54, Gn-20. D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLID AE 561 may be fitted. It is possible that the ratio Dt/Dc may prove of greater diagnostic value for species of Peneckiella than the ratio n\Dc\ unfortunately data for the European species are lacking at present. Frequency curves for the ratio L2/Z>c also show close correspondence between coralla. Plotted by classes 0-5 apart, curves for three coralla have a very strong mode at 4-8, and moderate right skew. This ratio also may prove a useful diagnostic parameter (com- pare P. boreensis, p. 562). text-fig. 21. Percentage frequency curves for L2/Dc from three coralla of Peneckiella mesa, and the holotype of P. boreensis sp. nov. Comparison. The trabecular and dissepimental structure of this species, as shown very clearly by topotypic material, are those characteristic of Peneckiella Soshkina, as the genus is interpreted (from the type species) by recent workers who have examined the type material (Fliigel 1956; Schouppe 1958; Rozkowska 1960). P. mesa differs from P. minor (Romer) in greater size, more septa, and the possession of strongly mesa-shaped tabulae; the mode of budding also appears to be somewhat different (see Rozkowska 1960). P. mesa does not resemble any of the species described by Soshkina from the Russian Upper Devonian. P. teicherti Hill 1954, from the middle Frasnian Ladjia saltica zone of the Fitzroy Basin, is tentatively placed in synonymy with P. mesa : its full range of variation is unknown, there being but one specimen. The only difference that I can detect is in the accessory dissepiments. In P. teicherti there is one series of small, vertical inner accessory dissepiments, but an outer series is rarely developed (PI. 77, figs. 2 a-c ). P. teichertH of Lenz (1961) is cerioid or phacelo-cerioid, with flat or sagging tabulae, and so differs strongly from both P. mesa and P. teicherti. Its tabulae are closer to those of the type species, and to some of the Russian species. P. boreensis sp. nov. differs from P. mesa in having very strongly dilated septa within the dissepimentarium ; the minor 562 PALAEONTOLOGY, VOLUME 8 septa are shorter, and the tabulae are more variable, seldom forming the high ‘mesas’ of P. mesa. Known localities. Gn-20 (type locality); BN-2; Cr-42, -46, -54; Ct-53, -64; MM-6; P-43. Peneckiella boreensis sp. nov. Plate 78, figs. 2 a-c\ text-fig. 21 Holotype. SU 12118, loc. BN-1. Derivation of name. From the parish of Boree Nyrang, in which occurs the type locality. Diagnosis. Phaceloid Peneckiella in which the septa are extremely dilated in the dissepi- mentarium, often forming a wide stereozone; with short minor septa. Description. The corallum is apparently dendroid; the only known specimen is inter- grown with a stromatoporoid. Maximum Dc is 10 mm., most being 6-8 mm. The calix is generally deep, with steep sides and a wide, flat base. There is usually a low, flat- topped axial boss. The epitheca is thin, and apparently does not extend distally as far as the calical edge, as in several sections of adult corallites the wall of the corallite is formed by the rounded outer edges of the septa, as in Macgeea and Thamnophyllum. Several sections also show sharp rejuvenescence rims. Budding occurs at the outer margin of the tabularium, and is parricidal, as in P. mesa. The number of buds formed is not known. Dimensions in mm. Dc Dt Dt/Dc n u u Mean 4-5 2-8 0-65 32 0-76 R 0-25 R Max. 101 6-5 0-78 56 0-94 R 0-41 R Min. 0-9 0-7 0-51 6 0-47 R 00 R No. of Readings 24 24 24 19 24 24 The long major septa extend unequally to the axis ; the minor septa end at the margin of the tabularium. Septal dilatation is strong in the dissepimentarium, where the septa are generally considerably wider than the interseptal loculi, and are frequently in con- tact to form a wide stereozone ; the dilatation is fusiform. In the tabularium the septa are attenuate and generally wavy. Carinae are lacking. The tabularium is wide; the tabular floors are gently domed to moderately mesa-shaped, often sagging axially. The tabulae are both complete and incomplete; at the margins of the ‘mesas’ there are frequent moderately domed tabellae. In juvenile corallites the tabulae are generally EXPLANATION OF PLATE 78 Figs. 1 a-c. Peneckiella mesa (Hill). Topotype SU 20103, loc. Gn-20. a, b, oblique transverse, and c, longitudinal sections; note the combination of horseshoe and peneckielloid dissepiments; > 4. Specimen collected E. M. Basnett. Figs. 2 a-c. Peneckiella boreensis sp. nov. Holotype SU 121 18, loc. BN-1. a. Longitudinal sections — note calix and horseshoe dissepiments in corallite at top left, horseshoe dissepiments in corallite at top right; b, longitudinal, and c, transverse sections; corallum invested by a stromatoporoid; x4. Figs. 3 a-c. Peneckiella sp. cf. minor kunthi (Dames) sensu Rozkowska, 1960. SU 11273, loc. BC-6; a, longitudinal and oblique transverse sections ; b, longitudinal, and c, oblique transverse sections ; x 4. Compare text-fig. 1 6a. Corallum invested by a stromatoporoid. Palaeontology, Voi. 8 PLATE 78 STRUSZ, Australian Devonian Rugosa m D. L. STR.USZ: DISPHYLLIDAE AND PHACELLOPH YLLID AE 563 complete, flat to sagging, and only rarely mesa-shaped. The narrow dissepimentarium consists of peneckielloid dissepiments, some horseshoe dissepiments, and occasional accessory dissepiments. The dissepiments are confined to narrow interseptal loculi, and are frequently immersed in the trabecular tissue of the zone of great septal dilata- tion. Variation. There are insufficient corallites available for a detailed study of variation; however, the available data show that for the plots of Dt : Dc and n : Dc, P. boreensis cannot be separated from P. mesa, except in so far as n and Dc have higher maxima. The major difference between the two species lies in the ratio L2/Dc. For P. mesa, frequency curves for this ratio have a strong peak at 4-8 mm., and are moderately right-skewed (see text-fig. 21). For P. boreensis the mode of 6-8 mm. is not marked, being the peak of a broadly symmetrical curve. Comparison. P. boreensis differs from P. mesa principally in its considerably greater degree of septal dilatation in the dissepimentarium. The minor septa are shorter, being invariably confined to the dissepimentarium; the tabulae are more variable, and generally do not form the extreme ‘mesas’ characteristic of P. mesa. Finally, the maxi- mum diameter is greater. Besides P. mesa, P. boreensis is most readily distinguished from all other described species of Peneckiella by the great dilatation of its septa. Known localities : Loc. BN-1. Peneckiella minor (F. A. Romer) kunthi (Dames 1868) 1868 Cyathophyllum kunthi Dames, p. 699 ( fide Fliigel 1956, p. 360, and Rozkowska 1960 p. 29). 1960 Peneckiella minor (Romer) kunthi (Dames); Rozkowska, p. 29, figs. 20-29. Contains a complete synonymy. Diagnosis. ‘Phaceloid colony; corallites straight, covered by thick epitheca, locally touching. Diameter 2-5 to 4-8 mm. Number of septa ranging from 12x2 to 16x2, only exceptionally 18x2. Major septa long with bent axial ends, thick and zigzagged, within the dissepimentarium frequently carinate. Minor septa short. Double row of diversely shaped dissepiments (horizontal, horseshoe, sigmoidal, peneckielloid). Tabulae usually complete, horizontal or concave. Trabecular fans, trabeculae thick (0-08-0T6 mm.). Budding latero-thamnophylloid.’ (Rozkowska 1960, p. 29.) Peneckiella sp. cf. minor kunthi (Dames) sensu Rozkowska 1960 Plate 78, figs. 3 a-c Material. SU 11273, loc. BC-6. Description. The available material consists of corallites irregularly scattered through a stromatoporoid; the corallum is probably dendroid. The corallites are initially ceratoid. later cylindrical, and show strong growth irregularities. Adults are about 4-5-5 mm. in diameter. The deep calix apparently has a rather narrow everted rim, steeply inclined sides, and a flat floor. Budding is apparently peripheral, the initial stage being marked by slight withdrawal of two or three adjacent septa from the periphery. 564 PALAEONTOLOGY, VOLUME 8 There are 34-36 strongly fusiform septa in adults. The major, sometimes more dilated than the minor, extend very unequally to the axis : some are about § R, while one or two meet or pass at the axis. The septa are attenuate in the tabularium. The minor septa end at the margin of the tabularium, at about \ R. The septa are slightly wavy in the tabularium, and irregular or zigzag in the dissepimentarium, particularly peripherally. The sides of the septa may be rough, particularly in dilated portions, but carinae are not developed. The septal dilatation spreads over the dissepimental surfaces just outside the tabularium, forming an inner wall as in Thamnophyllum. The trabeculae are arranged in nearly symmetrical phacellophylloid fans, but are always vertical or nearly so. The tabularium, a little over \ Dc, is composed of thin, generally complete tabulae. These are flat or a little irregular, and are partly supplemented periaxially by an inter- mittent series of axially inclined tabellae. The dissepimentarium consists predominantly of sigmoidal dissepiments (see p. 555), with some peneckielloid and horseshoe dissepiments. In addition, there are occasional outer accessory dissepiments, and an intermittent series of inner accessory dissepiments, which are generally small and vertical. Comparison. This specimen closely resembles the Polish species in the structure of the septa, tabularium, and dissepimentarium, but is slightly larger, with correspondingly more septa. It differs from P. mesa principally in having flat tabulae, supplemented periaxially by strongly inclined tabellae. P. boreensis is larger, with shorter minor septa, and also differs in the structure of the tabularium. Remarks. The Polish species is discussed further in conjunction with P. mesa. See text- figs. 16, 17, 19. Known localities. BC-6. Genus phillipsastrea d’Orbigny 1849 1849 Phillipsastrea d’Orbigny, p. 12 ( fide Lang, Smith, and Thomas 1940, p. 99). 1850 Pachyphyllum Milne-Edwards and Haime, p. lxviii ( fide Lang, Smith, and Thomas 1940, p. 92). 1958 Phillipsastrea d’Orbigny; Schouppe, p. 234. Contains an extensive synonymy. 1961 Pachyphyllum Edwards and Haime; Semenoff-Tian-Chansky, Lafuste, and Delga, p. 304. Type species. Astraea hennahi Lonsdale 1840, p. 697 ( partim — pi. 58, figs. 3-3 b only); subsequent designation Milne-Edwards and Haime 1850, p. lxxi {fide Lang, Smith, and Thomas 1940, p. 99). Diagnosis. Astraeoid, thamnastraeoid, or aphroid phacellophyllid. Remarks. The lectotype of P. hennahi has recently been shown to possess a series of horseshoe dissepiments (Schouppe 1958, p. 234), and so Phillipsastrea becomes a senior synonym of Pachyphyllum. As noted by Lang, Smith, and Thomas (1940, p. 99), much confusion has been caused by the invalid designation, by Milne-Edwards and Haime (1851, p. 173), of Erismatolithus Madreporites radiatus Martin 1809, as ‘Exemple’ of Phillipsastrea. That this confusion still exists is shown by the fact that this species is still quoted as ‘type species ’ by Russian authors — e.g. Soshkina (1954, p. 46) and Bulvanker (1958, p. 118). D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLIDAE 565 Phillipsastrea oculoides Hill 1942 Plate 73, fig. 7 ; text-fig. 22 1942c Phillipsastraea ocm/oiV/cs Hill, p. 186, pi. 6, fig. 9. Wellington, N.S.W. ; Garra Fm.,Emsian? Holotype. SU 5281, loc. MM-10. Diagnosis. Partly aphroid Phillipsastrea with short major septa; septa strongly dilated around tabularium; tabulae complete and incomplete, gently concave. text-fig. 22. Longitudinal section of Phillipsastrea oculoides Hill, x4. Traced from photograph of holotype, SU 5281. Description. The type (and only) specimen is poorly preserved. Additional longitudinal sections have been made, and so the following comments may be added to Hill's description. (1) There are one or two series of elongate, steeply axially inclined dissepiments inside the series of horseshoe dissepiments. (2) The dissepiments between tabularia are very irregular in size, but those near the horseshoe dissepiments are frequently very small. (3) The tabular floors are gently sagging, with somewhat upturned edges ; the tabulae are both complete and incomplete, and vary from sagging to moderately domed. (4) In transverse section, the septal dilatation around the tabularium may be in a single zone 3 mm. wide, or may divide into two narrow zones on either side of the series of horseshoe dissepiments. Comparison. P. oculoides resembles P. ibergense (F. A. Romer) in its short major septa dilated around the tabularium, but differs in having more septa, and a partly aphroid corallum. In the latter character it resembles P. ibergense var. progressa Rozkowska 1953, which however is smaller. Both of these species are from the Frasnian of Poland (Rozkowska 1953). Known localities. MM-10. 566 PALAEONTOLOGY, VOLUME 8 FOSSIL LOCALITIES CITED The numbers given to localities are those used during field-work, and do not form part of any locality numbering system of the Department of Geology, University of Sydney; an extended list may be found in Strusz (1963). Numbers were allotted according to parishes, being consecutive within each parish, and included a letter prefix indicating the appropriate parish. In the following list, the localities are listed alphabetically by parishes. In each case, the exact loca- tion is given : ( a ) by reference to portion numbers ; ( b ) by reference to the Dubbo (SI 55-4) and Bathurst (SI 55-8) 1: 250,000 topographic sheets, with grid references to the nearest 100 yards; and (c) by nearby geographic features. In addition, the lithology from which the specimens were collected is given. Parish Boree Cabonne, co . Ashburnham. BC-6: portion 70 (north side, centre); grid reference 1796.8890 (Bathurst sheet); in Mousehole Ck., 1810 yds. along creek, east from Orange-Parkes road. ‘Rubbly’ limestone (thinly interbedded calcarenite and shale.) Parish Bell, co. Ashburnham. Be-10: portion 81 (north-west corner), 75 yds. east of portion 62, and near portion 82; grid ref. 1749.9103 (Bathurst); in gully. Coarse calcirudite. Parish Boree Nyrang, co. Ashburnham. BN-1 : portion 222 (south-west sector), 170 yds. north of portion 201, and 380 yds. west of portion 111 ; grid ref. 1796.8956 (Bathurst); in Walkers Ck., east of road. Coarse calcirudite. BN-2 : junction of portions 4, 9, and 120; grid ref. 1803.8928 (Bathurst); west bank of Walkers Ck., just west of road bridge. Calcarenite (biostromal ?). Parish Curra, co. Gordon. Cr-4: portion 171 (east side), extending into south-west corner of portion 153, parish Gundy, co. Gordon; grid ref. 1837.9680 (Dubbo); in bed, and on east bank, of Curra Ck. Fine calcarenite. Cr-1 2: portion 90 (south-west corner); grid ref. 1823.9674 (Dubbo); Curra Ck. ‘Rubbly’ limestone. Cr-36 b: portion 173, just west of portion 1 12, and 150 yds. south of portion 172; grid ref. 1836.9672 (Dubbo) ; near gully, south of Wellington-Parkes road. Calcarenite. Cr-42: boundary between portions 175 and 176, 160 yds. west of portion 89; grid ref. 1833.9661 (Dubbo); pile of ‘floaters’ collected from nearby area, beside fence, just south-west of bend in gully. All the floaters are calcarenite. Cr-46: portion 59 (south-west sector), 110 yds. north of portion 83, and 280 yds. east of portion 74; grid ref. 1818.9575 (Dubbo). Calcarenite. Cr-54: portion 172 (south-west sector), 208 yds. north-east of junction, portions 166, 167, 172, 173; grid ref. 1833.9673 (Dubbo); south bank of road cutting, Wellington-Parkes road. Hill’s (1942c) ‘ Fingerpost ’ locality. ‘ Rubbly ’ limestone. Cr-94: portion 39, just north of boundary with portion 1, and 570 yds. west of the Bell River; grid ref. 1868.9641 (Dubbo); hillside west of road. Biostromal limestone (coralline) — same horizon as Iocs. Cr-100, -111, -113, and BR7177, and probably as P-26, -43. Cr-1 00: portion 80, 5 yds. west of portion 6, and 70 yds. north of boundary between portions 6 and 13; grid ref. 1864.9625 (Dubbo); hillside west of road. Coralline biostrome. Cr-1 06: boundary between portions 10 and 80, 70 yds. west of junction with portion 9; grid ref. 1861.9614 (Dubbo); west of road, west bank of Bell R. Black foetid fossiliferous pellet cal- carenite. Cr-1 1 1 : portion 1 1 1 (southern end), 70 yds. north of portion 39, and 180 yds. west of Bell R. ; grid ref. 1870.9646 (Dubbo). Coralline biostrome. Cr-1 13 : portion 50 (southern end) ; grid ref. c. 1871.9651 ; west bank of Bell R. Coralline biostrome. Parish Catombal, co. Gordon. Ct-18: portion 45 (north-west sector), c. 100 yds. east of portion 38; grid ref. 1755.9444 (Dubbo); in gully, a tributary of Back Ck. Pink crinoidal/coralline biostrome. D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLIDAE 567 Ct-28 : portion 40 (south-west corner), c. 30 yds. east of junction with portions 30, 45 ; grid ref. 1747.9437 (Dubbo); Yellow biostromal limestone; same horizon as loc. Ct-18. Ct-40: portion 125 (east side), 200 yds. due west from portion 48; grid ref. 1761.9417 (Dubbo): north bank of Sawpit Gully (tributary of Loombah Ck.), c. 1 ,000 yds. north-east of road crossing. ‘Rubbly’ limestone (biostromal?). Ct-53: portion 58 (north side, centre), midway between portions 77 and 79; grid ref. 1767.9391 (Dubbo); south bank of Loombah Ck., extending over the interval 75-200 yds. north of road. Coralline/brachiopodal biostrome. Ct-64: portion 65 (east centre), just west of boundary with portion 14, and 250 yds. south of portion 58; grid ref. 1771.9373 (Dubbo). Coralline biostrome. Parish Eurimbula, co. Gordon. E-16: portion 123 (west side), extending for 70 yds. east of portion 70, and 90-280 yds. north of portions 13 and 52; grid ref. c. 1765.9240 (Dubbo). ‘Rubbly’ limestone. E-21: portion 57, 580 yds. south of portion 27; grid ref. c. 1787.9280 (Dubbo); in gully, 400 yds. east of road. ‘Rubbly’ limestone; same horizon as loc. E-16. Parish Geurie, co. Lincoln. Ge-3: portion 210 (north-east corner); grid ref. 1777.9896 (Dubbo); south bank of road cutting, Mitchell Highway, on either side of mile-post 12 miles north-west of Wellington. Hill’s (1942c, p. 183) locality ‘ Wellington-Dubbo Road near Geurie, 12 miles from Wellington’. Calcarenite and calcilutite. Parish Gundy, co. Gordon. Gn-10: portion 132 (south-east sector), 60 yds. west of portion 115; grid ref. 1830.9644 (Dubbo); 600 yds. north of road. ‘ Rubbly ’ limestone. Gn-20: portion 30 (north-west corner); grid ref. c. 1795.9699 (Dubbo); south side of Wellington- Parkes road, c. 900 yds. east of Suntop Public School. Calcarenite (biostromal?). Parish Mickety Mulga, co. Gordon. MM-6: portion 35 (west), 35 yds. east of portion 36; grid ref. 1811.9812 (Dubbo); in gully, 100 yds. west of minor access track, and c. 1,100 yds. south of access road to ‘The Holmes’ homestead. Calcarenite. MM-10: boundary of portions 60 and 247, c. 500 yds. west of portion 208; grid ref. 1810.9863 (Dubbo); outcrop extends south from fence (portion 60), 200 yds. towards road. Hill’s (1942c) locality ‘ Portion 247, north of road ’. Calcarenite, in a succession of unfossiliferous calcilutites and pellet calcarenites. Parish Ponto, co. Gordon. P-13: boundary of Water Reserve 33680 and portion 142, 590 yds. north of portion 131, parish Gundy; grid ref. 1821.9740 (Dubbo). Poorly fossiliferous silty calcarenite. P-26: portion 142 (centre), c. 530 yds. south of portion 104: grid ref. 1819.9761 (Dubbo); south of gully, and 570 yds. due east of road. Coralline biostrome. P-43: portion 103 (west side); grid ref. c. 1818.9786 (Dubbo); hillside east of ‘Macquarie Park’ homestead and outbuildings. Complex of thin coralline, bryozoan, and brachiopodal biostromes, calcarenites, lenticular pellet calcarenites, and quartzose arenites and lutites. Same horizon as loc. P-26, and probably as the Cr-100 biostrome. Measured Sections Disphylloid corals are herein described from two localities along measured sections, full details of which may be found in Strusz (1963). The two localities are: BR7177: 177 ft. west from start of section BR1 (a marked tree c. 20 yds. west of the Bell R.); sec- tion measured along gully, southern edge of portion 50, parish Curra. This locality is just south of loc. Cr-113, and on the same horizon. CAT/255 : 255 ft. east from start of section CAT (marked point on east side of base of a pair of wheat silos, ‘ Catombal ’ property, portion 29, parish Catombal) ; section measured east across fields. Olive-grey calcarenite. pp B 6612 568 PALAEONTOLOGY, VOLUME 8 Acknowledgements. This paper is based on work submitted as part of a Ph.D. thesis in the University of Sydney with funds provided under a Sydney University Post-graduate Research Studentship. I would like to thank Professor C. E. Marshall for allowing me the facilities of his department. Further work has been done in the University College of Townsville, with the aid of research funds from the University of Queensland. I would like to thank Dr. T. B. H. Jenkins and Professor Dorothy Hill for their encouragement and assistance and for the advice and criticism they have offered during the preparation of this paper. I have also benefited from discussions with Dr. G. H. Packham and Mr. A. J. Wright, Dr. G. M. Philip and Mr. A. E. H. Pedder, Dr. J. R. Conolly, Dr. K. H. Wolf, and Mr. A. H. Rizvi. Alan Pedder very kindly lent me the typescript of his paper on Zelolasma and Sulcorphyllum. For hospitality during field-work, I wish to thank Mr. R. Sutherland, Mr. L. Munroe, and Mr. T. Wilson. REFERENCES basnett, e. m. and colditz, m. j. 1946. General Geology of the Wellington District, N.S.W. J.Proc. Roy. Soc. N.S.W. 79 (1945), 37-47. bulvanker, E. z. 1958. Devonskie Chetyrekhluchevye Korally okrain Kuznetzkogo basseyna. VNIGRI, Leningrad. chapman, E. J. 1893. On the Corals and Coralliform Types of Palaeozoic Strata. Trans. Roy. Soc. Canada, 39, 4. chapman, f. 1912. Reports on Fossils (Middle Devonian of the Buchan District). Rec. Geol. Surv. Victoria, 3, 218-23, pi. 33-36. 1914. New Silurian Fossils of Eastern Victoria, Part 3. Ibid. 3, 301-16, pi. 46-61. conolly, j. r. 1963. Upper Devonian Stratigraphy and Sedimentation in the Wellington-Molong District, N.S.W. J. Proc. Roy. Soc. N.S.W. 96, 73-106, pi. 1-3. dames, w. 1868. Letter to A. Kunth. Z. deutsch. geol. Gesell. 21, 699-700. d’orbigny, a. 1849. Note sur des polypiers fossiles. 1-12; Paris. etheridge, r., jr. 1892. Descriptions of four Madreporaria Rugosa — Species of the General Phillips- astraea, Heliophyllum, and Cyathophyllum — from the Palaeozoic Rocks of N.S. Wales. Rec. Geol. Surv. N.S.W. 4, 11-18, pi. 3-4. 1895a. Additional Notes on the Palaeontology of Queensland. Part I — Palaeozoic. Proc. Linn. Soc. N.S.W. (ser. 2), 9, 518-39, pi. 29-41. 18956. An Undescribed Coral from the Wellington Limestone, N.S. Wales. Rec. Geol. Surv. N.S.W. 4, 160-2, pi. 21-22. 1898. On a New Form of Syringopora, Allied to Syringopora tabulata. Van Cleve. Rec. Geol. Surv. N.S.W. 5, 149-53, pi. 16. ’ 1902. Additions to the Middle Devonian and Carboniferous Corals in the Collections of the Australian Museum. Rec. Aust. Mus. 4, 253-62. 1903. An Unusually Large Form of Rhizophyllum, lately discovered in N.S. Wales. Rec. Geol. Surv. N.S. W. 7, 232-3, pi. 47. 1 907. A Monograph of Silurian and Devonian Corals of N.S.W. Part II — The Genus Tryplasma. Geol. Surv. N.S.W., Palaeont. Mem. 13 (2), i-ix+41-102, pi. 10-28. flower, r. h. 1961. Montoya and Related Colonial Corals. Mem. N. Mex. Inst. Min. Technol., St. Bur. Mines Miner. Resour. 7, 1-97, pi. 1-52. flugel, H. 1956. Kritische Bemerkungen zum Genus Peneckiella Soshkina. Neues Jahrb. Geol. u. Palaont., Abh. 8 ( Monatschr . 1956), 355-65. fontaine, h. 1961. Les madreporaires paleozoiques du Viet-Nam, du Laos, et du Cambodge. Archs. Geol. du Viet-Nam, 5, 1-276, pi. 1-35. fromentel, e. de. 1861. Introduction a V etude des Polypiers fossiles. Paris. geinitz, H. B. 1845. Grundriss der Versteinerungskunde. 1-224 (-400?), pi. 1-8 (-16?); Dresden and Leipzig. goldfuss, G. A. 1826. Petrefacta Germaniae. 1. 76 pp., 24 pis. Diisseldorf. grabau, a. w. 1917. Stratigraphic Relationships of the Tully Limestone and the Genessee Shale in Eastern North America. Bull. Geol. Soc. Amer. 28, 945-58. gurich, G. 1896. Das Palaeozoicum des Polnischen Mittelgebirges. Verb. Russ.-Kais. Min. Gesell. St. Petersburg, (2), 32, i-vi+ 1-539, pi. 1-15. D. L. STRUSZ: DISPHYLLIDAE AND PH ACELLOPH YLLIDAE 569 hill, d. 1936. Upper Devonian Corals from Western Australia. J.Roy.Soc. West. Aust. 22, 25-39, pi. 1 . 1939. The Devonian Rugose Corals of Lilydale and Loyola, Victoria. Proc. Roy. Soc. Victoria (N.S.), 51, 219-56, pi. 13-16. 1940cl The Silurian Rugosa of the Yass-Bowning District, N.S.W. Proc. Linn. Soc. N.S.W. 65, 388-420, pi. 11-13. 19406. The Lower Middle Devonian Rugose Corals of the Murrumbidgee and Goodradigbee Rivers, N.S.W. J. Proc. Roy. Soc. N.S.W. 74, 247-76, pi. 9-11. — 1942a. The Middle Devonian Rugose Corals of Queensland — III. Burdekin Downs, Fanning River, and Reid Gap, North Queensland. Proc. Roy. Soc. Qd. 53, 229-68, pi. 5-11. 19426. The Devonian Corals of the Tamworth District, N.S.W. J. Proc. Roy. Soc. N.S.W. 76, 142-64, pi. 2-4. 1942c. Middle Palaeozoic Rugose Corals from the Wellington District, N.S.W. Ibid. 182-9, pi. 5-6. 1954a. Coral Faunas from the Silurian of New South Wales and the Devonian of Western Australia. Bull. Bur. Miner. Resour. Geol. Geophys. Aust. 23, 1-51, pi. 1-4. 19546. Devonian Corals from Waratah Bay, Victoria. Proc. Roy. Soc. Victoria (n.s.), 66, 105- 18, pi. 6-9. and jones, o. a. 1940. The Corals of the Garra Beds, Molong District, New South Wales. /. Proc. Roy. Soc. N.S.W. 74, 175-208, pi. 2-8. jack, r. l. and etheridge, r. jr. 1892. The Geology and Palaeontology of Queensland and New Guinea. Geol. Surv. Qd., Publ. 92, i-xxxii+ 1-768, pi. 1-68. jones, o. a. 1936. The Controlling Effect of Environment upon the Corallum in Favosites : with a Revision of some Massive Species on this Basis. Ann. Mag. Nat. Hist., ser. 10, 17, 1-24, pi. 1-3. 1944. Tabulata and Heliolitida from the Wellington District, N.S.W. J. Proc. Roy. Soc. N.S.W. 77 (1943), 33-39, pi. 1. and hill, d, 1940. The Heliolitidae of Australia, with a Discussion of the Morphology and Systematic Position of the Family. Proc. Roy. Soc. Qd. 51, 183-215, pi. 6-11. joplin, g. a. and culey, A. g. 1938. The Geological Structure and Stratigraphy of the Molong- Manildra District. J. Proc. Roy. Soc. N.S.W. 71 (1937), 267-81, pi. 2. joplin, g. a. and others. 1952. A Note on the Stratigraphy and Structure of the Wellington-Molong- Orange-Canowindra Region. Proc. Linn. Soc. N.S.W. 77, 83-88, pi. 1. lang, w. d. and smith, s. 1934. Ludwig’s ‘Corallen aus palaolithischen Formationen’ and the Geno- type of Disphyllum de Fromentel. Ann. Mag. Nat. Hist., ser. 10, 13, 78-81. 1935. Cyathophyllum caespitosum Goldfuss, and other Devonian Corals considered in a Revision of that Species. Q. J. Geol. Soc. London, 91, 538-90, pi. 35-37. and thomas, h. d. 1940. Index of Palaeozoic Coral Genera. Brit. Mus. (Nat. Hist.), London; vii + 231 pp. lenz, a. c. 1961. Devonian Rugose Corals of the Lower Mackenzie Valley, Northwest Territories. Geology of the Arctic, 500-14, pi. 1-3. Univ. Toronto Press. lonsdale, w. 1840. Determinations of Corals. In Sedgwick, A., and Murchison, R. I., On the Physical Structure of Devonshire, . . . Trans. Geol. Soc. London, ser. 2, 5, 697. mclaren, d. j. 1959. A Revision of the Devonian Coral Genus Synaptophyllum Simpson. Bull. Geol. Surv. Canada, 48, 15-33, pi. 7-10. mansuy, h. 1913. Paleontologie de l’Annam et du Tonkin. Mem. Serv. Geol. Indochine, 2 (3), 1-48, pi. 1-6. milne-edwards, h. and haime, j. 1850. A Monograph of the British Fossil Corals. Part I. Intro- duction, &c. Palaeontogr. Soc. i-lxxxv, 1-71, pi. 1-11. 1851. Monographic des Polypiers fossiles des terrains palaeozoiques. Arch. Mus. Hist. nat. Paris, 5, 1-502, pi. 1-20. moenke, m. 1954. Genus Hexagonaria in the Devonian of the Holy Cross Mountains. Acta Geol. Polonica, 4 (4), pars palaeont. 445-83, pi. 1-2. (Polish; English summary 138-44.) moore, r. c. (ed.) 1956. Treatise on Invertebrate Paleontology, Part F — Coelenterata. Geol. Soc. Amer. and Univ. Kansas Press. packham, g. h. 1954. A New Species of Hadrophyllum from the Garra Beds at Wellington, N.S.W. J. Proc. Roy. Soc. N.S.W. 87, 121-3. 570 PALAEONTOLOGY, VOLUME 8 packham, g. h. 1960. Sedimentary History of Part of the Tasman Geosyncline in South Eastern Australia. Rep. XXI Int. Geol. Congr. 7 (Regional Palaeogeogr.), 74-83. peddhr, a. e. h. 1964. Two New Genera of Devonian Tetracorals from Australia. Proc. Linn. Soc. N.S.W. 88 (1963), 364-7, pi. 19. penecke, k. A. 1894. Das Grazer Devon. Jb. K.-K. Geol. Reichsanst., Jahrg. 1893, 43, 567-616, pi. 7-12. philip, g. m. 1960a. Victorian Siluro-Devonian Faunas and Correlations. Rep. XXI Int. Geol. Congr. 7, 143-57. 19606. The Middle Palaeozoic Squamulate Favositids of Victoria. Palaeontology, 3, 186-207, pi. 30-34. 1962. The Palaeontology and Stratigraphy of the Siluro-Devonian Sediments of the Tyers Area, Gippsland, Victoria. Proc. Roy. Soc. Victoria, 75, 123-246, pi. 11-36. piveteau, J. (ed.) 1952. Traite de Paleontologie. Masson et Cie., Paris. romer, c. F. 1883. Lethaea geognostica. I. Theil. Lethaea palaeozoica. 1 (2), 1 1 3-544 ; Stuttgart. romer, f. a. 1855. Beitrage zur geologischen Kenntnis des nordwestlichen Harzgebirges. Dritte Abtheilung. Palaeontographica, 5, i-iv, 1-44, pi. 1-8. rozkowska, m. 1953. Pachyphyllinae et Phillipsastraea du Frasnian de Pologne. Palaeont. Polonica, 5 (1952), i-vi, 1-89, pi. 1-8. 1956. Pachyphyllinae from the Middle Devonian of the Holy Cross Mts. Acta Palaeont. Polonica, 1, 271-322. 1957. Considerations on Middle and Upper Devonian Thamnophyllidae Soshkina in Poland. Part II. Ibid. 2, 81-146. 1 960. Blastogeny and Individual Variations in Tetracoral Colonies from the Devonian of Poland. Ibid. 5, 3-64. schluter, c. 1885. Diinnschliffe von Zoantharia rugosa, Zoantharia tabulata und Stromatoporiden aus dem Palaontologischen Museum der Universitat Bonn, Aussteller Professor Dr. C. Schluter in Bonn. Cat. de V Exposition geol., Congr. geol. int., 3rd. session, Berlin, 52-56. schouppe, A. von. 1958. Revision des Formenkreises um Phillipsastraea d’Orb., ‘ P achy phy Hum' E. & H., Macgeea (Webst.), ‘ Thamnophyllum ’ Pen., Peneckiella Soshk. und verwandter Formen. Abh. Neues Jb. Geol. Palaont. 106, 139-244, pi. 5-6. semenoff-tian-chansky, p., lafuste, j., and delga, m. d. 1961. Madreporaires du Devonien du Chenoua (Algerie). Bull. Soc. geol. France, ser. 7, 3, 290-319, pi. 9. simpson, g. b. 1900. Preliminary Descriptions of New Genera of Paleozoic Rugose Corals. Bull. N.Y. St. Mus. 8, 199-222. smith, s. 1945. Upper Devonian Corals of the Mackenzie River Region, Canada. Special Pap. Geol. Soc. Amer. 59, i-viii, 1-126, pi. 1-35. soshkina, E. D. 1939. Verkhnedevonskie korally Rugosa Urala. Trudy Paleont. Inst. Akad. Nauk SSSR, 9 (2), 1-88, pi. 1-14. (In Russian.) 1949. Devonskie korally Rugosa Urala. Ibid. 15, 1-160, pi. 1-58. (In Russian.) 1951. Pozdnedevonskie korally Rugosa, ikh sistematika i evolyutsiya. Ibid. 34, 1-1 18, pi. 1-24. (In Russian.) 1952. OpredeliteT devonskikh chetyrekhluchevykh korallov. Ibid. 39, 1-127, pi. 1-49. (In Russian; French transl. S. Ketchian, C.E.D.P., 1249.) 1954. Devonskie chetyrekhluchevye korally russkoy platformy. Ibid. 52, 1-76, pi. 1-19. (In Russian.) stoll, N. r. and others (eds.). 1961. International Code of Zoological Nomenclature adopted by the XV International Congress of Zoology, London, July 1958. Int. Trust for Zool. Nomenclature, London. strusz, d. l. 1963. Studies in the Palaeontology, Petrography and Stratigraphy of the Garra Beds. Unpublished Ph.D. thesis, Univ. Sydney. — 1964. Devonian Trilobites from the Wellington-Molong District of New South Wales. J. Proc. Roy. Soc. N.S.W. 97, 91-97 pi. 1. — 1965. A Note on the Stratigraphy of the Devonian Garra Beds of N.S.W. Ibid, (in press). stumm, e. c. 1948. Lower Middle Devonian Species of the Tetracoral Genus Hexagonaria of East- Central North America. Contr. Mus. Paleont. Univ. Michigan, 7, 7-49, pi. 1-14. D. L. STRUSZ: DISPHYLLIDAE AND PHACELLOPHYLLIDAE 571 stumm, e. c. 1949. Revision of the Families and Genera of the Devonian Tetracorals. Mem. Geol. Soc. Amer. 40, i-viii+ 1-92, pi. 1-25. sun, y. c. 1958. The Upper Devonian Coral Faunas of Hunan. Palaeont. Sinica (144), N.s. B, 8, 1-28, pi. 1-12. teichert, c. 1948. Middle Devonian Goniatites from the Buchan District, Victoria. /. Paleont. 22, 60-67, pi. 16. wang, H. c. 1950. A Revision of the Zoantharia Rugosa in the Light of their Minute Skeletal Struc- tures. Phil. Trans. Roy. Soc. London, ser. B, 234, 175-246. wedekind, R. 1922. Zur Kenntnis der Stringophyllen des oberen Mitteldevon. Sitz.-ber. Ges. Beforde- rung Ges. Naturw. Marburg, 1 (1921), 1-16. yoh, s. s. 1937. Die Korallenfauna des Mitteldevons aus der Provinz Kwangsi, Siidchina. Palaeonto- graphica, 87, A, 45-76, pi. 4-1 1 . D. L. STRUSZ Department of Geology and Geophysics, University of Sydney, Sydney, N.S.W. Present address : Department of Geology, University College of Townsville, Manuscript received 17 September 1964 Townsville, Queensland. OERTLIANA, A NEW OSTRACOD GENUS FROM THE UPPER JURASSIC OF NORTH-WEST EUROPE by T. I. KILENYI Abstract. A new genus of the family Cytherideidae is established under the name Oert liana; four species of the genus are described and their stratigraphical range in north-west Europe is examined. In 1955 G. Schmidt, describing the Upper Jurassic ostracods of north-west Germany, mentioned an ovoid-shaped ostracod with a rather simple hinge structure as Ostracod 102. In 1957 Oertli found a similar form in the Lower Kimmeridgian of the Paris Basin and recognized it as a hitherto unknown genus of the family Cytherideidae. The preserva- tion of his specimens, however, made it impossible to establish a new genus as most of the internal features were unknown. The study of the ostracods of the Kimmeridge Clay type section in Dorset yielded specimens of this form in excellent preservation and the establishment of the genus became possible. So far four species of the genus Oert- liana are known from the Upper Jurassic of north-west Europe; O. kimmeridgensis occurs in the Lower Kimmeridgian of Dorset and the Paris Basin, Oertliana sp. 1. was found in the Upper Kimmeridgian of Dorset. Of the remaining two species O. sp. 2 occurs in the Kimmeridgian of north-west Germany and O. sp. 3 in the Purbeck of southern Sweden. Acknowledgements. I wish to acknowledge the kindness and help of Dr. J. W. Neale (University of Hull) and Dr. H. J. Oertli (S.N.P.A., Pau, France) in the preparation of this paper and for the loan of specimens. Abbreviations. In giving the dimensions of ostracod valves the following abbreviations are used throughout the text : L, length ; H, height ; W, width ; Hi, hinge length ; Mja, width of anterior margin. Dimensions are given in millimetres. Repository. All the figured and described specimens are stored in the Geology Department of the University of Hull. The specimen numbers are indicated by the prefix hu. SYSTEMATIC DESCRIPTIONS Subclass ostracoda Latreille 1806 Order podocopida Muller 1894 Suborder podocopina Sars 1866 Superfamily cytheracea Baird 1850 Family cytherideidae Sars 1925 Subfamily unknown Genus oertliana gen. nov. Type species. Oertliana kimmeridgensis sp. nov. Diagnosis. Carapace almost elliptical in side view, anterior end slightly more rounded than posterior. Surface smooth or finely punctate. Hinge adont, consisting of a faint [Palaeontology, Vol. 8, Part 3, 1965, pp. 572-6, pi. 79.] T. I. KILENYI: OERTLIANA, A NEW OSTRACOD GENUS 573 curved ridge on the right valve and a corresponding groove on the left. Muscle scar pattern consists of a vertical row of four equally sized scars with one anterior scar that is larger than the other four. Stratigraphical range: Upper Jurassic, Kimmeridgian- Purbeckian. Oertliana kimmeridgensis sp. nov. Plate 79, figs. 1-12; text-fig. 1 1957 Cytherideinarum ? sp. 1. Oertli, p. 661, pi. 3, figs. 86-91. Holotype. A female left valve, hu 2.J.1.31. Paratypes. 100 valves and carapaces, hu 3.J.29.1-100. Derivatio nominis. From typical horizon. Stratum typicum. Rasenia mutabilis Zone, Lower Kimmeridgian. Locus typicus. Black Head, Dorset. Diagnosis. Carapace elliptical, anterior end rounded, posterior slightly pointed. Valves subequal, left being slightly larger. Surface of valves finely punctate. Radial pore canals relatively thick in the middle, narrowing suddenly near both ends. Sexual dimorphism strong. Measurements. L H W Hi Mia Holotype 0-38 0-21 0-08 0-20 0-03 $ Left valve 0-36-0-38 0-20-0-21 0-08 0-20 0-03 $ Right valve 0-35-0-37 0-1 8-0-20 0-07 0-20 0-03 3 Left valve 0-43-0-45 0-21-0-23 0-08 0-25 0-04 3 Right valve 0-43-0-45 0-20-0-22 0-07 0-25 0-03 Description. Carapace elongate, elliptical. Left valve slightly larger than the right, with a slight overlap ventrally and dorsally. In dorsal view the carapace is lanceolate, the greatest width being at the middle in the females and somewhat nearer to the posterior end in the males. In side view the two valves are similar, although the right one is less high. The greatest height is at mid-point. The anterior end is rounded, the ventral and dorsal margins straight, running approximately parallel. The posterior end is slightly angular, more so in the right valve, where the postero-dorsal margin is straight and not rounded as in the left valve. Sexual dimorphism is pronounced, the males being much longer than the females. Their shape is essentially the same, but often the ventral margin of the male valve is slightly concave in the middle. The surface of the valve is finely punctate. This ornamentation is strongest on the central part of the valve and disappears gradually towards the peripheral regions. It is connected with the finer structure of the shell, which in transmitted light seems to be perforated by irregularly spaced cavities. The size of these cavities decreases towards the periphery of the valve. The normal pore canals are few. The duplicature is bilamellar. The inner lamella is moderately wide on the anterior, and narrow on the posterior part of the valve. Inner margin and line of concrescence coincide, the selvage is faint, but is better developed on the ventral parts, where it forms a fairly wide selvage lip. The radial pore canals are straight and simple. They are widest at their middle, narrowing suddenly at the ends. They number between ten and fifteen on the anterior margin. The hinge is adont and rather weakly developed. It consists of a curved ridge on the right valve and a corresponding shallow groove on the opposite valve. 574 PALAEONTOLOGY, VOLUME 8 The muscle scar pattern consists of a slightly oblique row of four circular and equally sized scars. There seems to be only one anterior scar, which is larger than the others and lies in line with the second and third scars in the row. Occurrence. R. mutabilis Zone, Lower Kimmeridgian, in Dorset and from the ‘ Lower Kimmeridgian ’ of the Paris Basin. Remarks. The genus Oertliana does not seem to fit into any of the subfamilies of the family Cytherideidae, due to its primitive adont hinge. The muscle scars, however, are without doubt typical of the Cytherideidea. text-fig. 1 . Oertliana kimmeridgensis gen. et sp. nov. a. Anterior margin, male left valve, x 200. b, Hinge margin, left valve, x 120. c. Hinge margin, right valve, x 120. d. Muscle scars, left valve, x 375. Oertliana sp. 1 Plate 79, figs. 13-16 Material. 39 valves, hu 2.J.34.1-39. Measurements. L H Hi Left valve 0-45 0-24 0-38 Right valve 0-44 0-23 0-38 EXPLANATION OF PLATE 79 Figs. 1-12. Oertliana kimmeridgensis sp. nov., Rasenia mutabilis Zone, Lower Kimmeridgian. I, Right valve, female, external view, hu 3.J.29.2, x 100. 2, Left valve, female, external view, Holotype, hu 2.J.31.1, x 100. 3, Right valve, male, external view, hu 3.J.29.6, x 100. 4, Left valve, male, external view, hu 3.J.29.7, x 100. 5, Right valve, female, external view, hu 3.J.29.3, x 100. 6, Left valve, female, external view, hu 3 J.29.4, x 100. 7, Right valve, male, external view, hu 3.J.29.17, x 100. 8, Left valve, male, external view, hu 3J.29.2, x 100. 9, Right valve, male, transmitted light, hu 3.J.29.15, X110. 10, Left valve, male, polarized light, hu 3.J.29.16, xllO. II, Carapace, male, dorsal view, hu 3.J.29.9-10, x 100. 12, Carapace, female, dorsal view, hu 3.J.29.1 1-12, x 100. Figs. 1 3-1 6. Oertliana sp. 1 , Pavlovia rotunda Zone, Upper Kimmeridgian. 13, Right valve, female?, external view, hu 2.J.34.3, x 100. 14, Left valve, female?, external view, hu 2.J.34.2, x 100. 15, Right valve, male?, external view, hu 2.J.34.1, x 100. 16, Left valve, male?, external view, hu 2.J.34.3, x 100. Palaeontology, Vol. 8 PLATE 79 KILENYI, Upper Jurassic ostracod T. I. KILENYI: OERTLIANA, A NEW OSTRACOD GENUS 575 Description. Carapace elongate, almost oblong shaped. The two valves are equal in size. Anterior end rounded, posterior slightly pointed. Dorsal margin straight and parallel to ventral margin, which is straight or gently concave. The surface of the valve is very finely punctate, almost smooth. There is no eye depression. Little of the interior of the valve could be observed because of the bad state of preser- vation. The inner lamella seems to be rather broad, and the inner margin and line of OERTLIANA KIMMERIDGENSIS OERTLIANA sp.l. OERTLIANA sp.2. OERTLIANA sp3. PURBECK ? i PORTLAND UPPER KIMMERIDGE MIDDLE KIMMERIDGE LOWER KIMMERIDGE 1 1 1 text-fig. 2. The stratigraphical distribution of species of Oertliana gen. nov. in north-west Europe. concrescence probably coincide. The selvage lip is developed on both valves. The hinge is adont; a ridge on the right valve and a corresponding groove on the left. Sexual dimorphism is doubtful. Occurrence. Pavlovia rotunda Zone, Upper Kimmeridgian, Chapmans Pool, Dorset. Remarks. The muscle scars could not be observed owing to the rather bad preservation, but similarity in shape, hinge, and duplicature with O. kimmeridgensis makes it certain that this is a species of Oertliana. Oertliana sp. 2 1955 Ostracod 102, Schmidt, p. 62, pi. 3 a, figs. 39, 40. Measurements. L, 0-45 ; H, 0-24. 576 PALAEONTOLOGY, VOLUME 8 Remarks. Ovoid carapace with very finely punctate surface. The left valve is larger than the right. Hinge adont. Occurrence. Oert liana sp. 2 was described by Schmidt (1955) from boreholes in north- west Germany ranging from the ‘Middle Kimmeridgian ’ to the top of the ‘Gigas Schichten almost the whole of the Kimmeridgian (in the English sense). It occurs both in marine and in brackish sediments. Oertliana sp. 2 closely resembles O. kimmeridgensis sp. nov., but in side view it appears much more rounded at the cardinal angles. Sexual dimorphism, which is strong in O. kimmeridgensis, seems to be absent. Oertliana sp. 3 1961 Cytherideinarum sp. 2, Oertli, Brotzen, and Bartenstein, p. 13, pi. 1, figs. 3 a, b. Measurements. L, 0-45-0-47 ; H, 0-26-0-28. Remarks. Bean-shaped carapace in side view, surface distinctly punctate, anterior and posterior end well rounded. Occurrence. From a borehole at Landskrona (Hilleshog), southern Sweden. The horizon is given by Oertli, Brotzen, and Bartenstein (1961) as Wealden. According to their correlation this is the equivalent horizon to the middle Purbeck in English stratigraphy. The environment is marine-brackish. As the material is rather poorly preserved only the external characteristics are known. Oertliana sp. 3 differs from the other species of the genus in its much higher shell. REFERENCES klinger, w. 1955. Mikrofaunistische und stratigraphisch-fazielle Untersuchungen im Kimmeridge und Portland des Weser-Aller-Gebietes. Geol. Jb. 70, 167-246. moore, R. c. (ed.) 1961. Treatise on Invertebrate Paleontology. Part Q. Arthropoda 3. Geol. Soc. Amer. and University of Kansas Press. oertli, h. J. 1957. Ostracodes du Jurassique superieur du Bassin de Paris (Sondage Vernon 1). Revue Inst. /rang. Petr. 12, fasc. 6. 647-95. brotzen, f., and bartenstein, h. 1961. Mikropalaontologisch-FeinstratigraphischeUntersuchung der Jura-Kreide-Grenzschichten in Siidschweden. Arsb. Sverig. Geol. Unders. 55, no. 3. schmidt, g. 1955. Stratigraphie und Mikrofauna des Mittleren Malm im nordwest-deutschen Berg- land. Abb. senckenb. naturf. Ges. 491, 1-76. T. I. KILENYI Department of Geology, Sir John Cass College, Jewry Street, London, E.C. 3 Manuscript received 14 October 1964 THE PALAEONTOLOGICAL ASSOCIATION COUNCIL 1965-6 President Dr. L. R. Cox, British Museum (Natural History), London Vice-Presidents Dr. W. S. McKerrow, University Museum, Oxford Professor F. H. T, Rhodes, University College, Swansea Treasurer Dr. C. Downie, Department of Geology, The University, Mappin Street, Sheffield 1 Secretary Dr. C. H. Holland, Department of Geology, Bedford College, London, N.W.l Editors Mr. N. F. Hughes, Sedgwick Museum, Cambridge Dr. Gwyn Thomas, Department of Geology, Imperial College of Science, London, S.W.7 Dr. I. Strachan, Department of Geology, The University, Birmingham, 15 Dr. M. R. House, University Museum, Oxford Other members of Council Dr. C. G. Adams, British Museum (Natural History), London Professor P. M. Butler, Royal Holloway College, Surrey Dr. W. J. Clarke, British Petroleum Company, Sunbury on Thames Dr. G. Y. Craig, The University, Edinburgh Dr. T. D. Ford, The University, Leicester Dr. B. M. Funnell, The University, Cambridge Dr. J. M. Hancock, King’s College, London Dr. G. A. L. Johnson, The University, Durham Dr. F. A. Middlemiss, Queen Mary College, London Dr. M. Mitchell, Geological Survey and Museum, London Dr. W. D. I. Rolfe, The University, Glasgow Dr. A. J. Rowell, The University, Nottingham Professor Scott Simpson, The University, Exeter Dr. L. B. Tarlo, The University, Reading Dr. H. Dighton Thomas, British Museum (Natural History), London Overseas Representatives Australia: Professor Dorothy Hill, Department of Geology, University of Queensland, Brisbane Canada : Dr. D. J. McLaren, Geological Survey of Canada, Department of Mines and Technical Surveys, Ottawa, Ont. India: Professor M. R. Sahni, 98, The Mall, Lucknow (U.P.), India New Zealand: Dr. C. A. Fleming, New Zealand Geological Survey, P.O. Box 368, Lower Hutt West Indies and Central America: Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad, Inc., Point k Pierrie, Trinidad, West Indies Eastern U.S.A. : Professor H. B. Whittington, Museum of Comparative Zoology, Harvard Univer- sity, Cambridge 38, Mass. Western U.S.A. : Professor J. Wyatt Durham, Department of Paleontology, University of California, Berkeley 4, Calif. PALAEONTOLOGY VOLUME 8 ' PART 3 CONTENTS Ostracoda from the Sutterby Marl (U. Aptian) of south Lincolnshire. By p. KAYE and d. barker 375 Upper Jurassic and Lower Cretaceous microfossils from the Hautes-Alpes. By JUDITH TURNER 391 Origin, limits, and systematic position of Scaphites. By j. wiedmann 397 Isorthis and Salopina (Brachiopoda) in the Ludlovian of the Welsh border- land. By v. G. WALMSLEY 454 A new fenestrate bryozoan from the Lower Carboniferous of County Fermanagh. By r. tavener-smith 478 Dionella gen. nov. (Superfamily Membraniporacea) from the Upper Creta- ceous of Europe. By A. w. medd 492 Disphyllidae and Phacellophyllidae from the Devonian Garra Formation of New South Wales. By d. l. strusz 518 Oertliana, a new ostracod genus from the Upper Jurassic of North-West Europe. By t. i. kilenyi 572 PRINTED IN GREAT BRITAIN AT THE UNIVERSITY PRESS, OXFORD BY VIVIAN RIDLER, PRINTER TO THE UNIVERSITY VOLUME 8 • PART 4 Palaeontology DECEMBER 1965 PUBLISHED BY THE PALAEONTOLOGICAL ASSOCIATION LONDON Price £3 THE PALAEONTOLOGICAL ASSOCIATION The Association was founded in 1957 to further the study of palaeontology. It holds meetings and demonstrations, and publishes the quarterly journal Palaeontology. Membership is open to individuals, institutions, libraries, &c., on payment of the appropriate annual subscription: Institute membership £5. 5s. (U.S. $15.50) Ordinary membership £3. 3s. (U.S. $9.50) Student membership £2. 2s. (U.S. $6.50) There is no admission fee. Student members will be regarded as persons receiving full-time instruction at educational institutions recognized by the Council; on first applying for membership, they should obtain an application form from the Secretary or the Treasurer. Subscriptions are due each January, and should be sent to the Treasurer, Dr. C. Downie, Department of Geology, The University, Mappin Street, Sheffield 1, England. Palaeontology is devoted to the publication of papers (preferably illustrated) on all aspects of palaeontology and stratigraphical palaeontology. Four parts are published each year and are sent free to all members of the Association. Members who join for 1965 will receive Volume 8, Parts 1 to 4. All back numbers are still in print and may be ordered from B. H. Blackwell, Broad Street, Oxford, England, at the prices shown below (post free) : Vol. 1 (for 1957-8) in 4 parts at £2 or U.S. $6.00 per part. Vol. 2 (for 1959) in 2 parts at £2 or U.S. $6.00 per part. Vol. 3 (for 1960) in 4 parts at £2 or U.S. $6.00 per part. Vol. 4 (for 1961) in 4 parts at £2 or U.S. $6.00 per part. Vol. 5 (for 1962) in 4 parts at £3 or U.S. $9.00 per part. Vol. 6 (for 1963) in 4 parts at £3 or U.S. $9.00 per part. Vol. 7 (for 1964) in 4 parts at £3 or U.S. $9.00 per part. A complete set. Volumes 1-7, consists of 26 parts and costs £64 or U.S. $192. Manuscripts on all aspects of palaeontology and stratigraphical palaeontology are invited. They should conform in style to those already published in this journal, and should be sent to Mr. N. F. Hughes, Department of Geology, Sedgwick Museum, Downing Street, Cambridge, England. A sheet of detailed instructions for authors will be supplied on request. © The Palaeontological Association, 1965 TRILOBITES FROM THE ALBANY DIVISION (ORDOVICIAN) OF THE GIRVAN DISTRICT, AYRSHIRE by RONALD PEARSON TRIPP Abstract. Two new genera, and nine new species of trilobites from the Albany mudstones with nodular lime- stones, east of Doularg, near Girvan, are described. The fauna is most closely allied to that of the superstes Mudstones, Aldons, having twenty-two of twenty-five genera in common, and fourteen species closely related. 0 Outside the district the closest resemblance is to the lower Edinburg Formation (Porterfield Stage) of the Appalachian Valley of the U.S.A. The trilobites are from the Albany mudstones with nodular limestones, exposed in a stream section, 950 yards east-north-east of Doularg Farm, Stinchar Valley, Girvan (Nat. Grid Ref. NX 269929). Correlation with the mixed shelly-graptolitic facies of the Lower Barr Series indicates that the fossiliferous member of the Albany mudstones is basal Caradoc (within the Nemagraptus gracilis zone): but the trilobite fauna per se is best compared with those typical of the Porterfield Stage of the standard Ordovician for NE. America. The exposure was discovered by Professor Alwyn Williams in the course of his field work in the district, and has been described by him (Williams 1962, pp. 45-47). I am greatly indebted to Professor Alwyn Williams for showing me the locality. Sincere thanks are also due to Dr. John Temple and Professor H. B. Whittington for help in the preparation of this paper. The terminology is essentially that adopted in the Treatise on Invertebrate Paleon- tology, Part O. Almost all the specimens are preserved as internal and external moulds. All the specimens were collected by the author ; the type and figured specimens have been presented to the Hunterian Museum, Glasgow University. The following is a list of the species recorded in this paper: Page Figures Trinodus doulargensis sp. nov. 578 Plate 80, figs. 1-4 Trinodus sp. 579 Plate 80, figs. 5-7 Carrickia pelagia gen. et sp. nov. 580 Plate 81, figs. 17-21 Remopleurides sp. A 581 Plate 80, figs. 8, 9, 17 Remopleurides sp. B 581 Plate 80, figs. 10, 1 1 Remopleurides sp. C 582 Plate 80, figs. 12-15 Hypodicranotus sp. 582 Plate 80, fig. 16 Isotelus sp. 583 Plate 80, figs. 18, 19 Nileus sp. 583 Plate 80, figs. 20, 21 Raymondaspis sp. 583 Plate 80, figs. 22 a, b Illaenus sp. 584 Plate 80, figs. 23-25 Bumastoides scoticus sp. nov. 585 Plate 80, figs. 26-32 Hibbertia whittingtoni sp. nov. 586 Plate 81, figs. 1-3 Otarion sp. 586 Dimeropyge hystrix sp. nov. 587 Plate 81, figs. 4-10 Toernquistia sp. A 588 Plate 81, figs. 11a, b [Palaeontology, Vol. 8, Part 4, 1965, pp. 577-603, pis. 80-83.] B 6612 Q q 578 PALAEONTOLOGY, VOLUME 8 Page Figures Toernquistia sp. B 588 Plate 81, fig. 12 Mesotaphraspis sp. 589 Plate 81, figs. 13, 14 Ampyx sp. 589 Plate 81, figs. 22, 23 Lonchodomas pernix sp. nov. 590 Plate 82, figs. 1-7 Ceraurinella sp. 591 Plate 82, fig. 8 Sphaerexochus sp. 591 Plate 82, figs. 10-17 Sphaerocoryphe sp. 592 Plate 82, fig. 9 Encrinuroides obesus sp. nov. 593 Plate 82, figs. 18-28 Encrinuroides sp. 594 Plate 82, figs. 29-31 Cybele ? sp. 595 Plate 83, fig. 1 Quinquecosta williamsi gen. et sp. nov. 595 Plate 83, figs. 2-10 Encrinuridae indet. 598 Plate 82, fig. 32 Plate 83, fig. 11 Amphilichas priscus sp. nov. 598 Plate 83, figs. 12-16 Amphilichas sp. 599 Plate 83, figs. 17-19 Hemiarges sp. 600 Plate 83, fig. 20 Ceratocephala sp. 600 Plate 83, fig. 21 Unassigned free cheek 600 Plate 81, fig. 15 Unassigned hypostome 601 Plate 81, fig. 16 SYSTEMATIC DESCRIPTIONS Family geragnostidae Howell 1935 Genus trinodus M’Coy 1846 Trinodus doulargensis sp. nov. Plate 80, figs. 1-4 Diagnosis. Convexity moderately strong. Glabella two-thirds length of cephalon, parallel sided for most of its length; glabellar tubercle and furrows lacking. Posterior borders produced backwards into short spines. Surface smooth. Holotype. A. 5830a, b (cephalon). Plate 80, figs. 1 a, b. Other material. One cephalon, one thoracic segment, three pygidia. Dimensions of holotype (in mm.). Length of cephalon . . . 2-2 Maximum width of cephalon (est.) . 2T Length of glabella . . . 1-5 Maximum width of glabella . . 0-9 Description. Cephalon narrowly rounded, as wide as long. Glabella parallel sided for most of its length, two-thirds length of cephalon, moderately convex; glabellar tubercle and furrows lacking. Basal lobes small, triangular, not touching mesially. Axial and pre- glabellar furrows deep and broad. Cheeks convex, uniformly narrow. Border moderately wide and convex, narrowing out towards back. Border furrow well defined. Posterior borders swollen, produced backwards into short triangular genal spines. Surface smooth. Thoracic segment with axis occupying three-quarters total width. Median lobe twice as wide posteriorly as anteriorly. Oval lateral lobes approximately one-third width of axis, their long axes inclined inwards and forwards, strongly demarcated. Axial furrows deep and broad. Pleurae narrow (tr.), longitudinally truncate, weakly swollen; anterior and posterior lobes scarcely developed. R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 579 Pygidium subquadrate. Articulating half-ring and pleural facets well developed. Axis slightly less than half length of pygidium. First ring short, with small, oval lateral lobes. Second ring long; elogate median lobe rising steeply towards back to form a rounded node. Terminal piece about one-third length of axis. Axial furrows deep and broad. Pleural lobes convex, narrowing towards front. Border weakly convex, extremely wide posteriorly particularly opposite the pair of small lateral spines. Border furrow well defined. Surface smooth. Remarks. T. doulargensis bears a close resemblance to T. elspethi (Raymond) (see Cooper 1953, pp. 7-8, pi. 1, figs. 1-12) from the Edinburg Formation, in all parts. The glabella is more parallel sided and lacks the median tubercle, the pygidium has a broader axis and border, and the dorsal exoskeleton is smooth, not granular. Trinodus sp. Plate 80, figs. 5-7 Material. Two cephala, two pygidia. Description. Glabella moderately long, narrowing slowly forwards ; glabellar tubercle and furrows lacking. Basal lobes small, sloping steeply backwards. Axial and preglabellar furrows moderately deep. Border narrows out towards back. Posterior borders strongly swollen, widening (exs.) laterally, not produced backwards as spines. Surface smooth. Axis slightly less than half length of pygidium, strongly swollen. Two rings, subequal in length, occupy slightly more than half length of axis. First ring with large, well-defined lateral lobes. Elongate median tubercle on second ring rises towards back. Axial furrows moderately deep. Pleural lobes of uniform width. Border wide posteriorly. A pair of oblique ridges cross border at postero-lateral angles, produced into short, broad-based spines. Surface smooth. Remarks. The cranidium figured lacks glabellar tubercles and furrows and appears to be referable to Trinodus ; it differs from the foregoing species in the shape of the glabella and absence of fixigenal spines. The long axis of the pygidum suggests reference to Geragnostus rather than Trinodus. However, it may be closely compared with Trinodus sp. from the Ashgillian of Poland (Kielan 1959, p. 62, pi. 1, fig. 5, text-fig. 14). Family komaspididae Kobayashi 1935 Genus carrickia gen. nov. Diagnosis. Glabella subquadrate, weakly convex; lateral glabellar furrows absent. Preglabellar field short. Anterior border wide (tr.). Palpebral lobes narrow, depressed, two-thirds length of glabella, reaching almost to back of cheeks. Type species. Carrickia pelagia sp. nov. Remarks. So far as is known the genus is monotypic. The genus is referred to the Family Komaspididae on account of the long palpebral lobes, wide fixed cheeks, and short preglabellar field. Both cranidium and pygidium bear a general resemblance to Gonio- phrys prima Ross (1951, pp. 81-82, pi. 18, figs. 9, 15, 17-20, 22, 27) from the Garden City Formation, but the new form differs conspicuously in the broader, less convex, glabella, and transversely wider anterior border. In the aforementioned features, Carrickia bears a resemblance to the dimeropygid Chomatopyge Whittington and Evitt (1954, pp. 49-53) 580 PALAEONTOLOGY, VOLUME 8 from the Lincolnshire and Edinburg Limestones, but the length of the palpebral lobes, short preglabellar field, absence of median preglabellar pit, and the conspicuously distinctive pygidium preclude affiliation with this genus. Carrickia pelagia gen. et sp. nov. Plate 81, figs. 17-21 Diagnosis. As for genus. Holotype. A. 5878 (cranidium). Plate 81, figs, \la-c. Paratype. A. 5882 (pygidium). Other material. Fourteen cranidia, nine pygidia. Material from other horizons. Cranidia from the platy upper Stinchar Limestone, and from the super stes Mudstones, Aldons. Dimensions (in mm.). Holotype A. 5880 Length of cranidium (sag.) 3-4 2-3 Length of glabella . . 2-4 1-7 Width of glabella . . 3-2 2-0 EXPLANATION OF PLATE 80 All the figured specimens are from the Albany mudstones with nodular limestones, 950 yards east- north-east of Doularg Farm, Stinchar Valley, Girvan; they have been deposited in the Flunterian Museum, Glasgow. The photographs are of internal moulds unless otherwise stated; the specimens were coated with ammonium chloride before being photographed. Figs. 1-4. Trinodus doulargensis sp. nov. 1 a, b, Cephalon (holotype A. 58306). Latex cast from external mould. Dorsal and oblique lateral views, x 10. 2, Thoracic segment (A. 5831). Latex cast from external mould, xlO. 3, Pygidium (A. 5832). Latex cast from external mould, x8. 4, Pygidium (A. 58336). External mould, x 8. Figs. 5-7. Trinodus sp. 5, Cranidium (A. 58346). External mould, x 8. 6, Pygidium (A. 5835a), x 10. 7, Pygidium (A. 58366). External mould, x 10. Figs. 8,9, 17. Remopleurides sp. A. 8, Left free cheek (A. 5837). Lateral view showing doublure, x4. 9, Right free cheek (A. 5838a), showing vincular ledge and ridge on doublure, x4. 17, Hypostome (A. 58396). External mould, x 8. Figs. 10, 11. Remopleurides sp. B. 10, Left free cheek (A. 5840) x5. 11, Right free cheek (A. 5841). External mould showing raised lines, x 10. Figs. 12-15. Remopleurides sp. C. 12, Small cranidium (A. 5842), x 8. 13, Right free cheek (A. 5843), x6. 14, Hypostome (A. 5844), X8. 15, Pygidium (A. 5845), x8. Fig. 16. Hypodicranotus sp. Left free cheek (A. 58466). Latex cast from external mould, x 8. Figs. 18, 19. Isotelus sp. 18, Cranidium (A. 5847). Oblique lateral view, showing elevated palpebral lobe, x 1. 19, Pygidium (A. 5848), with doublure exposed, x 1. Figs. 20, 21. Nileus sp. 20, Left free cheek (A. 5849), x 5. 21, Left free cheek (5850a) belonging to an individual with a shorter cephalon, x 5. Figs. 22 a, 6. Raymondaspis sp. 22 a, Cranidium (A. 5851a), X4. 226, The same, external mould (A. 58516), showing raised lines on glabella, and right basal lateral furrows, x8. Figs. 23-25. Illaenus sp. 23 a-c, Cranidium (A. 5852a). Dorsal, lateral, and frontal views, x 4. 24, Left free cheek (A. 5853). External mould, x 10. 25, Pygidium (A. 5854a), x 5. Figs. 26-32. Bumastoides scoticus sp. nov. 26 a, 6, Cranidium (A. 5855a). Dorsal and lateral views, x3. 27, Small cranidium (A. 5856), x6. 28, Right free cheek (A. 5857), x6. 29, Rostral plate (A. 58586). External mould, x 5. 30, Hypostome (A. 5859), probably belonging to this species. External mould. Oblique lateral view showing left anterior wing with longitudinal raised lines, x 10. 31a, Pygidium (holotype, A. 5860a). Dorsal view, X7. 316, The same. Lateral view, x5. 32, Elongate pygidium (A. 5861a), x4. Palaeontology, Vol. 8 PLATE 80 TRIPP, Girvan trilobites R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 581 Description. Cranidium broadly rounded anteriorly, moderately convex in both direc- tions. Glabella subquadrate, broadly rounded anteriorly, moderately convex. Lateral glabellar lobes and furrows absent. Occipital ring little longer mesially than laterally. Occipital furrow transverse but sinuous, deep, narrower than axial furrows. Preglabellar and axial furrows continuous, moderately deep and narrow. Preglabellar field short laterally, usually narrowing out mesially, weakly convex, continuous with fixed cheeks. Anterior border much wider (tr.) than glabella, of uniform length, rolled. Anterior border furrow broad. Fixed cheeks gentle convex, widening steadily towards back, slop- ing away from glabella. Palpebral lobes narrow, depressed, weakly rounded in outline, two-thirds length of glabella, reaching almost to back of fixed cheeks, posterior extremi- ties much further apart than anterior extremities. Palpebral furrows broad, with outer wall lower than inner. Posterior borders convex, expanding little laterally. Posterior border furrows deep and broad, widening laterally. Anterior branches of facial sutures curve forwards and slightly inwards ; posterior branches short, directed backwards and outwards. Surface smooth; glabella, occipital ring, and inner areas of fixed cheeks shal- lowly pitted. Pygidium elliptical in outline, strongly convex in both directions. Axis occupies more than half anterior width, transverse curvature slightly greater than that of pygidium, broad and undefined posteriorly; articulating half ring and two rings well defined, bowed backwards, terminal piece with strong independent convexity. Axial furrows sharp, bowed outwards alongside rings and terminal piece, flaring posteriorly and dying out before reaching margin. Two pairs of pleurae; interpleural furrows sharp, extending to margin ; pleural furrows oblique, deep. Surface finely granular. Remarks. The diagnostic characters of the genus clearly distinguish this form. The py- gidia described are attributed to this species mainly on the grounds of comparable frequency, but they are appropriate to the family. Family remopleurididae Hawle and Corda 1847 Genus remopleurides Portlock 1843 Remopleurides sp. A Plate 80, figs. 8, 9, 17 Material. Eight free cheeks, three hypostomes. Remarks. These free cheeks and hypostomes closely resemble a form which occurs in the confinis Flags (Tripp 1962, pp. 4-5, pi. 1, figs. 10-14), the Stinchar Limestone, and the superstes Mudstones. The free cheeks differ in lacking the posterior border furrow, in their non-striate surface, and in the presence of an oblique ridge on the doublure, tan- gential to the vincular ledge which is situated further back (PL 80, fig. 9). The hypostome is narrower posteriorly, and the anterior boss is more prominent. Remopleurides sp. B. Plate 80, figs. 10, 1 1 Material. Two free cheeks, one pygidium. Description. Free cheek moderately wide posteriorly; librigenal spine straight, slender, 582 PALAEONTOLOGY, VOLUME 8 moderately long ; subgenal notch extends half-way across border. Surface with moderately spaced raised lines. Pygidium elongate, with axis defined only by independent convexity. Pleurae fused except for two pairs of free points on posterior margin, outer pair extending slightly further backwards than inner pair. Remarks. An allied species occurs in the superstes Mudstones, Aldons. The affinities of these two forms are with R. caelatus Whittington (1959, pp. 401-11, pis. 1-3; pi. 4, figs. 1-25; text-figs. 4, 5) from the Edinburg Formation, and with R. plaesiourus Whit- tington (1959, pp. 412-14, pi. 4, figs. 26-30; pi. 5; pi. 6, figs. 1-15) from the Lower Martinsburg shale, both from northern Virginia. Remopleurides sp. C Plate 80, figs. 12-15 Material. One cranidium, two free cheeks, two hypostomes, one pygidium. Description. One small, incomplete cranidium; glabella abruptly expanded, weakly convex; anterior tongue long, moderately wide, almost horizontally extended. Occipital ring long and narrow. Palpebral lobes broad (tr.). Surface apparently smooth. Eye short, broad, and strongly convex transversely, weakly rounded in outline. Ex- ternal rim of eye lobe broad and weakly rounded, depressed anteriorly where marked off by broad furrows, dying out posteriorly. Border convex anteriorly, with a steep, salient anterior extension. Border widens and merges with eye rim posteriorly, produced into a strong, straight, broad based, librigenal spine. Hypostome elongatedly trapeziform in outline, almost as long as wide. Middle body convex, well defined. Anterior boss not prominent. Lateral and posterior borders narrow; postero-lateral forks absent. Pygidium subquadrate. Axis swollen, composed of two segments. Anterior pleurae narrowly pointed, extending backward as far as second pair. Space between inner pair of pleurae long and narrow. Doublure broad, curling upwards along inner margin. Remarks. The short eye is a commanding feature in both cranidium and free cheek. All the parts described may not belong to one species. Genus hypodicranotus Whittington 1952 Hypodicranotus sp. Plate 80, fig. 16 Material. One free cheek. Description. Free cheek narrow anteriorly, lateral outline almost straight. External rim of eye lobe marked off by sharp furrows. Cheek weakly convex. Lateral and posterior border furrows lacking. Genal angle acute. Subgenal spine narrow, directed almost straight backwards, extending a short way beyond posterior margin. Subgenal notch narrow, tapering slowly, more than half length of subgenal spine. Surface of posterior area and subgenal spine with faint, anastomosing, subconcentric terrace lines, widely spaced. R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 583 Remarks. The discovery of a Hypodicranotus type of hypostome in the upper platy Stin- char Limestone, Auchensoul Hill, establishes the occurrence of the genus in the Girvan District at a considerably earlier date than any of the North American records. The single free cheek from the Albany Mudstones differs from both H. striatulus (Walcott) (see Whittington 1952, pi. 1, figs. 1-6, 8, 10) and from H. missouriensis (Foerste) (see Bradley 1930, p. 30, figs. 6-8) mainly in the shorter (exs.) subgenal notch, more slender subgenal spines, and the presence of terrace lines on the posterior area. Family asaphidae Burmeister 1843 Genus isotelus DeKay 1824 Isotelus sp. Plate 80, figs. 18, 19 Material. Two cranidia, two free cheeks, three pygidia. Description. Cranidium narrow, gently convex in both directions. Glabella ill defined, weakly expanded anteriorly. Palpebral lobes placed far back, strongly elevated, with cheeks drawn up at base. Axial furrows faint. [Free cheek with elevated eye. Vincular notches weakly developed. Surface pitted.] [Pygidium subtriangular, gently convex in both directions, unsegmented. Axis broad at front, faintly defined by broad depressions anteriorly. Anterior half pleurae swollen, marked off by broad depressions; articulating facets strong. Doublure broad; inner area broad and gently concave at apex, narrowing and becoming convexly upturned antero- laterally. Terrace lines well marked.] Remarks. The elevated eye lobes are a peculiar feature of this species. Family nileidae Angelin 1854 Genus nileus Dalman 1 827 Nileus sp. Plate 80, figs. 20, 21 Material. One cranidium, five free cheeks. Material from other horizons. All parts from the superstes Mudstones, Aldons. Remarks. The genus Nileus has not been previously recorded from the Girvan District. The long eyes are one of the diagnostic characters of this species which will be fully described on material from Aldons in a subsequent paper. Family scutelluidae Richter and Richter 1955 Genus Raymond aspis Pribyl 1949 Raymondaspis sp. Plate 80, figs. 22a, b 1931 Bronteopsis cf. nitens (Wiman) Reed, pp. 26-27. Material. One cranidium, one hypostome. Material from other horizons. All parts common in the superstes Mudstones, Aldons. 584 PALAEONTOLOGY, VOLUME 8 Remarks. The well-marked basal lateral glabellar furrows and the raised lines on the cranidium occur also in R. brumleyi (Cooper 1953, p. 25, pi. 9, figs. 8-10) but are lacking in R. nitens (Wiman) (see Skjeseth 1955, p. 22, pi. 4, figs. 1,3); the closer affinity appears to be with the former species. Family illaenidae Hawle and Corda 1847 Genus illaenus Dalman 1824 Illaenus sp. Plate 80, figs. 23-25 Material. Five cranidia, one free cheek, four pygidia. Description. Cranidia strongly convex longitudinally. Glabella wide, short. Axial furrows shallow, bowed strongly inwards. Cheeks with slight independent convexity. Palpebral lobes long, placed far back, weakly rounded in outline. Surface smooth except for a median tubercle near base of glabella. Free cheek wide, gently convex. Lateral margin rounded, basal angle acute, but not produced into a spine. Pygidium moderately convex in both directions, broadly rounded in outline. Axis short, gently swollen, more than one-third anterior width, projecting anteriorly, defined by broad, shallow furrows. Doublure short, monocuspid. Remarks. There is a similarity between this species and I. devexus Tripp (1962, p. 12, pi. 2, figs. 10 a, b ) in convexity, and in the proportions of the glabella, but the cranidium is shorter in the Doularg form, and the glabella does not widen anterior to the eyes. EXPLANATION OF PLATE 81 Figs. 1-3. Hibbertia whittingtoni sp. nov. 1 a-c, Cephalon without upper lamella (holotype, A. 5862 a). Dorsal, frontal, and lateral views, x6. 2, Cephalon (A. 5863a), X4-5. 3, Fragmentary cephalon (A. 5864). External mould, showing pits on glabella, xl5. Figs. 4-10. Dimeropyge hystrix sp. nov. 4 a, Cranidium (holotype, A. 5865a). Dorsal view, X6. 4 b, c. The same. Anterior and lateral views, x 8. 5, Cranidium (A. 5866), with uniformly tuberculate occipital ring, x 6. 6, Cranidium (A. 5867a), with a pair of exceptionally large swellings at inner extremities of posterior borders, x 8. 7, Cranidium (A. 58686). Latex cast from external mould, showing three aciculate tubercles at back of occipital ring, a pair of prominent tubercles on pre- glabellar field, and aciculate tubercle at left anterior margin of anterior border, x 10. 8, Left free cheek (A. 5869). External mould, showing aciculate tubercles on margin, x 8. 9, Right free cheek (A. 5870), showing vincular swelling on doublure, x6. 10, Pygidium (A. 5871a). Oblique posterior view, x 10. Figs. 11a, b. Toernquistia sp. A. Cranidium (A. 5872). Dorsal and frontal views, X 14. Fig. 12. Toernquistia sp. B. Cranidium (A. 5873a), x8. Figs. 13, 14. Mesotaphraspis sp. 13a, b, Cranidium (A. 5874). Dorsal and lateral views, X12. 14, Right free cheek (A. 5875a), possibly belonging to this form, X 12. Fig. 15. Unassigned free cheek (A. 5876), x6. Fig. 16. Unassigned hypostome (A. 5877 a), x8. Figs. 17-21. Carrickia pelagia gen. et sp. nov. 17a-c, Cranidium (holotype, A. 5878). Dorsal, frontal, and lateral views, x8. 18, Cranidium (A. 5879 b). External mould, showing pits on glabella and inner part of fixed cheek, x 12. 19, Cranidium (A. 5880), x7. 20, Small cranidium (A. 58816), with narrow glabella. External mould showing smooth surface, x 12. 21a, 6, Pygidium (A. 5882). Dorsal and posterior views, x 18. Figs. 22, 23. Ampyx sp. 22, Cranidium (A. 5883). Oblique lateral view, x 3. 23, Pygidium (A. 5884a) attributed to this species, x 4. Palaeontology , Vol. 8 PLATE 81 W?- • •• li*# mams 5 ;«SS3 TRIPP, Girvan trilobites R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 585 Genus bumastoides Whittington 1954 Bumastoides scoticus sp. nov. Plate 80, figs. 26-32 Diagnosis. Cranidium subquadrate, strongly convex longitudinally, weakly so trans- versely. Palpebral lobes weakly rounded, forwardly placed. Anterior margin of pygidium almost transverse; doublure short, with a minute median cusp. Holotype. A. 5860a, b (pygidium) Plate 80, figs. 31a, b. Paratypes. A. 5855a, b (cranidium); A. 5857 (free cheek). Other material. Twelve cranidia, seven free cheeks, two rostral plates, one hypostome, thirty-seven pygidia. Dimensions (in mm.). Holotype A. 5861a, b Length of pygidium . .3-1 6-7 Width of pygidium . . 5-3 8-3 Description. Cranidium subquadrate, strongly convex longitudinally particularly near back, weakly so transversely. Lunate muscle impressions shallow, situated near back of cranidium and four-sevenths of cranidial width apart. Shallow apodemes on posterior margin, slightly wider apart. Palpebral lobes weakly rounded, forwardly placed. An- terior branches of facial sutures run almost straight foswards at first, bending inwards anteriorly; posterior branches run backwards and slightly outwards. Surface smooth; one or two terrace lines run parallel to anterior margin and curve round parallel to facial sutures. Free cheek weakly convex. Eye large, seven-tenths of its own length from posterior margin, situated at about the same distance from lateral as from posterior margin; lens surface convex, marked off by a shallow depression. Posterior and lateral margins almost straight; genal angle broadly rounded. Doublure lies close to dorsal surface posteriorly, increasingly convex anteriorly. Rostral plate broadly rounded anteriorly, bowed backwards posteriorly, weakly con- vex longitudinally. Terrace lines about sixteen in number, most of which are continuous from side to side. Hypostome rounded in posterior outline. Middle body evenly convex; crescentic posterior lobe slightly depressed, with small maculae laterally. Anterior wings large, broadly pointed, sloping dorsally upwards ; continuous with narrow, swollen posterior border. Surface smooth, except for anterior wings which bear longitudinal terrace lines, widely spaced. Pygidium gently convex proximal to doublure, sloping steeply downwards laterally. Anterior margin almost transverse. Axis indistinguishable except for slight forward convexity and depressions at positions of axial furrows, half width of pygidium, or more, apart. Doublure two-sevenths length of pygidium, weakly convex, with a faint longi- tudinal median depression; anterior margin simple except for a minute median cusp, distinguishable only on well-preserved specimens. Surface smooth; terrace lines on doublure closely spaced, faint. Remarks. The pygidium of B. scoticus differs from the type species, B. milled (Billings) (see Whittington 1954, pp. 138-9, pi. 62, figs. 16-18, 20, 25, 26, 29), in the almost 586 PALAEONTOLOGY, VOLUME 8 transverse anterior outline and in the monocuspid, not bicuspid, anterior margin of the doublure. The doublure is monocuspid also in B. billingsi (Raymond and Narraway) (see Bradley 1930, pi. 28, fig. 2), but the doublure is longer in that species. Family harpidae Hawle and Corda 1847 Genus hibbertia Jones and Woodward 1898 Hibbertia whittingtoni sp. nov. Plate 81, figs. 1-3 Diagnosis. Eye tubercles forwardly placed. Brim widest antero-laterally, forepart flexed upwards. Pits fine, radially arranged on cheek roll. Short genal caeca proximally on brim. Holotype. A. 5862a, b (cephalon). Plate 81, figs. la-c. Other material. Four cephala and fragments of brim. Dimensions of holotype (in mm.). Length of cephalon (sag.) 5 0 Length of cephalon (exs.) 7-7 Width of cephalon . 