CO TION z < CO co — CO £ CO v £ CO NOlinillSNI NVIN0SH1HNS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION z co isf i i i § O t ^ Nouniiii _ > LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1I WS^S 3 I B V B B y - m - vv - CO X CO _ TION NOlinillSNI NVIN0SH1IIAIS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOlinilll co z co z co 2 co < /VVx 2 *. y i ^coc^/ > 2 ^ ‘ Z C O Z CO * Z cO v z 811 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IINS S3iavaan LIBRAR ■7 ^ 2 ^ -7 vX ^ 2 - ” /Ki cr < o N^Viy “ ■'4F' o NS£dC^ “ '^jn^y 5 V " 5 N^T5v riON^ NOlinillSNI ^NVINOSHII INS ^S3 I a VB a 13 L I B R A R I ES"2 SMITHSONIAN^ INSTITUTION ^ NOlinillJ z r- v z r- z ~ ^ i= f— viVuarw ^ v^.^y i= ^ co *■ m x. m — — u/ £ co _ an LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1I1NS S3iavaan LIBRAR X CO Z CO z - .^V^OX 5 < , § 00 Iz to _ (/> \ tl CO — to 0SH1IINS S3IHVyail LIBRARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3ld\ IT LIBRARIES SMITHSONIAN ~ CO HSONIAN INSTITUTION NOliniliSNI NVINOSH1IWS SBIHVHa 2 ^ _____ --?» x ^ O 5 “ X^oiis^y o “ W' O 0SH1IWS S3iavyan LI B RAR I ES^SMITHSONIAN^INSTITUTION Z NOlinillSNI^NVINOSHlIWS S3IU\ m xo 7 t '^4^ ? % y >' 2 VP ■*> v z to N 0SH1IINS S3 I ava a IT LIBRARIES SMITHSONIAN INSTITUTION NOliniliSNI _N VI N0SH1UAJS S 3 I a 1 O O 'sivosv^ s 2 _J ZZL I — J HSONIAN INSTITUTION NOliniliSNI NVIN0SH1IWS S3iaVHail LIBRARIES SMITHSONIAN INSTIT z r- 2 r- z _ ° O xC^vTT'x jjj 5 m . - 73 > / 73 rn to ir to \ z ^ X to lOSHiiiNS saiavaan libraries Smithsonian institution NOliniliSNI nvinoshiiws sbih' 2! V////4W r > 5 VoNoty' Z to Z T LIBRARIES SMITHSONIAN INSTIT to HSONIAN ^ INSTITUTION^NOliniliSNI^NVINOSHlIlNS^S 3 I B V d 3 to “ to to o Dt^y _ NQQjusgy Q “ XOQns^y Q 0SH1IINS S3iyvaan LI BRAR I ES^SMITHSONIAN'-’lNSTITUTION^NOlinillSNI^NVINOSHimS S3IB\ Z r~ z r- z r^x, z 2 m >&v - m x^os>iy y ^vam^ m to to _ to HSONIAN INSTITUTION NOliniliSNI NVIN0SH1IWS S3IBVBan LIBRARIES SMITHSONIAN INSTIT z r to z to z CO Z Palaeontology VOLUME 17 PART 3 OCTOBER 1974 Published by The Palaeontological Association London Price £8 THE PALAEONTOLOGICAL ASSOCIATION The Association publishes Palaeontology and Special Papers in Palaeontology. Details of member- ship and subscription rates may be found inside the back cover. PALAEONTOLOGY The journal Palaeontology is devoted to the publication of papers on all aspects of palaeontology. Review articles are particularly welcome, and short papers can often be published rapidly. A high standard of illustration is a feature of the journal. Four parts are published each year and are sent free to all members of the Association. SPECIAL PAPERS IN PALAEONTOLOGY This is a series of substantial separate works. Members may subscribe to the Series; alternatively, Ordinary and Student members only may obtain individual copies at reduced rates. The following Special Papers are available : 1. (for 1967): Miospores in the Coal Seams of the Carboniferous of Great Britain, by a. h. v. smith and m. a. butterworth. 324 pp., 72 text-figs., 27 plates. Price £8 (U.S. $22.00), post free. 2. (for 1968): Evolution of the Shell Structure of Articulate Brachiopods, by A. williams. 55 pp ., 27 text- figs., 24 plates. Price £5 (U.S. $13.00). 3. (for 1968): Upper Maestrichtian Radiolaria of California, by Helen p. foreman. 82 pp., 8 plates. Price £3 (U.S. £8.00). 4. (for 1969): Lower Turonian Ammonites from Israel, by R. freund and m. raab. 83 pp., 15 text-figs., 10 plates. Price £3 (U.S. $8.00). 5. (for 1969) : Chitinozoa from the Ordovician Viola and Fernvale Limestones of the Arbuckle Moun- tains, Oklahoma, by w. a. m. jenkins. 44 pp., 10 text-figs., 9 plates. Price £2 (U.S. $5.00). 6. (for 1969): Ammonoidea from the Mata Series (Santonian-Maastrichtian) of New Zealand, by r. a. Henderson. 82 pp., 13 text-figs., 15 plates. Price £3 (U.S. $8.00). 7. (for 1970): Shell Structure of the Craniacea and other Calcareous Inarticulate Brachiopoda, by a. williams and a. d. wright. 51 pp., 17 text-figs., 15 plates. Price £1-50 (U.S. $4.00). 8. (for 1970): Cenomanian Ammonites from Southern England, by w. J. Kennedy. 272 pp., 5 tables, 64 plates. Price £8 (U.S. $22.00). 9. (for 1971): Fish from the Freshwater Lower Cretaceous of Victoria, Australia, with Comments on the Palaeo-environment, by m. waldman. 130 pp., 37 text-figs., 18 plates. Price £5 (U.S. $13.00). 10. (for 1971): Upper Cretaceous Ostracoda from the Carnarvon Basin, Western Australia, by R. h. bate. 148 pp., 43 text-figs., 27 plates. Price £5 (U.S. $13.00). 11. (for 1972): Stromatolites and the Biostratigraphy of the Australian Precambrian and Cambrian, by m. R. Walter. 268 pp.. 55 text-figs., 34 plates. Price £10 (U.S. $26.00). 12. (for 1973): Organisms and Continents through Time. A Symposium Volume of 23 papers edited by n. f. hughes. 340 pp., 132 text-figs. Price £10 (U.S. $26.00) (published with the Systematics Asso- ciation). 13. (for 1974): Graptolite studies in honour of O. M. B. Bulman. Edited by r. b. rickards, d. e. jackson, and c. p. hughes. 261 pp., 26 plates. Price £10 (U.S. $26.00). 14. (for 1974): Palaeogene foraminiferida and palaeoecology, Hampshire and Paris Basins and the English Channel, by J. w. Murray and c. a. wright. 45 text-figs., 20 plates. In press. SUBMISSION OF PAPERS Typescripts 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 The Secretary, P.A. Publications Committee, Department of Geology, Sedgwick Museum, Downing Street, Cambridge, CB2 3EQ, England, who will supply detailed instructions for authors on request (these are published in Palaeontology, 15, pp. 676-681). © The Palaeontological Association, 1974 Cover: Dactylioceras commune (J. Sowerby), Upper Lias, Jurassic, Whitby, Yorks. In collection of Professor J. E. Hemingway. The 'head' was carved by a Whitby jet worker to illustrate the traditional story of St. Hilda of Whitby (614-80), who turned snakes into stones by the power of prayer. THE UPPER PALAEOZOIC TETRACORAL GENERA LOPHOPHYLLIDIUM AND TIMORPHYLLUM by JERZY FEDOROWSKI Abstract. The morphology of Lophophyllidium- like genera and their relationships are described and discussed. In early growth stages, Lophophyllidium has a zaphrentoid arrangement of septa with an elongate counter septum ; the genus is characterized by a pseudocolumella that is extremely variable in morphology, both ontogenetically and between individuals; septal microstructure is trabecular. The synonymy of Lophophyllidium includes Sinophyllum Grabau, Malonophyllum Okulitch and Albritton, Stereostylus Jeffords, Agarikophyllum Fomitshev, and Khmero- phyllum Fontaine; each was originally distinguished on the basis of pseudocolumella morphology. Timorophyllum differs from Lophophyllidium in ontogeny and microstructure; nracrostructural similarities reflect homeomorphy. Specimens having a pseudocolumella are the most common components of the Permian tetracoral fauna from the Glass Mountains, Texas. The extraordinary individual variability of some species of Lophophyllidium and a study of the type material of the genera Lophophyllidium, Stereostylus, Lophamplexus, Malonophyllum , and topotypes of Timorphyllum led to the conclusion that the first four genera are synonyms. A complete study of phylogeny and the species of the discussed genera is beyond the scope of this paper. The main purpose is to show the variability of some skeletal elements of this group of corals and to indicate that some of these elements are not generic characters in the group discussed. Only papers about Lophophyllidium- like or Timorphyllum corals that include new or controversial opinions are discussed. Some opinions of other authors, which coincide with mine are not discussed. Species descriptions will be included in a mono- graphic study of Glass Mountains tetracorals, now in preparation. Figured material is housed in the United States National Museum (USNM) and the Kansas University Museum (KUM). THE SYSTEMATICS OF LOPHOPHYLLIDIUM Genus lophophyllidium Grabau, 1928 Type species. Cyathaxonia prolifera McChesney, 1860. Synonyms. Cyathaxonia McChesney, 1860 e.p., non Michelin in Gervais, 1840. Lophophyilum White, 1875, 1877, 1884; Martin, 1881; Keyes, 1894; Stuckenberg, 1904 e.p.; Duerden, 1906; Lorenz, 1906; Brown, 1909; Grabau, 1922; Morningstar, 1922; Soshkina, 1928 e.p.; Kelly, 1930; Huang, 1932; Yoh and Huang, 1932; Merla, 1934; ?Vojnovsky-Krieger, 1934, non Milne- Edwards and Haime, 1850. Sinophyllum Grabau, 1928. Malonophyllum Okulitch and Albritton, 1937. Soshkineophyllum Moore and Jeffords, 1941, non Grabau, 1928. Stereostylus Jeffords, 1947. Agarikophyllum Fomitshev, 1953. [Palaeontology, Vol. 17, Part 3, 1974, pp. 441-473, pis. 60 70.] A 442 PALAEONTOLOGY, VOLUME 17 Khmerophyllum Fontaine, 1961. Rotiphyllum Ivanovsky, 1967, non Hudson, 1942. Stratigraphic and geographic range. Lower Carboniferous to Upper Permian, cosmopolitan. Diagnosis. Solitary corals without dissepimentarium ; ontogeny zaphrentoidal, with cardinal septum unshortened in youngest stages ; axial end of counter septum is the main component of pseudocolumella but either axial tabellae or septal lamellae or both may be additional elements; minor septa extremely variable in length; growth lines in septa very steeply arranged; trabeculae very short, ordinarily crossing two growth lines only. Review of the synonymized genera Lophophyllum Milne-Edwards and Haime, 1850 (type species: L. konincki Milne-Edwards and Haime, 1850). The taxonomic position of this genus and its morphology was preliminarily discussed by Lecompte (1955) who considered it to be a distinct genus which differs from Koninckophyllum in lacking dissepiments. I agree with Lecompte’s conclusions and disagree with Carruthers’s (1913) concept of this genus which was based on erroneously identified material. Consequently, all species having a dissepimentarium were excluded from this study. Many of the species assigned to Lophophyllum were discussed by Schouppe and Stacul (1955). Most of the other species that lack a dissepimentarium and have a pseudocolumella belong to Lophophyllidium (see synonymy in Schouppe and Stacul). Lophophyllidium Grabau, 1928 (type species: Cyathaxonia prolifera McChesney, 1860). Grabau (1928, p. 99) following Carruthers (1913) wrote: ‘The corals of Lophophyllum proliferum type cannot be con- sidered congeneric with Lophophyllum tortuosum.' The correctness of this position is not in doubt, although Huang (1932, pp. 22, 23) disagreed with it. Huang, however, presented mostly historical and nomenclatural arguments. The ontogenetic development of C. prolifera was described by Duerden (1906). A neotype for this species was chosen by Jeffords (1942) but this neotype and most of the original syntypes are probably lost. The last two syntypes have been restudied and one has been illustrated (PI. 60, fig. 1 ; PI. 62, fig. 1 a-c\ PI. 63, figs. 4, 5). Moore and Jeffords (1945) established the family Lophophyllidiidae. Both genus and family (sometimes as a subfamily) are generally accepted. The most complete synonymy of the genus was given by Scrutton (1971), who agreed that the structure of the pseudocolumella was not taxonomically significant. Sinophyllum Grabau, 1928 (type species: Lophophyllum pendulum Grabau, 1922). Grabau (1928, p. 100) proposed the new generic name for corals having septa accelerated strongly in counter quadrants and a ‘pseudocolumella formed by the excessive thickening of the original pali-columella, which still remains as the central lamina. In section the pali-columella shows a zone of radial structure or an irregular series of rod shaped bodies.’ He also indicated that the axial end of the counter septum wraps around the pseudo- columella in some stages of growth (Grabau 1928, pi. IV, figs, lc, 3c). The last feature is very common in many of the later described species of Lophophyllidium- like corals. Huang (1932) and Wang (1947) studied topotypes of the American Cyathaxonia prolifera and compared the structure of its pseudocolumella with that of Lophophyllum pendulum Grabau, 1922. Both authors did not see any differences between these EXPLANATION OF PLATE 60 Fig. 1. Lophophyllidium proliferum (McChesney, 1860). Syntype KUM 52878 chosen by Jeffords 1942, x 3. Fig. 2. Lophophyllidium sp. nov. C. USNM 189807, Glass Mountains, Texas. Road Canyon Fm., Upper Leonard, x 3. Figs. 3-6. Lophophyllidium sp. nov. A. Same locality and age, different stages of growth, x 3. 3, USNM 189808, early neanic stage. Regular arrangement of the major septa. 4, USNM 189809, neanic stage. Regular arrangement of septa. 5, USNM 1 898 1 0, late neanic stage. Major septa started to be differentiated in length. 6, Holotype USNM 189811, ephebic stage with differentiated length of the major septa. Fig. 7. Lophophyllidium sp. USNM 189812. Same locality and age. Specimen without tabulae. Continuation from the major septa into septal lamellae and outstanding ridge of median lamella on the top of the pseudocolumella is visible, x 5. PLATE 60 . FEDOROWSKI, Lophophyllidium 444 PALAEONTOLOGY, VOLUME 17 structures. Huang (1932) proposed to use the name Sinophyllum as a subgenus of Lophophyllum , while Wang (1947) synonymized Sinophyllum with Lophophyllidium. Schouppe and Stacul (1955) discussed the structure of the genera Lophophyllidium and Sinophyllum , but made no final decision on their relationships. They used the name Lophophyllidium and also gave a list of synonyms of the genus Sinophyllum. Some authors (e.g. Jeffords 1942, 1947; Wang 1947, 1950; Minato 1955) thought Sinophyllum to be a junior synonym of Lophophyllidium , while others separated the two genera (e.g. Fontaine 1961 ; Pyzhjanov 1966). A strange concept of the genera Lophophyllum , Lophophyllidium , and Sinophyllum given by Heritsch (1936) was critically discussed by Fomitshev (1953a) and Schouppe and Stacul (1955). The structure of the pseudo- columella interpreted by Fontaine (1961) is the only qualitative difference between Sinophyllum and Lophophyllidium. The interpretation of Fontaine (1961) differs, however, from that given originally by Grabau (1928). The great variability of this structure found in American specimens suggests that both of these interpretations are valid, but that they do not have taxonomic importance. This variability will be discussed below. Fliigel (1972) introduced a new concept of the family Timorphyllidae Soshkina, 1941. According to him, this family can be divided in two subfamilies: Timorphyllinae and Lophophyllidiinae. The structure of the pseudocolumella, with septal lamellae in Lophophyllidinae and without them in Timorphyllinae, is the only difference between them. Fliigel (1972) also synonymized Stereostylus with Sinophyllum and assigned Sinophyllum to the Timorphyllinae. The author agrees neither with Fliigel’s (1972) family concept nor with the assignment of Sinophyllum (= Lophophyllidium according to the author) to Timorphyllinae. The structure of the pseudocolumella in both groups of corals is variable and the ontogeny, as well as the micro- structure, is different. These similarities and differences between Timorphyllidae and Lophophyllidiidae will be discussed below. Malonophyllum Okulitch and Albritton, 1937 (type species: M. texanum Okulitch and Albritton, 1937). This genus, originally inadequately described and illustrated, was discussed by Moore and Jeffords (1941), who described the second species of this genus, M. kansasense Moore and Jeffords, 1941. Lack of tabulae is the only difference between Malonophyllum and Lophophyllidium. Unfortunately all of Okulitch and Albritton’s original material is lost. The originals of M. kansasense, re-examined by the present writer, do not seem to have tabulae, even in the proximal ends of corallites, but the preservation of the specimens and their small number (three only) are not adequate to be sure of this feature. It must be emphasized also that there are laminae in the pseudocolumella very similar to those made by tabellae or tabulae in other specimens (PI. 64, fig. 2). The genus Malonophyllum is not accepted by any authors except Moore and Jeffords (1941). It was included in the synonymy of Lophophyllidium by Wang (1947, 1950), Hill (1956), and others but none of these authors have restudied either the originals of Okulitch and Albritton (1937) or of Moore and Jeffords (1941). Soshkineophyllum mirabile Moore and Jeffords, 1941 was included by Jeffords (1947, p. 40) in the synonymy of Stereostylus. The original material has been restudied by the present writer and is here synonymized with Lophophyllidium. Stereostylus Jeffords, 1947 (type species; S. lenis Jeffords, 1947). The type specimen and some of the EXPLANATION OF PLATE 61 Figs. 1, 2. Lophophyllidium sp. nov. D. Glass Mountains, Texas. Road Canyon Fm„ Upper Leonard. Relationship between septal lamellae and axial tabellae in pseudocolumella. 1, USNM 189813. Septa are continued on surface of tabella as distinct ridges. No interruption between free tabella and pseudo- columellar lamina is observed, x 5. 2, USNM 189814. Septal lamellae increased as a twisted fold of tabella. Their connections with particular septa are not distinct. Continuation of basal elements is very clear, x 5. Fig. 3. Lophophyllidium sp. nov. C. USNM 189815, neanic stage. Almost all major septa reach the distinct, thick pseudocolumella, which remains simple, x 3. Fig. 4. Timorphyllum wanner i Gerth, 1921. USNM 189816. Basleo, Timor, Middle Permian, x3. a, bottom of the calice, where only major septa are visible, b, the surface of the corallite with twice as many grooves as major septa. Fig. 5. Timorphyllum wanneri Gerth, 1921. USNM 189817. Same locality and age. Surface of the corallite without epitheca preserved, x 3. PLATE 61 FEDOROWSKI, Lophophyllidium and Timorphyllum 446 PALAEONTOLOGY, VOLUME 17 illustrated specimens are missing. Other illustrated specimens, and quite a few specimens not illustrated, were available for restudy. Ontogeny was almost fully elaborated and illustrated by Jeffords (1947, text- fig. 8). The genus Stereostylus was included in the synonymy of Timorphyllum by Wang (1947, 1950), but was generally accepted by most later authors, except Duncan (1962) and Rowett and Sutherland (1964), who synonymized it with LophophyUidium , and Fliigel (1972) who synonymized it with Sinophyllum. According to Jeffords (1947, p. 38), the most important differences between Stereostylus and Lopho- phyllidium are: external features of the corallites, smaller apical areas filled by stereoplasm, thinner or more rhopaloid septa, laterally compressed axial column, and lack of radiating and circumscribing laminae in the column. According to Hill (1956), the main character of Stereostylus is the pseudocolumella, which is free of the counter septum in the mature stage. Agarikophyllum Fomitshev, 1953, subgenus of LophophyUidium Grabau, 1928 (type species: A. pavlovi Fomitshev, 1953). This subgenus was founded on one incomplete specimen. It was cited and accepted in Osnovy Paleontologii (1962) only. The septal lamellae are not completely connected to each other and the small open spaces between them in the pseudocolumella in the mature stage are the main characters of this subgenus. The value of these characters will be discussed below. Khmerophyllum Fontaine, 1961 (type species: K. cambodgense Fontaine, 1961). The structure of the pseudocolumella is the main character of Khmerophyllum. It is composed of axial lamella not connected directly with the middle dark line of the counter septum and with radially arranged fascicles of calcite fibres. Kato (1963) connected this genus with LophophyUidium. Fontaine (1964) stated, however, that the microstructure of the septa of Khmerophyllum is trabecular and connected it (as a subgenus) with Sino- phyllum. Fliigel (1972) synonymized Khmerophyllum with LophophyUidium. Rotiphyllum Ivanovsky, 1967 does not show any of the characters indicated by Hudson, 1942 for this genus. On the contrary, Ivanovsky’s specimen has a very well developed, complex pseudocolumella con- nected with the counter septum, a shortened cardinal septum, and a LophophyUidium- like arrangement of the major septa; these are characteristic of the genus LophophyUidium. My reasons for placing all of these genera in synonymy with LophophyUidium are discussed in the following sections. IMPORTANCE OF BIOMETRIC CRITERIA The relationship between number of septa and corallite diameter is the most popular, and is generally considered the most important, biometric criterion in tetracorals. However, text-fig. 1 shows that this criterion is not very useful in the group of corals discussed. The holotypes of the type species of the discussed ‘genera’ included in EXPLANATION OF PLATE 62 Fig. 1. LophophyUidium proliferum (McChesney, 1860). Syntype KUM 52878, x3. a, transverse section of neanic stage, b, transverse section of ephebic stage, c, longitudinal section. Fig. 2. LophophyUidium sp. nov. C. Glass Mountains, Texas. Road Canyon Fm., Upper Leonard. Marginal part of corallite USNM 189807 showing the absence of minor septa, x 10. Figs. 3-5. LophophyUidium sp. nov. A. Same locality and age. Different stages of development of the contra- tingent minor septa, x 3. 3, USNM 189818; 4, USNM 189819; 5, USNM 189820. Fig. 6. Stereostylus lenis Jeffords, 1947. Paratype KUM 37361, illustrated by Jeffords (1947, pi. 14, fig. 13 a-c). Pseudocolumella compound and connected with the counter septum, x 3. Fig. 7. Stereostylus lenis Jeffords, 1947. Paratype KUM 37327. Two successive transverse sections, x3. a, through the tabula with elongated counter septum, b, beneath tabula with free pseudocolumella. Figs. 8 12. LophophyUidium sp. nov. D. Glass Mountains, Texas. Road Canyon Fm., Upper Leonard. Some examples of the pseudocolumella from simple (fig. 8) to ‘Khmerophyllum’- type (fig. 12), x3. 8, USNM 189821 ; 9, USNM 189822; 10, USNM 189823; 11, USNM 189824; 12, USNM 189825. Fig. 13. Timorphyllum wanneri Gerth, 1921. USNM 189826. Basleo, Timor, Middle Permian, x 3. a-d, successive transverse sections of neanic stage. Note extremely short cardinal septum, e, transverse section of ephebic stage. PLATE 62 FEDOROWSKI, Lophophyllidium and Stereostylus 448 PALAEONTOLOGY, VOLUME 17 LophophyUidium are marked by capital letters. Malonophyllum and Khmer ophy Hum appear to be sharply separated from the others, but it is quite easy to change this impression when some American species, indicated by small letters, are compared. At the same time, these last points are arranged very much like the data of one species only. Text-fig. 1 does not show the amounts of individual variability of the separate species. It shows, however, that: 1, simple comparison of n/d ratio of holotypes or a small number of paratypes is not adequate as a specific character; 2, the n/d ratio must be very carefully used, mainly as a secondary rank character; 3, there are some differences among certain groups of species, for example the American and Timor species of LophophyUidium. At the same time the plotted points of Timor Lopho- phyUidium (data after Schouppe and Stacul 1955) show that the Timor specimens of Timorphyllum and LophophyUidium are hardly differentiated in the diagram. One more observation should be made: the biometric data normally used in literature are mostly accidental. It has not been agreed which part of the corallite should be cut and measured for comparison. Hence, the available data seldom indi- cate which parts of corallites are compared. Data on the diagram (text-fig. 1) from the holotype of ‘Khmer ophy Hum' cambodgense emphasize how important this is. In fact, it is no less important in most solitary rugose corals. Two regions should be measured and identified : the margin and base of the calice. The measured regions should be indicated. Other uses of biometry for generic or specific characters (Jeffords 1947), which look quite spectacular, are not meaningful. Relations like length of corallite to its diameter at the calyx and cumulative frequency of length are too closely dependent on the environment to be useful. The length of septa in relation to diameter of corallite (another ratio proposed by Jeffords 1947) has no more importance than the n/d ratio, or even less. ONTOGENY Fortunately almost all of the discussed genera have juvenile stages more or less com- pletely described. In Stereostylus lenis Jeffords ( 1 947, fig. 8) and LophophyUidium pro- liferum (McChesney 1860) (Duerden 1906) the ontogeny has been investigated from the nepionic (6 septa) stage. Agarikophyllum is the only genus in which juvenile stages are unknown. It is not necessary to redescribe the ontogeny. 1, all the fully investigated speci- mens have early stages that are typical for the suborder Streptelasmatina— six connected protosepta and paired metasepta in quadrants. 2, arrangement of septa in the early neanic stage is zaphrentoid, with the cardinal and counter septa con- nected at the corallite axis. 3, counter septum remains elongated and extended to the axial part of corallite while the cardinal septum, which is long during the early part of the neanic stage, becomes gradually shortened in the later part of this stage. 4, axial part of the counter septum may remain simple or some septal lamellae (1-2 or more) may be fused with it making a compound pseudocolumella. 5, in many specimens or species the middle line of the counter septum does not continue into the pseudocolumella, whereas in many others the pseudocolumella becomes free of FEDOROWSKI: LO PHOPHYLLIDIUM AND TIMORPHYLLUM 449 text-fig. 1 . Diagram showing the relationship between number of septa (n) and diameter of corallite (d) in holotype specimens of ‘genera’ synonymized with Lophophyllidium, in Timorphyllum, and in some American species of Lopho- phyllidium. A, Agarikophyllunr, K, Khmerophyllum; l, Lophophyllidium ; L', Lophophyllidium described by Schouppe and Stacul (1955) from Timor; m, Malophyllum; s, Sino- phyllum; st, Stereostylus ; t, Timorphyllum ; o, specimens from Upper Leonard of the Glass Mountains, Texas; a-u, holotypes (with superscript ‘h’ added) and paratypes of some American species of Lophophyllidium described by Moore and Jeffords 1941, 1945 and Jeffords 1942, 1947. 450 PALAEONTOLOGY, VOLUME 17 the counter septum in the mature stage; neither the first nor second possibility is con- nected with a particular pseudocolumella structure. 6, minor septa appear normally very early in ontogeny and are well developed. In some species they appear only in the calice and then disappear into the thick stereoplasmatic layer on the inside surface of the external wall (PI. 60, fig. 2b; PI. 62, fig. 2). FINE STRUCTURE The existence of uni- and multitrabecular fine structure of septa in Lophophyllidium was noted by Kato (1963). The arrangement of trabeculae and septal growth lines, however, was not pointed out and illustrated. This structure in one of the original syntypes of L.proliferum (McChesney 1 860) (PI. 63, fig. 4), as well as in other American species investigated, should be considered as characteristic for the genus. Septal growth lines in most of the septa are very steeply, almost vertically arranged. In a few species these lines are almost horizontal in the external part of the septum when the curvature of the upper septal margin is distinctly marked. No septal growth lines running down the external margin of the septum (close to epitheca) were seen. In longitudinal section, trabeculae are very fine. They intersect mostly two, seldom three or more, septal growth lines. The arrangement of trabeculae is perpen- dicular to the growth lines. In cross-section they are not distinctly separated from each other and form the structure called lamellar (Schindewolf 1942). No multi- trabecular fine structure of septa was found in American specimens investigated. Instead, there are quite a few corals with zigzag structure, which is considered to be a result of recrystallization (see Oekentorp 1972, for discussion). It is surprising that few of the specimens investigated with the stereoscan micro- scope have well-organized crystalline structure even when the fine structure in trans- mitted light is well preserved. Crystals are mostly differentiated into two shapes only : fine crystals in the organic structures of the corallite and larger crystals in the inorganic interspaces. No trabecular arrangement of crystals was found in these septa, pre- sumably a result of recrystallization (PI. 70, fig. 1). The trabeculae were found in a few other septa, however (PI. 70, fig. 2). In some parts of the septa the mid-line of the septum can be seen (PI. 70, fig. 3), but this line is not constant and not visible in every part of the septum. The interruption of the middle line could not be attri- buted to the trabeculae-like organization of crystals. However, the so-called ‘dark EXPLANATION OF PLATE 63 Fig. 1. Timorphyllum wanneri Gerth, 1921. USNM 189827. Basleo, Timor, Middle Permian. Longitudinal section along middle line of septum. External part of corallite on left, x 50. Fig. 2. Timorphyllum wanneri Gerth, 1921. Same specimen, x 3. a, b, successive transverse sections of the ephebic stage. Fig. 3. Lophophyllidium simulans (Moore and Jeffords, 1941). Flolotype KUM 52869. Glass Mountains, Texas. Leonard. Distinct minor septa are visible, x 10. Figs. 4, 5. Lophophyllidium proliferum (McChesney, 1860). Syntype KUM 52878. 4, longitudinal section along septum. External part of corallite on left, x 50. 5, complex structure of pseudocolumella, x 30. PLATE 63 FEDOROWSKI, Timorphyllum and Lophophophyllidium 452 PALAEONTOLOGY, VOLUME 17 line’ visible in transmitted light is, apparently, the result of the vertical arrangement of crystals in this part of the septum. I do not agree with Schindewolf’s (1942) interpretation of the fine structure of some septa. Since lamellar structure was indicated by Schouppe and Stacul (1955) for Lophophyllidium and Timorphyllum it should be discussed here. Schindewolf’s specimens were not studied and so it cannot be stated definitely that the whole con- cept of lamellar structure is wrong. It is wrong, however, so far as Lophophyllidium and Timorphyllum are concerned and those Polycoelaceae studied by Ilina (1965), who identified trabecular microstructure in her specimens. The transverse section is not the best way to find trabeculae in corals. The longi- tudinal section made correctly through the middle part of the septum is the only one that can show the presence or absence of trabeculae. Sectioning specimens of Lopho- phyllidium and Timorphyllum showed trabeculae in both (PI. 63, figs. 1, 4). The septal growth lines are shown in the same picture. In my opinion the arrangement of those lines and the arrangement of trabeculae are diagnostic characters for the discussed genera. Schindewolf’s (1942, text-fig. 6 and text-fig. 2 a herein) original picture was changed slightly by Schouppe and Stacul (1955, text-fig. 9 and text-fig. 2b herein), making the interpretation of the real architecture more difficult. It is clear in both of those pictures, however, that the sections of the septum were cut improperly, not perpen- dicular to the trabeculae (text-fig. 2c, line A'B ), but perpendicular to the septum (text-fig. 2c, line AB). As a result of this incorrect angle, the structure visible in the section is misleading (text-fig. 2 d, e). Fibres on one side of the trabecula are cut more or less perpendicular to their crystal axes, while on the opposite side the cut is made almost parallel to the axes. Transmitted light going through the thin section is much more absorbed by fibres cut perpendicularly and one can see the fascicles of fibres shown by Schindewolf (1942) as a lamina. This is especially clear in the case when the trabeculae are slightly flattened (text-fig. 2d, e). The real architecture is also obscured by secondary septal accretion, which is often a continuation of the primary septal deposits. The internal margin of the septum, where the growth lines are almost vertical and trabeculae almost horizontal, is the best place to demonstrate what was stated above. There the points of the septal trabeculae are cut very obliquely and there- fore the ‘laminae’ are very distinct. In Timorphyllum , where the growth lines of the septa are convex in the middle part, one can observe the ‘laminae’ in the external, EXPLANATION OF PLATE 64 All figures x 30. Fig. 1. Lophophyllidium sp. nov. A. Glass Mountains, Texas. Road Canyon Fm., Upper Leonard. USNM 189828. Three successive transverse sections of the pseudocolumella beneath the bottom of the calice. a, complex, mostly septo-lamellar pseudocolumella, b, septo-lamellar structure in internal part and tabulo-laminar structure in external part of pseudocolumella, c, tabulo-laminar part of pseudocolumella. Fig. 2. Malonophyllum kansasen.se Moore and Jeffords, 1942. Holotype KUM 52848. Part of pseudo- columella with weakly septo-lamellar structure filled up secondarily by tabulo (?)