8-9 Width of cheek roll 4-4 Length of glabella . 1-9 Description. Width of cephalon approximately equal to length, but variable; glabella moderately long, narrowing forwards, swollen. Basal lateral lobes well developed. Occipital ring short; occipital furrow shallow. Alae small, depressed. Eye tubercles large, forwardly placed, strongly raised. Eye ridges indistinct. Preglabellar field extends less than half distance to brim. Cheek roll convex, strongly rounded in outline antero- laterally. Posterior borders raised, weakly defined. Border strong, extending to tip of brim. Brim widest antero-laterally, gently concave except for prolongations ; forepart of brim anterior to cheek roll flexed upwards. Brim prolongation shorter than median length of cephalon, narrowing steadily, upwardly inclined. Cheek roll prolongation narrows slowly at first, extending to tip of brim prolongation. Outer rim moderately developed on both lamellae. Glabella punctate medianly. Alae smooth. Pits on cheek roll and brim small, radially arranged on cheek roll; short genal caeca proximally on brim, either one or two pits between caeca. Outer rim smooth. Remarks. This species is referred to Hibbertia rather than Selenoharpes because the girder extends to the tip of the cheek prolongation; shape and pitting of the glabella, and flexure of brim are other features of resemblance. The occurrence of genal caeca on the inner part of the brim, and the fineness of the pitting of the brim are characters of Selenoharpes, which serve to distinguish this from other species of Hibbertia. Family otarionidae Richter and Richter 1926 Genus otarion Zenker 1833 Otarion sp. Material. One incomplete cranidium. R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 587 Remarks. The distinctive feature of this specimen is the weak convexity of the glabella and of the large basal lobes. Family dimeropygidae Hupe 1953 Genus dimeropyge Opik 1937 Dimeropyge hystrix sp. nov. Plate 81, figs. 4-10 Diagnosis. Glabella strongly swollen ; margin of anterior border angled mesially ; palpe- bral lobes situated anterior to mid-length of glabella; posterior borders with a pair of swellings proximally and several small tubercles laterally. Holotype. A. 5865a, b (cranidium). Plate 81, figs. 4 a-c. Paratypes. A. 5869 (free cheek); A. 5871a, b (pygidium). Other material. Seventeen cranidia, six free cheeks, two pygidia. Dimensions (in mm.). Holotype A. 5867a, b Length of cranidium . 30 30 Length of glabella . 20 1-9 Width of glabella (maximum) 1-6 VI Description. Glabella narrowly ovate, strongly convex in both directions, standing well above cheeks, sloping forwards and downwards. Muscle areas absent. Preglabellar and axial furrows continuous and deep ; axial furrows wide at back. Occipital ring convex longitudinally, moderately arched transversely. Occipital furrow broad and shallow. Preglabellar field moderately long, continuous with cheeks, steeply inclined, convex, with a shallow longitudinal median depression. Fixed cheeks broad, convex, highest near back. Palpebral lobes small, elevated, situated anterior to mid-length of glabella and twice their own length from posterior border ; one- and two-thirds maximum width of glabella apart. Palpebral furrows well defined. Margin of anterior border angled mesially, convex, almost as long as preglabellar field. Anterior border furrow broad and shallow. Posterior borders short (exs.) proximally, increasing in length and inflation laterally. Anterior branches of facial sutures converge slightly forwards; posterior branches run obliquely outwards and backwards to posterior border furrows, thence curving more strongly outwards and cutting posterior margin immediately inside libri- genal spines. Cranidium set with aciculate tubercles directed upwards on glabella, backwards on occipital ring, and forwards on preglabellar field and on forefront of anterior border. Occipital ring more finely tuberculate than glabella, sometimes with three large tubercles placed towards back of ring. Fixed cheeks slightly more coarsely tuberculate than glabella; usually a single tubercle on median depression, with a pair of tubercles on either side, hind ones prominent. Posterior part of anterior border granu- lar. Posterior borders with a pair of swellings proximally surmounted by one or more diminutive tubercles; lateral parts swollen, with several small tubercles; parts between low and smooth. Eye small, strongly convex. Inner area of cheek broad, weakly convex. Lateral border set at an angle to inner area, narrow anteriorly, widest opposite posterior border furrow. Lateral border furrow moderately deep, curving inwards posteriorly to join posterior 588 PALAEONTOLOGY, VOLUME 8 border furrow. Librigenal spine moderately long, tapering steadily, curving strongly inwards near tip. Doublure narrow and produced anteriorly, as wide as free cheek oppo- site posterior border where a (concave) vincular swelling is developed, flattened under spines. Inner area of cheek coarsely tuberculate; border granular, with a marginal row of long slender tubercles ; librigenal spine granular. Pygidium short, flattened, with a deep, almost vertical, border. Axis broad anteriorly, narrowing strongly, composed of three segments. First ring well formed, moderately arched; second ring with transversely oval lateral areas more swollen than median part; third ring represented only by a pair of small, transversely oval swellings. Ring furrows shallow but continuous. Axial furrows shallow. Three pairs of narrow, slightly swollen pleurae; pleural furrows oblique, shallow. A pair of tubercles behind axis is the only indication of a fourth segment. Border deep, sloping steeply and uniformly downwards. Pygidium granular ; a low tubercle placed proximally on each pleuron, a pair of aciculate tubercles, posterior much the larger, at extremities. Remarks. In cephalic characters D. hystrix bears a close resemblance to D. spinifera Whittington and Evitt (1954, pp. 42-46, pi. 22; pi. 23; text-figs. 9, 10) from the Lin- colnshire Limestone; the main points of similarity are the strongly swollen glabella, the shallow median depression crossing the preglabellar field, the incurved tips of the free cheeks, the aciculate tuberculation, and the arrangement of tubercles on the posterior borders. In pygidial characters the resemblance is less close. D. minuta 6pik (1937, pp. 32-33, pi. 3, figs. 1, 2; pi. 4, fig. 5; pi. 12, figs. 1, 2; pi. 19, fig. 1) from the Kukruse Shales possesses even larger swellings than D. hystrix proxi- mally on the posterior borders, but is quite distinct in other respects. Genus toernquistia Reed 1896 Toernquistia sp. A Plate 81, figs. 11 a,b Material. Two cranidia. Description. Glabella broadly rounded in outline anteriorly, convex; lateral glabellar furrows absent. Preglabellar and axial furrows continuous, deep and broad; median preglabellar pit large, indenting preglabellar field. Preglabellar field almost half length of glabella, convex, sloping steeply downwards. Anterior border moderately long, sloping forwards, border furrow well defined. Fixed cheeks moderately wide, gently convex. Shallow depressions run outwards and forwards from antero-lateral angles of glabella. Palpebral lobes large. Posterior borders widen (exs.) steadily laterally. Sur- face finely tuberculate. Remarks. The cranidium figured bears some resemblance to the type species, T. nichol- soni (Reed 1896, pp. 433-5, pi. 21, figs. 3, 3 a) but the glabella is much less swollen, and the divergent antero-lateral furrows are shallower. Toernquistia sp. B Plate 81, fig. 12 Material. One cranidium. Remarks. This cranidium differs from the foregoing species in its shorter, more swollen glabella, and longer anterior border. R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 589 Genus mesotaphraspis Whittington and Evitt 1954 Mesotaphraspis sp. Plate 81, figs. 13, 14 Material. One cranidium [two free cheeks]. Description. Cranidium weakly convex in both directions. Glabella lanceolate in out- line, weakly convex. Lateral lobes and furrows lacking. Preglabellar and axial furrows continuous, lightly impressed; shallow median furrow extends a short way across pre- glabellar field. Occipital ring moderately long; occipital furrow shallow, transverse. Preglabellar field approximately three-eighths length of glabella, weakly inclined, con- tinuous with fixed cheeks. Palpebral lobes moderately long, weakly defined. Anterior border of uniform width. [Free cheek narrow, weakly rounded in outline. Eye large; lens surface convex. Anterior branches of facial sutures diverge slightly forwards; posterior branches run outwards and then curve strongly backwards to cut posterior margin just inside genal angle. Inner area gently convex, at narrowest part slightly wider than border. Lateral border of uniform width, well defined. Librigenal spine continuous with lateral border in curvature, tapering strongly. Surface smooth.] Remarks. This species resembles M. inornata Whittington and Evitt (1954, pp. 48-49, pi. 24, figs. 1-39) from the Lincolnshire Limestone more closely than it does M. parva Whittington and Evitt (1954, pp. 46-48, pi. 3, figs. 31-36; pi. 4, fig. 11) from the Edin- burg Limestone. The following are the main differences from M. inornata: (1) The glabella is narrower and more pointed anteriorly; (2) the lateral lobes and furrows are lacking; (3) the preglabellar furrow is much more shallow and the median pit is lacking. The free cheeks described agree better with Mesotaphraspis than with Toernquistia (see Warburg 1925, pi. 5, fig. 41). Family raphiophoridae Angelin 1854 Genus ampyx Dalman 1827 Ampyx sp. Plate 81, figs. 22, 23 Material. One cranidium, one pygidium. Description. Cranidium convex in both directions, but distorted. Glabella swollen, widening steadily forwards; anterior part narrowing rapidly and projecting beyond cephalic margin. Anterior spine broken off at base, rounded in cross-section. First pair of lateral muscle areas faintly defined. Lateral glabellar areas not developed. Preglabel- lar and axial furrows shallow. Occipital ring convex, curved weakly backwards in out- line, continuous with posterior borders, which slope steeply forwards laterally. Fixed cheeks wide (tr.) and sloping steeply downwards anteriorly, connected in front of glabella; anterior border flexed forwards. Pygidium elliptical in outline, weakly convex. Axis about one-quarter anterior width, narrowing rapidly at first, then slowly, weakly defined. Inner parts of pleural lobes flattened; border slopes gently outwards, narrow (exs.), slightly embayed mesially. 590 PALAEONTOLOGY, VOLUME 8 One pair of faint pleural furrows, concave forwards ; area anterior to furrows depressed. Surface smooth, except for shallow terrace lines closely spaced on border. Remarks. The cranidium resembles the type species A. nasutus Dalman (see Whitting- ton 1950, pp. 554-6, pi. 74, figs. 3-9; text-figs. 6a, b ) and also A. camurus Raymond (see Cooper 1953, p. 16, pi. 5, figs. 1-2, 6-7) in shape of glabella and absence of lateral glabellar areas. The pygidium is much shorter than in either of these species. Genus lonchodomas Angelin 1854 Lonchodomas pernix sp. nov. Plate 82, figs. 1-7 Diagnosis. Glabella narrow, tapering slowly anteriorly, carina absent. Posterior borders curve slightly forwards ; posterior border furrows sharp. Basal part of glabella pitted. Holotype. A. 5885a, b (cranidium). Plate 82, figs. 1 a-d. Paratypes. A. 5887 (free cheek); A. 5890 (pygidium). Other material. Sixteen cranidia, two free cheeks, one hypostome, fourteen pygidia. Dimensions of holotype (in mm.). Length of cranidium to base of spine . 8-9 Width of cranidium . . . .9-3 Maximum width of glabella . . 3-7 Basal width of glabella . . .2-0 Description. Length of cranidium great compared with width, but variable. Glabella narrow, tapering slowly anteriorly, moderately convex longitudinally, strongly arched particularly at back. Hindmost part of glabella swollen to form a transverse ridge. Frontal spine square in cross-section, straight, horizontally extended; upper pair of angular ridges converge for a short distance on glabella but do not meet ; carina absent. Axial furrows broad and shallow, not crossing occipital segment; elongate fossulae just posterior to greatest width of glabella. Occipital ring bowed gently backwards, sloping forwards and downwards, continuous with posterior borders. Posterior borders become narrower laterally, and curve slightly forwards. Posterior border furrows sharp; lateral pits near extremities. Fixed cheeks narrow, convex, sloping forwards. Courses of facial sutures gently sinuous. Faint, longitudinal raised lines on antero-lateral slopes of glabella near base of frontal spine. Cranidium shallowly but irregularly pitted. Basal part of glabella strongly pitted; smooth, triangular, basal muscle areas strongly developed, other muscle areas indistinct. Free cheek narrow, sloping steeply outwards; inner area convex. Librigenal spine long and slender, rectangular in cross-section, horizontally extended, curving gently inwards. Doublure of cheeks continuous beneath glabella, longest (sag.) mesially, where doublure gently convex, narrowing laterally. Distal band flattened. Proximal band abruptly depressed, becoming wider and increasingly convex towards median line ; out- line of inner margin rounded, with a slight median embayment. Surface of cheek and doublure smooth. Hypostome trapeziform in outline, weakly convex longitudinally, strongly arched transversely. Middle body broadly rounded in posterior outline. Posterior lobe and R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 591 maculae not developed. Border narrow, convex, thickened and bent down at posterior wings. Posterior border furrow shallow. Thorax with axis three-eighths total width, weakly arched. Articulating half ring faint ; articulating furrow deep and broad. Axial furrows shallow. Inner parts of pleurae wide (tr.), directed straight outwards, and horizontally extended. Pleural furrows transverse, placed at two-thirds length from front, widening laterally. Outer parts of pleurae bent vertically downwards, incomplete. Pygidium subtriangular. Axis broad, ill defined, arched transversely, extending a short distance onto border. Axial furrows indistinct. Inner parts of pleural lobes weakly con- vex; border gently convex, sloping steeply downwards, wide (exs.) laterally, embayed mesially. One pair of pleural furrows, concave forwards, sharply defined laterally; area anterior to furrows depressed. Terrace lines closely spaced on border. Remainder of surface smooth. Remarks. Professor H. B. Whittington has pointed out that the name L. tumidum (Tripp 1962, p. 15) is a secondary homonym of A. tumidus Forbes (1849, pi. 10, p. 4). I hereby propose minuntionensis as a substitute name. L.pernix differs fromL. minuntionensis in its weaker longitudinal convexity and in the shape of the glabella. It is distinguished from most other species by the absence of carina and by the slowly tapering anterior part of the glabella. Family cheiruridae Salter 1864 Genus ceraurinella Cooper 1953 Ceraurinella sp. Plate 82, fig. 8 Material. One cranidium. Material from other horizons. Cranidia from the superstes Mudstones, Aldons. Remarks. The affinities of this cranidium are with Ceraurinella from which it differs in the weak definition and sinuous course of the axial furrows, and the absence of tuber- culation on both glabella and cheeks. Genus sphaerexochus Beyrich 1845 Sphaerexochus sp. Plate 82, figs. 10-17 Material. Twelve cranidia, two free cheeks, five pygidia. Material from other horizons. All parts from the superstes Mudstones, Aldons. Description. Glabella variable in proportions, frequently much wider than long, sloping steeply downwards with strong convexity. Basal lateral glabellar furrows deepest near anterior inner angles, proximal parts almost longitudinal, shallow. Palpebral lobes placed opposite basal lateral furrows. Fixed cheeks with genal angles bluntly pointed in large specimens (PI. 82, fig. 12), produced into short spines in smaller specimens (PI. 82, fig. Mb); lateral and posterior margins sigmoidal in outline. Inner area of free cheek and lateral border narrow; vincular notch strongly developed, lateral border furrow shallow and narrow compared with other species. 592 PALAEONTOLOGY, VOLUME 8 Pygidium twice as wide as long. Axis narrow. Third ring fused with terminal piece and with third pleurae. Axial furrows shallow anteriorly, moderately deep and very broad alongside terminal piece, interrupted opposite third ring. Three pairs of short, narrow, swollen pleurae, extending successively further backwards. Free points long, narrow, bluntly pointed, or rounded, with broad, rounded notches between. Dorsal surface and doublure granular, pygidium more densely so than remainder of exoskeleton. Remarks. This species differs from S. eurys Tripp (1962, pp. 19-20, pi. 3, figs. 1-7) in that successive pygidial pleurae extend further backwards. Cranidia are extremely variable and hard to distinguish reliably. Genus sphaerocoryphe Angelin 1854 Sphaerocoryphe sp. Plate 82, fig. 9 Material. One cranidium, one hypostome, one fragmentary pygidium. Remarks. Cranidium small (1-6 mm. in length) with elongate bulbous glabellar lobe. Hypostome subrectangular, with elongate middle body. The anterior pair of pygidial pleurae end in comparatively long, outwardly directed free points. The species is too inadequately known for close comparison. EXPLANATION OF PLATE 82 Figs. 1-7. Lonchodomas pernix sp. nov. 1 a, 6, Cranidium (holotype, A. 5885a). Dorsal and oblique lateral views, x 3. lc. The same, external mould (A. 58856), x 3. Id, Enlargement of external mould of basal part of glabella to show pitted surface with smooth areas, x 12. 2, Short cranidium (A. 5886), x 3. 3, Right free cheek (A. 5887), showing doublure and dorsal surface at base of librigenal spine, x 6. 4, Posterior part of hypostome (A. 5888), x 10. 5, Thoracic segment (A. 58896), external mould, x4. 6a, 6, Pygidium (A. 5890). Dorsal and posterior views, x5. 7, Pygidium (A. 5891). External mould showing anastomosing terrace lines on border, x 8. Fig. 8. Ceraurinella sp. Cranidium (A. 5892). External mould, X 6. Fig. 9. Sphaerocoryphe sp. Hypostome (A. 5893), X 8. Figs. 10-17. Sphaerexochus sp. 10, Cranidium with broad glabella (A. 5894a), X4. 11a, Small crani- ■ dium with narrow glabella (A. 5895a). Dorsal view, x6. 116, The same. Lateral view, showing short fixigenal spine, x 8. 12, Large cranidium (A. 5896) devoid of fixigenal spines, x 4. 13, Crani- dium (A. 5897a). Lateral view, x 12. 14, Cranidium (A. 5898a). Lateral view showing diminutive fixigenal spine, x5. 15, Right free cheek (A. 5899a), x6. 16, Pygidium (A. 59006). Latex mould from external cast, x 5. 17, Pygidium (A. 5901). Specimen with rounded pleurae and notches, x 4. I Figs. 18-28. Encrinuroides obesus sp. nov. 18a, 6, Cranidium (holotype, A. 5902a). Dorsal and lateral views, x 4. 19a-c, Small cranidium (A. 5903a). Dorsal, frontal, and lateral views. Note tall palpe- bral lobe in lateral view, x8. 20, Left free cheek (A. 5904a), x8. 21, Hypostome (A. 5905), x8. j 22, Hypostome (A. 5906). Latex mould from external mould showing granulation of surface and pits on anterior part of middle body, X 10. 23, Cranidium (A. 5907). External mould, showing unpitted eye ridges, bordered by large pits, x 12. 24, Cranidium (A. 59086). External mould, showing three pairs of sizable tubercles on glabella, x 12. 25, Pygidium (A. 5909a), x 10. 26, j Pygidium (A. 5910a). Comparatively broad axis emphasized by slightly anterior view, x 8. 27, Small pygidium (A. 5911a), Xl2. 28, Pygidium (A. 5912a). Lateral view showing steep doublure, x6. Figs. 29-31. Encrinuroides sp. 29, Cranidium (A. 5913). Left fixed cheek is closely associated, x6. 30, Left free cheek (A. 5914), x 6. 31, Pygidium (A. 5915a), laterally compressed, x 5. Fig. 32. Encrinuridae indet. Pygidium (A. 59166). External mould, x6. Palaeontology, Vol. 8 PLATE 82 TRIPP, Girvan trilobites R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 593 Family encrinuridae Angelin 1854 Genus encrinuroides Reed 1931 Encrinuroides obesus sp. nov. Plate 82, figs. 18-28 Diagnosis. Glabella slightly wider anteriorly than posteriorly, inflated; frontal lobe strongly rounded in outline. Eyes backwardly placed. Fixigenal spines slender. Tuber- culation weak. Pygidium short, with broad axis, and six pairs of short pleural ribs. Holotype. A. 5902a, b (cranidium). Plate 82, figs. 18a, b. Paratypes. A. 5904a, b (free cheek) ; A. 5905 (hypostome) ; A. 5909a, b (pygidium). Other material. Twenty-two cranidia, eighteen free cheeks, five hypostomes, ten pygidia. Dimensions of holotype (in mm.). Length of cranidium . . . .4-4 Width of cranidium (estimated) . . 9-6 Length of glabella . . . .3-5 Width of glabella across frontal lobe . 2-9 Width of glabella across basal lobes . 2 0 Description. Cephalon elliptical in outline, strongly convex in both directions. Glabella pyriform, narrowing little towards back, rising high above cheeks, strongly convex in both directions. Frontal lobe almost half length of glabella, strongly rounded anteriorly; longitudinal median furrow shallow, extending backward from preglabellar furrow for one-fifth length of glabella. Posterior lobe short, marked off by posterior lateral furrows which are faintly connected across central lobe; anterior and middle lateral lobes faintly defined by short furrows on steep lateral slopes of glabella, middle lobes slightly longer than anterior. Occipital ring short, strongly arched transversely. Occipital furrow shallow, transverse. Axial furrows deep and narrow; apodemes proximally at points of junction with middle and posterior lateral and occipital furrows; deep fossulae at forefront. False preglabellar field flattened, short. Preglabellar furrow deep and broad. Fixed cheeks convex, sloping inwards proximally. Positions of eye ridges indicated by absence of pitting. Palpebral lobes small, elevated, situated opposite hind part of middle lateral lobes, near mid-width of cheeks. Posterior borders short (exsag.), depressed, directed straight outwards. Posterior border furrows narrow. Laterally posterior bor- ders widen slightly, and join short lateral borders; fixigenal spines slender, moderately long, directed outwards and more or less strongly backwards. Anterior branches of facial sutures run obliquely forwards and inwards from eyes, cutting axial furrows immediately anterior to fossulae and meeting in front of false preglabellar field ; posterior branches curve outwards and slightly backwards. Eye lobe elongatedly oval, moderately high; lens surface convex, half height of lobe, marked off by a shallow furrow on internal moulds. Inner area of free cheek weakly convex, almost three times width of border. Pseudoglabellar area and anterior border fused, continuing curve of lateral border. Lateral border convex, weakly rounded in outline. Lateral border furrow deep and broad, slightly stronger anteriorly than pos- teriorly. Doublure convex, as wide as border. Hypostome subtriangular, anterior outline strongly rounded. Middle body oval, strongly swollen ; longitudinal median lobe short, widening rapidly backwards. Maculae B 6612 R r 594 PALAEONTOLOGY, VOLUME 8 indistinct. Anterior border moderately broad, flexed upwards. Anterior wings slope dorsally upwards. Lateral borders horizontally extended, widening slowly backwards. Posterior tongue pointed, comparatively short. Surface granular; parts of middle body shallowly pitted. Pygidium triangular, strongly convex, composed of twelve or more axial rings and six pairs of pleurae. Axis occupies about half maximum width of pygidium, convex longi- tudinally, with comparatively long post-axial ridge. Ring furrows continuous but faint mesially at back. Axial furrows straight, strongly convergent, becoming faint posteriorly. Pleural lobes narrow, sloping gently outwards with weak convexity. Suc- cessive pleural ribs directed increasingly backward; first four pairs of pleurae end in short, out-turned free points; fifth pair subparallel, sixth pair indistinctly marked off from axis, hardly reaching margin. Inter-pleural furrows broad, moderately deep. Articulating facets weakly developed. Doublure narrow, sloping steeply downwards. Cephalon within borders, including pseudoglabellar area, sparsely covered by com- paratively small tubercles of various sizes; a pair of conspicuous tubercles placed between basal lateral lobes. Inner areas of cheeks closely pitted, fixed cheeks more strongly so than free cheeks. Palpebral lobes and occipital ring smooth; posterior borders with a few small tubercles. Pseudoglabellar areas of free cheeks finely tuberculate; anterior and lateral borders granular, devoid of tubercles. Surface of pygidium granular ; a few large granules on axial rings. Doublure finely granular. Remarks. The squat pygidium, with broad axis, distinguishes E. obesus from all de- scribed species except E.fallax (Reed 1899, pp. 753-5, pi. 49, figs. 9-12) from theTramore Limestone, which it closely resembles in many respects. The main differences are that the eye lobes are situated further apart and further back, the tuberculation is weaker, the pygidium is more strongly convex, and the sixth pair of pleurae are less developed in E. obesus. Encrinuroides sp. Plate 82, figs. 29-31 Material. One cranidium, two free cheeks, four pygidia. Description. Glabella weakly convex, rising no higher than cheeks. Basal lateral lobes exceptionally short; lateral furrows short and broad. Palpebral lobes situated near mid- width of cheeks, not quite as far back as in E. obesus. Fixigenal spines comparatively long and stout, curving outwards and backwards. Free cheek gently convex; eye stalk incomplete. Inner area broad. Pseudoglabellar area and anterior border fused, set at an angle to lateral border, which narrows towards back. Lateral border furrow becomes weaker posteriorly. Surface of inner area sparsely tuberculate and shallowly pitted; pseudoglabellar area and lateral border granular. Pygidium narrow, strongly vaulted. Axis narrow compared with pleurae, convex longitudinally, composed of about fourteen rings. Six pairs of pleurae, last pair short ; first four pleurae end in long horizontally extended free points. Pleural ribs narrow, strongly swollen, rib furrows broad. Surface smooth except for a few low tubercles on axial rings. Remarks. This species differs conspicuously from E. obesus in the less convex glabella, the outline of the free cheek, and narrow, vaulted pygidium. R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 595 Genus cybele Loven 1846 Cybele ? sp. Plate 83, fig. 1 Material. One incomplete cranidium. Description. Glabella unknown except for swollen middle and posterior lateral lobes, standing higher than fixed cheek. Fixed cheek broad, weakly convex. Palpebral lobe elevated, placed far forwards. Eye ridge almost transverse, moderately long (tr.). Pos- terior border short (exs.), furrow broad, flat bottomed, bordered on anterior side by a raised ridge. Surface granular. Inner area of cheek except for eye ridge and posterior ridge shallowly and closely pitted. Entire inner area of cheek very sparsely tuberculate. Remarks. The unpitted ridge at the back of the inner area of the fixed cheek is an unusual feature, which occurs also in Cybele bellatula Dalman (see Opik 1937, pp. 120-1, text- fig. 34), a species in which the eye ridges are similarly developed. The specimen is not sufficiently complete to justify an unqualified determination. Genus quinquecosta gen. nov. Diagnosis. Glabella widens steadily forwards, strongly convex. Frontal lobe at least half length of glabella. Three pairs of lateral glabellar lobes ; posterior pair much the smallest with lateral nodes ; anterior lateral glabellar furrows bifurcate, anterior branches short. Anterior border short, swollen laterally, dying out mesially. Eyes moderately large, placed opposite middle lateral furrows and close to glabella. Pygidium composed of about twelve axial rings and five pairs of backwardly directed pleural ribs which terminate bluntly on the arc of a circle. Type species. Q. williamsi sp. nov. Remarks. Quinquecosta is most closely related to certain members of the Subfamily Cybelinae; Atractopyge xipheres (Opik 1925, pp. 11, 12; pi. 1, figs. 10, 11 ; 1937, p. 121, pi. 7, fig. 3; pi. 21, figs. 3, 4), for instance, presents a comparable bifurcation of the anterior lateral glabellar furrows, and the free cheeks are similarly constructed. The five pairs of pygidial ribs (unfurrowed pleurae) distinguish Quinquecosta from Cybele , which has five pairs of furrowed pleurae, and from Atractopyge, which has four pairs of pleurae or ribs. Pygidia of E. quinquecostatus Mannil 1958 and E. pilisiverensis Rosen- stein 1941 possess only five pairs of ribs, but the new genus is quite different in the longer, less divergent ribs, quadrate rather than triangular outline, and absence of axial tubercles. Certain features are evocative of the family Phacopidae — for instance, the glabella widening steadily forwards ; long frontal lobe ; apodemes on middle and posterior, not the anterior, lateral furrows; short posterior lateral glabellar lobes, with lateral nodes. The general stamp of the cranidium, and the five pairs of long pygidial ribs suggest a pliomerinid relationship, and as such this form was recorded in Williams’s monograph (1962, p. 47). Quinquecosta williamsi gen. et sp. nov. Plate 83, figs. 2-10 Diagnosis. Glabella as long as wide. Lateral lobes wide ; middle lateral lobes almost as large as anterior pair. Occipital ring long, as wide as base of glabella. 596 PALAEONTOLOGY, VOLUME 8 Holotype. A. 5918a, 6 (cranidium). Plate 83, figs. 2a, b. Paratypes. A. 5920 (free cheek); A. 5923 a, b (hypostome); A. 5924 (pygidium). Other material. Twenty-five cranidia, forty-five free cheeks, seven hypostomes, twenty-nine pygidia. Material from other horizons. Cranidia, hypostomes, and pygidia from the super stes Mudstones, Aldons. Dimensions of holotype (in mm.) Length of cranidium . . . .6-5 Width of cranidium .... 14-5 Length of glabella . . . .5-5 Width of glabella across frontal lobe . 61 Wdith of glabella across basal lobes . 3-4 Description. Glabella subpentagonal in outline, evenly convex in both directions. Fron- tal lobe half length of glabella (sag.); a shallow median pit near mid-length; antero- lateral angles sharply rounded. Lateral glabellar lobes wide, with some independent convexity but continuous with central lobe; anterior pair narrower proximally than distally; posterior pair much the shortest, expanding laterally. Anterior lateral furrows bifurcate within short distance ; anterior branches short, convergent forwards, posterior branches long, transverse, shallow. Middle lateral furrows short and deep, converging forwards. Posterior lateral furrows directed inwards and backwards for a short distance, then bending forwards. Occipital ring moderately long (sag.) and convex, as wide as basal lobes; occipital tubercle centrally placed, low. Occipital furrow shallow, bowed forwards. Transversely elongate apodemes at distal extremities of middle and posterior lateral furrows; rounded apodemes at extremities of occipital furrow. Axial furrows narrow, much deeper than lateral furrows; deep, rounded fossulae opposite anterior border furrow. Anterior border extremely short, well defined and swollen laterally, EXPLANATION OF PLATE 83 Fig. 1. Cybele ? sp. Cranidium (A. 5917), showing ridge running alongside posterior border furrow, x4. Figs. 2-10. Quinquecosta williamsi gen. et sp. nov. 2a, b, Cranidium (holotype, A. 5918a). Dorsal and oblique lateral views, x3-75. 3, Cranidium (A. 59196). External mould, x8. 4, Right free cheek (A. 5920), x4. 5, Right free cheek (A. 5921). External mould ,x6. 6, Cranidium and reversed thoracic segment (A. 5922a). Oblique lateral view, x 2-5. 7, Hypostome (A. 5923a), x 8. 8a, b. Pygidium (A. 5924). Dorsal and lateral views; axis extends almost as far as posterior margin, X3-5. 9 a-e, Pygidium (A. 5925). Latex cast from external mould. Dorsal, postero-ventral, and oblique lateral views; note short axis and post-axial ridge, x4. 10, Small pygidium (A. 5926 b). Latex cast from external mould; note short axis and incurved pleurae, x 6. Fig. 11. Encrinuridae indet. Right free cheek (A. 59276). External mould, x5. Figs. 12-16. Amphilichas priscus sp. nov. 12a, Cranidium (holotype, A. 59286). Latex cast from ex- ternal mould. Dorsal view, x4. 126, c. The same. Lateral and frontal views, x3. 13, Small cranidium (A. 5929), x 10. 14a, Hypostome (A. 5930a), attributed to this species, x5. 146, The same, external mould (A. 59306), x 10. 15, Pygidium (A. 5931a), attributed to this species, x6. 16, Pygidium (A. 5932), attributed to this species. External mould, x 12. Figs. 17-19. Amphilichas sp. 17, Cranidium (A. 5933a). Lateral view, X 3. 18, Hypostome (A. 5934) attributed to this species, x 3. 19, Pygidium (A. 59356) attributed to this species. External mould, x 8. Fig. 20. Hemiarges sp. Cranidium (A. 5936a), X 10. Fig. 21. Ceratocephala sp. Cranidium (A. 5937). External mould, X 10. Palaeontology, Vol. 8 PLATE 83 TRIPP, Girvan trilobites R, P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 597 becoming indistinct mesially, cut off by axial furrows. Anterior border furrow well marked laterally, becoming shallow mesially, where indistinguishable on internal moulds. Outline of anterior border transverse for median one-quarter width, oblique and gently embayed laterally. Fixed cheeks proximal to eyes narrow, strongly convex transversely, extending to anterior extremities of cranidium; postero-lateral areas short and broad (tr.). Palpebral lobes moderately large, about anterior width of glabella apart, extending from opposite mid-length of anterior lateral lobes almost to posterior border furrows, sloping steeply inwards. Palpebral furrows start near mid-length of palpebral lobes and curve round posterior extremities of eyes. A large pit at mid-point of lobe. Posterior borders short, strongly convex longitudinally, not depressed, bending sharply downwards and slightly backwards at two-fifths width, widening near genal angles which are rounded. Posterior border furrows deep and wide. Anterior branches of facial sutures run almost straight forwards at first, curving inwards and converging at anterior extensions of free cheeks, presumably joining rostral suture mesially; pos- terior branches long, almost transverse except for a backward sigmoidal bend at lateral border. Glabella and fixed cheeks within borders finely and sparsely granular. Occipital ring and posterior borders smooth. Anterior border strongly granular. Fixed cheeks coarsely pitted. Lens surface occupies two-thirds of short eye-lobe, marked off by a furrow on internal moulds only. Free cheek within border broad, gently convex. Border convex, narrow, strongly rounded in outline, widening forwards; anterior extensions large, strongly convex, bent at an angle to border, almost vertical. Border furrow uniformly narrow and shallow. Doublure two-thirds width of border, upcurved. External surface of inner area shallowly and densely pitted, and sparsely granular; border and doublure closely granular. Hypostome subtriangular. Middle body oval, convex, not clearly defined. A pair of small maculae just posterior to mid-length. Anterior border narrow and flattened mesi- ally, widening rapidly laterally and marked off from middle body by a broad furrow. Anterior wings large, steeply inclined. Lateral borders narrow anteriorly, widening slightly and joining posteriorly, horizontally extended, rounded in outline. Middle body granular. Axis one-quarter total width of thorax, strongly arched transversely. Articulating furrow broad, with deep, rounded apodemes near extremities. Axial furrows deep and narrow. Pleurae straight and horizontally extended to fulcra at mid-width, curving downwards and slightly backwards laterally, free points tapered. Pleural ribs strongly swollen proximally, becoming narrower (exs.) to fulcra, expanding and less swollen laterally. Anterior bands short (exs.), depressed. Pleural furrows sharp on external, broad on internal moulds, curving forwards and dying out half-way beyond fulcra. Surface granular. Pygidium as long as wide. Axis almost half maximum width, tapering rapidly to a pointed termination near posterior margin, strongly convex transversely at front, flattened posteriorly; in some specimens hind-part of axis is replaced by a post-axial ridge. Twelve rings, ring furrows transverse, deepest laterally, successively shallower towards back, deeper on internal than on external moulds. Axial furrows sharp. Pleural lobes convex, steep. Articulating facets extend for full width of lobes, marked off by broad furrows ; anterior slopes of first ribs widen markedly laterally. Five pairs of 598 PALAEONTOLOGY, VOLUME 8 unfurrowed pleural ribs directed backwards for most of their length, terminating bluntly. Successive ribs increasingly convergent, and extending slightly further backwards. First four pairs of ribs correspond with axial rings. Interpleural furrows deep and broad, extending successively more lightly across doublure. Doublure uniformly narrow, horizontally extended, arcuate in outline. Pygidium granular, median area of axis coarsely so. Remarks. Only one other species is known — a form with broader glabella from the platy upper Stinchar Limestone, Auchensoul Hill. Encrinuridae indet. Plate 82, fig. 32; Plate 83, fig. 11 Material. One free cheek, one pygidium. Description. One free cheek bears some resemblance to that of Quinquecosta but the eye lobe is much taller, the inner area is smaller, and, to judge by the course of the facial suture, the eye must have been forwardly placed and close to the glabella. Pygidium strongly convex. Axis narrow, composed of numerous well-defined rings. Axial furrows deep and narrow. Five (?) pairs of pleural ribs extend successively further backwards; fifth pleurae long, convergent. First three pairs of pleurae corre- spond with first three axial rings. Ribs widen laterally. Rib furrows deep and broad. Surface smooth. Remarks. Although the posterior part of the pygidium is lacking, it seems unlikely that more than five pairs of ribs were developed. Nevertheless, the affinities of both free cheek and pygidium are equivocal, and it is possible that both belong to some encrinurid genus. Family lichidae Hawle and Corda 1847 Genus amphilichas Raymond 1905 Amphilichas priscus sp. nov. Plate 83, figs. 12-16 Diagnosis. Cranidium moderately convex. Central glabellar lobe weakly expanded anteriorly. Lateral lobes narrow posteriorly. Longitudinal furrows shallow throughout. Basal lateral furrows faintly indicated. Holotype. A. 5928 a, b (cranidium). Plate 83, figs. 12 a-c. Other material. Five cranidia [four hypostomes, four pygidia]. Dimensions of holotype (in mm.). Length of cranidium .... 9 0 Maximum width of glabella . . 9-0 Width of central lobe opposite eyes . 2-9 Width of lateral lobes opposite eyes . 2-7 Description. Cranidium weakly convex for most of its length, more strongly so at front. Frontal lobe projects strongly anterior to lateral lobes, swollen, not expanded to full width of glabella, narrowing rapidly backwards, widening slowly posteriorly. Lateral lobes broad and swollen anteriorly, narrowing opposite posterior extremities of palpe- R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 599 bral lobes, where narrower than central lobe, widening posteriorly and extending back- wards far beyond central lobe; posterior lateral angles broadly pointed. Longitudinal furrows much deeper on internal than on external moulds. Basal lateral furrows repre- sented by deep notches formed by axial furrows at one-quarter length from back, and by independent convexity of basal lateral lobes. Axial furrows deep, bowed inwards. Occipital ring long, narrowing rapidly behind lateral lobes, not extending sideways beyond them. Occipital furrow deep and narrow. Anterior border ill defined. Palpebral lobes large, horizontally extended, slightly more than their own length from posterior border. Palpebral furrows lacking. Fixed cheeks slope outwards. Cranidium closely covered with tubercles of various sizes ; occipital ring more weakly tuberculate, a single large tubercle placed posteriorly. [Hypostome three-quarters as long (exs.) as wide, broadly rounded anteriorly. Middle body five-eighths length of hypostome, extending to anterior margin, moderately con- vex. Posterior lobes slightly swollen. Middle furrows short (tr.), narrow, bending abruptly backwards proximally, dying out before reaching posterior furrow. Lateral furrows deep, extending some way across posterior border. Lateral borders convex anteriorly; posterior wings and posterior lobes narrow. Posterior border long, weakly convex, with a median embayment extending for about one-third the length (sag.). Doublure of posterior border extends to posterior furrow, convex except in front of median embayment where a flattened channel reaches almost to anterior margin. Sur- face of antero-lateral parts of central body and lateral borders with anastomosing, sharply raised lines, convergent backwards.] [Fragmentary pygidium weakly convex; axis pointed, contracting abruptly posteriorly, with one ring clearly defined. Short pleural furrows on first and second pleurae.] Remarks. The closest resemblance is to A. wahlenbergi Warburg (1925, pp. 315-19, pi. 8, figs. 27-35, 26?, 41 ?) from the Kallholn Limestone; the new species differs mainly in the greater anterior width of the lateral lobes, and in the vestiges of the basal lateral lobes. Hypostomes and pygidia provisionally attributed to A. wahlenbergi and to A. priscus are comparable. Amphilichas sp. Plate 83, figs. 17-19 Material. Two cranidia, one hypostome, three pygidia. Description. Cranidia differ markedly from A. priscus in much stronger longitudinal convexity, wide expansion of the frontal lobe and deep longitudinal furrows which die out abruptly before reaching occipital furrow. Hypostome attributed to this form wider, particularly postero-lateral borders, and middle furrows do not bend backwards; anastomosing raised lines fainter. Pygidium attributed to this species has axis tapering backwards but undefined posteriorly; three axial rings strongly marked and a fourth faintly. Axial furrows straight. First and second pairs of pleurae broad, strongly furrowed, ending in long free points. Third pair of pleurae fused, unfurrowed, extending back- wards much further than tips of second pleurae, gently embayed for a short way mesially, but without free points. Surface strongly tuberculate. Remarks. Hypostome and pygidium have been attributed to this species on the grounds of their greater width, compared with A. priscus. 600 PALAEONTOLOGY, VOLUME 8 Genus hemiarges Giirich 1901 Hemiarges sp. Plate 83, fig. 20 Material. Three cranidia. Material from other horizons. Cranidia from the superstes Mudstones, Aldons. Remarks. The long, rounded frontal lobe and small bicomposite lobes are distinctive features of this small species. Identical cranidia from the superstes Mudstones, Aldons, are better preserved, and justify specific description. Family odontopleuridae Burmeister 1843 Genus ceratocephala Warder 1838 Ceratocephala sp. Plate 83, fig. 21 Material. Two cranidia. Description. Glabella broad posteriorly narrowing forwards, ill defined; longitudinal convexity strong. Central lobe with independent transverse convexity. Anterior lateral lobes obsolete. Middle and posterior lateral lobes and furrows ill defined except for apodemes. Occipital ring moderately long, strongly convex longitudinally, fused with fixed cheeks laterally, without posterior band. A pair of strong, divergent occipital spines at back of ring with prominent median tubercle anterior to them. Fixed cheeks only slightly wider than basal lobes. Palpebral lobes elevated, placed opposite middle lateral lobes ; a row of small tubercles just above base of lobe. Eye ridges comparatively short. Larger tubercles on cranidium are symmetrically arranged as follows: six pairs of tubercles on central lobe of glabella, fourth pair from back placed on a transverse ridge, and two tubercles one behind the other on middle and posterior lobes. Remarks. This form bears some resemblance to a new species from the platy upper Stinchar Limestone, but differs (1) in having cranidial tubercles smaller; (2) only one pair of tubercles on central lobe placed on a transverse ridge, compared with the upper Stinchar species in which all six pairs are situated on ridges ; (3) two pairs of tubercles on the lateral lobes. Unassigned free cheek Material. Three free cheeks. Plate 81, fig. 15 Description. Free cheek extremely broad, strongly rounded in outline, gently convex. Eye (incomplete) short, weakly rounded. Inner area much wider posteriorly than an- teriorly. Lateral border widens towards back. Short (exs.), wide (tr.) subgenal notch indents posterior border laterally. Lateral and posterior border furrows shallow, meet- ing at an acute angle. Librigenal spine continuous with lateral border in curvature, taper- ing steadily. Inner area of cheek finely and closely tuberculate. Remarks. It seems unlikely that the free cheek figured belongs to any of the genera re- corded in this paper. R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 601 Material. One hypostome. Unassigned hypostome Plate 81, fig. 16 Description. Hypostome elongate. Middle body oval, moderately swollen, extending to anterior border. Posterior lobe short, strongly defined laterally, weakly demarcated mesially. Lateral and posterior border furrows deep. Anterior wings small. Lateral borders consists of narrow bands anteriorly, widening slowly and becoming more convex backwards. Posterior border long (exs.), flattened, with rounded median inden- tation (best seen on external mould) extending half-way across border. Area anterior to indentation swollen. Longitudinal raised lines on anterior parts of lateral borders; remainder of surface smooth. Remarks. This hypostome does not seem to be attributable to any one of the foregoing genera. A pygidium of Carrickia is closely associated with this specimen, but there is no evidence that they are part of one individual, although appropriate in size. Ross (1951, pi. 16, figs. 21-29) figured five unassigned hypostomes from the same horizon as Gonio- phrys, but none of these has an embayed posterior border. CONCLUSIONS 1. The Albany mudstones with calcareous nodules yield a shelly fauna, without grap- tolites, in which both trilobites and brachiopods are moderately common. All the specimens occur as isolated parts, but the preservation is good, and there is no sign of abrasion. None of the species described has been recognized outside the Girvan area. 2. Williams (1962, p. 58) correlated the fossiliferous beds of the Albany Group with the top of the Stinchar Limestone. The trilobite fauna is most closely allied to that from nodules in the superstes Mudstones at Aldons Quarry. The following species occur at both horizons : Carrickia pelagia gen. et sp. nov. Sphaerexochus sp. Nileus sp. Quinquecosta williamsi gen. et sp. nov. Raymondaspis sp. Hemiarges sp. Ceraurinella sp. The following are closely allied to forms from the superstes Mudstones : Trinodus sp. Dimeropyge hystrix sp. nov. Remopleurides sp. B. Toernquistia sp. A. Hypodicranotus sp. Encrinuroides obesus sp. nov. Hibbertia whittingtoni sp. nov. All the genera except Bumastoides, Mesotaphraspis, and Ceratocephala occur in the superstes Mudstones, Aldons. However, the superstes is a much larger fauna, and the degree of relationship cannot be assessed until work on the Lower Barr trilobites has been completed. 3. The Albany mudstone fauna consists mainly of genera which occur earlier in the 602 PALAEONTOLOGY, VOLUME 8 district. The only new appearances of any significance are Nileus and Toernquistia, the former widely dispersed at earlier horizons, the latter a first record. 4. Outside the Girvan area the closest relations are with the lower Edinburg Formation of the Appalachian Valley of the U.S.A., as is the case with all the Lower Barr trilo- bite faunas. Seventeen genera are common to both horizons, the most significant being Mesotaphraspis. There is a possible fink with the Kukruse Stage (Cn) of Estonia (. Dimeropyge , Quinquecosta), but no connexion with contemporary Anglo- Welsh faunas. 5. The new genus Carrickia constitutes the only record of the rare family Komaspididae in the Girvan district. The most closely allied genus is Goniophrys from the Garden City Formation. The second new genus Quinquecosta provides an interesting link between the Encrinuridae, Pliomeridae, and Phacopidae. Unlike Carrickia, it does not appear to have North American antecedents. Hypodicranotus occurs considerably earlier in the Girvan District than in North America, as far as is known at present. REFERENCES bradley, j. h. Fauna of the Kimmswick limestone of Missouri and Illinois. Contr. Walker Mus. 2, 219-90, pi. 23-30. cooper, b. n. 1953. Trilobites from the Lower Champlainian formations of the Appalachian Valley. Mem. geol. Soc. Amer. 55, 1-69, pi. 1-19. forbes, e. 1849. Figures and Descriptions illustrative of British Organic Remains. Dec. II. Mem. Geol. Surv., pi. 10, pp. 1-4. kielan, z. 1959. Upper Ordovician Trilobites from Poland and some Related Forms from Bohemia and Scandinavia. Palaeont. polon. 11, 1-198, pi. 1-36. mannil, r. 1958. Trilobites of the families Cheiruridae and Encrinuridae from Estonia. (In Russian with Estonian and English summaries.) Eesti NSV Teaaduste Akad., Geol. Inst. Uurimused. 3, 165— 212, pi. 1-8. OPIK, a. a. 1925. Beitrage zur Kenntnis der Kukruse — (C2-) Stufe in Eesti, I. Acta Comment. Univ. Tartu. A, 8, 1-18, pi. 1-2. 1937. Trilobiten aus Estland. Acta Comment. Univ. Tartu. A, 32, 1-163, pi. 1-26. reed, f. r. c. 1896. The Fauna of the Keisley Limestone. — Part I. Quart. J. geol. Soc. Lond. 52, 407-37, pi. 20, 21. 1899. The Lower Palaeozoic Bedded Rocks of County Waterford. Ibid. 55, 718-71, pi. 49, 50. 1931. The Lower Palaeozoic Trilobites of the Girvan District, Ayshire. Palaeont ogr. Soc. ( Monogr .). Suppl. 2, 1-30. rosenstein, e. 1941. Die Encrinurus-Acrten des estlandischen Silurs. Geol. Inst. Univ. Tartu. 62, 49-77, pi. 1-4. ross, J. r. 1951. Stratigraphy of the Garden City Formation in Northeastern Utah, and its Trilobite Faunas. Peabody Mus. Nat. Hist. 6, 1-161, pi. 1-36. skjeseth, s. 1955. The Middle Ordovician of the Oslo Region, Norway. Norsk geol. Tiddsskr. 35, 9-28, pi. 1-5. tripp, r. p. 1962. Trilobites from the confinis Flags (Ordovician) of the Girvan District, Ayrshire. Trans. Roy. Soc. Edin. 65, 1-40, pi. 1-4. warburg, e. 1925. The Trilobites of the Leptaena Limestone in Dalarne. Bull. geol. Instn Univ. Uppsala, 17, 1-446, pi. 1-11. Whittington, H. B. 1950. Sixteen Drdovician genotype trilobites. J. Paleont. 24, 531-65, pi. 68-75. 1952. A unique remopleuridid trilobite. Breviora, Mus. Comp. Zool. 4, 1-9, pi. 1. 1954. Ordovician trilobites from Silliman’s Fossil Mount, in Miller, A. K., Youngquist, W., and Collinson, C. Ordovician cephalopod fauna of Baffin Island. Mem. Geol. Soc. Amer. 62, 119-49, pi. 59-63. R. P. TRIPP: TRILOBITES FROM THE ALBANY DIVISION 603 Whittington, h. b. 1959. Silicified Middle Ordovician Trilobites: Remopleurididae, Trinucleidae, Raphiophoridae, Endymioniidae. Bull. Mus. Comp. Zool. Harv. 121, 371-496, pi. 1-36. andEvrrr, w. r. 1954. Silicified Middle Ordovician Trilobites. Mem. Geol. Soc. Amer. 59, 1-137, pi. 1-33. williams, a. 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, pi. 1-25. R. P. TRIPP High Wood, Botsom Lane, West Kingsdown, Sevenoaks, Manuscript received 29 September 1964 Kent SENSORY SPINES IN THE JURASSIC BRACHIOPOD AC ANT HOT HI RI S by M. J. S. RUDWICK Abstract. The rhynchonelloid Acanthothiris has tubular external spines similar to the spines of productoids and chonetoids. Their structure, arrangement and mode of growth are described in detail, on the basis of a study of exceptionally well-preserved specimens of A. spinosa. They are compared to the spines of the related genus Acanthorhynchia, and to those of the living oyster Crassostrea echinata. On the basis of the known sensory physiology of living brachiopods, it is inferred that the spines near the valve-edges of Acanthothiris contained extensions of the mantle, and that their tips bore portions of the highly sensitive mantle-edge tissue. These sensitive spine-tips could have provided the brachiopod with effective ‘early warning’ protection against poten- tially harmful agents in the environment ; and the radiating spines themselves could have formed a protective grille straddling the apertures leading into the mantle cavity. Spines further from the valve-edge are blocked, and are regarded as having been superseded functionally. This interpretation is discussed in relation to the spines of other brachiopods and Crassostrea, and in relation to the ontogeny, ecology and phylogeny of Acanthothiris. Tubular external spines were developed in great abundance and diversity in one sub- order of the Brachiopoda, namely the Upper Palaeozoic Productoidea. Similar spines, developed only in one particular position on the shell, characterize another Upper Palaeozoic suborder, the Chonetoidea. Outside these two suborders, tubular spines are rare and sporadic. This paper is concerned with one of the best known of these ano- malous genera, the Jurassic rhynchonelloid Acanthothiris. This is separated by a span of about 100 million years from the last of the spiny productoids and chonetoids, and is even more decisively separated from them by morphology and taxonomy. In all charac- ters except the spines, it is a ‘normal’ rhynchonelloid. It thus illustrates a common phenomenon in brachiopod evolution : a structure characteristic of some major group is often found to have been evolved independently by one or a few anomalous genera which clearly belong (on every other criterion) to a quite different group (cf. Williams 1956). A functional analysis of the spines of Acanthothiris may help to throw light on this phenomenon. Material. For this study, I have used a small collection of very well-preserved specimens of A. spinosa (Schlotheim), the genotype, from the Inferior Oolite (Middle Jurassic) of ‘Collier’s Lane, nr. Bath’ (Sedgwick Museum, Cambridge, SM J57751— 71), together with some similar specimens from another old collection, labelled ‘Gt. Oolite, Bath’ (Museum of Comparative Zoology, Harvard, MCZ 5794). Despite the difference of labelling, all these specimens may come from the same locality. ‘ Collier’s Lane ’ is un- fortunately not mentioned by name in Richardson’s (1907) account of the stratigraphy of the Inferior Oolite around Bath. Richardson records A. spinosa from most units of the Inferior Oolite; but the matrix suggests that the specimens came from the Upper Trigonia Grit. The matrix is a hard oolitic limestone, but it has weathered in pockets to a much softer material which, with care, can be dissected away from the spines. Normally only the stumps of the spines are preserved. For comparative purposes, I have also studied the closely related species Acantho- [Palaeontology, Vol. 8, Part 4, 1965, pp. 604-17, pis. 84-87.] M. J. S. RUDWICK: SENSORY SPINES IN ACANTHOTHIRIS 605 rhynchia senticosa (von Buch) from the Inferior Oolite of Dorset. There are no living brachiopods with true tubular spines. But the mollusc Crassostrea echinata (Quoy and Gaimard) is a possible functional analogue. I have collected this oyster from intertidal localities on the Queensland coast and offshore islands, and studied it alive in a small aquarium. The photographs in this paper were taken with a Leitz Aristophot on Ilford Pan F film; all except Plate 86, fig. 5 were lightly coated with ammonium chloride. Acknowledgements. I am indebted to Mr. A. G. Brighton (Sedgwick Museum, Cambridge) and Pro- fessor H. B. Whittington (Museum of Comparative Zoology, Harvard) for access to collections, and to Mr. M. J. Collins for the loan of an exceptionally well-preserved specimen of Acanthorhynchia. A grant from the Daniel Pidgeon Fund, awarded by the Geological Society of London, was used to defray the expenses of studying marine biology, inter alia that of Crassostrea, on the Great Barrier Reef and adjacent coastline. STRUCTURAL ANALYSIS General description. The shells of Acanthothiris (the ‘corrected’ spelling Acanthothyris is invalid: see Ager 1960) are typically ‘ rhynchonelloid ’ in form. They are strongly biconvex in profile, broadly elliptical in plan, and have a non-strophic hinge. There is a small incurved ventral umbo ; the dorsal umbo is tucked inside the delthyrium. There is a dorsal median deflexion in the commissure, a low median fold on the dorsal valve, and a shallow median sulcus on the ventral. The surface of each valve is covered with a pattern of fine costellae radiating from the umbo. On most specimens, the costellae appear to be swollen at intervals into small pustules (PI. 84, fig. 1); but well-preserved specimens show that these pustules are in fact the broken stumps of slender spines. The whole array of spines covers almost all parts of the valve-surfaces. The spines are arranged in rows corresponding to the costellae from which they project, radiating from the umbo of each valve towards the valve-edge. The number of rows, like the number of costellae, increases progressively in this direction. Within each row, the spines project from the costella at fairly regular intervals (PI. 84, figs. 1-3 ; PI. 85, figs. 1, 3, 4). They are not arranged regularly in concentric fashion, but it will be convenient to refer to a set of spines as those projecting from any narrow concentric zone of the valve-surface (cf. Rudwick 1959). On most parts of the shell the spines project almost tangentially from the curved surface of the valves ; but in the lateral sectors, and especially near the commissure, they project at higher angles and even perpendicularly (PI. 84, fig. 5). The spines are very rarely preserved to their full length ; some have been dissected from the matrix to a length of 8 mm. (PI. 85, fig. 5) but were certainly longer than this. They are cylindrical; at the base they are about 0-5 mm. in diameter, but they taper gradually to about 02 mm. diameter on those preserved to the greatest length (PL 84, fig. 6; PI. 85, figs. 1, 5; PI. 86, figs. 1, 5). They are straight or slightly curved. They are composed of an external sheath of dense primary-layer shell, of matt appearance, within which there is generally an almost solid core of shiny fibrous secondary-layer shell. The con- trast between the two layers is easily seen at the broken tips of the spines (PI. 86, fig. 5) or on the broken stumps (PI. 85, fig. 2). Mode of growth of spines. Since the spines are sheathed in primary-layer shell, which is continuous with that of the rest of the valve, it is clear that they are derivatives of the valve-edge, at which the primary-layer was secreted. Their mode of growth can be 606 PALAEONTOLOGY, VOLUME 8 reconstructed from a study of individual spines preserved at various stages of growth, and from the growth-lines on the primary-layer. The commissure was affected throughout ontogeny by a serial deflexion of small wavelength and low amplitude. Thus the valve-edges were slightly crinkled and inter- locked with one another (PI. 84, fig. 2; PI. 86, figs. 6, 7), while the cumulative effect of the deflexion was to produce the costellae on the valve surfaces. Each spine first appeared at the valve-edge on one of the crests (text-fig. 1a) of the deflexion. Initially it was marked only by a slight expansion of the valve-edge at this point (text-fig. 1b). Successive increments of primary-layer accentuated this into a small ‘collar’ (text-fig. lc). At the next stage observed the ‘collar’ had been completed into a shelly ring surrounding a small foramen, while the main growing-edge had resumed its normal form (text-fig. Id). In other words, a small circle of valve-edge had, in effect, been ‘budded off’ from the A B C D E text-fig. 1 . Initial stages in the growth of a tubular spine at a valve-edge in Acanthothiris. For further explanation, see text. Semi-diagrammatic, about x 20. rest of the valve-edge. Thereafter it grew independently, adding circular increments of primary-layer, which lengthened the ring into a tube ; while the continued growth on the main growing-edge isolated the base of the incipient spine further from the valve-edge (text-fig. 1e). All these early stages in the growth of a spine can be recognized near the valve-edges of well-preserved specimens (PI. 84, figs. 3, 4; PI. 86, figs. 2, 3, 4). Further growth at the distal end of the spine, with a gradual reduction in the diameter of the circle of growing-edge, would have produced the long, gently tapering form of the full- grown spines. Relation of spines to mantle-tissue. These growth-stages can be interpreted in terms of the secretory activity of the mantle-tissue by using the standard homological relation between shell and mantle as known in living brachiopods (cf. Williams 1956). Thus the budding-off of a ring of valve-edge material reflects an identical budding-off of a ring of mantle-edge tissue, by which the primary-layer was secreted. Each hollow-growing spine must have been lined with an inner tube (or possibly a solid cylinder) of mantle-tissue, connecting the distal ring of mantle-edge tissue to the rest of the mantle. Obviously the spine could have grown in length only so long as this connexion was maintained. If the spines are examined in cross-section near the base (e.g. where they have all been broken off) it is clear that their walls were progressively thickened with secondary-layer, until the central lumen was constricted to vanishing point. Only the spines nearest the valve-edge (i.e. the youngest spines) have a clear central lumen ; those further from the valve-edge (formed earlier in ontogeny) are blocked with concentric layers of shell- material (PI. 84, fig. 3 ; PL 85, figs. 1, 2; PI. 86, fig. 4). This secondary-layer material must have been secreted by the inner tube or core of mantle-tissue, which must thereby have been progressively constricted. Ultimately the tissue must have been abandoned; the spine could then have grown no more. The completion of this process is seen also in M. J. S. RUDWICK: SENSORY SPINES IN ACANTHOTHIRIS 607 the blocking of the internal foramina leading into the spines : these are clearly open near the valve-edges, but are blocked further in from the valve-edges (PI. 86, figs. 2, 3, 6, 7). Time-sequence of growth of spines. This distinction between open and blocked spines leads to an interpretation of the time-relations of the whole array of spines. The spines in each row are those that were successively budded off, during ontogeny, from the same crest of the serial deflexion. The spines in each set, lying on or near any single growth- line on the valve-surface, are those that were formed all round the valve-edge at or about the same time during ontogeny. In any given row, at the moment when a new spine began to be formed at the valve-edge, the immediately preceding spines would still be open, but the earlier ones would have been blocked. On the valve as a whole, at any given moment in ontogeny, the most recently formed set or sets of spines, all round the valve- edge, would still be open, but all those nearer the umbo would have been blocked (PL 86, figs. 6, 7). Although a spine might have remained open after ceasing to grow, the converse is of course impossible ; hence the actively growing spines must also have been restricted to the peripheral zone of each valve. Each individual spine must therefore have grown relatively quickly to its full length. Orientation and arrangement of spines. If the shell is described ‘ statically ’, the orientation of the spines appears to be rather haphazard : especially in the lateral sectors they seem to project at many different angles (PL 84, fig. 2). A more ‘dynamic’ description, taking into account the time-relations of the spines, reveals a simpler and more orderly arrange- ment. Seen in plan, each spine projects perpendicularly from the growth-lines on the valve-surface at its base (PL 84, fig. 2; PL 85, figs. 1-4); in other words, during the time of its formation and growth it projected radially from the valve-edge. Seen in profile, each spine projects obliquely from the valve-surface and curves gently towards the opposite valve (PL 84, fig. 6; PL 86, fig. 1); during the time of its formation and growth it would have projected at a low angle across the plane of the commissure. The most recently formed sets of spines on the two valves would thus have projected radially and interdigitated with each other (text-fig. 2). Since all the spines were formed from the crests of the serial deflexion, and the crests necessarily alternate in position on the two valves, the projecting spines would have alternated likewise, and therefore would not have obstructed the closing of the shell. The spacing of the spines around the valve-edges was markedly uniform at any given growth-stage. In the later stages there was generally a gap of about 0-7 mm. between the bases of adjacent spines. Since the spines diverge and also taper, this spacing would obviously have increased away from the valve-edge. Ontogeny of spines. The formation of radiating spines began very early in ontogeny. On large shells the stumps of the earliest spines are preserved very close to the umbo of each valve (PL 84, fig. 1 ; PL 85, fig. 2). The growth-lines show that they correspond to a stage at which the shell was only about 1-5 mm. broad. A rough estimate from the size of the stumps suggests that the spines themselves were about 0T5 mm. in diameter near the base, and were there separated from each other by spaces about 0-25 mm. across. During the next stage of ontogeny the diameter and spacing of the spines increased rather rapidly to the adult size (PL 84, fig. 2). Thereafter, notwithstanding the progressive increase in the size of the shell, both the diameter and the spacing of the spines remained remarkably constant. This probably reflects a similar constancy in the original length of the spines. The average spacing of the peripheral spines (about 0-7 mm. at the base) was 608 PALAEONTOLOGY, VOLUME 8 maintained, in spite of the increasing length of the valve-edges, by the progressive intro- duction of new rows of spines (borne on new costellae) intercalated between the earlier rows. The effect of this is to be seen in the overall appearance of the whole shell: the bases of the spines appear to be scattered at a uniform density over all parts of the valve- surfaces (PI. 84, figs. 1-3, 5; PI. 85, figs. 1-4). VIII text-fig. 2. Reconstruction of parts of the valve-edges in Acanthothiris, gaping apart as in life, showing grille of spines straddling the aperture, and sensitive spine-tips projecting outwards from the shell; about x 10. Only the peripheral zone of ‘ open ’ spines is shown : note apertures on inner surface of lower valve. Costellae I, III, ... on one valve alternate with costellae II, IV, ... on the other; hence spine-rows 1, 3, ... interdigitate with spine-rows 2, 4, ... In each row there is one fully-grown ‘open’ spine (a); spines of the next set (b), which would ultimately supersede them, are at various stages of growth (9b and 8 b are most advanced, 5b and 4b least; 1 b has not yet appeared). The spacing between successive spines in a single row was also kept remarkably con- stant during ontogeny. This is particularly striking if a row on the anterior sector of the EXPLANATION OF PLATE 84 Figs. 1-6. Acanthothiris spinosa (Schlotheim) ; Inferior Oolite (Middle Jurassic); Bath, England. 1. Left postero-lateral view of SM J57759, X 3, showing general form of shell, costellae, and stumps of broken spines. 2. Spines on left antero-lateral sector of dorsal valve of SM J57757, x 6. 3. Spines on left lateral sectors of MCZ 5794a, x 6, showing short ‘active’ or ‘open’ spines near commissure (two with white pointers), and ‘blocked’ stumps of earlier spines. 4. Fragment of valve-edge MCZ 5794c, x 6, showing two very young spines. 5. Spines on left postero-lateral sector of SM J57757, x 6, showing radial orientation of later spines (one complete young spine with white pointer). 6. Row of spines on left lateral sector of dorsal valve of SM J57753, seen in profile (valve broken), x 6, showing slight curvature towards commissure. Palaeontology, Vol. 8 PLATE 84 RUDWICK, Spines of Acanthothiris M. J. S. RUDWICK: SENSORY SPINES IN ACANTHO THIRIS 609 shell is compared with one of the much shorter rows on the postero-lateral sectors. The total number of spines projecting from a long row is much greater than from a short row, although both of course represent the same span of time. Yet the spacing between successive spine-bases is approximately the same (about T2-T5 mm.). This implies that a new spine began to be formed at the valve-edge whenever the preceding spine had be- come distant from the edge by that amount, regardless of the length of time that had elapsed since the preceding spine was formed. Comparison with spines in other shells. The related genus Acanthorhynchia has spines which are closely similar to those of Acanthothiris, except that they are on a smaller scale. Thus, where only the stumps of the spines are preserved, they form a similar but much finer pattern on the valve-surface (PI. 87, fig. 2; compare PI. 84, fig. 1, at same magnification). The spines themselves are very similar in appearance, but are only about 0T mm. in diameter near the base, though they reach over 4 mm. in length (PI. 87, fig. 1 ; compare PI. 85, fig. 2, at same magnification). No other rhynchonelloid has comparable spines. The living species Tegulorhynchia doderleini (Davidson) has radial rows of very short projections, scarcely deserving the name of spines, each of which is a short incomplete tube with a slit or seam on the side nearest the valve-edge (Leidhold 1922, Taf. XIII). In this they resemble the earliest stage in the formation of a true tubular spine; but the budded-off portion of the valve-edge evidently fails to fuse into a complete isolated ring. As Leidhold pointed out, such short spines have more resemblance to the short spines of some bivalve molluscs (e.g. Pinna ) than to the tubular spines of fossil brachiopods. Holder (1958) has likewise emphasized the fundamental difference between such molluscan spines and the spines of Acanthothiris. Nevertheless, at least one mollusc shows a moderately close, though imperfect, analogy. This is the oyster Crassostrea echinata (Quoy and Gaimard), which in the juvenile stage develops moderately long quasi-tubular spines. They are not truly tubular, for they invariably have a narrow slit or seam on the side nearest the valve-edge (PI. 87, figs. 6, 8) ; but they may reach a length of 5 mm. at an external diameter of about 0-7 mm. They are formed by the development of an upturned ‘collar’ of valve-edge (PI. 87, figs. 1,2), which fails to fuse into a complete ring; but the further addition of incomplete rings of shell-material builds up a cylinder which, from the side furthest from the valve-edge, appears to be a complete tube (PI. 87, figs. 5, 7). The spines project almost perpendicularly from the surface of the free (uncemented) right valve of the oyster. They are not arranged in regular radial rows, but generally occur in roughly concentric sets. Each set consists of spines which, at their time of formation, were spaced at fairly regular intervals around the edge of the valve. The set nearest the valve-edge has open apertures to the interior ; earlier sets have been blocked off- at the base (PI. 87, figs. 4, 6, 8). Rather similar spines seem to have developed in one orthoid brachiopod, Spinorthis geniculata Wright (1964), of Upper Ordovician age. These spines are arranged in radial rows and concentric sets, and project almost perpendicularly from the valve-surfaces. They are short, and appear to be incomplete tubes. Spines with a greater similarity to those of Acanthothiris have been found in an atrypoid brachiopod, Atrypa cf. aspera, of Middle Devonian age (Jux 1962). They are arranged in radial rows, and seem to have been long and tapering; it is not clear whether they were true or incomplete tubes. s s B 6612 610 PALAEONTOLOGY, VOLUME 8 Apart from such sporadic occurrences, the only genuine structural analogues to the spines of Acanthothiris are the spines of productoids and chonetoids. These are truly tubular spines, and were apparently formed in an identical manner, by the ‘budding-off’ of small portions of the valve-edge. The spines of some spiriferoids, which may be ‘ double-barrelled ’ and bear hooks or barbs, do not seem to be comparable. FUNCTIONAL ANALYSIS Rejected interpretations. Many functional interpretations have been suggested for spines in general, or for the spines of brachiopods in particular. Most of them, however, are clearly inapplicable to the spines of Acanthothiris. These are so narrow in bore, and would have been even narrower when lined with an inner tube of mantle-tissue, that it is in- conceivable that they could have acted as channels for the water-currents flowing through the mantle-cavity. For the same reason, and since few of them were open at any given growth-stage, it is inconceivable that they could have increased significantly the area of mantle-tissue available for gaseous exchange. Moreover, as Holder (1958) has pointed out, Acanthothiris is generally found in calcareous sediments, which are unlikely to have been deposited in poorly aerated conditions. The spines are fragile (at least in the fossil state : they would have been less so when the organic component of the shell- material was still present) ; and once they were broken they could not be repaired (un- less they were still ‘open’). It is therefore difficult to believe that they performed any porcupine-like function of protection against large predators. On the other hand, it is perhaps possible that they might have served to camouflage the shell, by breaking up its otherwise sharp and distinctive outline. This fails to account, however, for most of the detailed characteristics of the spines. Finally, whatever the natural orientation of the shell, most of it must have been above the level of the substratum, in order to provide the apertures with clear access to food- and oxygen-bearing water. Hence only a few of the spines (probably the earliest, in the umbonal region) could ever have served to anchor the shell in a soft substratum (cf. text-fig. 3b). Unlike the rhizoid spines of productoids (Muir- Wood and Cooper 1960), the spines of Acanthothiris have never been found cemented to substratal material. Spines as a sensory mechanism. A more promising interpretation can be derived from knowledge of the sensory mechanisms of living articulate brachiopods. Although the exact nature of the sensory receptors is at present unknown, they are apparently confined to the extreme edge of the mantle. The mantle-edge is highly sensitive to touch, light, and some chemical stimuli, and is richly supplied with nerve-endings. (If setae are pre- sent, they too are sensitive, but only to tactile stimuli.) A simple reflex nerve circuit from the mantle-edge causes the shell to be snapped shut in response to stimuli (Rudwick EXPLANATION OF PLATE 85 Figs. 1-5. Acanthothiris spinosa (Schlotheim) ; Inferior Oolite (Middle Jurassic); Bath, England. 1. Later spines on dorsal valve of SM J57756, x 6, showing radial orientation and gently tapering form. 2. Umbonal region and right postero-lateral sectors of SM J57759, x 9, showing structure of stumps of spines, and their early origin. 3. Spines on left lateral sector of dorsal valve of MCZ 5794a, x 6, showing radial orientation. 4. The same, on SM J57757, x 6. 5. Later spines on anterior sector of (broken) dorsal valve of MCZ 5794 b, x 6, showing gently tapering form. Palaeontology, Vol. 8 PLATE 85 RUDWICK, Spines of Acanthothiris M. J. S. RUDWICK: SENSORY SPINES IN ACANTHOTHIRIS 611 1961). Hence any modification of the form of the mantle-edges can be considered as a possible modification of this protective system. For example, sharply zigzagged valve- edges would have reduced the distance between the sensitive mantle-edges without any corresponding reduction in the area of the apertures (Rudwick 1964a). This would have improved the quality of protection at the apertures, but would not have given the brachiopod any earlier warning of the approach of potentially harmful agents. Early warning could only be provided by extending the sensitive mantle-edges, or parts of them, outwards away from the rest of the shell. This would have been precisely the effect of the spines of Acanthothiris. Each of the ‘open’ spines would have carried at its tip a small piece of the sensitive mantle-edge tissue (and possibly, but not necessarily, one or more setae). Assuming only that at least one of the many peripheral nerve-fibres extended along each spine, the tip would have formed a highly sensitive ‘outpost’ of the brachiopod’s protective system, giving early warning of the approach of harmful agents and enabling the shell to be snapped shut well in advance. The paradigm, or optimal specification for this function (Rudwick 19646), is fairly easy to determine. Ideally the tips of the spines might be situated very far from the shell, but in reality the length of the spines would presumably be limited by the supply of material available for their construction and by the strength of the material itself. They would need to be as slender as possible, consistent with strength, since they would otherwise tend to obstruct the free flow of the water-currents in and out of the mantle-cavity. A cylindrical form is of course the most effective shape by which to com- bine strength with economy. Their sensitive tips would be disposed most effectively if they were situated at uniform intervals all round the approaches to the apertures. With the continuing growth of the shell, the tips of these ‘active’ spines would inevitably be shifted further and further from the apertures, and would gradually become less effective; and they would need to be replaced by new spines, projecting from the valve- edges in their new position. The observed structure and arrangement of the spines of Acanthothiris fit this speci- fication very well. If the tips of the ‘open’ spines are identified as the points of chief functional importance, much that is otherwise obscure about the spines becomes intelligible. This interpretation explains the radial orientation and regular spacing of the spines in each set; the time-sequence of growth and blockage that marks each row; and their slender cylindrical form. It implies that only the most recently formed sets of spines were fully functional at any given growth-stage ; and that all the earlier spines had been superseded and had become functionally redundant through the inevitable shift in their relative position (text-fig. 3a, b). The range of sensitivity around each spine-tip would depend on the sensitivity of the mantle-edge tissue. This in turn would probably depend on properties at the cellular level, which would be unlikely to change in proportion to the growth of the whole organ- ism. There would therefore be certain optimal distances for the spacing of adjacent spine- tips, both between spines in adjacent rows, and between successive spines in the same row ; and these optimal values would remain relatively constant during ontogeny. Although the validity of this aspect of the paradigm cannot be demonstrated conclusively, it is at least probable, and it would explain the remarkable constancy in the spacing of the spines. This interpretation is compatible with another, slightly different, concept of protection. 612 PALAEONTOLOGY, VOLUME 8 The peripheral spines, by their orientation and slight curvature towards the oppo- site valve, would have interdigitated when the shell was closed. When the valves gaped apart through a small angle (as they do in living brachiopods), and water-currents were being pumped by the lophophore through the mantle-cavity, the same spines would therefore have straddled the apertures with a protective grille (text-fig. 2). No ‘particle’ (e.g. a swimming animal or piece of floating debris) larger than the spaces between text-fig. 3. Reconstructions of Acanthothiris (a, b) and Acanthorhynchia (c), with analogous section of Crassostrea (d), showing functional interpretation of spines as sensory ‘antennae’. Note peripheral ‘open’ spines and earlier ‘blocked’ spines. ‘Contours’ of chemo-sensitivity (radii arbitrary) shown around open spine-tips and mantle-edges ; for tactile sensitivity all the spines would be operative, and the protection would surround the whole shell. Reconstructions based on conservative estimates of original length of spines. ‘Body’ of animals close-stippled; lophophore, gills, &c., omitted. About x 1-5. a, young Acanthothiris, with pedicle attached to shell fragment, b, adult Acanthothiris, with pedicle atrophied, resting loosely on unconsolidated substratum, c, Acanthorhynchia, with similar but more slender spines, and smaller shell, d, young Crassostrea echinata, with broad tubular spines on upper (right) valve, left valve cemented to rock surface. EXPLANATION OF PLATE 86 Figs. 1-7. Acanthothiris spinosa (Schlotheim) ; Inferior Oolite (Middle Jurassic); Bath, England. 1. Rows of spines on right antero-lateral sectors of SM J57757, x6, showing slight curvature towards commissure. 2. Interior of right antero-lateral sector of dorsal valve of SM J57753, x 6, showing broad peripheral zone of unblocked spine-apertures. 3. Oblique view of same valve-edge, x 6, showing connexion between two marginal apertures and their respective spines. 4. Very young spines on peripheral part of left antero-lateral sector of dorsal valve of SM J57756, x 6. 5. Close-up of spines on anterior sector of dorsal valve of SM J57757, x 18, showing shell-structure and growth- lines (not whitened). 6. Interior of left antero-lateral sector of MCZ 5794 b, x 6, showing narrow peripheral zone of unblocked spine-apertures. 7. The same, anterior sectors, x 6. Palaeontology , Vol. 8 PLATE 86 RUDWICK, Spines of Acanthothiris M. J. S. RUDWICK: SENSORY SPINES IN ACANTHO THIRIS 613 adjacent spines could have penetrated to the apertures. The slender and uniformly spaced spines would have been highly effective as a protective grille of this kind. In living brachiopods the shell itself is often highly sensitive to tactile stimuli (ap- parently by transmission of vibrations through the shell to the receptors in the mantle- edge). Thus another animal, alighting or crawling on to the brachiopod, even some way from the shell edge, may cause an immediate and rapid protective closure of the shell. In Acanthothiris, assuming the same properties, this type of tactile sensitivity would have extended to the whole length of the spines, in addition to the broader sensi- tivity (including photo- and chemo-sensitivity) of the mantle-edge tissue at the tips of the spines. Thus the protective grille of spines would have been more than a passive sorting device, barring the passage of particles larger than a certain size. Any but the lightest particles that collided with any part of any of the spines forming the grille would probably have caused a closure of the shell, and the grille would thus have acted as a protective warning device. (Even the ‘blocked’ spines further from the apertures could, of course, have acted in this way.) It is not possible to decide conclusively between these two related interpretations of ‘protection’, because their paradigms are so similar (cf. Rudwick 19646). In any case, given the validity of the anatomical reconstruction suggested here, the spines must necessarily have provided both varieties of protection, simply by virtue of their form and arrangement. Functional analogues. The spines of Crassostrea echinata appear to serve a similar func- tion of early warning protection. In the younger growth-stages, this species is cemented by the whole of its left valve to a substratum such as a rock surface or the root of a man- grove. When the shell is open, therefore, its apertures are immediately adjacent to the substratum. The peripheral set of spines projects perpendicularly from the edge of the right valve, and encloses quasi-tubular extensions of the right mantle. At the tip of each of these ‘open’ spines is a portion of the mantle-edge, which, as in brachiopods, is highly sensitive. Hence the peripheral set of spines extends the sensitivity of the oyster away from the shell in the most strategic direction, that is, away from the surface of the sub- stratum (text-fig. 3d). When, during the growth of the shell, a set of spines is no longer close to the valve-edge, a new set is formed and the older set is blocked off at the base. This oyster provides a fairly close functional analogue to the interpretation suggested here for the spines of Acanthothiris. The spines that Jux (1962) has described on a Devonian atrypid may also have been analogous in function. Though they are more irregular than the spines of Acantho- thiris, they are roughly radial and would have interdigitated; and Jux suggests that they could have formed a grille guarding the apertures and excluding larger particles of shell-debris, &c. On the present interpretation this grille would have been not merely a kind of sieve but an active sensory device, causing the protective closure of the shell. The tubular spines of productoids are diverse in form and arrangement, and were probably equally diverse in function, though all seem to have been formed in essentially the same manner. Of the varieties distinguished by Muir-Wood and Cooper (1960), the rhizoid spines manifestly served to anchor the shell to the substratum, and the halteroid spines are reasonably interpreted as serving to stabilize the shell on or in a soft substra- tum. The spines most closely analogous to the spines of Acanthothiris are the prostrate 614 PALAEONTOLOGY, VOLUME 8 spines, which are generally slender and project radially at a low angle to the valve- surface. These are interpreted as ‘serving partly for protection or as a strainer when radiating out from the anterior of the shell’ (p. 16). This statement is not amplified, and is not entirely clear. ‘Protection’ seems to imply protection by camouflage, or else what I have termed porcupine-like protection. A ‘strainer’ probably implies a structure similar to what I have termed a protective grille. But the primary function assigned to these spines in Waagenoconcha, which is one of the best examples, is ‘to prevent the shell from sinking into soft mud’ (p. 17). This could not be applied, however, to the prostrate spines on the dorsal valve, which are structurally identical to those on the ventral (assuming the shell lay on the convex ventral valve). But if they are interpreted instead as providing early warning protection, their distinctive arrangement becomes intelligible. They are very slender (about 0T5 mm. diameter) but relatively long (pre- served to 5 mm. length or more) ; and they are arranged radially, projecting outwards around both valve-edges, even posteriorly across the hinge-line (Muir-Wood and Cooper 1960, pi. 90). If the peripheral sets of spines were ‘open’ and contained extensions of the mantle, with sensitive mantle-edge tissue at the tip of each, they could have provided this productoid with efficient all-round early warning protection. The complex spines of some spiriferoids are different in structure, and presumably in mode of growth, from those of Acanthothiris. But they might have been similar in function. They are generally oriented radially, and are spaced uniformly all round the valve-edges. The internal marginal spines of Uncinulus and Hypothyridina covered the apertures with a protective grille (Schmidt 1937), but could not have given early warning protec- tion. Ecology of Acanthothiris. Young shells of A. spinosa have a hypothyrid pedicle fora- men, and were evidently attached to the substratum by a pedicle (text-fig. 3a). In larger shells, however, the ventral umbo has become tightly incurved towards the dorsal valve, leaving no space for the emergence of the pedicle. Thus the pedicle probably atrophied during ontogeny, and the adult brachiopods would have lain freely on the substratum. They were probably stabilized in orientation by shell-thickening in the umbonal region, and perhaps, adventitiously, by some of the earliest and superseded spines. The com- missural plane was probably held in an oblique or even vertical orientation, so that all the apertures were clear of the substratum (text-fig. 3b). The current system may have been the same as in adult living rhynchonelloids (Rudwick 1962), the water entering the mantle- cavity laterally and being ejected through an anterior aperture corresponding in position EXPLANATION OF PLATE 87 Figs. 1-2. Acanthorhynchia senticosa (von Buch); Inferior Oolite (Middle Jurassic); Loder’s Hill, near Bridport, Dorset, England. 1. Spines on right lateral sector of dorsal valve of specimen in collection of M. J. Collins, x9. 2. Ventral valve of same specimen, x 3, showing arrangement of stumps of spines at all growth-stages. Figs. 3-8. Crassostrea echinata (Quoy and Gaimard); Recent; intertidal rock-surfaces at Eimeo, near Mackay, Queensland. (SM). 3, 4. External and internal views of right (free) valve of young shell, x 6, showing one set of ‘ open ’ spines, with previous set blocked. 5, 6. External and internal views of part of right valve of another young shell, x 6, showing ‘ seam ’ on outer side of each spine. 7, 8. External and internal views of ventral edge of right valve of a larger shell, x 6, showing several sets of blocked spines. Palaeontology, Vol. 8 PLATE 87 RUDWICK, Spines of Acanthorhynchia and Crassostrea M. J. S. RUDWICK: SENSORY SPINES IN A CANT HO THIRIS 615 to the median deflexion of the commissure. In any case, whatever the arrangement of the apertures around the commissure, it is clear that the spines provided the same degree of putative protection to all the apertures. It is impossible to infer, with any supporting evidence, the exact nature of the ‘harm- ful agents’ against which the protection would have been effective. They might have included ‘ particles ’ as diverse as saltating shell-fragments, sand-grains or ooliths, pieces of organic debris in suspension, and small floating or swimming animals. Any of these might have clogged the delicate ciliary feeding apparatus of the brachiopod. The tips of the spines might also have been able, through the chemosensory properties of the mantle-edge, to detect the approach of potential predators, noxious water-conditions, &c. If Acanthothiris was equipped with such an effective sensory apparatus, it is perhaps surprising that it is generally found in association with other, more ‘normal’ brachic- pods (rhynchonelloids and terebratuloids without spines). This, however, overlooks the fact that such sensory apparatus as a brachiopod may possess is necessarily co-ordi- nated with other functional systems, especially the rejection mechanisms, which leave no structural trace in the fossil state. It is therefore invalid to assert that Acanthothiris must have been better adapted to its particular environment than the more ‘normal’ brachiopods with which it is associated, or (conversely) that since it is so associated its spines cannot have been an effective adaptation. Phylogeny of Acanthothiris. Buckman (1907) asserted that a spiny shell was the phylo- genetic terminal stage of a trend from a smooth shell through a stage in which the shell was ribbed. In the case of Acanthothiris there would seem to be no evidence for this. Its evolutionary origin is in fact obscure. Throughout its ontogeny (except in the earliest stage of growth), it seems to have been equipped with fully functional sensory spines. Unless some other rhynchonelloid is found, in which spines are less well developed or confined to a shorter part of ontogeny, it may be necessary to infer that the ancestor of Acanthothiris was a brachiopod of very small adult size, i.e. that it evolved by paedo- morphosis. The development of an early warning protective system in Acanthothiris was made pos- sible by the evolution of the capacity for forming tubular spines at the valve-edges. This enabled portions of the sensitive mantle-edges to be extended outwards from the shell. This capacity had not, apparently, been evolved previously among the rhyncho- nelloids, and only rarely (e.g. productoids and chonetoids) in the whole history of the brachiopods. Considering the great adaptational potential of tubular spines (as shown in the history of the productoids), it is surprising that the development of spines in this rhynchonelloid stock should have been so short-lived and ‘unsuccessful’. Possibly Acanthothiris and its allies became extinct, through the chances of environmental change, too soon for that potential to be realized. Possibly the many ecological niches that were open to the productoids in the Upper Palaeozoic were no longer open by the Jurassic, having been occupied perhaps by the bivalve molluscs. In any case, this is a problem that deserves closer analysis. Without the capacity for forming spines, it would be difficult for a brachiopod to evolve an effective early warning system. One possibility, which appears to have been realized on two separate occasions, involved the formation of long narrow projections 616 PALAEONTOLOGY, VOLUME 8 of the mantle- and valve-edges. This gave rise to the well-known and astonishingly close homoeomorphic resemblance between the Middle Triassic athyracean Tetractinella and the Upper Jurassic terebratellacean Cheirothyris (Cloud 1941, Rudwick 1965). Given the limited anatomical ‘materials’ out of which any brachiopod adaptation must be formed, the only other means of providing early warning protection would be by the use of the sensitive setae projecting from the mantle-edge. This possibility is in fact utilized by several living brachiopods, and was almost certainly exploited by many fossil groups ; but it suffers from the inherent limitation that the setae are only sensitive to tactile stimuli. CONCLUSION The structure and arrangement of the spines of Acanthothiris accord with a functional interpretation based on the known sensory properties of the mantle-edge in living brachiopods. On this interpretation, the spines projecting from the periphery of each valve acted like inflexible antennae, projecting outwards from the shell into the environ- ment, and enabling the brachiopod to receive early warning of the approach of poten- tially harmful agents. They would also have straddled all parts of the gape into the mantle-cavity with a protective grille. The brachiopod would have had this protection throughout life; the spines formed earlier in ontogeny were progressively superseded by those formed later, the functional spines remaining close to the valve-edges. Such spines would have been an effective adaptation in any brachiopod. Yet in fact they seem to have been evolved only once among the rhynchonelloids, and even on that occasion the adaptation had a relatively short history. They may have been evolved sporadically in other groups of brachiopods at other times, but were only common among the Upper Palaeozoic productoids. In view of their adaptive utility, their rela- tive rarity needs some explanation. Their formation depended on the ability of the shell and mantle to form tubular spines, and this may have been a property that was only rarely evolved. But this fails to account for their generally short history, once they had been evolved. This problem may remain insoluble, until the ecological environment of these brachiopods has been rigorously analysed. REFERENCES ager, d. v. 1960. Nomenclatural problems in the Mesozoic Rhynchonelloidea. Geol. Mag. 97, 157-62. buckman, s. s. 1907. Brachiopod morphology: Cincta, Eudesia and the development of ribs. Quart. J. Geol. Soc. Load. 63, 338-43. cloud, p. e. 1941. Homeomorphy, and a remarkable illustration. Amer. J. Sci. 239, 899-904. holder, h. 1958. fiber das Nadelkleid einiger Jura-Brachiopoden. Neues Jb. Geol. Palaont. Mh., 1958, 151-4. jux, u. 1962. Atrypiden mit erhaltener Bestachelung aus der Sotenicher Mulde (N.-Eifel). Ibid., 1962, 505-13. leidhold, c. 1922. Rhynchonella Doderleini Dav., eine kritische Brachiopodenuntersuchung. Neues Jb. Min. Geol. Palaont. 45, 423-70. Muir- wood, h. m. and cooper, g. a. 1960. Morphology, classification and life habits of the Produc- toidea (Brachiopoda). Mem. Geol. Soc. Amer. 81, pp. xi+447. richardson, l. 1907. The Inferior Oolite and contiguous deposits of the Bath-Doulting district. Quart. J. Geol. Soc. Lond. 63, 383-426. M. J. S. RUDWICK: SENSORY SPINES IN ACANTHO THIRIS 617 rudwick, m. j. s. 1959. The growth and form of brachiopod shells. Geol. Mag. 96, 1-24. 1961. ‘Quick’ and ‘catch’ adductor muscles in brachiopods. Nature, 191, 1021. 1962. Filter-feeding mechanisms in some brachiopods from New Zealand. J. Linn. Soc. Lond. (. Zool .), 44, 592-615. 1964#. The function of zigzag deflexions in the commissures of fossil brachiopods. Palaeontology, 7, 135-71. 19646. The inference of function from structure in fossils. Brit. J. Philos. Sci. 15, 27-AO. 1965. Adaptive homoeomorphy in the brachiopods Tetractinella and Cheirothyris. Palaont. Zeitschr. 39, Heft 3/4 (in the press). schmidt, h. 1937. Zur Morphogenie der Rhynchonelliden. Senckenbergiana, 19, 22-60. williams, a. 1956. The calcareous shell of the Brachiopoda and its importance to their classification. Biol. Rev. 31, 243-87. wright, a. d. 1964. The fauna of the Portrane Limestone, II. Bull. Brit. Mus. (Nat. Hist.), Geol. 9, 157-256. M. J. S. RUDWICK Sedgwick Museum, Cambridge Manuscript received 28 October 1964 SOME NORTH AMERICAN SPECIES OF THE DEVONIAN TETRACORAL SMITHIPHYLLUM by A. E. H. PEDDER Abstract. The tetracoral genus Smithiphyllum is emended. It is shown to be closely related to Tabulophyllum and consequently is transferred from the Spongophyllidae to the Endophyllidae. Additional material of the type species, S. imperfectum (Smith), is described and three new Frasnian species, S. belanskii from Iowa, S. kindlei from Alberta, and S. whittakeri from the Northwest Territories, are erected. Species now assigned to Smithiphyllum were first erected for specimens from Timan, at the turn of the century. Since then, others have been named for North American specimens, but recent workers have referred these to quite different genera and even famihes. Apart from the description of new species, the purpose of this paper is to redefine the genus and elucidate its relationships. The following abbreviations are used : GSC for Geological Survey of Canada (Ottawa) type number SUI for State University of Iowa (Iowa City) type number N.W.T. for Northwest Territories. SYSTEMATIC PALAEONTOLOGY Family endophyllidae Torley 1933 Genus smithiphyllum Birenheide 1962, emend. Type species ( original designation). Spongophyllum imperfectum Smith. See below, p. 622. Description. Fasciculate to subcerioid tetracorals with a relatively thick wall. The septa are radially arranged, typically smooth and clearly differentiated into two orders. The major may be withdrawn both axially and peripherally and the minor are commonly represented merely by ridges on the interior of the wall and short crests on the dissepi- ments. The dissepimentarium is narrow and may be lonsdaleoid; in some species, in- cluding the type, it is only intermittently developed. The tabulae are broad and in narrow corallites many are complete. They are generally gently sinuous in longitudinal section and are commonly downturned peripherally. The minute skeletal structure is imperfectly preserved in much of the material studied. The skeletal material of the wall is almost entirely lamellar. As seen in transverse section the lamellae are deflected towards the axis at the bases of the septa and in longitudinal section they lie oblique to the wall surfaces with their upper edge away from the peri- phery. A thin dark line (axial plate of Flower 1961, p. 28) separates contiguous corallites; however, no trace of this is present in most free corallites. Fibre fascicles have not been observed, but sparse, dark, and apparently structureless spine-like bodies within the septa and walls may represent trabeculae. [Palaeontology, Vol. 8, Part 4, 1965, pp. 618-28, pis. 88-89. A. E. H. PEDDER: NORTH AMERICAN SPECIES OF SMITHIPHYLLUM 619 text-fig. 1. Known distribution of Smithiphyllum in North American Frasnian beds. Occurrences indicated by a black circle. PALAEONTOLOGY, VOLUME 8 620 Additional species assigned Smithiphyllum belanskii sp. nov. See below, p. 623. Smithiphyllum lcindlei sp. nov. See below, p. 625. Spongophyllum lituus Smith 1945, p. 56, pi. 11, figs. 5 a-d. Redknife Formation (late Frasnian), Jean Marie River, N.W.T. Spongophyllum martinense Stumm 1948, p. 41, pi. 11, figs. 9, 12; pi. 12, fig. 6. Martin Limestone (Middle Frasnian), Dear Creek Valley, Arizona. Cyathophyllum stuckenbergii Lebedew 1902, pp. 179, 180, pi. 2, figs. 21-24. Uchta River, Timan. Probably Frasnian. Cyathophyllum weberi Lebedew 1902, p. 178, pi. 4, figs. 49-53. Uchta River, Timan. According to Soshkina (1952, p. 70) this occurs in the Middle Frasnian DJ 2-3 beds. Smithiphyllum whittakeri sp. nov. See below, p. 626. Species requiring further description Spongophyllum expansum Stumm 1937, pp. 436, 437, pi. 53, fig. 1 1 ; pi. 55, figs. 6a, b. Basal 500 feet of the Nevada Limestone, Eureka region, Nevada. Stumm considered this to be a lower Middle Devonian species; it is now thought to be Emsian (Johnson 1962). Spongophyllum nevadense Stumm 1937, pp. 435, 436, pi. 53, fig. 10; pi. 55, figs. 5a, b. Basal 500 feet of the Nevada Limestone (Emsian), Eureka region, Nevada. Calophyllum tschernyschewi Lebedew 1902, p. 148, pi. 2, figs. 29-31. Uchta River, Timan. Probably Frasnian. See under the discussion of S. imperfectum. Distribution. At present the genus is known definitely only from the Frasnian of North America and Timan. Its distribution in the former is indicated in text-fig. 1 . In addition to the described species the figure indicates the occurrence of some undescribed forms as well as others listed by Sloss and Laird (1945), Crickmay (1962, p. 4), and by McLaren and Mountjoy (1962, pp. 8, 23). Discussion. Birenheide (1962, p. 82) included completely cerioid species, such as Spongo- phyllum alpenense Ehlers and Stumm, S. breviseptatum Stumm, and S. missouriense Ehlers and Stumm, in the genus. The present writer prefers to regard these species as constituting either a new genus, or at least a subgenus. Another species included in Smithiphyllum by Birenheide is Spongophyllum pax Smith. This is based on a specimen found in a river boulder and is therefore of unknown stratigraphical origin. Crickmay (1960, pp. 878, 879) listed it in Givetian faunas from British Columbia, but Norford (1962, p. 27) has suggested that the species is Silurian and placed it in Columnaria. In recent years the species now grouped in Smithiphyllum have most frequently been referred to Spongophyllum. The latter cannot be said to be a fully understood genus since the minute structure of the type species is imperfectly known, but it does differ from Smithiphyllum in being cerioid and by having essentially concave tabulae and a well- developed dissepimentarium, even in very narrow corallites. Smithiphyllum is close to Tabulophyllum (text-figs. 2b, d ; 3a, b) and in fact Soshkina (1952, p. 70) assigned S. weberi to Tabulophyllum. Both genera possess a lamellar wall, broad typically peripherally downturned tabulae, and similar septa. Furthermore, the discontinuity, which is seen in the dissepimentarium of such species as Smithiphyllum imperfectum and S. belanskii, has also been figured in various species of Tabulophyllum (e.g. Freeh 1885, pi. 10, fig. 2; Sloss 1939, pi. 11, fig. 18). Smithiphyllum is distinguished from Tabulophyllum , which is solitary, by its fasciculate to subcerioid form. Birenheide (1962, p. 69) placed Smithiphllym, and also doubtfully Tabellaephyllum, in the Spongophyllidae. However, this writer feels that the discontinuous crest-like septa and the peripherally downturned tabulae indicate a closer relationship with the Endophyllidae. A. E. H. PEDDER: NORTH AMERICAN SPECIES OF SMITHIPHYLLUM 621 text-fig. 2. Minute skeletal structure (as far as preserved) in Smithiphyllum and Tabulophyllum. All X 15. A, c. Smithiphyllum belanskii sp. nov., A from a transverse section of the paratype, SUI 11617, c from a longitudinal section of the holotype, SUI 11616. B, d. Tabulophyllum mcconnelli (Whiteaves), B from a transverse section of topotype, GSC 17547, from the Escarpment Formation (middle Frasnian) on Hay River, N.W.T. d from a longitudinal section of GSC 17548 from the Ferques Limestone (middle Frasnian) in the Parisienne Quarry, near Ferques, Boulogne region, France, e, f. Smithiphyllum imperfectum (Smith), from transverse and longitudinal sections, respectively, of GSC 17543. A B text-fig. 3. Tabulophyllum rotundum Fenton & Fenton, x2. Based on topotype GSC 17549 from Cerro Gordo Member of the Lime Creek Formation at Rockford, Iowa. This specimen combines features considered by Fenton and Fenton (1924) as being diagnostic of both T. rectum, the type species of Tabulophyllum, and T. rotundum. A, transverse section, b, longitudinal section. 622 PALAEONTOLOGY, VOLUME 8 Smithiphyllum imperfectum (Smith 1945) Plate 88, fig. 7; Plate 89, figs. 1-3, 10-12; text-figs. 2e, f, 4a-i 1945 Spongophyllum imperfectum Smith, pp. 55, 56, pi. 11, figs. 3 a-g. Holotype. GSC 6307. Upper Devonian, about one mile below the upper end of the middle gorge of Jean Marie River, N.W.T. In current stratigraphical terms (Belyea and McLaren 1962, pp. 6, 7) this specimen was obtained from the Redknife Formation (late Frasnian). Other material. GSC 17543. Kakisa Formation (late Frasnian), 4 miles (direct) above the mouth of Poplar River, N.W.T. Collected by the writer in 1960 with Endothyra sp., Mictophyllum modicum Smith, M. semidilatum Smith, Hexagonaria magna (Webster and Fenton), Disphyllum sp., Tabulo- phyllum sp. close to T. mcconnelli (Whiteaves), Thamnopora spp.. Alveolites sp., Syringopora sp., Schizophoria sp., Nervostrophia sp., Atrypa sp., Cyrtospirifer sp., Cranaena sp., and Scutellum sp. Description. The growth form is dendroid ; in most cases the offsets diverge widely from the parent corallite so that the corallum has the appearance of being well spread out. Specimen GSC 17543, which was not completely collected, measured about 150x100x70 mm. before sectioning. The corallites are sub-cylindrical and have an adult diameter of between 7 and 9 mm. Increase is lateral. The exterior of the exposed corallites bear rugae, but not interseptal ridges. Although the wall is generally about 0-4 mm. thick, it varies between 0-25 and 0-75 mm. (Smith gives the range as 0-5 to 0-75 in the holotype) ; it consists of a very thin dark outer axial plate and an inner much thicker lamellar layer. The septa are smooth, radially disposed and well differentiated into two orders. The major taper towards their axial extremity where they are very thin; most are continuous lamellae typically extending from one- to two-thirds the distance to the axis; some are peripherally withdrawn. The minor septa are very short, most are less than 0-5 mm. long and in young corallites they may be almost imperceptible. Number of major septa ranges from nineteen at 6-5 mm. diameter to twenty-two at 9-0 mm. diameter. The dissepimentarium is intermittently developed and is entirely absent in some longitudinal sections ; where it is developed it consists of a single row of elongate dissepi- ments numbering about ten in 10 mm. The tabulae are broad and may be predominantly complete. Some are very nearly flat, but most are very gently sinuous as seen in longitudinal section. Discussion. The synonymy and distribution of the species is fully treated elsewhere (Predder in press) and is not repeated here. The cylindrical corallite described by Lebedew (1902, p. 148, pi. 2, figs. 29-31) under the name Calophyllum tschernyschewi is similar to isolated corallites of S. imperfectum. EXPLANATION OF PLATE 88 All figures x 2 Figs. 1-3, 5. Smithiphyllum belanskii sp. nov. from the Shellrock Formation, Iowa. 1-3, Holotype, SUI 11616. 5, Paratype, SUI 11617. Fig. 4. Smithiphyllum whittakeri sp. nov. from the Redknife Formation, N.W.T. Holotype, GSC 17545. Fig. 6. Smithiphyllum kindlei sp. nov. from the Mount Hawk Formation, Alberta. Holotype, GSC 17544. Fig. 7. Smithiphyllum imperfectum (Smith) from the Redknife Formation, N.W.T. GSC 17543. Palaeontology, Vol. 8 PLATE 88 PEDDER, Smithiphyllum A. E. H. PEDDER: NORTH AMERICAN SPECIES OF SMITHIPHYLLUM 623 If Lebedew’s species is based on an individual corallite of a fasciculate corallum, it appears to differ from S. imperfectum only in having as many as twenty-seven major septa at its full diameter of 9 mm. text-fig. 4. Smithiphyllum imperfectum (Smith), x 2. Based on figures given by Smith and Birenheide of the holotype, GSC 6307, and original figures based on GSC 17543. A-c, e, transverse sections of the holotype. d, transverse section of GSC 17543. f, i, longitudinal sections of GSC 17543. g, h, longitudinal sections of the holotype. Smithiphyllum belanskii sp. nov. Plate 88, figs. 1-3, 5; Plate 89, figs. 13, 16; text-figs. 2a, c; 5a-d Name derivation. The species is named for the late C. H. Belanski in recognition of his contribution to the study of the Shellrock Formation and its fauna. Holotype. SUI 11616. Mason City Member of the Shellrock Formation (early Frasnian) at Nora Springs, Floyd County, Iowa. Collected by the writer in 1956 with calcispheres, stromatoporoids and Phillipsastrea websteri (Belanski). Paratype. SUI 11617. Same horizon, locality, and collector as the holotype. Diagnosis. Maximum corallite diameter about 14 mm. Major septa up to twenty-five in number and withdrawn both axially and peripherally. Minor septa very suppressed. Dissepimentarium discontinuous. Description. The corallum is dendroid to phaceloid. The largest specimen studied is the holotype, which was incompletely collected and measured about 100 x 90 x 60 mm. before sectioning. Free corallites are subcylindrical, but contiguous ones may have their shape modified by pressure from adjacent corallites. They appear to have attained their full diameter of from 14 to 16 mm. slowly, so that most transverse sections show numerous 624 PALAEONTOLOGY, VOLUME 8 corallites of less than maximum diameter. Increase is lateral. The external features of the type specimens are completely masked by matrix. The wall consists of an inner lamellar layer and a dark outer axial plate; the latter is extremely thin and only clearly visible between touching corallites. The thickness of the wall varies between 0-2 and 1 mm. text-fig. 5. Smithiphyllum belanskii sp. nov., X2. Based on the holotype, SUI 11616 and the paratype, SUI 11617. a, d, longi- tudinal sections of the holotype. b, transverse section of the holotype. c, transverse section of the paratype. The septa are meagrely developed, radially arranged, and are highly differentiated into two orders. The major are smooth and spring from either the interior of the wall or a dissepiment; in addition to being peripherally discontinuous, they are withdrawn axially, so that few extend more than two-thirds the distance to the axis and many EXPLANATION OF PLATE 89 All figures x 2 Figs. 1-3, 10-12. Smithiphyllum imperfectum (Smith) from the Redknife Formation, N.W.T. GSC 17543. Figs. 4, 6, 19, 20. Smithiphyllum whittakeri sp. nov. from Redknife Formation, N.W.T. 4, 6, 20, Paratype, GSC 17546. 19, Holotype, GSC 17545. Figs. 5, 7-9, 14, 15, 17, 18. Smithiphyllum kindlei sp. nov. from the Mount Hawk Formation, Alberta. Holotype, GSC 17544. Figs. 13, 16. Smithiphyllum belanskii sp. nov. from the Shellrock Formation, Iowa. 13, Paratype, SUI 11617. 16, Holotype, SUI 11616. Palaeontology, Vol. 8 PLATE 89 PEDDER, Smithiphyllum A. E. H. PEDDER: NORTH AMERICAN SPECIES OF SMITH IPHYLLUM 625 considerably less. The minor septa are mere ridges on the interior of the wall and some are totally suppressed. Number of major septa ranges from none at 3 mm. diameter to twenty-five at 15-5 mm. diameter. The dissepimentarium is lonsdaleoid and as viewed in longitudinal section typically consists of a single row of elongate dissepiments ; two rows occur in parts of some coral- lites and in places the dissepimentarium is entirely absent. The tabulae are broad and many are quite complete ; in longitudinal section they are mostly very slightly convex or slightly sinuous. Discussion. The new species resembles Smithiphyllum weberi (Lebedew), particularly specimens figured by Soshkina (1952, pi. 4, fig. 16). It is distinguished by having a slightly more persistant dissepimentarium and by having fewer septa, which number up to seventy in S. weberi. S. martinense (Stumm) differs from S. belanskii in being slightly smaller and having fewer and considerably less differentiated septa. B text-fig. 6. Smithiphyllum kindlei sp. nov., X 2. Based on the holotype, GSC 17544. A, transverse section, b, longitudinal section. Smithiphyllum kindlei sp. nov. Plate 88, fig. 6; Plate 89, figs. 5, 7-9, 14, 15, 17, 18; text-figs. 6a, b Name derivation. From the late E. M. Kindle in recognition of his work on the succession of faunas in the Jasper Park region. Holotype. GSC 17544. Mount Hawk Formation, 350 to 370 feet above base, Roche Ronde, Bosche Range, Jasper National Park, Alberta, 53° 14' N., 118° OF W. Collected by P. B. Jones in 1960. Elements of the Calvinaria albertensis fauna, including Thamnophyllum colemanense (Warren) and Smithiphyllum imperfectum (Smith) were collected above and below the type horizon. Diagnosis. Known maximum width of corallite 15 mm. Maximum septal count fifty. Major septa long and commonly not withdrawn. Dissepimentarium well developed. Description. The corallum is dendroid. Before sectioning the incomplete fragment of the holotype measured about 90 x 65 x 55 mm. The corallites are subcylindrical where free and have adult diameters of from 12 to 15 mm. Increase is lateral. External features in the type material are largely obscured by matrix but the poorly preserved surface of one corallite bears traces of rugae. t t B 6612 626 PALAEONTOLOGY, VOLUME 8 The wall, consisting of a very thin axial plate and an inner lamellar layer, is about 04 or 0-5 mm. thick in the regions farthest from the septal bases. The septa are thin, smooth, and gently to moderately sinuous; their arrangement is radial and they are strongly differentiated into two orders. The major septa are generally well developed, some are peripherally discontinuous, and a few are entirely withdrawn from the dissepimentarium. Most, however, are continuous lamellae extending variably from about one-half to the entire distance to the axis. The feebly developed minor septa are typically represented by low crests on the dissepiments and to a lesser extent the tabulae, or by ridges on the interior of the wall. Numbers of major septa range from fourteen at 4 mm. diameter to twenty-five at 15 mm. diameter. The dissepimentarium is from 1 to 3 mm. wide; in places the dissepiments are limited to a single row as seen in longitudinal section, elsewhere they may be up to three deep; they number about ten in 10 mm. The tabulae are broad and may be quite complete; in some cases their median profile is irregularly sinuous, in others it is peripherally downturned, or there is a peripheral trough. Discussion. Smithiphyllum kindlei does not closely resemble any previously described species. It is, however, close to an undescribed species also occuring in the Frasnian of Alberta. S. kindlei is smaller, has fewer septa and a less pronounced dissepimentarium, furthermore the septa in S. kindlei are less withdrawn both axially and peripherally. Smithiphyllum whittakeri sp. nov. Plate 88, fig. 4; Plate 89, figs. 4, 6, 19, 20; text-figs. 7a, b Name derivation. From the late E. J. Whittaker, the pioneer geologist and collector on Jean Marie and neighbouring rivers. Holotype. GSC 17545. Redknife Formation (late Frasnian), middle Jean Marie Gorge, N.W.T., about 60° 24' N., 121° 10' W. Collected by R. de Wit in 1957. Specimens of Tabulophyllum sp. close to T. mcconnelli (Whiteaves), Smithiphyllum imperfectum (Smith), Devonoproductus sp., Atrypa ciliipes Crickmay, Spinatrypa sp., Vandergrachtella kobayashii Crickmay, and Paracyclas ? sp. were also present in the collection. Paratype. GSC 17546. Same horizon, &c. as holotype. Diagnosis. Corallum probably loosely aggregate. Corallite wall thin. Corallites up to 14-5 mm. in diameter. Major septa number up to twenty-seven, regularly withdrawn axially. Minor septa comparatively well developed. Dissepimentarium narrow, but regular. Description. Fragments only of the corallum are available and it is not known whether the coral is weakly aggregate or truly fasciculate. The corallites are subcylindrical and have an adult diameter of from 12 to 14-5 mm. Increase is lateral. Fine rugae are present on the exterior, but there are no interseptal ridges. The wall consists of a very narrow exterior axial plate and a lamellar layer; in the parts furthest from the septal bases it is about 0-5 mm. thick. The septa are thin, smooth, and most are gently sinuous as seen in transverse section; a few, however, are crenulate. They are radially arranged and clearly differentiated into two orders. The major septa are typically continuous lamellae, although some are A. E. H. PEDDER: NORTH AMERICAN SPECIES OF SMITHIPHYLLUM 627 discontinuous in the dissepimentarium; they are withdrawn from the axis leaving a region from 2 to 4 mm. wide devoid of septa. The minor septa, which are mere ridges in young stages, become well developed in later stages ; in adult corallites they are about one-half as long as the major and occur as discontinuous crests on the dissepiments and tabulae. Numbers of major septa range from nineteen at 4-5 mm. diameter to twenty-seven at 14-5 mm. diameter. As seen in longitudinal section the dissepimentarium is typically a single row of com- paratively large dissepiments, numbering from ten to twelve in 10 mm. ; it is about 2 mm. A B text-fig. 7. Smithiphyllum whittakeri sp. nov. X 2. Based on the holotype, GSC 17545. a, transverse section, b, longitudinal section. wide and fairly constant, although it is absent immediately below a lateral offset in the holotype. The tabulae are broad and many are quite complete ; in median profile they tend to be flat, or gently sloping, in the axial region, and sigmoidal at the periphery. Discussion. The species is reminiscent of Smithiphyllum weberi (Lebedew), but differs in having considerably more prominent minor septa and a less lonsdaleoid dissepimen- tarium. Acknowledgements. The Canadian specimens described in this paper were collected by geologists working for Triad Oil Co. Ltd., Calgary, Alberta, and were originally studied by the writer while employed by the same company. The management of the company has consented to the publication of this work and has kindly allowed the type specimens to be transferred to the collection of the Geological Survey of Canada. B. Whan of the technical staff of the Department of Geology of the University of New England drafted text-fig. 1. REFERENCES belyea, h. r. and mclaren, d. j. 1962. Upper Devonian formations, southern part of Northwest Territories, northeastern British Columbia, and northwestern Alberta. Pap. geol. Surv. Can. 61-29. birenheide, r. 1962. Revision der koloniebildenden Spongophyllidae und Stringophyllidae aus dem Devon. Senckenbergiana lethaea, 43, 41-99, pi. 7-13. crickmay, c. h. 1960. Studies of the western Canada Stringocephalinae. J. Paleont. 34, 874-90, pi. 113-15. 1962. New Devonian fossils from western Canada. Calgary. 16 pp. 9 pis. fenton, c. l. and fenton, m. a. 1924. The stratigraphy and fauna of the Hackberry Stage of the Upper Devonian. Contr. Mus. Geol. Univ. Mich. 1. B 6612 T 1 2 628 PALAEONTOLOGY, VOLUME 8 flower, r. h. 1961. Montoya and related colonial corals. Mem. New Mexico State Bur. Mines, 7, 1-97, pi. 1-52 (in atlas). frech, f. 1885. Die Korallenfauna des Oberdevons in Deutschland. Z. dtsch. geol. Ges. 37, 21-130, pi. 1-11. Johnson, j. g. 1962. Lower Devonian-Middle Devonian boundary in central Nevada. Bull. Amer. Ass. Petrol. Geol. 46, 542-6. lebedew, n. 1902. Die Bedeutung der Korallen in den devonischen Ablagerungen Russlands. Mem. Comite geol. Russie, 17, 137-80, pi. 1-5. mclaren, d. j. and mountjoy, e. w. 1962. Alexo equivalents in the Jasper region, Alberta. Pap. Geol. Surv. Can. 62-23. norford, b. s. 1962. Columnaria pax (Smith) and the Silurian Columnaria Columbia n, sp. from British Columbia. Bull. geol. Surv. Can. 92, 25-30, pi. 9-12. pedder, a. e. h. 1965 (in press). Upper Devonian tetracorals from the Canadian Rocky Mountains. Bull. geol. Suvr. Can. sloss, l. l. 1939. Devonian rugose corals from the Traverse Beds of Michigan. J. Paleont. 13, 52-73, pi. 9-12. — — and laird, w. m. 1945. Mississippian and Devonian stratigraphy of northwestern Montana. Oil Gas Invest. Chart, U.S. geol. Surv. 15. smith, s. 1945. Upper Devonian corals of the Mackenzie River region, Canada. Spec. Pap. geol. Soc. Amer. 59. soshkina, e. d. 1952. Key to the Devonian tetracorals (in Russian). Trans. Paleont. Inst., Akad. Nauk S.S.S.R. 39. stumm, e. c. 1937. The Lower Middle Devonian tetracorals of the Nevada Limestone. J. Paleont. 11, 423-43, pi. 53-55. 1948. Upper Devonian compound tetracorals from the Martin Limestone. Ibid. 22, 40-47, pi. 10-13. A. E. H. PEDDER, Department of Geology, University of New England, Armidale, N.S.W., Manuscript received 17 March 1964 Australia ARCOSCALPELLUM COMPTUM (WITHERS), A SPECIES OF CIRRIPEDE NEW TO THE GAULT by J. S. H. COLLINS Abstract. A species of cirripede, Arcoscalpellum comptum (Withers), hitherto known only from terga and carinae from the Lower Greensand, is here recorded from the Gault. Further valves are attributed to this species and the present known vertical and geographical distribution is tabulated. Withers (1910) founded his species Scalpellum comptum on two terga (I 13403 — holo- type, and 1 13404 — paratype, in the collection of the British Museum (Natural History) ) from the Lower Greensand of Sevenoaks, Kent. In 1935 he referred it to the genus Arcoscalpellum Hoek, and in 1945 described further terga and two carinae (C. W. and E. V. Wright Coll.) from the ‘Crackers’ of Atherfield, Isle of Wight. A comparison of these specimens with valves from the Gault collected by the author reveals that they belong to the same species. Although there are distinct differences between the terga (the commonest valves found) of A. comptum and those of A. arcua- tum (Darwin), a well-known Gault species, their superficial similarity is undoubtedly the reason why the presence of A. comptum in the Gault has been overlooked until now. The scutum and upper latus, here described, are sufficiently distinct from the corre- sponding valves of A. arcuatum and any other known Gault species, yet so resemble in structure and ornament the terga and carinae of A. comptum that they may with certainty be included within that species. DESCRIPTION 1910 Scalpellum comptum Withers, p. 153, fig. 5. 1935 Scalpellum ( Arcoscalpellum ) comptum Withers, p. 202, pi. xxiii, figs. 7, 8. 1945 Arcoscalpellum comptum (Withers), p. 552, pi. ii, figs. 1, 2. Terga. (Plate 90, figs. 1-5.) All the terga examined from the Gault agree closely with the types and have, as Withers (1935) states, a prominent ridge extending from the apex to the middle of the scutal margin, on either side of which the valve is almost smooth or with only weakly developed longitudinal ridges. The number of these ridges on the carinal side and on the extreme occludent side of the valves varies slightly among indi- viduals. The length of the valve slightly exceeds twice its width. The terga closely resemble those of A. arcuatum (Darwin), but may at once be dis- tinguished by the smaller number of longitudinal ridges. Moreover, while in A. arcuatum these ridges are sometimes bifurcated, and shorter finer ridges which do not extend to the apex are always intercalated, the holotype of A. comptum (I 13403) and all the terga examined from the Gault have uninterrupted ridges that pass continuously from the apex to the margin of the valve. In the paratype of A. comptum (I 13404), however, two of the ridges on the carinal side are bifurcated. The terga of both species have the same size range and the inner side is similar. [Palaeontology, Vol. 8, Part 4, 1965, pp. 629-33, pi. 90.] 630 PALAEONTOLOGY, VOLUME 8 Carina (Plate 90, figs. 6a, b, 7). The specimens figured by Withers from the ‘Crackers’ of Atherfield are embedded in matrix and Withers was of the opinion (1945) that the intraparietes, which are hidden, were inflected at right angles, causing the upper part of the valve to be solid. A further small carina (21512, C. W. Wright Coll.) from the ‘Crackers’ of Atherfield, and the author’s specimens from the Gault show that this is not so. In these a weakly developed ridge separates the parietes from the intraparietes, which are narrow, being at their widest part equal in width to the parietes. They are short and sometimes faintly longitudinally striated, and are inflected inwards to about 45°, the valve remaining open almost to the apex. The inner margin is thickened. The Gault specimens examined are larger than those recorded from the ‘Crackers’ and range in length from 9 to 13 mm., and are about four times longer than wide. The carina is distinguished from that of A. arcuatum by the rounded tectum, the laterally visible intraparietes and the general absence of ridges. Scutum (Plate 90, figs. 10, 11). The scutum, which is trapezoidal in outline, is about twice as long as wide. It is moderately convex transversely, the convexity being slightly accentuated towards the occludent margin. The apex is acute and somewhat produced. The apico-basal ridge is moderately developed and becomes splayed out a little towards the base ; its margins are rounded throughout its length and it is slightly produced at the basi-lateral angle. The basal margin may be either convex or slightly concave, and at right angles to the lower part of the occludent and lateral margins. The rostral angle is very slightly produced downwards. The occludent margin is regularly convex and there is a shallow depression parallel to its edge. The tergal margin, which is short, may be either slightly convex or concave and the edge is slightly arched. The tergo-lateral angle is barely produced and is inclined with the lateral margin at about 40°. (In A. arcuatum this angle averages 36°.) There is a lateral depression parallel to the tergal margin. The lateral margin is regularly convex, except close to both the tergo-lateral and basi-lateral angles, where it is slightly excavated. Generally there are longitudinal ridges on the lateral side and a few on the extreme part of the occludent side of the valve. Other valves may be almost smooth, although there is a tendency for the ridges on the extreme lateral side to remain. None of the valves examined shows intercalated or bifurcated ridges. EXPLANATION OF PLATE 90 All specimens, except Fig. 1 1 , are from the Upper Albian, H. orbignyi subzone (between Beds 50 and 64 of Milbourne 1963), Rugby Portland Cement Co., Ford Place, Wrotham, Kent. Fig. 11 is from the Middle Albian, D. cristatum subzone, of the same locality. All photographed specimens x3. Specimens deposited in Palaeontological Department, British Museum (Natural Flistory). Figs. 1-11, 18. Arcoscalpellum comptum (Withers). 1, Tergum, left, In64248. 2, Tergum, right, In64249. 3, 4, 5, Terga, left, In64250-2. 6a, b, Carina, outer and side views, In64253. 7, Carina, outer view, In64254. 8, Upper latus, left, In64255. 9, Upper latus, right, In64256. 10, Scutum, left, In64257. 11, Scutum, right, In64258. 18, Carina, transverse section x5, 8 mm. from apex, just below maximum development of the intraparietes, In64265. Figs. 12-17, 19. Arcoscalpellum arcuatum (Darwin). 12, Tergum, left, bifurcated and intercalated longitudinal striae clearly seen on occludent side of valve, In64259. 13 a, b, Carina, outer and side views, In64260. 14, Upper latus, left, In64261. 15, Upper latus, right, In64262. 16, Scutum, left, In64263. 17, Scutum, right, In64264. 19, Carina, transverse section x5, 7-5 mm. from apex, In64266. Palaeontology, Vol. 8 PLATE 90 COLLINS, Gault cirripede J. S. H. COLLINS: ARCOSCALPELLUM COMPTUM (WITHERS) 631 On the inner surface a prominent pit, inclined towards the occludent margin, extends above the adductor muscle pit to the apex. As growth advances a pit, bounded by a sharp ridge extending along the tergal margin, is developed. Upper latus (Plate 90, figs. 8, 9). The valve is thin and subtriangular in outline. It is slightly inclined towards the scutum, almost flat transversely, and slightly convex longi- tudinally. The umbo is slightly produced with a thickened ledge, which is more pro- nounced on the scutal side, and which thins out towards the lateral angles. The tergal margin is slightly convex and is shorter than the scutal margin, which is moderately concave. The angle enclosed by two lines drawn from the apex to the lateral angles is approximately 65°. The basal margin is boldly convex ; the portion occupying the dis- tance from a poorly developed apico-basal ridge to the sharp basi-tergal angle, is almost straight. The basi-scutal angle is bluntly rounded. The portion of the valve on the tergal side of the apico-basal ridge is slightly raised and a pronounced ridge extends parallel to the scutal margin. A few very fine uninterrupted ridges complete the surface ornament. The growth-lines are distinct and present a slightly overlapping appearance. On the thickened edges of the scutal and tergal margins the growth-lines are upturned, forming fine longitudinal ridges. CONCLUSIONS Withers (1910) compared the terga of A. comptum with those of A. arcuatum and A. fossula (Darwin) and, with knowledge of the carina, further implied (1945) that A. comptum was the ancestor of A. arcuatum. With better preserved specimens from the Gault, it is possible to advance these opinions further. As stated above, the length of the terga of A. comptum slightly exceeds twice the width, a feature approaching the valves of A. angustatum (Geinitz) of the A. fossula group. The terga of A. arcuatum are generally just twice as long as wide. The ridge extending from the apex to the scutal margin is, to a greater or lesser degree, present in both groups (contrary to Withers’s statement (1910) that it is never present in A. arcuatum). The presence of small, laterally visible intra- parietes in the carina of A. comptum, together with the rounded tectum devoid of a median ridge in all but the earliest stage of growth, are both charac- teristic of the A. fossula group. The intraparietes of A. arcuatum, which are turned under, are not visible in the lateral view, and the tectum is strongly arched with a median ridge. The outline of the scutum is similar to that in text-fig. 1. Evolution of Arcoscalpellum both groups, but the valve resembles that of A. comPtum (Withers^ and allied sPecies* angustatum in being nearly devoid of ridges. The upper latus is flatter, thinner, and smoother than that of A. arcuatum and in this 632 PALAEONTOLOGY, VOLUME 8 respect approaches that of A. angustatum. The concave scutal margin and broadly rounded basi-scutal angle also suggest that this valve more nearly approaches the A. angustatum form than the A. arcuatum. The latest known occurrence of A. comptum in the Upper Gault, H. varicosum subzone, overlaps the lowest known record of A. angustatum (Withers 1935, and speci- mens in author’s collection). The earliest record of A. arcuatum is from the A. intermedins subzone, from where it continues at least to the H. varicosum subzone. From the evidence thus far available, it would seem that A. comptum was more prob- ably the ancestral form of the A.fossula group, as originally suggested by Withers (1910) rather than that of the A. arcuatum group. Both groups presumably arose from a common ancestor, yet to be found. table 1. Vertical and geographical distribution of Arcoscalpellum comptum (Withers) in the Gault. Subzones H. varicosum H. orbignyi D. cristatum A. daviesi E. nitidus D. doris/ E. neglectus D. niobe A. intermedius H. spathi Ford Place, Kent Reservoir Pit, {Rugby Portland Ford Place, Folkestone, Cement Co.) T Kent Kent CLST CLST T T ST Greatness Lane, Kent T CST T T CT CT T C: carinae, L: latera, S: scuta, T: terga. Subzones after Milbourne 1963. Small Dole, Sussex T Range (Table 1). The earliest valves of A. comptum so far recorded from the Gault are terga and a carina collected from the A. intermedius subzone of Ford Place, Wrotham, Kent, and it is evident that the upward range extends at least as far as the H. varicosum subzone. In some subzones the tergum (the commonest valve found) occurs fairly fre- quently. For example, from the lower part of the H. orbignyi subzone (Beds 50-64 of Milbourne 1963) of Ford Place, one tergum of A. comptum was collected to about ten of A. arcuatum. The greatest development of A. comptum appears to have occurred in the H. orbignyi subzone. Acknowledgements. I wish to thank the Officers of the British Museum (Natural History) for providing access to specimens ; Mr. C. W. Wright for the opportunity to examine his specimens and for valuable advice; Mrs. J. I. Collins, Messrs. A. T. Buffery, R. A. Milbourne, and N. Stevens for assistance in the field; and Mr. E. Kentish for preparing the photographs. REFERENCES darwin, c. r. 1851. A monograph on the fossil Lepadidae, or pedunculated cirripedes of Great Britain. Palaeontogr. Soc. ( Monogr .). milbourne, r. a. 1963. The Gault at Ford Place, Wrotham, Kent. Proc. Geol. Assoc. 74,55-80. J. S. H. COLLINS: ARCOSCALPELLUM COMPTUM (WITHERS) 633 withers, t. h. 1910. Some new species of the cirripede genus Scalpellum from British Cretaceous rocks. Geol. Mag. 7, 151-9. 1935. British Museum Catalogue of Fossil Cirripedia in the Department of Geology. 2, Cretaceous, xiii-433, 50 pis. London. 1945. New Cretaceous cirripedes and crab. Ann. Mag. Nat. Hist. (2) 12, 552-61, 2 pis. j. s. H. COLLINS, 63 Oakhurst Grove, East Dulwich, Manuscript received 17 October 1964 London, S.E. 22 AN ALMOST COMPLETE SKULL ROOF AND PALATE OF THE DIPNOAN DIPNO RHYNCHU S SUSSMILCHI (ETHERIDGE) by K. S. W. CAMPBELL Abstract. A preliminary account is given of the cranial structure of Dipnorhynchus sussmilchi (Etheridge) based on a newly found topotype from the Middle Devonian of Taemas, New South Wales. U p till now the Devonian Dipnoan genus Dipnorhynchus Jaekel has been known from two specimens only, one D. lehmanni Westoll (Lehmann and Westoll 1952; Lehmann 1956) from the Lower Devonian Hunsriickschiefer of Germany, and the other D. sussmilchi (Etheridge 1906; Hills 1933, 1941) from the Middle Devonian of Taemas, New South Wales. The genus is of particular interest since it is the oldest known member of the Order Dipnoi, and the structure of its skull table displays many primitive features which have been of value in comparative studies of the Dipnoi and the Rhipistidia, as well as in studies of dipnoan evolution (Westoll 1949). A new specimen of D. sussmilchi has been discovered in the ‘ Spirifer yassensis Lime- stone’ at the locality known as Shearsby’s Wallpaper, Taemas, Burrinjuclc Dam, close to the locality at which the original specimen was obtained. The skull roof is almost complete, and is only slightly distorted. The only parts missing are the bones 4+X and the postero-lateral tips of the bones H (for terminology see Westoll 1949). The bones of the snout, which are covered by a layer of cosmine obliterating the sutures in the other two specimens, can be seen with great clarity. There are no cheek bones. Most of the palate is present, and a large part of the quadrate ramus appears to be embedded in the matrix on one side. Little has yet been done to prepare the specimen, but from a preliminary study the following conclusions can be drawn : 1 . The relative length of the otic region is even greater than had been anticipated in the reconstructions of Hills (1941, fig. 6). This is a primitive feature in which D. suss- milchi is now known to be very similar to D. lehmanni. 2. This relative increase in length is due to the larger size of bones ‘H’ and ‘I’. ‘ H ’ especially is larger, and in this it differs from the same bone in D. lehmanni. 3. There is no bone ‘A’ as was inferred by both Hills (1941) and Westoll (1949), but the bones ‘ I ’ join medially as Westoll was able to show (by means of X-ray photographs) in the case of D. lehmanni. However, a more or less symmetrical gap in the medial portion of the back of the skull may have been the site of a loosely articulated bone ‘A’. 4. Although the bones ‘B2’ are medially fused in the type specimen (Hills 1941, p. 51), they are divided by a clear suture in this individual. 5. Bones ‘ Jx’, ‘J2\ ‘K’, ‘Lx’, ‘L2’, ‘M’, ‘N’, and ‘Q’ are all present as separate entities as in the type specimen, and show no sign of fusions. There is only one minor difference — the enlargement of ‘L2’ on the right side. [Palaeontology, Vol. 8, Part 4, 1965, pp. 634-7, pi. 91.] Palaeontology, Vol. 8 PLATE 91 CAMPBELL, Dipnorhynchus from Australia K. S. W. CAMPBELL: SKULL ROOL OP DIPNORHYNCHUS SUSSM1LCHI 635 6. There is one lateral line bone anterior to ‘Q’, and three or four lying medially to this one. All these bones are irregularly shaped and as yet have not been named. Another set of unnamed bones — three on the left side and four on the right — lie in front of bone 2. This latter bone is itself split into two elements on the right side. 7. Between the lateral line bones and anterior to B2 there are twenty-nine separate irregularly shaped bones which are only roughly symmetrically arranged. This is by far the greatest number known in any member of the Dipnoi. On their exposed surfaces the bones show the same characters as the others and there is no reason to believe that they are histologically different in any way (cf. Westoll 1949, p. 141). 8. The anterior tip of the snout is very thick and is cosmine covered. It shows the characteristic depressions for the external nares. 9. From the arrangement of the sensory pores both sub- and supra-orbital lateral line canals are inferred to be present and show a pattern similar to that of the type. However, in many places it is not possible to identify the position of the canals with pre- cision, and this must await further preparation of the specimen. The anterior part of the supra-orbital canal is somewhat more flexed than that of the type (compare text-fig. la with Hills 1941, figs. 5-6). The sub-orbital canal appears on the snout in the position indicated on Westoll’s diagram (1949, fig. 5a). It can be traced as a line of very coarse pores running around the under edge of the snout across the top of the external naris and gradually dying out toward the mid-fine. 10. The occipital commissure of the lateral fine system is not clearly marked, but the presence of several large pores near the postero-median corners of the bones ‘I’ suggests that it probably passes out of these bones in this region. This supports the view that a movable bone ‘A’ was originally present in this region, but further preparation is needed. 11. It is difficult to distinguish between fine cracks and pit fines in some areas, but the pit fines shown in text-fig. 1 are definite. 12. The bone of the palate is an enormously thick structure and its edges are knobbled where they were embedded in cartilage. All the bones are fused, and while the vomers can be distinguished from each other and from the pterygoids by emarginations and deep grooves, it is not possible to separate the pterygoids and the parasphenoids. The vomers are short and do not reach to the snout. At the posterior end of the palate the bone thins out rapidly and rises abruptly. There is clearly no parasphenoidal stalk. In general form the palate is very similar to that of D. lehmanni , though in that species Westoll (1952, fig. 4b) claimed to be able (with difficulty) to delineate a narrow parasphenoid. 13. There is no ceratodont dentition. The teeth form three large, slightly elongate bulbous masses. The central one is the most anterior and is itself subdivided into three segments by the remnants of the sutures between the vomers and the pterygoids, and between the two pterygoids themselves. Where the sutures join there is a deep pit. The two lateral teeth are at the antero-lateral margins of the pterygoids and are formed entirely from these bones. Teeth of this shape and the massive palatal bones suggest that EXPLANATION OF PLATE 91 Figs. 1-3. Dipnorhynchus sussmilchi (Etheridge), dorsal, ventral, and lateral views of the skull. The dorsal view was photographed with the long axis horizontal, and the ventral view with the palate horizontal. All approx, x 1 . 636 PALAEONTOLOGY, VOLUME 8 the animal was a shell feeder. The plate figured by Hills (1933, pi. 12, fig. 4) is not a tooth plate of Dipnorhynchus. 14. The associated fauna is very rich and consists mainly of brachiopods, orthoconic cephalopods, high spired gastropods, bivalves, and trilobites. These undoubtedly indicate a marine environment. As has been suggested by Hills (1958) the fine state of text-fig. 1 a-b. Dorsal and dorso-lateral views of the skull prepared by inking in the sutures and the sensory pores on enlarged photographs, and then bleaching them. preservation of the fish fauna in the Murrumbidgee limestones in general indicates that they probably have not been transported far after death. The interpretation of the struc- ture of the teeth and palate given above supports the view that it was a marine organism. Work on the specimen is proceeding and it is intended to produce a more complete account as soon as possible. Note. Throughout this paper the bone nomenclature originated by Foster-Cooper (1937) and developed for Dipnorhynchus by Westoll (1949), has been used without modification for the sake of simplicity. It is not implied that this nomenclature is accepted in its entirety. K. S. W. CAMPBELL: SKULL ROOF OF DIPNO RHYNCHUS SUSSMILCHI 637 All except the posterior bones of the skull were found on the one piece of rock. During a search at a subsequent date these bones were found in place only a few feet from the original site by Mr. R. W. Day of this department. His assistance is grate- fully acknowledged. REFERENCES foster-cooper, c. 1937. The Middle Devonian Fish Fauna of Achanarras. Trans. Roy. Soc. Edinb., 59, 223-39, pi. 1-8. hills, e. s. 1933. On a Primitive Dipnoan from the Middle Devonian Rocks of New South Wales. Ann. Mag. nat. Hist., ser. 10, 11, 634-43, pi. 11-12. 1941. The Cranial Roof of Dipnorhynchus sussmilchi (Eth. fil.). Rec. Aust. Mus. 21, 45-55, pi. 9. 1958. In westoll, t. s. ed., Studies in Fossil Vertebrates. The Athlone Press, London. lehmann, w. m. 1956. Dipnorhynchus lehmanni Westoll, ein primitiver Lungenfisch aus dem rheinischen Unterdevon. Palaont. Z. 30 (1-2), 21-25, pi. 1. westoll, t. s. 1949. In jepson, g. l. et al. Genetics, Palaeontology and Evolution. Princeton Univer- sity Press. 1952. In lehmann, w. m. and westoll, t. s. A Primitive Dipnoan Fish from the Lower Devon- ian of Germany. Proc. Roy. Soc. Lond. (B), 140, 403-21, pi. 24. K. S. W. CAMPBELL Department of Geology, Australian National University, Canberra, A.C.T., Manuscript received 6 November 1964 Australia SOME EUROPEAN PROVIVERRINI (MAMMALIA, DELTATHERIDI A) by LEIGH VAN VALEN Abstract. A generic revision of the Proviverrini is made, with special attention to European material from the middle Eocene. Phylogeny and parallel evolution are discussed. Proviverra is made identifiable; Sinopa , Geiselo- therium, Leonhardtina, and Prorhyzaena are synonymized with it. Specimens of the Miacidae and Oxyaenidae, and one of unknown affinities, are described from the Eocene of Europe. New taxa are the following : Delta- theridia, new order; Arfia, new genus; Prototomus torvidus, Cynohyaenodon trux, Prodissopsalis phonax, P. theriodis, and Miacis ? macintyri, new species. While in Europe in 1963 I had the opportunity to examine material of a number of species of European creodonts, especially middle Eocene proviverrines. The following account, together with another paper (Van Valen, in press) and one by R. J. G. Savage (in press), provides the framework for a revision of the known Old World proviverrines. Early Eocene specimens recently collected by D. E. Russell, D. E. Savage, and P. Louis will provide further information when described. I have elsewhere (Van Valen, in press) reduced the subfamily Proviverrinae to the rank of a tribe in the Hyaenodon- tinae, family Hyaenodontidae, which I include in a new order, Deltatheridia. Because the publication date of that long paper is still uncertain although it was submitted for publication in the spring of 1962, I will give here a brief diagnosis of the Deltatheridia. The drawings of teeth were made on quadrille paper using a reticule in a binocular micro- scope unless otherwise stated. The following abbreviations are used: BM, British Museum (Natural History), Department of Palaeontology; MME, Museum fur Mitteldeutsche Erdgeschichte, Halle/Saale; NMB, Naturhistorische Museum, Basel. Acknowledgements. I am indebted to a number of people for hospitality and access to collections during this study. These are especially Prof. H.