-laminae. Fig. 3. Lophophyllidium sp. nov. D. USNM 189822. Glass Mountains, Texas. Road Canyon Fm., Upper Leonard. Weak, small, complex pseudocolumella. PLATE 64 FEDOROWSKI, Lophophyllidium and Malonophyllum L / 454 PALAEONTOLOGY, VOLUME 17 text-fig. 2. Interpretation of lamellar fine structure of septa. a, Schindewolf’s (1942) schematic picture. b, Schouppe and Stacul’s (1955) modification of the previous picture. c, author’s interpretation of Schindewolf’s (1942) picture with trabeculae perpen- dicular to the growth lines. A-B section given by Schindewolf is very oblique to trabeculae. A'-B' section more or less perpendicular to the trabeculae. d, supposed irregular trabecula with the line of section given obliquely. e, the same trabecula cut along the oblique plane showing in d. EXPLANATION OF PLATE 65 Fig. 1. Lophophyllidium extumidum Moore and Jeffords, 1945. Paratype USNM 140325 (figured by Moore and Jeffords, 1945, fig. 21a, b), x 30. a, transverse section of pseudocolumella, b, longitudinal section between septa showing connection of free tabellae and pseudocolumellar laminae. Fig. 2. Lophophyllidium extumidum Moore and Jeffords, 1945. Paratype USNM 140324 (figured by Moore and Jeffords 1945, fig. \ la~c), x 50. Longitudinal section along septum and through pseudocolumella, showing connections between growth lines of septum and pseudocolumella. Fig. 3. Lophophyllidium hadrum Jeffords, 1947. Holotype KUM 32279 (figured by Jeffords 1947, pi. 3, fig. 3 a-d. Compare also PI. 67, fig. 3 in this paper). Complex and compact pseudocolumella in the neanic stage of corallite, x 30. PLATE 65 FEDOROWSKI, Lophophyllidium 456 PALAEONTOLOGY, VOLUME 17 as well as in the internal, part of the septum, although oppositely directed (PI. 66, fig. 3), and there are no laminae’ at all in the more perpendicularly cut, middle part of the septum. MACROSTRUCTURES Major septa Major septa are not very important systematically in the genera discussed, in con- trast to their usefulness in many other groups of corals. The arrangement of septa in the young stage is similar in all synonymized genera, as discussed above. Arrangement in the mature stage varies from pinnate through bilateral to radial in the species investigated. At the same time, the length of particular septa in cross-section and in the calice can be changeable. The cardinal septum is generally shortened, but can be almost as long as the rest of the major septa in some species. In this case, however, the last pair of metasepta in the cardinal quadrants is generally shortened. Alar septa in many described species are slightly or much elongated. There are many other species in which these septa are equal in length to other major septa and can be distinguished only by the presence of the last pair of shortened metasepta in the counter quadrants as seen in cross-section or on the surface of the corallite. Alar septa are never shortened. Metasepta may be more or less equal in length in some species, while in other ones they are very much differentiated. This character, as well as the total length of major septa in comparison to the corallite diameter is very often consistent in particular species, but only in comparable growth stages. Changes of arrangement and length of major septa during ontogeny in one species are shown on Plate 60, figs. 3-6. The counter septum will be discussed together with the pseudocolumella. Minor septa The minor septa are commonly free and quite short, as illustrated by many authors. It is possible, however, to find among Lophophyllidium species, quite different types: 1. Underdeveloped minor septa, which are characteristic, for example, of Lopho- phyllidium sp. nov. C. They are quite visible in the calices and more or less (PI. 60, fig. 2a, b) on the surface of corallites, but they are completely covered by stereo- plasm beginning from half the depth of calices down. No minor septa are visible in cross-sections of the specimens of this species (PI. 62, fig. 2), except in sections through the uppermost part of the calice. 2. Completely or partly contratingent minor septa. It is not clear which kind of minor septa should be called contratingent, those with dark lines of major and minor septa fused or those in which only the light tissue is in contact. Only the second case EXPLANATION OF PLATE 66 Fig. 1. Lophophyllidium sp. nov. A. USNM 189829. Glass Mountains, Texas. Road Canyon Fm., Upper Leonard. Longitudinal section along tabello-lamellar pseudocolumella, x 30. Figs. 2, 3. Timorphyllum wanneri Gerth, 1921. USNM 189826. Basleo, Timor, Middle Permian. 2, SEM photomicrograph of zigzag arrangement of trabeculae, x 350. 3, transverse section of septum (a piece of middle part of septum was cut out) showing ‘lamellar’ structure both in external (down) and internal part of septum and the zigzag arrangement of the trabeculae in its middle part, x 50. PLATE 66 FEDOROWSKI, Lophophyllidium and Timorphyllum 458 PALAEONTOLOGY, VOLUME 17 has been observed in Texas specimens. On Plate 62, figs. 3-5 a few corallites from the same species and locality with normal and contratingent septa are shown. The full intraspecific variability of this character will be discussed and illustrated in a future paper. Of course, as in every case of contratingency, the triad at the counter septum is developed. Pseudocolumella This is the structure on which most speculations about this group of corals have been based. It is closely related to the counter septum in all of the discussed genera. It may be separated from the axial end of the counter septum or be permanently a part of it. The pseudocolumella is monoseptal in all neanic stages studied. Beginning from the late neanic stage there are several types of gradual change, most of which were considered typical for particular genera : simple pseudocolumella, which can be free of the counter septum; in Stereostylus, pseudocolumella composed of septal lamellae, in Lophophyllidiimr, pseudocolumella, pendulum-like, composed of fibres (according to Fontaine 1961; non Grabau 1928), in Sinophyllum ; pseudocolumella composed of radially arranged fascicles of fibres, in Khmerophyllum ; pseudocolumella with septal lamellae not closely packed, in Agarikophyllum; pseudocolumella ‘con- centrically lamellar’ (Okulich and Albritton 1937) or composed of septal lamellae and layers of stereoplasm (in Moore and Jeffords’ revised specimens), in Malono- phyllum. The following observations resulting from a study of North American specimens can be easily transferred to the type specimens of the discussed ‘genera’. Simple pseudocolumella. This is typically developed in the juvenile stages of all specimens in this group of corals, but can remain unchanged to the end of ontogeny. It can be built as a simple elongation of the counter septum, sometimes not even thickened (PI. 60, fig. 5; PI. 62, fig. 8) or with stereoplasmatic covers (PI. 61, fig. 3; PI. 62, fig. la). It can also be completely free of the counter septum (PI. 62, fig. lb). The successive stages of the process of its separation will be discussed below. The young corallites are the best proof that the pseudocolumella can be very distinct and stereoplasmatically thickened, remaining monoseptal at the same time. It is true even in specimens having a composite pseudocolumella in the mature stage (PI. 61 , fig. 3). Many major septa reach the pseudocolumella here, but do not penetrate EXPLANATION OF PLATE 67 All figures x 30. Fig. 1. Lophophyllidium sp. nov. B. Glass Mountains, Texas. Road Canyon Fm., Upper Leonard. USNM 1 89830 showing secondary zigzag fine structure of septa and one of the possibilities of breaking the connection between pseudocolumella and counter septum on the successive transverse sections, a, middle line of counter septum (horizontal on the upper part of the picture) is connected at right angles with the middle lamella, b, counter septum (vertical, left part of picture) is overpassing pseudocolumella at its side. Fig. 2. Lophophyllidium sp. nov. D. USNM 189831. Same locality and age. A few incipient septal lamellae in thin pseudocolumella connected with counter septum. Fig. 3. Lophophyllidium hadrum Jeffords, 1947. Holotype KUM 32279 (compare also PI. 65, fig. 3 in this paper). Cross-section of the pseudocolumella of the adult part of specimen showing weak, ‘ Agariko - phyllum'- type structure. PLATE 67 FEDOROWSKI, Lophophyllidium 460 PALAEONTOLOGY, VOLUME 17 it. The process of the gradual complication of the structure of the pseudocolumella can be observed both in the ontogeny of the same specimen and as individual variation in different specimens belonging to the same species. Lophophyl/idium sp. nov. D is a good example of the last case (PI. 62, figs. 8-12). The pseudocolumella changes here from simple to very complex while the other structural elements remain stable and similar. Composite pseudocolumella. (1) A lamellar pseudocolumella is composed of the axial end of the counter septum (medial lamella) and septal lamellae. The medial lamella in well-preserved specimens always rests on the thickened part of the pseudo- columella (PI. 60, figs. 2b , 7). The ‘pure’ lamellar pseudocolumella can be investi- gated mainly in the calices of specimens without tabulae. One can see there that the septal lamellae are situated vertically or obliquely on the surface of the medial lamella. They begin to develop close to the upper part of the medial lamella as very tiny folds, which become gradually larger downwards (PI. 60, fig. 7). Septal lamellae in such specimens are directly fused with major septa at the same level, i.e. the apparent base of the calice. It is the same level, common for each of the major septa, at which their direction of growth changes from being directed into the corallite centre through a short section of upward growth (apparent base of the calice) to growth outwards. The last type of growth is characteristic for septal lamellae. In addition almost all major septa in most of the investigated specimens, turn right (counter-clockwise) on the apparent base of the calice. All the described changes can be observed both in the calice (PI. 60, fig. 7), and in the transverse section of the columella, where the fascicles of fibres are situated opposite to those in the septa (PI. 69, fig. 2). Indeed, the septa and septal lamellae cannot be separated from each other. The origin of both is the same. Both are secreted in the same fold of ectoderm, which began at the epitheca and ended close to the top of the pseudo- columella, where it connected with the fold of ectoderm which secreted the medial lamella. Septal lamellae in specimens without tabulae are usually distinct and clearly separated from each other under the apparent bottom of calice. Beneath it they can be connected to each other if numerous. If there are only a few septal lamellae, the spaces between them are secondarily filled by the stereoplasmatic thickening of the medial lamella. The fine structure of these parts is fibrous, or can be zigzag, when recrystallized (PI. 69, fig. 1). Between widely separated septal lamellae, however, the EXPLANATION OF PLATE 68 All figures x 30. Fig. 1. Lophophyllidium sp. nov. A. USNM 189832. Glass Mountains, Texas. Road Canyon Fm., Upper Leonard. Transverse section of pseudocolumella showing closely packed septal lamellae in the internal part and laminar structure with a few septal lamellae in the left part. Fig. 2. Lophophyllidium sp. nov. A. USNM 189829. Same locality and age. Transverse section of the pseudocolumella showing regular, lamello-tabellar structure. Fig. 3. Lophophyllidium sp. nov. D. USNM 189825. Same locality and age. a, transverse section of pseudo- columella in neanic stage with two septal lamellae only, b, transverse section of the pseudocolumella in the adult stage showing 'Khmerophy llum' -type structure. PLATE 68 FEDOROWSKI, Lophophyllidium 462 PALAEONTOLOGY, VOLUME 17 axial, steeply arched tabellae commonly penetrate. These tabellae could be elongated without any boundary into the structure of the pseudocolumella. Laminar structure of the pseudocolumella between septal lamellae may be observed in this case (PI. 68, figs. 1, 2). The latter structure can be called the lamello-tabellar pseudocolumella. (2) Lamello-tabellar pseudocolumella. A pseudocolumella of this type is derived from that mentioned above by a change of rhythm of the secretion of the structural elements. The surfaces of axial tabellae become the bases for septal lamellae (PI. 61, figs. 1, 2). This change of rhythm can be observed in many cross-sections of the pseudocolumellae (PI. 68, fig. 3b). Many of the intermediate stages observed suggest that this passing from one structure to another was probably not genetically deter- mined and thus not helpful for taxonomic purposes. Most probably this change is a function of time (= stage of growth) and very much dependent on individual variability. There are quite a few corallites in the investigated collection that did not build lamello-tabellar pseudocolumellae. There are also many others in which only very narrow or incomplete parts of laminar structure can be observed ; in these cases, growth of lamellae was faster than growth of the basal elements, which did not cover the previous ones (PI. 69, fig. 2). At least there are such corallites in the collection possessing a pseudocolumella in which axial tabellae predominate, while the lamellae are settled in them as short sparse fascicles or fibres (PI. 69, fig. 3). A good example of the entire change of the structure of the pseudocolumella in the same specimen is shown on Plate 64, fig. 1 a-c. The pseudocolumella is monoseptal at the beginning, as in all other specimens of the genus, and then becomes lamellar, lamello-tabellar, and simple-lamellar close to the calice, but still beneath the last tabula. The lamello-tabellar pseudocolumella can have sometimes a little non-typical structure, which has been interpreted as a generic character ( Agarikophyllum , Khmero- phyllum). The pseudocolumella of the Khmer ophy Hum- type appeared when septal lamellae are spirally arranged and very closely packed on the particular tabellae (PI. 68, fig. 3b). The cross-section of such a pseudocolumella characteristically shows more or less regularly arranged fascicles of fibres in a comparatively long row. This occurred when the septal pockets of ectoderm on the successive tabellae increased immediately one after another. The degree of complication of this type of pseudo- columella can be very great. In one Glass Mountains species, very great variability of this character was noted, and three extreme forms of pseudocolumella for this species are illustrated (PI. 64, fig. 3; PI. 67, fig. 2; PI. 68, fig. 3a, b). The pseudocolumella of the Agarikophyllum-type appears in specimens which do not have thick, secondarily secreted stereoplasmatic sheets on the primary structural EXPLANATION OF PLATE 69 All figures x 30. Figs. 1-3. Lophophyllidium sp. nov. A. Glass Mountains, Texas. Road Canyon Fm„ Upper Leonard. 