-W. Matthes of Halle, Dr. H. Schaefer and Prof. J. Hiirzeler of Basel, Dr. D. E. Russell of Paris, M. P. Louis of Reims, Dr. A. J. Sutcliffe of London, and Dr. R. J. G. Savage of Bristol. Part of this study was supported by a NATO post-doctoral fellowship. Order deltatheridia, nov. Diagnosis. Insectivorous or carnivorous placental mammals usually with the following characters: tall cusps on the teeth; distinct paraconid; relatively close paracone and metacone; relatively wide stylar shelf; persistent embrasure shear on the molars, which have a prominent metastyle and no hypocone ; no interradicular crests ; premolars not molariform; fissured ungual phalanges; primitive placental pattern of internal carotid circulation ; alisphenoid canal present. Included groups. The order is based equally on the Palaeoryctidae (including Delta- theridiidae), Hyaenodontidae (including Limnocyonidae), and Oxyaenidae, with ordinal rank separate from the Insectivora because of the inclusion of the latter two families and separate from the Carnivora because close phyletic affinity with the Miacidae is improbable and predation clearly originated independently in the two groups. The Didy- [Palaeontology, Vol. 8, Part 4, 1965, pp. 638-65.] L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 639 moconidae are also included, and the Micropternodontidae and some or all zalambdo- donts, especially the Tenrecidae (including Potamogalidae and Apternodontidae) and Chrysochloridae, may also fall here. The origin of the Oxyaenoidea from the Palaeoryc- tidae is shown by Van Valen (in press), where other phylogenetic and anatomical mat- ters are also discussed. GENERA AND SPECIES OF PROVIVERRINI The following genera and species have previously been described or are described in the present paper. Taxa I regard as of uncertain validity are preceded by a question mark ; type species of genera considered valid are preceded by an asterisk. Only papers giving original information or interpretations are cited. IPrototomus Cope 1874, p. 13 [? = Protoproviverra (not Protoviverra) Lemoine 1891 (May), p. 272; not Protoproviverra Ameghino 1891 (August), which was replaced by Amphiproviverra Ameghino 1891, and is now included in Thylacodictis Mercerat, 1891], validation of name pending. * Prototomus viverrinus Cope 1874, p. 13. Early Eocene, Wasatchian; New Mexico and Wyoming. See also Cope (1877, pp. 109, 112), Matthew (1901, p. 27), Matthew (1909, p. 470), Matthew (1915, p. 83), McKenna (1960, p. 92), Gazin (1962, p. 54), Van Valen (in press). Validation of the name is pending. Prototomus multicuspis (Cope 1875, p. 10). Early Eocene, Wasatchian; New Mexico and Wyoming. See also Cope (1877, pp. 112, 116), Matthew (1901, p. 27), Matthew (1915, p. 80), Gazin (1962, p. 55). ? Prototomus vulpecula (Matthew 1915, p. 80). Early Eocene, Wasatchian; Wyoming and Colorado. See also Gazin (1952, p. 53), Robinson (1960, p. 1), Gazin (1962, p. 55). Prototomus mordax (Matthew 1915, p. 73). Early Eocene, Graybullian; Wyoming. See also Denison (1938, p. 241), Van Valen (in press). ? Prototomus secundaria (Cope 1875, p. 9). Early Eocene, Wasatchian; New Mexico, ?Wyoming. See also Cope (1877, p. 115), Matthew (1901, p. 27), Matthew (1915, p. 82), McKenna (1960, p. 91), Van Valen (in press). Prototomus palaeonictides (Lemoine 1880, p. 5) [= Protoproviverra pomelii Lemoine (1891, p. 265), based on the same specimen]. Early Eocene, late Ypresian; France. See also Teilhard (1921, p. 50 or 58), Louis (1963, p. 87), Van Valen (in press). Prototomus torvidus, new species. Middle Eocene, Lutetian ; Switzerland. Arfia, new genus. * Arfia opisthotoma (Matthew 1901, p. 28). Early Eocene, Graybullian; Wyoming. See also Matthew (1915, p. 73). ? Arfia shoshoniensis (Matthew 1915, p. 73). Early Eocene, Graybullian; Wyoming. Tritemnodon Matthew (1906, p. 25). Tritemnodon strenua (Cope 1875, p. 10). Early Eocene, Wasatchian; New Mexico and Wyoming. See also Cope (1877, p. 117), Matthew (1901, p. 26), Matthew (1915, p. 74), Gazin (1952, p. 54), Gazin (1962, p. 56). Tritemnodon hians (Cope 1877, p. 118). Early Eocene, Wasatchian; New Mexico and Wyoming. See also Matthew (1901, p. 25), Matthew (1915, p. 75). Tritemnodon whitiae (Cope 1882, p. 161). Early Eocene, Lostcabinian ; Wyoming. See also Cope (1884, p. 292), Matthew (1901, p. 26), Matthew (1906, pp. 207, 209), Matthew (1915, p. 84). *Tritemnodon agilis (Marsh, 1872, p. 204) [= Stypolophus brevicalcaratus Cope (18726, p. 469); = Sinopa gracilis Wortman (1902, pp. 17,201)]. Middle Eocene, Bridgerian; Wyoming. See also Cope (1884, p. 291), Matthew (1901, p. 24), Wortman (1902, p. 437), Matthew (1906, pp. 207, 209), Matthew (1909, p. 475), Thorpe (1923, p. 37). Proviverra Riitimeyer 1862, p. 80 [= Sinopa Leidy 1871, p. 116; = Stypolophus Cope 1872a, p. 1; 640 PALAEONTOLOGY, VOLUME 8 = Prorhyzaena Rutimeyer 1891, p. 105; - Leonhardtina Matthes 1952, p. 223; = Geiselotherium Matthes 1952, p. 225], Proviverra rapax (Leidy 1871, p. 116). Middle Eocene, Bridgerian; Wyoming. See also Wortman (1902, p. 435), Matthew (1906, pp. 207, 209), Matthew (1909, pp. 464, 465, 471). IProviverra rapax lania (Matthew 1909, p. 472). Middle Eocene, Bridgerian; Wyoming. IProviverra grangeri (Matthew 1906, p. 206). Middle Eocene, Bridgerian; Wyoming. jProviverra pungens (Cope 1872a, p. 1). Middle Eocene, Bridgerian; Wyoming. See also Cope (1884, p. 291), Matthew (1901, p. 24), Wortman (1902, p. 436), Matthew (1906, p. 209), Matthew (1909, p. 472). IProviverra minor (Wortman 1902, p. 17). Middle Eocene, Bridgerian; Wyoming. See also Matthew (1906, pp. 207, 209), Matthew (1909, p. 473), Thorpe (1923, p. 35). If this species is shown to be valid and if both it and Filhol’s Cynohyaenodon minor prove referable to Proviverra, then Wort- man’s species must be renamed as it would be a secondary homonym. Probably at least one Bridgerian species other than P. rapax is valid, but which this (or these) may be is now uncertain. ^.Proviverra major (Wortman 1902, p. 18). Middle Eocene, Bridgerian; Wyoming. See also Matthew (1906, pp. 207, 209), Matthew (1909, p. 473), Thorpe (1923, p. 36). *Proviverra typica Rutimeyer 1862, p. 80 [= Prorhyzaena egerkingiae Rutimeyer, 1891, p. 105]. Middle Eocene, Lutetian; Switzerland. See also Schlosser (1887, p. 213), Rutimeyer (1891, p. 102). Proviverra gracilis (Matthes 1952, p. 223) [= Geiselotherium pilzi Matthes 1952, p. 225]. Middle Eocene, Lutetian; Germany. Proviverra minor (Filhol 1877, p. 28). Late Eocene, Ludian; France. See also Schlosser (1887, p. 216), Filhol (1888, p. 191), Martin (1906, pp. 417, 420, 421). Cynohyaenodon Filhol 1873, p. 87 [= Pseudosinopa Deperet 1917, p. 172]. Cynohyaenodon trux, new species. Middle Eocene, Lutetian; Switzerland. Cynohyaenodon ruetimeyeri (Deperet 1917, p. 172). Middle Eocene, Lutetian; Switzerland. * Cynohyaenodon cayluxi (Filhol 1873, p. 87). Late Eocene, Ludian; ?Early Oligocene, ?Sannoisian; France. See also Filhol (1877, p. 7), Gaudry (1878, p. 20), Cope (1884, p. 289), Schlosser (1887, p. 217), Filhol (1888, p. 190), Smith (1902, p. 400), Martin (1906, pp. 420, 422), Matthew (1906, p. 209), Edinger (1929, p. 151), Russell and McKenna (1962, p. 277). Cynohyaenodon leenhardti Martin 1906, p. 420. Late Eocene, Ludian; France. Prodissopsalis Matthes 1952, p. 206 [= Imperatoria Matthes 1952, p. 214]. *Prodissopsalis eocaenicus Matthes 1952, p. 206 [= Imperatoria gallwitzi Matthes 1952, p. 214; = Imperatoria hageni Matthes 1952, p. 220; = Geiselotherium robustum Matthes 1952, p. 228], Middle Eocene, Lutetian; Germany. Prodissopsalis phonax, new species. Middle Eocene, Lutetian; Switzerland. Prodissopsalis theriodis, new species. Middle Eocene, Lutetian; Switzerland. Paracynohyaenodon Martin 1906, p. 424. *Paracynohyaenodon schlosseri Martin 1 906, p. 426. Late Eocene, Ludian ; France. IMetasinopa Osborn 1909, p. 423. Metasinopa ethiopica (Andrew 1906, p. 233). Late Eocene, lower Fluvio-marine beds; Egypt. See also Schlosser (1911, p. 77), Van Valen (in press). * Metasinopa fraasi Osborn 1909, p. 423. Early Oligocene, upper Fluvio-marine beds; Egypt. See also Schlosser (1911, p. 77). Propterodon Martin 1906, p. 455. *Propterodon morrisi (Matthew and Granger 1924, p. 1) [= P. irdinensis Matthew and Granger 1925, p. 4]. Late Eocene, Irdin Manha Formation; Mongolia. See also Van Valen (in press). cf. Propterodon Propterodon minutus (Douglass 1901, p. 19). Early Oligocene, Chadronian; Montana. See also Matthew (1903, p. 208), Scott and Jepsen (1936, p. 53), Van Valen (in press). L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 641 Dissopsalis Pilgrim 1910, p. 64. *Dissopsalis carnifex Pilgrim 1910, p. 64 . Early Pliocene, Chinji beds, West Pakistan. See also Pilgrim (1914, p. 267), Pilgrim (1932, p. 11), Colbert (1933, p. 1), Colbert (1935, p. 75). ? Dissopsalis ruber Pilgrim 1910, p. 64. Locality and references as for D. carnifex. Quercitherium (not Quercytherium) Filhol 1880, p. 48. * Quercitherium tenebrosum Filhol 1880, p. 48. Late Eocene, Ludian; France. See also Filhol (1882, p. 30; reprint of Filhol 1880), Filhol (1892, p. 135), Martin (1906, p. 429), Deperet (1917, p. 165), Piveteau (1935, p. 84), Russell and McKenna (1961, p. 277); Van Valen (in press). The following genus and species have most recently been referred to the Proviverrini (as a subfamily or family) but are here not recognized as such: Galethylax Gervais 1850. The type specimen is of uncertain relationships but is apparently not a pro- viverrine. See Van Valen (in press). Proviverra americanus (sic) Scott 1892, is an artiodactyl (cf. Van Valen, in press). The foregoing classification has a number of innovations, in addition to the new taxa. These may be summarized as follows. The type species of Sinopa is placed in Proviverra and other species are placed in Proviverra and three other genera. Prototomus is provi- sionally revived; it is probably a senior synonym of Protoproviverra. Pseudosinopa is synonymized with Cynohyaenodon , and Prorhyzaena, Geiselotherium, and Leonhardtina are synonymized with Proviverra ; the synonymy of species in Proviverra is also new. Geiselotherium robustum is synonymized with Prodissopsalis eocaenicus, Cynohyaenodon minor is transferred to Proviverra, and Paracynohyaenodon morrisi is made conspecific with Propterodon irdinensis. Propterodon does not include the species on which it was originally based; see below. GENERIC DIAGNOSES The lower dentition is known for all species of proviverrines, but the upper dentition is unknown in a number of species. For this reason I will emphasize characters of the lower teeth. The skull, most of which is unknown in several genera, does not as yet pro- vide characters that are clearly diagnostic at the generic level except for Quercitherium and Dissopsalis, genera that are recognizable at a glance from the teeth. All relative statements used in the diagnoses can be interpreted from published figures of the genera involved. Prototomus, as tentatively used here, is known mainly from American species. It is the most primitive known genus of proviverrines and was probably the stem genus of the entire family Hyaenodontidae. For this reason the allocation of some species to Prototomus ox other genera is arbitrary. Intergradation is particularly evident on the basis of species now known with Tritemnodon, which is, however, generically distinct at least by the middle Eocene. In the lower molars of Prototomus the metaconid is not or is only slightly reduced; it is larger than or equal to the paraconid, and may be considerably larger. The paraconid is not shifted anteriorly. The labial height of the protoconid above the talonid is about the same as the length of the trigonid. The molar talonids are two- thirds to fully as long as the trinonids, and the talonid rim is high or moderately high. The size of the trigonid of Mx is less than or usually equal to that of M2, which is about the same as that of M3 but may be somewhat larger or smaller. The existence and length of diastemata between the canine and P3 are variable. P3 is relatively narrow and high: 642 PALAEONTOLOGY, VOLUME 8 its width is about a third of its length, and its labial height (above the base of the enamel on the roots) is two-thirds to four-fifths its length. The paracone and metacone of the upper molars are more or less close together but are not twinned, and are of about the same size as each other on M1 and M2 (either may be slightly larger). The European species P. palaeonictides is distinguished especially by the low angle (about 60°) that the posterolingual corner of the trigonid of at least Mx makes with the horizontal plane of the tooth, and the fact that the metaconid of at least Mx is about as tall as the proto- conid when the sides of the trigonid are taken as vertical; one or occasionally both of these characters are present also in some American specimens. Arfia has and M2 with structure similar to that of Prototomus, but on M3 the meta- conid is moderately reduced and is somewhat smaller than the paraconid, which has shifted anteriorly to a variable degree. The M3 protoconid height is variable with respect to the trigonid length. The molar talonids vary from as long as the trigonids to a little over half that length ; their rims are high on and M2, but reduced on M3. The trigonid of M3 is much larger than that of M2, which is moderately larger than that of Mx. There are no diastemata or only one short one between the canine and P3. P3 is moderately wide and high; its width is two-fifths to half its length and its height is four-fifths of its length. The paracone and metacone of the upper molars are about as separated as in Prototomus and are of about equal size on M1 and M2. P4 differs from that of all other proviverrines for which the upper dentition is known except Quercitherium in the reduc- tion of the metastyle, which is a weak crest not more than a fifth of the labial length of the tooth. Tritemnodon has lower molars with the metaconid slightly to (middle Eocene) much reduced ; at least on M3 the metaconid is smaller than or equal to the paraconid, and is almost always smaller. The paraconid is not to moderately shifted anteriorly. The labial height of the protoconid above the talonid is greater than or equal to the length of the trigonid, distinctly greater in the middle Eocene. The molar talonids are half to three- fourths as long as the trigonids in the early Eocene, not more than a third as long in the middle Eocene. The talonid rim, moderate to high in the early Eocene, becomes much reduced. The trigonid of M) is smaller than that of M2, which is about the same size as that of M3 but may be slightly larger or smaller. There are moderate diastemata between the canine and P3. P3 is narrow and low: its width is one-third to two-fifths of its length and its height is three-fifths of its length. The paracone and metacone of the upper molars are moderately to (middle Eocene) highly connate ; the paracone is larger than the metacone, considerably larger in the middle Eocene. Proviverra has a lower dentition like that of Prototomus ; the molar talonids are some- times even longer than the trigonids. The only consistent structural difference I have found between the genera, if they are regarded as distinct, is the greater separation of the paracone and metacone in Proviverra. Cynohyaenodon has lower molars with the metaconid slightly to moderately reduced and the paraconid moderately to not shifted anteriorly. The M3 protoconid height is slightly less than the trigonid length. The talonids are two-fifths to two-thirds the length of the trigonids, and have moderately high rims. The trigonids increase moderately and regularly in size from Mx to M3. There are no diastemata in the premolar region. P3 is relatively wide and high: its width is about half its length and its height is about four- fifths its length (three-fifths in C. trux). The paracone and metacone of the upper molars L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 643 are only moderately (Lutetian) to closely appressed and are similar to each other in size on M1 and M2. Prodissopsalis has the metaconid of the lower molars moderately to much reduced, it being smaller than the paraconid, which has shifted far anteriorly. The protoconid of M3 is somewhat lower than the length of the trigonid. The talonids are a third to a half the length of the trigonids, and have low or intermediate rims. There is a moderate and pro- gressive increase in the size of the trigonids from Mx to M3. Diastemata are absent. P3 is relatively wide and low: its width is about half its length and its height is about three- fifths its length. The paracone and metacone of the upper molars are only moderately close together; the metacone is distinctly larger than the paracone on M1 and M2. Paracynohyaenodon has the metaconid of the lower molars only slightly reduced on Mj and M2, moderately reduced on M3. The paraconid is moderately to far anterior; the height of the protoconid of M3 is probably about the same as the trigonid length. The talonids are half to two-thirds of the trigonid length, and have moderate to high rims. The trigonid of Mx is moderately smaller than that of M2, which is much smaller than that of M3. The nature of the anterior and upper dentition is unknown. Metasinopa is not certainly distinct at the generic level from Paracynohyaenodon. The metaconid of the lower molars is much reduced on Mx and M2, and nearly absent on M3. The paraconid is moderately to far anterior; the height of the protoconid of M3 is about the same as the trigonid length. The talonids are a third to a half the trigonid length, and have moderately high rims (at least on M2). The trigonid of Mx is moderately smaller than that of M2, which is much smaller than that of M3. Diastemata are absent. The P3 is wide, its width in P. fraasi being about three-fifths its length. The height of P3 and the nature of the upper dentition are unknown. Propterodon has much reduced metaconids on the molars, and the paraconid is far anterior. The height of the protoconid of M3 is unknown, but by analogy with Mx and M2 would be about the same as the trigonid length, perhaps slightly less. The talo- nid of Mx is about four-tenths the length of the trigonid, that of M2 three-tenths, and that of M3 one-tenth. The talonid rim is moderately high labially and posteriorly, but is ab- sent lingually. There is a considerable but regular increase in the size of the trigonid from Mx to M3. There are no diastemata in the premolar region. P3 is relatively wide and high : its width is half its length and its height is four-fifths its length. The paracone and metacone of the upper molars are unknown. Dissopsalis has the metaconid of Mi moderately reduced, that of M2 much reduced, and that of M3 absent. It should be noted that the only known Mx is from a different individual than the M2 and M3. The paraconid of Mx is only slightly shifted anteriorly, while that of M2 and M3 is far anterior. The protoconid of M3 is incompletely known but was probably considerably lower than the length of the trigonid. The talonid of Mi and M2 is about half the length of the trigonid, while that of M3 is less than a fifth. The talonid rim is high or moderately high on Mx and M2, and nearly absent on M3. The trigonid of M2 is somewhat larger than that of Mx, and that of M3 is considerably larger than that of M2. The lower premolars anterior to P4 are unknown, but by analogy with the maxilla there were probably one or two diastemata. P4 is relatively wide. The para- cone and metacone of the upper molars are not particularly close together; the paracone is much smaller than the metacone on M1 and especially M2. The greatly enlarged anteroposterior shearing surface on M2 is diagnostic. 644 PALAEONTOLOGY, VOLUME 8 Quercitherium is immediately recognizable by its greatly enlarged and bulbous pre- molars, especially the second, which are larger than the molars. The metaconids of M2 and M3 are only slightly reduced; that of Mx has not been described. The paraconids of and M2 are not or are only slightly shifted anteriorly, while that of M3 is moder- ately far anterior. The protoconid of M3 is considerably lower than the length of the trigonid. The talonids of Mx and M2 are two-thirds to three-fourths the length of the trigonid, while that of M3 is more or less a third the trigonid length. The talonid rim is high on at least Mi and M2. There is a moderate and progressive increase in the size of the trigonids from Mx to M3. Diastemata are absent. P3 is tall and very wide : its width is about two-thirds its length and its height is about four-fifths its length. The paracone and metacone of the upper molars are not particularly close together; the paracone is smaller than the metacone on M1 and especially M2. ADAPTATION, PHYLOGENY, AND PARALLELISM Of the genera here included in the Proviverrini, only Dissopsalis, Propterodon, and Metasinopa are derivable from any other genus (except Prototomus) without invoking a reversal and subsequent re-establishment of carnassial specialization. The major cri- terion I use for a genus in proviverrines (aside from probable monophyly) is a significant difference in adaptive specializations from the most recent possibly ancestral genus. Prototomus may be regarded as a primitive proviverrine in all characters. Querci- therium is specialized in its crushing premolars (with concomitant changes in the skull and mandible), with moderate carnassial specialization in the anterior paraconid of M3 and the size of the metacone of M2. Proviverra has somewhat less carnassial specializa- tion than Prototomus. This difference is not great and is perhaps not of generic rank, but it represents a definite trend in opposition to that of all other proviverrines. Pro- viverra minor (Filhol) has crushing premolars, but these are not as large as those of Quercitherium and there are no indications of the carnassial specialization of the latter genus ; in fact the molars of P. minor (Filhol) are nearly inseparable from those of P. typica. It is possible that the European and American species of Proviverra are indepen- dent developments from the stage of Prototomus , but this is simply speculation. Even evo- lutionary classification must follow the dictates of static morphology until reasonable evidence for a phylogeny is available. At least five of the remaining eight genera present an interesting picture of independent adaptations to increase the efficiency of shear between the upper and lower molars. Each of the five has one or more specializations that are not present in any of the other four genera taken individually, although a single specialization may not distinguish a genus from all the other four genera taken together. This pattern is in fact an important reason for my recognizing these groups as genera. Arfia is characterized especially by the enlargement and specialization of M3 without particular carnassial specializations on the other lower teeth. Tritemnodon has tall trigonids with much reduced metaconids and talonids, moderately anterior paraconids, and connate paracone and metacone with the metacone reduced. Cynohyaenodon has only moderately reduced metaconids and talonids and moderately anterior paraconids, but has connate paracone and metacone by the Ludian. Prodissopsalis has reduced metaconids and anterior paraconids as its main specializations. Paracynohyaenodon, L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 645 as yet known from only one specimen, has a relatively anterior paraconid and an en- larged M3; the metaconid of M3 is less reduced than that of Arfia. Metasinopa could have evolved from either Arfia or an early Paracynohyaenodon. It is similar to the latter, differing especially in its more reduced metaconids. The metaconid of M3 of at least most specimens of Arfia is smaller than that of Paracynohyaenodon even in the Gray Bull. Propterodon could have been derived most easily from either Cyno- hyaenodon or Prodissopsalis ; it is advanced in all known respects except that the M3 is not notably enlarged. Dissopsalis seems almost equally well derivable from Prodis- sopsalis (perhaps the most likely candidate), Paracynohyaenodon, Metasinopa, Lutetian species of Cynohyaenodon, or even Arfia. It should, however, be noted that the upper teeth of Paracynohyaenodon and Metasinopa are unknown. The three genera discussed in this paragraph could easily have originated directly from Prototomus via unknown interme- diates. The phylogeny of hyaenodonts is poorly known, and parallelism is so prevalent in demonstrable cases that it very possibly occurred in cases in which it cannot now be shown. This last conclusion is particularly relevant to the Hyaenodontini ( Pterodon , Hyaeno- don, Hemipsalodon, Metapterodon, Ischnognathus, and just possibly Apterodon, a prob- able mesonychid). The Hyaenodontini is defined exclusively on the basis of the degree of carnassial specialization, and it seems probable that the threshold was reached more than once. In particular, it is difficult to see how Metapterodon could have evolved from an ancestor in the Hyaenodontini. Even the genera Pterodon and Hyaenodon are defined largely on the basis of degree of carnassial specialization, and polyphyletic origin of them would not be surprising (but it is not yet proved). The ancestors of the Hyaeno- dontini are unknown, although Prodissopsalis, Cynohyaenodon, or an early Propterodon are probably the most likely candidates among described genera; Metasinopa is also quite possible. One or more species from the middle Eocene of Egerkingen (‘ Hyaenodon ’ schlosseri and others not certainly distinct) are already of the hyaenodontine grade; ‘ H. ’ schlosseri lacks a metaconid of M3 and the talonid is nearly absent on this tooth. It is probable that Tritemnodon, at least by the Bridgerian, was too divergent to give rise to any other known genus. EGERKINGEN PROVIVERRINI The geology of the fissure-fillings of Egerkingen has been summarized by Stehlin (1903, 1905). J. Hiirzeler, who has also collected there, kindly provided some additional information. Three localities have been collected from since 1896; these are designated ot, /3, and y. Quarries a and ft are of late Lutetian age, while y is of early or middle Lute- tian age. The specimens collected before 1896 are largely or entirely from the early and middle Lutetian. I found only one specimen of a proviverrine (of Prototomus torvidus, from P) which was from either a or p ; presumably all the species described here were nearly or quite contemporaneous, despite their number. One or more species of miacid and one or two of a form resembling the Hyaenodontini are also present in Egerkingen y. The latter species are probably not referable to a described genus (cf. Martin, 1906). My schedule, arranged in advance under the belief that there were only about two species of Egerkingen proviverrines, was inflexible and did not permit complete analysis of the available material at Basel. It nevertheless seems desirable to present the information u u B 6612 646 PALAEONTOLOGY, VOLUME 8 obtained. The species were first distinguished on the basis of the upper dentition; the separation so obtained was confirmed by the lower dentition. Upper and lower teeth are referred to the same species on the basis of size, degree of carnassial specialization, and direct occlusal manipulation. There is only one mandible for Prodissopsalis phonax; in all other cases several to many specimens are available for both upper and lower den- titions. Proviverra typica Rutimeyer Text-figs. 1b, 2a, 3a, 4a Proviverra typica Rutimeyer 1862, p. 80. Prorhyzaena egerkingiae Rutimeyer 1891, p. 105. Discussion. This species is the smallest and most abundant Egerkingen proviverrine. It is moderately although not excessively variable; this variation was regarded by Rutimeyer (1891) as indicating the presence of two species and even of two genera. No specimen in the Basel Museum labelled Prorhyzaena gives any indication of the structure of the premolars anterior to P4, and from Riitimeyer’s figure (1891) this has probably always been true despite statements about them in the literature. P4 is not enlarged ; that of the type of Prorhyzaena egerkingiae differs from that of the type of Proviverra typica mainly in the presence of two cuspules rather than one on the lingual lobe. The talonid of Mx is as wide as the trigonid. The widest part of P3 is just posterior of the middle. The alveolar length from the posterior end of the upper canine to the posterior end of P4 is 18-0 and 17-7 mm. on the two sides of the type; to the posterior end of M3 it is 28-3 mm. Riitimeyer’s various figures of the teeth are highly inaccurate. His partly corrected figures of the somewhat crushed skull (Rutimeyer 1862, pi. 5) are better, but the skull was probably somewhat deeper; text-fig. 3 a of the present paper supplements them. The type skull of P. typica now contains the root of the right canine, the base of the crown of the left P1, all of both P4s, both M4s without most of the paracone and metacone and the distal end of the metacrista, the lingual half of the right M2, and the left M3 without the surface of the crown of the labial half. Prototomus torvidus sp. nov. Text-figs. 1a, 4b Type. NMB No. Eh. 522, partial right maxilla. See text-fig. 4b. Type locality. Egerkingen y. Diagnosis and discussion. P. torvidus is slightly more carnassial than Proviverra typica, but the two species may nevertheless be congeneric. Prototomus torvidus is somewhat larger. In the upper molars the main differences are that the protocone of Prototo- mus torvidus is lower and the metacrista is taller than in Proviverra typica ; the paracone and metacone are also closer together and taller. M3 is moderately reduced in Prototomus torvidus but not in Proviverra. The protocone lobe of the P4 of Prototomus torvidus is distinctly shorter anteroposteriorly than that of Proviverra typica, and lacks or nearly lacks cuspules on its lingual rim except for the protocone ; in Proviverra typica there are usually rather prominent cuspules there. The metacrista of P4 is cut off from text-fig. 1 . Lower molars of proviverrines, mainly in occlusal view, a, Prototomus torvidus, NMB unnumbered, right M3, X 6-8. b, Proviverra typica, NMB unnumbered, right M3, x 6-8. c, Proviverra minor (Filhol), NMB No. Q.C. 1028, right M3, x 6-8. d-f, Proviverra gracilis, MME No. 343, x 6-0 d, occlusal view right M2; e, occlusal view right M3; f, lingual view right M3. g, Cynohyaenodon trux, NMB unnumbered, left M3, X 6-8. h, Cynohyaenodon ruetimeyeri, NMB unnumbered, left M3, X 6-8. i, cf. Prodissopsalis sp., MME No. 3987, unerupted right M2, X 6-0. j-k, Paracynohyaenodon schlosseri, type, NMB No. Qu. B. 393, X 6-8. J, right M2; k, right M3. l, Prodissopsalis theriodis, NMB No. Em. 14, right M 3, x6-8. M, Prodissopsalis phonax, NMB unnumbered, left M3, x6-8. In all text-figures dotted lines are inferred, dotted regions are wear surfaces, dashed lines are valleys, large dots are cusp apices, and jagged lines and regions with X’s are broken. 648 PALAEONTOLOGY, VOLUME 8 the paracone by a sharp carnassial notch; they do not more or less merge as in Provi- verra typica. The parastyle of P4 is somewhat wider also, causing the deep notch in the anterior border of the tooth to be about central rather than distinctly on the labial half of the tooth as in Proviverra typica. In the lower molars the talonid and metaconid of Prototomus torvidus are relatively slightly smaller than in Proviverra typica, but the meta- conid is still larger than the paraconid in lingual view. There is a reasonable possibility that Prototomus torvidus belongs to Proviverra even if Prototomus is generically distinct from Proviverra, but it is placed in Prototomus pending clarification of the early Eocene proviverrines of Europe. The specific name is a Latin word meaning savage or wild, in allusion to the presum- ably predaceous habits of the species and also with reference to the work of R. J. G. Savage on the Old World Hyaenodontidae and that of D. E. Savage on the European Eocene. Cynohyaenodon trux sp. nov. Text-figs. 1g, 2b, 3b Type. Unnumbered fragment of rostrum in Naturhistorische Museum, Basel. See text-fig. 3b. Type locality. Egerkingen, quarry unknown. Diagnosis and discussion. The paracone and metacone of the upper molars are much more separate than in C. cayluxi (the upper molars of C. ? leenhardti are unknown), being intermediate in this respect between those of Prototomus torvidus and Proviverra typica. The protocone lobe of the upper molars is slightly shorter anteroposteriorly than in Proviverra typica’, the protocone is relatively low and the metacrista is relatively high and long, as in Prototomus torvidus. The protocone lobe of P4 is distinctly longer anteroposteriorly than that of Prototomus torvidus, being similar in this respect to that of Proviverra typica, but this lobe is narrow transversely and extends anteriorly nearly as far as does the parastyle. The notch in the anterior border of P4 is shallow and central; the nature of the metacrista is unknown. P3 has a third root in at least the type, unlike at least Proviverra typica. The metaconid of M3 is moderately reduced and the paraconid is moderately anterior, and the cusps of the talonid are more or less merged into the talonid rim, unlike the situation in Proviverra and Prototomus. M4 is not quite as tall as Ma. The paraconid of P4 is relatively smaller than that of Proviverra typica but the talonid is relatively larger. The height of P3 is much less than that of M4. P2 is taller than P3; the apex of the protoconid of P2 is about a third of the distance from the anterior end of the tooth. P4 is widest at the talonid; P3 is widest just anterior of the talonid. The alveolar length from the posterior end of the upper canine to the posterior end of P4 on the type of 25-2 mm. The specific name is a Latin word meaning savage or fierce, given for the same reasons as Prototomus torvidus. Cynohyaenodon ruetimeyeri (Deperet) Text-figs. 1h, 2c, 4c Type. The mandible figured by Riitimeyer (1891) in his plate 7, fig. 10, is here designated as the type. It is in the Naturhistorische Museum, Basel. Diagnosis and discussion. The upper molars of C. ruetimeyeri are rather similar to those L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 649 of C. trux, but are somewhat larger and the protocone lobe is slightly smaller. The pro- tocone lobe of P4 is more transverse than in the three Egerkingen species described above, and also slightly more transverse than in C. cayluxi. It does not extend far anteriorly. The notch in the anterior border of P4 is moderately deep and distinctly on the labial half of the tooth. As in C. cayluxi, there is a distinct carnassial notch between the metacrista and the paracone. The embrasure pit (a pit in the maxilla in the embrasure text-fig. 2. P3 of proviverrines, labial view, a, Proviverra typica, NMB unnumbered, right P3, X 6-8. B, Cynohyaenodon trux, NMB unnumbered, left P3, X 6-8. c, Cyno- hyaenodon ruetimeyeri, NMB unnumbered, X 6-8. d, Proviverra gracilis, MME No. 343, left P3, x 6-0. e, Prodissopsalis theriodis, NMB unnumbered, x 6-8. Symbols as in text-fig. 1 . between two cheek teeth) between M2 and M3 is about the same size as that between M1 and M2, which is concordant with the fact that the trigonid of M3 is not much larger than that of Ma. The paraconid of the lower molars is somewhat more anterior than in C. trux, the metaconid is somewhat smaller, and Mi is somewhat more reduced. The paraconid is somewhat further forward and the metaconid slightly smaller than in C. cayluxi. The lower premolars are considerably wider and more robust than in C. trux. P4 is of about equal width at the middle and the talonid ; P3 is widest in the middle or just posterior of the middle. P3 is considerably larger relative to P4 than in C. trux; the height of P3 is almost equal to that of M4. As in C. trux, the height of P4 is greater than that of Mj but less than that of M2. Deperet (1917, p. 172) casually established the genus and species Pseudosinopa ruti- meyeri, relying on Riitimeyer’s figures. The two jaws mentioned as belonging to this 650 PALAEONTOLOGY, VOLUME 8 species are those figured by Rutimeyer (1891) in his plate 7, figs. 10 and 11. Both speci- mens figured there are reversed as to the side of the mouth. On comparison of these jaws with the other specimens of Egerkingen proviverrines, it became evident that text-fig. 3. Proviverrine skulls, not drawn with reticule, a, Proviverra typica, type, ventral view of rostrum with outlines of some alveoli and left P\ M1, and M3; x2-0. B, Cynohyaenodoii trux, type, ventral view of rostrum with PJ, M1, and outline of M2, x 2-0. c-d, Prodissopsalis theriodis, type, NMB No. Em. 1 2, x0-5. Left side and ventral view of left half of skull, with outlines of P-M1 ; additional orbital foramina may also be present. Symbols as in text-fig. 1 . the specimen of Riitimeyer’s fig. 10 is specifically and even generically different from that of his fig. 11, which belongs to Prodissopsalis phonax. As both the names Prodis- sopsalis and Cynohyaenodon are in current use while Pseudosinopa is not, I have res- tricted the type of P. ruetimeyeri to the specimen of the older genus to preserve the current names. The spelling of the specific name is altered because the species was named after Rutimeyer. L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 651 Prodissopsalis phonax sp. nov. Text-figs. 1m, 4d Type. NMB No. Em. 16, partial skull. See text-fig. 4d. Type locality. Egerkingen, quarry unknown. Diagnosis and discussion. The protocone of the upper molars is slightly smaller than that of Prototomus torvidus and Cynohyaenodon trux. The paracone and metacone are only moderately connate. The protocone of P4 and the upper molars is somewhat more anterior and lower and relatively slightly smaller than that of Cynohyaenodon rueti- meyeri. The paracone is slightly smaller relative to the metacone, and these cusps are usually slightly more connate than in C. ruetimeyeri. The embrasure pit between M2 and M3 is distinctly deeper than that between M1 and M2, which corresponds with the fact that the trigonid of M3 is considerably larger than that of M2. Mx is slightly more reduced in comparison to that of C. ruetimeyeri, and M3 is somewhat taller relative to M2. The paraconid is considerably more anterior, the metaconid is somewhat smaller, and the horizontal ramus is somewhat deeper and thicker than in C. ruetimeyeri. In the type skull the postvallum of M2 is almost entirely worn away; most of the meta- cone has been removed in the process. There is virtually no wear on the postvallum of P4 or M1, although the cusps of these teeth are somewhat worn from the apices. This wear pattern emphasizes the importance that M2-M3 shear had attained in Prodis- sopsalis. The skull is apparently somewhat deeper than that of Proviverra. The orbit ex- tends only to the anterior part of M1, not to the posterior part of P4 as in both American and European species of Proviverra. The species Cynodon helveticus Rutimeyer 1862, was founded on an isolated lower molar and has been variously referred to the Canidae and the Miacidae. The specimen in the Basel Museum now labelled as the type belongs to Prodissopsalis phonax. I do not regard the two species as necessarily synonymous, because the tooth mentioned is of a different size and morphology from the type of C. helveticus (or else, as is possible, the published measurement is ten per cent, off and the figure is quite unrecognizable). I have therefore proposed to the International Commission on Zoological Nomen- clature that the name Cynodon helveticus be suppressed. The specific name of P. phonax is a Greek word meaning bloodthirsty, given for the same reasons as Prototomus torvidus. Prodissopsalis theriodis sp. nov. Text-figs. 1l, 2e, 3c-d, 4e-g Type. NMB No. Em. 12, nearly complete skull. See text-figs. 3c-d, 4e. Type locality. Egerkingen, quarry unknown. Diagnosis and discussion. The upper molars differ from those of P. phonax, aside from greater size, in having a somewhat longer metacrista. The upper and lower premolars are broader and more bulbous than those of P. phonax. The only structural differ- ence in the lower molars is that in most specimens the talonid is relatively smaller in text-fig. 4. Upper cheek teeth of proviverrines, mainly in occlusal view. A, Proviverra typica , right P4 (from type), M1 (NMB, unnumbered isolated tooth), M2 (NMB, unnumbered maxilla with P4-M3), and M3 (NMB No. Eh. 554), all X 6-8. b, Prototomus torvidus, type, NMB No. Eh. 522, right P4-M1, X 6-8; P4 is tilted slightly anteriorly, c, Cynohyaenodon ruetimeyeri, NMB unnumbered, right P4-M4, X 6-8. d, Prodissopsalis phonax, type, NMB No. Em. 16, right P2-ML and outlines of M2~3, x 2-0; not drawn with reticule, e-g, Prodissopsalis theriodis ; not drawn with reticule. E, left P^M1 (type, NMB No. Em. 12) and M2 (NMB unnumbered), all x20. f, right M1 (NMB, unnumbered DP4-M‘), x 2-0. G, left M2 (NMB, unnumbered), X 3-2. H-i, Prodissopsalis sp., NMB No. Bchs. 266, right M?1, X 6-8. h, direct posterior view; i, occlusal view. Symbols as in text-fig. 1. L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 653 P. theriodis ; the horizontal ramus of the mandible is, however, relatively much deeper. The skull characters are as for P. phonax; it can be seen in P. theriodis that the infra- orbital foramen is rather high and the basicranium is short relative to these structures in Proviverra. The embrasure pit between M2 and M3 is very deep, in correspondence with the tall trigonid of M3. There is no evidence of a bulla. Martin (1906) recognized that P. theriodis was generically distinct from all named forms, so he proposed the name Propterodon for it. Unfortunately he did not also pro- pose a specific name, and by the vagaries of the rules of nomenclature Propterodon irdinensis is the type species of Propterodon. I believe that Propterodon morrisi (= P. irdinensis ) is generically distinct from Propterodon theriodis, which thereby excludes P. theriodis from the genus originally based on it. The specimen figured by Riitimeyer (1891, pi. 7, fig. 15a) on which Propterodon was largely based, is clearly referable to Prodissopsalis theriodis despite the inaccurate rendition of M3, the only tooth preserved. This inaccurate drawing led Matthew and Granger (1925) to believe that their Proptero- don irdinensis was congeneric with Prodissopsalis theriodis. P. theriodis is different from ‘ Hyaenodon' schlosseri, which is even more advanced. Stehlin (1940) mentions the presence in the late Ypresian Teredinasande (= sables a Unios) of Epernay, of a species related to P. theriodis but of greater size. The name theriodis is a Greek word meaning savage, and is given for the same reasons as Prototomus torvidus. GEISELTAL CARNIVORES The geology and biota of the famous middle Eocene brown coal of Geiseltal have been well discussed by Krumbiegel (1959) and will not be repeated here. In 1952 Matthes described six genera and eight species of carnivores, all as new, from the Geiseltal. Two of the monotypic genera have elsewhere (Van Valen, in press, b ) been transferred to the insectivore family Paroxyclaenidae and shown to represent only one species. I believe that only two genera and species are represented by the remaining named taxa, although as will be shown below there are also other carnivores present in the fauna. The preservation of microscopic details in the Geiseltal material is justly famous but macroscopically the specimens of mammals are usually crushed so badly that, aside from the teeth, only the most obvious features of their anatomy can be determined. The photographs published by Matthes (1952) indicate most of what can be seen of the two species he discussed; the drawings in the present paper are meant only to supple- ment these photographs in regions which would otherwise be obscure. All the known specimens of Gieseltal carnivores are in the Museum fur Mitteldeutsche Erdgeschichte, Halle/Saale. The absence of any species of proviverrines in common between the faunas of Eger- kingen and Geiseltal is somewhat surprising for neighbouring faunas supposedly of the same age. The ecology of a coal swamp is, however, presumably different from that of the Egerkingen fissures. No proviverrine species of either fauna is definitely more advanced than comparable species in the other fauna. The rarity of isolated teeth of small animals in the Geiseltal collection is perhaps due to the fact that the teeth are coloured shiny black like the matrix and are easily overlooked. 654 PALAEONTOLOGY, VOLUME 8 Proviverra gracilis (Matthes) Text-fig. 2d Leonhardtina gracilis Matthes 1952, p. 223. Geiselotherium pilzi Matthes 1952, p. 225. Diagnosis and discussion. Only the lower dentition of P. gracilis is surely known (see below). It is the largest known European species of Proviverra ; the length of M2 of two specimens is 6-1 and 64 mm. The structure of the lower molars is, as far as determinable, the same as that of Proviverra minor Filhol, but the premolars are narrow and not robust as in the latter species. P3 is unusual for European species of Proviverra in that it is relatively low; it lacks a paraconid and talonid. P4 also lacks a paraconid. One upper tooth from Leonhardt Quarry, MME No. 2832, is probably referable to P. gracilis. Its labial length is about 6-8 mm. ; the protocone lobe is missing, as is the junction be- tween the metacone and the metacrista. The paracone and metacone are of about equal size, and are separated to about the same degree as in P. typica. The diagnoses of Leonhardtina and Geiselotherium given by Matthes (1952) do not clearly distinguish between them. The only character in the diagnoses that is apparently different, the relative height of the paraconid and metaconid on the lower molars, is a result of breakage in the type of L. gracilis. I have compared the types with each other and find no reason to regard them as of distinct species. Prodissopsalis eocaenicus Matthes Text-figs. 5b-e Prodissopsalis eocaenicus Matthes 1952, p. 206. Imperatoria gallwitzi Matthes 1952, p. 214. Imperatoria hageni Matthes 1952, p. 220. Geiselotherium robustum Matthes 1952, p. 228. Diagnosis and discussion. P. eocaenicus is most similar to P. theriodis but is considerably larger. In P. eocaenicus the metacrista of the upper molars is somewhat shorter, and the paracone and metacone are slightly more connate. The upper premolars are not clearly different when the total samples are considered except that the metastylar lobe of P3 is somewhat larger in P. eocaenicus. The metaconid of the lower molars is slightly more posterior in P. eocaenicus and perhaps slightly smaller. The paraconid is also smaller and somewhat lower. A paraconid is present on the P3 of P. eocaenicus ; it is absent on the P3 of P. theriodis. The conspecificity of the upper dentitions (P. eocaenicus of Matthes) and the lower dentitions was determined by size, comparison with P. theriodis , morphological occlusal relations, and direct occlusal manipulation (allowing for crushing). The embrasure pit in the maxilla between M2 and M3 is larger than that between M1 and M2, and these are of the relative size expected with the M2 and M3 of the lower dentitions. Almost all the differences between the mandibles and lower teeth mentioned by Matthes (1952) are due to crushing, and the few remaining differences are of the magnitude expected in a relatively invariable species. Metrically, only two measurements could be taken on as many as six specimens : the diameter of the trigonid of M3 from the anterolingual base of the paraconid to the posterolabial base of the protoconid, and the length of P3. L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 655 The coefficients of variation (and the mean in parentheses) for these measurements are 2-8 (11-8) and 74 (12-6) respectively, both based on six specimens from two or three quarries. The upper jaws in fact differ more (in the structure of P3 and in size) than do the lower jaws, but Matthes placed them in a single species and in this I follow him. The degree of molarization of P3 is somewhat variable in P. eocaenicus. In the type (MME No. 94) there is no trace of a protocone, while in MME No. 93 a distinct proto- cone lobe bearing a separate cingulum is present (text-fig. 5c). A third specimen, MME text-fig. 5. A, cf. Oxyaena sp., BM No. Ml 3778, lingual view right M2 x 5-0; not drawn with reticule. b-e, Prodissopsalis eocaenicus, not drawn with reticule, b, left P3, type, occlusal view, x 5-0. c, left P3, MME No. 93, occlusal view, xl-1. d, metastylar lobe of left P4, type, occlusal view, x5-0. e, right M2, paratype, occlusal view, x 5 0. Symbols as in text-fig. 1. No. 3980, is intermediate. Presumably P3 is about at the anterior limit of possible expression of the protocone prepattern. The name Prodissopsalis has been chosen over Imperatoria by Savage (in press). I would have preferred the latter name to avoid phylogenetic commitments, and had in fact chosen it (Van Valen, in press), but Savage’s choice may stand. The Egerkingen species P. theriodis is in fact somewhat more similar to Dissopsalis than is P. eocaenicus. cf. Prodissopsalis sp. Text-fig. li One mandible, described by Heller (1930) as an oxyaenid and not mentioned by Matthes (1952), is not referable to either of the species described above. This specimen is MME No. 3987 ; no information is available as to which individual quarry yielded it. The specimen is one of the two specimens of proviverrines I can recall that have deci- duous teeth; the other is the type of ‘ Paracynohyaenodon ’ morrisi. DP3 and DP4 are 656 PALAEONTOLOGY, VOLUME 8 present, as is the unerupted and broken M2. No evidence is available as to the presence or absence of M3. The paraconid of M2 is slightly lower than the metaconid. The trigonid is relatively higher than that of the Ma of Proviverra gracilis, the protoconid is much longer, and the talonid is somewhat shorter. The paraconid is more anterior and rela- tively longer. The talonid cusps are more prominent than in P. gracilis, and the basin and grooves there are deeper. The prefossid (the depression in the trigonid) is a deep pit, whose bottom slopes only gradually down lingually. There is a possibility that this specimen belongs to the Limocyonini, but such a place- ment is not made here. The ancestors of Thereutherium, the only known European limno- cyonine, are as yet unknown, but it seems improbable that the species to which the present specimen belonged was ancestral because of its considerably greater size. Be- cause it fits readily into Prodissopsalis, although not into any described species, this specimen is tentatively referred to Prodissopsalis. It is somewhat the smallest known species of that genus if correctly allocated; the length of M2 is 7 -2 ±0-2 (approximate 95 per cent, confidence interval) mm., and the trigonid width of M2 is 4-2±0T mm. Miacisl macintyri sp. nov. Text-figs. 6a-c Type. MME No. 3849. Type locality. Geiseltal, quarry uncertain. Diagnosis and discussion. A true miacid, not previously mentioned, occurs at Geiseltal. The type is a compressed skull with the upper and lower jaws nearly in occlusion, to- gether with much of the skeleton. Without further preparation many diagnostic features are obscure, but enough is visible to show that an undescribed and interesting species is present. The number of molars cannot yet be determined, but analogy with other species suggests that three will be found in each jaw. The full completement of four premolars is present. There is no parastyle on P4; P4 has a strongly tilted paracone. The labial cingulum of P4 is incomplete, and the metacrista is highest anteriorly. The parastyle lobe of M1 is moderately strong and projects anteriorly as well as labially. At least a precingulum is present on M1 ; the presence of a postcingulum is uncertain. The paracone of M1 is much larger than the metacone, and overhangs the anterior part of the tooth or nearly does so. The paracone and metacone are rather sharp, as is the centrocrista. The stylar shelf is relatively narrow, and there is no distinct metaconule. M1 is relatively narrow transversely. The length of P4 is 9-8±(H (approximate 95 per cent, confidence interval) mm., the length of the labial border of M1 is 7-8±0-2 mm., and the slant height of the paracone of P4 from the base of the enamel is 6-3 ±0-2 mm. I have compared my notes and drawings of M. ? macintyri with the type of ‘ Viverra ’ hastingsiae (probably the same as ‘ Viverravus ’ angustidens, fide Teilhard 1915), BM No. 30203. The M1 of the latter is more transverse than in M. ? macintyri, the metacone is larger, the stylar shelf is wider, the parastyle projects more labially but less anteriorly, and lingual cingula are nearly absent. In labial view, the angle at the top of the para- cone of M1 is larger than in M. ? macintyri. A definite parastyle is present on P4, although it is weak; the paracone and metacrista are lower relative to the length of P4; and there is a rather large cuspule in the position of a mesostyle, low and posterolabial of the L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 657 paracone. Miacis invictus Matthew and Granger (1925), from the Irdin Manha, is smaller and structurally different in several respects. As far as I can determine, the species most similar to M. ? macintyri is M. ? exilis from the French Phosphorites. On this species see Teilhard (1915). For the upper denti- tion of M. ? exilis only a poor photograph of the lateral view of the skull has been published. Teilhard’s comments, however, refer to the two features that are the most striking in the upper dentition of M. ? macintyri, namely the much reduced metacone and the beak-like parastyle. These points are of course only suggestive without further knowledge of both species. M. ? macintyri is much larger than M. ? exilis. text-fig. 6. A-c, Miacis ? macintyri, MME No. 3849. A, occlusal view right M1, x4-2, not drawn with reticule, b, labial view paracone and metacone of right M1, x 4-2, not drawn with reticule, c, labial view right P4, x60. d-e, Eutheria incertae sedis, MME No. 2833, x6-0. d, occlusal view; e, directly labial view. Symbols as in text-fig. 1 . It is quite possible that M. ? macintyri will not prove to belong to the genus Miacis when better known. It may be noted that the Mx of M. ? exilis differs from that of the American species in having a lower trigonid and a broader talonid. The M1 of at least M. ? macintyri is squat and is small in comparison to P4. Probably relevant miacid mat- erial from Egerkingen is preserved in quantity at Basel. The species name is after G. T. MacIntyre, who is revising the Miacidae. cf. Miacidae, incertae sedis A single upper tooth figured by Heller (1930, pi. 1, fig. 6) appears to be the P4 of a miacid, but it is not closely similar to any P4 known to me. I could not find this tooth when I was in Halle in 1963. Eutheria incertae sedis Text-figs. 6d-e One isolated tooth, MME No. 2833, from Cecilie Quarry, is of peculiar morphology and is not closely similar to any other mammal with which I have compared it. It could be a P4, a DP4, or a molar. The talonid is relatively small and very low; its top is at the 658 PALAEONTOLOGY, VOLUME 8 same level as the base of the enamel on the lingual side of the trigonid. Talonid cusps are nearly absent. The posterior wall of the trigonid slopes far posteriorly, especially on the lingual side. A strong labial cingulum is present by the talonid. The length of the tooth is 4-2 mm., the trigonid width is 2-7 mm., and the slant height of the proto- conid from the base of the enamel on the anterior root is 3-7 mm. When in Halle I regarded this tooth as the molar of a proviverrine descended from Prototomus palaeonictides, but this allocation seems rather improbable. Its least unlikely affinities seem to be as the DP4 of an erinaceoid or the M2 of a miacine. The P4 or DP4 of a paroxyclaenid (Van Valen, in press, b) is also a possibility. REMARKS ON OTHER OXYAENOIDS Proviverra minor (Filhol) Text-fig. lc As noted above, the species described by Filhol (1877) as Cynohyaendon minor is referable to Proviverra. It is easily distinguishable from P. typica by the more bulbous premolars, but the molars are very similar. I could detect no consistent difference what- ever in the upper molars. The lower molars of P. minor are slightly wider than those of P. typica, the talonid of M3 (only) is slightly shorter, the paraconid of P4 is relatively smaller, and the talonid of P4 is slightly less developed lingually. Prodissopsalis sp. One upper molar (NMB No. Bchs. 266) from the Lutetian of Bouxwiller (Buchs- weiler), Alsace, is apparently referable to Prodissopsalis. It is probably a right M1 and represents an undescribed species, being considerably smaller than P. phonax and structurally different as well. The parastyle is larger than in P. phonax, but the most important structural difference from the named species of the genus is in the relatively large size of the protocone, particularly its height. This is a primitive character, as is presumably the small size, and suggests that the Bouxwiller species is the most primitive of its genus. The protocone region is rather similar to that of Oxyaena. The wear pattern is, however, markedly different (cf. Van Valen, in press) and reference to the Proviver- rini seems necessary. The relatively small protocone lobe and the moderate separation of the paracone and metacone indicate reference to Prodissopsalis rather than to Proto- tomus or Cynohyaenodon. The length of the labial border is 5-8 mm., of the anterior border 4-5±0T (approximate 95 per cent, confidence limits) mm., and of the posterior border 6-8 ±0T mm. cf. Oxyaena sp. Text-fig. 5a Three specimens possibly referable to Oxyaena are now known from the Blackheath Beds, probably of earliest Eocene age, at Abbey Wood, England. White (1931) has discussed the geology. The most diagnostic specimen is BM No. Ml 3778, an unworn isolated M2, from which the paraconid is missing. This tooth was figured inaccurately by Cooper (1932), who compared it especially with ‘ Sinopa'. It differs from Wasatchian proviverrines especially by the fact that the hypoconid is slightly lower than the ento- L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 659 conid (cf. Van Valen, in press). The metaconid is about a third of the width of the protoconid. The width of the trigonid is 8-5 mm., the width of the talonid is 6-2 mm., the labial slant height of the protoconid from the base of the enamel on the anterior root is 13-9 mm., and the length (parallel to the base of the enamel) from the carnassial notch of the paralophid to the posterior end of the tooth is 1 TO mm. The second specimen, BM No. M15128, is the unworn protocone lobe of an upper molar, probably the M1 of an oxyaenid. It differs from American species of Oxyaena, however, in having the lingual cingula, which are weak but joined lingually, definitely less than half the height of the protocone. Its anteroposterior length at the conules is 5-8 mm. A third specimen, BM No. Ml 5125, is a (D)P3 or (D)P4 which is worn nearly to the roots. The protocone is weak and slightly anterior of the midline ; there are weak lingual cingula, which are not joined on the lingual side of the protocone. The only known specimen of Argillotherium toliapicum, from the London Clay, has recently been identified (Van Valen, in press) as very probably an immature oxyaenine, cf. Oxyaena. No additional specimens referable to the Oxyaeninae have been reported from Europe (Van Valen, in press). The lower molar of cf. Oxyaena from Abbey Wood is of an appropriate size for the skull of ‘ Argillotherium ’, and is very possibly conspecific with it. BM No. Ml 5125 could be the last permanent or deciduous premolar of this species. BM No. Ml 5128 is, however, considerably too small to fit into the single molar alveolus in the type of ‘ Argillotherium ’, and is probably at least specifically distinct. It is moreover not excluded that all three teeth from Abbey Wood are conspecific with each other and different from 'Argillotherium' . cf. Prototomus sp. The presence of the Proviverrini in England is shown by a fragment of mandible with M3, BM No. M20212, from the Blackheath Beds at Abbey Wood. The tooth is that of a primitive proviverrine. The lingual length is 4-7 mm., the trigonid width is 2-8 mm., the talonid width is T9 mm., and the labial slant height of the protoconid from the base of the enamel on the anterior root is 4-6 mm. It is probable that Didelphisl colchesteri Owen (1846) from Kyson (probably equivalent to or a facies of the Pebble-beds) is also a proviverrine and may represent the same species, although the type of D. ? colchesteri may not be adequate for identification at the species level. On D. ? colchesteri see also Charlesworth (1839), Lyell (1840), and Owen (1840); on the geology of Kyson see Prest- wich (1850), Whitaker (1885), and White (1931). Prestwich (1850) and others ascribe the name D. ? colchesteri to Charlesworth, but I have been unable to find a paper in which he uses this name. Arfia gen. nov. Type species. Sinopa opisthotoma Matthew 1901, p. 28. Referred species. Sinopa shoshoniensis Matthew 1915, p. 73 (possibly a synonym of A. opisthotoma). Diagnosis and discussion. Arfia differs from other proviverrines by the following charac- ters: Mx and M2 unspecialized, metaconid of M3 moderately reduced and paraconid of M3 shifted anteriorly, M3 enlarged, metacrista of P4 much reduced. Further differences have been given above with the comparison of genera. The name is onomatopoeic. Arfia represents a distinct specialization among the Proviverrini and, although possibly 660 PALAEONTOLOGY, VOLUME 8 ancestral to Paracynohycienodon or even Quercitherium, is sufficiently different from these and from Prototomus to necessitate a separate genus. Metasinopa fraasi Osborn 1909 When describing this species, based on one mandible, Osborn (1909) figured a maxilla as possibly conspecific and stated that M3 was absent. That this was probably true in life is indicated by an apparently unbroken vertical surface of bone just posterior to M2. The presence of an M3 in M. fraasi is, however, proved by a strong shear facet on the postvallid (the posterior wall of the trigonid) of M3. Furthermore, the postvallid wear is markedly stronger than the prevallid wear in M. fraasi, while the reverse is true of the maxilla. Therefore the maxilla is not conspecific with the mandible ; the maxilla belongs to a small species of ‘ Hyaenodon’ as this genus is now defined, despite the presence of a nubbin on M1 in the place of a protocone. Propterodon morrisi (Matthew and Granger) Paracynohyaenodon morrisi Matthew and Granger 1924, p. 1. Propterodon irdinensis Matthew and Granger 1925, p. 4. It is clear that the two species of proviverrines described by Matthew and Granger from the Irdin Manha Formation of Mongolia are conspecific. In the type description of P. morrisi the two preserved molariform teeth were identified as M4 and M3, the latter incompletely erupted. Comparison with the specimens referred to P. irdinensis shows, however, that the ‘M3’ of P. morrisi would be a perfect M2 in size and morphology for P. irdinensis (Mx and M3 are known in specimens referred to the latter species) and that the morphology of the mandible is comparable; in particular the posterior mental foramen is under P3 in both species when the present tooth identifications are used. There is space in the type specimen of P. morrisi for an unerupted M3, and the tooth here identified as DP4 is low-crowned and narrow and has divergent roots and an anteriorly placed paraconid, all features that are characteristic of the DP4s of most primitive pla- centals. L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 661 table 1 . Measurements of upper teeth, in millimetres. With three or more specimens the standard deviation of the sample (not of the mean) is given, and the number of specimens is in parentheses. Anterior Posterior Labial Mid- Posterior height Labial height margin margin margin * width metacone paracone Prototomus torvidus P3 — — 5-5 2-1 P4 4-4, 4-3 60, 5-2 5-6, 5-6 4-5, 4-0 — 4-2 M1 4-6, 4-8 6-7, 6-6 5-5, 5-4 — 4-6 — M2 5-9±0-2 (3) 8-1, 6-9 5-4, 5-2 — 4-1, 4-4 — M3 5-4 3-2 2-4, 2-0 — 1-7 — Proviverra typica P4 3 -8 ±0-3 (3) 4-3±0-l (3) 4-5±0-4 (4) 3 -3 ±0-2 (6) — 3-4±0-3 (3) M1 4-8±0-4 (6) 6-3±0-6 (7) 4-6±0-2 (6) — 3-7, 3-0 — M2 5-7, 5-9 7-6, 6-8 5-2, 4-5 — 3-6 — M3 5-9±0-6 (3) 4-1, 3-8 2-4, 2-4 — — — Proviverra minor M1 4-6 5-9 4-3 — — — M2 5-4 6-8 4-5 — 3-2 — M3 6-lt — — — — — Cynohyaenodon trux M1 5-9 8-6 6-3 — 51 — M2 6-8 8-9, 8-9 5-8 — 4-1, 5-1 — Cynohyaenodon ruetimeyeri P4 5-7 7-4, 7-3 6-8 5-6, 5-6 M1 6-0±0-l (4) 9-1, 9-7 6-9, 7-7 — . — — M2 6-5, 7-5 101, 101 6-4, 7-4 — 5-7 — M3 7-8 4-9 30 — — — Prodissopsalis phonax P2 — 7-2 2-5 — — P3 — 6-6, 6-7 3-8, 3-4 — — P4 61 8-7 7-2, 8-1 6-1 — — M1 6-9±0-2 (3) 10-1, 10-9 7-8±0-l (3) — 5-7 — M2 8-7, 8-3 120, 11-6 7-4±0-2 (3) — — — M3 8-8, 8-4 5-7, 6-0 3-0±0-2 (3) — 2-6 — Prodissopsalis theriodis P2 — 8-0 8-7 — — P3 — 4-2 5-8 — — P4 8-5 11-5 11-3 8-5 — — M1 9-6, 8-7 13-0, 12-1 10-8, 9-7 — — — M2 11-2, 9-7 16-4 14-1 — — — * for M3, length at metacone. t approximate 95 per cent, confidence interval: ±0-4. x x 662 PALAEONTOLOGY, VOLUME 8 table 2. Measurements of lower teeth, in millimetres. Mandible depth Lingual Trigonid Talonid below length width width M2 protoconid Prototomus torvidus 7-2* m2 4-9f, 5-3 3-5 3-1 m3 5-0f, 5-8 3-7 2-8 Proviverra typica 6-2 P4 3-6 1-7 — M! 4-3, 4-2 2-3, 2-6 2-3, 2-1 m2 4-5, 4-5 2-8, 2-9 2-5, 2-3 m3 4-5, 4-7 2-8, 2-7 2-2, 1-9 Proviverra minor 8-4 P4 4-6 2-4 — 4-7 3-0 2-8 m2 50 3-4 2-9 m3 50 3-2 2-2 Paracynohyaenodon schlosseri 12-5 7-4 4-3 4-0 m2 8-4 4.9 4-2 m3 90 5-2 3-4 Cynohyaenodon trux 10-6 P2 5-4 2-1 — P3 5-4 2-4 — P4 5-8 2-6 — M1 5-7 3-2 2-7 m2 6-1 3-7 2-8 m3 6-5 3-6 2-2 Cynohyaenodon ruetimeyeri 14-4, 11-5 P3 6-2, 5-5 3-2, 2-3 — P4 7-2 3-6 — Mi 6-3, 6-3 3-5, 3-4 2-9 m2 7-2, 7-0 4-2, 4-2 3-6, 2-8 m3 7-0 3-8 2-4 Prodissopsalis phonax 14-5 Mi 7-1 3-8 3-6 m2 8-1 4.4 3-7 m3 8-8 4-6 3-6 Prodissopsalis theriodis — P2 8-3 4-1 — P3 7-8 4-1 — P4 9.4 4-8 — Mi 8-8 4.9 4-8 m2 9-6, 9-7 5-3, 5-6 5-1, 5-5 m3 11-4, 10-7 6-4, 5-7 3-8, 3-3 Approximate 95 per cent, confidence interval: ±0-4. f From Egerk ingen /3. L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 663 REFERENCES ameghino, f. 1891. Nueves restes de mamlferos fosiles descubiertos por Carlos Ameghino en el Eoceno inferior de la Patagonia austral. Rev. Argentina Hist. Nat. 1, 289-328. Andrews, c. w. 1906. A descriptive catalogue of the Tertiary Vertebrata of the Fayum, Egypt. British Museum (Natural History), London. charlesworth, e. 1 839. Illustrated zoological notices. 1 . On the discovery of a portion of an opos- sum’s jaw in the London Clay, near Woodbridge, Suffolk. Mag. Nat. Hist, (n.s.) 3, 448-51. colbert, e. h. 1933. The skull of Dissopsalis carnifex Pilgrim, a Miocene creodont from India. Amer. Mus. Novitates 603, 1-8. 1935. Siwalik mammals in the American Museum of Natural History. Trans. Amer. Phil. Soc. (n.s.) 26, 1-401. cooper, c. f. 1 932. On some mammalian remains from the lower Eocene of the London Clay. Ann. Mag. Nat. Hist. (10) 9, 458-67. cope, e. d. 1 872«. Second account of the new Vertebrata from the Bridger Eocene. Paleont. Bull. 2, 1-3 ; and Proc. Amer. Phil. Soc. 12, 466-8. 18726. Third account of new Vertebrata from the Bridger Eocene of Wyoming valley. Paleont. Bull. 3, 1-4; and Proc. Amer. Phil. Soc. 12, 469-72. 1874. Report upon vertebrate fossils discovered in New Mexico, with descriptions of new species. Ann. Rept. Chief Engineers, 1874, Appendix FF, 1-18. 1875. Systematic catalogue of Vertebrata of the Eocene of New Mexico collected in 1874. Rept. to Engineer Dept., U.S. Army, Geog. Explor. Surv. West of 100th Meridian (Wheeler Survey), 5-37. 1877. Report upon the extinct Vertebrata obtained in New Mexico by parties of the expedition of 1874. Geog. Surv. West of One Hundredth Meridian (Wheeler Survey), pt. 2, 1-370. Washington. 1882. Contributions to the history of the Vertebrata of the Lower Eocene of Wyoming and New Mexico, made during 1881. Proc. Amer. Phil. Soc. 20, 139-97. 1884. The Vertebrata of the Tertiary formations of the West. Book 1. Rept. U.S. Geol. Surv. Terr. 3, 1-1009. denison, R. H. 1938. The broad-skulled Pseudocreodi. Ann. New York Acad. Sci. 37, 163-256. deperet, c. 1917. Monographic de la faune de mammiferes fossiles du Ludien inferieur d’Euget-les- Bains (Gard). Ann. Univ. Lyon (n.s.), Div. 1, 40, 1-288. douglass, e. 1901. Fossil Mammalia of the White River Beds of Montana. Trans. Amer. Phil. Soc. (N.s.) 20, 237-78. EDiNGER, t. 1929. Die fossilen Gehirne. Ergebn. Anat. Entwickl. 28, 1-249. filhol, h. 1873. Sur les Vertebres fossiles trouves dans les depots de phosphate de chaux du Quercy. Bull. Soc. Philomath. Paris (6) 10, 85-89. 1877. Recherches sur les phosphorites du Quercy. Ann. Sci Geol. 8, 1-340. 1880. Memoire relatif a quelques mammiferes fossiles provenant des depots de phosphorite du Quercy. Bull. Soc. Sci. Phys. Nat. Toulouse 5, 19-156. 1882. Memoires sur quelques mammiferes fossiles des phosphorites du Quercy. Toulouse. 1888. Etude du squelette du Cynohyaenodon. Memoires publies par la societe philomathique a V occasion du centenaire de sa fondation, 179-92. Paris. 1892. Note sur le Quercitherium tenebrosum. Bull. Soc. Philomath. Paris (8) 4, 135-137. gaudry, A. 1878. Les enchainements du monde animal dans les temps geologiques. Mammiferes tertiaires. Paris. gazin, c. l. 1952. The lower Eocene Knight Formation of Western Wyoming and its mammalian faunas. Smithsonian Misc. Coll. 117 (18), 1-82. 1962. A further study of the lower Eocene mammalian faunas of south-western Wyoming. Ibid. 144 (1), 1-98. gervais, f. l. p. 1850. Zoologie et paleontologie frangaises. Paris. (Seen in second edition, 1859.) heller, f. 1930. Die Saugetierfauna der mitteleozanen Braunkohle des Geiseltales bei Halle a. S. Jahrb. Halleschen Verb. Erforsch. Mitteldeutschen Bodensch. Verwert. (n.f.) 9, 13-41. krumbiegel, g. 1959. Die tertiare Pflanzen- und Tierwelt der Braunkohle des Geiseltales. Wittenberg. leidy, J. 1871. Remains of extinct mammals from Wyoming. Proc. Acad. Nat. Sci. Philadelphia, 113-16. 664 PALAEONTOLOGY, VOLUME 8 lemoine, v. 1880. Communication sur les ossements fossiles des terrains tertiares inferieures des environs de Reims. Association Frangaise pour V Avancement des Sciences, Congres de Montpellier, 1879. Reims. (Pp. 1-17 in reprint.) 1891. Etude d’ensemble sur les dents des mammiferes fossiles des environs de Reims. Bull. Soc. Geol. France (3) 19, 263-90. louis, p. 1963. Mammiferes de l’Eocene inferieur des environs de Reims et d’Epernay. Res. Communic. LXXX Cong. Assoc. Frangaise Avanc. Sci., Reims, 1961, 85-90. lyell, c. 1840. On the occurrence of fossil quadrumanous, marsupial, and other Mammalia in the London Clay, near Woodbridge, in Suffolk. Ann. Nat. Hist. 4, 1 89-90. mckenna, m. c. 1960. Fossil Mammalia from the early Wasatchian Four Mile fauna, Eocene of north- western Colorado. Univ. California Publ. Geol. Sci. 37, 1-130. marsh, o. c. 1872. Preliminary descriptions of new Tertiary mammals. Part II. Amer. Journ. Sci. (3) 4, 202-10. martin, r. 1906. Revision der obereocaenen und unteroligocaenen Creodonten Europas. Rev. Suisse Zool. 14, 405-600. matthes, h. w. 1952. Die Creodontier aus der mitteleozanen Braunkohle des Geiseltales. Hallesches Jahrb. Mitteldeutsche Erdgesch. 1, 201-40. Matthew, w. d. 1901. Additional observations on the Creodonta. Bull. Amer. Mus. Nat. Hist. 14, 1-38. 1903. The fauna of the Titanotherium beds at Pipestone Springs, Montana. Ibid. 19, 197-226. 1906. The osteology of Sinopa, a creodont mammal of the middle Eocene. Proc. U.S. Nat. Mus. 30, 203-33. 1909. The Carnivora and Insectivora of the Bridger Basin, middle Eocene. Mem. Amer. Mus. Nat. Hist. 9, 289-567. 1915. A revision of the lower Eocene Wasatch and Wind River faunas. Part I. — Order Ferae (Carnivora), suborder Creodonta. Bull. Amer. Mus. Nat. Hist. 34, 4-103. — — and granger, w. 1924. New Carnivora from the Tertiary of Mongolia. Amer. Mus. Novitates 104, 1-9. . 1925. New mammals from the Irdin Manha Eocene of Mongolia. Ibid. 198, 1-10. osborn, h. f. 1909. New carnivorous mammals from the Fayum Oligocene, Egypt. Bull. Amer. Mus. Nat. Hist. 26, 415-24. owen, r. 1840. Description of mammalian remains found at Kyson in Suffolk, mentioned in the preceding notice. Ann. Nat. Hist. 4, 191-4. 1846. A History of British fossil mammals and birds. London. pilgrim, g. e. 1910. Notices of new mammalian genera and species from the tertiaries of India. Rec. Geol. Surv. India 40, 63-71. 1914. Description of teeth referable to the Lower Siwalik Creodont genus Dissopsalis. Ibid. 44, 265-79. 1932. The fossil Carnivora of India. Mem. Geol. Surv. India (n.s.) 18, 1-232. piveteau, j. 1935. Etudes sur quelques creodontes des Phosphorites du Quercy. Ann. Paleont. 24, 75-95. prestwich, j. 1850. On the structure of the strata between the London Clay and the Chalk in the London and Hampshire Tertiary systems. Quart. Jour. Geol. Soc. London 6, 252-8 1 . robinson, p. 1960. Sinopa from the Cuchara formation of Colorado. Postilla 44, 1-4. rutimeyer, l. 1 862. Eocaene Saugethiere aus dem Gebiet des schweizerischen Jura. Allg. Schweize- rische Gesell. (n.d.) 19, 1-98. 1891. Die eocane Saugethier-Welt von Egerkingen. Abhandl. Schweizerische Palaont. Gesell. 18, 1-151. russell, d. e., and mckenna, m. c. 1962. Etude de Paroxyclaenus, mammifere des phosphorites du Quercy. Bull. Soc. Geol. France (7) 3, 274-82. savage, R. J. G. (In press.) Tertiary Carnivora of East Africa. Fossil Mammals Africa. schlosser, m. 1887. Die Affen, Lemuren, Chiropteren, Insectivoren, Marsupialier, Creodonten und Carnivoren des europaischen Tertiars. Teil 1. Beitr. Palaont. Geol. Osterreich-Ungarns 6, 1-224. 1911. Beitrage zur Kenntnis der oligozanen Landsaugetiere aus dem Fayum: Agypten. Beitr. Palaont. Geol. Osterreich-Ungarns Orients 24, 59-167. L. VAN VALEN: SOME EUROPEAN PROVIVERRINI 665 scott, w. b. 1892. A revision of the North American Creodonta with notes on some genera which have been referred to that group. Proc. Acad. Nat. Sci. Philadelphia 44, 291-323. and jepsen, g. l. 1936. The mammalian fauna of the White River Oligocene. Parti. Insectivora and Carnivora. Trans. Amer. Phil. Soc. 28, 1-153. smith, G. E. 1902. On the morphology of the brain in Mammalia, with special reference to that of the lemurs, recent and extinct. Trans. Linnaean Soc. London ( Zool .) (2) 8, 319-432. stehlin, h. g. 1903. Die Saugetiere des schweizerischen Eocaens. Critischer Catalog der Materialen. Erster Teil. Abhandl. Schweizerischen Paldont. Gesell. 30, 1-153. 1905. Die Saugetiere des schweizerischen Eocaens. Critischer Catalog der Materialen. Dritter Teil. Ibid. 32, 447-595. 1940. Uber die Saugetierfauna der Teredinasande von Epernay und Umgebung. Eclogae Geol. Helvetiae 33, 292-8. teilhard de chardin, p. 1915. Les Carnassiers des phosphorites du Quercy. Ann.Paleont. 9,103-92. 