1, USNM 189833 showing middle line of counter septum just after interruption of the middle line of pseudocolumella. Septal lamellae rarely placed, with laminar (recrystallized in zigzag structure) layers between them. 2, USNM 189834. Pseudocolumella free of counter septum is built by closely packed septal lamellae. 3, USNM 189835. Middle line of counter septum is connected with median lamella. Pseudocolumella mostly lamellar with rarely arranged, short, fascicle-like septal lamellae. PLATE 69 FEDOROWSKI, Lophophyllidium sy.WAk 464 PALAEONTOLOGY, VOLUME 17 elements. It can be observed mostly just before the end of the individual develop- ment. Lophophyllidium hadrum Jeffords (PI. 65, fig. 3; PI. 67, fig. 3) is a good example of this kind of structure. The normal, closely packed lamello-tabellar pseudocolumella starts to weaken inside the calice region, without losing its primary external shape. This type of pseudocolumella should be considered as a gerontic, or at most, a specific character. Explanation of some of the terms used in this discussion. 1 . ‘Lamellae increasing on the surface of the tabellae’ can be taken literally when the new septal lamella appears after a separate period of only tabellar (laminar) secretion. This happens quite often, but not so generally as the second possibility, i.e. when both of the secretions, tabellar and lamellar, are practically contemporary. In the last case, ectodermal pockets secreting septal lamellae are formed after a very short period of flat, tabular secretion or even without it in some places. It can be observed in longitudinal section, when pieces of lamellae, intersecting a few tabellae are visible. It is difficult to distinguish the axial tabellae in cross-section, when septal lamellae are very closely packed; it is almost always possible in the longitudinal section, however (PI. 66, fig. 1). 2. The term ‘lamello-tabellar pseudocolumella’ could be restricted. It is difficult to identify with certainty the laminae with the axial tabellae in the pseudocolumella. In longitudinal sections of many corallites one can observe the continuation from free tabella into pseudocolumellar lamina (PI. 65, fig. la). In other corallites or in particular parts of them, however, only tabellae seem to reach the pseudocolumella. It is almost impossible to distinguish particular laminae in the pseudocolumella in these sectors of the pseudo- columella, but it is still possible in others. The analysis of quite a few specimens assured the author that both of the basal elements — free axial tabellae or tabulae and pseudocolumellar laminae— were secreted con- temporaneously by continuous basal ectoderm. Concerning the relationship between septal lamellae and pseudocolumellar laminae, Lophophyllidium extumidum Moore and Jeffords, 1945 gives more informa- tion. The pseudocolumella in cross-section of this species appears to be laminar only (PI. 65, fig. la). In longitudinal section, however (in the section made between septa) (PI. 65, fig. 16), the continuation from free tabula into pseudocolumellar lamina is visible, and in section made along the septum, close to its middle line, the growth lines of the septum are continued right up into the growth lines of the pseudocolumella (PI. 65, fig. 2). The described example proves the uniformity and replacement of such seemingly different structures as septal lamellae and axial tabellae (or pseudo- columellar laminae) in the pseudocolumella. The connection of the counter septum and pseudocolumella may be complete or only apparent. Many pseudocolumellae are isolated. Full connection occurs when the mid-line of the counter septum is continued into the mid-line of the pseudo- EXPLANATION OF PLATE 70 All figures x 500. Fig. 1. Timorphyllum wanneri Gerth, 1921. USNM 189826. SEM micrograph of longitudinal section of tabula (photo W. R. Brown). Figs. 2-4. Lophophyllidium sp. nov. A. SEM micrograph of transverse sections of septa. 2, USNM 189811. Two very small trabeculae in the middle part of septum (photo Dr. J. E. Sorauf). 3, USNM 189828. Primary septal structure (dark middle line in transmitted light) is visible as small crystals with apparent random arrangement. The big crystals at the margins are secondary organic septal structure. Both structures are changed by recrystallization. 4, USNM 189836. Secondary zigzag structure due to recrystallization of mid part of septum. Long crystals preserved in some places may be remnants of particular trabeculae. PLATE 70 3 4 FEDOROWSKI, Timorphyllum and Lophophyllidium 466 PALAEONTOLOGY, VOLUME 17 columella. Apparent connection occurs when the mid-lines are not fused or when the side of the counter septum rests against the pseudocolumella. The complete isolation of the pseudocolumella and counter septum is produced either by shortening of the counter septum or by change of direction of its growth; it can be parallel to the pseudocolumella without any connection with it. On Plate 67, fig. 1 a, b and Plate 69, figs. 1, 3 a few successive stages of the relationship between those two elements are shown. This character is not of great value for systematics even at the specific level. The shape of the counter septum in the calice and its connection with the pseudo- columella in this part of the corallite may have some value for specific identification. However, this character must be very carefully investigated both in ontogeny and individual variability before using it for this purpose. Some examples of shapes of the counter septum are shown on text-fig. 3. text-fig. 3. Different shape of counter septa and different kinds of connections with pseudocolumellae in species of Lophophyllidium from Glass Mountains, Texas (Leonardian). a-c, Lophophyllidium sp. nov. A ; counter septum and pseudocolumella in the specimens are in different stages of growth. d, e, Lophophyllidium sp. nov. B\ different shape of septa and pseudocolumellae in mature specimens. /, Lophophyllidium sp. nov. C; typical relationship between septa and pseudocolumella. g, h, Lophophyllidium sp. nov. D\ different shape of septa and pseudocolumella in mature specimens. FEDOROWSKI: LO P HO P HY LLIDIU M AND TIMORPHYLLUM 467 Tabularium Among the corals assigned by the author to Lophophyllidium are many individuals without a tabularium, i.e. without tabulae. All appear to have juvenile characters: pinnate arrangement of septa which reach a simple, not compound pseudocolumella. It is possible to arrange these specimens in a limited development line and compare them with the stages of mature specimens which do possess a tabularium. Arrangement and number of septa in both groups are approximately the same (septa were numbered on the apparent bottom of calices) when individual variability is considered. The cross-sections of young stages of mature specimens differ from the young corallites without tabularia in having thick, secondary stereoplasmatic sheets on the septa. The moment when the corallite started to secrete tabulae is very variable in this group of corals, but it happened mostly late in ontogeny. It seems that tabulae are not very important skeletal elements here. The apparent bottom of the calice, as well as the strong skeleton in the central part of corallite (pseudocolumella in calice and stereoplasmatic column beneath it), make the tabulae unimportant from a mechanical point of view. The secreted vertical elements were probably sufficient both for supporting the body of the polyp and for protecting the corallite against external pressure. No mature individuals without a tabularium were found among the corals studied by the author. It seems, however, that three specimens from Kansas described by Moore and Jeffords (1941) as Malonophyllum kansasense , have no tabularium even in the mature (or mature-like) stage. However, they have laminae in the pseudo- columella (PI. 64, fig. 2). It is not considered in this particular case that the lack of a tabulae is an adequate generic character, because of the poor preservation of the Kansas material, as well as wide heterochronism observed among very similar corals. Malonophyllum and Lophophyllidium are therefore considered synonyms. STRATIGRAPHIC POSITION Reports of species of the discussed genera in the Lower Carboniferous are very sporadic and poor. Vojnovsky-Krieger (1934) described one incomplete specimen from the Gornyi Altai as Lophophyllum micula. The original material in the Tscherny- shev Museum in Leningrad has been examined. It has a well-developed counter septum and a kind of pseudocolumella, but its very short septa and flat tabulae can hardly be compared with those of Lophophyllidium proliferum. The possibility of some relationship with Lophophyllidium cannot be excluded. Stuckenberg (1904) described Lophophyllum minimum , which may be a representative of the genus Lophophyllidium , but it may actually belong to Lophophyllum or be a young corallite (without dis- sepimentarium) of Koninckophyllum. The material in Leningrad consists of three pieces of corallite sectioned transversely and obliquely. In transverse section the septa are thin and uniform, except for a shortened cardinal and elongated counter septa. The pseudocolumella is well developed, not compound; quite a few sectioned tabellae surround it. In oblique section, arched tabulae and a well-developed pseudo- columella are visible. The material is not adequate to make a final decision about its generic position. Stuckenberg's stratigraphic designation (Lower Carboniferous) is also not certain. 468 PALAEONTOLOGY, VOLUME 17 e f 9 h text-fig. 4. Schematic drawings showing arrangement of septa and structure of pseudocolumella in particular genera synonymized in the paper. a , Lophophyllum Milne-Edwards and Haime according to Lecompte (1955) revision based on type material (type species Lophophyllum konincki Milne-Edwards and Elaime). Keyhole cardinal fossula and simple columella. ft, Lophophyllum Milne-Edwards and Haime according to Carruthers (1913) revision based on topotype material (type species Cyathaxonia tortuosa Michelin). Specimens possess dissepimentarium and simple columella. They were later synonymized with Koninckophyllum Nicholson and Thomson, 1876. c, Sinophyllum Grabau, 1928. c 1, pseudocolumella according to Grabau, 1928 possesses a ‘series of rod-shaped bodies’ which were compared by Huang (1932) and Wang (1947) with the septal lamellae of Lophophyllidium proliferum. c 2, pseudocolumella built with simple fibres of calcite, according to Fontaine 1961. d , Lophophyllidium and the next four ‘genera’ have the same arrange- ment of major septa with shortened cardinal septum, elongated counter septum, and more or less distinct pseudofossulae. Lopho- phyllidium has pseudocolumella complex, built with septal lamellae. e, Malonophyllum , pseudocolumella built with rare septal lamellae and probably tabular laminae. According to Okulitch and Albritton 1937 there are no tabulae. /, Stereostylus , pseudocolumella may be connected with counter septum or free of it. It may also be simple or complex in the different para- types of the type species. g, Agarikophyllum , pseudocolumella not completely compact. Par- ticular septal lamellae may not be connected with each other. /;, Khmer ophy llum, pseudocolumella built with short fascicles of calcite situated on the tabular laminae. FEDOROWSKI: LO P HO P HY LLID IU M AND TIMORPHYLLUM 469 Rotiphyllum sp. described by Ivanovsky (1967) from the Lower Carboniferous (Lower Miksin stage) of the Lena River seems to be definitely a representative of the genus LophophyUidium with a compound pseudocolumella. This, and Lophophyllidium sp. described by Kostic-Podgorska (1955) from the Lower Carboniferous of Yugo- slavia are presently the best illustrated and most definite representatives of Lopho- phyllidium from the Lower Carboniferous. Termier and Termier (1950) reported Stereostylus and a few other similar forms from the Visean and probably Namurian of Morocco. Among them, some cross- sections of Stereostylus maroccanus (Termier and Termier 1950, pi. 35, fig. 28) look similar to normally developed Lophophyllidium species, despite Termier and Termier having stated that a dissepimentarium is present. The dissepiment-like structures look very similar to sections of the external parts of tabulae. This specimen was described from the Visean/Namurian boundary. In the same paper Termier and Termier illustrated some other Lophophyllidium- like specimens, Cyathaxonia sp. (pi. 36, fig. 5) from the Visean and Cyathaxonia (?) sp. (pi. 36, fig. 21) from uppermost Visean or Namurian. All these specimens may belong to the genus Rylstonia or to Lopho- phyllidium. The material has not been seen and no longitudinal sections were illus- trated. Drawings made by Termier and Termier are inadequate. There are no reports of any Lophophyllidium-like corals from the Lower and Middle Namurian. However, the coral fauna of that age is one of the poorest known from the Carboniferous. Corals of Lophophyllidium and Stereostylus type became quite common in the Lower Morrowan (Namurian C) of the U.S.A. (Jeffords 1942; Moore and Jeffords 1945; Rowett and Sutherland 1964), but they are not recorded from Europe or Asia. Dr. N. P. Vassiljuk kindly informed the author (written communication 1973) that she has one specimen of Lophophyllidium from Namurian C (Limestone F2) of the Donets Basin and that there are some Lophophyllidium corals in the Bashkirian of the Petshora region and in the Moskovian of the North Ural. Fomitshev’s (1953a) paper is the first report of Lophophyllidium in Westphalian A (Upper Morrowan) of Donets Basin. Starting from this horizon and extending into the Permian, Lopho- phyllidium and Stereostylus- type corals are quite common in Eurasia. They are also one of the most important components of the coral fauna in America during the same period. Agarikophyllum was reported from beds of Westphalian D-Lower Stephanian age in the Donets Basin (Fomitshev 1953a). It has the same type of pseudocolumella as has commonly been reported from the Upper Pennsylvanian of the U.S.A. (Jeffords 1947) and is also observed in the Glass Mountains Permian collection. Species assigned to the genera Sinophyllum , Malonophyllum , and Khmerophyllum have been reported only from the Permian. Sinophyllum is reported mostly from Far East Asia (e.g. Grabau 1928; Huang 1932; Fontaine 1961; Pyzhjanov 1966), Malono- phyllum from the Permian of North America exclusively (Okulitch and Albritton 1937; Moore and Jeffords 1941), and Khmerophyllum from Cambodia (Fontaine 1961). This review suggests that : (a) There is no interruption in the stratigraphic distri- bution of Lophophyllidium-type corals, except in the Lower and Middle Namurian, but this time period is one of the poorest known periods of tetracoral history. ( b ) The oldest known species (except L. micula Vojnovsky-Krieger, 1934 and L. minimum 470 PALAEONTOLOGY, VOLUME 17 Stuckenberg, 1904) have compound pseudocolumellae, which seems rather sur- prising. Ontogenetic studies suggest that the most primitive forms should have simple pseudocolumellae. (c) More complicated types of pseudocolumellae are known in the Upper Carboniferous and Permian species. No succession of develop- ment of a particular type of pseudocolumella can be traced. It is also impossible to point out which type is more advanced. The same specimen can possess several types of pseudocolumellae