1921. Les mammiferes de l’Eocene inferieur frangais et leurs gisements. Ibid. 10, 171-6; 11, 1-108. thorpe, M. R. 1923. Notes on the Bridger (Eocene) Carnivora. Amer. Jour. Sci. (5) 5, 23-39. van valen, l. (In press, a.) Deltatheridia, a new order of mammals. Bull. Amer. Mus. Nat. Hist. (In press, b.) Paroxyclaenidae, an extinct family of Eurasian mammals. Jour. Mammal. Whitaker, w., dalton, w. h., and bennett, f. j. 1885. The geology of the country around Ipswich, Hadleigh, and Felixstowe. Mem. Geol. Surv. England and Wales (explanation of sheets 48 NW. and NE.), 1-156. white, E. i. 1931. The vertebrate faunas of the English Eocene. Vol. 1. British Museum (Natural History), London. wortman, j. l. 1902. Studies of Eocene Mammalia in the Marsh Collection, Peabody Museum. Part I. Carnivora. Amer. Jour. Sci. (4) 13, 39-46, 115-28, 197-206, 433-48; 14, 17-23. L. VAN VALEN, American Museum of Natural History, New York Manuscript received 26 October 1964 APICAL DEVELOPMENT IN TURRITELLID CLASSIFICATION WITH A DESCRIPTION OF CRISTISPIRA PUGETENSIS GEN. ET SP. NOV. by RICHARD C. ALLISON Abstract. The nature of primary spiral ribs on apical whorls of Turritella ( s.l .) is utilized for subdivision of the genus. However, variation causes some doubt as to the actual value of this criterion. Recognized variation is of two kinds: (1) ‘variation’ within a stock or group arising from improper grouping of species due to inadequate descriptive notation, and (2) actual variation displayed by various specimens of single species. Inadequate application of the terms ‘unicarinate’, ‘mesocostate’, ‘unicostate’, ‘bicarinate’, ‘bicostate’, &c. has confused efforts to arrange six Gulf Coast Eocene taxa in homogeneous groups. Notational systems proposed by Marwick (1957a, b) and Kotaka (1959) permit uniform description of apical developmental characters, and provide a firm base for taxonomic differentiation. The six taxa discussed belong to three separate apical types. Some species are constant in apical development, while others show considerable variation. A given spiral rib may appear at various distances from the apex, but always seems to appear in a constant order relative to other spiral ribs. As long as this sequential order of appearance remains constant, Kotaka’s notation allows continued use of apical ontogenetic features for taxonomy. This constancy of sequential appearance of primary spirals is not yet demonstrated for all apically variable turritellas; further investigation is needed to confirm apical developmental characters as reliable tools for all turritellids. A new turritellid from the Eocene of Washington, Cristispira pugetensis gen. et sp. nov., illustrating orderly apical variation, is described and represents a new stock seemingly unrelated to presently known species from the north Pacific. Recently, a new turritellid was collected from the Raging River Formation (Eocene) of Washington by James D. Vine of the U.S. Geological Survey. The specimens were submitted to me by F. Stearns MacNeil for study and description. The material consists of approximately eighty external moulds which have been studied by means of latex casts. The fossils are from a locality in the centre of the NE. \ of the SW. \ of section 15, Township 23 North, Range 7 East, Willamette Survey, in the Tiger Mountain-Taylor Mountain upland area, King County, Washington, U.S. Geological Survey locality M-648 (see USGS 7^-minute Hobart Quadrangle, 1953). The fossil locality is about 1,600 feet below the top of the Raging River Formation, and currently stands as the stratigraphically lowest recorded fossil occurrence in that formation. The subjacent sedimentary rocks of the Raging River Formation are approximately 1,000 feet thick but no fossil material is known from them. Vine (1962a, pp. 7-11) named the Raging River Formation and discussed its stratigraphic relationships and age. Lists of marine fossils identified by F. Stearns MacNeil and Welden W. Rau were included. Turritella n. sp. aff. T. yabei Kotaka of MacNeil’s checklist is the Cristispira pugetensis gen. et sp. nov. of this paper. MacNeil concluded that the Raging River Formation is of middle Eocene to early late Eocene (late Ulatisian to early Narizian) age. In view of the present trend towards subdivision of the genus Turritella Lamarck, it seems advisable to present at least a limited discussion of turritellid classification in connection with the description of the new genus. It is hoped that this discussion will emphasize the value of taxonomic subdivision of turritellids to palaeontologists. [Palaeontology, Vol. 8, Part 4, 1965, pp. 666-80, pi. 92.] R. C. ALLISON: APICAL DEVELOPMENT IN TURRITELLID CLASSIFICATION 667 The types and figured specimens of Cristispira are deposited in the U.S. National Museum. The hypotypes of Turritella temblorensis Wiedey and Turritella arenicola (Conrad) are in the Museum of Paleontology, University of California, Berkeley. TURRITELLID CLASSIFICATION Familial subdivision. The generic assignment of turritellas has been, and remains, diffi- cult. Systematists have attempted to subdivide the inclusive genus Turritella ( sensu lato ) with varying degrees of success, and several investigators have closely examined the morphologic criteria used to subdivide the family. Notable have been the works of Palmer (1937), Bowles (1939), Merriam (1941), Palmer in Harris and Palmer (1947), Marwick (1957a, b), and Kotaka (1959). Each of these papers contains a history of previous efforts, so it is not repeated here. Merriam (1941, p. 35) has clearly stated the need for subdividing Turritella Lamarck. He states (loc. cit.) : The major desideratum with regard to the establishment of a universal classification of Turritellidae is information concerning the apical development of those subdivisions, mostly European and austral, to which taxonomic designation has already been given. Figures, descriptions, and available material are not satisfactory for this purpose. It appears evident that if there is justification for subdivision of Lamarck’s genus — and there un- doubtedly is — a large number of units of at least subgeneric standing must ultimately be erected and adequately defined. In this manner alone can the classification be placed on a basis of actual genetic relationship. There is still much to be done before the existing taxonomic units are thoroughly understood and described, but studies such as those of Marwick (1957a) have laid a firm foundation for further elaboration of turritellid classification. Collectively these previously mentioned workers have given increasingly careful attention to the shape of the growth-line on the whorl sides and base, to the nature of the ontogeny of the primary spirals, to the type of protoconch, and to the details of the aperture. Such studies have shown an amazing diversity within the Turritellidae. One of the most significant advances has been the recognition of various groups or stocks of turritellids which have evolved along quite different lines. These groups are restricted in their zoogeographic distribution. Seemingly, no world-wide classification based on simple growth-line similarities (e.g. Guillaume’s Groupes) can adequately express phylogenetic relationships. Convergences in adult appearance are rather common even though apical developmental characters may be quite divergent. The morphologic criteria previously enumerated form the best basis for classifica- tion yet proposed, but some difficulties are still apparent, and the absolute significance of each of these characters in classification of all turritellids has not yet been demon- strated. Perhaps the most useful character has been the apical ontogeny. Many recog- nized groups, such as Merriam’s stocks, have been found to be amazingly consistent in their ontogenetic development. This constancy seems to indicate that the youthful development is a valuable criterion which may be applied to turritellid classification. Notation of apical whorls. The terminology applied to the apical whorls of gastropods is far from standardized. Cox (1955, pp. 195-8) and Cox in Moore et al. (1960, pp. I 1 1 1- 14) has reviewed these terms and pointed out some of the limitations in their usage. The early whorls which immediately follow the protoconch in turritellids have been 668 PALAEONTOLOGY, VOLUME 8 variously referred to as the ‘nuclear whorls’, the ‘nepionic whorls’, and the ‘neanic whorls’. Table 1 summarizes several authors’ usages of the terminology applied to the protoconch and early portion of the teleoconch as used herein. For several reasons none of the terms enumerated above is satisfactory. ‘Nuclear whorls’ neither agrees with Dali’s (1890-8) usage of ‘nucleus’ nor with Cox’s (1955 and 1960) use of the same term. In addition, as Cox (1955, pp. 196-7) has pointed out, ‘nucleus’ is not an appropriate term for a series of whorls. The terms ‘nepionic’ and ‘neanic’ are equally objectionable as applications of life-cycle terminology to the hard parts of an organism, the precise table 1. Comparison of terminology applied to apical whorls of gastropods Dali (mO- 1898) Palmer (1937, 1947) Merriam (1941) Cox (1955, 1960) Marwick (1957a, b) Kotaka (1959) This Paper Proto- conch = earliest caplike shell Nucleus Nucleus Proto- conch Nuclear whorls Proto- conch (nucleus == point of origin) Nucleus = semi- globular initial part Protoconch Proto- conch (nucleus = point of origin) Proto- conch (nucleus point of origin) Apical whorls Nepionic whorls (first 3 or 4) or post- nuclear whorls Early part of teleo- conch Neanic whorls Neanic whorls Early part of teleo- conch life history of which is unknown. Usage of ‘nepionic’ depends on correlation of the protoconch with the embryonic stage, an assumption that is clearly not warranted. If modern usage of ‘embryo ’ is to be followed, the embryonic stage must be that part of the ontogeny before the animal becomes self-supporting. While some turritellids (e.g. Gazameda gunnii (Reeve) and others) have been shown to be ovoviviparous, developing the protoconch and the early whorls of the teleoconch within the em- bryo, other species (e.g. T. communis Risso) seem to be oviparous, though remaining only a short time in the plankton (Merriam 1941, pp. 12-13). The usage of ‘neanic whorls’ for the early part of the teleoconch implies that the protoconch represents the ‘nepionic’ stage. This implication would clearly not be true for ovoviviparous turri- tellids. Cox (1955, p. 195) has also pointed out that ornamentation may appear before the termination of the planktonic larval stage in many gastropods. For this reason, the im- plication that the protoconch and the ‘nepionic’ stage are correlative may also be incorrect for oviparous species ; this latter usage would be correct only when the proto- conch is terminated at the end of the planktonic larval stage. The usage of ‘neanic whorls’ also contradicts the usual application of this term in cephalopod literature where most authors have followed Hyatt. It seems clear that unless the life cycle of the gastropod in question is understood, there is no firm basis for application of these life- R. C. ALLISON: APICAL DEVELOPMENT IN TURRITELLID CLASSIFICATION 669 j- protoconch (incomplete) cycle terms to the hard parts of gastropods. It is best to use the definition of ‘proto- conch’ given by Cox (1955, p. 197) and to apply the term ‘teleoconch’ to the remainder of the shell, wholly avoiding the life-cycle terms. The protoconch and the early whorls in which the principal developmental stages are observed may be referred to collectively as the ‘apical whorls’, though it must be remembered that spiral rib development often continues throughout life in the Turritellidae. Palmer (1937), Merriam (1941), Bowles (1939), and Palmer in Harris and Palmer (1947) have grouped species into unicarinate, mesocostate, unicostate, bicarinate, bi- costate, cingulate, tricarinate, tricostate, and multicostate groups or stocks. Palmer in Harris and Palmer (1947, p. 280) has pointed out that her terms ‘bicarinate’, &c. have priority over Bowles’s and Merriam’s ‘bicostate’, &c., and that the former terms are more appropriate (see p. 670). Finlay (1930), Ida (1952), Marwick (1957a, b), and Kotaka (1959) have described these same relationships using several systems of nota- tion with capital or capital and lower-case letters and numerals (see Marwick 1957a, p. 148; 19576, pp. 12-14; and Kotaka 1959, p. 59). These latter systems describe the actual ontogeny more precisely, thus provi- ding a more accurate basis for comparison. The notation of Kotaka is here used to designate the major elements of the spiral sculpture (see text-fig. 1). B is the medial primary, D the peribasal one involved with the anterior suture ; A is the first to appear adapical to B, and C is the first to appear abapical to B (between B and D). The con- ventions used by Kotaka (loc. cit.) and Marwick (loc. cit.) for secondary and tertiary spirals are not used in the present text-fig. 1. Diagrammatic representation of study because of their limited significance Cristispira pugetensis gen. et sp. nov. showing for generic differentiation, although they notational usage. The formula is a3B! C2d. The may be used advantageously in specific antispiral and spiral sinuses of the growth-line diagnoses. When the primaries are weakly immediately abapical to d. developed, they are noted by a lower-case letter (e.g. b) instead of the capital. Numerical subscripts indicate order of appearance (e.g. Bx indicates primary B appears first). D has been included in the formula when present, but no numerical subscript is given due to the difficulty in determining its point of origination. It should be capitalized when it forms a prominent element of the ornamentation visible on the whorl side, and designated by the lower-case d when it simply forms the angulation between the whorl base and side, remaining level or nearly level with the suture. In the present study ‘unicarinate’, ‘mesocostate’, ‘bicarinate’, ‘unicostate’, ‘bicostate’, &c. are used only to delimit the condition of the earliest suture antispiral sinus spiral sinus growth line angle 670 PALAEONTOLOGY, VOLUME 8 post-protoconch whorls. Weak primaries are not neglected (e.g. a species with ax Bx cx is considered tricostate, not unicostate or mesocostate). These terms should not be applied to whorls other than those showing the first ornamental stage. Notation and progressive ontogenetic change. Palmer (1937, pp. 188-9) and Palmer in Harris and Palmer (1947, pp. 280-1) has classified her Gulf Coast Eocene species accord- ing to the sculpture of (1) the ‘nepionic’ whorls (first 3 or 4) and (2) the first ‘post- nepionic’ whorls, with each of these stages furnishing a basis for subdivision into unicarinate, bicarinate, and tricarinate species groups. As a consequence, the same species may be considered unicarinate on the basis of the ‘nepionic’ whorls, or bicari- nate on the basis of the first ‘postnepionic’ whorls. This classication was an early attempt to provide a framework which describes the normal ontogenetic change in turritellids, and yet one which could accommodate certain unusual ontogenetic rever- sals (see discussion of T. arenicola danvillensis). This progressive change during ontogeny has been emphasized by Palmer (1947, p. 279): Fragments taken from different parts in the length of the shell, unless fitted into the total pattern [of ontogeny], may seem to belong to different species. The change of ornamentation in the ontogeny of gastropods is universal but such a process is particularly accelerated and profusedly developed in the Turritellidae. She also states (op. cit., p. 280) The factor of change enumerated in the first paragraph of this discussion [see above] is not accounted for in the Bowles-Merriam grouping. A species may be unicarinate on the nepionic or postnepionic whorls, becoming bicarinate and later multicarinate on the mature whorls. . . . There is no rigidity in their carination over the entire length of their shell. Therefore a qualifying statement must accom- pany any outline as to what part of the shell the ‘type carination’ may occur. From this data, one may easily visualize the results of tachygenesis and bradygenesis. A unicostate form might be ‘accelerated’ in ontogenetic development to the point of being apically tricostate, or a tricostate form might be ‘retarded’ to a unicostate condition. This progressive ontogenetic change may take three forms: (1) simple addition of spiral ribs, (2) changes in relative strength of ribs, either weakening or strengthening, and (3) loss of ribs present in earlier stages. All three features may occur independently of gerontism. Inconsistent notation and variation. While the recognition of the basic type carination has proved useful in arranging many species, difficulties have been encountered with others. This difficulty stems from two major causes : (1) inadequate or inconsistent appli- cation of the terms ‘unicarinate’, ‘mesocostate’, ‘unicostate’, ‘bicarinate’, ‘bicostate’, &c., and (2) actual apical variation. For example, Palmer has not considered weak primaries in her assignment of the terms ‘unicarinate’, &c. (i.e. ‘ unicarinate ’ may signify only one primary present, or one primary which is stronger than its adjacent cohorts). Bowles and Merriam generally have followed the practice of recognizing in their termi- nology all primaries present. This difference between Palmer’s scheme and that of Merriam and of Bowles results from a difference in emphasis. Palmer has emphasized the enlarged ribs and applied the term ‘carinate’, while Merriam and Bowles have emphasized all the ribs and have applied the term ‘costate’ (‘carinae’ referring to strong keeled ribs and ‘costae’ referring to any rib). Therefore a form with ax Bx cx might be considered uni- R. C. ALLISON: APICAL DEVELOPMENT IN TURRITELLID CLASSIFICATION 671 carinate in Palmer’s scheme and tricostate in Merriam’s and Bowles’s schemes (Palmer, personal communication, 6 April 1964). Where these terms are used, care should be taken to apply the proper termination, though usage of Kotaka’s notation is less apt to result in confusion. As a result of the differing emphasis of these schemes, and of the inconsistent application of these terms by various investigators, a variety of apical developmental types may bear a single descriptive term. In addition, different primary ribs may appear first in separate unicarinate or unicostate genera [e.g. Acutospira with Cx (unicostate) and Cristispira with Bx (mesocostate)]. Note that mesocostate implies the Bx condition; unicostate means a single primary (A, B, or C) appears first; unicarinate means a single primary rib which is keeled (A, B, or C) appears first, but may or may not be the only rib present. Merriam and Bowles have partially accounted for this problem by recognizing a ‘ cingulate ’ division for those species which have Ax Cl5 although the Bx Cx and Ax Bx conditions would both receive a ‘bicostate’ designation. This lack of precision in definition and usage of the terms ‘unicarinate’, ‘mesocostate’, ‘ unico state ’, ‘bicarinate’, ‘bicostate’, &c. and the resultant improper grouping of species creates the impression that there is apical variation in otherwise consistent stocks. While the costate or carinate terms are useful in classification, they do not provide a sufficiently refined basis for comparison of species. Because of its greater precision and simplicity, the notational system of Kotaka should be more generally adopted as a tool in turritellid classification. An example of the utility of this notation is provided by consideration of six Gulf Coast taxa ; all have been wholly or partially referred to as ‘ unicarinate ’ or ‘ unicostate ’ at some stage during their life cycle and therefore warrant comparison to Cristispira pugetensis. Three separate apical developmental types may be recognized among these species ; one, T. arenicola (Conrad), illustrates actual variation in the apical characters of different individuals. It is hoped that this view will explain some anomalies in classi- fication based on apical development. The features of these six taxa are summarized in the following discussion (note the inadequacy of simple ‘carinate’ and ‘costate’ classification). Turritella alveata Conrad 1855 Development. Data are from Palmer in Harris and Palmer (1947, pp. 288-90, pi. 36, figs. 7-12). The earliest post-protoconch whorls show ax Bx cx with ax very fine and much weaker than Cx at first, shortly progressing to ax Bx Cx, and finally to Ax Bx Cx d. cx is stronger than ax initially. The protoconch consists of one to one and a half whorls. Remarks. Palmer has considered this species apically unicarinate (op. cit., p. 281): ‘Obscure unicarinate stage, followed by a short bicarinate stage, succeeded by a tri- carinate stage.’ The weak primaries, a and c, are neglected in the ‘nepionic’ whorl designation. Bowles’s (1939, p. 271) tricostate diagnosis of the form is in agreement with the classification adopted here. The apical development apparently remains constant. T. creola and T. alveata appear to be closely allied species. Turritella creola Palmer in Harris and Palmer 1947 Development. Data are from Palmer in Harris and Palmer (1947, p. 286, pi. 35, figs. 2, 4, 5, 11). The development is ax Bx cx, i.e. the species is truly tricostate initially, soon 672 PALAEONTOLOGY, VOLUME 8 showing a stronger C, or ax Bx Cx d. cx is much stronger than the very weak ax initially. The protoconch consists of one whorl. Remarks. Palmer (loc. cit.) has classified this species as a unicarinate-bicarinate form, these statements applying to the ax Bx cx stage and the ax Bx Cx stage respectively. The species should be considered tricostate. Palmer (ibid.) states : This species [T. creola ] may be differentiated from T. arenicola in that the bicarinate feature has been formed as a definite character by the stage of the fifth whorl and the unicarinate phase is shortened and limited to the first four postnuclear whorls. In this respect T. creola seems to be more of a derivative of T. dutexata and allies of the lower Claiborne, than directly related to T. arenicola. She states again (op. cit., p. 287): T. creola is apparently the Jackson representative of the T. dutexata stock of the lower Claiborne. While tachygenic development of T. dutexata (a2 Bx Cx d) or T. arenicola (a3 Bx c2 d) could develop a species such as T. creola, the latter is better classified along with those species included in Bowles’s ‘tricostate group’ (Bowles 1939, pp. 270-1). Application of the terms ‘carinate’ and ‘costate’ therefore seems to have obscured the relationships of this species. T. creola has not been cited for apical variability. Turritella apita de Gregorio 1890 Development. Data taken from Palmer (1937, pi. 24, figs. 1, 3, 7, 10). The development is a2 Bx Cx d, soon becoming a2 Bx cx d. In other words, the species is actually bicostate apically ; a2 is always weak, and Cx (never as strong as Bx) becomes reduced in strength relative to Bx. The latter spiral remains strong and gives the whorls an angulate appear- ance. The protoconch consists of about two whorls. Remarks. Bowles (1939, p. 275) has classed T. apita as a unicostate species, although his illustration (pi. 31, fig. 4) does not show the earliest whorls. The text implies an a3 Bx c2 development. Palmer in Harris and Palmer (1947, p. 285, no fig.) classes T. apita as a bi-unicarinate species (referring to the ‘nepionic’ and first ‘postnepionic’ whorls respectively). She states (loc. cit.): It and T. arenicola danvillensis are the only two Turritellas [sic] of the Claiborne-Jackson Eocene which belong in such a group [bi-unicarinate]. T. apita passes into the unicarinate stage and retains it throughout life, while in T. arenicola, the unicarinate stage is a transitional phase, developing five and more spiral ribs, depending on the locality of its occurrence. The discrepancies in assignment seem to be the result of inadequate material and in- sufficiently precise terminology. The treatment of Palmer in Harris and Palmer ignores the adult a2 cx and utilizes only the young Bx Cx and the adult Bx characters. Turritella arenicola danvillensis Stenzel and Turner 1940 Development. Data taken from Stenzel and Turner (1942, card 58), and K. V. W. Palmer (personal communication). Bx Cx appear on the third whorl. By the seventh or eighth whorl Cx weakens (but does not disappear) and the species is ‘ unicarinate ’ though still ‘ biscostate ’ at this stage ; Cx strengthens on the ninth or tenth whorl ; a2 then appears along with several other secondaries on the posterior slope, d angulates the whorl at the R. C. ALLISON: APICAL DEVELOPMENT IN TURRITELLID CLASSIFICATION 673 suture. The protoconch consists of about two whorls (see Palmer in Harris and Palmer 1947, pi. 34, fig. 4). Remarks. Palmer in Harris and Palmer (1947, pp. 284-5, pi. 34, figs. 1, 4, 5; pi. 35, fig. 1) reports examination of over one thousand fragments with abundant apical material from the type locality at Danville Landing, Louisiana. She mentions no disagreement with Stenzel and Turner’s description of the apical ontogeny. Dr. Palmer has informed me (personal communication, 6 April 1964) that the double carination of the third whorl is sharp, after which the lower rib weakens until the whorls are ‘unicarinate’, and that the lower rib then increases in size until the whorls again become ‘bicarinate’. However, Cx does not actually disappear. The species should be considered apically bicarinate, but such a simple statement seems to be insufficient for evaluation of its relationships. T. apita and T. arenicola danvillensis are both classified as bicarinate-unicarinate by Palmer; on T. arenicola danvillensis Q is weakened temporarily on the early whorls, while it is weakened throughout adult whorls in T. apita. This seems to be evidence of close relation- ship, differing in degree rather than in kind. T. arenicola danvillensis could be sufficiently bradygenic to show an ancestral bicarinate condition, followed by a descendent unicari- nate stage with a ‘normal’ a3 Bx C2 d development. It is interesting to note that the apical development of T. arenicola danvillensis is different from that of T. arenicola. T. arenicola is apically unicostate, but some specimens (PI. 92, fig. 10) very closely approach the bicostate condition; such convergence may very well attest to the close relationship of the two taxa. However, it seems likely that T. arenicola danvillensis should be given independent specific and generic rank. Even if the two species are closely related, attainment of the unicostate condition in T. arenicola could form a convenient arbitrary boundary for discrimination of separate generic entities. Careful study of the apical stages from various demes of T. arenicola and T. arenicola danvillensis, coupled with a similar analysis of their stratigraphic relationships, should aid in the interpretation of the phyletic relationships and the dependent taxonomy. T. arenicola danvillensis apparently is consistent in its early development, for Palmer records no individual variation in her study of the abundant specimens from Danville Landing. Turritella arenicola (Conrad) 1865 Development. Data taken from Bowles (1939, pp. 275-6, pi. 31, figs. 5-7). The develop- ment is a3 Bj c2 d plus several more secondaries posteriorly. One or both secondaries posterior to a3 seem to appear concurrently with it. This is a truly unicarinate and meso- costate species. B is slightly anterior of the whorl midline apically, later migrating to a more central position on the adult whorls. The protoconch consists of about two whorls (figured for comparison on PI. 92, fig. 10). Remarks. Palmer (1937, pp. 197-8), Bowles (loc. cit.) and Palmer in Harris and Palmer (1947, pp. 281-3) have concurred in classifying T. arenicola as a unicarinate or unicostate species. However, Bowles has remarked on the individual variation seen on the apical whorls. He states (op. cit., p. 276): There is a certain amount of individual variation in the persistence of the unicarinate sculpture on the apical whorls. On some specimens the second prominent revolving rib appears as early as the third 674 PALAEONTOLOGY, VOLUME 8 whorl, while in others it is still absent on the eighth or ninth whorls. It invariably does appear, however, and it always appears anterior to the original carination. c2 appears as a fine thread almost immediately after Bx on the specimen figured for comparison. The point of initiation of a given spiral element (number of whorls from the protoconch) is quite variable in other species as well (e.g. Cristispira pugetensis), even in specimens from a single locality. Bowles does not clearly specify whether variants are common among specimens from a single locality or if the variation is more com- monly between individuals from separate localities. However, the order of rib appear- ance, and therefore the notation used here, remains constant in T. arenicola and in all other variable species with which the writer is acquainted. Bowles (1939, p. 276) notes the close resemblance of T. arenicola (Conrad) to T. dutexata Harris, which seems to be a bicostate (a2Bx Cxd) form. Palmer (1937, p. 199) classifies T. dutexata as tricarinate-bicarinate. Her illustration of an incomplete specimen (pi. 26, fig. 1) shows a tricarinate (tricostate) individual, which remains so, but the speci- men of fig. 4 on the same plate is clearly bicarinate (bicostate). Harris originally described the species as bicarinate to the very apex, mentioning no tricarinate stage. If T. dutexata or its relatives are ancestral to the T. arenicola group, we might think of the a3 and c2 of T. arenicola as having been retarded in their point of insertion (bradygenesis). Turritella arenicola branneri Harris 1894 Development. Data from Palmer in Harris and Palmer (1947, pi. 34, figs. 2, 3, 6, 7). The primary spirals develop in the order a3 Bi c2 d, plus several other secondaries posteriorly. This is a truly unicarinate and mesocostate subspecies. The protoconch consists of about two whorls. Remarks. Palmer (1937, p. 197, pi. 23, figs. 1, 2) and Palmer in Harris and Palmer (loc. cit.) has considered T. arenicola branneri unicarinate. Bowles (1939, p. 275) considers it a junior synonym of T. arenicola (Conrad), and Palmer (1947, p. 282) con- siders that it has been synonymized with good reason. Nevertheless, she lists it sepa- rately ‘ to tabulate the characters of extreme forms so that such may be used as criteria forjudging other variations and the value of named units.’ T. arenicola branneri seems properly considered as a junior synonym. T. arenicola branneri has not been specifically cited for apical variation. Summary Careful study of these species suggests that the apical variation previously implied is in part real and in part the result of inadequate terminology. T. alveata Conrad and T. creola Palmer in Harris and Palmer are better considered as members of Bowles’s ‘tricostate group’. T. apita de Gregorio and T. arenicola danvillensis Stenzel and Turner seem to be related and should either be placed in a group by themselves (bicarinate- unicarinate group of Palmer) or in Bowles’s ‘ bicostate group ’. T. arenicola (Conrad) and T. arenicola branneri Harris belong to Bowles’s ‘unicostate group’. Thus constituted, three separate groups are recognized, each group apparently warranting taxonomic recognition. Formal designation of these taxa should await review and inclusion of other Gulf Coast species. R. C. ALLISON: APICAL DEVELOPMENT IN TURRITELLID CLASSIFICATION 675 While these six taxa have not been classified as members of a single group, all have had the term ‘ unicarinate ’ or ‘unicostate’ applied to some part of their ontogenetic cycle. Classification of these six species in one group would necessitate inconsistent application of the criteria used to designate the class; i.e. the variation in such a group would be unnatural and in part the result of insufficiently refined descriptive notation for the comparison of apical stages. This ‘variation’ actually constitutes valid differences which we may recognize in taxonomy. Among these six taxa, only T. arenicola (Conrad) has been specifically cited for its individual apical variation. As I have previously emphasized, the various spirals appear at different distances from the protoconch in different individuals of the species, but the order of appearance remains constant. Cristispira pugetensis also is markedly variable in this same fashion, but again the sequential order of spiral rib appearance remains constant. Instances in which the sequential order of insertion of the primary spirals is reversed in individual variants are unknown to the writer ; such cases would create con- siderable difficulty for the uniform application of the sequential order of apical spirals to taxonomy, but in their absence such criteria seem to be of considerable value. Variable species may not be well adapted with respect to their apical shell characters, hence showing a wider phenotypic expression in the absence of direct selection pressure. Constant species may be more thoroughly adapted with a lesser range of phenotypic expression. Conversely, environmental influences may be more important than genetic factors in producing apical variation. Studies on living turritellids should be under- taken in order to evaluate these factors. Variation between separate demes (stratigraphic and geographic separation) of the same species should also be investigated. If compari- sons show the position of primary rib appearance to differ consistently between separate demes, we may find tachygenesis and bradygenesis to be significant mechanisms of evolution in the Turritellidae. Interpretation of such processes may provide a usable tool in reconstructing the broad outlines of phyletic history. The diversity of the many Gulf Coast turritellid species does not militate against use of apical ontogeny as a taxonomic key, but presents various examples of rather rapid diversification (radiation). In all likelihood, taxonomy based on these apical characters as well as growth-line characteristics, &c., would closely approach a true phyletic classification. Instances will be found in which arbitrary ‘cut-off's ’ will be necessary, but this is a natural result of well-documented phyletic relationships being expressed in taxonomy. It is impossible in this brief study to give a thorough review of all cases which may have bearing on the validity of using apical ontogeny in classification. The six taxa discussed have been chosen for examination because of their bearing on the new material from the Washington Eocene and because they illustrate the applicability of more refined notation in comparing apical stages of turritellids. SYSTEMATIC DESCRIPTION Phylum mollusca Class GASTROPODA Subclass PROSOBRANCHIA Order caenogastropoda 676 PALAEONTOLOGY, VOLUME 8 Family turritellidae Woodward 1851 Subfamily turritellinae Woodward 1851 Genus cristispira gen. nov. Type species. Here designated Cristispira pugetensis sp. nov. Diagnosis. Shell turritelliform ; protoconch of three or four (?) smooth whorls; primary spirals appearing in order a3 Bx C2 d; adult growth-line with double sinus, deepest part of antispiral sinus usually above whorl midline and substantially embayed; spiral sinus shallower and broader with deepest part of embayment on d or slightly above; growth- line angle variable, ranging from about 15° to 25°; base of whorl with prominent revolv- ing spirals ; aperture subovate with heavy parietal wash. Name. Latin crista, f. = crest ; spira, f. = coil, twist. Discussion. The very slightly effuse basal and columellar lips and the subovate aperture of Cristispira are vaguely reminiscent of Mesalia and other genera assigned to the Pareorinae Finlay and Marwick 1937. However, the concave basal growth-line and lack of a ridge on the adapical columellar lip show that it is a member of the Turritellinae (see Marwick 1957a, p. 164, under ‘ Zaria ’). Cristispira pugetensis sp. nov. Plate 92, figs. 1-9, 12-14 1962a Turritella n. sp. aff. T. yabei Kotaka, MacNeil in Vine, p. 9. Description. Shell of medium size; maximum observed length 56 mm.; maximum ob- served diameter 16 mm. Pleural angle averages about 19°, ranges from about 15° to 30°; apical angle usually about same as pleural angle, but on some specimens much wider than pleural angle. Spire profile normally conical to concave conical, but speci- mens with wider apical angle obconical adapically. Primary spirals develop in order a3Bx C2d; d forms angulation between side and base exactly at suture nearly throughout ontogeny, only rarely becoming raised on whorl side adjacent to anterior suture. Proto- conch of about three or four (?) sharply convex smooth whorls (multispiral); about EXPLANATION OF PLATE 92 Figs. 1-9, 12-14. Cristispira pugetensis gen. et sp. nov. USGS locality M-648, Raging River Formation, Washington. Rubber casts. USNM Cat. No. 132. Fig. 3, holotype; remainder, para- types. 1, specimen with Bx d alone (x 1-7), 648626. 2, specimen with Bx d alone (x 1-6), 648627. 3, specimen with a3BxC2d (xl-7), 648628, holotype. 4, specimen with Bx C2 d (xT7), 648629. 5, specimen with a3 Bx C2 d ( x 1 -5), 648630. 6, specimen with Bx C2 d showing slightly effuse columellar lip and parietal wash ( x 1-6), same as tig. 14, 648631. 7, specimen with Bx d alone showing proto- conch partially decorticated (x2-6), 648632. 8, specimen with Bx C2d (x 1-7), 648633. 9, specimen with Bx C2d (x 1-7), 648636. 12, specimen with Bx C2d (x 1-7), 648634. 13, gerontic whorl showing heavy growth-lines ( x 2), 648635. 14, same as fig. 6, showing parietal wash ( x 1-3), 648631. Figs. 10, 11. Turritella arenicola (Conrad). UCMP locality A-1043, Jackson Group, Louisiana. 10, specimen with a3Bx c2d plus posterior secondary; accelerated specimen with c2 as a fine thread almost immediately after Bx (x9-3), UCMP 36491. 11, specimen showing growth-line (xl-9), UCMP 36492. Fig. 15. Turritella temblorensis Wiedey. UCMP locality B-7853, Topanga Formation, California. Specimen showing double sinused growth-line with ? a3Bx C2d (x0-95), UCMP 36493. Palaeontology, Vol. 8 PLATE 92 ALLISON, Turritellids R. C. ALLISON: APICAL DEVELOPMENT IN TURRITELLID CLASSIFICATION 677 fifth whorl a strong medial angulation (BJ appears, increasing in strength and angulating whorls throughout ontogeny until last one or two, when it may decrease in strength slightly with onset of gerontism; normally with five or six unicarinate apical whorls before appearance of C2 on lower third of whorl, though point of insertion of C2 is quite variable, C2 failing to appear at all on some smaller specimens; C2 variable in strength from coarse thread to slightly subordinate to B! ; a3 even more variable, making its appearance in only about one-third of specimens, apparently never earlier than in young adult whorls ; a3 variable from coarse thread, when discernible, to weak primary, but always subordinate to Bx and C2. Whorl profile variable with diverse development of primaries, from strongly angulate medially on specimens with no a3 or C2 to ‘pagoda- form’ on specimens with strong, subequal Bx and C2 with no a3, subrounded with slight medial angulation on specimens with a^Qd all developed. Well-preserved specimens with many fine spiral threads. Body whorls of occasional large specimens with gerontic features such as slight obsolescence of primaries, heavy strengthening of growth-lines, and a slight tendency toward looser clasping of preceding whorl causing d to be weakly exposed and suture more deeply impressed. Suture normally moderately impressed and clearly discernible. Base flattened and ornamented with four or five coarse secondary spirals. Aperture subovate with heavy parietal wash; basal and columellar lips very slightly effuse, continuing to parietal wash; peristome incomplete. Growth-line moder- ately variable ; antispiral sinus moderately deep with deepest part of embayment usually between a3 and B^ growth-line usually spirally convex just below posterior suture; spiral sinus shallower and broader than antispiral sinus with maximum at d or slightly above; growth-line with broad shallow antispiral concavity on base; growth-line angle moderate and variable, ranging between 15° and 25°. Name. The species name refers to the Puget Sound region and Puget Lowlands of western Washington. Holotype. USNM 648628. Paratypes, USNM 648626-7, 648629-35. All material is from U.S. Geo- logical Survey locality M-648. Discussion. Only three primary spiral developmental combinations occur : Bx d, Bx C2 d, or a3 Bj C2 d (i.e. a3 Bx d or a3 C2 d are never found in this species) in spite of the variation in point of insertion of a3 and C2. Specimens lacking C2 invariably lack a3, although C2 may be present without a3. a3 is therefore always retarded first, and C2 next ; the primary spirals are both inserted and retarded in an orderly manner. This sequential order of rib appearance has also been described in T. arenicola (Conrad) (Bowles 1939, p. 276, also quoted on page 673 of this paper). A fundamental order of this nature seems to be usual in turritellids ; the apical ontogeny may be variously retarded or accelerated, but the sequence of spiral appearance in different groups remains constant and is therefore a useful criterion in classification. MacNeil in Vine (1962a) has classified C. pugetensis as ‘ Turritella n. sp. aff. T. yabei Kotaka’ and Kotaka has included T. yabei under Acutospira. In 1959 Kotaka (pp. 101-2) proposed Acutospira as a new subgenus of Colpospira Donald 1900, and referred three Japanese Tertiary species to it. The type species, A. okadai Nagao 1928, develops the primaries in the order A2 b3 Q d, with Cx appearing just slightly before Aa. Therefore the apical ontogeny differs markedly from that of Cristispira. The type species of Colpo- spira Donald is multicostate apically with a deep antispiral sinus and a very shallow or negative growth-line angle (see Marwick 1957a, pp. 151-3). The growth-line of vy B 6612 678 PALAEONTOLOGY, VOLUME 8 Acutospira Kotaka is similar, but in view of the differences in apical development, the subgeneric relationship of Acutospira to Colpospira seems questionable. A. tashiroi Kotaka has a growth-line with a strongly negative angle; apparently it is tricostate, but the apical development is not known. Adult whorls have a very heavily developed A. A. yabei Kotaka seems to be bicostate, developing B and C at about the same point apically with A appearing later (d also present). This differs considerably from the development of typical Acutospira. The growth-line is sharply and deeply embayed and the deepest part lies on A; the growth-line angle is very small or negative. In general appearance ‘A.’ yabei resembles C. pugetensis, but the latter has a prolonged early development of B. In Cristispira, the maximum depth of the growth-line is usually between a3 and Bx and the growth-line angle is wider, never becoming negative. MacNeil (1964, pp. B-2, 3, pi. 1, figs. 5-8, 12-18) has described a bicostate turritellid, Turritella kotakai, with a strong B on the adult whorls, from the Miyara Formation (middle or late Eocene) of Ishigaki-shima, Ryukyu Islands. This new species seems to be related to ‘A.’’ yabei Kotaka. Its apical development is Bx Cx d with no A appearing. While some of the early Tertiary Japanese turritellids are similar in gross aspect to C. pugetensis, the refined generic concepts used here preclude congeneric assignment. C. pugetensis seems to have no definite relatives among described species from the Pacific Coast Tertiary. Various subspecies assigned to the T. variata and T. diversi- lineata branches of the T. uvasana stock of Merriam (1941) are similar in general appearance and have the coarsely ornamented whorl base, but they are bicostate on the apical whorls. Among other Pacific Coast species, those referable to the T. broderi- piana stock of Merriam (1941) are unicostate, but these have a broad, shallow growth- line with a single sinus. This latter group seems to be referable to Archimediella (Torculoidella) Sacco 1895 (see Marwick 1957a, pp. 159-60). One species, T. temblor ensis Wiedey, from the Californian Miocene, warrants compari- son. It is apparently a unicostate form which may develop the primary spirals in the order a3 Bx C2 d as in Cristispira, but the very earliest apical whorls are not known. Merriam has considered the growth-line to be single sinused (1941, p. 116), but better material from the Topanga Formation shows it to be double sinused and concave on the base (see PI. 92, fig. 15). Loel and Corey (1932, p. 265) have called attention to the simi- larity in appearance of T. temblor ensis and Zaria duplicata (Linnaeus), but the latter bears the characteristic convex basal growth-line of the Pareorinae, while the former belongs to the Turriteliinae. As Merriam has suggested (op. cit., p. 117), the similarity is one of homeomorphy. In general appearance T. temblorensis is very similar to Cristispira. The whorl base bears the coarse revolving ribs, the growth-line angle is about the same, and a small parietal wash seems to be present. However, the growth-line has a much broader and shallower antispiral sinus with its deepest part falling near B. The segment of the growth-line above the maximum flexure is spirally concave rather than convex as in Cristispira. C apparently develops very soon after B and a follows after several whorls, along with a secondary just below the posterior suture (concurrently?). A unicostate stage does not seem to be prolonged as in C. pugetensis. In spite of the several similari- ties between these two species, I am inclined to judge the difference in growth-line as of supraspecific importance. Discovery of more complete apical material of T. temblorensis and stratigraphically and geographically intermediate forms may further elucidate their relationship. R. C. ALLISON: APICAL DEVELOPMENT IN TURR1TELLID CLASSIFICATION 679 The developmental patterns of several Gulf Coast Eocene species have been reviewed in the preceding pages. T. apita de Gregorio differs from Cristispira in its apical develop- ment (a2 Bx Cx becoming a2 Bx cx d) although the adult shell is reminiscent of some variants of the Washington species. T. arenicola danvillensis Stenzel and Turner also differs markedly in the early development (Bx Q with Q weakening for a few whorls, then strengthening and followed by a2) and has more evenly convex and regularly orna- mented adult whorls. T. creola Palmer and T. alveata Conrad are apically tricostate (a! Bx c1 d) and therefore not closely related. T. arenicola (Conrad) and T. arenicola branneri Harris are the only Gulf Coast taxa which warrant careful comparison with C. pugetensis. Apically their development is agBiCad; a3 and a secondary just posterior to it seem to appear concurrently. The proto- conch consists of about two whorls and a parietal wash is present. The growth-lines have their maximum flexure near the midline of the whorl above B, and below a... The antispiral sinus is deeply embayed, though not so sharply as in Cristispira. The growth-line, as in Cristispira , has a very slight spiral convexity just below the suture, and the growth-line angle measures about 10°. The base of the whorl bears revolving ribs of moderate strength (Palmer 1937, pi. 23, fig. 2) but the complete apertural details are not known. Conrad originally assigned T. arenicola to Mesalia ?, but this assignment seems to have been based on the rather shallow whorl height and broad pleural angle rather than apertural characteristics. T. arenicola branneri Harris differs from T. arenicola only in the number of secondaries on the adult whorls and in the pleural angle. Bowles (1939, p. 275) has properly considered it a junior synonym of T. arenicola (Conrad). While T. arenicola (Conrad) and T. arenicola branneri Harris are similar to Cristispira in many aspects, I am somewhat hesitant to assign them to the latter genus. The Gulf Coast taxa have a much rounder whorl profile and more numerous revolving spirals, with one of the posterior secondaries appearing concurrently with a3; the basal spirals are not so well developed as in Cristispira. Perhaps the most significant difference is the more broadly rounded antispiral sinus with the narrower growth-line angle. As Marwick (1957a, pp. 156, 158) has emphasized in his discussions of Kurosioia Ida, Maoricolpus Finlay, and Stiracolpus Finlay, geographic and stratigraphic factors should not be ig- nored for purposes of classification. Convergences are quite common in the Turritel- lidae. Marwick (1957a, p. 158) states: ‘Close agreements in outer lip characters and in primary spiral ontogeny are essential for generic grouping, and even then the possi- bility of convergences of distant stocks with simple characters must be considered.’ Unfortunately, the phyletic relationships of Cristispira remain obscure. No certain ancestor or congener is yet known from the Tertiary deposits of the Americas and the western Pacific. While T. temblorensis Wiedey and T. arenicola (Conrad) are similar mor- phologically, both differ in the details of the growth-line. Wide geographic or strati- graphic separation, or both, in the absence of intermediate forms, contributes additionally to the speculative nature of congeneric assignments. In all probability relatives of Cristispira will be found in as yet unknown north Pacific fossil faunas. Acknowledgements. The writer is indebted to F. Stearns MacNeil, James D. Vine, J. Wyatt Durham, Stanley R. Primmer, Joseph T. Gregory, Katherine V. W. Palmer, Charles W. Merriam, and K. Sakomoto for their help in the preparation of this paper. 680 PALAEONTOLOGY, VOLUME 8 REFERENCES allison, r. c. 1965. Value of neanic ontogenetic development in taxonomy of the Turritellidae (Gastropoda) with reference to a new stock from the Eocene of Washington. [Abstract] Abstracts for 1964. Spec. Pap. Geol. Soc. Amer. 82, 235-6. bowles, e. 1939. Eocene and Paleocene Turritellidae of the Atlantic and Gulf Coastal Plain of North America. J. Paleont. 13, 267-336, pi. 31-34. cox, l. r. 1955. Observations on gastropod descriptive terminology. Proc. malac. Soc. Lond. 31 (5, 6), 190-202. 1960. In R. c. moore (ed.), Treatise on Invertebrate Paleontology. Part I, Mollusca 1, I 84-169. Lawrence, Kansas. dall, w. H. 1890-8. Contributions to the Tertiary fauna of Florida. Trans. Wagner free Inst. Sci. Philad. 3 (1-4), 1-947, pi. 1-35. finlay, h. j. 1930. Additions to the Recent molluscan fauna of New Zealand. Trans. Proc. N.Z. Inst. 61 (2), 222-47, pi. 1-4. ida, K. 1952. A study of fossil Turritella in Japan. Rep. geol. Surv. Japan , no. 150, 1-64, pi. 1-7. kotaka, t. 1959. The Cenozoic Turritellidae of Japan. Sci. Rep. Tohoku Univ., 2nd ser. (Geol.), 31, 1-135, pi. 1-15. loel, w. and corey, w. h. 1932. The Vaqueros formation, lower Miocene of California. 1, Paleonto- logy. Univ. Calif. Pubis Bull. Dep. Geol. Sci. 22, 31-410, pi. 4-65. macneil, f. Stearns. 1964. Eocene megafossils from Ishigaki-shima, Ryukyu-retto. Prof. Pap. U.S. geol. Surv. 399-B, Bl-14, pi. 1-3. marwick, j. 1957a. Generic revision of the Turritellidae. Proc. malac. Soc. Lond. 32 (4), 144-66. 19576. New Zealand genera of Turritellidae, and the species of Stiracolpus. Paleont. Bull. Wel- lington, 27, 1-55, pi. 1-5. merriam, c. w. 1941. Fossil turritellas from the Pacific Coast region of North America. Univ. Calif. Pubis Bull. Dep. Geol. Sci. 26, 1-214, pi. 1-41. palmer, k. van winkle. 1937. The Claibomian Scaphopoda, Gastropoda and dibranchiate Cephalo- poda of the southern United States. Bull. Am. Paleont. 7, 1-730, pi. 1-90. ■ in Harris, g. d. and palmer, k. van winkle. 1947. The Mollusca of the Jackson Eocene of the Mississippi Embayment (Sabine Riber to the Alabama River), 2nd section, including part II, uni- valves and index. Bull. Am. Paleont. 30, 209-466, pi. 26-65. stenzel, h. B. and turner, f. e. 1942. Type invertebrate fossils of North America, cards 1-148. Bur. of Econ. Geol., Austin, Texas. vine, j. d. 1962a. Stratigraphy of Eocene rocks in a part of King County, Washington. Rep. Invest. Div. Geol. Wash. St. Dep. Conserv., no 21, 1-20. 1962 b. Preliminary geologic map of the Hobart and Maple Valley quadrangles, King County, Washington. Div. Geol. Wash. St. Dep. Conserv., Geol. Map GM-1. RICHARD C. ALLISON Museum of Paleontology, University of California, Berkeley 4, California Manuscript received 28 October 1964 A NEW CALAMITALEAN CONE FROM THE MIDDLE PENNSYLVANIAN OF SOUTHERN ILLINOIS by f. A. hibbert and D. a. eggert Abstract. A new species of the calamitalean cone genus Paracalamostachys Weiss, P. cartervillei, is described. The holotype consists of a group of cones found in an ‘ironstone’ nodule of Middle Pennsylvanian age. The cones are small, ranging from 1 1 to 17 mm. in length, and 4-0 to 4-5 mm. in diameter. Several cones are present at a node on the cone-bearing stem, they lack peduncles, and consist of whorls of bracts and sporangiophores arranged alternately, the fertile whorls being superposed. Approximately six sporangiophores are present in the fertile whorl and twelve bracts in the sterile whorl. Each sporangiophore bears four horizontally elongate sporangia arranged in the form of a Maltese cross around a central stalk. The sporangiophores are enclosed by the upturned ends of the subtending bracts. Spores ranging in size from 40 to 100 /x were isolated from sporangia and are referable to the form genus Calamospora Schopf, Wilson, and Bentall, being comparable to the dispersed spore species, C. breviradiata Kosanke. The cone is apparently homosporous and is compared with previously described species of the genus Paracalamostachys Weiss and related genera. The holotype of a new species of the genus Paracalamostachys Weiss (Weiss 1884) was recently discovered at an abandoned mine site about half a mile West of Carterville, Illinois. It consists of a compressed axis bearing cones at three nodes preserved in an ‘ironstone’ nodule. At this Middle Pennsylvanian plant locality in Williamson County (Sec. 4, T 9 S, R 1 E), Herrin Quadrangle, nodules containing plant material are found in place in shales above the Herrin (Illinois No. 6) coal, and in weathered shales forming the upper part of the overburden material removed from above the coal during mining. The horizon from which the holotype is derived is in the Carbondale Formation of the Kewanee Group (Kosanke et al. 1960). The original appearance of the specimen is shown in Plate 93, figs. 1 and 2. The fossil was more fully exposedby etching with a 10 per cent, solution of phosphoric acid which removed the nodular matrix. This etching action of the acid was allowed to proceed for varying lengths of time, while being observed under a binocular dissecting microscope. Periodic flushings with distilled water were applied to halt the action of the acid so that critical examination of the progress of the etching could be determined. The above technique revealed that the cones were preserved in some depth in the matrix and that they were not so compressed as the original appearance of the specimen suggested. This method of study has yielded a considerable amount of information concerning the struc- ture of the cones and may prove to be particularly useful with reproductive structures found in comparable nodules. In some cases a fine white material, apparently consisting of phosphates produced in the treatment, may form on the nodular surface. Application of a dilute solution of ammonium hydroxide followed by thorough washing with distilled water removes most of this unwanted by-product of the acid treatment. Spore samples were taken from sporangia in various positions along the axis of the cone. These were picked off, as spore masses, as the phosphoric acid etching proceeded. Extremely short oxidation (2 to 3 minutes) with concentrated HN03 followed by a washing with 10 per cent, potassium hydroxide solution released the individual spores [Palaeontology, Vol. 8, Part 4, 1965, pp. 681-686, pi. 93.] 682 PALAEONTOLOGY, VOLUME 8 from the spore mass and made them more readily examined, without affecting their structure. Genus paracalamostachys Weiss Paracalamostachys cartervillei sp. nov. Plate 93, figs. 1-5; text-fig. 1 Diagnosis. Small cone composed of alternating whorls of bracts and sporangiophores with sporangiophores of successive whorls superposed. Approximately six sporangio- phores and twelve bracts in each of the two types of whorls. Each sporangiophore having a central stalk bearing four horizontally elongate sporangia. Cones homosporous, having spores of the Calamospora breviradiata Kosanke type. Position of attachment of sporangiophore stalk to cone axis not known. Stratigraphic occurrence. Shales above the Herrin (Illinois No. 6) coal, Carbondale Formation, Kewanee Group, Middle Pennsylvanian. Holotype. Specimen No. 563 and slide preparations in the Paleobotanical Collections, Botany Depart- ment, Southern Illinois University. Description of cone. The overall length of the entire specimen (PI. 93, figs. 1, 2) is 3-8 cm., the diameter of the axis bearing the cones ranging from 2-5 mm. just above a node, to about 3-0 mm. immediately below a node. The length of the cones, which are present at three nodes along the axis, varies from 11 to 17 mm. while their maximum diameters range from 4-0 to 4-5 mm. Three cones are present at each of the two upper nodes and two cones are present at the lowest node. The type of foliage borne by the fertile stem cannot be determined ; however, a few elongate structures suggestive of Aster ophyllites type foliage are present in the nodule (PI. 93, fig. 1, b) and have an orientation con- sistent with their having been leaves of the stem. Prior to treatment with phosphoric acid, each cone showed several horizontal rows of brown, somewhat spherical structures which subsequent study showed to be sporangia and spore masses. Numerous bracts, present in whorls, alternate with the whorls of fertile parts and have broad basal portions and acute tips. The exact number of bracts in each whorl is estimated at being twelve on the proportion of the diameter of the cone which one bract occupies. Except where the bracts are obviously displaced due to crushing, they turn abruptly upward immediately to the outside of the sporangia and cover the fertile parts, giving the cone a compact appearance. Selective etching of the matrix has made it possible to determine some features of the sporangiophores. The part and counterpart of the cone indicated by the letters A in Plate 93, figs. 1 and 2 respectively were chosen for this purpose since they appeared to be the most complete. Fig. 3 shows this cone (PI. 93, fig. 1, at a) at the same magnification after EXPLANATION OF PLATE 93 Figs. 1-5. Paracalamostachys cartervillei sp. nov. 1, 2, Part and counterpart of holotype showing appearance before treatment to expose material. Cone which was more fully exposed by acid treat- ment indicated by a; possible foliar structures indicated by b. x 3. 3, Cone indicated by A in fig. 1 after acid treatment, x 3. 4, Enlarged view of exposed cone showing details of sporangiophores. Indications explained in text, x 9. 5, Isolated spore of Calamospora breviradiata type from holotype specimen, x 500. Palaeontology, Vol. 8 PLATE 93 HIBBERT and EGGERT, Pennsylvanian Calamitalean cone HIBBERT AND EGGERT: A NEW CALAMITALEAN CONE 683 removal of the matrix. Portions of six whorls of sporangiophores have been uncovered. The upper and lower whorls are very incomplete, but the four central whorls are exposed in a fair degree of completeness (PI. 93, fig. 4). Each sporangiophore bears four sporan- gia which are circular in outline, except for flattened areas where they are in mutual contact, when viewed in the tangential aspect (PI. 93, fig. 4, at c). The horizontally elongate form of the sporangia can be seen in views that are more radial (PI. 93, fig. 4, at d). The sporangiophore stalks are generally poorly preserved and their positions are marked by hole-like depressions in most cases. No distal expansion or head was found on the stalks of the sporangiophores, but the preservation does not preclude their having been present originally. In the centrally located sporangiophore of the third whorl from the top of the cone (PL 93, fig. 4, at c) the tissue of the sporangiophore stalk is preserved and consists of a slightly raised area having a circular outline. The sporangia of each sporangiophore occur in two superposed pairs. Superposition is also evident between the sporangiophores of the second and third whorls from the top of the cone, while those of the fourth whorl are less clearly superposed with the members of the higher whorls (PI. 93, fig. 4). This is due to vertical separation of the sporangia in the centre of the whorl, which occurred during the acid treatment. Although it is difficult to delimit some of the members of the whorls of sporangiophores, it is most likely that they were super- posed throughout the cone. In cases where the sporangiophore stalk (or the depression marking its position) is viewed in radial aspect this structure has an essentially hori- zontal orientation (PI. 93, fig. 4, at d). The form and orientation of the parts of the sporangiophores suggest that the sporangiophore stalks were attached to the cone axis approximately midway between bract whorls. Unfortunately, no clear instance of the attachment of a sporangiophore stalk to the cone axis was found. Parts of the cone which were uncovered most completely suggest that the number of sporangiophores was probably six per whorl. These sporangiophores are closely en- veloped by the bracts which appear as fine lines of dark material in the nodular matrix in side view (PI. 93, fig. 4, above the arrow at d). Poor preservation of cuticular remains present in the specimen does not allow a deter- mination of the structure of the epidermis. Description of spores. Spores isolated from the cones are radial, trilete, and were probably spherical prior to compression. The size range is from 40 to 100 p with a peak at 55-60 p. The trilete rays are short and vary in length from 11 to 16 p. The exine is uniformly thin (less than 2 p) and the area contagionis is present in most spores. The spores are referable to the form genus Calamospora S.W. and B. (PI. 93, fig. 5). The trilete mark is distinct, the lips slightly to more prominently elevated and the commissure is thin and attenuate. In most specimens the spore coat shows numerous folds, which may be due to the thinness of the spore wall. The spore coat is laevigate and translucent. Spores within sporangia from all parts of the cones exhibit a similar size range. Of three hundred spores, the size ranged from 40 to 100 p as shown in the histogram (text- fig. 1). The largest spores generally show little or no folding even though their wall thickness is the same as that of the smaller spores. It is likely that P. cartervillei was homo- sporous since the cones exhibit a similar range in spore size throughout their lengths. Kosanke (1950) has previously described spores of Calamospora from Illinois. Two 684 PALAEONTOLOGY, VOLUME 8 species, C. breviradiata Kosanke and C. hartungiana S.W. and B. occur at the horizon from which our cones were taken. In many specimens isolated from the cones the struc- ture of the trilete was similar in every respect to that of C. breviradiata (e.g. the specimen in Kosanke 1950, pi. 9, fig. 4). In other spores isolated from the cones the lips of the trilete were lower, but definite, and extended from 1 to 2 ^ on either side of the suture (PI. 93, fig. 5). In wall structure and mean size the spores are similar to C. breviradiata as described by Kosanke but exhibit a greater size range than that reported by him for this species. The difference in structure of the trilete, present in spores from the same sporangium, may be due to varying degrees of compression, or of spore maturity. For the present we are assigning the spores of P. cartervillei to the dispersed spore species Calamospora breviradiata Kosanke. Discussion. Information obtained from the cones and their included spores clearly indicates a botanical affinity with the order Calamitales. The cones from Carterville are similar to several species of the genera Paracalamostachys and Calamostachys described by Weiss (1876, pi. 16, figs. 1, 2; 1884, pi. 21, fig. 6 and pi. 22, figs. 1, 2, 5, 10-14). The former genus is maintained as distinct from Calamostachys on the grounds that the exact position of sporangiophore attachment is not known. In Calamostachys the sporangio- HIBBERT AND EGGERT: A NEW CALAMITALEAN CONE 685 phore stalk is attached to the cone axis approximately midway between successive bract whorls. As the new material did not allow the position of sporangiophore attachment to be determined, appropriate comparison can best be made with species of Paracalamo- stachys. Among the various species of this genus, the new form is somewhat similar to P. minor Weiss (1884) but differs from that species in that the spores of the new cone are distinguishable from those of P. minor as described by Hartung (1933). P. striata Weiss is distinguished from the Illinois cones in part by being heterosporous (Hartung 1933), while P. po/ystachya (Sternberg) Weiss has elongate peduncles and exposed sporangia. P. rigida Weiss is a smaller cone, but little is known concerning this form and it cannot, therefore, be adequately compared with the cones from Illinois. The species P. hetero- spora recently described by R. and W. Remy (1958) is a very large cone, about 15 cm. long, and is heterosporous. It seems most likely that the new species, P. cartervillei, may have been of the Calamostachys type since compression specimens of Palaeostachya often show some indication of the axillary attachment of the sporangiophores in the form and orientation of the sporangia even when the sporangiophore stalk is not visible (e.g. figures in Weiss 1884, pi. 21, figs. 3, 4). The new cone is also distinct in numerous ways from the recently described genus Calamocarpon (Baxter 1963). The more than twenty-five species of Calamostachys currently recognized encompass a very wide range in form, size, and mode of preservation. The only feature common to all of these forms is the mid-internodal attachment of the sporangiophore stalks to the cone axis between successive bract whorls. The structural details of some species, primarily petrified forms, such as C. binneyana, C. ludwigi, C. casheana, and C. ameri- cana are known. In the case of most of the compression forms much less is known and comparison among the various supposed species is limited to obvious features such as cone length, diameter, and the spacing of fertile and sterile whorls. In a few instances the number of bracts and sporangiophores present in the whorls can be estimated and has been used in distinguishing some of the compression species. Paracalamostachys cartervillei resembles Calamostachys binneyana in the approximate number of bracts and sporangiophores present in the cone whorls, but it is smaller in length and diameter. The spores of C. binneyana persisted in tetrads which often con- tained aborted spores and were enclosed in a perisporial membrane. There is no evidence that these features were present in P. cartervillei. On the basis of measurements presented by Hartung (1933) for the spores of C. binneyana, it appears that P. cartervillei produced spores of greater diameter and with a greater size range than did C. binneyana. Of the remaining forms of Calamostachys which may be compared with the Carterville cones, C. ludwigi should be noted for its similarity. C. ludwigi differs from the Carterville cones in the same features which set them apart from C. binneyana. All the remaining species of Calamostachys are clearly distinct fromP. cartervillei on the basis of numerous features. Inability to determine the exact position of sporangiophore stalk attachment to the cone axis makes it necessary to refer the Carterville cone for the present to the genus Paracalamostachys in which this structural feature is not known. Acknowledgements. The authors thank Dr. W. S. Lacey for helpful suggestions and a critical reading of this paper prior to publication while serving as Visiting Professor in the Botany Department, Southern Illinois University. This research was supported in part by National Science Foundation Grant GB 1540. 686 PALAEONTOLOGY, VOLUME 8 REFERENCES Baxter, r. w. 1963. Calamocarpoti insignis, a new genus of heterosporous, petrified calamitean cones from the American Carboniferous. Amer. Journ. Bot. 50, 469-76. hartung, w. 1933. Die Sporenverhaltnisse der Calamariaceen. Inst. Paldobot. u. Petrog. d. Brenn- steine Arb. 3, 95-149. kosanke, r. m. 1950. Pennsylvanian spores of Illinois and their use in correlation. Illinois State Geol. Surv. Bull. 74, 128 pp. simon, j. a., wanless, h. r., and willman, h. b. 1960. Classification of the Pennsylvanian strata of Illinois. Illinois State Geol. Surv. Rept. of Invest. 214, 84 pp. remy, r. and remy, w. 1958. Beitrage zur Kenntnis d. Rothliegendflora Thiiringens. Teil III. Sitzber. Dtsch. Akad. fViss. 3, 1-16. weiss, ch. E. 1876. Steinkohlen-Calamarien. Abhand. Geol. Specialkarte. v.PreuJSen u.d. Thuringischen Staaten 2, Heft 1, Text 149 pp.; Atlas 19 pis. 1884. Steinkohlen-Calamarien II. Ibid. 5, Heft 2, Text 204 pp.; Atlas 28 pis. F. A. HIBBERT Botany School, Downing Street, Cambridge D. A. EGGERT Dept, of Botany, University of Iowa, Manuscript received 16 October 1964 Iowa City LOWER AND MIDDLE DEVONIAN SPORES OF NORTH AND CENTRAL VESTSPITSBERGEN by K. C. ALLEN Abstract. From the Devonian succession in North and Central Vestspitsbergen, seven hundred rock-specimens collected by the author and over three hundred collected by previous Expeditions have been examined macro- scopically; two hundred and fifty of these were prepared for microscopic investigation. These included speci- mens from every Spitsbergen formation, and consisted typically of medium to fine-grained clastic sediments. Eighty-two dispersed spore species are recorded, of which forty-eight new species are systematically described. Five new genera ( Bullatisporites , Craspedispora, Cymbosporites, Aulicosporites, and Chelinospora) are proposed, and another genus ( Archaeozonotriletes ) is emended. Twenty-five morphologically complex species were sec- tioned, in order to elucidate and interpret their structure. Consideration is given to corrosion and preservation of the spore exine. This paper describes only the preparation and systematics of the dispersed spores from the Devonian succession of North and Central Vestspitsbergen. In a later paper the microfloral assemblages and their stratigraphical applications will be discussed. The study of the dispersed spores from the Devonian succession of Spitsbergen was undertaken at the joint suggestion of Mr. N. F. Hughes and Mr. W. B. Harland. An initial study by Dr. P. F. Friend and Mrs. M. Mortimer had shown the presence of spores from horizons in the Mimer Valley Series. All the samples studied were collected by members of various Spitsbergen Expedi- tions organized from the Sedgwick Museum, although prior to 1959 few samples were collected for this purpose ; samples for palynological preparation and examination were collected by Dr. P. F. Friend and Dr. D. J. Gobbett in 1959, and by the author in 1961. The spore genera and species are based solely on morphological criteria, their botanical affinities are for the most part unknown. The Devonian succession in Spitsbergen is confined to the island of Vestspitsbergen, where the Devonian outcrops over a large area (Friend 1961, p. 79, fig. 1). It consists of about six thousand metres of predominantly Old Red Sandstone facies. Numerous sections were collected for palynological study throughout the area by the author, in the hope of producing a stratigraphical succession based on dispersed spores. Well- localized samples collected by previous Cambridge Spitsbergen Expeditions were also used. For general stratigraphy and stratigraphical nomenclature, I have followed Friend (1961) throughout. Acknowledgements. The writer expresses his sincere gratitude to Mr. N. F. Hughes for constant advice and encouragement throughout the course of this study. The investigation was made possible through the fieldwork of members of various Cambridge Spitsbergen Expeditions. The writer is parti- cularly indebted to Mr. W. B. Harland who organized most of these Expeditions, and who jointly with Mr. Hughes suggested the research topic; to Dr. P. F. Friend who led the 1961 Cambridge Spitsbergen Expedition and who in 1959 with Dr. D. J. Gobbett, collected many of the samples investigated. Grateful acknowledgement is extended to Professor O. M. B. Bulman, F.R.S., for the use of research facilities of the Sedgwick Museum, Cambridge, where the study was carried out. The writer also [Palaeontology, Vol. 8, Part 4, 1965, pp. 687-748, pis. 94-108.] 688 PALAEONTOLOGY, VOLUME 8 thanks Dr. M. E. Dettmann for helpful discussions, and Mr. A. Barlow for kindly assisting with much of the photography. Through the courtesy of Professor O. A. Hoeg, of the University of Oslo, the writer was able to examine collections of Devonian plants from Spitsbergen housed in the Palaeontological Museum, Oslo. The writer is indebted to the Shell International Petroleum Company Limited for financial assistance during the tenure of one of their Postgraduate Studentships. PREPARATION AND EXAMINATION OF SAMPLES Specimens were examined macroscopically, and information on colour, grain-size, calcareous and micaceous content, plant fragments, and any weathering was recorded. A representative selection of two hundred and fifty specimens, comprising a variety of lithological types ranging from coals to coarse-grained sandstones, and all of probable continental origin, were prepared for microscopic examination. Red oxidized shales and sandstones failed to produce any spores, but all other lithological types, though not all other samples, produced spores ; the grey and green siltstones yielded the best preserved and most diverse assemblages. Although the same basic technique was used throughout, with such a variety of litho- logies every sample required individual attention and timing at each stage of its prepara- tion. The macroscopic record of each sample prepared was studied, so that any constant reaction might perhaps be related to some visible feature, thus helping to improve the selection of suitable samples. Each specimen for microscopic investigation was scrubbed in distilled water to remove any Recent contamination. About 3 gm. of sediment were broken up with a hammer into approximately 3-5 mm. fragments. To avoid any con- tamination, the specimen was crushed within sheets of newspaper on an anvil, both the hammer and anvil surfaces being washed after each crushing. If the specimen was super- ficially weathered, care was taken to use only the central part. Calcareous samples were treated with 20-30 per cent, hydrochloric acid, and left for at least 2 hours (and frequently overnight). All clastic samples were then treated with cold 50-60 per cent, hydrofluoric acid for 2-14 days; the remaining sediment was then transferred to a nickel crucible to which fresh hydrofluoric acid was added and boiled for 30-40 minutes. Insoluble fluorides resulting from the previous treatment were removed in warm 10-30 per cent, hydrochloric acid in a water-bath. Frequently as many as ten of these 5-minute treatments were needed to remove all the fluorides. To the washed residue in 10 cc. of distilled water were added 2-3 drops of 50 per cent. Stergene (or other non-ionic detergent). This was then subjected to a 5-20 second treat- ment with an ultrasonic disintegrator (1:1 end ratio steel probe vibrating at 20 kilo- cycles per second). This disaggregated clumps of organic and mineral matter. The residue was then washed in distilled water until the top fluid remained clear. Oxidation of the humic material was then carried out in Schulze solution. Maceration time varied from 10 minutes to 5 hours, with different samples. Frequently, further clearing and concentration with an alkali was necessary, and a few drops of 5 per cent, potassium hydroxide, sodium hydroxide, or ammonium hydroxide were used, the latter seemingly gave the best results. The residue was left in alkali for about 15 seconds. K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 689 Often minerals which survived the HF treatments were in such concentration that they required removal by a heavy liquid separation. Acidified zinc bromide (S.G. 2-2) was used, the preparation being centrifuged for 20 minutes at 1,800 revolutions per minute. Only two coals are present in the Spitsbergen Devonian succession ; these were crushed and then macerated in Schulze solution for 24-48 hours, thereafter following the preparation technique for clastic sediments. From these residues at least two strew-slides were prepared for each productive sample, the residue being mounted in unstained glycerine jelly on standard glass slides, and covered with No. 0 coverslips. The slide was tapped lightly, to encourage the residue to settle in one plane, and after allowing the glycerine jelly at least three days to set, the coverslips were sealed with gold size. Single megaspore and microspore mounts were made of several species. Single spores were also separated from residues for serial section- ing, following the embedding and sectioning techniques of Wigglesworth (1959) and Hughes, Dettmann, and Playford (1962). From one to four specimens of twenty-three species were sectioned at 2-3 /x intervals, perpendicular to the equatorial plane, mounted in glycerine jelly and sealed as before. From unproductive samples, one slide was usually prepared as a negative record of the sample. Surplus residue was stored in polythene- stoppered glass tubes, in 50 per cent, glycerine containing a few drops of phenol to pre- vent fungal growth. All the strew-slides were first surveyed at x 125 magnification. However, after all pro- ductive samples had been studied, the interpretation of morphological features and specific identification was completed at magnifications of x 500 and x 1250, using as many well-preserved specimens as possible and the sections where available. After the specific descriptions were complete, a quantitative estimation was made of the species present, by counting under high power 200 specimens from each sample yielding a suffi- cient number of well-preserved spores. DISPERSED SPORE SYSTEMATICS Nomenclature and Classification. The nomenclature of the described Spitsbergen Devon- ian spores follows the rules of priority and typification of the International Code of Botanical Nomenclature (1961, Montreal). No botanical affinities are implied with generic designations, which relate to form genera based solely on their morphological features. The nomenclature of suprageneric categories is the artificial classification pro- posed by Potonie and Kremp (1954), later amplified by them (1955, 1956a) and by Potonie (1956, 1958, 1960), and subsequently revised in part by Dettmann (1963). Terminology. In the systematic section, the descriptive terms used are for the most part those which have been well defined, clearly understood, and widely accepted by previous authors. An effort has been made to keep the number of terms to a minimum. The author uses the terms intexine and exoexine (Potonie and Kremp 1955) to denote respectively the inner and outer layers of a two-layered spore wall (exine). Where the exoexine is stratified (occasionally separate, but never truly cavate), for example in Cirratriradites avius sp. nov. (PI. 99, fig. 1 1), the terms outer exoexine and inner exoexine are used. In Spitsbergen Devonian species referred to the genus Perotrilites (Erdtman) 690 PALAEONTOLOGY, VOLUME 8 ex Couper 1953, the exact nature of the outermost layer is unknown; it may be a true perine, or it may be an outer exoexine, and the non-committal term outer sculptine is used. The term cavate (Dettmann 1961) is used for asaccate spores in which the spore wall layers are separated from each other by a cavity, the width of which is at least 20 per cent, of the total spore radius, or if less, then the outer layer is thinner and loosely enveloping, as for example in Perotrilites pannosus sp. nov. (PI. 102, fig. 14). This excludes Densosporites and Anulatisporites which, as demonstrated by Smith (1960, pi. 20, figs. 1-8), occasionally have a small cavity between the intexine and exoexine, and Gemino- spora in which the intexine sometimes separates wholly or partially from the exoexine. Equatorial flange is used for what appears in proximo-distal aspect to be a membranous zona, but proves from sections to be a sharply tapering cingulum. Granulate pseudosculpture is used where corrosion of the infra-granulate structure of a spore wall has resulted in an upstanding ‘ornament’ of granules. A crumina is defined as a separation and extension of the outer part of the exine (or exoexine). This may occur primarily in the distal region, as for example in Hystrico- sporites monosaccus (Archangelskaya) comb. nov. (PI. 96, figs. 1, 2), where it is a separation of the outer exoexine ; or in the equatorial region as in Labiadensites fimbriatus (Waltz) Hacquebard and Barss 1957, see Dettmann and Playford 1963 (pi. 96, figs. 1, 2), where the outer part of the exine is separated and extended. Miospore is used for dispersed spore species, in which the mean diameter is less than 200 i, and megaspores for dispersed spore species in which the mean diameter is more than 200 p. New species have been erected only where fifteen or more adequately preserved speci- mens were available for study. Assignment to previously described species is made only when there is reasonable evidence for identity from published illustrations and descrip- tions. Frequently, however, descriptions are short and illustrations poor, and although the Spitsbergen specimens may appear similar, I cannot be certain that their construction is identical. In most of these cases I have erected new species, recording in my com- parisons where conspecifity may occur; in the event of a re-examination of poorly described species, any conspecifity may be easily located. All type and other figured specimens here described, are housed in the Sedgwick Museum, Cambridge, and referred to by a preparation slide number, followed by the ‘ east-west ’ and ‘ north-south ’ mechanical stage readings, and then by the Sedgwick Museum Spitsbergen Specimen number and a Sedgwick Museum number. In single mount preparations, the slide num- ber is prefixed ‘M’, with sections the slide number is prefixed ‘S’. The stage readings are from the Leitz Dialux microscope No. 3 (serial no. 526724) in the Sedgwick Museum, Cambridge. Anteturma sporites H. Potonie 1893 Turma triletes (Reinsch) Dettmann 1963 Supersubturma acavatitriletes Dettmann 1963 Subturma azonotriletes (Luber) Dettmann 1963 Infraturma laevigati (Bennie and Kidston) H. Potonie 1956 Genus leiotriletes (Naumova) Potonie and Kremp 1954 Type species. Leiotriletes sphaerotriangulus (Loose) Potonie and Kremp 1954. 691 K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES Leiotriletes parvus Naumova 1953 Dimensions. (Twenty specimens) Equatorial diameter 16-24 p (mean 19 p). Occurrence. Fraenkelryggen Division, and Lower Reuterskioldfjellet Sandstone; Gedinnian and Siegenian. Leiotriletes pyramidalis (Luber) comb. nov. 1941 Azonotriletes pyramidalis Luber, in Luber and Waltz, p. 54, pi. 12, fig. 182. 1955 Filicitriletes pyramidalis Luber, p. 60, pi. 3, fig. 70. Dimensions. (Eighteen specimens) Equatorial diameter 54-95 p (mean 70 p). Elevated lips, individually 2-3 /x wide, 4-9 p high. Remarks. Filicitriletes lacks type species designation, and embraces within it spores of several previously well-established genera (Potonie 1958, p. 35). Occurrence. Reuterskioldfjellet Sandstone and Mimer Valley Series; Siegenian to Eifellian. Comparison. Leiotriletes pulvereus Balme and Hassell 1962 (p. 5, pi. 1, fig. 182) has a larger size range, weaker lip development, and is punctate. L. tumidus Butterworth and Williams 1958 (p. 359, pi. 1, figs. 5, 6) has a smaller size range and is tumid. L. ornatus Ishchenko 1956 (p. 22, pi. 2, figs. 18-21) is smaller and has thicker lips. Leiotriletes pagius sp. nov. Plate 94, figs. 1, 2 Diagnosis. Miospores trilete; amb convexly triangular with broadly rounded apices. Laesurae distinct, simple, straight, almost reaching the equatorial margin. Exine 2-5-4 /x thick, homogeneous, laevigate. Dimensions. (Twenty-two specimens) Equatorial diameter 46-63 p (mean 54 /x). Holotype. Preparation KA 251/9, 21-5 94-8, K905, N231. Locus typicus. South Mimerdalen, Central Dicksonland, Spitsbergen; Reuterskioldfjellet Sandstone, Siegenian. Description. Holotype diameter 58 /lx. Exine 4 p thick. Comparison. Leiotriletes dissimilis McGregor 1960 (p. 27, pi. 11, fig. 1) has punctate contact areas, low distinct lips and a thinner exine. Leiotriletes confertus McGregor 1960 (p. 27, pi. 1 1, fig. 2) also has distinct lips and a thinner exine, and the only difference between these two McGregor species is the punctate contact area and open commissure in L. dissimilis, which might be the result of preservation. Leiotriletes devonicus Naumova 1953 (p. 22, pi. 1, fig. 5) and L. trivialis Naumova 1953 (p. 45, pi. 5, fig. 14) are both considerably smaller. Naumova (1953) in describing L. devonicus states ‘exine dense, thick, spore margin thickened’. This thick wall might be misinterpreted as a cingulum, unless seen in broken specimens or in oblique aspect, and many form species included within Stenozonotriletes, may in fact be thick-walled representatives of Leiotriletes or Punctatisporites. If Stenozonotriletes simplex Naumova 1953 (p. 36, pi. 3, fig. 17; p. 69, pi. 10, fig. 3; and p. 130, pi. 19, figs. 16, 17) should prove to have a thick wall and not a cingulum, then Leiotriletes pagius sp. nov. would be synonymous with it. Occurrence. Lower Reuterskioldfjellet Sandstone; Siegenian. 692 PALAEONTOLOGY, VOLUME 8 Genus punctatisporites (Ibrahim) Potonie and Kremp 1954 Type species. Punctatisporites punctatus Ibrahim 1933. Punctatisporites glaber (Naumova) Play ford 1962 Dimensions. (Twenty specimens) Equatorial diameter 29-44 /x (mean 38 p). Occurrence. Throughout the Lower and Middle Devonian succession. Punctatisporites laevigatus (Naumova) comb. nov. 1953 Stenozonotriletes laevigatus Naumova, p. 70, pi. 10, figs. 9, 10. Dimensions. (Twenty specimens) Equatorial diameter 51-71 p (mean 61 p). Exine 3-5-6 p thick. Remarks. Variation in length of laesurae frequently occurs within a single specimen. Naumova (1953, p. 70) in her description, records ‘a thick exine’, and the narrow border to which she refers (which is approximately 4 p thick in her illustration) is probably an optical section through this thick exine, rather than a cingulum as suggested by her generic placing; the species is therefore more appropriately placed within Punctati- sporites. Comparison. Punctatisporites aerarius Butterworth and Williams 1958 (p. 360, pi. 1, figs. 10, 11) is larger and has a thinner exine. Stenozonotriletes facilis Ishchenko 1956 var. facilis (p. 73, pi. 14, figs. 162-4) may be conspecific, but it is not clear from the description whether the otorochka (5 p wide) is a cingulum or the optical section of a thick wall. Occurrence. Fraenkelryggen Division, Dicksonfjorden Sandstone, Reuterskioldfjellet Sandstone, and Mimer Valley Series ; Gedinnian to Givetian. Punctatisporites flavus (Kosanke) Potonie and Kremp 1955 1950 Calamospora flava Kosanke, p. 41, pi. 9, fig. 2. 1955 Punctatisporites flavus (Kosanke) Potonie and Kremp, p. 42. Dimensions. (Twenty-two specimens) Equatorial diameter 95-152 p (mean 120 p). Occurrence. Lower Mimer Valley Series and probable Upper Reuterskioldfjellet Sandstone, Emsian and Eifelian. Genus calamospora Schopf, Wilson, and Bentall 1944 Type species. Calamospora hartungiana Schopf, in Schopf, Wilson and Bentall 1944. Calamospora microrugosa (Ibrahim) Schopf, Wilson, and Bentall 1944 Synonymy. See Playford 1962, p. 579. Dimensions. (Twenty specimens) Equatorial diameter 50-1 10 p (mean 86 p). Exine 2 p or less thick. Occurrence. Kapp Kjeldsen Division, Dicksonfjorden Sandstone, Reuterskioldfjellet Sandstone, and Mimer Valley Series ; Siegenian to Givetian, more common in the Givetian. K. C ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 693 Calamospora nigrata (Naumova) comb. nov. 1953 Leiotriletes nigratus Naumova, p. 23, pi. 1, fig. 9. 1958 Leiotriletes nigratus Ishchenko, p. 35, pi. 1, fig. 5; non Naumova 1953. Dimensions. (Twenty-six specimens) Equatorial diameter 60-80 /x (mean 76 /x). Exine 1 /x or less thick. Remarks. The short laesurae, and thin folded exine suggest correct inclusion within Calamospora. Naumova (1953) quotes dimensions of 60-70 /x, and Ishchenko (1958) 90-1 10 /x. Neither author states the number of spores measured. Naumova’s readings are probably two equatorial axes of the same spore. The Spitsbergen spores are of inter- mediate size. In some specimens there is a tetrad impression of narrow arcuate folds (the curvaturae and curvaturae imperfectae of Potonie and Kremp 1955). Comparison. Leiotriletes atavus Naumova 1953 (p. 23, pi. 1, fig. 8) is smaller and lacks folding. Calamospora saariana Bhardwaj 1957 (p. 81, pi. 22, figs. 13-15) is also smaller and possesses lips. Occurrence. Present in the majority of well-preserved samples ; Gedinnian to Givetian, most common in the Emsian. Calamospora witneyana Chaloner 1963 Dimensions. (Fifteen specimens) Equatorial diameter 1 16-212 /x (mean 151 /x). Occurrence. Reuterskioldfjellet Sandstone and Lower Mimer Valley Series; Siegenian and Emsian. Genus trileites (Erdtman 1945, 1947) ex Potonie 1956 Type species. Trileites (al. Triletes) spurius (Dijkstra) Potonie 1956. Trileites oxfordiensis Chaloner 1963 Dimensions. (Twenty specimens) Equatorial diameter 186-530 /x (mean 290 p). Occurrence. Dicksonfjorden Sandstone, Reuterskioldfjellet Sandstone and Mimer Valley Series; Siegenian to Givetian. Infraturma apiculati (Bennie and Kidston) R. Potonie 1956 Genus granulatisporites (Ibrahim) Potonie and Kremp 1954 Type species. Granulatisporites granulatus Ibrahim 1933. Granulatisporites muninensis sp. nov. Plate 94, figs. 3-5 Diagnosis. Miospores trilete; amb triangular with straight to slightly convex sides and rounded apices. Laesurae straight, length two-thirds to full spore radius, accompanied by smooth, raised lips, individually up to 2 p wide. Exine 1 /x thick, homogeneous to in- fragranulate, contact areas laevigate, proximo-equatorial and distal surfaces sculptured with evenly spaced granules. Dimensions. (Twenty specimens) Equatorial diameter 24-30 /x (mean 27 /x). Holotype. Preparation KA 293/1, 56-7 105-1, K556, N233. z z B 6612 694 PALAEONTOLOGY, VOLUME 8 Locus typicus. East Munindalen, Central Dicksonland, Vestspitsbergen ; Plantekl0fta Conglomerate, probable Givetian. Description. Holotype triangular with straight sides, diameter 25 p. Laesurae four-fifths spore radius, accompanied by lips individually 1 p wide, slightly separating equatorially. Slight curvatural ridges in radial regions. Remarks. The lips are frequently inaperturate, and are sometimes separated equatorially. Slight curvaturae are occasionally present in the radial regions. Comparison. Granulatisporites triconvexus Staplin 1960 (p. 15, pi. 3, figs. 11, 12) is larger, more finely granulose, the lips are less prominent and there is no suggestion of laevigate contact areas. Granulatisporites planiusculus (Luber) Playford 1962 (p. 533, pi. 79, fig. 18) is larger, has a thinner exine and lacks the laevigate contact areas. Archaeozonotriletes parvibasilaris Naumova var. triangulatus Chibrikova 1959 (p. 71, pi. 11, fig. 5) has a thicker exine and sculptured contact areas. Anapiculatisporites devonicus var. azonatus (Chibrikova) Yigran 1964 (p. 13; pi. 1, figs. 21-23) differs only in having a sculpture of coni. Occurrence. Present in most productive samples; Gedinnian to Givetian, most common in the Givetian. Genus cyclogranisporites Potonie and Kremp 1954 Type species. Cyclogranisporites leopoldi (Kremp) Potonie and Kremp 1954. Cyclogranisporites rotundus (Naumova) comb. nov. 1953 Lophotriletes rotundus Naumova, p. 58, pi. 7, fig. 19; p. 108, pi. 16, fig. 34. 1953 Lophotriletes rotundus Naumova var. minor, p. 96, pi. 15, fig. 10; p. 108, pi. 16, figs. 29, 30. No type designated. Dimensions. (Twenty-five specimens) Equatorial diameter 29-45 p (mean 35 p). Remarks. Lophotriletes rotundus Naumova var. minor, lacks both holotypic designation and description. Its separation from L. rotundus Naumova is made on its slightly smaller EXPLANATION OF PLATE 94 All figures x 500 unless otherwise stated; from unretouched negatives. Figs. 1, 2. Leiotriletes pagius sp. nov. 1, Holotype, sectional focus; KA 251/9, 21-5 94-8, K905, N231. 2, Proximal surface; KA 251/7, 40-2 97-5, K905, N232. Figs. 3-5. Granulatisporites muninensis sp. nov. 3, 4, Holotype, proximal and distal surfaces respec- tively; KA 293/1, 56-7 105T, K556, N233. 5, Holotype, X 1000; distal surface, showing granules. Figs. 6-9. Cyclogranisporites plicatus sp. nov. 6, Holotype, KA 258/3, 504 100T, K922, N234. 7, Showing curvaturae; KA 258/2, 494 914, K922, N235. 8-9, Specimens showing variation in size of granules. 8, KA 258/4, 35-1 91-2, K922, N236. 9, KA 258/2, 49-1 91-6, K922, N237. Figs. 10-18. Geminospora spp. 10-11. G. tuberculata (Kedo) comb. nov. 10, Proximal surface; KA 229/3, 22-1 894, K550, N238. 11, Distal surface; KA 255/2, 41-3 87-5, K760, N239. 12-16. G. svalbardiae (Vigran) comb. nov. 12, Distal surface, showing intexine; KA 255/2, 48-2 93-5, K760, N240. 13, 14, Proximal and distal surfaces respectively; KA 243/2, 35-8 91-5, K767, N241. 15, Distal surface; KA 243/2, 47-0 97-2, K767, N242. 16, Section, showing lips and uniformly thick exine; KA 286/S2, 52-5 94-9, K773, N243. 17-18. G. spinosa sp. nov. Holotype, proximal and distal surfaces respectively; KA 243/1, 304 107-3, K767, N244. Figs. 19-20. Acanthotriletes raptus sp. nov. Holotype, proximal and distal surfaces respectively; KA 258/3, 47-0 108-7, K922, N245. Palaeontology, Vol. 8 PLATE 94 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 695 size. However, as only two specimens of L. rotundus var. minor were measured and as the Spitsbergen specimens span the size range of both, the variety is included within the synonymy. Its circular amb and granulose sculpture indicate more appropriate inclusion within Cyclogranisporites. Occurrence. Upper Mimer Valley and Wijde Bay Series; Givetian. Cyclogranisporites plicatus sp. nov. Plate 94, figs. 6-9 Diagnosis. Miospores trilete, originally spherical; amb circular to subcircular. Laesurae distinct to discernible, straight, length spore radius, simple, or accompanied by smooth often sinuous lips, individually 0-5-2 g wide. Exine thin, 1 g or less, homo- geneous, supporting a granulose ornament of variable density. Low, narrow curvaturae or curvaturae imperfectae are present in approximately one-third specimens. Major compressional folds always present. Dimensions. (Forty-five specimens) Equatorial diameter 51-82 g (mean 66 g). Holotype. Preparation KA 258/3, 50-4 100 1, K922, N234. Locus typicus. Mimerdalen, Central Dicksonland, Spitsbergen; ? Reuterskioldfjellet Sandstone, Emsian. Description. Holotype subcircular, diameter 53 g. Laesurae simple, length two-thirds spore radius. Exine less than 1 g thick, densely sculptured. Three compressional folds. Remarks. Variation in size and density of ornament is demonstrated in Plate 94, figs. 8, 9. Curvaturae imperfectae seen in Plate 94, fig. 7. Comparison. Lophotriletes rugosus Naumova 1950 (pi. 2, fig. 6) var. rugosus Naumova 1953 (p. 27, pi. 2, fig. 1 and p. 54, pi. 7, fig. 1), is smaller, and may have a sculpture of cones. Cyclogranisporites lasius (Waltz) Playford 1962 (p. 585, pi. 79, figs. 19, 20) has more distinct laesurae and a thicker exine. Retusotriletes verruculatus Naumova 1953 (p. 29, pi. 2, fig. 10) is smaller, and is roundly triangular. R. punctatus Chibrikova 1959 (p. 52, pi. 5, fig. 7) is also smaller and roundly triangular. Miospores recorded from lEnigmophyton superbum fructification Hoeg 1942 (p. 118, pi. 49, figs. 9-11) appear to be very similar, but are probably laevigate. Occurrence. Present in most well-preserved samples, but extremely common in the Reuterskioldfjellet Sandstone and Lower Mimer Valley Series; Gedinnian to Givetian. Genus geminospora Balme 1962 Type species. Geminospora lemurata Balme 1962. Discussion. Although the type species has a slightly thicker distal surface (a feature com- mon to many Devonian spores), the difference is not appreciable and the genus cannot be regarded as truly patinate. Although the species described below do not show this feature, they are regarded as being of similar construction, and are included within this genus. 696 PALAEONTOLOGY, VOLUME 8 Geminospora tuberculata (Kedo) comb. nov. Plate 94, figs. 10, 1 1 1955 Archaeozonotriletes tuber culatus Kedo, p. 35, pi. 5, figs. 6, 7. Description of specimens. Miospores trilete; amb roundly triangular to subcircular. Laesurae straight, length two-thirds to full spore radius, simple, or rarely accompanied by narrow, sharp lips, individually 0-5-1 ft wide. Exine two-layered; intexine very thin, 1 ft or less, homogeneous, separated at least in part from the exoexine; exoexine 2-3 ft thick, finely infragranulate. Proximal surface laevigate, distal surface sculptured with granules and small cones 1 ft or less high. Major folding common, frequently the intexine is more strongly folded. Dimensions. (Twenty specimens) Equatorial diameter 40-65 ft (mean 56 ft). Remarks. Chibrikova (1959, p. 58), includes Geminospora tuberculata (Kedo) in syno- nymy with Archaeozonotriletes meonacanthus Naumova nom. nud. The latter, however, lacks both holotypic designation and description. Comparison. Archaeozonotriletes plicatus Naumova nom. nud. in Chibrikova 1959 (p. 64, pi. 8, fig. 10) is very similar, and may prove to be conspecific. Occurrence. Upper Mimer Valley Series; Givetian. Geminospora svalbardiae (Vigran) comb. nov. Plate 94, figs. 12-16 1964 Lycospora svalbardiae Vigran, p. 23, pi. 3, figs. 4, 5; pi. 4, figs. 1, 2. Description of specimens. Miospores trilete; amb roundly triangular to subcircular. Laesurae usually straight, occasionally slightly sinuous, length three-quarters to full spore radius, frequently accompanied by low lips, individually 1-3 ft wide. Exine, at least in some specimens visibly two-layered ; intexine thin, less than 1 ft thick, homogeneous, rarely folded, closely appressed to or slightly separate from the exoexine ; exoexine 3-7 ft thick, finely infragranulate. Contact areas laevigate, occupying most of the proximal surface, proximo-equatorial (where the contact areas are short of the equatorial margin) and distal surfaces densely granulate. Exoexine frequently with tangential and acuate folds. Dimensions. (Thirty-four specimens) Equatorial diameter 50-88 ft (mean 72 ft). Remarks. The arcuate folding (PI. 94, fig. 13), often gives the impression of a cingulum under lower power, however, the section (PI. 94, fig. 16) although very compressed, clearly demonstrates the constant exine thickness ; also the exclusively distal ornament, and lips formed by an upturning of the exoexine; no intexine is visible in the sections. This constant exine thickness and exclusively distal sculpture indicate more appropriate inclusion in Geminospora. Comparison. Apiculatisporis sp. Hoffmeister, Staplin, and Malloy 1955a (pi. 1, fig. 13) appears to be very similar. However, it is not possible to see from the photograph whether it is sculptured proximally. Retusotriletes parvimammatus Naumova 1953 var. famenensis Naumova 1953 (p. 124, pi. 18, fig. fig. 22) is clearly similar to some specimens, but there is no mention of a laevigate proximal surface, lips, or the presence of major K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 697 folds. Geminospora lemurata Balme 1962 (p. 5, pi. 1, figs. 5-10) is smaller, lacks lips, and has a slightly thicker distal surface. Archaeozonotriletes atratus Naumova 1953 (p. 99, pi. 15, fig. 25) has much wider and higher lips, and has an ornament of verrucae. Archaeo- zonotriletes notatus Naumova 1953 (p. 84, pi. 13, fig. 12) lacks lips, and there is no men- tion of a laevigate proximal surface. Occurrence. Very common throughout the Upper Mimer Valley Series, but also occurs in the Wijde Bay Series, Lower Mimer Valley Series and Reuterskioldfjellet Sandstone; Emsian to Givetian. Geminospora spinosa sp. nov. Plate 94, figs. 17-18 Diagnosis. Miospores trilete ; amb roundly triangular to subcircular. Laesurae straight, length three-quarters to full spore radius, simple or more usually accompanied by narrow sharp lips, individually 0-5-2 p wide, up to 5 /x high. Exine two-layered; intexine thin, 0-5-2 p wide, homogeneous, usually slightly separate at least in part from the exoexine; exoexine 2-5 p thick, coarsely infragranulate. Contact areas laevigate, fre- quently depressed, occupying from two-thirds to the whole of the proximal surface, proximo-equatorial (where the contact areas are short of the equatorial margin) and distal surfaces sparsely sculptured with spines, 0-5-1 -5 p wide, 3-6 p long. Dimensions. (Fifteen specimens) Equatorial diameter 59-91 p (mean 70 p). Holotype. Preparation KA 243/1, 30-4 107-3, K767, N244. Locus typicus. North ridge of Kinanderfjellet, Central Dicksonland, Spitsbergen; Upper Mimer Valley Series, probable Upper Givetian. Description. Holotype subcircular, diameter 80 p. Laesurae indistinct, masked by ele- vated lips totalling 4 p wide. Intexine very thin, partly separate from exoexine; exoexine 5 p thick, radial diameter of depressed contact area 30 p. Proximo-equatorial and distal spines 1-2 p wide, 3-5 p long. Remarks. The spines are usually appressed against the exoexine. Darkening at the equa- torial end of the laesurae in some specimens is usually due to slight folding, and gives the spores an appearance similar to Pulvinispora Balme and Hassell. However, in this genus the darkened appearance is due to a thickening rather than a folding (Balme and Hassell 1962, p. 10). Comparison. Archaeozonotriletes comans Chibrikova 1959 (p. 70, pi. 19, fig. 3) has shorter simple laesurae, and a denser ornament. Occurrence. Mimer Valley Series; Eifelian and Givetian. Genus acanthotriletes (Naumova) Potonie and Kremp 1954 Type species. Acanthotriletes ciliatus (Knox) Potonie and Kremp 1954. Acanthotriletes raptus sp. nov. Plate 94, figs. 19-20 Diagnosis. Miospores trilete ; amb circular, subcircular to roundly triangular. Laesurae distinct to discernible, simple, straight, length |-| spore radius. Exine 1-2 p thick, homo- 698 PALAEONTOLOGY, VOLUME 8 geneous to finely infragranulate; proximal surface laevigate, distal surface evenly crowded with spines and subordinate cones, 0-5-2-5 p wide, 1-5-4 p high, with polygonal bases, and often only tapering at their apices. Dimensions. (Fifteen specimens) Equatorial diameter 34-42 p (mean 38 /x). Holotype. Preparation KA 258/3, 47-0 108-7, K922, N245. Locus typicus. South Mimerdalen, Central Dicksonland, Vestspitsbergen; Reuterskioldfjellet Sand- stone, Eifelian. Description. Holotype subtriangular with convex sides and broadly rounded apices, diameter 38 p. Laesurae approximately half spore radius. Exine 1 /x thick, spines 0-5- 1-5 /x wide, up to 4 p high. Remarks. This species is included within Acanthotriletes rather than Apiculatisporis on the basis of its dominant, albeit short, spinose ornament. Comparison. All species which may be closely compared with Acanthotriletes raptus sp. nov. have no record of a laevigate proximal surface. This feature alone I regard as suffi- cient for specific separation. However, I am aware that unless studied under oil immersion, it is not always easy to identify this feature in small, proximo-distally compressed speci- mens, and an absence of proximal sculpture may have been overlooked by some authors. I have assumed that this might be the case, and therefore fist below other differing features. A. usitatus Naumova 1953 (p. 24, pi. 1, fig. 15) is triangular and has longer laesurae. Apiculatisporis pineatus Hoffmeister, Staplin, and Malloy 19556 (p. 38, pi. 38, fig. 3) has a larger, more varied ornament. Acanthotriletes parvispinosus Naumova 1953 (p. 24, pi. 1, fig. 16) is smaller and has longer laesurae. A. tenuispinosus Naumova var. tenuispinosus has longer laesurae and narrower spines. Occurrence. Reuterskioldfjellet Sandstone; Eifelian. Genus hystricosporites McGregor 1960 Type species. Hystricosporites delectabilis McGregor 1960. Hystricosporites porrectus (Balme and Hassell) comb. nov. Plate 95, figs. 1-3 1962 Archaeotriletes porrectus Balme and Hassell, p. 10, pi. 5, figs. 1-4. Description of specimens. Miospores trilete; amb circular to subcircular. Laesurae obscured by membranous, often sinuous, elevated lips, 32-60 p high, length equal to full spore radius. Exine three-layered; intexine approximately 1 p thick, homogeneous, closely appressed to the exoexine and often indistinct; exoexine 4-9 /x thick, inner exoexine infragranulate, outer exoexine and sculptural elements homogeneous. Contact areas laevigate, proximo-equatorial and distal surfaces support a dense ornament of grapnel-tipped spines, typically 25-35 p long (range 15-40 p) rarely with a bulbous base, the shaft tapering gradually towards the apex. The majority are preserved in lateral compression. Dimensions. (Twenty specimens) Equatorial diameter 90-1 84 p (mean 118 p), polar diameter (including apical processes) 102-170 p (mean 141 /x). K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 699 Remarks. Since the submission for publication of the paper by Balme and Hassell (1962) there has been further generic subdivision of spores with grapnel-tipped appendages, and the Australian species may now more suitably be included within Hystricosporites McGregor. Sections (PI. 95, figs. 2, 3) demonstrate clearly the homogeneous intexine, the infragranulate inner exoexine and the homogeneous outer exoexine and sculptural elements. The elevated lips are formed only from the outer layer of the exoexine. Comparison. Hystricosporites delectabilis McGregor 1960 (p. 32, pi. 11, figs. 13, 14, text- fig. 2) is larger, and lacks the greatly elevated lips. Hystricosporites corystus Richardson 1962 (p. 173, pi. 25, figs. 1, 2) has a sparse ornament of spines, and a thinner exine. Occurrence. Upper Mimer Valley Series; Givetian. Hystricosporites porcatus (Winslow) comb. nov. Plate 95, figs. 4-6 1962 Dicrospora porcata Winslow, p. 52; pi. 11, figs. 4, 5, 5a; pi. 12, fig. 5; pi. 22, fig. 15. 1964 Hystrosporites costatus Vigran, p. 14, pi. 5, figs. 3-5. Description of specimens. Miospores trilete; amb circular to subcircular. Laesurae in- distinct, obscured by sinuous, elevated lips, 10-34 p high, 2-5 p wide, length f-f spore radius. Exine three-layered; intexine 2p or less thick, homogeneous, usually closely appressed to the exoexine and often indistinct; exoexine 6-12 p thick, inner exoexine infragranulate, outer exoexine and sculptural elements homogeneous. Contact areas, radial diameter f-f spore radius, each supporting 8-15 low, slightly sinuous muri, 2-8 p wide, 3-6 p high, 1-8 p apart, bordered by low curvaturae 6-10 p wide. Proximo- equatorial and distal surfaces support bulbous based grapnel-tipped spines, typically 15-30 p long (range 10-45 p), 17-23 round the equatorial margin. Dimensions. (Twenty-six specimens) Equatorial diameter 84-164 p (mean 111 p); radial diameter of contact area 26-44 p. Remarks. In ill-preserved specimens, the intexine separates from the exoexine, and is often folded. Section (PL 95, fig. 6) demonstrates clearly the low, wide, proximal radial muri; but in this strongly compressed specimen the stratification of the exine is not seen. Dimensions recorded for the Spitsbergen specimens are somewhat smaller than those recorded by Winslow (1962, p. 52). Comparisons. This is the only species recorded to date, which reports the presence of proximal, radial muri, though it would appear from the illustrated holotype that Hystricosporites delectabilis McGregor 1960 (pi. 11, fig. 13) may possess this feature. Occurrence. Upper Mimer Valley Series; Givetian. Hystricosporites mitratus sp. nov. Plate 95, figs. 7, 8 Diagnosis. Miospores trilete; amb circular to subcircular. Laesurae indistinct, length approximately half of the spore radius, accompanied by smooth, narrow often sinuous, elevated lips, 2-4 p wide, 16-21 p high. Exine 5-12 p thick, finely infragranulate; contact areas laevigate, small, radial diameter approximately half of the spore radius. 700 PALAEONTOLOGY, VOLUME 8 bounded by smooth, narrow elevated curvaturae, basally 3-7 p wide, 12-20 /x high; proximo-equatorial and distal surfaces support an ornament of grapnel-tipped spines, typically 20-35 p long (range 12-40 p), with enlarged, often bulbous bases, 12-25 round the equatorial margin. Dimensions. (Twenty-three specimens) Equatorial diameter 88-136 /x (mean 1 13 p); radial diameter of contact area 23—40 /x (mean 32 /x). Holotype. Preparation KA 148/5, 28-8 92-7, FI 598, N252. Locus typicus. Huginaspiskardet, Central Dicksonland, Vestspitsbergen ; Lower Mimer Valley Series, Eifelian. Description. Holotype subcircular, diameter 120 /x. Lips 18 /x high, each 2-3 /x thick. Curvaturae 18 p high, radial diameter of contact area 30 p. Grapnel-tipped spines 24- 44 /x long. Remarks. This species was not sectioned, and although no intexine was visible it seems very probable that the exinal stratification is similar to sectioned species of the genus Hystricosporites McGregor. The curvaturae are of the same dimensions as the lips, and probably formed during the same developmental stages within the tetrad, and formed only from the outer homogeneous layer of the exoexine. Comparison. Archaeotriletes honestus Naumova 1953 (p. 124, pi. 18, figs. 24, 25) has a thicker exine, thicker, lower curvaturae, and lacks lips, a feature associated with the curvaturae in H. mitratus sp. nov. Occurrence. Upper Reuterskioldfjellet Sandstone; and Lower Mimer Valley Series; Emsian and Lower Eifelian. Hystricosporites monosaccus (Archangelskaya) comb. nov. Plate 96, figs. 1, 2 1963 Archaeotriletes monosaccus Archangelskaya, p. 19, pi. 2, figs. 1-5. 1964 Hystricosporites coronatus Vigran, p. 15, pi. 3, figs. 1-3. Description of specimens. Miospores trilete; amb circular to subcircular. Laesurae obscured by high, membranous, sinuous lips, 20-72 /x high, individually 2-4 p wide at their base, narrowing upwards, length three-quarters to full spore radius. Exine three- layered ; intexine 1-3 /x thick, homogeneous, frequently partly separated from the exoexine; exoexine 8-18 /x thick, inner exoexine coarsely infragranulate, outer exoexine EXPLANATION OF PLATE 95 All figures x 500 unless otherwise stated ; from unretouched negatives. Figs. 1-8. Hystricosporites spp. 1-3. H. porrectus (Balme and Hassell) comb. nov. 1, Lateral view; KA 123/4, 35-6 94-0, G1358, N246. 2, Section showing the three-layered exine ; KA 123/S9, 47-5 102-7, G1358, N247. 3, Section (xlOOO) showing the homogeneous outer exoexine forming the spines, and the infra-granulate inner exoexine; KA 123/S10, 52-2 105-5, G1358, N248. 4-6. H. porcatus (Winslow) comb. nov. 4, Proximal surface; KA 261/4, 49-4 89-3, K891, N249. 5, Proximal oblique aspect (most of the grapnel-tipped spines have been eroded); KA 133/1, 42-3 93-1, G1385, N250. 6, Section showing proximal radial muri; KA 123/S4, 42-3 93-8, G1358, N251. 7-8. H. mitratus sp. nov. 7, Holotype, central oblique focus; KA 148/5, 28-8 92-7, FI 598, N252. 8, Proximal oblique aspect; KA 148/5, 25-0 98-4, F1598, N253. Palaeontology , Vol. 8 PLATE 95 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 701 homogeneous to infragranulate. Distally, the outer exoexine separates from the inner exoexine, and distends distally as a crumina, typically 70-90 p long (range 23-109 p). Proximal surface supports an ornament of radial muri, 6—10 yu- wide, up to 8 p high; distal surface with grapnel-tipped spines, 10-32 p high, with enlarged and frequently bulbous bases, more sparsely disposed on the crumina. Dimensions. (Twenty-two specimens) Overall equatorial diameter 63-143 fx (mean 99 p), polar diameter (including apical processes and crumina), 77-248 p (mean 172 /a). Remarks. This species is included within Hystricosporites McGregor 1960, rather than Archaeotriletes Naumova 1953, because of the absence of an equatorial extension. The great difference between the equatorial and polar axial measurements, results in the spores lying parallel to the longer polar axis, and because the proximal surface is flat, details of the contact areas are rarely seen. Sections (PI. 96, fig. 2) clearly demonstrate the construction of the distal crumina; this feature is not seen in other Spitsbergen Devonian species. Sections of Labiadensites fimbriatus (Waltz) Hacquebard and Barss 1957, in Dettmann and Playford 1962 (p. 679, pi. 96, figs. 1, 2) demonstrates the slight development of a cruminate cingulum. Comparison. H. monosaccus (Archangelskaya) comb. nov. differs from all other grapnel- tipped species, in possessing a crumina. Occurrence. Upper Mimer Valley Series; Givetian. Hystricosporites corystus Richardson 1962 Description of specimens. Miospores trilete, amb subcircular to subtriangular. Laesurae straight, length approximately three-quarters to full spore radius, frequently obscured by membranous, often sinuous, elevated lips, 34-68 p high, extending to the equatorial margin. Exine three-layered; intexine rarely seen, approximally 1 p thick, homogeneous; exoexine 14-30 p thick, inner exoexine infragranulate, outer exoexine and sculptural elements homogeneous. Contact areas laevigate, proximo-equatorial and distal surfaces support a sparse ornament of grapnel-tipped spines 30-60 p long, 8-13 p wide, 10-13 spines round the equatorial margin. Dimensions. (Fifteen specimens) Equatorial diameter 64-210 p (mean 141 p), polar diameter (includ- ing apical processes) 96-214 p (mean 151 p). Occurrence. Upper Mimer Valley Series; Givetian. Genus raistrickia (Schopf, Wilson, and Bentall 1944) Potonie and Kremp 1954 Type species. Raistrickia grovensis Schopf in Schopf, Wilson, and Bentall 1944. Raistrickia aratra sp. nov. Plate 96, figs. 3, 4 Diagnosis. Miospores trilete ; amb circular to subcircular. Laesurae indistinct, straight, length spore radius, simple or rarely accompanied by low narrow lips. Exine 2-6 p thick, homogeneous ; supporting a very variable sculpture of predominantly high baculo-verrucae, 3-12 p wide, 6-14 p high, with occasional rugulae, 5-7 p wide, 7-10 p 702 PALAEONTOLOGY, VOLUME 8 high, up to 25 /x long, cones and spines 2-10 p wide, 4-12 p high, the cones often with papillate tips. Concentration of sculptural elements variable, but they are always sparser and more reduced proximally, where spines and cones are more frequent. Dimensions. (Twenty specimens) Equatorial diameter 38-92 p (mean 61 p). Holotype. Preparation KA 243/1, 27-0 92-3, K767, N256. Locus typicus. East Munindalen, Central Dicksonland, Vestspitsbergen ; Planteklofta Conglomerate ; probable Upper Givetian. Description. Holotype circular, diameter IS p. Laesurae simple, straight, length two- thirds of spore radius. Exine 4 p thick, variable ornament of baculo-verrucae 9-13 /x high, rugulae, up to 8 /x wide and 25 /x long, cones up to 9 p high, often with papillate tips, and a few small spines. Sculpture reduced proximally. Remarks. Although the specimens appear well preserved, evidence from occasional speci- mens suggests that many of the verrucae, spines and cones, are the result of corrosion of bacula and rugulae by a splitting and ‘rounding off’ of originally high, flat-topped sculptural elements. This species is included with the genus Raistrickia on the basis of predominant baculo-verrucae and variable sculptural elements. Comparison. Lophozonotriletes macrogrumosus Kedo 1957 (pi. 4, figs. 21, 22) has a thicker wall, and lower more rounded processes. Raistrickia irregularis Kosanke 1950 (p. 47, pi. 11, fig. 5) has longer laesurae accompanied by lips, and smaller sculptural elements. Raistrickia ? gibberosa Hacquebard 1957 (p. 310, pi. 2, fig. 1) has longer lae- surae and a more regular ornament. R. cf. clavata Vigran 1964, p. 16, pi. 2, fig. 10, is smaller, and never has sculptural elements over 9 p high, but is otherwise similar. Occurrence. Upper Mimer Valley Series ; Givetian. Genus bullatisporites gen. nov. Type species. Bullatisporites bullatus sp. nov. Diagnosis. Miospores trilete; amb circular to subcircular. Exine sculptured with pila, the caput frequently supporting a small spine ; sculpture somewhat reduced proximally. Contact areas occasionally depressed, and sometimes bounded by curvatural folds. EXPLANATION OF PLATE 96 All figures x 500 unless otherwise stated; from unretouched negatives. Figs. 1-2. Hystricosporites monosaccus (Archangelskaya) comb. nov. 1, Lateral view; KA 255/M2, 28-6 108-0, K760, N254. 2, Section showing three-layered exine and crumina; KA 255/S3, 54-4 101-0, K760, N255. Figs. 3-4. Raistrickia aratra sp. nov. 3, Holotype, proximal oblique aspect; KA 243/1, 27-0 92-3, K767, N256. 4, Lateral view; KA 243/1, 52-3 96-8, K767, N257. Figs. 5-7. Bullatisporites bullatus gen. et sp. nov. 5, Holotype, proximal surface; KA 258/1, 55-7 101-8, K922, N258. 6, Proximal surface; KA 258/1, 27-8 88-3, K922, N259. 7, (x 1000) Details of the pila on the holotype. Figs. 8-13. Convolutispora spp. 8, C. vermiformis Hughes and Playford, Distal surface ; KA 261/4, 49-0 98-2, K891, N260. 9-13. C. disparalis sp. nov. 9, Holotype, optical section; KA 243/2, 54-9 103-8, K767, N261. 10, Distal surface; KA 293/1, 48-9 93-4, K556, N262. 11, Proximal surface; KA 243/1, 41-8 105-9, K767, N263. 12, (x 1000) Details of sculpture; KA 243/3, 54-4 88-9, K767, N264. 13, Holotype (x 1000). Palaeontology , Vol. 8 PLATE 96 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 703 Comparison. Differs from other genera within the Infraturma Apiculati, in being both circular and pilate. Dibolisporites Richardson (1965) includes species with variable orna- ment which may possess pila, but the sculptural elements are predominantly biform. Derivation of name. L. bulla — knob, boss. Bullatisporites bullatus sp. nov. Plate 96, figs. 5-7 Diagnosis. Miospores trilete ; originally spherical, amb circular. Laesurae straight, length J-f spore radius, accompanied and frequently obscured by lips, 1-3 p wide, up to 5 p high. Exine 1-2 p thick, infragranulate ; proximo-equatorial and distal surfaces densely covered with pila, caput 1-2 p wide, collum 0-5-1 -5 p wide, 1-3 p high, the caput fre- quently supports at its apex a minute spine. The contact areas support a somewhat re- duced sculpture of variable concentration and extent; curvaturae sometimes present in the radial regions. Major and minor folding common. Dimensions. (Twenty-two specimens) Equatorial diameter 84-112 p (mean 100 p). Holotype. Preparation KA 258/1, 55-7 101-8, K922, N258. Locus typicus. South Mimerdalen, Central Dicksonland, Spitsbergen; Reuterskioldfjellet Sandstone, Emsian. Description. Holotype 98 p. Laesurae just perceptible (under oil), length approximately half of the spore radius. Lips each 1-5 p wide, slightly raised. Exine 1 p thick, two major distal folds; proximo-equatorial and distal pila 1-2-5 p high, caput 1-2 p wide, collum 0-5-1 -5 p wide. In the contact areas, pila reduced only in the region adjacent to the lips. Remarks. Lack of preferred orientation suggests a spherical shape. Curvaturae imper- fectae are present in approximately one-quarter of the specimens. Only under oil immersion can the finer details of the pila be seen. Comparison. Retusotriletes gibberosus Naumova nom. nud. 1953 (pi. 22, fig. 110) see Kedo 1955 (p. 21, pi. 1, fig. 16) lacks typification and description; however, the specimen illustrated by Kedo is clearly similar to B. bullatus sp. nov. and may be conspecific. Kedo states that the ‘protuberances are round’ but his illustration indicates that they may be pila. Occurrence. Dicksonfjorden Sandstone, Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series ; Siegenian to Eifelian. Infraturma murornati Potonie and Kremp 1954 Genus convolutispora Hoffmeister, Staplin, and Malloy 1955 Type species. Convolutispora florida Hoffmeister, Staplin, and Malloy 1955. Convolutispora vermiformis Hughes and Playford 1961 Plate 96, fig. 8 1957 Convolutispora flexuosa forma minor Hacquebard, p. 312, pi. 2, fig. 10. Dimensions. (Seventeen specimens) Equatorial diameter 37-52 p (mean 46 p). 704 PALAEONTOLOGY, VOLUME 8 Remarks. The Spitsbergen Devonian specimens are somewhat smaller than those described from other areas. Occurrence. Upper Mimer Valley Series; Givetian. Convolutispora disparalis sp. nov. Plate 96, figs. 9-13 Diagnosis. Miospores trilete; amb circular to subcircular, equatorial margin undulating. Laesurae indistinct, simple, straight, length |-f spore radius. Exine excluding ornament 1-2 p thick, homogeneous ; distal and proximo-equatorial surfaces sculptured with a variable ornament of sinuous, narrow, frequently anastomosing rugulo-cristae, 1-3 p wide, decreasing in width from their base, 2-5 p high, usually with an undulating crest, often forming an imperfect reticulum, the intersections of which occasionally support a small papillum or spine. Lumina irregular, up to 6 p in longest diameter. Contact areas support a reduced, sparse ornament, of low, smooth rugulae, small cones and granules. Dimensions. (Sixty-one specimens) Equatorial diameter (excluding muri) 31-52 /x (mean 40 /x). Holotype. Preparation KA 243/2, 54-9 103-8, K767, N261. Locus typicus. East Munindalen, Central Dicksonland, Spitsbergen; Planteklofta Conglomerate, probable Upper Givetian. Description. Holotype subcircular, diameter 44 p. Laesurae length § spore radius. Exine approximately 1 p thick; rugulo-cristae 1-2 p wide, 1-3 p high, the intersections often with papillate tips. Comparison. The presence of a reduced proximal sculpture, and of papillae on many of the intersections, separates this species from Convolutispora fromensis Balme and Hassell 1962 (p. 8, pi. 1, figs. 14-16) which in addition has lower muri; and Convoluti- spora venusta Hoffmeister, Staplin, and Malloy 19556 (p. 385, pi. 38, fig. 1 1) which also has wider, lower ridges. Occurrence. Common throughout the Upper Mimer Valley Series; Givetian. Convolutispora mimerensis (Vigran) comb. nov. Plate 97, figs. 1-3 1964 Reticulatisporites mimerensis Vigran, p. 17, pi. 2, figs. 16, 17. Description of specimens. Miospores trilete; amb convexly subtriangular to circular. Laesurae simple, straight, length |-| spore radius, frequently obscured by sculptural elements. Exine 5-8 p thick (including muri), supporting low, smooth, slightly sinuous muri, 2-5 p broad, 1-3 p high, most of which anastomose, forming an imperfect reticu- lum. Lumina irregular, 7 p wide in longest diameter. Dimensions. (Twenty specimens) Equatorial diameter 36-68 p (mean 53 p). Remarks. Variation occurs in the orientation and spacing of the muri which are often radially directed and more widely spaced equatorially (PI. 97, fig. 2). I have included this species in Convolutispora on the basis of its imperfect rather than perfect reticulum. K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 705 Comparison. Convolutispora crassa Playford 1962 (p. 594, pi. 81, figs. 10-12) is larger, has a thicker exine and flat-topped muri. Camptotriletes simplex Naumova nom. nud. in Kedo 1955 (p. 26, pi. 2, fig. 14) may be conspecific, but the illustration is too poor to see details of the sculpture, and the species lacks holotypic designation. Occurrence. Upper Mimer Valley Series; Givetian. Convolutispora tegula sp. nov. Plate 97, figs. 4-8 Diagnosis. Miospores trilete; amb circular to subcircular, occasionally oval. Laesurae frequently indistinct, simple, straight, length §-§ spore radius. Exine including ornament 5-9 p thick, homogeneous, often punctate, with a crowded sculpture of short, broadly rounded, partly anastomosing muri, 2-7 p wide, 2-4 p high, the shorter muri often poly- gonal in outline, lumina subordinate, less than 1 /x wide. Dimensions. (Thirty-five specimens) Equatorial diameter 41-70 p (mean 53 p). Holotype. Preparation KA 286/2, 27-4 107-8, K773, N267. Locus typicus. East Munindalen, Central Dicksonland, Spitsbergen; Planteklofta Conglomerate, probable Upper Givetian. Description. Holotype circular, diameter 68 /x. Laesurae indistinct, length approximately two-thirds spore radius. Exine punctate, 8 p thick, muri 3-7 p wide, 2-4 /x high. Remarks. Plate 97, fig. 7, shows corrosion of the muri. Comparison. Convolutispora fromensis Balme and Hassell 1962 (p. 8, pi. 1, figs. 14-16) closely resembles C. tegula sp. nov. in sculptural pattern, but has a much thinner exine and narrower muri. Convolutispora florida Hoffmeister, Staplin, and Malloy 19556 (p. 384, pi. 38, figs. 5, 6) has a more extensively anastomosing muroid pattern, and wider lumina. Convolutispora usitata Playford 1962 (p. 595, pi. 82, figs. 4, 7, and 8) has similar sculpture, but is much larger. Occurrence. Upper Mimer Valley Series; Givetian. Genus reticulatisporites (Ibrahim) Potonie and Kremp 1954 Type species. Reticulatisporites reticulatus Ibrahim 1933. Reticulatisporites emsiensis sp. nov. Plate 97, figs. 9-1 1 Diagnosis. Miospores trilete ; amb circular to sub-circular. Laesurae distinct to discernible, length |-f spore radius, simple or more usually accompanied by low narrow folds. Exine 2-4 p thick (excluding muri), infra-granulate ; proximal surface laevigate or more frequently sparsely sculptured with small verrucae or granules, 2 /x or less in height and width, distally sculptured with strongly developed muri 1-4 p wide, 3-8 p high, enclosing more or less uniform lumina 8-20 /x in longest diameter, polygonal in outline centrally, more rounded equatorially. 706 PALAEONTOLOGY, VOLUME 8 Dimensions. (Twenty-three specimens) Equatorial diameter 49-82 p (mean 65 p). Number of equa- torial muri 1 1-20, number of distal lumina 14-36. Holotype. Preparation KA 240/2, 37-4 101 5, K582, N269. Locus typicus. Manchesterbreen spur, Central Dicksonland, Spitsbergen; Lower Mimer Valley Series, Emsian. Description. Holotype circular, diameter 72 p. Laesurae simple, length two-thirds spore radius. Exine 4 p thick, very sparse proximal ornament of granules, distal muri 2-3 p wide basally, narrowing upwards, 5-8 p high, lumina 12-20 p wide. Remarks. Both Naumova (1953) and Chibrikova (1959) include species assignable to Reticulatisporites, within Archaeozonotriletes. Comparison. Small size and exclusively distal reticulate sculpture, separates Reticulati- sporites emsiensis sp. nov. from the majority of other well-described species of Devonian and Carboniferous age. Reticulatisporites speciosus Hacquebard and Barss 1957 (p. 18, pi. 2, fig. 7) has an exclusively distal reticulum, but is larger, has longer laesurae, wider lips and a thicker exine. Archaeozonotriletes retiformis Naumova 1953 (p. 87, pi. 14, fig. 4) is very similar and may be conspecific, but no mention is made of a granulate proximal surface or a reticulum confined to the distal surface. Occurrence. Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Emsian. Reticulatisporites sp. cf. Dictyotriletes minor Naumova 1953 Plate 97, figs. 12, 13 Description of specimens. Miospores trilete ; amb circular to subcircular. Laesurae simple, straight, length approximately two-thirds spore radius. Exine 1-2 p thick, finely infra- EXPLANATION OF PLATE 97 All figures x 500 unless otherwise stated ; from unretouched negatives. Figs. 1-8. Convolutispora spp. 1-3. C. mimerensis (Vigran) comb nov. 1, Proximal surface; KA 286/3, 20-8 97-0, K773, N265. 2, Distal surface; KA 227/1, 38-0 98-9, K623, N266. 3, (x 1000) Sculptural details (proximal). 4-8. C. tegula sp. nov. 4-6, Holotype, proximal, sectional and distal foci; KA 286/2, 27-4 107-8, K773, N267. 7, Poorly preserved specimen; KA 241/3, 28-5 95-0, K767, N268. 8, (x 1000) Sculptural details of the holotype. Figs. 9-13. Reticulatisporites spp. 9-11. R. emsiensis sp. nov. 9, 10, Holotype, proximal and distal surfaces respectively; KA 240/2, 37-4 101-5, K582, N269. 11, Distal surface; KA 281/6, 57-0 107-5, K850, N270. 12-13. R. sp. cf. Dictyotriletes minor Naumova, Proximal and distal surfaces respectively; KA 21 All, 36-0 107-3, K872, N271. Fig. 14. Perforosporites sp. Proximal surface; KA 209/M2, 32-0 97-7, K519, N272. Figs. 15-21. Emphanisporites spp. 15-18. E. decoratus sp. nov. 15, 16, Holotype, proximal and distal surfaces respectively; KA 251/3, 33-5 88-9, K905, N274. 17, 18, Proximal and distal surfaces respec- tively; KA 251/9, 20-6 96-5, K905, N275. 19. E. neglectus Vigran, Proximal surface; KA 162/1, 37-5 89-1, G1356, N273. 20. E. minutus sp. nov. Holotype, proximal surface; KA 262/4, 56-8 91-5, K908, N276. 21. E. patagiatus sp. nov. Holotype, proximal surface; KA 271/3, 38-1, 99-3, K897, N277. Figs. 22-23. Craspedispora craspeda gen. et sp. nov. 22, Holotype, proximal surface; KA 258/3, 42-8 89-4, K922, N278. 23, Holotype ( X 1000). Figs. 24-25. Diatomozonotriletes sp. 24, Distal surface (x 1000); KA 243/2, 54-1 103-6, K767, N279. 25, Proximal surface; KA 229/1, 38-0 94-8, K550, N280. Palaeontology, Vol. 8 PLATE 97 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 707 granulate, contact areas laevigate, proximo-equatorial and distal surfaces sculptured with narrow, low muri, basally 1 /x or less wide, narrowing upwards, approximately 1 /x high; lumina 3-7 /x in longest diameter. Dimensions. (Five specimens) Equatorial diameter 18-32 /x (mean 27 /x). Comparison. Too few specimens are present for exact comparison: Dictyotriletes minor Naumova 1953 (p. 28, pi. 2, fig. 7) is clearly similar, but is triangular, and the muri are somewhat higher. However, the illustration of D. minor in Kedo 1955 (p. 2, pi. 2, fig. 13) has lower muri, is subcircular, and compares very closely with the Spitsbergen specimens. Occurrence. Reuterskioldfjellet Sandstone and Mimer Valley Series ; Emsian to Givetian. Genus perforosporites Scott and Rouse 1961 Type species. Perforosporites robustus Scott and Rouse 1961. Discussion. Separable from Foveosporites Balme 1957 on its regular distribution of fovea. Perforosporites sp. Plate 97, fig. 14 Descriptions of specimens. Miospores trilete; amb circular. Laesurae simple, straight, length J-f spore radius. Exine 6-10 /x thick, infragranulate ; contact areas laevigate, proximo-equatorial and distal surfaces with a sparse, regular ornament, of circular to oval fovea, 1-4 /x wide, 3-6 /x long, 3-24 /x apart. Distally the exine slightly ‘overhangs’ the fovea. Dimensions. (Two specimens) Equatorial diameter 116-17 p. Comparison. Perforosporites robustius Scott and Rouse 1961 (p. 978, pi. 113, figs. 1-6, pi. 114, figs. 1-5) the only species previously attributed to this genus, is roundly tri- angular, smaller, has less distinct and longer laesurae, and frequently has blunt, short papillae. Reticulatisporites textilis Balme and Hassell 1962 (p. 9, pi. 2, figs. 11, 12) has sculptured contact areas, indistinct laesurae, and more numerous fovea. Occurrence. Top of the Reuterskioldfjellet Sandstone, and bottom of the Lower Mimer Valley Series; Upper Emsian. Genus emphanisporites McGregor 1961 Type species. Emphanisporites rotatus McGregor 1961. Emphanisporites rotatus McGregor 1961 1962 Radforthia radiata Winslow, p. 72, pi. 16, figs. 15, 15a, pi. 22, fig. 17. Dimensions. (Ten specimens) Equatorial diameter 34-47 /x (mean 39 f). Comparison. Stenozonotriletes ornatissimus Naumova nom. nud. (pi. 22, fig. 11) lacks holotypic designation and description, but may have proximal muri, in which case it would be synonymous. Radiaspora sp. A Balme 1962 (p. 6, pi. 1, figs. 11, 12), may prove to have proximal rather than distal muri, and would then be conspecific with E. rotatus 708 PALAEONTOLOGY, VOLUME 8 McGregor. Radforthia radiata Winslow 1962 (p. 72, pi. 16, figs. 15, 15a, pi. 22, figs. 17) appears to differ only in its greater size range, and although neither McGregor (1961) nor Winslow (1962) give the mode of their measurements, R. radiata must be synony- mous at least in part, with E. rotatus. Occurrence. Reuterskioldfjellet Sandstone and Lower Mimer Valley Series; Siegenian to Lower Eifelian. Emphanisporites neglectus Vigran 1964 Plate 97, fig. 19 Description of specimens. Miospores trilete ; amb circular to roundly triangular. Laesurae distinct, straight, length §-f spore radius, narrow, occasionally accompanied by slightly sinuous lips, each up to 1 p wide. Exine 0-5-2* 5 p thick, infragranulate ; depressed con- tact areas support a distinct to discernible ornament of low, narrow muroid folds, sur- rounded by low, sharp, curvaturae ; proximo-equatorial and distal surfaces laevigate or very finely granulose. Dimensions. (Twenty-six specimens) Equatorial diameter 30-47 p (mean 40 p). Radial diameter of contact area approximately four-fifths of total spore radius. Comparison. Emphanisporites obscurus McGregor 1961 (p. 5, pi. 1, fig. 14), is larger and lacks curvaturae. Occurrence. Dicksonfjorden Sandstone, Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Siegenian, Emsian. Emphanisporites decoratus sp. nov. Plate 97, figs. 15-18 Diagnosis. Miospores trilete; amb roundly triangular to subcircular and occasionally oval. Laesurae distinct, straight, length two-thirds to full spore radius, occasionally accompanied by low, smooth lips, individually less than 1 p wide. Exine 1-4 p thick, homogeneous to infragranulate; proximal surface with distinct to discernible radially disposed muroid folds of variable number, 2 p or less wide, distal surface supporting an ornament of cones and spines, 1-5-5 p high, 0-5-2 p wide. Dimensions. (Twenty-nine specimens) Equatorial diameter 34-61 p (mean 49 p). Holotype. Preparation KA 251/3, 33-5 88-9, K905, N274. Locus typicus. South Mimerdalen, Central Dicksonland, Spitsbergen; Reuterskioldfjellet Sandstone, Siegenian. Description. Holotype roundly triangular, diameter 51 p. Laesurae simple, almost reach- ing equatorial margin. Exine 3 p thick; proximal muri indistinct, distal surface support- ing a predominance of cones, 1 p wide, up to 1-5 /x high. Remarks. Although cones usually predominate, specimens with a dominance of spines exist. Comparison. Differs from other species of Emphanisporites yet described in having a strongly sculptured distal surface. Emphanisporites neglectus Vigran (PI. 97, fig. 19) is occasionally sculptured, but then, only with very fine granules. Occurrence. Lower Reuterskioldfjellet Sandstone, Siegenian. K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 709 Emphanisporites minutus sp. nov. Plate 97, fig. 20 Diagnosis. Miospores trilete ; amb roundly triangular to subcircular. Laesurae straight, length three-quarters to full spore radius, accompanied by smooth, narrow lips, in- dividually less than 1 p wide. Exine 2-3 p thick, homogeneous to infragranulate ; proximal surface with 15-30 radially disposed muri, approximately 1 p wide, distal surface laevigate. Dimensions. (Twenty specimens) Equatorial diameter 16-27 p (mean 24 p). Holotype. Preparation KA 262/4 56-8 91-5, K908, N276. Locus typicus. South Mimerdalen, Central Dicksonland, Spitsbergen; Reuterskioldfjellet Sandstone, Siegenian. Description. Holotype subtriangular, diameter 26 p. Lips extend almost to the equatorial margin. Exine 2 p thick; proximal surface with twenty-five radially disposed ribs, 0-5—1 p wide. Comparison. Emphanisporites rotatus McGregor 1961 (p. 3, pi. l,figs. 1-4) is considerably larger and the laesurae are frequently unaccompanied by lips. Occurrence. Fraenkelryggen Division and Reuterskioldfjellet Sandstone, Gedinnian to Lower Emsian. Emphanisporites patagiatus sp. nov. Plate 97, fig. 21 Diagnosis. Miospores trilete; amb subcircular to subtriangular, periphery often un- dulating. Laesurae simple, straight, length |— § spore radius. Exine 2-6 p thick, homo- geneous to infragranulate; proximal surface with 8-15 radially disposed muri, 3-14 p wide, up to 4 p high, extending spore radius from the equatorial margin, proximal polar region and distal surface laevigate. Dimensions. (Fifteen specimens) Equatorial diameter 33-50 p (mean 45 p). Holotype. Preparation KA 271/3, 38-1 99-3, K897, N277. Locus typicus. South Munindalen, Central Dicksonland, Spitsbergen; Reuterskioldfjellet Sandstone, Emsian. Description. Holotype roundly triangular, diameter 49 p. Laesurae length two-thirds of spore radius. Exine 6 p thick, finely infragranulate, nine proximal radial ribs, 4-13 p wide, extend spore radius from the equatorial margin. Comparison. Emphanisporites robustius McGregor 1961 (p. 4, pi. 1, fig. 13) is larger and lacks the laevigate proximal polar region. Occurrence. Reuterskioldfjellet Sandstone; Siegenian and Emsian. Subturma zonotriletes Waltz 1935 Infraturma tricrassati Dettmann 1963 Genus craspedispora gen. nov. Type species. Craspedispora craspeda sp. nov. Diagnosis. Miospores trilete; amb subcircular to roundly triangular. Laesurae usually 3 a B 6612 710 PALAEONTOLOGY, VOLUME 8 well defined, simple or accompanied by lips. Central area encompassed inter-radially by a narrow zona. Central area sculptured proximo-equatorially and distally. Discussion. Differs from both Reinschospora Schopf, Wilson, and Bentall 1944 and Diatomozonotriletes (Naumova) Playford 1962 in having a more roundly triangular shape, and a nonfimbriate zona rather than a corona. Camarozonotriletes (Naumova) Potonie 1958 has a thick interradial crassitude. Derivation of name. Gr. krespedon — edge, border. Craspedispora craspeda sp. nov. Plate 97, figs. 22, 23 Diagnosis. Miospores trilete ; amb subcircular to roundly triangular. Laesurae straight, distinct, length ■£-§ spore radius, simple or accompanied by narrow, smooth lips in- dividually 0-5-1 p wide. Exine of central area 1-2-5 p thick, sometimes slightly thicker inter-radially, homogeneous to finely infragranulate, extending interradially as a thin, membranous zona 2-5 p wide, the zona never develops in the radial regions. Contact areas often slightly darkened, laevigate, proximo-equatorial and distal surfaces of the central area with a distinct ornament of cones, 1-5 p or less in height and basal diameter. Zona laevigate, or occasionally with a few small cones. Curvatural folds sometimes present. Dimensions. (Thirty-four specimens) Equatorial diameter (including zona) 35-42 p (mean 39 p). Holotype. Preparation KA 258/3, 42-8 89-4, K922, N278. Locus typicus. South Mimerdalen, Central Dicksonland, Spitsbergen; Reuterskioldfjellet Sandstone, Emsian. Description. Holotype roundly triangular, diameter 41 p. Laesurae length two-thirds of spore radius. Exine 2 p thick, slightly less at the apices, zona 2-5 p wide. Remarks. Corrosion of the thin zona frequently occurs, giving the outer margin an irregular appearance. In oblique aspect, the zona is very difficult to see. Occurrence. Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Emsian and Eifelian. Genus diatomozonotriletes (Naumova) Playford 1962 Type species. Diatomozonotriletes saetosus (Hacquebard and Barss 1957) Hughes and Playford 1961. Diatomozonotriletes sp. Plate 97, figs. 24, 25 Description of specimens. Miospores trilete; amb triangular, with straight to slightly concave sides, and rounded apices. Laesurae indistinct, simple, straight, length approxi- mately three-quarters spore radius. Corona composed of discrete, closely spaced, pointed saetae, 2-5 to 3 p long interradially, diminishing in size towards the triangular apices, from which they are absent. Exine of central area approximately 1 p thick, finely infragranulate, proximally laevigate, distally supporting an ornament of small discrete cones, 1-5 p or less in height and basal diameter. Exine folded. K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 711 Dimensions. (Three specimens) Overall equatorial diameter 25-27 p. Remarks. Although too few specimens are present to warrant specific assignment, these specimens are interesting in that they provide the lowest stratigraphical record for this genus. Occurrence. Planteklofta Conglomerate and Fiskeklofta Formation; probable Upper Givetian. Infraturma cingulati (Potonie and Klaus) Dettmann 1963 Genus stenozonotriletes (Naumova) Potonie 1958 Type species. Stenozonotriletes conformis Naumova 1953. Stenozonotriletes furtivus sp. nov. Plate 98, figs. 2, 3 Diagnosis. Miospores trilete; amb roundly triangular to subcircular. Laesurae distinct, straight, length f-f spore radius, simple or more frequently accompanied by low, flat- topped lips, individually 2-3 p wide at the polar end, decreasing gradually equatorially. Exine laevigate to finely punctate ; cingulum smooth, uniform or slightly narrower inter- radially. Dimensions. (Seventeen specimens) Overall equatorial diameter 65-108 p (mean Sip); width of cingulum 10-19 p (mean 12 p). Holotype. Preparation KA 274/4, 44-8 94-4, K872, N281. Description. Holotype roundly triangular, overall diameter 70 p. Laesurae half spore radius, accompanied by low thickened lips, individually 3 p wide. Cingulum uniform, 10 p wide. Remarks. Frequently the cingulum is eroded, giving it an irregular appearance. Comparison. Archaeozonotriletes subcompactus Naumova 1955 (p. 84, pi. 13, fig. 14 and p. 35, pi. 13, fig. 9) is circular and has a sculpture of small protuberances. Triletes dubius Eisenack 1944 (p. 115, pi. 2, fig. 7, text-fig. 14) has a narrower cingulum and the exine thickens towards the polar axis, forming a darkened triangular area, unlike the lip formation in S. furtivus sp. nov. Occurrence. Dicksonfjorden Sandstone, Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series ; Upper Siegenian to Eifehan. Stenozonotriletes insessus sp. nov. Plate 98, fig. 1 Diagnosis. Miospores trilete; amb convexly subtriangular. Laesurae straight, extending almost to the equatorial margin, accompanied by narrow lips, each approximately 0-5 p wide, up to 2-5 p high. Exine homogeneous to finely infragranulale, laevigate. Cingulum uniform. Dimensions. (Twenty specimens) Overall equatorial diameter 24-36 p (mean 30 p) ; width of cingulum 3-6 p (mean 4 p). Holotype. Preparation KA 223/2, 35-5 89-7, K829, N283. 712 PALAEONTOLOGY, VOLUME 8 Locus typicus. South Mimerdalen, Central Dicksonland, Spitsbergen; Reuterskioldfjellet Sandstone, Emsian. Description. Holotype overall diameter 36 p, lips individually 0-5 p wide, 2 p high. Cingulum 4 p wide. Comparison. Archaeozonotriletes pusillus Naumova 1953 (p. 86, pi. 13, fig. 19) has simple laesurae which often have ‘clavate endings’; but is otherwise similar. Stenozonotriletes extensus Naumova var. minor Naumova 1953 (p. 37, pi. 3, fig. 18; p. 72, pi. 10, fig. 21 ; p. 130, pi. 19, figs. 19, 20) is more coarsely infragranulate and has simple laesurae. Occurrence. Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series ; Siegenian to Eifelian. Stenozonotriletes sp. Plate 98, fig. 4 Description of specimens. Miospores trilete; amb convexly subtriangular. Laesurae straight or slightly sinuous, length three-quarters to full central area radius. Exine finely punctate; cingulum smooth, usually slightly thicker interradially. Dimensions. (Six specimens) Overall equatorial diameter 54-66 p (mean 58 p), width of cingulum 6-8 p interradially, 4-6 p radially. Comparison. Too few specimens are present to warrant the erection of a new species. Stenozonotriletes perforatus Playford 1962 (p. 607, pi. 86, figs. 8-9, text-fig. 5c), has a narrower, laevigate cingulum, but is otherwise similar. Stenozonotriletes sp. cf. S. recognitus var. recognitus Naumova in Balme and Hassell 1962 (p. 14, pi. 3, figs. 3, 4) differs only in having a darker contact area. Occurrence. Upper Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Upper Emsian, Lower Eifelian. Genus lycospora (Schopf, Wilson, and Bentall) Potonie and Kremp 1954 Type species. Lycospora micropapillata (Wilson and Coe) Schopf, Wilson, and Bentall 1944. EXPLANATION OF PLATE 98 All figures x 500 ; from unretouched negatives. Pigs. 1-4. Stenozonotriletes spp. 1, S. incessus sp. nov. Holotype, sectional focus; KA 223/2, 35-5 89-7, K829, N283. 2-3. S. furtivus sp. nov. 2, Holotype, proximal surface; KA 274/4, 44-8 94-4, K872, N281. 3, Proximal surface; KA 209/2. 42-7 92-1, K519, N282. 4. S. sp. Sectional focus; KA 230/4, 36-0 101-7, K638, N284. Pigs. 5-6. Densosporites devonicus Richardson. 5, Proximal surface; KA 209/Ml, 63-8 90-8, K519, N285. 6, Section; KA 286/S12, 32-6 95-8, K773, N286. Figs. 7-8. Lycospora culpa sp. nov. Holotype, proximal and distal surfaces respectively; KA 262/4, 57-7 106-4, K908, N287. Figs. 9-16. Samarisporites spp. 9-10. S. praetervisus (Naumova) comb, nov., Proximal and distal surfaces respectively; KA 243/1, 52-2 98-4, K767, N288. 11, S. senotus sp. nov. Holotype, sectional focus; KA 243/2, 32-8 105-3, K767, N289. 12-16. S. hesperus sp. nov. 12, 13, Holotype, proximal and distal surfaces respectively; KA 243/3, 51-7 102-7, K767, N290. 14, Distal surface; KA 243/3, 24-7 93-4, K767, N291. 15, Distal surface showing cristo-reticulate sculpture; KA 243/3, 37-6 103-4, K767, N292. 16, Section, showing the cingulate exine and distal sculpture; KA 243/S11, 43-4 90-8, K767, N293. Palaeontology, Vol. 8 PLATE 98 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 713 Lycospora culpa sp. nov. Plate 98, figs. 7, 8 Diagnosis. Miospores trilete; amb roundly triangular with acute to broadly rounded apices, occasionally subcircular, conformable with the central area outline. Laesurae indistinct, straight, length two-thirds to full central area radius, accompanied by narrow, smooth, low lips, individually 0-5-1 -5 p wide, length three-quarters to full spore radius. Exine two-layered; intexine approximately 1 p thick, closely appressed to the exoexine and usually indistinct, exoexine of both central area and equatorial flange finely infra- granulate. Proximal surface laevigate, distal surface of central area sparsely covered with small cones and granules 1 -5 p or less in height and basal diameter, equatorial flange laevigate. Proximal surface usually with numerous folds. Dimensions. (Twenty specimens) Overall equatorial diameter 33-60 p (mean 47 p), width of cingulum 2-14 p (mean 6 p). Holotype. Preparation KA 262/4, 57-7 106-4, K908, N287. Locus typicus. South Mimerdalen, Central Dicksonland, Spitsbergen; Lower Reuterskioldfjellet Sandstone, Siegenian. Description. Holotype roundly triangular with convex sides and acute apices, diameter 51 /x, width of cingulum 6 /x. Laesurae indistinct, lips 1 p wide, extending full radius of the central area. Intexine 1 p thick, exoexine proximally folded, distally with grana and small cones on the central area. Remarks. Some specimens are slightly thicker at the very inner margin of the cingulum, but in no specimen is it truly bizonate. This species provides a very low stratigraphical record for the genus Lycospora. Comparison. Lycospora uber (Hoffmeister, Staplin, and Malloy) Staplin 1960 (p. 20, pi. 4, figs. 13, 17, 18, 20) is smaller, and only faintly granulose. Hymenozonotriletes mille- granus Naumova 1953 (p. 126, pi. 18, figs. 31, 32) is bizonate, and more densely sculp- tured. Hymenozonotriletes mancus Naumova 1953 (p. 63, pi. 8, fig. 17) has a laevigate distal central area. Hymenozonotriletes limpidus Naumova 1953 (p. 98, pi. 15, fig. 21) is circular, lacks lips, and is more densely sculptured. Occurrence. Lower Reuterskioldfjellet Sandstone, Siegenian. Genus densosporites (Berry) Potonie and Kremp 1954 Type species. Densosporites covensis Berry 1937. Densosporites devonicus Richardson 1960 Plate 98, figs. 5, 6 Dimensions. (Forty specimens) Overall equatorial diameter 66-156 p (mean 99 p)\ total width of cingulum 18-40 p (mean 26 p), inner zone 12-27 p (mean 18 p). Remarks. The section of D. devonicus Richardson (PI. 98, fig. 6) demonstrates the raised lips, distal sculpture, and thick exoexine. The intexine cannot be seen in this section. Occurrence. Upper Mimer Valley Series; Givetian. 714 PALAEONTOLOGY, VOLUME 8 Genus samarisporites Richardson 1965 Type species. Samarisporites ( Cristatisporites ) orcadensis (Richardson) Richardson 1965. Discussion. Samarisporites triangulatus sp. nov. and Samarisporites inusitatus sp. nov. (see below) both have equatorial structures which vary from a zona to a cingulum (here referred to as an equatorial flange), giving evidence in support of Dettmann’s (1963) emendation of the infraturma Cingulati (Potonie and Klaus 1954), to include zonate forms. Samarisporites praetervisus (Naumova) comb. nov. Plate 98, figs. 9, 10 1953 Hymenozonotriletes praetervisus Naumova, p. 40, pi. 4, fig. 8. Description of specimens. Miospores trilete; amb roundly triangular to subtriangular. Laesurae indistinct, obscured by smooth elevated lips, individually 1-2 p wide, 5-10 p high, almost extending to the equatorial margin. Exine two-layered; intexine homo- geneous 2 p or less wide, closely appressed to the exoexine and often indistinct; exo- exine of both central area and cingulum finely to coarsely infragranulate, proximally laevigate, distal surface densely ornamented with cones, 2-7 p wide, 3-8 p high. Cones on the central area variable both in shape and distribution, commonly with rounded apices, often supporting an apical capitate spine, the cones are usually closely spaced, with rounded polygonal or occasionally fused bases. Cones on the cingulum less dense, with pointed to slightly rounded apices, 33-52 cones round the equatorial margin. Cingulum uniform, tapering, differentiation from the central area distinct to obscure. Dimensions. (Sixteen specimens) Overall equatorial diameter 70-124 p (mean 86 p); width of cingulum 16-30 p (mean 21 p). Comparison. Samarisporites orcadensis (Richardson 1960, p. 58, pi. 14, fig. 12, text-fig. 8) Richardson 1965, is considerably larger. Occurrence. Upper Mimer Valley Series; Givetian. Samarisporites senotus sp. nov. Plate 98, fig. 1 1 Diagnosis. Miospores trilete; amb convexly triangular, irregular, central area outline circular. Laesurae straight, length two-thirds to full radius of the central area, accom- panied by smooth, narrow, often sinuous lips, individually 1-1-5 p wide, extending on to the cingulum, usually to the equatorial margin. Exine two-layered; intexine thin, approximately 1 p wide, closely appressed to the exoexine, and often indistinct; exoexine of both central area and cingulum finely infra-granulate. Proximal surface laevigate, distal surface very sparsely covered with small spines and cones, 1-4 p wide, 2-4 p high, the cones frequently support a small apical spine. Cingulum of uniform width, proximally occasionally raised above the central area, the inner distal margin may be slightly thickened, but the cingulum is never distinctly bizonate. Dimensions. (Sixteen specimens) Overall equatorial diameter 50-64 p (mean 57 p); width of cingu- lum 12-16 p (mean 15 p). Holotype. Preparation KA 243/2, 32-8 105-3, K767, N289. K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 715 Locus typicus. East Munindalen, Central Dicksonland, Spitsbergen; PlantekMta Conglomerate, probable Upper Givetian. Description. Holotype diameter 57 p, cingulum 12 p wide. Laesurae straight, length two- thirds central area radius, lips each 1 p wide, extending almost to the equatorial margin. Distal spines and cones 1-3 p wide, 2-3 p high, somewhat corroded. Remarks. In corroded specimens, the sculptural elements may be partially or completely eroded. Comparison. Hymenozonotriletes spinosus Naumova 1953 (p. 41, pi. 14, fig. 9) is much larger, and has a more regular ornament. Zonotriletes deliquescens Luber, in Luber and Waltz 1941 (p. 50, pi. 1, fig. 6), is larger, and the inner part of the cingulum is not thickened. Hymenozonotriletes pusillus (Ibrahim) Ishchenko 1952 (p. 50, pi. 13, fig. 122) is smaller, the lips are shorter, and there is a sparser ornament on the cingulum. Densosporites landesii Staplin 1960 (p. 25, pi. 5, fig. 11) has an ornament only of cones, shorter lips, and a distinctly bizonate cingulum. Occurrence. Upper Mimer Valley Series; Givetian. Samarisporites hesperus sp. nov. Plate 98, figs. 12-16 Diagnosis. Miospores trilete; amb circular to subcircular conformable with the central area outline. Laesurae indistinct, length three-quarters to full central area radius, accom- panied and usually obscured by narrow, often slightly sinuous, elevated lips, individually 1-2-5 p wide, extending to the inner margin of the cingulum. Exine one-layered, finely infra-granulate, proximally laevigate, distally with an even distribution of cones on both central area and cingulum. The cones 2-6 p wide, 2-6 p high, have rounded apices, and a capitate apical spine in well-preserved specimens ; the bases of the cones on the central area are usually fused, often to such an extent as to form an imperfect reticulum. Cingulum uniform, tapering abruptly at the outer margin, differentiation from the central area is sometimes indistinct. Dimensions. (Fifty-three specimens) Overall equatorial diameter 52-89 p (mean 73 p); width of cingulum 7-12 p (mean 10 p). Holotype. Preparation KA 243/3, 51-7 102-7, K767, N290. Locus typicus. East Munindalen, Central Dicksonland, Spitsbergen; Planteklofta Conglomerate, probably Upper Givetian. Description. Holotype subcircular, diameter 68 p, cingulum 9 p wide. Laesurae in- distinct, lips approximately 1 p wide, cones 4-6 p wide, 3-6 p high, occasionally fused near the distal pole. Capitate spines eroded from many cones. Remarks. The section (PL 98, fig. 16) clearly demonstrates the laevigate proximal surface, raised lips, one-layered exine, and abruptly tapering cingulum. In specimens where the cones are eroded, the fused bases are clearly seen (PI. 98, fig. 15). Comparison. Archaeozonotriletes crassispinosus Chibrikova 1959 (p. 62, pi. 8, fig. 5) is roundly triangular, lacks lips, and there is no evidence of small spines on the cones. 716 PALAEONTOLOGY, VOLUME 8 Lycospora magnified McGregor 1960 (p. 35, pi. 12, fig. 5, pi. 13, figs. 2-4) is much larger, and has a smaller sculptural pattern. Lycospora rugulatus Vigran 1964 (p. 23, pi. 1, figs. 17, 18; pi. 2, fig. 15) is clearly similar, but lacking cones, has a rugulate rather than a cristate sculpture. The lack of cones in Vigran’s specimens may be due to poor preserva- tion; if this is proved to be so, then Samarisporites hesperus sp. nov. would be synony- mous with L. rugulatus. Occurrence. Upper Mimer Valley Series, particularly common in the Planteklofta Conglomerate; probable Givetian. Samarisporites triangulatus sp. nov. Plate 99, figs. 1-6 Diagnosis. Miospores trilete, amb triangular, with straight to moderately convex sides, central area outline circular to roundly triangular. Laesurae indistinct, length where seen two-thirds to full central area radius, accompanied by smooth, elevated lips, individually 0-5-3 p wide, extending on to the equatorial flange, and frequently to the equatorial margin. Exine two-layered; intexine 1-2 p wide, closely appressed to the exoexine, and often indistinct ; exoexine of both central area and equatorial flange finely infragranulate. Proximal surface laevigate, distally the central area supports a closely spaced ornament of cones, 2-5 p in height and basal diameter, the cones occasionally support a small apical spine. Variation occurs in the basal sculpture of the elements, which may be separate, occasionally fused, or more comprehensively associated, forming an imperfect reticulum; distally the equatorial flange is laevigate, or rarely with a sparse ornament of small cones. Equatorial flange acutely tapering, irregular, reaching its maximum width radially, frequently being only just perceptible in the inter-radial regions. Dimensions. (Forty-five specimens) Overall equatorial diameter 46-78 p (mean 62 j u); diameter of central area 32-54 p (mean 43 p); width of cingulum radially 10-20 p (mean 13 p), interradially 1-9 p (mean 5 p). Holotype. Preparation KA 243/1, 39-6 97-6, K767, N294. Locus typicus. East Munindalen, Central Dicksonland, Spitsbergen; Planteklofta Conglomerate, probable Upper Givetian. EXPLANATION OF PLATE 99 All figures x 500; from unretouched negatives. Figs. 1-9. Samarisporites spp. 1-6. S. triangulatus sp. nov. 1, 2, Holotype, proximal and distal sur- faces respectively; KA 243/1, 39-6 97-6, K767, N294. 3, Proximal surface; KA 243/3, 57-8 108-3, K767, N295. 4, Lateral view; KA 243/3, 35-1 87-6, K767, N296. 5, 6, Sections showing the equa- torial flange; KA 242/S3, 36-5 100-0, K772, N297 and KA 242/S3, 51-4 97-7, K772, N298 respectively. 7-9. S. inusitatus sp. nov. 7, 8, Holotype; proximal and distal surfaces respectively; KA 243/3, 52-2 101-3, K767, N299. 9, Lateral view; KA 243/1, 30-6 98-0, K767, N300. Figs. 10-13. Cirratriradites avius sp. nov. 10, Holotype, proximal surface; KA 242/1, 21-2 89-7, K772, N301. 11, 12, Sections showing the three-layered exine; KA 290/S11, 53-3, 105-0, K681, N302, and KA 290/S12, 32-9 90-5, K681, N303. 13, Proximal oblique aspect; KA 290/2, 45-9 108-4, K681, N304. Figs. 14-15. Camptozonotriletes asaminthus sp. nov. 14, Holotype, proximal oblique aspect; KA 290/4, 53-0 91-2, K681, N307. 15, Proximal surface; KA 290/1, 38-8 103-2, K681, N308. Figs. 1 6-1 7. Cirratriradites dissutus sp. nov. 1 6, Holotype, the equatorial flange is somewhat eroded, as in all specimens; KA 251/3, 33-0 89-2, K905, N305. 17, Specimen showing the flange characteris- tically eroded in the inter-radial areas; KA 262/M5, 36-6 99-6, K908, N306. Palaeontology, Vol. 8 PLATE 99 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 111 Description. Holotype amb triangular, central area circular, overall diameter 54 p, cen- tral area 40 p, width of equatorial flange radially 12 p, inter-radially 1-4 p. Lips indi- vidually 2-5 p wide, extend almost to equatorial margin. Distal central area cones 3- 4 p wide, 2-5 p high, with separate bases, distal equatorial flange laevigate. Sculptural elements obscure the equatorial flange completely in one inter- radial region. Remarks. There is some variation in the proximo-distal thickness of the cingulum, which may be very thin (PI. 99, fig. 4) and represent a zona, or more definitely a cingulum, as demonstrated in the sections (PL 99, figs. 5, 6). Sections also show the intexine and exclusively distal ornament. Comparison. The distinctive equatorial flange separates this species from others assign- able to Samarisporites. Occurrence. Upper Mimer Valley Series; Givetian. Samarisporites inusitatus sp. nov. Plate 99, figs. 7-9 Diagnosis. Miospores trilete ; amb circular to roundly triangular, conformable with the central area outline. Laesurae obscured by smooth, elevated lips, individually 1-2 p wide, 3-8 p high, extending to the equatorial margin. Exine homogeneous to finely infragranulate, proximally laevigate, distally the central area supports an ornament of cones, 1-4 p wide, 1-6 p high, the cones often supporting a small apical spine. Consider- able variation occurs in the basal sculpture of the cones which may be separate, fused, or more comprehensively associated as an imperfect reticulum; distally the equatorial flange is laevigate or rarely with a sparse ornament of small cones. Equatorial flange acutely tapering, uniform. Dimensions. (Eighteen specimens) Overall equatorial diameter 60-66 p (mean 62 p) ; width of equa- torial flange 5-9 p (mean 6-5 p). Holotype. Preparation K243/3, 52-2 101-3, K767, N299. Locus typicus. East Munindalen, Central Dicksonland, Spitsbergen ; Planteklofta Conglomerate, probable Upper Givetian. Description. Holotype circular, overall diameter 62 p, equatorial flange 7 p. Lips each approximately 1 p wide, 3 p high. Distal cones 1-3 p in height and diameter, occasionally fused basally. Remarks. Specimens in oblique aspect (PI. 99, fig. 9) demonstrate the thin equatorial flange. Comparison. Samarisporites hesperus sp. nov. (PI. 98, figs. 12-16) has a larger, denser sculpture, and a strong cingulum. Hymenozonotriletes celeber Chibrikova 1959 (p. 77, pi. 13, fig. 3) is considerably larger, and has a much wider equatorial flange. Occurrence. Fiskeklofta Formation, Planteryggen Sandstone, and Planteklofta Conglomerate; probable Upper Givetian. 