or different types of pseudocolumellae can appear in the same species. There is no reason to consider the structure of the pseudocolumella as a generic character. Genus timorphyllum Gerth, 1921 Type species. T. wanneri Gerth, 1921. Synonyms. See Schouppe and Stacul 1955, p. 151 ( non Timorphyllum Gerth, 1921 ; Moore and Jeffords 1941). ? Timorphyllum Gerth of Soshkina, 1941. non Timorphyllum Gerth of Fomitshev, 1953a. Diagnosis. Solitary corals with simple or compound, variable pseudocolumella and without dissepimentarium ; minor septa mostly in the form of septal grooves ; cardinal septum extremely short in the early neanic stage; fine structure trabecular; in longi- tudinal section trabeculae arranged fan-like, perpendicular to semicircular growth lines. Remarks. The genus was discussed by Schouppe and Stacul (1955) in great detail. Regarding the micro-structure and fine-structure, the American species T. simu/ans Moore and Jeffords, 1941 (PI. 63, fig. 3), synonymized by Schouppe (1957) with T. wanneri ajermatiensis, has well-developed minor septa, thus negating Moore and Jeffords’s (1941) statement that they are absent. It was impossible to prepare a good longitudinal section through the septum, because the material is sparse (two speci- mens only), poorly preserved, and mostly silicified. However, the arrangement of the growth lines in the septa in a broken specimen is similar to Lophophyllidium , to which this species is now assigned. In Soshkina, Dobroljubova, and Porfiriev 1941, Soshkina described two species from the Lower Permian of the western slope of the Urals and called them Timor- phyllum. According to her, the Ural specimens do not have minor septa. The real generic position of these species (represented each by one specimen only) cannot be decided without careful restudy. Timorphyllum maichense Fomitshev, 1953 from the Doliolina Beds of Far East Asia has all Lophophyllidium- type structures. However, the microstructure has not been investigated. Schouppe and Stacul (1955) emphasized the absence of minor septa in Timor- phyllum, supposing this character to be the main difference between Lophophyllidium and Timorphyllum. From study of many topotypes of the genus Timorphyllum , I concluded that most are actually monoseptal (PI. 61, fig. 5; PI. 62, fig. 13c; PI. 63, fig. 2a, b). However, one specimen (PI. 61, fig. 4a, b ) has the surface of the external wall well preserved. On this surface there are twice as many septal grooves as there are major septa. Thus, Timorphyllum possesses incipient minor septa, but only as septal grooves, which are very easily destroyed during fossilization or weathering. FEDOROWSKI: LO P HO P HY LLIDIU M AND TIMORPHYLLUM 471 The ontogeny of the genus Timorphyllum has never been fully described. Two specimens in the Smithsonian Institution collections in which young parts are pre- served are figured (PI. 62, fig. 13 a-e). The arrangement of septa is more or less similar to that described in Lophophyllidium , but the unusual shortening of the cardinal septum is noteworthy. It is so short in the youngest stage investigated (14 septa at 4-2 mm diameter) that its primary connection with the counter septum is doubtful. The true phylogenetic position of Timorphyllum cannot be determined without complete investigation of its ontogeny. At present one can only say that none of the Lophophyllidium- like species investigated has such a shortened cardinal septum in such a young ontogenetic stage. On the contrary, this septum is mostly connected with the counter during the entire neanic stage in Lophophyllidium. Schouppe and Stacul (1955, text-fig. 21) show clearly the arrangement of septal growth lines in Timorphyllum. This arrangement may be slightly variable in the sense of more or less convexity. Some septa may be so convex that the external parts of the growth lines go down approximately 10 mm before intersecting the external wall. Schouppe and Stacul (1955) accepted Schindewolf’s (1942) concept of the lamellar structure of this type of septum. The present writer cut a few topotypes more or less along the mid-lines of the septa and discovered well-developed trabeculae (PI. 63, fig. 1 ) crossing few of the growth lines and arranged fan-like. The arrangement of septal growth lines and trabeculae, together with the ontogeny, are the main dif- ferences between the genera Timorphyllum and Lophophyllidium. The microstructure was also investigated in cross-section. Photography by transparent light and by scanning electron microscope gave similar results (PI. 66, figs. 2, 3), and showed that the trabeculae are arranged almost at 90° to the mid-line of the septum, producing a structure similar to that of zigzag carinae. This structure is visible only in the middle part of the septum and is completely covered by the secondary stereoplasmatic sheets. The information given above permits clarification of some aspects of the morpho- logical status of the genus Timorphyllum and a comparison with Lophophyllidium. Timorphyllum differs from Lophophyllidium in its ontogeny and microstructure. Individual variability in Timorphyllum appears to be very great. The pseudocolumella and the axial structure are especially variable. Some topotypes have both a solitary pseudocolumella and Verbeekiella-\ike axial structure in different parts of their growth. Moreover, the change from one structure to another is not related to change in growth stage. It seems necessary to check this character as well as the relationship between Verbeekiella and Wannerophyllum, which differ only by the presence or absence of minor septa. The example of Timorphyllum shows that this character is very easy to miss. CONCLUSIONS 1. The genus Lophophyllidium seems to be one of the most variable and widely stratigraphically and geographically distributed tetracoral genera. 2. The ontogeny of Lophophyllidium , with zaphrentoid arrangement of septa, elongated counter septum (and, during most of ontogeny, also the cardinal), can be compared with and related to corals of the Fasciculophyllum omaliusi group. The lack of a ‘ Calophyllum stage in ontogeny seems to be one of the most important differences between this genus and Soshkineophyllum. 472 PALAEONTOLOGY, VOLUME 17 3. The structure of the pseudocolumella, used hitherto as a generic character, is variable even in the same specimen. The genus is not well known in the Lower and Middle Carboniferous, and the absence of corals with Stereostylus- type pseudo- columella at that time does not mean that it appears only in the Upper Carboniferous. Even so, other modifications of the pseudocolumella occur at that time and tran- sitional stages can be found. 4. The fine- and microstructure of Lophophyllidium and Timorphyllum is trabecular. The arrangement of trabeculae and the type of growth of the septa are different and should be used as generic distinctions. 5. The genus Timorphyllum is very similar to Lophophyllidium when only the macro- structure is considered. The microstructure and ontogeny indicate that they are not synonyms but homeomorphs. 6. Timorphyllum, in contrast to Lophophyllidium, seems to be endemic and characteristic only of Timor Island and Western Australia. Acknowledgements. I am particularly grateful to Drs. G. Arthur Cooper and Richard E. Grant for pro- viding Glass Mountains specimens for study, to Drs. William A. Oliver, Jr. and William J. Sando for helpful discussions and for correcting the manuscript, Dr. A. J. Rowell for lending me type material of the Lophophyllidium- like genera for restudy, and to Dr. N. P. Vassiljuk for information about the distri- bution of undescribed species of Lophophyllidium in the U.S.S.R. This study was made possible by a Smithsonian Institution fellowship. REFERENCES brown, t. c. 1909. Studies on the morphology and development of certain rugose corals. Ann. New York Acad. Sci. 19, 45-97. carruthers, R. G. 1913. Lophophyllum and Cyathaxonia : Revision notes on two genera of Carboniferous corals. Geol. Mag. 10, 49-56, pi. 3. duerUen, j. e. 1906. The morphology of the Madreporaria, VIII. The primary septa of the Rugosa. Ann. Mag. Nat. Hist. (7), 18, 226-242. duncan, H. 1962. Phylum Coelenterata, in Uppermost Pennsylvanian and lowermost Permian rocks in Kansas. U.S. Geol. Survey, Prof. Pap. 323, 64-67. flugel, H. w. 1972. Die palaozoischen Korallenfaunen Ost-Irans. 2. Rugosa und Tabulata der Jamal- Formation (Darwasian?, Perm.). Jb. geol. Bundesanst. 115, 59-102, pis. 1-6. fomitshev, v. d. 1953u. Korally Rugosa i stratigrafija sredne i werchnekamennougolnych i permskich otlozenji Doneckogo basseina. Trudy Vsesojuz. Geol. Inst. 1-622, pis. 1-44. — 19536. Permskie korally dalnego Vostoka. Ibid. 1-55, 7 pis. fontaine, h. 1961. Les Madreporaires Paleozoiques du Viet-Nam, du Laos et du Cambodge. Arch. geol. Viet-Nam. 5, 1-276, pis. 1-35. — 1964. Madreporaires Paleozoiques du Viet-Nam, du Laos, du Cambodge et du Yunnan. Nouvelles determinations et notes bibliographiques. Ibid. 6, 75-90, pi. 8. gerth, H. 1921. Die Anthozoen der Dyas von Timor. In wanner, j. Paldontologie von Timor. 9, pt. 16, 67- 147, pis. 145-150. Stuttgart. grabau, a. w. 1 922. Palaeozoic corals of China. Part I. Tetraseptata. Palaeont. Sinica( B), 2, pt. 1 , 1 -76, pi. 1 . — 1928. Palaeozoic corals of China. Part I. Tetraseptata. II. Second contribution to our knowledge of the Streptelasmoid Corals of China and adjacent territories. Ibid. 2, pt. 2, 1-175, pis. 1-6. HERITSCH, F. 1936. Lophophvllum, Lophophyllidium und Sinophyllum. Zbl. Mineral. Geol. Palaont. (B), 408-415. hill, d. 1 956. Rugosa, in moore, r. c. (ed.). Treatise on Invertebrate Paleontology, F, Coelenterata. Pp. 233- 324, Geol. Soc. America and Kansas Univ. Press. huang, t. k. 1932. Permian corals of Southern China. Palaeont. Sinica (B), 8, 1-163, pis. 1-16. FEDOROWSKI: LO P HO P HY LLIDIU M AND TIMO RPHYLLU M 473 HUDSON, R. G. s. 1942. Fasciculophyllum Thompson and other genera of the Zaphrentis omaliusi group of Carboniferous corals. Geol. Mag. 79, 257-263. ilina, T. G. 1965. Chetyrekhluchevye korally pozdnoj permi i rannego triasa Zakavkazja. Trudy paleont. Inst. 107, 1-104, pis. 1-20. ivanovsky, A. b. 1967. Etjudy o rannekamennougolnych rudozach. Moskva, 92 pp., 22 figs., pis. 1-22. jeffords, m. r. 1942. Lophophyllid corals from Lower Pennsylvanian rocks of Kansas and Oklahoma. Bull. geol. Survey Kansas , 41, 185-260, pis. 1-8. 1947. Pennsylvanian Lophophyllidid corals. Paleont. Contr. Univ. Kansas , 1, 1-84, pis. 1-28. kato, M. 1963. Fine skeletal structure in Rugosa. J. Fac. Sci. Hokkaido Univ. (4), 11, 571-630, pis. 1-3. kostic-podgorska, v. 1955. Unterkarbonische Korallen aus dem Palaozoikum des Sanagebietes (Bosnien). Rec. trav. geol. Inst. 8, 169-177, pis. 1-4. (In Serbian with German summary.) lecompte, m. 1955. Note introductrice a la revision du genre Lophophyllum M. E. & H. Publ. Ass. Etud. Paleont. (21), 8, 401-414, pis, 1, 2. mcchesney, J. H. 1860. Descriptions of new species of fossils from the Paleozoic rocks of the Western States. Chicago, 96 pp. pis. 1-11. milne-edwards, h. m. and haime, j. 1850. A monograph of the British fossil corals. Paleontogr. Soc. [Monogr.]. minato, m. 1955. Japanese Carboniferous and Permian corals. J. Fac. Sci. Hokkaido Univ. (4), 9, 1-202, pis. 1-43. moore, R. c. and jeffords, r. m. 1941. New Permian corals from Kansas, Oklahoma and Texas. Bull. geol. Surv. Kansas, 38, 65-120, pis. 1-8. 1945. Description of Lower Pennsylvanian corals from Texas and adjacent States. Univ. Texas Publ. 4401, 77-208, pis. 1-14. oekentorp, k. 1972. Sekundar-strukturen bei palaozoischen Madreporaria. Forsch. geol. Palaont. Munster , 24, 35-108. okulitch, v. j. and Albritton, c. c., jr. 1937. Malonophyllum, a new tetracoral from the Permian of Texas. J. Paleont. II, 24-25, pi. 4. pyzhjanov, i. v. 1966. Nekotorye predstaviteli Rugosa iz niznepermskich otlozenji Severnogo Pamira. Trudy Upr. geol. Tadzh. SSR, 2, 265-297, pis. 1-6. rowett, c. l. and Sutherland, p. k. 1964. Wapanucka rugose corals. Bull. geol. Survey Oklahoma , 104, 1-124, pis. 1-9. schindewolf, o. H. 1942. Zur Kenntms der Polycoelien und Plerophyllen. Abh. Reichsamts Bodenforsch. (n.f.), 204, 1-324, pis. 1-36. schouppe, a. v. 1957. Beitrage zur Palaontologie des Ostindischen Archipels. 12. Zwei Pterocorallia aus dem Perm von Portugiesisch Timor. N. Jb. Geol. Palaont. Abh. 104, 359-381, 13 pis. and STACUL, p. 1955. Die Genera Verbeeliella Penecke, Timorphyllum Gerth, Wannerophyllum n. gen., Lophophyllidium Grabau aus dem Perm von Timor. Palaeontographica , Suppl. 4 (5), 95-196, pis. 7-8. scrutton, c. t. 1971. Palaeozoic coral faunas from Venezuela, I. Silurian and Permo-Carboniferous corals from the Merida Andes. Bull. Brit. Mus. nat. Hist. Geol. 20, 5, 183-227, pis. 1-5. soshkina, e. d. 1928. Niznepermskie (artinskie) korally zapadnogo sklona Severnogo Urala. Bjull. mosk. Obsc. Ispyt. Prir. (Otd. geol.), 6, 3-4, 337-393, pi. 12. dobroljubova, t. a. and porfiriev, G. s. 1941. Permskie Rugosa evropeiskoj chasti SSSR. Paleont. SSSR , 5, 1-304, pis. 1-63. stuckenberg, a. a. 1904. Korally i msanki niznego otdela srednerusskogo kamennougolnogo izvestnjaka. Trudy Geol. Kom. (n.s.), 14, 11 09, pis. 1-9. termier, g. and termier, h. 1950. Paleontologie Marocaine II. Invertebres de l’Ere Primaire. Foraminiferes, Spongoaires et Coelenteres. Serv. Mines Carte Geol. Maroc. Notes et mem. 73, 1-220, pis. 1-51. vojnovsky-krieger, k. g. 1934. Niznekamennougolnye korally iz okrestnostej Archangelskogo zavoda na zapadnom sklone Juznogo Urala. Trudy Vses. geol.-razv. objedin. 107, 1-64, pis. 1-4. wang, h. c. 1947. Notes on some Permian Rugose corals from Timor. Geol. Mag. 84, 334-344, pi. 9. 1950. A revision of the Zoantharia Rugosa in the light of their minute skeletal structures. Phil. Trans. roy. Soc. London , (B), 234, 175-246, pis. 4-9. JERZY FEDOROWSKI Palaeozoological Institute of the Polish Academy of Sciences Poznan Branch Typescript received 7 August 1973 Poznan, ul. Mielzynskiego 27/29 C FORAMINIFERAL BIOSTRATIGRAPHY OF THE OLIGOCENE - MIOCENE LIMESTONES OF CHRISTMAS ISLAND (INDIAN OCEAN) by c. G. adams and d. j. belford Abstract. Foraminifera indicative of the Tertiary Lower e, Upper e, and Lower /‘Stages’ of the East Indian Letter Classification are recognized in the post-Eocene limestones of Christmas Island. The local ranges of Spiroclypeus globulus Nuttall (here regarded as a junior synonym of S. margaritatus), Miogypsina neodispansa (Jones and Chap- man), Lepidocyclina ( Eulepidina ) ephippioides (J. and C.), and L. ( E .) andrewsiana (J. and C.) — for all of which this small island is the type area— are determined. Five faunal assemblages are recognized, and one new species, Hetero- stegina barriei, described. Christmas Island lies almost 320 km south of Java and has an area of about 140 km2. It is basically a truncated volcanic cone, capped with about 190 m of mainly flat-lying Cenozoic limestones, rising some 2450 m from the floor of the eastern part of the Indian Ocean. The island is heavily forested, the best natural exposures of limestone being in the steep inland cliff's. The area was active tectonically throughout Tertiary times, and the sedimentary succession is much affected by faulting. The geology was originally described by Andrews (1900), the only recent accounts being a paper by Trueman (1965) and an unpublished report by Barrie (1967) for the British Phosphate Commissioners. The Tertiary foraminifera of the island were first described by Jones and Chapman (in Andrews 1900) on the basis of 58 poorly localized rock samples which failed to yield a recognizable faunal sequence owing to the uncertainty of their stratigraphical relationships. However, a number of new species were described, the most important being Lepidocyclina (E.) andrewsiana , L. ( E .) ephippioides, L. (E.) murrayana, L. insulaenatalis, and Orbitoides ( Lepidocyclina ) neodispansa', some of these names subsequently came into general use throughout the region. The main part of the lime- stone was thought to be of early Miocene age, but of the few samples which appeared to have been taken from at or near the base of the succession in the area of Flying Fish Cove, no. 595 contained planktonic foraminifera (including Orbulina) now known not to occur below Blow’s zone N. 9 (approximately base of Lower /), while samples 924 and 220 yielded Miogypsina, a genus which appears first in strata of Upper e age in Indonesia. The faunas from these three samples therefore conflicted with the evidence for the age (Lower e ) of the overlying beds. Nuttall (1926) revised the orbitoids with the aid of additional sections cut from Andrews’s samples, and cleared up some of the confusion caused by the poor original descriptions. However, Tan (1936) noted that the published descriptions of Miogypsina neodispansa were still inadequate for the satisfactory establishment of its systematic position. Ludbrook (1965) described the fauna from 22 isolated samples collected from different parts of the island. Although Nuttall and Ludbrook both contributed to our knowledge of the faunas, neither had access to material which would have enabled them to solve [Palaeontology, Vol. 17, Part 3, 1974, pp. 475-506, pis. 71-74.] 