718 PALAEONTOLOGY, VOLUME 8 Genus cirratriradites Wilson and Coe 1940 Type species. Cirratriradites saturni (Ibrahim) Schopf, Wilson, and Bentall 1944. Discussion. Cirratriradites avius sp. nov. (PI. 99, figs. 10-13) and Cirratriradites dissutus sp. nov. (PI. 99, figs. 16-17) are included within this genus on the basis of their apparently thin equatorial flange, and reduced sculpture. Cirratriradites avius sp. nov. Plate 99, figs. 10-13 Diagnosis. Miospores trilete; amb roundly triangular, with convex sides and sharply to broadly rounded apices. Laesurae indistinct, accompanied by smooth elevated lips, individually 1-3 p wide, length three-quarters to full spore radius. Exine three-layered; intexine finely infra-granulate, approximately 2 p thick, closely appressed to the exo- exine, and seen only in over-macerated specimens or sections; exoexine two-layered, inner exoexine coarsely infra-granulate, outer exoexine homogeneous. Proximal surface laevigate, distal surface laevigate, or very sparsely ornamented with cones 1-2 p wide, 1-4 p high; equatorial flange moderately to acutely tapering, uniform, differentiation from the central area often indistinct. Dimensions. (Twenty specimens) Overall equatorial diameter 84-172 p (mean 123 p), width of equa- torial flange 16-36 p (mean 27 p). Holotype. Preparation KA 242/1, 21-2 89-7, K772, N301. Locus typicus. East Munindalen, Central Dicksonland, Spitsbergen; Planteklofta Conglomerate, probable Upper Givetian. Description. Holotype diameter 121 p, equatorial flange approximately 22 p wide. Lips narrow, total width 5 p, inaperturate, extend on to the inner margin of the cingulum. Exine laevigate, equatorial flange corroded. Remarks. Sections (PI. 99, figs. 1 1, 12) show the three-layered exine, and the elevated lips formed as an extension of the homogeneous outer exoexine only, which in this specimen are inaperturate. The inner exoexine is very thick distally, which probably accounts for the indistinct intexine and central area outline. As demonstrated in Cirratriradites eiegans by Hughes, Dettmann, and Playford (1962), sections of what appears in proximo-distal aspect to be a thin equatorial flange, is in fact surprisingly thick. Comparison. Cirratriradites eiegans (Waltz) Potonie and Kremp 1956 (p. 126) is clearly similar, but has a more distinct and scabrate central area, and as seen in section (Hughes, Dettmann, and Playford 1962, p. 251, pi. 38, figs. 6, 7), has a thinner intexine, and an undivided exoexine. Occurrence. Upper Mimer Valley Series ; Givetian. Cirratriradites dissutus sp. nov. Plate 99, figs. 16, 17 Diagnosis. Miospores trilete; amb circular, subcircular to oval, conformable with the central area outline. Laesurae straight often open, length equal to full central area K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 719 radius, accompanied by membranous, elevated lips, individually 1-5-6 p wide, up to 8 /x high, extending on to the equatorial flange, and frequently to the equatorial margin. Exine apparently one-layered, coarsely infra-granulate, central area 4-10 p thick, slightly thicker distally, punctate, occasionally microreticulate proximally, laevigate distally; equatorial flange much lighter in colour, thin, uniform, laevigate, often broken radially. Proximal and distal minor folds occasionally present. Dimensions. (Twenty-one specimens) Overall equatorial diameter 71-123 p (mean 90 p), diameter of equatorial flange 4-20 /x (mean 1 1 yu.). Holotype. Preparation KA 251/3, 33-0 89-2, K905, N305. Locus typicus. South Mimerdalen, Central Dicksonland, Spitsbergen; Lower Reuterskioldfjellet Sandstone, Siegenian. Description. Holotype subcircular, diameter 96 p, flange 10 p wide. Laesurae open, lips up to 6 p wide extend to the equatorial margin. Exine of central area 8 p thick, proxi- mally microreticulate. Remarks. The majority of specimens are dark and ill preserved, and are very difficult to macerate satisfactorily. The breakdown of the equatorial flange radially (PL 99, fig. 17) in the majority of specimens, gives the spore a very distinctive appearance. Corrosion of the coarsely infra-granulate exine gives the spore a pseudosculpture of granules. Comparison. Hymenozonotriletes varius Naumova 1953 var. varius (p. 38, pi. 4, fig. 10) is smaller, and has a very different lip construction. Occurrence. Common in the Lower Reuterskioldfjellet Sandstone, but extends up into the Lower Mimer Valley Series; Siegenian and Emsian. Genus camptozonotriletes Staplin 1960 Type species. Camptozonotriletes vermiculatus Staplin 1960. Camptozonotriletes asaminthus sp. nov. Plate 99, figs. 14, 15 Diagnosis. Miospores trilete; proximally flattened, distally convex, amb circular, con- formable in outline with the central area. Laesurae indistinct, straight, length two-thirds to full central area radius, accompanied and frequently masked by smooth, often sinuous, elevated lips, individually 3-5 p wide, 7-12 p high, extending on to the equa- torial flange, and occasionally reaching the equatorial margin. Exine one-layered, coarsely infra-granulate, central area exine 2-10 p thick, slightly thicker distally, equa- torial flange lighter in colour, tapering equatorially ; proximal surface laevigate, distal surface densely sculptured with low verrucae and occasional cones, 2-5 p wide, 2-4 p high, somewhat reduced and occasionally absent from the equatorial flange. Proximal surface frequently with conspicuous radially directed folds. Dimensions. (Twenty-eight specimens) Overall equatorial diameter 80-108 p (mean 89 p) equatorial flange 6-24 p wide (mean 13 p). Holotype. Preparation KA 290/4, 53 0 91-2, K681, N307. 720 PALAEONTOLOGY, VOLUME 8 Locus typicus. West Lagercrantzberget, Central Dicksonland, Spitsbergen; Upper Mimer Valley Series, Upper Givetian. Description. Holotype diameter 96 p, equatorial flange 1 1 p wide. Laesurae indistinct, lips each 3 p wide, 7 p high, extending to equatorial margin. Distal verrucae 3-4 p in height and basal diameter, only slightly reduced on the equatorial flange. Remarks. The majority of specimens are only slightly compressed, and as the proximal surface is flat, and the distal surface deeply convex, they are preserved in oblique aspect. The coarsely infra-granulate structure of the distal verrucae often results in their gradual corrosion and specimens with only a slightly undulose distal surface are frequent. Comparison. Hymenozonotriletes trichomirovii Naumova 1953 (p. 62, pi. 8, fig. 12) is smaller, and has an ornament of spines on the equatorial flange. Cirratriradites ornatus Neves 1960 (p. 269, pi. 33, fig. 3) has a punctate central exine, and an ornament of cones. Occurrence. Upper Mimer Valley Series; Givetian. Camptozonotriletes aliquantus sp. nov. Plate 100, figs. 1, 2 Diagnosis. Miospores trilete ; amb roundly triangular, with moderately to strongly convex sides and well rounded or occasionally acute apices, conformable in outline with the central area. Laesurae distinct to discernible, simple, straight, length two-thirds to full central area radius. Exine two-layered; intexine 1-3 p thick, infra-punctate, often slightly separate from the exoexine and usually distinct; exoexine homogeneous to infra- punctate extending beyond the intexine as a membranous flange. Proximal surface laevigate, distal central area comprehensively sculptured with high, narrow muri 1-3 p wide, which frequently anastomose to form an imperfect reticulum; on the equatorial flange, the muri are somewhat lower, and are radially directed, occasionally extending to the equatorial margin. Major compressional folding frequent, the intexine is sometimes folded independently of the exoexine. All specimens are slightly corroded. Dimensions. (Fifteen specimens) Diameter of exoexine 57-97 p (mean 76 p) ; diameter of intexine 40-74 p (mean 55 p); equatorial flange 9-16 p (mean 12 p). Holotype. Preparation KA 281/2, 37-0 96-4, K850, N309. Locus typicus. Estheriahaugen, Central Dicksonland, Spitsbergen; Emsian. Description. Holotype roundly triangular, diameter of exoexine 80 p, of intexine 50 p, equatorial flange 10-12 p wide. Laesurae indistinct, only one laesura clearly seen. Intexine 3 p thick, unfolded ; distal exoexine with an imperfect reticulum centrally, muri slightly corroded; radially directed muri do not reach the equatorial margin of the flange. One major distal fold. Remarks. The radially directed muri on the equatorial flange are an easily recognizable feature of this species. Camptozonotriletes aliquantus sp. nov. is included in this genus and not in Cirratriradites Wilson and Coe 1940, on the basis of the very prominent distal sculpture. Occurrence. Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Siegenian to Lower Eifelian. K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 721 Infraturna patinati Butterworth and Williams 1958 Genus archaeozonotriletes (Naumova) emend. 1953 Archaeozonotriletes Naumova, p. 30. 1958 Archaeozonotriletes (Naumova) Potonie, p. 28. Emended diagnosis. Miospores trilete; amb circular, subcircular to subtriangular. Laesurae usually long, simple, or accompanied by lips. Exine one- or two-layered, acavate, laevigate, or punctate; distally patinate. The patina may be of uniform thick- ness, or thickest in the distal polar region. Type species. Archaeozonotriletes variabilis Naumova 1953, p. 30, pi. 2, fig. 12 (designated by Potonie 1958, p. 28). Other species. The following species can probably be included within Archaeozono- triletes (Naumova) emend. 1 . Archaeozonotriletes accitus Chibrikova 1959, p. 66, pi. 9, fig. 5. Occurrence : Western Bashkiria, U.S.S.R.: Givetian. 2. Archaeozonotriletes tschernovii Naumova 1953, p. 81, pi. 12, fig. 12. Occurrence: Chkalov province, U.S.S.R. ; Lower Frasnian. Discussion. The type species selected by Potonie 1958, is somewhat atypical amongst the large number of species recorded within Archaeozonotriletes by Naumova (1953). More typical, are thick-walled apiculate forms, which have an intexine partly separated from the exoexine ; many of these latter species, however, can be included within Gemino- spora Balme 1962. Potonie 1958 (p. 28) interpreted the construction of A. variabilis as an irregular cingulum. Evidence from the Spitsbergen specimens, however, shows that A. variabilis has a very thick distal patina, and therefore is frequently preserved in oblique aspect, thus giving the impression of an irregular cingulum. Comparison. The validation of Archaeozonotriletes by Potonie 1958, precedes by one month the erection of Tholisporites Butterworth and Williams 1958. In Tholisporites, however, the patina is thickest in the equatorial region, and is, according to Butterworth and Williams 1958 (p. 382) closely related to Densosporites and Anulatisporites; whereas the uniform or distal polar thickened patina of Archaeozonotriletes shows little construc- tive similarity to Densosporites. Also in Tholisporites, the patina appears to end abruptly on the proximal area, whereas in Archaeozonotriletes, the exine gradually thins over the proximal surface. Archaeozonotriletes variabilis (Naumova) emend. Plate 100, figs. 3-6 1953 Archaeozonotriletes variabilis Naumova 1953, p. 30, pi. 2, figs. 12, 13, pi. 12, figs. 8-11; p. 83, pi. 13, figs. 7-9. Emended diagnosis. Miospores trilete, amb circular to subcircular, conformable with the central area outline. Laesurae straight, length three-quarters to full central area radius. Exine homogeneous, laevigate to finely punctate, proximally 1-5-4 p thick; distally strongly patinate, 5-23 p thick. 722 PALAEONTOLOGY, VOLUME 8 Dimensions. (Nineteen specimens) Equatorial diameter 42-60 p (mean 53 /x); polar diameter 49-74 p (mean 61 p). Locus typicus. Kaluga province, Starooskol beds, U.S.S.R. : Givetian. Description. (From the illustration in Naumova 1953, pi. 2, fig. 12.) Holotype diameter in oblique aspect 50 p. Laesurae simple, straight, length equal to full central area radius exine patinate, proximal oblique aspect 6 p, distal oblique aspect 20 p. Remarks. The thickness of the patina is often as much as 30 per cent, of the total polar diameter (PI. 100, fig. 4). In well-preserved specimens, the punctate nature of the exine is not always obvious, whilst in poorly preserved specimens the punctae are clearly seen (PI. 100, fig. 5). Spores with a thicker distal hemisphere, are common both in Devonian samples from Melville Island, Arctic Canada, McGregor 1960 (p. 38) and from Spits- bergen. It is quite possible that some of the species included within the genus Stenozono- triletes by many Russian authors, may in fact possess a patina. Specimens in oblique aspect are the most informative for distinguishing species of Archaeozonotriletes from thick-walled Punctatisporites, and Stenozonotriletes. Russian authors who use the term ‘otorochka’, appear to use it both for equatorial structures and normal exine thickness. Comparison. Tholisporites tenuis McGregor 1960 (p. 38, pi. 13, fig. 9) has a proximal membranous veil and a thinner patina. T. scoticus Butterworth and Williams 1958 (p. 382, pi. 3, figs. 48-50) is smaller, and the patina has its greatest thickness equatorially. Trematozonotriletes irregularis (Andrejeva, in Luber and Waltz 1941) Ishchenko 1959 (p. 79, pi. 9, fig. 116) appears very similar, but the punctae are confined to the inner margin of the patina. Stenozonotriletes fixus Ishchenko 1952 (p. 56, pi. 16, fig. 141) with its punctate exine, appears to be similar to poorly preserved specimens of A. variabilis (Naumova), but is cingulate rather than patinate. Occurrence. Upper Mimer Valley Series; Givetian. Previous records. Recorded by Naumova (1953) from the Frasnian and Givetian of Kaluga, Chkalov, and Voronezh provinces, U.S.S.R., and by Kedo (1957) from the Famennian of Belorussia, U.S.S.R. EXPLANATION OF PLATE 100 All figures x 500 unless otherwise stated ; from unretouched negatives. Figs. 1-2. Camptozonotriletes aliquantus sp. nov. 1, Holotype, distal surface, KA 281/2, 37 0 96-4, K850, N309. 2, Proximal surface, showing more distinct laesurae; KA 209/1, 34- 1 97-2, K519, N310. Figs. 3-10. Archaeozonotriletes spp. 3-6. A. variabilis (Naumova) emend. 3, Proximal oblique aspect; KA 286/2, 27-5 107-3, K773, N311. 4, Lateral view, showing thick distal patina; KA 286/1, 20-6 99-6, K773, N312. 5, Showing corroded punctate exine; KA 286/3, 20-7 96-0, K773, N313. 6, Proximal oblique aspect; KA 286/1, 52-4 108-3, K773, N314. 7, A. sarus sp. nov. Holotype; KA 261/1, 32-4 103-2, K891, N315. 8-10. A. columnus sp. nov. 8, Holotype, proximal oblique aspect; KA 278/1, 27-6 101-8, K855, N316. 9, Section; KA 286/S4, 40-1 101-7, K773, N317. 10, (X 1000) Showing punctate exine; KA 286/2, 28-8 87-6, K773, N318. Figs. 11-12. Cymbosporites catillus gen. et sp. nov. 11, Holotype; KA 287/2, 34-1 94-2, K846, N329. 12, Distal surface; KA 243/1, 28-5 95-1, K767, N330. Figs. 13-14. Archaeozonotriletes meandricus sp. nov. 13, Holotype, proximal surface; KA 209/2, 35-7 88-6, K519, N319. 14, Proximal surface; KA 209/M4, 24-7 100-5, K519, N320. Palaeontology, Vol. 8 PLATE 100 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 723 Archaeozonotriletes sarus sp. nov. Plate 100, fig. 7 Diagnosis. Miospores trilete; amb circular. Laesurae short, distinct, length spore radius, accompanied at least in part by narrow lips, individually 1 /x or less wide, the laesurae often extend equatorially beyond the lips. Exine infra-punctate, laevigate; proximally 2-3 p thick, distal surface slightly patinate, 3-5 p thick. Dimensions. (Twenty-five specimens) Equatorial diameter 42-57 /x (mean 50 p). Holotype. Preparation KA 261/1, 32-4 103-2, K891, N315. Locus typicus. North Mimerdalen, Central Dicksonland, Spitsbergen ; Fiskelofta Formation, Givetian. Description. Holotype diameter 56 p. Laesurae approximately half spore radius, lips individually 0-5 p wide, length one-third spore radius. Proximal oblique aspect 2-5 p2 distal oblique aspect 4-5 p. Comparison. Trematozonotriletes irregularis (Andrejeva, in Luber and Waltz 1941) Ishchenko 1958 (p. 79, pi. 9, fig. 116) is smaller, has longer laesurae and a thicker patina. Tholisporites tenuis McGregor 1960 (p. 38, pi. 13, fig. 9) has simple laesurae and a thin membranous proximal surface. T. densus McGregor 1960 (p. 37, pi. 13, figs. 6, 7) has longer, simple laesurae, a thicker patina and a proximal membranous veil. A. variabilis (Naumova) emend, has longer, simple laesurae, and a thicker patina. Occurrence. Upper Mimer Valley Series ; Givetian. Archaeozonotriletes columnus sp. nov. Plate 100, figs. 8-10 Diagnosis. Miospores trilete; amb circular, subcircular to oval. Laesurae simple, straight, length three-quarters to full central area radius. Exine thick, homogeneous to infra-granulate, finely to coarsely punctate; proximally 4-10 p thick, distally patinate 8-28 p thick. Dimensions. (Forty specimens) Equatorial diameter 76-145 p (mean 107 p). Holotype. Preparation KA 278/1, 27-6 101-8, K855, N316. Locus typicus. Estheriahaugen, Central Dicksonland, Spitsbergen; PlantskMta Sandstone, probably Upper Givetian. Description. Holotype 116 p, laesurae extend full central area radius. Proximal oblique aspect 9 p, distal oblique aspect 21 p. Remarks. The section (Plate 100, fig. 9) clearly demonstrates the overall thickness of the exine, together with the thickened distal hemisphere. Corrosion of the exine results in an increase in the diameter of the punctae and more drastically in their fusion. Comparison. Tholisporites punctatus McGregor 1960 (p. 38, pi. 13, fig. 10) is somewhat smaller, and has the suggestion of a thin proximal membrane. Archaeozonotriletes variabilis (Naumova, p. 30, pi. 2, fig. 12) emend, is smaller, and in relation of its proximal hemisphere, is more strongly patinate. A. vivax Chibrikova 1959 (p. 69, pi. 10, fig. 4) has according to Chibrikova, a tuberculate ornament, though it would appear from the 724 PALAEONTOLOGY, VOLUME 8 illustration to be foveo-reticulate, rather than punctate. Foveosporites pertusus Vigran 1964 (p. 18, pi. 4, figs. 3, 4; pi. 5, figs. 1, 2 a-d) is clearly similar, but although the thick- ness of the exine varies from 4-17 p, Vigran gives no indication that this species is patinate. Occurrence. Upper Mimer Valley Series; Givetian. Archaeozonotriletes meandricus sp. nov. Plate 100, figs. 13, 14 Diagnosis. Miospores trilete, amb circular. Laesurae straight, length spore radius, accompanied by distinctive, smooth, sinuous, elevated lips, which increase markedly in width equatorially (total width at polar end 4-8 p, at equatorial end 21-26 yu.), length f-f spore radius. Exine probably one-layered, 4-9 p thick equatorially and distally, thinning proximally, coarsely infra-granulate, laevigate (punctate in corroded specimens). Contact areas depressed, radial diameter 26-50 p, bounded by distinct, low, curvatural ridges. Dimensions. (Eighteen specimens) Equatorial diameter 82-173 p (mean 117 p). Holotype. Preparation KA 209/2, 35-7 88-6, K519, N316. Locus typicus. West Odellfjellet, North Dicksonland, Spitsbergen; Lower Mimer Valley Series, Lower Eifelian. Description. Holotype 137 p. Laesurae straight, lips totalling 5 p wide proximally, 24 p wide equatorially, length two-thirds spore radius, extending beyond the contact areas. Radial diameter of contact areas 40-42 p. Comparison. The very distinctive sinuous lips which increase in thickness equatorially, separate A. meandricus sp. nov. from other laevigate, patinate species. Occurrence. Upper Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Emsian and Eifelian. Genus tholisporites Butterworth and Williams 1958 Type species. Tholisporites scoticus Butterworth and Williams 1958. Tholisporites ancylus sp. nov. Plate 101, figs. 1-7 Diagnosis. Miospores trilete ; amb subcircular to roundly triangular, with convex sides and broadly rounded apices, occasionally oval. Laesurae straight, length J-f spore radius, rarely accompanied by lips. Exine two-layered; the inner layer (? intexine) 2-5 p thick, homogeneous, proximally minutely granulose, distally laevigate (seen only when the outer layer is absent); outer layer (? exoexine) laevigate, patinate, equatorially 4-18 p wide, decreasing in thickness towards the distal pole, extends proximally only a short distance, its maximum extension being in the radial regions. The outer patinate layer is corroded in all specimens, and probably because of its structure, corrodes in a constant pattern, giving the spores a ‘sculptured’ appearance. Equatorially the corroded patina appears as large broadly rounded segments, and distally as irregular, partly anastomosing ridges, often radially directed. K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 725 Dimensions. (Twenty-one specimens) Overall equatorial diameter 80-1 16 /u. (mean 94 p), diameter of the inner layer 64-100 p (mean 78 p). Maximum proximal enclosure of the inner layer by the outer layer, radially 18 /x, inter-radially 6 p. Holotype. Preparation KA 240/2, 50-2 100-6, K582, N321. Locus typicus. Manchesterbreen Spur, Central Dicksonland, Spitsbergen; Lower Mimer Valley Series ; Upper Emsian. Description. Holotype roundly triangular, with broadly rounded apices and convex sides, overall diameter 96 p, diameter of inner layer 80 p. Laesurae indistinct, length approximately one-third of spore radius. Inner layer 2 p thick, outer layer equatorially approximately 16 p thick, absent from the distal pole, where the laevigate inner layer can be seen. Proximal overlap of the outer layer 12 p radially, 3 p inter-radially. Remarks. The author regards the ‘sculpturing’ as the effect of corrosion of a finely punctate exine rather than a true sculptural pattern primarily for two reasons. The con- siderable variation in the size of the ‘sculptural’ elements, being much larger and less numerous in the best preserved specimens (PL 101, fig. 5) suggesting further breakdown of the exinal layer with increased corrosion. True sculptural elements in miospores are frequently more homogeneous than the exine which supports them, and with corrosion they tend to become shorter and rounder, rather than breaking up into more numerous elements as in Tholisporites ancylus sp. nov. If these were true sculptural elements, it is unlikely that they would extend completely to the base of the outer exinal layer, as demonstrated in the sections (PI. 101, figs. 3, 4). The corrosion pattern occurs in all specimens, and is a morphological feature easily recognizable, and is here included in the diagnosis. Sections (PI. 101, figs. 3, 4) show the distal and proximo-equatorial extension of the outer layer. The outer layer occasionally breaks away from the inner layer. Occurrence. Dicksonfjorden Sandstone, Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Emsian and Eifelian. Genus cymbosporites gen. nov. Type species. Cymbosporites cyathus sp. nov. Diagnosis. Miospores trilete ; amb circular, subcircular, to roundly triangular. Laesurae long, usually accompanied by lips. Exine thin proximally, equatorially and distally patinate, the patina of even thickness, or with its greatest thickness in the distal polar region. Patina variably sculptured with cones, spines, and granules. Comparison. Archaeozonotriletes (Naumova) emend, has a laevigate or punctate patina. Tholisporites Butterworth and Williams 1958 has its greatest thickness equatorially, and is laevigate or with very small sculptural elements. Derivation of name : L. cymba (Gr. kymbe) — cup, bowl. Cymbosporites cyathus sp. nov. Plate 101, figs. 8-11 Diagnosis. Miospores trilete; amb subcircular to roundly triangular with convex sides and broadly rounded apices, conformable with the central area outline. Laesurae 3 B B 6612 726 PALAEONTOLOGY, VOLUME 8 straight often indistinct, extending full central area radius, accompanied by smooth, elevated lips, individually 0-5-3 p wide, 2-5 p high. Exine proximally 1-3 p thick, equatorially and distally patinate, 8-12 p thick, homogeneous to finely infra-granulate. Proximal surface laevigate ; patina supporting a variable concentration of cones, 2-5 p wide, 1-5 p high, often supporting a small apical spine; where densely packed, the cones have polygonal bases. Arcuate folding of the proximal exine at the central area margin is very common. Dimensions. (Twenty-four specimens) Equatorial diameter 53-80 p (mean 63 p). Holotype. Preparation KA 287/2, 55-2 94-0, K846, N326. Locus typicus. Estheriahaugen, Central Dicksonland, Spitsbergen; Estheriahaugen Formation, Givetian. Description. Holotype subcircular, diameter 61 p. Laesurae indistinct, accompanied by narrow lips, individually 0-5 p wide. Exine very thin proximally, patina equatorially 12 p thick; patina densely covered with cones, 3-5 p wide, 2-5 p high, often with poly- gonal bases. Remarks. The very thin proximal exine often sags into the ‘ cup-shaped ’ patina, resulting in arcuate folding along the central area margin. Comparison. It is difficult to make exact comparisons with Russian species of similar appearance in proximo-distal aspect ; unless mentioned in the description, or illustrated in oblique aspect, these species may appear cingulate. Many of the Russian species may in fact be patinate rather than cingulate. Archaeozonotriletes famennensis Naumova 1953 (p. 117, pi. 17, figs. 31-34) is probably of similar construction, but is smaller, and has lips which extend on to the thicker equatorial exine. A. pustulatus Naumova 1953 (p. 35, pi. 3, fig. 10) is smaller, lacks lips, and has a sparser ornament. A. decorus Naumova 1953 (p. 35, pi. 3, figs. 11, 12) lacks lips, and the equatorial exine is thicker. Lepidozonotriletes subtriquetrus (Luber, in Luber and Waltz 1941) Luber 1955 (pi. 5, EXPLANATION OF PLATE 101 All figures x 500, unless otherwise stated ; from unretouched negatives. Figs. 1-7. Tholisporites ancylus sp. nov. 1, 2, Holotype, proximal and distal surfaces respectively, the proximal view shows the extent of the outer layer; KA 240/2, 50-2 100-6, K582, N321. 3, Section showing extent of the outer layer; KA 274/S3, 24-2 94-8, K872, N322. 4, (x 1000) the same. 5, Corrosion pattern; KA 202/5, 30-6 92-2, K555, N323. 6, Pseudo-ornament of verrucae formed by corrosion of the patina; KA 271/3, 23-8 87-8, K897, N324. 7, Corrosion pattern, much of the outer layer has corroded away; KA 274/6, 52-7 104-2, K872, N325. Figs. 8-11. Cymbosporites cyathus gen. et sp. nov. 8, 9, Holotype, proximal and distal surfaces respectively; KA 287/2, 55-2 94-0, K846, N326. 10, Proximal oblique aspect, showing patina and distal sculpture; KA 293/1, 42-5 94-5, K556, N327. 11, Proximal surface; KA 243/3, 23-4 95-7, K767, N328. Figs. 12-20. Chelinospora concinna gen. et sp. nov. 12, 13, Holotype, proximal and distal surfaces respectively, KA 293/1 , 31-791-2, K556, N33 1.14, Paratype, proximal oblique aspect, showing patina ; KA 286/1, 50-3 105-0, K773, N332. 15, 16, Specimens showing proximally reduced sculpture; KA 243/2, 35-6 106-3, K767, N333, and KA 243/2, 55-0 90-0, K767, N334 respectively. 17, Distal surface; KA 243/1, 54-8 96-5, K767, N335. 18, Lateral view showing very thick corroded patina; KA 293/1, 26-9 90-7, K556, N336. 19, 20, Specimens showing corrosion of the patina; KA 243/2, 24-8 103-5, K767, N337, and KA 293/1, 35-4 89-7, K556, N338 respectively. Palaeontology, Vol. 8 PLATE 101 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 727 figs. 94 and 95) is smaller, and is probably cingulate. Lycospora magnified McGregor 1960 (p. 35, pi. 12, fig. 5, pi. 13, figs. 2-4) is considerably larger, and has some fusion of the basal ornament. Occurrence. Upper Mimer Valley Series; Givetian. Cymbosporites catillus sp. nov. Plate 100, figs. 11-12 Diagnosis. Miospores trilete; amb circular, subcircular to roundly triangular, with moderately convex sides and broadly rounded apices, conformable with the central area outline. Laesurae straight, often indistinct, extend almost to the central area margin, accompanied by smooth, slightly elevated lips, individually 0-5-2 y wide. Exine proxi- mally 1-2 y thick, distally and equatorially patinate, 6-9 y thick, homogeneous. Proximal surface laevigate, patina sculptured with densely packed granules or small verrucae, 1-5 ju. or less in height. Frequent arcuate folding of the thin proximal exine occurs at the central area margin. Dimensions. (Thirty-five specimens) Equatorial diameter 34-50 y (mean 40 y). Holotype. Preparation KA 287/2, 34-1 94-2, K846, N329. Locus typicus. Estheriahaugen ; Central Dicksonland, Spitsbergen; Estheriahaugen Formation, Givetian. Description. Holotype roundly triangular, with convex sides and broadly rounded apices, diameter 44 y. Laesurae indistinct, accompanied by lips, individually 0-5 y wide. Exine very thin proximally, equatorial patina 7 y thick, patina densely covered with granules. Remarks. Specimens in oblique aspect clearly demonstrate the nature of the uniform distal patina. As in Cymbosporites cyathus sp. nov. the thin proximal surface sags into the ‘cup-shaped’ patina, resulting in arcuate folding of the proximal surface at the central area margin, which tends to support the thin proximal wall at this point. Comparison. Cymbosporites cyathus sp. nov. (PI. 101, figs. 8-11) is larger, and has an ornament of cones. Several Russian species included within Archaeozonotriletes and Retusotriletes have a similar proximo-distal appearance, descriptions make it difficult to interpret their construction, which may be patinate, cingulate, or thick-walled apiculate. Below are listed species with somewhat similar proximo-distal appearance, size range, and sculptural elements, which may prove to have a patinate construction similar to C. catillus sp. nov. ; Retusotriletes verrucosus Kedo 1955, p. 22, pi. 1, fig. 17. Archaeozonotriletes basilaris Naumova 1953, p. 81, pi. 13, fig. 16; p. 33, pi. 3, fig. 30; p. 128, pi. 19, fig. 8. Retusotriletes accuratus Chibrikova 1959, p. 51, pi. 5, fig. 6. Archaeozonotriletes subpusillus Chibrikova 1959, p. 61, pi. 8, fig. 2. Archaeozonotriletes truncatus Naumova 1953, p. 34, pi. 3, fig. 7. Occurrence. Upper Mimer Valley Series, particularly common in shale samples; Givetian. 728 PALAEONTOLOGY, VOLUME 8 Genus chelinospora gen. nov. Type species. Chelinospora concinna sp. nov. Diagnosis. Miospores trilete; amb circular to roundly triangular. Laesurae distinct, usually long, simple or accompanied by narrow folds. Exine one or two-layered, acavate, thin proximally, equatorially and distally patinate, the patina may be of even thickness, or with its maximum thickness either equatorially or distally. Patina reticulate or foveo- reticulate, contact areas laevigate or with a reduced sculpture of muri, granules, and cones. Comparison. Differs from other patinate genera in having a reticulate or foveo- reticulate patina. Derivation of name. Gr. Chelinos — netted. Chelinospora concinna sp. nov. Plate 101, figs. 12-20 Diagnosis. Miospores trilete; amb circular to subcircular, conformable with the central area outline. Laesurae distinct to discernible, straight, length three-quarters to full central area radius, simple or accompanied by low, smooth, narrow lips, less than 1 p wide. Exine homogeneous to infra-punctate, equatorially 3-12 p thick (including muri), proximal oblique aspect 2-5 p thick, distal surface variably patinate 5-32 p thick, the patina is usually thickest in the distal polar region, and occasionally extends on to the proximal surface. Contact areas support a sparse ornament of small verrucae or granules, 2 p or less in width and height, and less frequently rugulae or muri, 4-6 p long, 0-5-2 p wide. Patina coarsely reticulate, muri 0-5-2 p wide, 1-4 p high, enclosing large polygonal to irregularly rounded lumina 5-22 p in longest diameter. Dimensions. (Thirty specimens) Equatorial diameter 32-54 p (mean 41 p), polar diameter 36-74 p. Holotype. Preparation KA 293/1, 31-7 91-2, K556, N331. Paratype. Preparation KA 286/1, 50-3 105-0, K773, N332. Locus typicus. North ridge of Kinanderfjellet, Central Dicksonland, Spitsbergen; Upper Mimer Valley Series, Givetian. Description. Holotype subcircular, equatorial diameter 48 p. Laesurae straight, length three-quarters central area radius, accompanied by narrow lips. Exine 6-11 p thick equatorially (including muri), patina thins abruptly on the proximal surface. Contact areas with a sparse ornament of verrucae and granules 2 p or less wide, distal muri 1 p or less wide, 2 p high, lumina 10 to 22 p in longest diameter. Paratype in oblique aspect, proximally 2-3 p thick, distally 12-13 p thick. Laesurae j straight, length two-thirds spore radius. Muri 1 p wide, approximately 1 p high, lumina 10-24 p wide. Reduced muroid ridges proximally. Remarks. There is considerable variation in width of the patina, thickness of muri and size of lumina. Corrosion of the reticulum and patina is common, giving spores an unusual appearance (PI. 101, figs. 19, 20). K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 729 Comparison. Knoxisporiites reticulatusV igran 1964 (p. 22, pi. 1, figs. 10-12; pi. 2, figs. 8, 9) is similar but has a laevigate proximal surface. Verrucosisporites variabilis McGregor 1960 (p. 30, pi. 11, fig. 15) resembles ill-preserved specimens, in which the patina has broken down, and the variable ornament recorded, including large flat-topped elements, may be the result of a corroded patina. However, there is no mention by McGregor of any reticulate sculpture. Archaeozonotriletes strangulatus Naumova 1953 (p. 73, pi. 11, figs. 22, 23) is subtriangular, has shorter laesurae, is proximally laevigate, and there is no mention of variable exine thickness. Occurrence. Mimer Valley Series; Givetian and probable Upper Eifelian. Chelinospora ligurata sp. nov. Plate 102, figs. 1-7 Diagnosis. Miospores trilete ; amb circular to roundly triangular, with convex sides and broadly rounded apices. Laesurae simple, straight, length three-quarters to full central area radius. Exine at least in some specimens two-layered; intexine 1-2 p thick, homo- geneous; exoexine homogeneous to infra-punctate, proximally thin, distally patinate 10-21 p thick (including muri), maximum thickness either equatorially or in the distal polar region; the patina occasionally extends on to the proximal surface, where it may thin gradually or abruptly. Proximal surface laevigate or with a sparse ornament of small verrucae or cones, 2 p or less in height and basal diameter; patina with a high indistinct reticulum, muri 1-3 p wide, 2-7 p high, lumina 4-9 p wide. Dimensions. (Thirty specimens) Equatorial diameter 42-70 p (mean 55 p). Holotype. Preparation KA 243/2, 53-5 91-0, K767, N339. Locus typicus. East Munindalen, Central Dicksonland, Spitsbergen; Planteklofta Conglomerate, probable Upper Givetian. Description. Holotype roundly triangular, diameter 58 p. Laesurae equal full central area radius. Intexine indistinct; exoexine proximal surface sparsely sculptured with small cones and verrucae 1-2 p in height and basal diameter, equatorially patina 10 p wide (excluding muri), muri 2-3 p wide, 5-6 p high; distal reticulate pattern indistinct. Remarks. Corrosion in the majority of specimens results in both the reticulum and the patina breaking down, giving the spore a very distinctive ‘sculptured’ appearance of high, close bacula (PI. 102, figs. 6, 7). Comparison. Chelinospora concinna sp. nov. (PI. 101, figs. 12-20) has a coarser, lower reticulum. Occurrence. Fiskeklofta Formation, Planteryggen Sandstone, and Planteklofta Conglomerate; probable Upper Givetian. Chelinospora perforata sp. nov. Plate 102, figs. 8-10 Diagnosis. Miospores trilete; amb roundly triangular, to subcircular, conformable with the central area outline. Laesurae straight, length three-quarters to full central area radius, accompanied and frequently obscured by smooth, narrow lips, individually 0-5- 730 PALAEONTOLOGY, VOLUME 8 2 y wide. Exine infra-granulate, thin proximally, distally patinate, equatorially 11-17 y thick, gradually thinning towards the distal pole. Contact areas laevigate, patina foveo- reticulate, fovea 2-6 y in longest diameter, circular, oval, or occasionally irregular, 1-5 y apart. Dimensions. (Sixteen specimens) Equatorial diameter 58-86 y (mean 71 y). Holotype. Preparation KA 272/2, 30-0 97-6, C626, N344. Locus typicus. Huginaspiskardet, Central Dicksonland, Spitsbergen; Lower Mimer Valley Series, Lower Eifelian. Description. Holotype roundly triangular, diameter 58 y. Laesurae length two-thirds of spore radius, lips each 1-5 y wide. Equatorial patina 11 y thick. Fovea 2-5 y in longest diameter. Comparison. Perforosporites robustus Scott and Rouse 1961 (p. 978, pi. 113, figs. 1-6, pi. 114, figs. 1-5) is smaller, lacks lips, is frequently sculptured with papillae, and lacks a patina. Occurrence. Upper Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Emsian and Eifelian. Subsuperturma perinotriliti (Erdtman) Dettman 1963 Genus auroraspora Hoffmeister, Staplin, and Malloy Type species. Auroraspora solisortis Hoffmeister, Staplin, and Malloy 1955. Auroraspora macromanifestus (Hacquebard) Richardson 1960 Dimensions. (Six specimens) Diameter of exoexine 192-216 y, diameter of intexine 107-26 y. Occurrence. Fiskeklofta Formation and Planteryggen Sandstone; Upper Givetian. Genus perotrilites (Erdtman) ex Couper 1953 Type species. Perotrilites granulatus Couper 1953. EXPLANATION OF PLATE 102 All figures x 500 unless otherwise stated ; from unretouched negatives. Figs. 1-10. Chelinospora spp. 1-7. C. ligurata sp. nov. 1-3, Holotype, proximal, sectional and distal foci; KA 243/2, 53-5 91-0, K767, N339. 4, Proximal oblique aspect, showing thick corroded patina; KA 243/1, 41-8 105-3, K767, N340. 5, Showing distal reticulum; KA 243/2, 46-8 91-8, K767, N341. 6, 7, Corroded specimens; KA 243/1, 26-0 96-4, K767, N342, and KA 243/3, 35-8 107-5, K767, N343 respectively. 8-10. C. perforata sp. nov. 8, 9, Holotype; proximal and distal surfaces respec- tively, KA 272/2, 30-0 97-6, C626, N344. 10, (x 1000) detail of distal foveo-reticulate exine; KA 271/3, 37-9 109-3, K897, N345. Figs. 1 1-20. Perotrilites spp. 1 1-13. P. eximius sp. nov. 11, Holotype, proximal surface; KA 295/M4, 49-2 96-1, K922, N346. 12, Section showing the three-layered sclerine; KA 258/S2, 22-8 92-1, K922, N347. 13, Section showing outer sculptine only partly separate from inner sculptine; KA 258/S3, 28-5 90-2, K922, N348. 14-15. P. pannosus sp. nov. Holotype, proximal and distal surfaces respectively; KA 243/1, 29-2 96-6, K767, N353. 16-20. P. ergatus sp. nov. 16, 17, Holotype, proximal and distal surfaces respectively; KA 261/4, 53-0 98-5, K891, N349. 18, Showing distal muroid fold pattern; KA 290/5, 54-9 103-6, K681, N350. 19, 20, Sections showing the very thick inner sculptine; 19, (x 1000) KA 290/S8, 30-8 106-4, K681, N351. 20, KA 290/S9, 35 0 99-5, K681, N352. Palaeontology, Vol. 8 PLATE 102 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 731 Discussion. The species listed below are included within Perotrilites (Erdtman) ex Couper 1953, on the basis that they possess a thin homogeneous or finely infra-granulate outer layer, in comparison with the much thicker infra-granulate inner layer. As the exact nature of this outer layer is not known (it may in fact be an outer exoexine layer), the term perine is not used, and the terms inner and outer sculptine (Erdtman 1952, p. 468) are employed for these two layers. Where an intexine is also demonstrated, the term sclerine (Erdtman 1952, p. 468) is used for all three layers. The cavity occurs between the outer and inner sculptine, and not between the inner sculptine and intexine. Sectioned species included in Grandispora (PL 103, fig. 9) and Calyptosporites (PI. 103, fig. 1 1), show an exoexine and intexine of approximately equal thickness. Perotrilites eximius sp. nov. Plate 102, figs. 11-13 Diagnosis. Miospores trilete ; cavate ; amb and inner sculptine outline convexly triangular to circular. Laesurae straight, short, length |— § central area radius, often accompanied and exceeded equatorially by triradiate folds of the outer sculptine, total width 4-12 /x. Exine three-layered; intexine thin, 05-1-5 /x, homogeneous, closely appressed to the inner sculptine, and seen only in sections ; inner sculptine 4-9 p thick, homogeneous to finely infra-granulate, laevigate; outer sculptine 1-2-5 /x thick, homogeneous. Proximal surface laevigate, distal surface supporting a variable ornament of cones and occasional spines, 3-7 p wide, 4-8 /x high, the majority of cones supporting a smaller apical cone or bifurcate-tipped spine. Proximal surface often with numerous minor folds. Dimensions. (Forty-four specimens) Diameter of outer sculptine 80—160 yu. ; diameter of inner sculp- tine 52-105 /x (mean 90 /x). Holotype. Preparation KA 295/M4, 49-2 96-1, K922, N346. Locus typicus. South Mimerdalen, Central Dicksonland, Spitsbergen; Reuterskioldfjellet Sandstone, Emsian. Description. Holotype roundly triangular, outer sculptine diameter 136 yu,; inner sculptine diameter 102 /x. Laesurae approximately half inner sculptine radius, elevated lips 1 1 /x high, length four-fifths spore radius. Inner sculptine 9 /lx thick, outer sculptine 2 p thick. Distal cones 3-6 p wide, 3-8 p high, somewhat smaller on the outer sculptine extension. Remarks. The section (PI. 102, fig. 12) clearly shows the three-layered sclerine, composed of an intexine and an inner and outer sculptine. Although the outer sculptine is not diaphanous, the thick inner sculptine is undoubtedly an exoexinal layer, and the species is included within Perotrilites (Erdtman) ex Couper 1953, on the basis of having a much thinner layer outside this exoexinal layer. Occurrence. Dicksonfjorden Sandstone, Upper Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series ; Emsian and Eifelian. Perotrilites ergatus sp. nov. Plate 102, figs. 16-20 Diagnosis. Miospores trilete ; cavate ; amb and inner sculptine outline convexly triangular to circular. Laesurae straight, length f-f central area radius, usually obscured by 732 PALAEONTOLOGY, VOLUME 8 triradiate folds of the outer sculptine, totalling 1-2 p wide, 2-7 p high, extending to the equatorial margin. Inner sculptine inter-radially 6-10 /a thick, radially 4-7 /a thick, coarsely infra-granulate, laevigate; outer sculptine 1-2 /a thick, homogeneous to finely infra-granulate, extends 4-14 /a beyond the inner sculptine, proximally laevigate, distally supporting a sparse ornament of cones, 1-3 p in height and basal diameter. Distal surface frequently with small muroid folds, 1-3 /x wide, occasionally anastomosing as an imperfect reticulum. Dimensions. (Thirty-two specimens) Diameter of outer sculptine 71-124 /a (mean 91 /a); diameter of inner sculptine 60-108 /a (mean 79 p). Holotype. Preparation KA 261/4, 53-0 98-5, K891, N349. Locus typicus. Estheriahaugen, Central Dicksonland, Spitsbergen: Fiskeklofta Formation, Givetian. Description. Holotype amb and inner sculptine roundly triangular, outer sculptine diameter 84 /a, inner sculptine diameter 69 /a. Laesurae length three-quarters inner sculptine radius. Inner sculptine 4 p thick radially, 8 /a thick inter-radially, outer sculp- tine 1-2 p thick. Distal surface with occasionally anastomosing muroid folds. Remarks. Sections (PI. 102, figs. 19, 20) show clearly the cavate nature of the sculptine, the thick infra-granulate inner sculptine, and the triradiate folds. It is not certain from the strongly compressed sections whether an intexine is present as seen in Perotrilites eximius sp. nov. (PI. 102, fig. 12). Comparison. Perotrilites eximius sp. nov. (PI. 102, figs. 1 1-13) has much larger sculptural elements, and a uniformly thick inner sculptine. Diaphanospora sp. Balme and Hassell 1962 (p. 22, pi. 4, figs. 8, 9) is smaller, and has a laevigate outer sculptine. Archaeozono- triletes arduus Archangelskaya 1963 (p. 23, pi. 8, figs. 1-5) has a thicker distal surface, and the muroid folds are in the contact areas and not proximo-equatorial and distal. Occurrence. Dicksonfjordon Sandstone, Upper Reuterskioldfjellet Sandstone, and Mimer Valley Series ; Emsian to Givetian. Perotrilites pannosus sp. nov. Plate 102, figs. 14, 15 1964 Perotrilites cf. perinatus Vigran, p. 19, pi. 3, jfigs. 7, 8. Diagnosis. Miospores trilete ; cavate ; amb and inner sculptine outline convexly triangular to circular. Laesurae straight, §— f central area radius, usually masked by triradiate folds of the outer sculptine, totalling 2-3 /a wide, 2-5 /a high. Inner sculptine 3-5 p thick, infra-granulate, laevigate; outer sculptine 1-2 p thick, homogeneous, loose-fitting. Contact areas laevigate, proximo-equatorial and distal surfaces with a distinct pattern of muroid folds, 0-5-2 p wide, which may be arcuate, or anastomosing to form an im- perfect reticulum. Distal surface with a sparse ornament of small cones, 1-2-5 p in height and basal diameter, the cones frequently support a small apical spine. Dimensions. (Forty-one specimens) Diameter of outer sculptine 40-74 p (mean 62 p) ; diameter of inner sculptine 37-64 p (mean 54 p). Holotype. Preparation KA 243/1, 29-2 96-6, K767, N353. Locus typicus. East Munindalen, Central Dicksonland, Spitsbergen; Planteklofta Conglomerate, probable Upper Givetian. K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 733 Description. Holotype amb and inner sculptine outline subcircular, outer sculptine diameter 63 p, inner sculptine diameter 56 p. Laesurae straight, length half inner sculp- tine radius, triradiate folds ^-§ outer sculptine radius. Muroid folds on proximo- equatorial and distal surfaces, forming an imperfect reticulum. Distal cones 1-2 p wide, 1-2-5 p high. Comparison. Perotrilites bifur catus Richardson 1962 (p. 174, pi. 25, figs. 4, 5, text-fig. 3) is considerably larger, has a larger ornament of spines and the outer sculptine is not folded in a definite pattern. Perotrilites sp. McGregor 1960 (p. 35, pi. 12, fig. 8) has a granulate outer sculptine. Diaphanospora riciniata Balme and Hassell 1962 (p. 22, pi. 4, figs. 1-4; text-fig. 5) has a laevigate outer sculptine, and a cingulate inner sculptine. Perotrilites ergatus sp. nov. (PI. 102, figs. 16-20) is larger and lacks a uniformly thick inner sculptine. Occurrence. Dicksonfjorden Sandstone, Upper Reuterskioldfjellet Sandstone, and Mimer Valley Series; Emsian to Givetian. Genus grandispora Hofffneister, Staplin, and Malloy 1955 Type species. Grandispora spinosa Hoffmeister, Staplin, and Malloy 1955. Grandispora diamphida sp. nov. Plate 103, figs. 1-6 Diagnosis. Miospores trilete; cavate; amb and intexine outline circular to subcircular. Laesurae straight, often indistinct, length three-quarters to full radius of the intexine, accompanied and usually masked by narrow elevated lips, 0-5-2 p wide, 10-26 p high, extending on to the exoexinal extension, frequently reaching the equatorial margin and occasionally separating equatorially. Intexine 1-4 p thick, finely infra-granulate. Exoexine 2-4 p thick, infra-granulate, proximal surface laevigate, distal surface sculp- tured with cones and spines of variable shape, 3-6 p wide basally, 3—10 yu. high, the cones are often flask-shaped, and may possess a small bifurcate tip. The cones and spines are often more concentrated on the exoexinal extension. Dimensions. (Twenty-three specimens) Diameter of exoexine 67-90 p (mean 75 p); diameter of in- texine 54-66 p (mean 59 p). Holotype. Preparation KA 281/5, 55-3 98-8, K850, N354. Locus typicus. Estheriahaugen, Central Dicksonland, Spitsbergen; Reuterskioldfjellet Sandstone, Emsian. Description. Holotype circular, exoexine diameter 71 p, intexine diameter 55 p. Laesurae indistinct, raised lips 1-5 p wide, 8 p high, separating equatorially and extending to the equatorial margin. Distal spines and cones 3-5 p wide, 3-6 p high; few cones have a small bifurcate tip. Remarks. Grandispora diamphida sp. nov. resembles species of Calyptosporites Richard- son 1962 in having cones and spines with bifurcate tips; but is here assigned to Grandi- spora Hoffmeister, Staplin, and Malloy on the basis of its smaller size and circular outline. 734 PALAEONTOLOGY, VOLUME 8 Comparison. Hymenozonotriletes polyacanthus Naumova 1953 (p. 41, pi. 4, figs. 11, 12) has similar ornament, but appears to possess a bizonate cingulum. Occurrence. Upper Reuterski0ldfjellet Sandstone, and Lower Mimer Valley Series; Emsian and Eifelian. Grandispora inculta sp. nov. Plate 103, figs. 7-9 Diagnosis. Miospores trilete; cavate; amb and intexine outline subcircular to roundly triangular. Laesurae indistinct, length J-f intexine radius, accompanied and frequently masked by smooth, narrow lips, total width 1-4 p, 2-6 p high, extending one-third to full radius of the exoexine. Intexine 1-5-3 p thick, infra-granulate, laevigate, rarely folded; exoexine 1-3 p thick, homogeneous to infra-granulate, often folded. Proximal surface laevigate, distal surface densely covered with cones 1-2 p in height and basal diameter. Minor folds frequent on the exoexine. Dimensions. (212 specimens) Diameter of exoexine 5 1—86 /lx. (mean 70 p); diameter of intexine 34- 63 p (mean 49 p). Holotype. Preparation KA 261/4, 39-6 88-8, K891, N359. Locus typicus. Estheriahaugen, Central Dicksonland, Spitsbergen; Fiskeklofta Formation, Givetian. Description. Holotype amb roundly triangular, exoexine diameter 72 p, intexine outline subcircular, diameter 42 p. Laesurae indistinct, lips individually 1-1-5 p wide, length of two lips one-third of exoexine radius, the third extending almost to the exoexinal margin. Distal cones 1-2 p in height and basal diameter. Remarks. The section (PI. 103, fig. 9) demonstrates the cavate nature of the exine, the approximately equal thickness of the exoexine and intexine, and the exclusively distal ornament. The short laesurae and infra-granulate nature of the intexine is well seen in broken specimens or where the exoexine is removed. Comparison. Zonotriletes explanatus Luber, in Luber and Waltz 1941 (p. 10, pi. 1, fig. 4) has ‘the entire surface of the exine covered with small tubercles ’, but is otherwise similar, although it may not be cavate. Hymenozonotriletes brevimammus Naumova 1953 (p. 39, pi. 4, fig. 3) has a verrucose ornament. Spore number 4 Apiculatozonales Group Thomp- son 1940, figured in Thompson 1952 (p. 10, fig. 14) is larger, and although the exine is clearly two-layered, it appears to have an equatorial flange. Occurrence. Upper Mimer Valley Series; Givetian. EXPLANATION OF PLATE 103 All figures x 500 unless otherwise stated ; from unretouched negatives. Figs. 1-9. Grandispora spp. 1-6. G. diamphida sp. nov. 1 , 2, Holotype, proximal and distal surfaces respectively; KA 281/5, 55-3 98-8, K850, N354. 3, Proximal surface; KA 281/5, 43-0 100-1, K850, N355. 4, Showing distal arcuate folds; KA 240/Ml, 39-6 98-7, K582, N356. 5, 6, Sections showing raised lips and cavate exine; 5, KA 274/S7, 25-6 103-7, K872, N357. 6, (x 1000) KA 274/S7, 43-4 88-9, K872, N358. 7-9. G. inculta sp. nov. 7, Holotype, sectional focus; KA 261/4, 39-6 88-8, K891, N359. 8, Distal surface; KA 261/1, 37-5 98-4, K891, N360. 9, Section showing cavate exine and exclusively distal sculpture; KA 290/S2, 59-3 93-6, K681, N361. Figs. 10-11. Calyptosporites proteus (Naumova) comb. nov. 10, distal surface; KA 290/Ml, 48-4 96-6, K681, N362. 11, Section showing cavate exine and exclusively distal sculpture; KA 290/S21, 34-5 91-6, K681, N363. Palaeontology , Vol. 8 PLATE 103 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 735 Genus calyptosporites Richardson 1962 Type species. Calyptosporites velatus (Eisenack) Richardson 1962. Calyptosporites microspinosus (Richardson) emend. Richardson 1962 1960 Cosmosporites microspinosus Richardson, p. 53, pi. 14, figs. 5, 6. 1962 Calyptosporites microspinosus Richardson, p. 192. Dimensions. (Fifteen specimens) Diameter of exoexine 216-301 p (mean 264 p) ; diameter of intexine 70-1 20 p (mean 100 p). Remarks. Occasionally one or more of the lips are absent, presumably the result of compression or corrosion. Cone coverage is never dense, but in some corroded speci- mens as few as two cones are present. None of the Spitsbergen specimens exhibit the bifurcate tips noted by Richardson (p. 53). Occurrence. Fiskektefta Formation; Givetian. Calyptosporites proteus (Naumova) comb. nov. Plate 103, figs. 10, 11 1953 Hymenozonotriletes proteus Naumova, p. 40, pi. 14, fig. 5. 1955 Hymenozonotriletes proteus Naumova var. eximius Kedo, p. 31, pi. 4, fig. 3. Description of specimens. Miospores trilete ; cavate, amb subtriangular with moderately to strongly convex sides and acute to well-rounded apices, intexine outline roundly triangular to subcircular. Laesurae often indistinct, straight, length half to full intexine radius, usually masked by smooth, elevated lips, individually 2-5-6 p wide, 4-9 p high, length half to full exoexine radius. Intexine distinct, 1-4 p thick, infra-granulate, laevi- gate ; exoexine 2-4 p thick, infra-granulate. Proximal surface laevigate, distal surface sculptured with cones 3-5 p wide, 3-7 p high, the rounded apices frequently supporting a small spine. The ornament is usually sparse, but is occasionally more dense. Distal folding of the exoexine frequent. Dimensions. (Twenty-three specimens) Diameter of exoexine 1 10-170 p (mean 144 p); diameter of intexine 58-100 p (mean 79 p). Remarks. Sections demonstrate the cavate nature of the spore, the absence of a limbus, the approximately equal thickness of the two exine layers, and the infra-granulate intexine. Comparison. Calyptosporites velatus (Eisenack) Richardson 1962, p. 192, in Richardson 1960 (p. 52, pi. 14, fig. 4, text-fig. 3) has very similar dimensions, but has an ornament of much smaller cones, which have acute rather than rounded apices. Hymenozono- triletes ventosus Kedo 1957 (pi. 3, fig. 1) is much smaller. Calyptosporites microspinosus Richardson 1962 (p. 192) in Richardson 1960 (p. 53, pi. 14, figs. 5, 6) is considerably larger (the mean of the Spitsbergen specimens of C. microspinosus being over 100 p greater than the mean of C. proteus). Occurrence. Mimer Valley Series ; Givetian and probably Upper Eifelian. 736 PALAEONTOLOGY, VOLUME 8 Calyptosporites optivus (Chibrikova) comb. nov. Plate 104, figs. 1A 1959 Archaeozonotriletes optivus Chibrikova, p. 60, pi. 7, fig. 9. 1960 Retusotriletes sp. Taugourdeau-Lantz, p. 145, pi. 1, fig. 5. 1964 Biharisporites spitsbergensis Vigran, p. 12, pi. 2, figs. 1-4. Description of specimens. Megaspores trilete; cavate; amb and intexine outline sub- circular to roundly triangular, with convex sides and rounded apices. Laesurae straight, often indistinct, length three-quarters to full intexine radius, accompanied by smooth, elevated lips, individually 6-17 p wide, 7-30 p high, extending on to the exoexinal extension, and occasionally to the equatorial margin. Intexine 2-5 p thick, homo- geneous, laevigate; exoexine 3-5 p thick, infra-granulate. Contact areas laevigate, proximo-equatorial and distal surfaces support an ornament of cones of very variable size and distribution 1—10 yu, wide, 2-8 p high; the ornament may be sparse, or closely packed and with polygonal bases, often supporting a small apical spine or cone. Arcuate folds 7-20 p wide are usually present, which separate the raised contact area from the proximo-equatorial region; the intexine is positioned within this raised central region. Dimensions. (Forty-five specimens) Diameter of exoexine 1 60-384 p (mean 273 p); diameter of in- texine 112-230 p (mean 160 p). Remarks. The section (PI. 104, fig. 4) demonstrates the cavate nature of the spore, but not the proximal arcuate folds. This species is included within Calyptosporites on the basis of its large size, ornament of cones, and approximately equal thickness of the two exine layers. Biharisporites Potonie 1956 includes specimens with a very thin, mem- branous intexine (mesosporium). Comparison. Triletes sp. A. Winslow 1962 (p. 38, pi. 19, fig. 10) is clearly very similar, but there is no mention of an intexine, and the exine is considerably thicker. Bihari- sporites submamillarius McGregor 1960 (p. 33, pi. 11, fig. 16, pi. 12, figs. 1-3) is larger, lacks the prominent arcuate folds and raised contact area, and has a much thinner intexine. Occurrence. Upper Mimer Valley Series ; Givetian. EXPLANATION OF PLATE 104 All figures x 500 unless otherwise stated ; from unretouched negatives. Figs. 1-4. Calyptosporites optivus (Chibrikova) comb. nov. 1, (x200) Proximal surface; KA 269/Ml, 48-8 104 1, K558, N364. 2, (x 500) the same. 3, (x 1000) Showing cones with papillate apices; KA 242/Ml, 32-7 98-2, K772, N365. 4, Section showing the cavate nature of the exine; KA 203/S2, 50-8 105-1, K555, N366. Figs. 5-12. Rhabdosporites spp. 5-8. R. cymatilus sp. nov. 5, 6, Holotype, proximal and distal surfaces respectively; KA 274/4, 21-9 104-2, K872, N372. 7, Distal surface; KA 274/2, 51-6 93-3, K872, N373. 8, Section, showing cavate exine and folded exoexine; KA 251/S2, 57-2 88-4, K905, N374. 9-12. R. scamnus sp. nov. 9, 10, Holotype, proximal and distal surface respectively; KA 290/2, 38-7 100-0, K681, N369. 11, Section, very compressed; the cavate nature of the exine is not clearly seen; KA 290/S5, 29-8 91-6, K681, N370. 12, Distal surface showing numerous folds; KA 290/4 37-8 100-3, K681, N371. Palaeontology, Vol. 8 PLATE 104 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 737 Calyptosporites indolatus sp. nov. Plate 106, figs. 1, 2 Diagnosis. Megaspores trilete ; cavate ; amb irregular, subcircular to roundly triangular, occasionally oval or roundly rectangular. Laesurae straight, length intexine radius, frequently obscured by smooth lips, individually 2-5 y wide, extending half to full spore radius. Intexine approximately 6-12 /x thick, laevigate, outline usually indistinct; exoexine 2-4 y thick, infra-granulate, with a sparse distal ornament of cones and spines, 2-10 y wide, 5-24 y high. Exoexine strongly folded. Dimensions. (Twenty-seven specimens) Diameter of exoexine 156-540 y (mean 290 y) ; diameter of intexine 90-278 y (mean 155 y). Holotype. Preparation KA 235/Ml, 45-7 104- 1, K590, N367. Locus typicus. Gonvillebreen-Horbyebreen Col, Central Dicksonland, Spitsbergen; Lower Mimer Valley Series, Lower Eifelian. Description. Holotype diameter 408 y, exoexinal extension 98 y. Laesurae indistinct, elevated lips totalling 7 y wide. Distal cones and spines 2-10 y wide, 5-24 y high. Exo- exine with irregular prominent folds. Remarks. Sections of this species appear to be zonate and prove to be somewhat con- fusing: evidence from broken and dissected specimens, from excentric intexines, and from independent cross-folding on the exoexinal extension, demonstrate that the spore is clearly cavate. It would appear that specimens are sometimes so compressed, that there has been fusion of the equatorial exoexine; even the spore cavity (PI. 106, fig. 2) is represented only by a faint line. Occurrence. Upper Mimer Valley Series ; Givetian. Genus rhabdosporites Richardson 1960 Type species. Rhabdosporites langii (Eisenack) Richardson 1960. Rhabdosporites scamnus sp. nov. Plate 104, figs. 9-12 Diagnosis. Miospores trilete; cavate; amb and intexine outline roundly triangular to circular. Laesurae straight, length two-thirds to full intexine radius, accompanied by low thickened bps, 1-2 y wide, frequently extending beyond the laesurae on to the exoexinal extension. Intexine distinct, 1 -5—4 yu. thick, homogeneous, laevigate; exoexine 1-5-3 y thick, infra-granulate, sculptured with a dense ornament of minute granules. Distal surface with a distinctive, variable distribution of muroid folds, frequently three in number, radially directed and usually situated inter-radially, but occasionally with a more comprehensive pattern. Short minor folds particularly on the equatorial exoexine are common. Dimensions. (Fifty-eight specimens) Diameter of exoexine 54-1 19 y (mean 82 y); diameter of in- texine 42-84 y (mean 60 y). Holotype. Preparation KA 290/2, 38-7 100 0, K681, N369. 738 PALAEONTOLOGY, VOLUME 8 Locus typicus. West Lagercrantzberget, Central Dicksonland, Spitsbergen; Upper Mimer Valley Series, Givetian. Description. Holotype roundly triangular, exoexine diameter 80 p, intexine diameter 58 p. Exoexine and intexine of approximately equal thickness. Three large distal inter- radial muroid folds present. Remarks. The section (PI. 104, fig. 11) demonstrates the two-layered exine, both layers of approximately equal thickness. Because of strong compression, the cavate nature of the equatorial exine is not clearly seen. Comparison. Camptozonotriletes velatus (Waltz) Playford 1963 (p. 645, pi. 93, figs. 1-3) closely resembles Rhabdosporites scamnus sp. nov. Sections of C. velatus (Waltz) Play- ford, in Dettmann and Playford 1962 (p. 680, pi. 96, figs. 10-12) are clearly similar, and are probably cavate, as they have tentatively suggested. However, R. scamnus sp. nov. lacks the ‘spanner like’ lips, and except in a few specimens, also lacks the more com- prehensive distal folding. Rhabdosporites langi (Eisenack) Richardson 1960 (p. 54, pi. 14, figs. 8-9) is larger, and lacks lips. Rhabdosporites parvulus Richardson 1965, is of similar size, but lacks the major folding so prominent in R. scamnus sp. nov. Naumova (1953) assigned to Archaeozonotriletes and Hymenozonotriletes probable cavate forms, several of which exhibit major distal folding. Archaeozonotriletes notatus Naumova 1953 (p. 84, pi. 13, fig. 12) and Hymenozonotriletes angulatus Naumova 1953 (p. 65, pi. 8, fig. 21) both have three major radially directed distal folds, but are considerably smaller. Naumova has clearly misinterpreted the radially directed folds as ‘the bordered aperture of the perisporium’, and the real trilete mark as the ‘aperture of the spore body’ only. Size alone is not usually a criterion for specific separation, but insufficient details of the exact construction of A. notatus Naumova and H. angulatus Naumova makes closer comparison impossible. Hymenozonotriletes facetus Archangelskaya 1963 (p. 28, pi. 15, figs. 5, 6) has muroid folds proximally and lacks the major distal folds. Occurrence. Upper Mimer Valley Series ; Givetian. Rhabdosporites cymatilus sp. nov. Plate 104, figs. 5-8 Diagnosis. Miospores trilete; cavate; amb and intexine outline roundly triangular, subcircular to oval, undulating. Laesurae often indistinct, straight, |—f spore radius, frequently accompanied and often obscured by sinuous lips, individually 1-3 p wide, 5-10 p high. Intexine indistinct, 1-3 g thick, infra-granulate, unfolded; exoexine EXPLANATION OF PLATE 105 All figures x 500 unless otherwise stated; from unretouched negatives. Figs. 1-6. Aulicosporites spp. 1-3. A. aulicus gen. et sp. nov. 1, Holotype, proximal surface; KA 209/Ml, 63-3 98-8, K519, N375. 2, Section, showing cavate exine and exclusively distal sculpture; KA 240/S 1, 37-6 101-6, K582, N376. 3, (xlOOO) Details of distal sculpture; KA 209/M3, 36-7 91-2, K519, N377. 4-6. A. vitabilis sp. nov. 4, Holotype, proximal surface ( X 250) ; KA 243/Ml, 30-3 94-5, K767, N378. 5, (X 1000) Details of the cristo-reticulate ornament; KA 243/Ml, 41-0 100-2, K767, N379. 6, Section, showing sharp raised lips, and intexine; KA 243/S4, 40-7 100-4, K767, N380. Palaeontology, Vol. 8 PLATE 105 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 739 2-4 i~i thick, coarsely infra-granulate, laevigate, with a regular pattern of slightly sinuous muroid folds, 2-8 p wide, 2-4 p high, more or less radially directed (at least equatorially), and frequently absent from the contact areas. Dimensions. (Twenty-four specimens) Diameter of exoexine 64-172 /x (mean 112/x); diameter of intexine 68-92 p (mean 79 p). Holotype. Preparation KA 274/4, 21-9 104-2, K872, N372. Locus typicus. Reuterskioldfjellet, Central Dicksonland, Spitsbergen; Reuterskioldfjellet Sandstone, Emsian. Description. Holotype subcircular, exoexine diameter 64 p, intexine diameter 58 p. Laesurae simple, straight, length approximately half intexine radius. Exoexine and in- texine of approximately equal thickness. Muroid folds confined to the proximo- equatorial and distal surfaces, being very sinuous in the distal polar region, and being straighter and radially directed equatorially. Remarks. The majority of the specimens are dark; and it is often difficult to see details of the intexine. Further maceration fails to clear the spores; however, even in dark specimens the regularly undulating exoexinal margin, the result of the radially directed folds, is an easily recognizable feature. Sections are difficult to cut from spores in this low horizon, but Plate 104, fig. 8 shows clearly the intexine and the folded exoexine. Comparison. Rhabdosporites langii (Eisenack) Richardson 1960 (p. 54, pi. 14, figs. 8, 9, text-figs. 4, 6b) lacks the intricate and regular fold pattern, a feature present in all specimens of Rhabdosporites cymatilus sp. nov. ; a distinctive diagnostic feature justifying its specific separation. Occurrence. Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Siegenian, Emsian and probably Lower Eifelian. Genus aulicosporites gen. nov. Type species. Aulicosporites aulicus sp. nov. Diagnosis. Megaspores trilete ; cavate ; amb circular to roundly triangular. Intexine and exoexine of approximately equal thickness, intexine laevigate, exoexine sculptured at least distally with a cristate or cristo-reticulate ornament, with cones or occasional spines. Discussion. Differs from both Calyptosporites Richardson 1962 and Biharisporites Potonie 1956, in having a cristate or cristo-reticulate ornament. Derivation of name. Gr. aulakos — with small furrows. Aulicosporites aulicus sp. nov. Plate 105, figs. 1-3 Diagnosis. Megaspores trilete ; cavate ; amb and intexine outline circular, subcircular to roundly triangular. Laesurae distinct, straight, length one-third to full radius of the intexine, accompanied by conspicuous, smooth, elevated lips, 5-12 p wide, 6-10 p high, length three-quarters to full spore radius. Intexine 6-9 p thick, infra-granulate ; exoexine 740 PALAEONTOLOGY, VOLUME 8 7-12 /a thick, coarsely infra-granulate, proximally laevigate, distally sculptured with cones, variable both in size and distribution. The cones (6-40 p wide, 6-24 p high) have rounded apices, occasionally supporting a very small cone or spine ; usually more densely packed in the distal polar region, where they are fused basally or have a polygonal out- line. Distally the exoexinal extension is often laevigate. Dimensions. (Thirty-eight specimens) Diameter of exoexine 1 80-306 ju. (mean 217 /x) ; diameter of intexine 90-190 p (mean 146 /x). Holotype. Preparation KA 209/Ml, 63-3 98-8, K519, N375. Locus typicus. Odellfjellet, North Dicksonland, Spitsbergen; Lower Mimer Valley Series, Lower Eifelian. Description. Holotype diameter of exoexine 208 p, diameter of intexine 120 /x. Laesurae approximately half intexine radius. Lips each 8-10 p wide, extend almost to the equa- torial margin. Distal cones 6-12 p wide, 6-10 p high, with both separate polygonal and fused bases, and occasionally with small apical spines. Small folds present on exoexinal extension. Remarks. The section (PI. 105, fig. 2) demonstrates the cavate nature of the spore, the thick intexine, and the exclusively distal sculpture of cones. Occurrence. Dicksonfjorden Sandstone, Upper Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Emsian and Eifelian. Aulicosporites vitabilis sp. nov. Plate 105, figs. 4-6 Diagnosis. Megaspores trilete, amb circular to roundly triangular with convex sides and broadly rounded apices. Laesurae straight, length f-f spore radius, accompanied and often obscured by acute, thickened lips, individually 5-25 p wide, 25-44 p high at the proximal pole, 16-26 p high equatorially. Intexine seen only in sections, approximately 5 p thick, homogeneous, laevigate; exoexine 8-10 p thick (excluding ornament), infra- punctate. Contact areas occupying most of the proximal surface, laevigate or punctate, except adjacent to the lips, where they may support an ornament of cones ; contact areas bounded by curvaturae, up to 20 p wide, formed by a fusion of cones. Distal surface cristo-reticulate, with cones and occasional spines 6-12 p wide, 7-20 p high. Dimensions. (Seventeen specimens) Diameter of exoexine 265-364 p (mean 322 p); diameter of intexine approximately 70 p. Holotype. Preparation KA 243/Ml, 30-3 94-5, K767, N378. EXPLANATION OF PLATE 106 All figures x 500 unless otherwise stated ; from unretouched negatives. Figs. 1-2. Calyptosporites indolatus sp. nov. 1, Holotype (x250), proximal surface; KA 235/Ml, 45-7 104-1, K590, N367. 2, Compressed section; KA 235/SI, 49-7 105-0, K590, N368. Figs. 3-4. Retialetes sp. 3, KA 251/1, 18-9 94-1, K905, N381. 4, (x 1000) the same. Figs. 5-7. Ancyrospora langii (Taugourdeau-Lantz) comb. nov. 5, Distal surface; KA 243/3, 47-7 92-1, K767, N386. 6, 7, Sections, showing the raised lips, equatorial flange, and intexine; KA 243/S8, 52-3 104-6, K787, N387 and KA 243/S8, 33-9 103-4, K767, N388, respectively. Palaeontology, Vol. 8 PLATE 106 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 741 Locus typicus. East Munindalen, Central Dicksonland, Spitsbergen; Planteklofta Conglomerate, probable Upper Givetian. Description. Holotype subcircular, exoexinal diameter 296 p, intexine not seen. Laesurae straight, length four-fifths spore radius, accompanied by lips each up to 24 p wide. Con- tact areas laevigate, except for an ornament of cones adjacent to the lips. Distal cristo- reticulum well developed. Remarks. The section (PI. 105, fig. 6) clearly shows the intexine, the raised sharp lips formed by an upward extension and thickening of the exoexine, the infra-punctate (here corroded) nature of the exoexine, and the homogeneous sculptural elements. Comparison. Biharisporites ellesmerensis Chaloner 1959 (p. 322, pi. 55, fig. 2, text-fig. 1) is smaller, and has smaller sculptural elements, which are not fused basally. No intexine appears to be present, but this feature is very indistinct in Aulicosporites vitabilis sp. nov. Occurrence. Planteryggen Sandstone, and Planteklofta Conglomerate; probable Upper Givetian. Turma aletes Ibraham 1933 Subturma azonaletes (Luber) Potonie and Kremp 1954 Infraturma reticulonapita (Erdtman) Yimal 1952 Genus retiaietes Staplin 1960 Type species. R. radforthii Staplin 1960. Retiaietes sp. Plate 106, figs. 3, 4 Description of specimens. Spores alete, ellipsoidal. Exine (excluding ornament) thin, 1-5 p or less, finely sculptured with low, narrow muri, 1 p or less wide, 0-5-2 p high. Lumina uniform, usually polygonal, small, typically 2-4 p wide (range 1-8 p). Exine strongly plicated with major folds. Dimensions. (Seven specimens) 40-64 p (mean 51 p) by 31-44 p (mean 38 p). Comparison. Retiaietes radforthii Staplin 1960 the only other species recorded for this genus, is con- siderably larger. Occurrence. Lower Reuterskioldfjellet Sandstone ; Siegenian. Previous records. The genus has previously been recorded only from the Lower Carboniferous. INCERTAE SEDIS Genus nikitinsporites Chaloner 1959 Type species. Nikitinsporites canadensis Chaloner 1959. Nikitinsporites spitsbergensis sp. nov. Plate 108, figs. 1-5 Diagnosis. Megaspores trilete, amb subtriangular to subcircular. Laesurae obscured by greatly elevated lips (sometimes broken), individually 7-15 p wide, 74-160 p high. Exine 3c B 6612 742 PALAEONTOLOGY, VOLUME 8 two layered; intexine 2-4 /x thick, homogeneous, closely appressed to the exoexine, and seen only in sections; exoexine 18-45 p thick, coarsely infra-granulate. Contact areas laevigate, proximo-equatorial and distal surfaces supporting thick, more or less parallel- sided spines, structurally composed of rod-shaped elements parallel to the long axis; apically the spines narrow abruptly, ending with an homogeneous grapnel-tip, this tip is never wider than the main shaft of the spine; typically the spines are 100-50 /x long (range 48-250 p), typically 40-60 /x wide (range 14-80 /x wide) the equatorial margin supports 8-20 spines. A pseudoflange up to 80 /x wide, formed by the fused bases of the spines, is occasionally present. Dimensions. (Twenty-four specimens) Equatorial diameter (excluding spines) 240-440 /x (mean 342 /x) ; polar diameter (including apical prominance) 240-416 /x (mean 303 p). Holotype. Preparation KA 203/M4, 30-4 107 1, K555, N382. Locus typicus. North ridge of Kinanderfjellet, Central Dicksonland, Spitsbergen ; Upper Mimer Valley Series, Givetian. Description. Holotype equatorial and polar diameter (excluding elevated lips) both 256 /x, elevated lips 160 /x. Spines 28-40 /x wide, 80-240 p long, only a few with the grapnel-tips preserved, basal fusion of spines (pseudoflange) up to 20 p wide. Remarks. Sections (PI. 108, figs. 4, 5) show the very thick exoexine, and very thin closely appressed intexine ; the spore cavity appears somewhat irregular in shape. This species is included within Nikitinsporites Chaloner 1959, primarily on the shape and structure of the grapnel-tipped spines. Large size and greatly elevated lips, are features of Nikitin- sporites which are also seen in some species included within Hystricosporites McGregor 1960 and Ancyrospora Richardson 1960 emend Richardson 1962. For example Hystrico- sporites porrectus (Balme and Hassell) comb. nov. (PI. 95, figs. 1-3) has elevated lips, but the spines are homogeneous throughout, and with their large grapnel-tips, are clearly different from the spines of Nikitinsporites which are homogeneous only at the very apex, with a small grapnel tip never exceeding the diameter of the main shaft of the spine. A pseudoflange (also present in some species of Ancyrospora) may be absent, partially developed or strongly developed (PI. 108, fig. 2) in otherwise identical speci- mens, and Nikitinsporites Chaloner 1959 is therefore included within Incertae sedis. Comparison. Nikitinsporites sp. Yigran 1964 (p. 20, pi. 2, figs 11—13) has a variable sculpture of conical-based spines, coni and verrucae. Nikitinsporites canadensis Chaloner 1959 (p. 328, pi. 55, fig. 5, text-fig. 4) is circular, and is considerably larger. Dicrospora sp. Winslow 1962 (p. 55, pi. 10, figs. 1 and la) illustrated but not described, is probably assignable to N. spitsbergensis sp. nov. Occurrence. Fiskeklofta Formation and Planteryggen Sandstone; Givetian. EXPLANATION OF PLATE 107 All figures x 500 unless otherwise stated ; from unretouched negatives. Figs. 1-6. Ancyrospora spp. 1, A. trocha sp. nov., Holotype, proximal surface; KA 209/M6, 39-2 103-3, K519, N389. 2-5. A. reuta sp. nov. 2, Holotype, lateral view; KA 209/4, 34-0 88-1, K519, N390. 3, 4, Sections, showing equatorial flange and thick exine; KA 204/SI, 21-0 90-4, K838, N391 and KA 204/SI, 62-0 105-5, K838, N392. 5, ( x 1000) Details of grapnel-tipped spines; KA 204/M2, 32-3 106-6, K838, N393. 6. A. sp. ( X 250) Proximal surface; KA 242/Ml, 33-2 95-1, K772, N394. **3 Palaeontology, Vol. 8 PLATE 107 ALLEN, Lower and Middle Devonian miospores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 743 Genus ancyrospora Richardson 1960 emend Richardson 1962 Type species. Ancyrospora grandispinosa Richardson 1960 emend Richardson 1962. Ancyrospora langii (Taugourdeau-Lantz) comb. nov. Plate 106, figs. 5-7 1960 Archaeotriletes langii Taugourdeau-Lantz, p. 145, pi. 3, figs. 33, 34, 39. 1964 Ancyrospora cf. simplex Vigran, p. 26, pi. 6, figs. 1-3. Dimensions. (Twenty specimens) Overall equatorial diameter 66-140 g (mean 86 g); central area diameter 40-81 /x (mean 55 g); equatorial flange (excluding ornament) 12-48 /x wide. Grapnel-tipped spines typically 12-25 g long (range 8-38 /x), typically 7-12 /x wide basally (range 5-20 /x). Remarks. The spines narrow gradually from the base to the apex, ultimately widening into a grapnel-tip. The spines except for the homogeneous grapnel-tip, are infra- granulate. Sections (PI. 106, figs. 6, 7) demonstrate the elevated membranous lips, thin intexine, and equatorial flange composed only of exoexine. In Archaeotriletes (Naumova) Potonie 1958 the spines are confined to the central area, and are absent from the flange. Comparison. Hymenozonotriletes incisus Naumova 1953 (p. 68, pi. 9, fig. 11) is similar, and should this species prove to have grapnel-tipped spines, then it would be conspecific. Occurrence. Upper Mimer Valley Series; Givetian. Ancyrospora trocha sp. nov. Plate 107, fig. 1 Diagnosis. Miospores trilete; amb and central area outline circular to subcircular. Laesurae straight, length three-quarters to full central area radius, accompanied and often obscured by sinuous, membranous, elevated lips, individually 2-4 g thick, usually closely appressed to the exoexine, but occasionally separated and folded; exoexine coarsely infra-granulate 7-20 g thick centrally, extending as an equatorial flange, typically 20-32 g wide (range 15-50 g). Contact areas occupy most of the proximal surface inside the flange, and support an ornament of low, broad, flat-topped radial muri, 6-13 g wide, 4-12 on each contact area; proximo-equatorial and distal surfaces sparsely sculptured with short spines, typically 15-30 g long (range 10-58 g), the majority with large bulbous bases, 10-24 g wide. Dimensions. (Twenty-five specimens) Overall equatorial diameter (excluding spines) 154-272 g (mean 189 g). Holotype. Preparation KA 209/M6, 39-2 103-3, K519, N389. Locus typicus. Odellfjellet, North Dicksonland, Spitsbergen; Lower Mimer Valley Series, Lower Eifelian. Description. Holotype circular, diameter 178 g. Laesurae straight, length equals full intexine radius, elevated Ups 16 g high. Intexine 3 g thick, folded; exoexine 16 g thick, equatorial flange 22 g wide. Radial muri 8-13 g wide, 6-8 on each contact area ; proximo- 744 PALAEONTOLOGY, VOLUME 8 equatorial and distal spines sparse, short, 14-18 y high, frequently with the grapnel tip broken. Comparison. Ancyrospora grandispinosa Richardson 1960 emend Richardson 1962 (p. 175, pi. 27, figs. 3-5, text-fig. 4) is clearly similar, but lacks the proximal radial muri. Hystricosporites porcatus comb. nov. has longer spines, a curvatural ridge formed in part from the bases of the spines (zonarial ridge of Winslow 1962, p. 52), and lacks the mem- branous equatorial flange. Occurrence. Upper Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Emsian and Eifelian. Ancyrospora reuta sp. nov. Plate 107, figs. 2-5 Diagnosis. Miospores trilete; amb and central area outline circular to subcircular. Laesurae indistinct, length half to full radius of the central area, occasionally accom- panied by smooth lips, individually 1-3 y wide. Exine two-layered; intexine infra- granulate 1-4 y thick; exoexine infra-granulate 7-15 y thick, extending as an equatorial flange, typically 25-40 y wide (range 14-54 y); proximal surface laevigate, distal surface densely covered with short grapnel-tipped spines, 6-20 y long, frequently with large bul- bous bases, 6-10 y wide. Dimensions. (Sixteen specimens) Overall equatorial diameter 1 11-1 80 y (mean 151 y); central area diameter 83-128 y (mean 96 y). Holotype. Preparation KA 209/4, 34 0 88-1, K5 19, N390. Locus typicus. Odellfjellet, North Dicksonland, Spitsbergen; Lower Mimer Valley Series, Lower Eifelian. Description. Holotype in oblique aspect, subcircular, overall equatorial diameter 128 y, central area diameter 100 y. Laesurae indistinct. Exine 10 y thick centrally, equatorial flange 14-22 y wide. Distal spines 6-10 y long, only a few grapnel-tips being preserved. Remarks. Specimens are very dark, and difficult to macerate, tending to break down with increased maceration, rather than becoming lighter in colour. Frequently, the grapnel-tip is not preserved. The sections (PI. 107, figs. 3, 4) demonstrate the equatorial flange and thick exine. No intexine is seen in this specimen. Comparison. Ancyrospora grandispinosa Richardson 1960 emend Richardson 1962 (p. 175, pi. 27, figs. 3-5, text-fig. 4) is considerably larger, and has much longer spines. Ancyrospora trocha sp. nov. (PI. 107, fig. 1) has proximal radial muri. Occurrence. Dicksonfjorden Sandstone, Upper Reuterskioldfjellet Sandstone, and Lower Mimer Valley Series; Emsian to Eifelian. EXPLANATION OF PLATE 108 All figures x 200 unless otherwise stated; from unretouched negatives. Figs. 1-5. Nikitinsporites spitsbergensis sp. nov. 1, Holotype, lateral view; KA 203/M4, 30-4 107T, K555, N382. 