476 PALAEONTOLOGY, VOLUME 17 the basic stratigraphical problems. In 1965 Mr. J. M. Barrie, then with the Common- wealth Bureau of Mineral Resources, carried out a geological survey of the island for the British Phosphate Commissioners; material collected by him showed con- clusively that the base of the ‘Miocene’ limestones was older than insueta Zone, but otherwise added little to our knowledge of the island’s stratigraphy. One of us (D. J. B.) therefore visited the island in 1967 and collected more than 200 samples in stratigraphical order along the five traverses shown on text-fig. 1. Although the results presented here are based primarily on this material, all the previous collec- tions have been re-examined and evaluated in interpreting the biostratigraphy. Of the three anomalous samples collected by Andrews and referred to earlier, that containing Orbulina (595) was undoubtedly obtained from one of the plankton-rich fissure infillings which occur in Flying Fish Cove. The occurrence of Miogypsina neodispansa in samples 220 and 924 is more difficult to explain. Mr. P. J. Barrett and Mr. D. A. Powell recently collected further material (G. 840; G. 852-860) from a poorly exposed yellowish limestone in contact with the basalt at the point where Andrews obtained rock no. 924. Nine of these samples yielded M. neodispansa', the fauna of the yellowish limestone cannot be traced laterally, and two samples collected immediately above (G. 861-862) contain a Lower e stage fauna. Mr. Barrett con- siders the yellowish limestone to be different from others in the area, and regards it as being formed also as a result of fracture filling, perhaps of a tension fracture which was gradually opening as the island underwent adjustment. This seems to be the only explanation for a younger limestone directly against the basalt with an older limestone at a higher level. It is unfortunate that the limestone occurs at the northern end of a fault zone where the succession is obscure. Sample 220 is probably from the same locality. Throughout this paper the East Indian Letter Stages are used in the sense of Adams (1970). Table 1 equates the Oligocene and early Miocene parts of this classifi- cation with the current European terminology. table 1. Approximate correlation of the mid- Tertiary Letter Stages with the conventional stage terminology of Europe. (Only part of the / stage is shown here.) Epoch European Stage Letter Stage M ID z LD U 0 1 E Serravall ian Langhian Burdigalian Aquitanian f Upper e (=e5) UJ L z Cl O o 3 M O E Chattian Lower e ( — e 1 - 4 ) Rupelian d Latto rfian c ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 477 text-fig. 1. Map of Christmas Island showing principal traverses and other important sample localities. 478 PALAEONTOLOGY, VOLUME 17 LITHOLOGY, STRATIGRAPHY, AND FAUNAL SUCCESSION The greater part of the succession is made up of foraminiferal and algal debris in a matrix of carbonate mud. Molluscs, corals, and coral debris are present in many samples, but only in ‘G’ Traverse is there direct evidence for a coral reef. The lime- stone contains few planktonic foraminifera except in fissure infillings. Assemblages of miliolids and peneroplids known to be characteristic of shallow-water sheltered environments (e.g. lagoons) occur at many levels, whereas foraminifera believed to be typical of higher energy environments are virtually restricted to Assemblage 1. The post-Eocene limestones are seen to rest on basalt or tuff wherever they are exposed, although the contact is usually obscure; in some places it is certainly a fault plane but in others may be an erosion surface with a limestone breccia or con- glomerate above. In the traverses detailed below, the base of the limestone rests on basalt. Some traverses have been named after localities in which they occur; others are differentiated by letters assigned by Barrie (1967). The discontinuity of sample numbers in the traverses resulted either from difficulty of access, particularly in lD’ Traverse, so that several visits were necessary at different times, or from further collecting in order to check particular parts of traverses after preliminary examination of samples on the island. On discovering that the faunal sequence was difficult to interpret, we decided to take no risks with the identification of species thought to be of age-diagnostic value. The distribution charts (text-figs. 3, 5, 7, 9, 10) are, therefore, more complex than usual. Whenever we have been unable to see the diagnostic characters of a species (a common situation when random thin sections of limestone are studied), we have recorded it simply as ‘X’ sp. Thus, in ‘D’ Traverse (text-fig. 3) we recognize Mio- gypsina ( Miogypsinoides ) bantamensis, M. ( Miogypsinoides ) complanata, M. ( Mio - gypsinoides) cf. complanata , and M. ( Miogypsinoides ) sp. This, we believe, fairly reflects the difficulties encountered in distinguishing individual species when the critical characters vary in the degree to which they are visible. "D" Traverse (text-fig. 2). A stream traverse on the north side of the island immediately south-west of Flying Fish Cove. Fifty samples were available from the south side of the stream and nine from the north side. A further 15 samples had been collected previously by Barrie. The sequence here has a vertical thickness of about 165 m. The lower part of the sequence is composed entirely of skeletal calcarenites and lime- stone breccias, the faunas of which show no signs of reworking. Above sample 1170 the limestones are calcarenitic micrites or true skeletal calcarenites. Calcareous algae are common throughout. The lowermost 73 m of rock contains a fauna which includes Lepidocyclina (. Eulepidina ) spp., Miogypsina ( Miogypsinoides ) bantamensis, Spiroclypeus margari- tatus. Sorites cf. orbiculus , Borelis spp., and rare specimens of Austrotrillina striata. From about 122 m the hillside is covered with limestone boulders and basalt rubble, no exposure of solid rock being visible. Thin sections cut from several of the boulders revealed a fauna similar to that occurring at higher levels in the succession, and it seems probable that the rocks have fallen more than 45 m to their present positions. At about 114 m the sequence is faulted, about 4-5 m of Eocene limestone being intro- duced at this level. These Eocene beds are terminated abruptly by basalt, which ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 479 ‘O' TRAVERSE LEGENO P- Present X=Rore (<3 specimens) 0= Few (3-10 •• ) #- Common (11-25 " ) ■ ^Abundant (>25 •• ) ON [ Si? 1345 1346 5 1348 8 8 s e s S S xt r~- £ SI o $ $ s 1 1 1 § ’T O 1 I s § 1 S § 1 1 1^53 1 1 i § s 1031 Sj s s s FEOI 1 o | 1080 o S s s s ACERVULINA, BORODIN /A , KANAKA !A spp P P P P P p p P p P P p p p p p p ? ? p p p CARPENTERIA, SPORADOTREMA spp P p p p p p P p p p p P P p p OPERCUUNA sp X O LEP/DOCYCL/NA (E) sp x x X ? ? X ? X x x x O x x x X X X x x X x o x o x X x x x x n a ■ CYCL0CLYPEU5 cf El DAE X x x ? ? ? X X ? • O M/UOUDS ■ • o X • O X o • o o o • o • X o o o o O O O AM PH IS TEGINA spp X o • o Q O • o • 0 a GYPS/NA GLOBULA X X X x x X X x x X X X 0 X X X X x x x BO REUS PYGMAEUS X X X X X X X X X X o o o x O o x X X O BO RE US spp X X o X X x X o o o X X x x SORITES cf ORBICULUS X X o x X o X o x x x X x o o o o x X o x x k x X x G o o o o x X HETEROSTEG/NA BARRIEI • a • • ■ X • ■ ■ ■ • • o • a • • O • ■ o X ? ? ? • x ? HETEROSTEGINA sp X 0 SPIROCLYPEUS MARGAR/TATUS • ■ • • o o x x • • o o o o o o o o M ( M/OG YPS 1 NO IDES) BANTAMENSIS X ■ • o o C.) • • o o X • • • o o • • o O o o a o LEPIDOCYCLINA (E) EPHIPPIOIDES p p ! X X o X o o X LEPIDOCYCLINA sp (B form) p p P p p P p p P P p p p p P M (MIOGYSINOIDES) cf COMPLANATA ? x X x ? , ? t r SPIROCLYPEUS MARGAR/TATUS (“GLOBULUS type) ■ ■ • • ? ■ X • • □ ■ • . ■ ■ ■ ■ ■ ■ ■ X AUS TROTRILL IN A STRIATA X X r x X o X x x GYPS/NA DISC A X LEPIDOCYCLINA (N) sp , X T SPIROCLYPEUS sp o o M. (MIOGYPSINOIDES) COMPLANATA ■ M (MIOGYPSINOIDES ) sp x ASSEMBLAGE !• 2 4—- I 4 text-fig. 3. Foraminiferal distribution chart for ‘D’ Traverse. See text-fig. 2 for sample order. 480 PALAEONTOLOGY, VOLUME 17 text-fig. 4. Profile of ‘G’ Traverse. produces a steep slope about 43 m high. The nature of the contact between the lime- stone and basalt is unknown. Immediately above the basalt the slope has been graded for the construction of a pilot washing and screening plant, and new ‘C’ grade calcination plant. The nearest continuous limestone exposure at this level is in a cliff about 10 m high beginning near Flying Fish Cove some 380 m to the north. This may be Andrews’s locality ‘at 500 feet running south from Flying Fish Cove’. If so, his sample 549 (containing Borelis and Eulepidina) must have been taken somewhere in this vicinity. The fauna in this small outcrop is dominated by Heterostegina barriei sp. nov. and Spiroclypeus-, many of the foraminifera appear to be rolled and abraded. Barrie’s sample D. 2 collected from the site of the washing and screening plant con- tains the same fauna, as do several samples collected between 189 and 195 m (i.e. just below the South Point Road). ‘G’ Traverse (text-fig. 4). This, in effect, is a continuation of ‘D’ Traverse to the top of the island, but offset a little to the south. Twenty-seven samples were collected by 'G' TRAVERSE LEGEND P = Present X = Rqre (<3 specimens) O = Few (3-10 specimens) • = Common( 11-25 specimens) ■ = Abundont(>25 specimens) SAMPLE NO 1064 O 1059 1060 9501 1057 1356 1056 1055 ,354 I 1353 I 1352 1 1 1053 1 1052 1076 1 1050 1 1075 | 1074 1 1073 ZL0\ j o SPIROCLYPEUS MARGAR/TATUS • o • O O • • • ■ M/UOL/DS ■ ■ X x X X o O O X O o O • o SORITES cf ORB/CULUS X O • X X X X X x X AUSTROTRILLINA STRIATA X o X X X I HETEROSTEGINA sp. x CARPENTER/A , SPORADOTREMA spp p p P P P P P P p p P P P P BO REUS spp o X X ACE RYU L IN A , BOROD/N/A, KAN AKA! A spp p P p p p p p P AMPHISTEGINA spp n ■ X * X o o X SP/ROCL YPEUS sp. O M (MIOGYPSINOl DES ) DEHA ART! ■ ■ ■ TAYA MAI A MARIANENSIS AUSTROTRILUNA sp X ASSEMBLAGE , p 3 • *■ " text-fig. 5. Distribution chart for ‘G’ Traverse. See text-fig. 4 for sample order. ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 481 Belford and seventeen by Barrie, through a vertical distance of 100 m. The succession is relatively well exposed between the railway line and the ‘C’ Grade access road, but is then rather poorly exposed up to the level of M. dehaarti. At higher levels the outcrop is fairly continuous. The lower and upper parts of the sequence are formed of skeletal calcarenites, but between 220 and 250 m corals and calcareous algae are well developed and may represent a true reef. The microfauna of the lower part of the section is characterized mainly by the presence of Sorites, Austrotrillina striata, and Spiroclypeus margaritatus. Eulepidina and Miogypsina ( Miogypsinoides ) are absent. At about 270 m there is a facies change and M. ( Miogypsinoides ) dehaarti occurs in abundance, accompanied by Carpen- teria and Amphistegina. At this level the rock becomes a true foraminiferal coquina for the first time. The combined thicknesses of lD’ and ‘G’ Traverses total about 270 m, of which at least 4-5 m is Eocene limestone and 44 m basalt. Waterfall Traverse (text-fig. 6). Thirty-seven samples were collected by Belford and Barrie from about 161 m of strata exposed in this traverse on the eastern side of the island. The lithology is remarkably uniform throughout the traverse, the limestone consisting essentially of calcilutitic skeletal calcarenites. Six samples from the lowest 15 m yielded no species of age-diagnostic value. Heterostegina barriei appears in sample 1095 and continues up to the 121m level (sample 1284). The upper part of the traverse is characterized by the occurrence of Tayamaia marianensis, a species which has not been found in situ on the western side. Miogypsina ( Miogypsinoides ) is fairly common throughout the section. A single specimen of Lepidocyclina in 1095 is clearly redeposited. The Waterfall Traverse thus seems to be equivalent to kG’ Traverse plus the top of ‘D’ Traverse (Eocene beds excluded) on the north-west side of the island. text-fig. 6. Profile of Waterfall Traverse. 482 PALAEONTOLOGY, VOLUME 17 WATERFALL TRAVERSE LEGEND P= Present x=Rore (<3 specimens) 0= Few (3-10 • ) •= Common (11-25 ■ ) ■ = Abundant (>25 * ) SAMPLE No 1 1 1 g 1 i 8 1 1 s 1 1 o Z82I g g 1 g 8 8 2.601 s s 8 1 g s g 1 1 8 a $ 8 CARPENTER! A, SPORADOTREMA spp P p p p p p P p p P P p p p p p p P P P p p P P P AMPHISTEGINA spp • • o • • • X X o O • • • ■ O a GYPS! N A GLOBULA X X X X X X X O ? MIL! OLIOS • X ■ ■ X o o X o X X X o o o ■ • • ■ • O BOR ELIS spp X X X X X X SORITES cf ORBICULUS X X X o X t X X o X X X X X O M (MIOGYPSINOIDES) DEHAARTI ■ a ■ ■ O M (MIOGYPSINOIDES) cf BANTAMENS/S ? X ACERVULINA, BORODINIA, KANAKA! A spp. P p p e ? p p p p p p LEPIDOCYCLINA sp (B form) X SPIROCLYPEUS MARGAR/TATUS (GL°BpgUS) • X X ■ ■ • ■ X AUSTROTRILLINA STRIATA X o X X X X HETEROSTEG/NA BARRIEI X X X ? • ■ a • BOREL/S PYGMAEUS X X M (MIOGYPSINOIDES) cf DEHAARTI o o ■ X o O o SPIROCLYPEUS MARGAR/TATUS X X X X X X X o HETEROSTEG/NA sp X TAYAMAIA MARIANENSIS • • X o o • ? LEPIDOCYCLINA (N) sp ’ ASSEMBLAGE « 3 ► * 2 ► text-fig. 7. Foraminiferal distribution chart for Waterfall Traverse. See text-fig. 6 for sample order. Ross Hill Traverse (text-fig. 8). Sixty-one samples were collected by Belford and Trueman from the traverse, which covers about 200 m of limestone on the eastern side of the island. The twelve samples taken from the lowermost 91 m came from lime- stone blocks in basalt rubble and were not necessarily in situ, although their faunas suggest that they are not far out of position. The greater part of this sequence consists of fine- to coarse-grained skeletal calcarenites with varying amounts of calcilutite and secondary calcite. The small ‘steps’ in the profile above 130 m probably repre- sent faults, and in this connection it should be noted that Miogypsina ( Miogyp - sinoides) dehaarti occurs in sample 1304 only 6 m above M. ( M .) complanata in 1303 (see p. 496). The lowermost 80 m is characterized by an association of Spiroclypeus and Hetero- stegina barriei. This is followed at 1 40 m (sample 1 304) by Miogypsina ( Miogyp sinoides ) dehaarti and at 164 m by Tayamaia marianensis. At 213 m (sample 1228) M. ( Mio- gypsina) cf. neodispansa occurs. However, it disappears again 7-5 m higher in the sequence. Austrotrillina howcliini occurs in sample 1317 (220 m) and continues almost to the top of the section, where Flosculinella bontangensis appears. A. howchini and F. bontangensis have not been seen in the same samples, and the latter has not been found associated with any other age-diagnostic species. Sydney s Dale Traverse. The 21 samples from this stream traverse on the western side of the island represent about 60 m of limestone consisting of limestone breccias and calcarenitic muds. The lower half of the sequence is quite well exposed and is represented by 16 samples, at least 4 of which (between 1244 and 1246, see ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 483 ROSS HILL TRAVERSE LEGEND P- Present X = Rore (< 3 specimens ) 0= Few (3-10 " ) • = Common (11-25 « ) ■ = Abundontj >25 « ) fi? o 5 2 o> 320 1239 240 J « JO * K CD g s CD g r~ JO s tfj ■V a a o o. 2 2 u, po OJ g 1309 6 g O 1304 s o 3 o 1208 - o § g. a CARPENTERIA, SPORADOTREMA spp p p p p p p P P p p p p p p P p p p p p p p p p P p P p p P P p p p p p p p p P P P p p p p p AM PH IS TEG/NA spp ’ O oo • O o x o • o o o o • • • o o x x o X o o o • X • o o o o o X ■ B O O X O X o o X ■ M(MIOGYPSINCIDES) sp x ■ o • ■ ■ • • • o o o o o X HETEROSTEG/NA BARRIEI ■ • MIL! OLIOS ■ x • s a o ■ o x ■ • ■ O o o o • o • x o o o o x O o x • o ■ o x o O • o • x SPIROCLYPEUS MARGARt TA TUS (G^°^pgUS) O o o o M(MIOGYPSINOIDES) BANTAMENS/S ’ ’ ’ > > X > p M(MIOGYPSINOIDES) OEHAARTI o p ■ p p p p p p p p X o B p ’ ACERVUUNA, BORODINIA, KANAKA/A spp p p p p p p p p p P p p p p p P p p p p p p p p p p p p p p p p p P p p GYPSINA GLOBULA X X X B OREL IS spp X X X x x x x x x x - x x X X X SORITES cf ORBICULUS x x o O X O o o x o x o x x x o o o o o • x x X X x x x o SPIROCLYPEUS sp o "ROTA LI A" sp ffl ■ • o AUSTROTRILLINA STRIATA x x x x o X M (MIOGYPSINOIDES) COMPLANATA O o BORE LIS PYGMAEUS X X O X X TAYAMA/A MARIANENSIS x x x o ■ X • o • o x ■ o ■ • o • • o PE/VEROPL/OS O o MARGINOPORA VERTEBRAUS o x X ? X OPERCULINA sp X M(MIOGYPSINA) cf NEODISPANSA o x o AUSTROTRILLINA HOWCHINI o • • * o FLOSCUL !NE LLA BONTANGENS/S O x X x ? ARCH A! AS sp X I - D ASSEMBLAGE L 5 ► 4 4. 