2, Proximal surface showing pseudoflange; KA 282/M2, 33-9 96-4, K854, N383. 3, ( x 1000) bifurcate tipped spine from the holotype. 4, 5, Sections showing the thick exoexine and thin intexine. 4, (x500) KA 235/S17, 43-2 99-4, K590, N384. 5, (xlOOO), KA 235/S11, 20-7 106-0, K590, N385. Palaeontology, Vol. 8 PLATE 108 ALLEN, Middle Devonian megaspores K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 745 Ancyrospora sp. Plate 107, fig. 6 Description of specimens. Megaspores trilete; amb triangular to roundly triangular, deeply incised, central area outline triangular. Laesurae indistinct, straight, length two- thirds to full spore radius, accompanied by smooth elevated lips, individually 4-12 p wide, 10—40 yu. high. Exine two-layered; intexine approximately 3 p thick, closely appressed to the exoexine and usually indistinct; exoexine infra-granulate, centrally 15- 20 p thick, thinning and extending equatorially as a pseudoflange up to 60 p wide. Proximal surface laevigate; equatorially and distally sculptured with flexuous spines, 40-100 p long, basally 10-40 p wide, each narrowing gradually towards the apex, where it widens slightly into a grapnel-tip. The spines, except for the homogeneous grapnel-tip, are infra-granulate. Dimensions. (Six specimens) Overall equatorial diameter (excluding spines) 222-80 p ; central area diameter 120-44 p. Remarks. Structurally the spines are similar to those in Nikinisporites Chaloner 1959, but differ in sculpture. Comparison. Archaeotriletes villosus Chibrikova 1959 (p. 44, pi. 2, fig. 1) is clearly similar, and may prove to be identical, but the sculpture is reduced in size towards the distal pole, and there is no evidence of grapnel-tips to the spines, although this may be due to preservation. Occurrence. Fiskeklofta Formation, Planteryggen Sandstone, and Planteklofta Conglomerate; probable Middle and Upper Givetian. SOME GENERAL COMMENTS The Spitsbergen succession is one of the few Devonian continental successions which extends from the Lower Devonian into the Middle Devonian without a stratigraphical break. Botanical affinities can tentatively be suggested on the basis of spores of similar con- struction found in situ in Devonian and Carboniferous plants, but many species, includ- ing those with proximal radial muri, grapnel-tipped spines, and patinate thickenings, which are restricted to, or abundant in, the Devonian, are as yet unassigned to any plant group. Dispersed spores of diverse construction, included within the Infraturmae Laevigati, Apiculati, Murornati, Tricrassati, Cingulati, Patinati, and many cavate forms are described from the Lower Devonian. The majority of Lower Devonian plants described by Hoeg (1942) from Spitsbergen are assignable to the Psilophytales, and from the evidence of spores in situ described from the Devonian, the majority of Psilophytales produced simple laevigate or apiculate forms. This diversity of dispersed spores recovered from the Lower Devonian, suggests that there must have been a greater diversity of parent plants than have as yet been described from, or preserved in, the Lower Devonian of Spitsbergen or elsewhere. 746 PALAEONTOLOGY, VOLUME 8 REFERENCES Ames, h. t. (ed.). 1961. Catalog of Fossil spores and pollen: vol. 1, Translations. The Pennsylvania State University. archangelskaya, a. d. 1963. New spore finds from Devonian deposits of the Russian platform. Ministerstvo geol. i ochroni Nedr SSSR. Moscow, 37, 18-30, 16 pis. (in Russian). balme, b. e. 1962. Upper Devonian (Frasnian) spores from the Carnarvon Basin, Western Australia. Palaeobotanist, 9, 1-10. and hassell, c. w. 1962. Upper Devonian spores from the Canning Basin. Western Australia. Micropaleontology, 8, 1-28. bhardwaj, d. c. 1957a. The palynological investigations of the Saar coals (Part I — Morphography of Sporae Dispersae). Palaeontographica, B101, 73-125. 19576. The spore flora of Velener Schichten (Lower Westpahlain D) in the Ruhr Coal Measures. Ibid. B102, 110-38. butterworth, m. a. and williams, r. w. 1958. The small spore floras of coals in the Limestone Coal Group and Upper Limestone Group of the Lower Carboniferous of Scotland. Trans, roy. Soc. Edinb. 63, 353-92. chaloner, w. g. 1959. Devonian megaspores from Arctic Canada, Palaeontology, 1, 321-32. 1963. Early Devonian spores from a borehole in Southern England. Grana palynologica, 4, 1. 100-10. chibrikova, e. v. 1959. Spores of the Devonian and older deposits of Bashkir. Akad. Nauk. S.S.S.R., Materials on Palaeont. and Strat. of Devonian and older deposits of Bashkir, 3, 116 (in Russian). couper, r. a. 1958. British Mesozoic microspores and pollen grains, a systematic and stratigraphic study. Palaeontographica, B103, 75-179. dettman, m. e. 1961. Lower Mesozoic megaspores from Tasmania and South Australia. Micro- paleontology, 7, 71-86. 1963. Upper Mesozoic microfloras from South-Eastern Australia. Roy. Soc. Viet, 'll, 1-148. and playford, g. 1963. Sections of some spores from the Lower Carboniferous of Spitsbergen. Palaeontology, 5 , 679-81. dybova, s. and jachowicz, a. 1957. Microspores of the Upper Silesian Coal Measures. Inst. Geol. Prace, 23, 1-328. eisenack, a. 1944. Uber einige pflanzliche funde in Geschieben, nebst Bemerkungen zum Hystricho- sphaerideen-Problem. Z. Geschiebeforsch, 19, 103-24. erdtman, G. 1952. Pollen morphology and plant taxonomy — Angiosperms. Waltham, Mass. ■ 1957. Pollen and spore morphology I plant taxonomy; Gymnospermae, Pteridophyta, Bryophyta, Almqvist and Wiksells, Stockholm. faegri, K. and iversen, J. 1950. Text-book of modern pollen Analysis. Munksgaard, Copenhagen. friend, p. f. 1961. The Devonian stratigraphy of North and Central Vestspitsbergen. Proc. Yorks. Geol. Soc. 33, 77-118. hacquebard, p. a. 1957. Plant spores in coal from the Horton group (Mississippian) of Nova Scotia. Micropaleontology, 3, 301-24. and barss, m. s. 1957. A Carboniferous spore assemblage, in coal from the South Nahanni River area, Northwest Territories. Bull. geol. Surv. Canada, 40, 1-63. Harris, w. F. 1955. A manual of the spores of New Zealand Pteridophyta. Bull. New Zealand Dept. Sci. Indust. Res. 116, 186 pp. hoeg, o. A. 1942. The Downtonian and Devonian flora of Spitsbergen. Skr. Norges Svalb. og Ishavs — Unders. 83, 77-146. hoffmeister, w. s. 1959. Lower Silurian plant spores from Libya. Micropaleontology, 5, 331-334. > staplin, f. l., and malloy, r. e. 1955a. Geologic range of Paleozoic plant spores in North America. Ibid. 1, 9-27. 19556. Mississippian plant spores from the Hardinsburg Formation of Illinois and Kentucky. J. Paleont. 29, 372-99. hughes, n. f., dettmann, m. e., and playford, g. 1962. Sections of some Carboniferous spores. Palaeontology, 5, 247-52. K. C. ALLEN: LOWER AND MIDDLE DEVONIAN SPORES 747 hughes, n. f. and playford, g. 1961. Palynological reconnaissance of the Lower Carboniferous of Spitsbergen. Micropaleontology, 7, 27-44. ishchenko, a. m. 1952. Atlas of the microspores and pollen of the Middle Carboniferous of the western part of the Donetz Basin. Izd. Akad. Nauk. Ukrainian S.S.R., Inst. Geol. Nauk. 1-83 (in Russian). 1956. Spores and pollen of the Lower Carboniferous deposits of the western extension of the Donetz Basin and their stratigraphical importance. Akad. Nauk Ukrainian S.S.R., Tr. Inst. Geol. Nauk. Strat. Palaeont. 11, 1-185 (in Russian). 1958. Spore-pollen analysis of the Lower Carboniferous sediments of the Dnieper-Donetz Basin. Ibid. 17, 1-188 (in Russian). kedo, g. i. 1955. Spores of the Middle Devonian of the north-east of Belorussia S.S.R. Palaeont. i stratig. B. S.S.R. Minsk, Akad. Nauk B. S.S.R. 1, 5-47 (in Russian). 1957. Spores from salt deposits of the Pripiat depression and their stratigraphic significance. Ibid. 2, 3-43 (in Russian). kosanke, r. m. 1950. Pennsylvanian spores of Illinois and their use in correlation. Bull. III. State Geol. Surv. 74, 1-128. kremp, G. o. w. (ed.) 1960. Catalog of Fossil spores and pollen: vols. 11, 12, Devonian spores. The Pennsylvania State University. lanjouw, J. 1961. International Code of Botanical Nomenclature. Utrecht. luber, a. A. 1955. Atlas of the spores and pollen grains of the Palaeozoic deposits of Kazachstan. Tr. Akad. Nauk Kazach. S.S.R., Alma-Ata, 1-126 (in Russian). and waltz, i. e. 1938. Classification and stratigraphical value of the spores of some Carboni- ferous coal deposits in the U.S.S.R. Trans. Central Geol. Prosp. Inst. 105, 1-45 (in Russian). 1941. Atlas of microspores and pollen grains of the Palaeozoic of the U.S.S.R. Tr. All- Union Geol. Sci. Res. Inst. ( V.S.E.G.E.I. ), 139, 1-107 (in Russian). mcgregor, d. c. 1960. Devonian spores from Melville Island, Canadian Arctic Archipelago. Palaeontology, 3, 26-44. 1961. Spores with proximal radial pattern from the Devonian of Canada. Bull. geol. Surv. Canada, 76, 1-12. naumova, s. N. 1939. Spores and pollen of coals of the U.S.S.R. Rep. Int. Geol. Congr., 17th Session U.S.S.R. 1, 353-64. 1953. Spore-pollen complexes of the Upper Devonian of the Russian Platform and their strati- graphical value. Tr. Inst. Geol. Nauk Akad. S. S.S.R. 143 (Geol. Ser. no. 60), 1-204 (in Russian). neves, r. 1961. Namurian plant spores from the Southern Pennines, England, Palaeontology, 4, 247-79. playford, g. 1962. Lower Carboniferous microfloras of Spitsbergen, Part I. Ibid. 5, 550-618. 1963. Part II. Ibid. 5, 619-78. potonie, R. 1934. Zur Morphologie der fossilen Pollen und Sporen. Arb. Inst. Palaob. Petrogr. Brenn- steine, 4, 5-24. 1956. Synopsis der Gattungen der Sporae dispersae. I. Teil: Sporites. Beih. Geol. Jahrb. 23, 1-103. 1958. Synopsis der Gattungen der Sporae dispersae II. Teil: Sporites (Nachtrage), Saccites, Aletes, Praecolpates, Polyplicates, Monocolpates. Ibid. 31, 1-114. 1960. Synopsis der Gattungen der Sporae dispersae. III. Teil: Nachtrage Sporites, Fortsetzung Pollenites. Mit Generalregister zu Teil 1-111. Ibid. 39, 1-189. 1 962. Synopsis der Sporae in situ. Die Sporen der Fossilen Fruktifikationen (Thallophyta bis Gymnospermaphyta) im naturlichen System und im Vergleich mit den Sporae dispersae. Ibid. 52, 1-204. and kremp, g. 1954. Die Gattungen der palaozoischen Sporae dispersae and ihre Stratigraphie. Geol. Jahrb. 69, 111-94. 1955. Die Sporae dispersae des Ruhrkarbons, ihre Morphographie und Stratigraphie mit Ausblicken auf Arten anderer Gebeite und Zeitabschnitte ; Teil I. Palaeontographica, B98, 1-136. 1956. Idem; Teil II. Ibid. B99, 85-191. Richardson, j. b. 1960. Spores from the Middle Old Red Sandstone of Cromarty, Scotland. Palaeonto- logy, 3, 45-63. 748 PALAEONTOLOGY, VOLUME 8 richardson, j. b. 1962. Spores with bifurcate processes from the Middle Old Red Sandstone of Scotland. Palaeontology, 5, 171-94. 1965. Middle Old Red Sandstone Spore Assemblages from the Orcadian Basin, North-East Scotland. Ibid. 7, 559-605. schopf, j. m., wilson, l. r., and bentall, r. 1944. An annotated synopsis of Paleozoic fossil spores and the definition of generic groups. Rep. Inv. III. State Geol. Surv. 91, 1-72. scott, d. l. and rouse, g. e. 1961. Perforosporites, a new genus of plant spores from the Lower Devonian of Eastern Canada. J. Palaeont. 35, 977-80. smith, a. h. v. 1960. Structure of the wall in certain miospores belonging to the series Cingulati Pot. and Klaus 1954. Palaeontology, 3, 82-85. staplin, f. l. 1960. Upper Mississippian plant spores from the Golata Formation, Alberta, Canada. Palaeontographica, B107, 1-40. taugourdeau-lantz, j. 1960. Sur la microflore du Frasnian Inferieur de Beaulieu (Boulonnais). Rev. de Micropaleontologie, 3, 144-54. 1962. Remarque sur la structure du genre Hymenozonotriletes (Naoumova 1937 ? 1939) ex Naou- mova 1953. Ibid. 5, 51-53. Thompson, p. w. 1940. Beitrag zur Kenntnis der fossilen flora der Mitteldevons in Estland. Tartu Uelik Lood. Self. 45 (1-4), 195-216. 1952. Beitrag zu Kenntnis der Sporomorphenflora im Unter-und Mitteldevon. Palaeont. Z. 25, 155-9. vigran, j. 1964. Spores from Devonian deposits, Mimerdalen, Spitsbergen. Skr. Norsk Polarinst. 132, 1-30. wigglesworth, v. b. 1959. A simple method for cutting sections in the 0-5 to 1 j u. range, and for sections of chitin. Quart. J. micr. Sci. 100, 315-20. wilson, l. r. and coe, e. z. 1940. Descriptions of some unassigned plant microfossils from the Des Moines Series of Iowa. Amer. Midi. Nat. 23, 182-6. winslow, m. r. 1962. Plant spores and other microfossils from Upper Devonian and Lower Mississip- pian rocks of Ohio. Prof. Pap. U.S. geol. Surv. 364, 1-93. K. C. ALLEN Department of Geology, Sedgwick Museum, Cambridge Manuscript received 30 October 1964 FRESHWATER OSTRACODS FROM THE BATHONIAN OF OXFORDSHIRE by R. H. BATE Abstract. Eight species of freshwater ostracod are described from the Bathonian of Oxfordshire as exposed in the Old Cement Quarry, Kirtlington, near Oxford. Because of the number of specimens available only two, Timiriasevia mackerrowi sp. nov. and Theriosynoecum kirtlingtonense sp. nov., are identified specifically. Other genera represented are : — Bisulcocypris ; Darwinula ; and Limnocythere. The presence of these freshwater ostracods in marine sediments and associated with marine ostracods is considered to be due to their being brought into the area by rivers. The occurrence of two species of Timiriasevia from this country is the first recorded indenti- fication of the genus outside the U.S.S.R. During an excursion organized by the Palaeontological Association and led by Dr. W. S. McKerrow to the Oxford district in 1958, the marine marls and limestones of Bathonian age were examined in the Old Cement Quarry, Kirtlington (Grid Reference SP/495200). The presence of freshwater ostracods associated with an otherwise marine fauna led to a more detailed sampling the following year. Whilst it is intended to describe the complete ostracod fauna in a subsequent publication the presence of this freshwater fauna is considered to be of sufficient importance to warrant its prior description here. The Kirtlington quarry section has been previously described by Arkell (1931, p. 570 and 1947, p. 57) and need not be discussed in detail. The samples from which the ostra- cods are described were obtained from the Fimbriata-waltoni clay at the top of the Bladon Beds and from three horizons within the Wychwood Beds (see text-fig. 1). These beds are of Middle to Upper Bathonian age. During Bathonian times the period of marine deposition in the Oxford area was greatly influenced by the land nearby. This is evidenced by the presence of abundant lignite, reptilian bones, and charophytes ; all these being brought into the area by rivers draining the land surface. It is not surprising, therefore, that freshwater ostracods should be abundant at several horizons throughout the succession. All specimens referred to in the text have been deposited in the collections of the Department of Palaeontology, British Museum (Natural History). Acknowledgements. Dr. P. S. Ljubimova, V.N.I.G.R.I., Leningrad, U.S.S.R., and Dr. H. Oertli, S.N.P.A., Pau, France, kindly sent me material from their collections for comparison purposes. Their assistance in this way is gratefully acknowledged. Mr. S. H. Eagar printed the photographs taken by the author. SYSTEMATIC DESCRIPTIONS Subclass ostracoda Latreille 1806 Order podocopida Muller 1894 Suborder podocopina Sars 1866 Superfamily darwinulacea Brady and Norman 1889 Family darwinulidae Brady and Norman 1889 Genus darwinula Brady and Robertson 1885 Type species. Polycheles stevensoni Brady and Robertson 1870, by original designation. [Palaeontology, Vol. 8, Part 4, 1965, pp. 749-59, pis. 109-11.] 750 PALAEONTOLOGY, VOLUME 8 Text-fig. 1 . Section of north-east face of Old Cement Quarry, Kirtlington, showing the four horizons from which fresh-water ostracods have been obtained. R. H. BATE: BATHONIAN FRESHWATER OSTRACODS 751 Darwinula sp. A Plate 109, figs. 1^4 Material. Two specimens (Io2715-16), a right valve from horizon no. 3 and a complete carapace from horizon no. 4. Description. Carapace small, elongate-oval in outline with the greatest height and width in the posterior third. Greatest length passes through mid-point. Anterior narrowly rounded, posterior broadly rounded. Dorsal margin straight with convex antero-dorsal and postero-dorsal slopes. Ventral margin broadly incurved antero-medially. Left valve larger than the right which it overlaps strongly around the posterior and along the ventral margin. Hinge adont, consisting of a simple straight groove in the right valve. Other internal details not observed. Dimensions. Carapace, BMNH Io2715, length 0-64 mm.; height 0-30 mm. Right valve BMNH Io2716, length 0-64 mm.; height 0-27 mm. Remarks. Although close to Darwinula leguminel/a (Forbes in Lyell 1855, p. 294, text-fig. 334c) Darwinula sp. A may be distinguished by the more evenly rounded anterior margin, the line of greatest length passing through or slightly below mid-point. D. leguminella on the other hand has a more acuminate anterior margin, extended antero-ventrally so that the line of greatest length passes below mid-point. Until further material becomes available it is preferred to avoid giving the present series a specific name. Superfamily cytheracea Baird 1850 Family limnocytheridae Klie 1938 Genus limnocythere Brady 1868 Type species. Cy there inopinata Baird 1843 by subsequent designation Brady and Norman 1889. Limnocythere sp. A Plate 109, figs. 8, 9 Material. A single left valve (Io2713) occurring at horizon no. 4. Description. Valve subrectangular in outline with the ventral margin strongly concave and the greatest height situated in the posterior third. A deep, narrow, vertical sulcus is situated just anterior of mid-point. Greatest length of valve extends through mid-point. Anterior broadly rounded, posterior slightly more obliquely rounded. Dorsal margin long, slightly convex with prominent cardinal angles. The strong incurvature of the ventral margin is antero-median in position, the postero-ventral margin of the valve being strongly convex with a flat, distinct, marginal border. An oval swelling is situated just below and behind the vertical sulcus; posterior part of valve also prominently swollen. Shell surface granular, possibly originally punctate but now not clearly deter- minable because of preservation. Internally the lophodont type hinge consists of simple terminal sockets open ventrally to the inside of the valve, and a median element formed by the dorsal edge of the valve. Inner margin and line of concrescence coincide, duplica- ture of moderate width. Muscle scars and radial pore canals not determined. Dimensions. Left valve, BMNH Io2713, length 0-83 mm.; height 0-43 mm. 752 PALAEONTOLOGY, VOLUME 8 Remarks. Limnocythere sp. A appears to be close to Limnocy there fragilis Martin (1940, p. 348, pi. 7, figs. 105-9, pi. 9, fig. 152) in general appearance although it can be distin- guished by the much stronger ventral incurvature and correspondingly more convex postero-ventral margin. The posterior is also more obliquely rounded in the present species which is further differentiated by the presence of a small oval swelling below and behind the vertical sulcus. Limnocythere sp. B Plate 109, figs. 5-7 Material. A single left valve (Io2714) found at horizon no. 2 in the Kirtlington section. Description. Valve subrectangular with the greatest height in the anterior third. Greatest length through mid-point. Dorsal margin straight with prominent cardinal angles, the anterior angle in particular being noticeably upstanding. Ventral margin medially in- curved. Anterior broadly rounded, posterior more narrowly rounded. Valve bi-sulcate with a median vertical sulcus and a shorter antero-median sulcus. Shell surface reticulate and rather irregularly swollen, giving a warty appearance. A short, pointed spine, back- wardly projected, is situated below and slightly behind the median sulcus. Anterior and posterior margins flattened to form distinct marginal borders. Hinge lophodont, the terminal sockets being open to the interior of the valve. Median bar long and narrow. Inner margin and line of concrescence coincide, the duplicature being of moderate width. Radial pore canals long and straight, 8 anteriorly and 7 posteriorly. Muscle scars not observed. Dimensions. Left valve, BMNH Io2714, length 0-83 mm.; height 0-44 mm. Remarks. The ornamentation, warty appearance, and the presence of a ventro-lateral spine as well as the bi-sulcate development distinguish this species from all others previously described. Genus bisulcocypris Pinto and Sanguinetti 1958 Type species. Bisulcocypris pricei Pinto and Sanguinetti 1958, by original designation. Remarks. The genus Bisulcocypris, exclusively freshwater in habit, has a surface orna- mentation consisting of pitting or reticulation with a tendency towards the development of small nodes or tubercles. Possibly the degree of tubercular development might reflect some variation in the environment (e.g. salinity or pH) although some species appear to be characteristically ornamented in this way. However, the development of nodes EXPLANATION OF PLATE 109 All figures x 65 unless otherwise stated. Figs. 1-4. Darwinula sp. A. Right, left, dorsal, and ventral views, complete carapace, Io2715. Figs. 5-7. Limnocythere sp. B. External and internal views (fig. 7, showing radial pore canals), left valve, Io2714. Figs. 8, 9. Limnocythere sp. A. External and internal views, left valve, Io2713. Figs. 10-12. Bisulcocypris sp. B. Internal view to show hinge, external view (ventral margin broken on this illustration) and internal view to show radial pore canals, left valve. Io2718. Figs. 13, 14. Bisulcocypris sp. A. Muscle scars and external view, broken right valve, Io2717. Fig. 13, x 170. # % Palaeontology, Vol. 8 PLATE 109 BATE, Bathonian freshwater ostracods R. H. BATE: BATHONIAN FRESHWATER OSTRACODS 753 within Bisulcocypris is never carried to such an extent as is considered here to be found in Theriosynoecum. Of the number of species placed by Pinto and Sanguinetti (1962) into Bisulcocypris one, B.fittoni (Mantell), is tuberculate not only as an adult but also throughout its ontogeny (see Sohn and Anderson 1964) and is to be more accurately referred to Theriosynoecum. Both genera are freshwater in habit so there is no objection to this. Pinto and Sanguinetti’s statement (p. 76, which will be discussed later) that Theriosynoecum is a marine genus is not accepted here. The presence in the Kirtlington section of two species of Bisulcocypris and the occur- rence of B. tenuimarginata (Oertli 1957, p. 765) in Bathonian sediments of Poitou, France, extends the range of this genus below that indicated by Pinto and Sanguinetti (p. 76). As there is some confusion at the present time concerning the separate identity of the genera Bisulcocypris and Theriosynoecum it is relevant to state here that only strongly tuberculate species (throughout their ontogeny) are considered to belong to Therio- synoecum irrespective of the possession or absence of an accommodation groove. There appears to be no real justification for their separation on any other grounds. Bisulcocypris sp. A Plate 109, figs. 13, 14 Remarks. A broken right valve (Io2717) from horizon no. 1 is all that has been found of this particular species. Shell surface is strongly punctate with longitudinal ridges developed along the ventral and ventro-lateral surfaces. Bisulcocypris sp. B Plate 109, figs. 10-12 Material. Two left valves (Io2718-19) occurring at horizon no. 4. Description. Valve subquadrate in outline with the greatest length passing through mid- point and the greatest height in the posterior third. Anterior and posterior broadly rounded. Ventro-lateral margin convex, overhanging the ventral surface. Ventral margin with shallow antero-median incurvature; dorsal margin strongly concave medially with broadly rounded anterior cardinal angle and sharply acute posterior angle. Posterior part of valve strongly inflated, the inflation being obliquely angled below the posterior car- dinal angle to leave a flattened area. Shell surface uniformly reticulate with a series of short spines and nodes. In both left valves a short spine is present on the posterior car- dinal angle, two further spines occur on the upper part of the posterior and a number around the anterior margin where they are found just to the inside of the narrow mar- ginal border. The area between the two antero-dorsal sulci is noticeably swollen and two low nodes occur in the anterior half of the valve. In the posterior half three low nodes may occur one above the other just behind the second of the two sulci. These are present, however, only on one of the two valves and in this case a prominent tubercle is also developed just above the line of greatest length and in front of the two posterior spines. Longitudinal striae extend along the ventral and ventro-lateral surfaces and around the posterior. Hinge lophodont, the posterior socket in the left valve being short and trian- gular in shape whilst the anterior socket is elongate and rather narrow. Median bar 754 PALAEONTOLOGY, VOLUME 8 smooth, long, and strongly developed. An accommodation groove is present only over the posterior half of the hinge. Inner margin and line of concrescence do not coincide around the anterior margin where a narrow vestibule is developed. Radial pore canals (approximately 30 anteriorly) short, straight, and rather slender, posterior canals not seen. A narrow flange extends, outside the selvage, around the free margin from the anterior to the posterior cardinal angle. Muscle scars not observed. Dimensions. Left valve, BMNH Io2718, length 1-03 mm.; height 0-63 mm. Remarks. It is impossible to contrast sp. A and sp. B on the amount of material available but it would appear that in the case of sp. B the ornamentation is reticulate rather than punctate with the tendency to develop small tubercles, the two sulci are also separated by a distinct swelling. In sp. A the area between the sulci is flattened. Bisulcocypris tenuimarginata (Oertli 1957, p. 765, pi. 23, figs. 15-24) is a much smaller, compact ostra- cod, showing no tendency to develop small nodes. Genus theriosynoecum Branson 1936 Type species. Morrisonia wyomingensis Branson 1935, by monotypy. Remarks. The genus Theriosynoecum was originally described from Wyoming, U.S.A., under the preoccupied name of Morrisonia by Branson from freshwater sediments of the Morrison formation. Pinto and Sanguinetti (1962, p. 76) suggest that Theriosynoecum is a marine ostracod because of its association with marine forms in Bathonian and Callovian sediments of England. A few years ago when Professor Pinto was visiting this country he examined my material from Kirtlington, where the association of this ostracod with marine ostracods is beyond dispute. However, when the complete evidence, both stratigraphical and palaeontological, is taken into consideration it can be seen that the presence of Therio- synoecum, as well as specimens of Darwinula, Bisulcocypris , and Timiriasevia, in marine sediments at Kirtlington is due entirely to their being brought into the area by rivers. Theriosynoecum kirtlingtonense sp. nov. Plate 110, figs. 1-11, Plate 111, fig. 1 Material. Fourteen specimens (Io2720-33) from horizons nos. 1-4. Diagnosis. Carapace rectangular with well-rounded anterior and posterior margins. Posterior cardinal angle acute, anterior angle rounded. Shell surface strongly reticulate anteriorly, generally weakly reticulate posteriorly. Yentro-lateral margin projected as thinly developed ridge, particularly postero-laterally. 8 strong tubercles are developed in EXPLANATION OF PLATE 110 All figures x 65 unless otherwise stated. Figs. 1-11. Theriosynoecum kirtlingtonense sp. nov. Figs. 1-3, 11. External, internal, and dorsal views and internal view (x 110) to show anterior radial pore canals, male left valve, holotype Io2720. Figs. 4-7. External, internal, dorsal, and ventral views male right valve, paratype Io2726. Figs. 8, 9. External and internal views, female right valve, paratype Io2725. Fig. 10. Muscle scars (x 120), female right valve, paratype Io2730. Palaeontology, Vol. 8 PLATE 110 BATE, Bathonian freshwater ostracods R. H. BATE: BATHONIAN FRESHWATER OSTRACODS 755 both male and female dimorphs, 3 in the anterior half and 5 in the posterior half. Small additional nodes may be present close to the anterior and posterior margins and at the posterior cardinal angle. A narrow anterior vestibule is present. Radial pore canals short, thin, and numerous, approximately 27 anteriorly. Muscle scars an oblique row of 4 oval scars, no anterior scars observed. Holotype. A male left valve BMNH Io2720, from horizon no. 4. Description. Carapace bi-sulcate, oval-rectangular in outline with well-rounded anterior and posterior margins. Dorsal margin slightly convex in the left valve, the outline being broken posteriorly by the development of an accommodation groove, medially concave in the right valve. Ventral margin with a broad median incurvature. Ventro-lateral margin projected ventrally, overhanging the ventral surface in side view, particularly well developed postero-laterally. Cardinal angles prominent with the posterior angle of the left valve being particularly acute, and possessing a short spine. Sexual dimorphism strongly apparent, the females being quite swollen posteriorly. Greatest length of cara- pace through mid-point, greatest height in the male dimorph either in the anterior or the posterior third, in the female dimorph the greatest height occurs in the posterior third. Although a complete carapace has not been found the greatest width would be in the posterior third. A flattened marginal border extends round the anterior. Carapace strongly tuberculate, possessing a maximum of 8 tubercles, one of which separates the two antero-dorsal sulci. This oval tubercle may, however, be missing, in which case the two sulci fuse to form a single broad depression. In front of the short anterior sulcus a low oval tubercle is developed just below the anterior cardinal angle. A second rounded tubercle occurs just below this anterior sulcus. Behind the second of the two sulci 5 tubercles may be developed. In position there is a vertical row of 3 tubercles immediately behind the sulcus and of these the dorsal and ventral ones are the most strongly developed, being directed to the rear. The median tubercle is often found out of position, further towards the posterior, or may be missing altogether. Two prominent tubercles are situa- ted further back on the posterior part of the valve and again are situated one above the other. Shell surface is reticulate but often only in the anterior half, the ornamentation fading away towards the posterior. In some specimens, however, the reticulation is present over the whole of the valve, and in all specimens continues over the tubercles. Young instars show the same number and positioning of the tubercles. Ventral and underside of ventro-lateral surface strongly ornamented with about 4 longitudinal ridges. Particularly in the females further ridges may also be present on the ventro- lateral surface and around the posterior. Hinge lophodont: left valve with smooth terminal sockets of which the posterior socket is triangular in shape and the anterior socket is more elongate. Median bar long, smooth, and strongly developed. Above the median element an accommodation groove is developed in the posterior half of the valve only. Right valve with a terminal blade-like tooth and a smooth median groove which follows the outline of the dorsal margin by bending downwards medially and is most broadly developed in the anterior half. Inner margin and line of concrescence coincide except for around the anterior margin where medially a narrow vestibule is apparent. Radial pore canals numerous, short, and straight, approximately 27 in number an- teriorly. Occasionally the canals can be seen to occur in pairs diverging towards the outer margin. Selvage prominent with a well-developed flange outside. This flange 756 PALAEONTOLOGY, VOLUME 8 extends completely around the free margin but is only poorly developed in the region of the ventral incurvature. An oblique row of 4 oval adductor scars has been observed, but so far no scars anterior to these. Dimensions. Holotype: Male left valve, BMNH Io2720, length 1-03 mm.; height 0-56 mm. Paratypes: Male right valve, BMNH Io2726, length 1-05 mm.; height 0-59 mm. Female right valve, BMNH Io2725, length 1-15 mm.; height 0-70 mm. Remarks. Theriosynoecum kirtlingtonense is the first species of the genus to be described from Bathonian sediments and is readily distinguished from other members of the genus by the number of tubercles present and their position on the carapace. For example T. fittoni (Mantell 1844, p. 545, fig. 2) from the English Weald Clay (Lower Cretaceous) possesses 10 tubercles whilst T. wyomingense (Branson 1935, p. 521, pi. 57, figs. 17-21) from the Morrison formation (Upper Jurassic), U.S.A. possesses 4 tubercles in the dorsal region of the carapace and a crescentic ventro-lateral ridge. Genus timiriasevia Mandelstam 1947 Type species. Timiriasevia epidermiformis Mandelstam 1947, by original designation. Remarks. Timiriasevia is found in freshwater sediments of Mesozoic age and as such has only been described from the U.S.S.R., where some 16 species, all of relatively small size, are known. The species described here, T. mackerrowi sp. nov. and Timiriasevia sp. A constitute the first record of the genus outside the U.S.S.R. Timiriasevia mackerrowi sp. nov. Plate 111, figs. 2-12 Derivation. The species is named after Dr. W. S. McKerrow. Material. Ten specimens (Io2734-43) from horizons nos. 1-4. Diagnosis. Carapace oval in side view slightly constricted anterior of mid-point, greatly expanded posteriorly; heart-shaped in dorsal view. Valves taper to anterior. Left valve larger than right. Adult instars with prominently developed carapace in postero-ventral region. Ornamentation consists of a series of fine longitudinal ridges which follow the outline of the carapace to give a finger-print arrangement. Hinge lophodont. Inner margin and line of concrescence do not coincide anteriorly and posteriorly. Radial pore canals short, straight, simple, and fairly numerous, exact number not determined. EXPLANATION OF PLATE 111 All figures x 65. Fig. 1. Theriosynoecum kirtlingtonense sp. nov. Dorsal view of right valve to show terminal hinge teeth, female paratype Io2725. Figs. 2-12. Timiriasevia mackerrowi sp. nov. Figs. 2-5. Right, dorsal, ventral, and left views, complete carapace, holotype Io2734. Fig. 6. External view, right valve, paratype Io2740. Figs. 7, 12. Internal and dorsal views, right valve, paratype Io2741. Fig. 8. Internal view, juvenile right valve, showing narrow duplicature, paratype Io2737. Figs. 9-11. External, internal, and dorsal views, left valve, paratype Io2736. Figs. 13-15. Timiriasevia sp. A. External, internal, and dorsal views, right valve, Io2744. Palaeontology , Vol. 8 PLATE 111 BATE, Bathonian freshwater ostracods R. H. BATE: BATHONIAN FRESHWATER OSTRACODS 757 Muscle scars consist of a subvertical row of 4 oval adductor scars, a rounded antero- dorsal antennal scar, and a larger rounded antero-ventral mandibular scar. Holotype. A complete carapace, BMNH Io2734, from horizon no. 1. Description. Carapace oval in side view slightly constricted mid-dorsally just anterior of mid-point and with broadly convex dorsal margin and rounded anterior and posterior margins. Ventro-lateral margin convex, overhanging the ventral surface. Ventral margin convex, incurved antero-medially. Anterior narrower than posterior, greatest height being situated just behind valve mid-point. Greatest length of carapace passes very slightly below mid-point; greatest width in posterior third. Carapace heart-shaped in dorsal view tapering to the anterior, where there is a flattened marginal border; broadly expanded to the rear. The posterior expansion is most prominently developed in adult instars where the selvage moves inwards relative to the postero-ventral margin. Shell surface laterally ornamented by a series of fine longitudinal ridges which tend to follow the outline of the valve and are therefore somewhat concentrically arranged towards the outer margins. The result is to produce an ornament reminiscent of a finger-print. The ventral surface is similarly ornamented by fine longitudinal ridges. Left valve slightly larger than the right which it overlaps mid-ventrally and around the anterior and posterior. Postero-ventrally the left valve may have originally overlapped the right, but this is not clear in the present material. Hinge lophodont : left valve with long, narrow anterior socket and a shorter, wider, posterior socket. Both sockets are smooth and terminally open, being continuous with a narrow groove extending around the anterior and posterior, below the selvage. Median bar smooth and short. In the right valve the smooth terminal hinge elements are simply the anterior and posterior terminations of the prominent selvage. Median groove short but quite broad. Inner margin and line of concrescence do not coincide around the anterior and posterior margins, although the vestibule is extremely narrow. Radial pore canals short, straight, and simple anteriorly, the exact number has not been determined. Selvage very prominent around the free margin, in adults with a broad anterior and posterior flange outside. This flange is essentially composed of a broad flange groove and a narrower flange proper which extends along the ventral surface. Posteriorly, in adult specimens, because of the develop- ment of the postero-ventral part of the valve, the flange groove becomes broad and extends obliquely across that part of the valve. In more juvenile instars the flange groove is narrower and follows the outline of the posterior margin. It is only in these instars that the relationship of the inner margin to the fine of concrescence may be determined posteriorly, as in adults the flange groove completely extends over the inner termination of the duplicature. Muscle scars as seen in a single specimen consist of an oblique row of 4 oval adductor scars, a smallish, round, antero-dorsal antennal scar, and a larger, apparently also rounded, antero-ventral mandibular scar. Dimensions. Holotype: Carapace, BMNH Io2734, length 0-58 mm.; height 0-36 mm.; width 0-43 mm. Paratypes: Left valve, BMNH Io2736, length 0-60 mm.; height 0-37 m.m. Juvenile right valve, BMNH Io2737, length 0-51 mm.; height 0-34 mm. Right valve, BMNH Io2740, length 0-63 mm.; height 0-37 mm. Right valve, BMNH Io2741, length 0-57 mm.; height 0-34 mm. Remarks. Timiriasevia mackerrowi closely resembles T. polymorpha Mandelstam (in B 6612 3 D 758 PALAEONTOLOGY, VOLUME 8 Galeeva 1955, p. 61, pi. .15, figs. 4a, b, b) although it may, however, be distinguished by ornamentation which in T. polymorpha consists of a neat reticulation. Gomphocythere sp. 1. Oertli (1957, p. 763, pi. 22, figs. 24-31) which is, in fact, a species of Timiriasevia, has a similar although much weaker ornament to that of the present species but differs in outline, tending to be much more square with the greatest height in the anterior and not the posterior third. Timiriasevia sp. A Plate 111, figs. 13-15 Material. A single right valve (Io2744) from horizon no. 4. Description. Valve sub-rectangular in outline with the greatest length passing very slightly below mid-point and greatest height being in the anterior third. Dorsal margin convex with fairly prominent cardinal angles ; anterior and posterior rounded ; ventral margin convex with a shallow antero-median incurvature. Ornamentation consists of longitudinal low ridges along the ventral surface which laterally become terminally oblique V-ing towards the dorsal margin and tending to branch at about valve centre. A prominent lateral ridge extends along the ventro-lateral margin. Hinge lophodont, con- sisting of smooth terminal elements formed by the selvage, and an elongate, smooth, median groove. Inner margin and fine of concrescence appear to be almost coincident anteriorly, not clear posteriorly. Radial pore canals short, straight, and simple, exact number not determinable. Selvage prominent with a flange outside. Flange groove of moderate width postero-ventrally. Muscle scars not seen. Dimensions. Right valve, BMNH Xo2744, length 0-50 mm. ; height 0-27 mm. Remarks. Timiriasevia sp. A is found associated with T. mackerrowi sp. nov., from which it is easily distinguished by shape, position of greatest height, presence of a prominent ventro-lateral ridge, and a more pronounced V-ing of the lateral ornament. These chara- ters also distinguish this species from others previously described from the U.S.S.R. REFERENCES arkell, w. j. 1931. The Upper Great Oolite, Bradford Beds and Forest Marble of South Oxfordshire, and the succession of Gastropod faunas in the Great Oolite. Quart. J. Geol. Soc. Lond. 87, 563-629, pis. 48-49. 1947. The Geology of Oxford. 267 pp. Oxford. branson, c. c. 1935. Fresh-water Invertebrates from the Morrison (Jurassic?) of Wyoming. J. Paleont. Tulsa, 9, 514-22, pis. 56-57. galeeva, l. i. 1955. Cretaceous Ostracod succession of Mongolian Peoples Republic. Moscow [in Russian]. Reference not seen, details from Ljubimova 1956. ljubimova, p. s. 1956. Ostracods from the Cretaceous deposits of the eastern part of the Mongolian National Republic and their significance for stratigraphy. Trud. vses. nef.-nauch. issled. geol. Inst. ( VNIGRI ), Leningrad, (N.S.) 93, 1-174, pis. 1-25 [in Russian]. lyell, c. 1855. A Manual of Elementary Geology. 5th ed. 655 pp., 717 text-figs. London. mantell, G. A. 1844. The Medals of Creation, ix+1016 pp. London. martin, g. p. r. 1940. Ostracoden des norddeutschen Purbeck und Wealden. Senckenbergiana, Frankfurt a.M., 22, 275-361, pis. 1-12. oertli, h. j., 1957. In Bernard, f., bizon, J.-J., and oertli, H. J. Ostracodes lacustres du Bathonien du Poitou (Bassin de Paris). Bull. geol. Soc. France, Paris, 6, 753-70, pis. 21-23. R. H. BATE: BATHONIAN FRESHWATER OSTRACODS 759 oertli, h. j. 1963. Fames d'Ostracodes du Mesozoique de France. 57 pp. 90 pis. Leiden. pinto, i. d. and SANGUINETTI, y. 1962. A complete revision of the genera Bisulcocypris and Therio- synoecum (Ostracoda) with the world geographical and stratigraphical distribution. Esc. Geol. P. Alegre, 4, 1-165, pis. 1-17. sohn, i. G. and anderson, f. w. 1964. The ontogeny of Theriosynoecum filtoni (Mantell). Palae- ontology, London, 7, 72-84, pi. 15. R. H. BATE Department of Palaeontology, British Museum (Natural History) London S.W. 7 Manuscript received 25 November 1964 THE PALAEONTOLOGICAL SOCIETY Extracts from the Annual Report of the Council for 1964-5 Membership. On 31 December 1964 there were 1,269 members (646 Ordinary, 108 Student, and 515 Institutional), a net increase of 79 members during the year. Finance. The year’s working shows an excess of income over expenditure of £1,415 in contrast with the slight deficit estimated at the beginning of the year. This excess has been transferred to the Publi- cations Reserve Account to cover the cost of the increased numbers of Palaeontology to be printed in future. There has been a small under-expenditure on Palaeontology compared with the estimates, and a sum of £343 unspent from the last part of Volume 6 in the previous year has also been brought forward. Sales of back numbers, etc., from Messrs. Blackwells have brought in a much larger amount than anticipated, but this is partly due to their changing over during the year from paying for parts as sold, to paying for stocks as received. The income from subscriptions has been approximately as expected, and so has the interest received. Administrative expenditure has been slightly less than anticipated. ‘Palaeontology’. The four parts of Volume 7 (for 1964) were published during 1964-5. They contained 45 papers. Meetings. Four meetings took place during 1964-5. The Association is grateful to the Council of the Geological Society of London, Professor J. F. Kirkaldy (Queen Mary College, London), and Professor E. A. Vincent (University of Manchester) for generously granting facilities for meetings, and to the Local Secretaries for their efficient services. a. The Seventh Annual Genera! Meeting was held in the Rooms of the Geological Society of London, Burlington House, London, W. 1. on Wednesday, 4 March 1964, at 5.0 p.m. The Annual Report of the Council for 1963-4 was adopted and the Council for 1964-5 was elected. Professor T. Neville George delivered the Seventh Annual Address on ‘Morphogeny in the Spirifers’. b. A Field Demonstration Meeting was held at Dudley on Saturday, 23 May 1964. The theme was ‘The Wenlockian faunas of the Wren’s Nest’. Dr. I. Strachan was Leader and Local Secretary. c. A Discussion and Demonstration Meeting was held in the Department of Geology, Queen Mary College, London, E. 1 on Saturday, 17 October 1964 at 2.30 p.m. There were twenty-one exhibits and about seventy persons attended. The Local Secretary was Dr. F. A. Middlemiss. d. A Discussion Meeting on ‘Chemistry and Microstructure of Fossils’ was held in the Department of Geology, University of Manchester, on Friday/Saturday, 18/19 December 1964. About eighty persons attended. Ten papers were read during three sessions and there were twenty-three exhibits. A Palaeontological Association Dinner was held at Woolton Hall, University of Manchester on 18 December. Dr. F. M. Broadhurst was Local Secretary. Council. The following were elected members of the Council of the Association for 1964-5 at the Annual General Meeting on 4 March 1964: President : Dr. L. R. Cox. Vice-Presidents: Professor T. Neville George, Dr. W. S. McKerrow, Professor F. H. T. Rhodes. Treasurer: Dr. T. D. Ford. Assistant Treasurer: Dr. C. Downie. Secretary: Dr. C. H. Holland. Editors: Mr. N. F. Hughes, Dr. Gwyn Thomas, Dr. I. Strachan, Dr. M. R. House. Other members: Dr. C. G. Adams, Dr. F. M. Broadhurst, Professor O. M. B. Bulman, Professor P. M. Butler, Dr. W. J. Clarke, Dr. G. Y. Craig, Dr. W. T. Dean, Dr. B. M. Funnell, Dr. F. A. Middlemiss, Mr. M. Mitchell, Dr. A. J. Rowell, Dr. R. J. G. Savage, Professor Scott Simpson. BALANCE SHEET AND ACCOUNTS FOR THE YEAR ENDING 31 DECEMBER 1964 Balance Sheet Liabilities Publications Reserve Account £ s. d. £ s. d. Balance as per Annexed Account Amounts Received In Advance 6,083 2 3 Subscriptions for 1965 ........ Provision for Cost of Publication of Palaeontology , Vol. 7 as per Income 375 1 4 and Expenditure Account ....... 6,437 0 1 Less Expenditure incurred to 31 December 1964 .... 3,886 18 11 2,550 1 8 Sundry Creditors 10 10 0 Note: No amount has been included in these Accounts for Subscriptions unpaid at 31 December 1964. £9,018 15 3 Assets Office equipment at cost ........ 27 6 6 Less Depreciation to date . - . 4 5 4 23 1 2 Investments at cost: Equities Investment Fund for Charities — 958 Units 999 18 3 5% Defence Bonds ......... 4,000 0 0 4,999 18 3 Sundry Debtors and Payments in Advance: Authors for reprints 137 14 0 Advance payments re Palaeontology, Vol. 8 . 411 7 6 549 1 6 Cash at Bank — Deposit Account ....... 2,686 11 10 Current Account ....... 760 2 6 3,446 14 4 £9,018 15 3 Report of the Auditors to the Members of the Palaeontological Association. We have examined the above Balance Sheet and annexed Income and Expenditure Account which in our opinion give respectively a true and fair view of the state of the Association’s affairs as at 31 December 1964 and of its income and expendi- ture for the year ended on that date. BAKER BROS. HALFORD & CO. Chartered Accountants. 762 PALAEONTOLOGY, VOLUME 8 Income and Expenditure Account Expenditure £ s. d. £ s. d. To Provision for Cost of Publication of Palaeontology Vol. 7, Part 1 1,558 3 10 Part 2 1,578 16 9 Part 3 1,650 0 0 Part 4 1,650 0 0 6,437 0 7 Less Excess provision for Vol. 6 not required .... 343 0 6 6,094 0 1 Administrative Expenses : Postage and stationery 76 3 4 Insurance 1 16 3 Audit Fee 10 10 0 Miscellaneous .......... 5 3 6 Cost of distribution of publications ...... 44 16 6 Honorarium to Assistant Treasurer 25 0 0 Duplicating and dispatching circulars, etc 156 5 4 319 14 11 Excess of Income over Expenditure for the year transferred to Publica- tions Reserve Account ........ 1,415 15 7 £7,829 10 7 Income By Subscriptions .......... 4,923 18 4 Sales of publications ........ 1,785 10 2 Interest received: 5% Defence Bonds 200 0 0 Bank Deposit Account 60 12 1 Equities Investment Fund for Charities ..... 35 18 6 296 10 7 Special donations ......... 16 10 0 General donations: British Petroleum Co. Ltd. ....... 250 0 0 Texaco .......... 175 0 0 Burmah Oil Co. Ltd 100 0 0 Mobil Oil Co. Ltd 100 0 0 625 0 0 Miscellaneous receipts : Commission on sales of American Treatise .... 147 8 6 Sundries .......... 34 13 0 182 1 6 £7,829 10 7 Publications Reserve Account To Balance as per Balance Sheet ....... 6,083 2 3 £6,083 2 3 By Balance at 31 December 1963 4,667 6 8 Less General Donations not received ...... 0 0 0 Excess Income over Expenditure for the year transferred from Income and Expenditure Account ........ 1,415 15 7 £6,083 2 3 INDEX Pages 1 to 198 are contained in Part 1 ; pages 199 to 372 are in Part 2; 373 to 576 in Part 3; and 577 to 759 in Part 4. Figures in Bold type indicate plate numbers. A Acanthorhynchia, 609; senticosa, 87. Acanthothiris, 604; spinosa, 84-86. Acanthotriletes raptus, 697, 94. Accinctisporites lignatus, 304, 37. Acrocythere hauteriviana, 378. Adams, T. D., and Haynes, J. Foraminifera in Holocene marsh cycles at Borth, Cardiganshire (Wales), 27. Algae: Carboniferous Codiaceae, 192; Cretaceous Codiaceae, 199. Alisporites, 307, 345; circulicorpus, 309, 39; minuti- saccus, 310; nuthallensis, 346, 43; cf. parvus, 310, 39; toralis, 308, 38. Allen, K. C. Lower and Middle Devonian spores of North and Central Vestspitsbergen, 687. Allison, R. C. Apical development in turritellid classi- fication with a description of Cristispira pugetensis gen. et sp. nov., 666. Alps: Jurassic and Cretaceous microfossils, 391. Alvin, K. L. A new fertile lycopod from the Lower Carboniferous of Scotland, 281 . Amphilichas, 598; priscus, 598, 83; sp., 599, 83. Amplexopora? , 7; evenensis, 7, 3; sp., 9, 3. Ampyx sp., 589, 81. Ancyrospora, 743; langii, 743, 106; reuta, 744, 107; trocha, 743, 107; sp., 745, 107. Arabicodium, 200; aegagrapiloides, 23. Archaeozonotriletes, 721; columnus, 723, 100; meand- ricus, 724, 100; sarus, 723, 100; variabilis, 721, 100. Arcoscalpellum, 629; arcuatum, 90; comptum, 629, 90. Arfia, 659. Arthropoda: See Cirripedia, Ostracoda and Trilobita. Astiericeras, 438; asterianum, 59. Aulicosporites, 739; aulicus, 739, 105; vitabilis, 740, 105. Auroraspora macromanifestus, 730. B Babinka, 231 ; prima, 242, 26-28. Barker, D. See Kaye, P. Bate, R. H. Freshwater ostracods from the Bathonian of Oxfordshire, 749. Bates, D. E. B. A new Ordovician crinoid from Dol- gellau, North Wales, 355. Billingsastraea, 547; aperta, 547, 74, 75; speciosa, 549, 76. Bisulcocypris, 752; sp. A, 753, 109; sp. B, 753, 109. Bivalvia: Australian Carboniferous, 75; mode of life of ‘ Posidonia 156; Ordovician lucinoid, 231. Bositra buchi, 157, 19. Boueina, 199; hochstetteri var. liasica, 23; cf .pygmaea, 24. Brachiopoda: from New South Wales, 54; Ludlovian, 454; sensory spines, 604; unusual structures in atrypids, 358. Brodispora striata, 300, 36. Bryozoa: Caradocian from Shropshire, 5; Dionella gen. nov., 492; from New South Wales, 54; Ptilofenestella gen. nov., 478. Bullatisporites, 702; bullatus, 703, 96. Bulman, O. M. B. See Rickards, R. B. Bumastoides scoticus, 585, 89. Burgess, 1. C. Calcifolium (Codiaceae) from the Upper Visean of Scotland, 192. C Cadosina furca, 395, 51. Calamospora, 692; microrugosa, 692; nigrata, 693; witney ana, 693. Calamospora breviradiata type, 93. Calcifolium, 192; okense, 193, 21; punctatum, 194, 22. Calpionella, 393; alpina, 51, 52; elliptica, 51, 52. Calyptosporites, 735; indolatus, 737, 106; micro- spinosus, 735; optivus, 736, 104; proteus, 735, 103. 1 Camarotoechia ’ sp. B, 71, 11. Camerosporites, 312; secatus, 313, 36, 38. Campbell, K. S. W. An almost complete skull roof and palate of the dipnoan Dipnorhynchus sussmilchi (Etheridge), 634. Camptoionotriletes, 719; aliquant us, 720, 100; assamin- thus, 719, 99. Carboniferous : calami talean cone, 107, 681 ; calcareous algae, 192; corallum increase in Lithostrotion, 204; fauna from New South Wales, 54; fertile lycopod, 281; new fenestrate bryozoan, 478; Nuculoceras, 226; palaeoecology of Goniatite Bed, 186; West- phalian megaspores, 82. Carrickia, 579; pelagia, 580, 81. Cephalopoda: Australian Carboniferous, 74; Nuculo- ceras, 226; Scaphites, 397. Ceratocephala sp., 600, 83. Ceraurinella sp., 591, 82. Chaphekar, M. On the genus Pothocites Paterson, 107. Chelinospora, 728; concinna, 728, 101; ligurata, 729, 102; perforata, 129, 102. Chordasporites singulichorda, 306, 38. Cirratriradites, 718; avius, 718, 99; dissutus, 718, 99. Cirripedia: Arcoscalpellum comptum, 629. Clarke, R. F. A. British Permian saccate and mono- sulcate miospores, 322; Keuper miospores from Worcestershire, England, 294. Cleiothyridina australis, 65, 10. Codakia, 233; orbicularis, 27; ( Ctena ) sp., 26. Codiaceae: Carboniferous from Scotland, 192; Cretaceous, 199. 764 INDEX Coelenterata : Australian Devonian corals, 518; corallum increase in Lithostrotion, 204; North American Smithiphyllum, 618. Collins, J. S. H. Arcoscalpellum comptum (Withers), a species of cirripede new to the Gault, 629. Conbaculatisporites longdonensis, 299, 36. Convolutispora, 703; disparalis, 704, 96; mimerensis, 704, 97; tegula, 705, 97; vermiformis, 703, 96. Cookson, I. C. On a new species of Hoegisporis Cookson, 39. Copper, P. Unusual structures in Devonian Atrypidae from England, 358. Craspedispora, 709; craspeda, 710, 97. Crassostrea echinata, 609, 87. Cretaceous: Aptian ostracods, 375; calcareous algae, 199; cirripede, 629; Hoegisporis, 39; microfossils from Alps, 391 ; newbryozoan genus, 492; Scaphites, 397. Crinoidea: Devonian from Somerset, 11; Ordovician from Wales, 355; Saccocoma, 391. Cristispira pugetensis, 676, 92. Crust aesporites, 330; globosus, 331, 40. Cybele ?sp., 595, 83. Cycadopites, 312, 347; acerrimus, 312, 39; rarus, 348, 44; subgranulosus, 312, 39. Cyclogranisporites, 298, 694; congestus, 298, 35; oppressus, 298, 35; plicatus, 695, 94; rotundus, 694. Cymbosporites, 725; catillus, 727, 100; cyathus, 725, 101. Cynohyaenodon, 642; ruetimeyeri, 648; trux, 648. Cystosporites, 101; giganteus, 101, 16; varius, 102, 16. Cythereis, 384; bekumensis, 384, 50; sutterbyensis, 385, 50. Cytherella ovata, 385, 50. Cytherelloidea cf. ovata, 386, 50. Cytheropteron, 378; (C.) cf. inaequivalve, 378, 48; rugosa, 379, 48; ( Eocytheropteron ) nova reticulata, 379, 48; ( Infracytheropteron ) exquisita, 380; lindumensis, 380, 48, Cytherura reticulosa, 377. D Darwinula, 749; sp. A, 751, 109. Densosporites devonicus, 713, 98. Devonian: corals from New South Wales, 518; Dipnorhynchus, 634; new crinoid from Somerset, 11; Smithiphyllum from North America, 618; spores from Spitsbergen, 687 ; structures in brachio- pods, 358. Diatomozonotriletes sp., 710, 97. Dicaulograptus gen. nov., 278. Dicellograptus sp., 41. cf. Dictyotriletes minor, 706, 97. Dielasma picketti, 69, 12. Dimeropyge hystrix, 587, 81. Diodontopteria delicata, 78, 12. Dionella, 492; flacilla, 512, 69; simulacrum, 507, 70; suffragista, 513, 71; surculus, 501, 69; trifaria, 496, 67, 68; trigonopora, 503, 69; triminghamensis, 510, 71. Diplodonta punctata, 28. Dipnorhynchus sussmilchi, 634, 91. Disphyllum, 535; gregorii, 536; cf. gregorii, 536, 73. Divaricella sp., 28. Dolocythere rara, 377. Dolocytheridea minuta, 377, 48. E Echinodermata. See Crinoidea. Eggert, D. A. See Hibbert, F. A. Elliott, G. F. The interrelationships of some Cretaceous Codiaceae (Calcareous Algae), 199. Ellipsovelatisporites plicatus, 303, 39. Emphanisporites, 707; decoratus, 708, 97; minutus, 709, 97; neglectus, 708, 97; patagiatus, 709, 97; rotatus, 707. Encrinuridae indet., 598, 82, 83. Encrinuroides, 593; obesus, 595, 82; sp., 594, 82. England: Devonian Atrypidae, 358 ; Keuper miospores, 294; Permian miospores, 322; Westphalian mega- spores, 82; Siurian graptolites, 247; new Devonian crinoid, 11 ; freshwater Bathonian ostracods, 749; Upper Aptian ostracods, 375. Enzonalasporites vigens, 302, 37. Eomarginifera, 60; tenuimontis, 61, 10. Eoscaphites, 404; circularis, 404, 53; subcircularis, 407, 53-55 ; tenuicostatus, 410, 53. Eucytherura ornata, 378, 48. F Falcisporites zapfei, 345, 40. Felix, C. J. Neogene Tasmanites and leiospheres from southern Louisiana, U.S.A., 16. Fenestella allynensis, 58, 13. Foraminifera: Holocene from Wales, 27. Ford, T. D. The palaeoecology of the Goniatite Bed at Cowlow Nick, Castleton, Derbyshire, 186. G Gastropods: turritellid classification, 666. Geminospora, 695; spinosa, 691, 94; svalbardiae, 696, 94; tuberculata, 696, 94. Girtypecten, 75; ?sp., 16, 12. Globochaete alpina, 394, 52. Grandispora, 733; diamphida, 733, 103; inculta, 734, 103. Granulatisporites muninensis, 693, 94. Graptolithina: development of dicellograptid, 41; development of Lasiograptus, 272 ; new Silurian from England, 247. Gruenewaldtia latilinguis, 369, 47. H Halimeda, 200; praemonilis, 23; sp., 23. Hallam, A. Environmental causes of stunting in living and fossil marine benthonic invertebrates, 132. Haynes, J., See Adams, T. D. Hemiarges sp., 600, 83.' Hexagonaria, 543; approximans, 543; a. cribellum, 544, 74, 75. Hibbert, F. A., and Eggert, D. A. A new calamitalean INDEX 765 cone from the Middle Pennsylvanian of Southern Illinois, 681. Hibbertia whittingtoni, 586, 81. Hoegisporis, 39; lenticulifera, 9; uniformis, 39, 9. Holdsworth, B. K. The Namurian goniatite Nuculo- ceras stellarum (Bisat), 226. Homotrypa oweni, 5, 2, 3. Hypodicranotus sp., 582, 80. Hystricosporites, 698 ; corystus, 701 ; mitratus, 699, 95 ; monosaccus, 700, 96; porcatus, 699, 95; porrectus, 698, 95. I Jlionia prisca, 237, 27. Illaenus sp., 584, 80. Illinites, 341 ; delasaucei, 342, 44 ; klausi, 342, 40, 43 ; tectus, 342, 41. Iocrinus brithdirensis, 355, 45. Ireland: Oedicybele from Kildare Limestone, 1; new fenestrate bryozoan from Fermanagh, 478. Isorthis, 462; amplificata, 469, 61; clivosa, 471, 61, 62; orbicularis, 465, 63-65; scute formis, 474, 65; s. uskensis, 416, 65; slitensis, 467, 62. Isotelus sp., 583, 80. J James, J. The development of a dicellograptid from the Balclatchie Shales Of Laggan Burn, 41. Jeffries, R. P. S., and Minton, P. The mode of life of two Jurassic species of ‘ Posidonia ’ (Bivalvia), 156. Jull, R. K. Corallum increase in Lithostrotion, 204. Jurassic: freshwater Bathonian ostracods, 749; microfossils from Hautes-Alpes, 391 ; mode of life of ‘ Posidonia ’, 156; new ostracod genus, 572; sensory spines in brachiopod, 604. K Kaye, P., and Barker, D. Ostracoda from the Sutterby Marl (U. Aptian) of South Lincolnshire, 375. Kilenyi, T. I. Oertliana, a new ostracod genus from the Upper Jurassic of north-west Europe, 572. Kitakamithyris triseptata, 68, 11. Klausipollenites, 311, 344; devolvens, 311, 37; schaubergeri, 345, 43. Krausella minuta, 316, 48. L Labiisporites granulatus, 307, 344, 39, 41, 44. Laevigatisporites, 87 ; glabratus, 87, 14. Lagenicula, 91; irregularis, 95, 15; perverrucata, 93, 15; verrucata, 94, 15; verrurugosa, 91, 14, 15. Lagenoisporites, 90; rugosus, 91, 14. Lasiograptus harknessi, 274. Leiosphaeridia, 24; plicata, 24, 8; ralla, 25, 5. Leiotriletes, 690; pagius, 691, 94; parvus, 691; pyramidalis, 691. Limnocythere, 751; sp. A, 751, 109; sp. B, 752, 109. Lithostrotion, 204; arundineum, 212; columnare, 220; junceum, 208; cf. martini, 215; minus, 219; cf. portlocki, 218; sp., 222. Lonchodomus pernix, 590, 82. Lueckisporites, 305, 331; triassicus, 305, 38; virkkiae, 331, 43. ?Lycophoria rhenana, 371, 47. Lycopod: fertile from Carboniferous of Scotland, 281. Lycospora, 712; culpa, 713, 98. M Macrocypris parva, 375. Mammalia: European Proviverrini, 638. Mansuyphyllum, 525; bellense, 526, 72; catombalense, 529, 12, 73; parvulum, 521, 72, 73; sp. A, 531, 72; sp. B, 532, 73; sp. C, 533, 73. Marginatid patersonensis, 63, 10. McAlester, A. L. Systematics, affinities, and life habits of Babinka, a transitional Ordovician lucinoid bivalve, 231. Medd, A. W. Dionella gen. nov. (Superfamily Mem- braniporacea) from the Upper Cretaceous of Europe, 492. Mesotaphraspis sp., 589, 81. Metasinopa fraasi, 660. Miacis? macintyri, 656. Miller, T. G. Time in stratigraphy, 113. Mimatrypa, 361; desquamata, 364, 46, 47; cf. insquamosa, 47. Minton, P. See Jeffries, R. P. S. Monoceratina tricuspidata, 382, 48. Monoclimacis, 249; flumendosae, 252; fl. kingi, 253; griestonensis nicoli, 249, 30 ;haupti, 255 ; linnarssonii, 250, 30; shottoni, 254, 30. Monograptus, 261 ; danbyi, 266, 30; firmus sedberghen- sis, 262, 31; flexilis belophorus, 268, 29; minimus cautleyensis, 266, 30; radotinus inclinatus, 264; simulatus, 265, 29; sp. A, 269. N Neocythere ( Physocythere ) cf. bordeti, 383, 49. New South Wales: Devonian corals, 518; Lower Carboniferous fauna, 54. Nikitinsporites spitsbergensis, 741 , 108. Nileus sp., 583, 80. North America: Devonian corals, 618; calamitalean cone from Illinois, 681 ; Neogene Tasmanites from Louisiana, 16. Nuculoceras, 226; stellarum, 221, 25. Nuskoisporites, 329; dulhuntyi, 330, 40; cf. rotatus, 330, 40. O Oedicybele kildarensis, 1,1. Oertliana, 512; kimmeridgensis, 573, 79; sp. 1, 574, 79; sp. 2, 515; sp. 3, 516. Ordovician: Babinka, 231; bryozoa from Shropshire, 5 ; development of dicellograptid, 41 ; Lasiograptus harknessi, 272; new crinoid from Wales, 355; trilobites from Girvan, 577. Orthonotacythere, 381; inversa tuber culata, 381, 48; sp. B, 381, 49. Osmundacidites alpinus, 299, 37. 766 INDEX Ostracoda: Aptian, 375; freshwater Bathonian, 749; new Jurassic genus, 572. Otarion sp., 586. Ovalipollis, 306; breviformis, 307, 39; avails, 307, 39. Oxroadia gracilis, 281, 32-34. cf. Oxyaena sp., 658. P Palaeoecology: causes of stunting, 132; Goniatite Bed, Derbyshire, 186; mode of life of ‘Posidonia\ 156; life habits of Babinka, 231. Paracalamostachys cartervillei, 682, 93. Paradisphyllum, 537; liarundinetum, 538, 74. Patinasporites cf. densus, 302, 37. Pedder, A. E. H. Some North American species of the Devonian tetracoral Smithiphyllum, 618. Peneckiella, 555; boreensis, 562, 78; mesa, 556, 75, 77, 78; minor kunthi, 563, 78; teicherti, 561, 77. Perforosporites sp., 707, 97. Perisaccus, 328 ; granulosus, 329, 42 ; laciniatus, 329, 42. Permian: saccate and monoculcate miospores, 322. Pernopecten trevallynensis, 77, 12. Perotrilites, 730; ergatus, 731, 102; eximius, 731, 102; pannosus, 732, 102. Phillipsastrea, 564; oculoides, 565, 73. Platysaccus, 311, 347; radialis, 347, 43; sp., 311, 37. Pontocyprella rara, 376, 49. ‘ Posidonia ’ radiata, 179, 19. Pothocites, 107 ; grantonii, 108, 18. Potoniesporites, 323 ; novicus, 326, 40, 44. Pristiograptus, 256 ; auctus, 260, 31 ; dubius pseudolatus, 257, 31; meneghini, 258, 29; pseudodubius, 259, 29; watneyae, 256, 30; welchae, 261, 30. Prodissopsalis, 643; eocaenicus, 654; phonax, 651; theriodis, 651; sp., 655, 658. Prolecanites sp., 74, 12. Propterodon morrisi, 660. Protocythere, 383; derooi, 383, 50; mertensi langtonen- sis, 383, 49. Protohaploxypinus, 336; chaloneri, 337, 42; jacobii, 337, 41; microcorpus, 338, 41; cf. samoilovichii, 337, 42. Prototomus cf. sp., 659; torvidus, 646. Protozoa: Holocene foraminifera, 27; tintinnina from Alps, 393. Proviverra, 642; gracilis, 654; minor, 658; typica, 646. Pteridophyta: calami talean cones from Scotland, 107; calamitalean cone from U.S.A., 681 ; fertile lycopod from Scotland, 281. Ptilofenestella, 490; carrickensis, 491, 66. Punctatisporites, 692 ; flavus, 692; glaber, 692; laevigatus, 692. Q Quantoxocrinus, 12; ussheri, 13, 4. Quinquecosta williamsi, 595, 83. R Raistrickia aratra, 701, 96. Raymondaspis sp., 583, 80. Remopleurides, 581; sp. A, 581, 80; sp. B, 581, 80; sp. C, 582, 80. Retialetes sp., 741, 106. Reticulatisporites, 705; emsiensis, 705, 97; sp., 796, 97. Rhabdosporites, 737; cymatilus, 738, 104; scamnus, 737, 104. Rickards, R. B. New Silurian graptolites from the Howgill Fells (Northern England), 247. Rickards, R. B., and Bulman, O. M. B. The develop- ment of Lasiograptus harknessi (Nicholson 1867), 272. Roberts, J. A Lower Carboniferous fauna from Trevallyn, New South Wales, 54. Ross, J. R. P. Homotrypa and Amplexopora? from the Caradoc Series, Shropshire, 5. Rudwick, M. J. S. Sensory spines in the Jurassic brachiopod Acanthothiris, 604. S Saccocoma, 391, 51, 52. Salopina, 457 ; lunata, 459, 64, 65. Samarisporites, 714; hesperus, 715, 98; inusitatus, 717, 99; praetervisus, 714, 98; senotus, 714, 98; tri- angulatus, 716, 99. Scaphites, 411 ; ( Hyposcaphites ) stephanoceroides, 436, 59 ; ( Metascaphites ), 427 ; thomasi, 58 ; ( Otoscaphites ), 429; bladenensis, 432, 58; puerculus, 434, 59; ( Pteroscaphites ), 435 ; minutus, 58; ( Scaphites ), 412; collignoni, 427, 57; equalis, 417, 56; hilli, 421; hugardianus, 423, 54, 57; meriani, 426, 54, 57; obliquus, 415, 56; similaris, 422; simplex, 412, 54, 55; yonekurai, 421, 56. Schuleridea derooi, 377, 49. Scotland: Carboniferous algae, 192; new fertile lycopod, 281; Girvan trilobites, 577; dicellograptid from Girvan, 41 . Setosisporites pilatus, 88, 14, 15. Silurian: Ludlovian brachiopods, 454; new graphto- lites, 247. Smithiphyllum, 618; belanskii, 623, 88, 89; imperfec- tum, 622, 88, 89; kindlei, 625, 88, 89; whitt akeri, 626, 88, 89. Sphaerexochus sp., 591, 82. Sphaerocoryphe sp., 592, 82. Spinner, E. Westphalian D megaspores from the Forest of Dean Coalfield, England, 82. Spitsbergen: Lower and Middle Devonian spor-es, 687. Spores: Devonian from Spitsbergen, 687 ; Hoegisporis, 39; megaspores from Forest of Dean Coalfield, 82; Permian miospores, 322; Keuper miospores, 294; Tasmanites and leiospheres, 16. Stenoscisma, 72; laevis, 73, 13. Stenozonotriletes, 711; furtivus, 711, 98; insessus, 711, 98; sp., 712, 98. ?Stillina cf. fluitans, 380, 48. Stomiosphaera minutissima, 395, 51. Stratigraphy: time in, 113. Striatoabietites, 339; richteri, 340, 43. Striatopodocarpites, 338; antiquus, 339, 42; cancellatus, 339, 43 ; fusus, 339, 41. Strusz, D. L. Disphyllidae and Phacellophyllidae from INDEX the Devonian Garra Formation of New South Wales, 518. Succinctisporites, 303 ; grandior, 303, 37 ; radialis, 303, 35. T Tabulophyllum, 620. Taeniaesporites, 333; albertae, 334, 44; angulistriatus, 334, 44; bilobus, 335, 41; labdacus, 334, 44; noviaulensis, 333, 42; novimundi, 334, 44; nubilus, 336, 41. Tasmanites, 19; balteus, 22, 8; corrugatus, 20, 5; fissura, 19, 5 ; fulgidus, 21, 7;porosus, 19, 5; usitatus, 20, 6; validus, 22, 8. Tavener-Smith, R. A new fenestrate bryozoan from the Lower Carboniferous of County Fermanagh, 478. Temple, J. T. The trilobite genus Oedicybele from the Kildare Limestone (Upper Ordovician) of Eire, 1. Tertiary: apical development of gastropods, 666; Eocene mammals, 638; Holocene foraminifera, 27; Neogene Tasmanites, 16. Theriosynoecum kirtlingtonense, 754, 110, 111. Tholisporites ancylus, 724, 101. Timiriasevia, 756; mackerrowi, 756, 111; sp. A, 758, 111. Tintinnina, 393. Tintinnopsella, 393; carpathica, 52; sp., 51. Toemquistia, 588; sp. A, 588, 81; sp. B, 588, 81. Triangulatisporites regalis, 98, 14, 15. Triassic: Keuper miospores from England, 294. Trilei tes oxfordiensis, 693. Triletes subpalaeocristatus, 288, 34. Trilobita: Oedicybele from Eire, 1 ; Ordovician from Girvan, 577. Trinodus, 578; doulargensis, 578, 80; sp., 579, 80. Tripp, R. P. Trilobites from the Albany Division (Ordovician) of the Girvan District, Ayrshire, 577. Tuberculatisporites, 87 ; brevispiculus, 87, 14. 767 Turner, J. Upper Jurassic and Lower Cretaceous microfossils from the Hautes-Alpes, 391. Turritella, 666; alveata, 671 ; apita, 672; arenicola, 673, 92; a. branneri, 674; a. danvillensis, 672; creola, 671 ; temblorensis, 92. Tytthodiscus calif orniensis, 23, 6. V Valvisporites, 96 ; auritus, 96, 17 ; migrozonales, 97, 17 ; sofiaense, 97, 16. Van Valen, L. Some European Proviverrini (Mam- malia, Deltatheridia), 638. Verrucosisporites, 296; contactus, 297, 35; morulae, 296, 35. Vertebrata: European Proviverrini, 638; skull and palate of Dipnorhynchus, 634. Vestigisporites minutus, 327, 40. Vittatina, 340; hiltonensis, 341. W Wales: Holocene foraminifera, 27; new Ordovician crinoid, 355. Walmsley, V. G. Isorihis and Salopina (Brachiopoda) in the Ludlovian of the Welsh Borderland, 454. Webby, B. D. Quantoxocrinus, a new Devonian inadunate crinoid from West Somerset, 11. Welsh Borders : Ludlovian brachiopods, 454; Caradoc bryozoa, 5. Werriea, 59; australis, 60, 11. Wiedmann, J. Origin, limits, and systematic position of Scaphites, 397. Worthoceras, 439; gibbosum, 60; platydorsum, 59; rochatianum, 60; vermiculum, 59, 60; worthense, 59. Z Zelolasma gemmiforme, 534, 72. Zonalesporites, 99; brasserti, 100, 16; dentatus, 100, 17; rotatus, 101, 17. Zuluscaphites, 443. THE PALAEONTOLOGICAL ASSOCIATION COUNCIL 1965-6 President fDr. L. R. Cox, British Museum (Natural History), London Vice-Presidents Dr. W. S. McKerrow, University Museum, Oxford Professor F. H. T. Rhodes, University College, Swansea Treasurer Dr. C. Downie, Department of Geology, The University, Mappin Street, Sheffield 1 Secretary Dr. C. H. Holland, Department of Geology, Bedford College, London, N.W. 1 Editors Mr. N. F. Hughes, Sedgwick Museum, Cambridge Dr. Gwyn Thomas, Department of Geology, Imperial College of Science, London, S.W. 7 Dr. I. Strachan, Department of Geology, The University, Birmingham, 15 Dr. M. R. House, University Museum, Oxford Other members of Council Dr. C. G. Adams, British Museum (Natural History), London Professor P. M. Butler, Royal Holloway College, Surrey Dr. W. J. Clarke, British Petroleum Company, Sunbury-on-Thames Dr. G. Y. Craig, The University, Edinburgh Dr. T. D. Ford, The University, Leicester Dr. B. M. Funnell, Sedgwick Museum, Cambridge Dr. J. M. Hancock, King’s College, London Dr. G. A. L. Johnson, The University, Durham Dr. F. A. Middlemiss, Queen Mary College, London Mr. M. Mitchell, Geological Survey and Museum, London Dr. W. D. I. Rolfe, Hunterian Museum, The University, Glasgow Dr. A. J. Rowell, The University, Nottingham Professor Scott Simpson, The University, Exeter Dr. L. B. H. Tarlo, The University, Reading Dr. H. Dighton Thomas, British Museum (Natural History), London Overseas Representatives Australia: Professor Dorothy Hill, Department of Geology, University of Queensland, Brisbane Canada: Dr. D. J. McLaren, Geological Survey of Canada, Department of Mines and Technical Surveys, Ottawa India: Professor M. R. Sahni, 98 The Mall, Lucknow (U.P.), India New Zealand: Dr. C. A. Fleming, New Zealand Geological Survey, P.O. Box 368, Lower Hutt West Indies and Central America: Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad, Inc., Pointe-a-Pierre, Trinidad, West Indies Eastern U.S.A.: Professor H. B. Whittington, Museum of Comparative Zoology, Harvard Univer- sity, Cambridge 38, Mass. Western U.S.A.: Professor J. Wyatt Durham, Department of Paleontology, University of California, Berkeley 4, Calif. t Dr. Cox died in August. PALAEONTOLOGY VOLUME 8 • PART 4 CONTENTS Trilobites from the Albany division (Ordovician) of the Girvan district, Ayrshire. By R. p. tripp 577 Sensory spines in the Jurassic brachiopod Acanthothiris. By m. j. s. rudwick 604 Some North American species of the Devonian tetracoral Smithiphyllum. By A. E. H. PEDDER 618 Arcoscalpellum comptum (Withers), a species of cirripede new to the Gault. By j. s. H. COLLINS 629 An almost complete skull roof and palate of the dipnoan Dipnorhynchus sussmilchi (Etheridge). By k. s. w. Campbell 634 Some European Proviverrini (Mammalia, Deltatheridia). By l. van valen 638 Apical development in turritellid classification with a description of Cristi- spira pugetensis gen. et sp. nov. By r. c. allison 666 A new calamitalean cone from the Middle Pennsylvanian of southern Illinois. By f. a. hibbert and d. a. eggert 681 Lower and Middle Devonian spores of North and Central Vestspitsbergen. By K. c. ALLEN 687 Freshwater ostracods from the Bathonian of Oxfordshire. By r. h. bate 749 PRINTED IN GREAT BRITAIN AT THE UNIVERSITY PRESS, OXFORD BY VIVIAN RIDLER, PRINTER TO THE .UNIVERSITY X to o | g = .* - ' i § i riTUTION ^NOIinillSNI_NVlNOSHlMS S3I y VB Bll\lBRARI ES^SMITHSONIAN^INSTITUTION ^NOIl yvaan LIBRARIES SMITHSONIAN INSTITUTION NQIinillSNI NVINOSH1I1NS S3 I y \/y a 1 1 L I B|i > 33 \Z — X Z *Wr rn NTUTIOn" NOlinillSNI NVINOSH1MS S3iyvy rn ^S. co I n U B RAR I ES SMITHSONIAN ~ INSTITUTION ^ N0I1 z .v,. co X 1 = 1 CO \ ^ g ^ | %■ s - g > yvyan_LiBRARi es Smithsonian _ institution NoiinmsNi nvinoshiiws shi yvyan lib f o Z 'r o 5 •ITUTION^ NOIinillSNI^NVINOSHllWS^S3 I y vy a n^LI B RAR I eszsmithsonian"'institution^noui 2 2 jvy a n li brar i es^smithsonian institution NoiiniusNrNviNOSHiiws^ss 1 y vy a n “li br ~~ - * — — 7 fn z to Z \ ^ ^ 2 v >' s X^vos^x >• _ TUT!ON w NOIinillSNI_NVINOSHlllAIS^S3 i y vy a ll_LI BRAR i es^smithsonian^institution ^NOlil O ^ O Xjgvoi^X r ' o JVaan libraries smithsonian institution NoixniusNi^NviNosHims^ssiyvyan^LiB "• 2 X r- z r- z ^ JJJ w xjomsjx m z xgirsgx m ITUTION NOlinillSNI NVINOSH1IWS S3 i y vy a I*|“l I B r A R I ES^SMITHSONIAN-INSTITUTION^ NOIll 'V. 9 ^ ^ zl Z «/. to 7 to ivy a n_ li b rar i es smithsonian_ institution NoiimiiSNi NviNOSHiiws^ss i y vy a I "|2 LI B SMITHSONIAN INSIliuiiwn o > '" ~ ' i >' 2 NVIN0SH1IWSOTS3 I uVa a nZLI B RAR I ES^SMITHSONIAN INSTITUTION NOIinillSNI_NVINOSHlllNS _s 5S =} § v: o _ _ _ 2 SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSHIIINS S3IHVaail LIBRARIES SMITHSONIAN I * r- .. z _ Z E - w . fE c/> EE c/> nvinoshiiws S3 1 a va an libraries Smithsonian institution,,, NoiiniusNi nvinoshuins s CO 2 CO 2 ... CO 2 to ^ < xSX .. s . . s .< s: CO wid. xs- CO (f®- in N co ;v#^ mW to lifF. ^Is to nvinoshiiins S3 lava an libraries Smithsonian institution NouniiiSNi^NViNOSHiiiNS^ x r* 2i x ^ r-i T- — . rx _ .^TiTT^x O ^x \x CO 33 /§j< > [SJ 33 CO 2 f CO SMITHSONIAN INSTITUTION^NOIlfUllSNI NVINOSHIIINS S3 I a Va an LI B RAR I ES SMITHSONIAN 2 > CO Z <0 Z ^ 5 \X>' > ' 5 X£vos^ >' S ' ' VP > 5 2 CO x co A Z (O NVINOSHIIINS S3 lava an LIBRARIES SMITHSONIAN INSTITUTION NOIlfUllSNI_NVINOSHlllNS CO x to - Wj =± 5* -I I .. .. c / D • ' W' ~ O Xjy^jg/ _ O Z -J Z _J Z SMITHSONIAN INSTITUTION NOIlfUllSNI NVINOSHIIWS S3 I a Va 3 IT LIBRARIES SMITHSONIAN * r- C ^ 33 > |i^ ^31 £ If® f3|j > m ^ocy <0 NVINOSHIIINS S3iavaan_ LIBRARIES SMITHSONIAN_INSTITUTlON" NOIlfUllSNI NVINOSHIIINS CO 2 tO 2 V- to Z CO <^s° X 2 _ , < 1“ /^sv'iTx ^ .^ilxL'x iE . « - ^x - ™ &'i i W i#fPI 2 v > XOuasfx 2 ~ > ■%; INSTITUTION NOIlfUllSNI NVINOSHlIlNS^Sa I a Va a IT LIBRARIES SMITHSONIAN_ CO — to 5 to 2 /TaSO\7>v LU z a&CS' . UJ xx UJ //;