2 . text-fig. 9. Foraminiferal distribution chart for Ross Hill Traverse. See text-fig. 8 for sample order. 484 PALAEONTOLOGY, VOLUME 17 SYDNEY'S DALE TRAVERSE LEGEND P= Present X= Rare (< 3 specimens) 0= Few (3-10 » ) •= Common (ll_25 " ) ■= Abundant(>25 " ) SAMPLE No CO rO rO ro rO OJ CM CM 1246 CM (M 1245 CM 3 CM CM CM 1244 CARPENTER! A , SPORADOTREMA spp P P P P p P P P P P P AMPH/STEG/NA spp • • o X o o OPERCUL/NA sp. X X X X SORITES cf. ORB/CULUS X • X o X M ( MIOG YPSINOIDES) cf. BANTAMENS/S X X o X o AUSTROTR! LUNA STRIATA X X X M.( MIOG YPS! NO ID ES) DEHAARTI ■ o • tayamaia MARIANENSIS X X X X o DISCOCYCLINA spp X X X X "ROT A LI A" spp a ■ X o • s o • o O 19 • a BO RE LIS spp t X X X SP/ROCLYPEUS sp X X X X X o MILIOLIDS o • ■ GYPS! N A GLOBULA X ? X X X X • X X • ■ o X LEPIDOCYCUNA (E.) sp. o HETEROSTEG/NA cf. SA/PANENS/S ? X o GYPS/NA D/SCA X ACERVUL/NA -, BORODINIA, KANAKA! A spp P p p p LEPIDOCYCUNA (N.) sp X M. ( MIOG YPS/NO/DE S) cf. DEHAARTI o X X X M. (MIOGYPSINA) cf. NEOD/SPANSA X X BORE LIS PYGMAEUS X X X M. (MIOGYPSINA) sp. X X ASSEMBLAGE U 3 -I text-fig. 10. Foraminiferal distribution chart for Sydney’s Dale Traverse. Topographically highest sample on the left, lowest on the right. text-fig. 10) contain reworked Eocene foraminifera. These, however, are never numerous. The occurrence of Miogypsina ( Miogvpsinoides ) dehaarti and Tayamaia marianensis in the lowest sample (1244) indicates that no part of the sequence is older than Assemblage 3. The frequency of occurrence of each species on the distribution charts (text-figs. 3, 5, 7, 9, 10) has been determined as the maximum number observed in any one thin section from a sample. Because of their large size, B forms of Lepidocychna (. Eulepidina ) in the ’D’ Traverse are recorded only as present. Encrusting genera are similarly treated since individual specimens tend to be badly fragmented, rendering counts meaningless. ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 485 AGE OF THE FAUNAS Five faunal assemblages have been recognized within the post-Eocene limestones. Assemblages 1 and 2 probably being in part laterally equivalent (see below): 1. Lepidocyclina ( Eulepidina ) ephippioides/M. ( Miogypsinoides ) bantamensis/ Spiroclypeus margaritatus Assemblage. These species form the bulk of this Assemblage which is seen only in kD’ Traverse. Age: Lower e. 2. Heterostegina barriei/ Spiroclypeus margaritatus Assemblage. Well developed in ‘D’ Traverse and also seen in the Waterfall Traverse. Age : Lower e, on the presence of M. ( Miogypsinoides ) complanata in a few samples. 3. M. ( Miogypsinoides ) dehaarti/Tayamaia marianensis Assemblage. Seen in the Waterfall and Ross Hill Traverses where it is well developed; also occurs in Sydney’s Dale. Age: Upper e. 4. Miogypsina/A. howchini Assemblage. Seen in stratigraphical sequence only in the Ross Hill Traverse, although isolated samples are known from the vicinity of Flying Fish Cove and Sydney’s Dale. Age: Upper e. 5. Austrotrillina howchini / Flosculinella bontangensis Assemblage. Believed to be restricted to a thin zone at the top of the succession and so far observed only in the upper part of the Ross Hill Traverse. Age: Lower /. These faunas tend to grade into one another and their constituent species are not necessarily mutually exclusive. Hence, M. (M.) bantamensis and Spiroclypeus margaritatus may be found in Assemblage 3, as may Austrotrillina howchini. Long- ranging species such as Gypsina globula (Reuss), Borelis pygmaeus Hanzawa, and Sorites cf. orbiculus (Forskal) occur throughout the greater part of the succession. The faunal sequence shows certain peculiarities. Occurrences of M. (Miogyp- sinoides) complanata in sample 64 (near Jedda Cave), samples 1302 and 1303 (Ross Hill Traverse), and D. 3 in the ‘D’ Traverse, either within or above the range of M. ( M .) bantamensis , are anomalous. In each sample, all the numerous specimens of miogypsinids appear to have long nepionic spires, thus ruling out the possibility that we are dealing merely with a few reworked individuals. It might be argued that sample 64 has been raised to its present high position by faulting, but this explanation will not suffice for samples 1302 and 1303, sample D. 3, and 135B (Dolly Beach), in each of which M. ( M .) complanata occurs with foraminifera typical of Assemblage 2. Although the oldest Tertiary e limestones appear to be those in the lower part of ‘D’ Traverse on the north-west side of the island, they have evidently reached their present position by faulting. They are nowhere seen immediately beneath the Assemblage 2 faunas in a continuous section. There is, therefore, a definite possibility that Assemblage 1 may be partly or entirely the lateral equivalent of Assemblage 2, the composition of the faunas reflecting differences in the local environment of deposition rather than any significant stratigraphical change. However, the occur- rence of M. ( M .) complanata with Spiroclypeus and Heterostegina barriei at Dolly Beach, Ross Hill, and in the ‘D’ Traverse strongly suggests that the lower part of Assemblage 2 is slightly older than the lowest beds yielding Assemblage 1 in ‘D’ Traverse. This interpretation would explain the occurrence of rolled and abraded specimens of Eulepidina (nearly always microspheric forms) in Assemblage 2, and the absence or rarity of encrusting genera in Assemblage 1. 486 PALAEONTOLOGY, VOLUME 17 Lower e is now believed (Adams 1970) to be equivalent to the upper Oligocene (Chattian) of Europe, Upper e to the lower Miocene (Aquitanian and Burdigalian in the type areas), and Lower/ to the early middle Miocene. The minimum thickness of post-Eocene limestone on the island (assuming that Assemblages 1 and 2 are laterally equivalent) is about 190 m. If they are not equiva- lent, the figure increases to 265 m. In this connection, it may be noted that a borehole in the South Point area was abandoned while still in limestone at 244 m (pers. comm., D. A. Powell), whereas another hole (no. 14) on the plateau north-east of Smith Point reached basalt at 167 m. A stratigraphic bore ST. 1, Jones Spring, north of the Waterfall Traverse, passed through 23 m of Tertiary e limestone and 55-5 m of basalt before entering an upper Eocene (Tertiary b ) limestone. SYSTEMATIC PALAEONTOLOGY The limestones of Christmas Island are rich in microfossils, which, unfortunately, have had to be examined mainly by means of random thin sections. Only a few oriented sections could be prepared owing to the difficulty of freeing individual speci- mens from the hard rock matrix. This rendered specific determinations difficult, especially for genera such as Miogypsina, Cycloclypeus, and Lepidocyclina, in all of which the nature of the embryonic apparatus is of critical importance. It is usually impossible to determine the range of variation of species seen only in random sections, since two or more species of the same genus may be present in the rock. For this reason no serious taxonomic revisions are attempted here. Synony- mies are restricted to the original description, to previous records from Christmas Island and, where appropriate, to important recent redescriptions. Special mention must be made of the work of Jones and Chapman (1900). These authors based their descriptions on a very small number of thin sections (one or two per sample) and misinterpreted many of the specimens. Not only did they mistake Miogypsina ( Miogypsinoides ) for Heterostegina , and Miogypsina ( Miogypsina ) for Orbitoides ( Lepidocyclina ), but they failed to distinguish between specimens now referred to Spiroclypeus and Lepidocyclina. They also erected new species on shape and size alone, disregarding the possibility that these characters might be highly variable. Nuttall (1926), in revising the orbitoids, corrected most of their taxonomic errors. Figured specimens prefixed CPC are deposited in the Commonwealth Palaeonto- logical Collection, Bureau of Mineral Resources, Canberra, Australia; those pre- fixed P are deposited in the Palaeontology Department, British Museum (Natural History), London, England. Thin sections representative of the samples referred to in this paper are deposited in both Canberra and London. Family miliolidae Ehrenberg, 1839 Genus austrotrillina Parr, 1942 This genus was revised by Adams (1968) and nothing new can be added here. The commonly occurring species on Christmas Island is A. striata , but at high levels in the succession it is replaced by forms transitional to A. howchini. ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 487 Austrotrillina howchini (Schlumberger) Plate 73, fig. 7 1893 Trillina howchini Schlumberger, p. 119, text -fig. 1, pi. 3, fig. 6. 1968 Austrotrillina howchini (Schlumberger); Adams, p. 86, pi. 2, figs. 1-7 ; pi. 6, figs. 1-5, 7. Remarks. Associated with Miogypsina in the upper part of the Ross Hill Traverse. Ludbrook’s record (1965, p. 291) of A. howchini in association with Flosculinella bontangensis was an error; these species have not been observed together either in the original slides from sample P. 33 or in material subsequently obtained from this locality. There is, however, no reason why A. howchini should not be found with F. bontangensis , since their ranges are known to overlap elsewhere in the region (e.g. Australia; Crespin 1955). All the individuals seen so far are fairly primitive forms lacking the greatly thickened wall so characteristic of the end members of the lineage. Austrotrillina striata Todd and Post Plate 73, fig. 6 1954 Austrotrillina striata Todd and Post, p. 555, pi. 198, fig. 9. 1965 Austrotrillina howchini (Schlumberger); Ludbrook, p. 292, pi. 21, figs. 4-6. 1968 Austrotrillina striata Todd and Post; Adams, p. 92, pi. 4, figs. 1-13; pi. 6, fig. 9. Remarks. A. striata occurs at intervals throughout the Tertiary e limestones. Its first occurrence is in sample 1334 (‘D’ Traverse); its last, in sample 1228 (Ross Hill Traverse). At higher levels it is replaced by fairly primitive forms of A. howchini. Family soritidae Ehrenberg, 1839 Genus sorites Ehrenberg, 1839 Type species. Sorites dominicensis Ehrenberg = Nautilus orbiculus Forskal. Sorites cf. orbiculus (Forskal) Plate 74, figs. 2, 10 1775 Nautilus orbiculus Forskal, p. 125. 1965 Sorites martini (Verbeek); Ludbrook, pp. 290-292. 1965 Sorites orbiculus (Forskal); Cole, p. 20, pi. 6, figs. 1-5, 7, 9; pi. 7, figs. 1-8, 10-12; pi. 8, figs. 7-9. 1969 Sorites orbiculus (Forskal); Cole, p. C5, pi. 3, figs. 7, 8, 16; pi. 4, figs. 3-7. Remarks. This long-ranging species occurs throughout the entire Oligocene-Miocene sequence. It is never abundant, random sections usually showing one or two indi- viduals only. Cole (1969) gave reasons for regarding this form as S. orbiculus rather than S. martini, the name applied by most previous authors to Tertiary e specimens. In the absence of good equatorial sections it is impossible to be certain that some specimens do not belong to S. marginalis (Lamarck). Genus marginopora Blainville, 1830 Type species. Marginopora vertebralis Blainville. 488 PALAEONTOLOGY, VOLUME 17 Marginopora vertebralis Blainville Plate 74, fig, 1 1 1830 Marginopora vertebralis Blainville, p. 377. Remarks. Individuals referable to this species occur in a few samples from the upper part of the Ross Hill Traverse in Assemblages 3-5. Unfortunately, they are not numerous and no well oriented sections have been obtained. Family alveolinidae Ehrenberg, 1839 Genus borelis de Montfort, 1808 Type species. Nautilus melo var. B Fichtel and Moll, 1798. Borelis pygmaeus Hanzawa Plate 71, figs. 9-14 1900 Alveolina melo (Fichtel and Moll); Jones and Chapman, p. 255. 1930 Borelis ( Fasciolites ) pygmaeus Hanzawa, p. 94, pi. 26, figs. 14 and 15. 1965 Borelis pygmaeus Hanzawa; Ludbrook, p. 292, pi. 21, figs. 7 and 8. Remarks. This well-known species is common in Assemblages 1 to 3. Small inflated forms of Borelis , very like the Recent B. pulchrus, also occur at some horizons and seem to grade into B. pygmaeus. Cole (1969) referred all such specimens from Midway to B. melo. However, the typical B. melo , from the middle Miocene of the Mediter- ranean region and the Middle East, is a strongly inflated form (usually higher than wide) which shows no axial thickening and tends to develop supplementary chamber- lets ( B . melo curdiea). B. pulchrus and B. pygmaeus always show a tendency towards axial thickening, are rarely, if ever, higher than wide, and do not develop supplemen- tary chamberlets. This is not the place, nor is the present material appropriate, for a EXPLANATION OF PLATE 71 Figs. 1-4. Heterostegina barriei sp. nov. From ‘D’ Traverse; all x 60. 1, paratype A, CPC. 13734, median section, sample 1173. 2, paratype B, CPC. 13735, transverse section, sample 1 180. 3, holotype, CPC. 13733, slightly oblique median section, sample 1168. 4, paratype C, CPC. 13736, median section, sample 1170. Figs. 5-7. Heterostegina cf. borneensis van der Vlerk. All transverse sections, x 10. 5, CPC. 13737, oblique transverse section, sample G837, Flying Fish Cove. 6, CPC. 13738, transverse section, sample G838, Flying Fish Cove. 7, CPC. 13739, transverse section, sample G837, Flying Fish Cove. Fig. 8. Tayamaia marianensis (Hanzawa), CPC. 13740, sample 1216, Ross Hill Traverse, x 30. Figs. 9-14. Borelis pygmaeus Hanzawa. Specimens from ‘D’ Traverse showing variation in shell size and form. 9, CPC. 13741, sample 1356, off-centre, slightly oblique section, x40. 10, CPC. 13742, sample 1 334, slightly off-centre axial section, x 40. 11, CPC. 1 3743, sample 1334, slightly off-centre axial section, x 40. 12, CPC. 13744, sample 1037, axial section, x 48. 13, CPC. 13745, sample 1333, slightly off-centre axial section, x 40. 14, CPC. 13746, sample 1080, off-centre axial section, x 40. Fig. 15. Borelis melo (Fichtel and Moll). Axial section P. 49087, from the Miocene of Turkey, x 50, intro- duced for comparison with B. pygmaeus. Figs. 16- 18. Miogypsina ( Miogypsina ) neodispansa (Jones and Chapman). All from Andrews’s sample no. 220, south side of Flying Fish Cove. 16, P. 49088, transverse section, x 30. 17, P. 49089, oblique trans- verse section, x24. 18, P. 49090, slightly oblique median section showing embryonic apparatus, x 30. PLATE 71 ADAMS and BELFORD, Oligocene-Miocene foraminifera 490 PALAEONTOLOGY, VOLUME 17 revision of the genus Borelis. We are therefore retaining Hanzawa’s specific name, while drawing attention to the similarity between these specimens and the Recent B. pulchrus and B. pulchrus schlumbergeri. The difference between B. pulchrus schlum- bergeri and B. melo is very well illustrated by Reiss and Gvirtzman (1966, pis. 1 and 2). Genus flosculinella Schubert, 1910 Type species. Alveolinella bontangensis Rutten, 1913. Flosculinella bontangensis (Rutten) Plate 74, fig. 3 1913 Alveolinella bontangensis Rutten, p. 221, pi. 14, figs. 1-3. 1965 Flosculinella bontangensis (Rutten); Ludbrook, p. 292, pi. 21, fig. 13. Remarks. This species is known only from the uppermost beds in the Ross Hill Traverse. It has been found associated with Amphistegina, Sorites and encrusting genera. However, Austrotrillina howchini occurs at about the same level and is known to have co-existed with F. bontangensis elsewhere in the region. Flosculinella sp. Remarks. A single individual was seen in Barrie’s sample 69F (Batu Merah, Flying Fish Cove). It is impossible to decide whether this specimen should be referred to F. reicheli Mohler or to F. globulosa (Rutten). However, its occurrence with an Assemblage 2 fauna almost certainly means that the genus can no longer be relied on to mark the base of Upper e. Family nummulitidae Blainville, 1825 Subfamily cycloclypeinae Butschli, 1880 Genus cycloclypeus Carpenter, 1856 Type species. C. mammilatus Carter, 1861. Cycloclypeus cf. eiclae Tan Sin Hok 1932 Cycloclypeus eiclae Tan Sin Hok, p. 50, pi. 5, fig. 6: pi. 12, figs. 2-3; pi. 13, fig. 2. 1965 Cycloclypeus cf. eidae Tan Sin Hok ; Ludbrook, p. 291 . Remarks. This species is rather rare. It occurs in a number of samples from ‘D’ Traverse (Assemblage 1); Ludbrook reported it from sample P. 52, Flying Fish Cove. Genus heterostegina d'Orbigny, 1826 Type species. H. depressa d’Orbigny. Heterostegina barriei sp. nov. Plate 71, figs. 1 -4 1900 Heterostegina depressa d’Orbigny; Jones and Chapman, pp. 244 and 252. Not p. 229, pi. 20, fig. 1. This species is named after Mr. J. M. Barrie, in whose samples it was first recognized. ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 491 Description of holotype. Test small, with evolute primary chambers arranged in 3 rapidly expanding whorls. Proloculus and deuteroconch minute, followed by 5 operculine chambers. Secondary septa long, and well developed from their first appearance. No ornament visible. Dimensions. Diameter 10 mm (test incomplete). Internal diameter of proloculus 0 05 mm; external diameter of proloculus and second chamber 0 091 mm. Variation. Other specimens show that the diameter of the test ranges at least up to 21 mm, although the flange is almost always broken. The number of operculine chambers ranges from 5 to 8, and the number of whorls from 3 to 3j. Secondary septa are always long. Although no ornament has been observed, its presence cannot be entirely ruled out since small pustules do not always show up in random sections. Locality and horizon. Holotype from sample 1 168, ‘D’ Traverse, at base of the lime- stone outcrop at this locality. This species is characteristic of Assemblage 2. Age: Lower e. H. barriei appears to differ from all other described species of the genus in being unusually small (under 2 mm average diameter) and in having a very small embryonic apparatus followed by from 5 to 8 operculine chambers in the megalospheric form. It most closely resembles H. granulatestata subsp. praeformis Papp and Kupper from the middle Miocene of southern Europe, but is smaller and has long secondary septa only. H. suborbicularis d’Orbigny is the commonly reported Tertiary e species in the Indo-Pacific region, but this is involute and has very many more operculine chambers (cf. Cole 1969, pi. 3, figs. 1-5, 18). Heterostegina cf. borneensis van der Vlerk, 1929 Plate 71, figs. 5-7 Remarks. Specimens probably referable to this species have been seen in a few samples (G. 837, G. 838, and G. 862) from Flying Fish Cove. They are two or three times the size of the largest specimens of H. barriei , and the two species have not been seen in association. A positive identification is, unfortunately, impossible in the absence of sections showing the embryonic apparatus and first whorl. Genus operculina d’Orbigny, 1825 Remarks. Specimens occur at intervals throughout the succession. They are rarely numerous and appear to be specifically indeterminable in random sections. It is possible that more than one species is represented. Genus spiroclypeus Douville, 1905 Type species. S. orbitoideus Douville. At least eleven nominal species of this genus have been described from Tertiary e strata in the Indo-West Pacific region, and although attempts have been made to distinguish between them (e.g. Krijnen 1931), authors have found the greatest difficulty in naming specimens satisfactorily. It is undoubtedly significant that no 492 PALAEONTOLOGY, VOLUME 17 one has yet described a succession in which even a few of these species have been shown to succeed one another in time, and as Cole (1969) has pointed out, several authors have found two or three so-called species in the same beds. Cole (1969) therefore assigned seven of the ‘species’ occurring in the Tertiary e rocks of the region to S. margaritatus (Schlumberger). Although first inspection of the present material suggested that several species were represented, closer examination indicated that transitional forms occur. This, together with the absence of any obvious pattern of stratigraphical distribution, strongly suggests that only one species is present despite the wide range of morphological variation. Spiroclypeus occurs abundantly only in ‘D’ Traverse. It is common at some levels in ‘G’ Traverse, but well-oriented individuals have not been seen. It occurs in four samples near the base of the Ross Hill Traverse, but except in 1302, specimens are rare and specifically indeterminable. It is also present in four samples from the succession in Sydney’s Dale. Spiroclypeus margaritatus (Schlumberger) Plate 72, figs. Ill 1900 Orbitoides ( Lepidocyclina ) sumatrensis Brady; Jones and Chapman, p. 244, pi. 20, fig. 6. 1902 Heterostegina margaritatus Schlumberger, p. 252, pi. 7, fig. 4. 1926 Spiroclypeus globulus Nuttall, pp. 36, 37, pi. 5, figs. 5-7, text-fig. 1. 1965 Spiroclypeus globulus Nuttall; Ludbrook, p. 291, pi. 22, fig. 3. 1969 Spiroclypeus margaritatus (Schlumberger); Cole, p. C8, pi. 2, figs. 1-20; pi. 3, figs. 9-14, 19 (synonymy). Remarks. The highly inflated form ( S . globulus of Nuttall) of this species is common to abundant throughout the upper part of ‘D’ Traverse (Assemblage 2). However, it is usually abraded and shows signs of having been rolled and redeposited. The flange is rarely preserved except in the thicker and stronger B’ forms. It occurs also in Assemblage 2 at the base of the Ross Hill Traverse and in Assemblages 2 and 3 at Waterfall. The chief variation in the present material is in the strength of the ornament, the number of lateral layers, and the width of the walls between the lateral chambers. Some highly inflated forms, particularly those of the microspheric generation, seem to possess a single umbonal pillar (PI. 72, fig. 4); however, this effect can also be produced by sections through thickened lateral walls (see PI. 74, fig. 13). EXPLANATION OF PLATE 72 Figs. 1-11. Spiroclypeus margaritatus (Schlumberger). All except fig. 7 from ‘D’ Traverse. 1, CPC. 13747, sample 1172, off-centre transverse section through microspheric form, x 10. Note thick pseudo-pillars and compare with figs. 3 and 4. 2, CPC. 13748, sample 1 177, typical transverse section through inflated form (‘S. globulus ’ of Nuttall), x 20. Most individuals from Christmas Island are of this type. 3, CPC. 13749, sample 1171, oblique transverse section showing pseudo-pillars, x 10. Probably a microspheric form. 4, CPC. 1 3750, sample 1 332, off-centre transverse section, probably through a microspheric form, x 10. Note the massive umbonal pseudo-pillar and compare with Lepidocyclina sp., Plate 74, fig. 13. 5, CPC. 13751, sample 1026, slightly off-centre transverse section through megalospheric form, x 20. 6, CPC. 13752, sample 1169, slightly off-centre transverse section through megalospheric form, x 20. A more typical shape for the species, but not common on Christmas Island. 7, CPC. 13753, sample 1302, Ross Hill Traverse, median section through megalospheric form, x20. 8, CPC. 13754, sample 1172, transverse section, x 20. 9, CPC. 13755, sample 1332, median section, x 20. 10, CPC. 13756, sample 1036, median section, x 20. 11, CPC. 13757, sample 1079, median section, x 20. PLATE 72 ADAMS and BELFORD, Oligocene-Miocene foraminifera 494 PALAEONTOLOGY, VOLUME 17 The only differences between S. leupoldi and S', margaritatus seem to be the inflation of the test and the number of lateral layers. No differences can be seen in equatorial sections obtained from samples near the base and top of ’D’ Traverse (1332-1177). Family miogypsinidae Vaughan, 1928 Genus miogypsina Sacco, 1893 Type species. Nummulites globulina Michelotti, 1841. Subgenus miogypsinoides Yabe and Hanzawa, 1928 Type species. Miogypsina dehaartii van der Vlerk, 1924. Subgenus miogypsinoides Yabe and Hanzawa, 1928 Plate 73, figs. 1 -5 1900 Miogypsina complanata Schlumberger, p. 330, pi. 2, figs. 13-16. 1936 Miogypsinoides ubaghsi Tan Sin Hok, p. 48, pi. 1, figs. 1-7. Remarks. This primitive species is found at only four localities. It occurs in 1302 and 1303, Ross Hill Traverse (Assemblage 2); in Barrie’s samples D. 3 from the D’ Traverse and 1 35B above Dolly Beach (East Coast), each with an Assemblage 2 fauna ; and it forms a true foraminiferal coquinite in Barrie’s sample 64 near Jedda Cave in the centre of the island. Sample D. 3 is accurately located in the sequence ; 64 and 135B are isolated samples which cannot at present be located relative to the local strati- graphical succession. Samples 1302 and 1303 are from a boulder-strewn slope and must be older than other samples collected from boulders occurring at lower levels along the same traverse. There are two possible explanations for this peculiar distri- bution. Either all rocks containing M. ( M .) complanata are up-faulted relative to all those containing M. ( M .) bantamensis or the ability to produce a long periembryonic spire had not been entirely lost by the time M. ( M .) bantamensis evolved. The former explanation is considered to be the most likely since all the specimens seen in each sample appear to have the same grade of structure. Cole (1969) argued that because M. ( M .) dehaarti, M. (M.) lateralis , M. (M.) EXPLANATION OF PLATE 73 Figs. 1-5. Miogypsina ( Miogypsinoides ) complanata Schlumberger. 1, CPC. 13758, Ross Hill Traverse, sample 1302, oblique median section through microspheric form, x 30. 2, CPC. 13759, Ross Hill Traverse, sample 1303, transverse section through megalospheric form, x 30. 3, CPC. 13760, Ross Hill Traverse, sample 1303, median section through megalospheric form, x 30. 4, CPC. 13761, ‘D’ Traverse (Barrie’s sample D3), median section through megalospheric form, x 40. 5, CPC. 13762, same locality as 4, off-centre transverse section, x 40. Fig. 6. Austrotrillina striata Todd and Post, CPC. 13763, Ross Hill Traverse, sample 1303, off-centre transverse section. Fig. 7. Austrotrillina howchini (Schlumberger), CPC. 13764, Ross Hill Traverse, sample 1239, transverse section, x 50. Figs. 8-11. Miogypsina ( Miogypsinoides ) bantamensis Tan Sin Hok, all x 30. All from ‘D’ Traverse. 8, CPC. 13765, sample 1079, megalospheric form in median section. 9, CPC. 13766, sample 1334, off-centre transverse section. 10, CPC. 13767, sample 1079, transverse section of megalospheric form. 1 1, CPC. 13768, sample 1079, median section of megalospheric form. Figs. 12-14. Miogypsina ( Miogypsinoides ) dehaarti van der Vlerk. All from ‘G’ Traverse. 12, CPC. 13769, sample 1061, transverse section through megalospheric form, x 50. 13, CPC. 13770, sample 1059, median section through megalospheric form, x 30. 14, CPC. 13771, sample 1061, slightly oblique median section through megalospheric form, x 30. PLATE 73 ADAMS and BELFORD, Oligocene-Miocene foraminifera 496 PALAEONTOLOGY, VOLUME 17 mauretanicus, M. ( M .) fomiosensis, and M. ( M .) bantamensis could be shown to occur in association in one part of the region or another, only one species should be recog- nized. However, he failed to notice that the gradation is in time rather than in space. Hence, M. (M.) complanata is never found with M. ( M .) dehaarti, whereas either (but not both) may occur with M. ( M .) bantamensis. Variation. Nepionic spire long (16-23 chambers in the A form), test relatively small (up to 1-3 mm in maximum diameter), lateral walls nearly always thin. Miogypsina ( Miogypsinoides ) bantamensis Tan Sin Hok Plate 73, figs. 8-11 1936 Miogypsinoides complanata (Schlumberger) forma bantamensis Tan Sin Hok, p. 48, pi. 1, fig. 13. 1940 Miogypsinoides lateralis Hanzawa, p. 783, pi. 39, figs. 10-14. Remarks. Cole (1969) has argued that M. ( M .) bantamensis is a junior synonym of M. ( M .) dehaarti and that the two cannot be distinguished in the drill holes on Midway Atoll. However, if the principle of nepionic acceleration is valid, M. (M.) banta- mensis must be slightly older than M. ( M .) dehaarti. This is supported by the strati- graphical distribution of these two species on Christmas Island. M. ( M .) bantamensis is common in Assemblage 1 (‘D’ Traverse) and also occurs in Assemblage 3 (Ross Hill and Sydney’s Dale), although the paucity of well-oriented individuals renders identification difficult in many samples. Present evidence suggests that the earliest representative of M. ( Miogypsinoides ) in the Indo-Pacific region (M. ( M .) complanata) gradually underwent a shortening of the periembryonic spire leading to the condition seen in M. ( M .) bantamensis. This process continued until individuals recognizable as M. ( M .) dehaarti were produced. This shortening of the spire was accompanied at first by an increase in the number of equatorial chambers, thus producing a larger test, and later by an increase in the thickness of the lateral walls. The shortening of the spire may have been linked with an increase in the internal diameter (volume) of the proloculus. The evolutionary sequence is tabulated below : As these changes in shell form were gradual and progressive, transitional forms occur between 1 and 2, and 2 and 3. However, M. ( M .) complanata and M. (M.) dehaarti have never been found in natural association. Cole’s evidence from Midway is not opposed to this hypothesis. His figured speci- mens show an overall increase in the length of the spire with age. 3. M. (M.) dehaarti 2. M. ( M .) bantamensis 1. M. (M.) complanata I Late Lower e Upper e Data from Cole (1969, pi. 1) Depth 595-600 ft 901-906 ft No. of chambers in spire ( — ) figs. 3 and 4 (7); figs. 1,11, and 12 (9) figs. 9 ( 1 0) ; fig. 20 (9 + ) ; fig. 8 is a micro- spheric form 926-927 ft figs. 5 and 6(10); figs. 13, 14, 16, and 17 (13); figs. 18 and 19(12 + ) ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 497 Hanzawa (1957) and others have attributed importance to the attitude of the em- bryonic chambers relative to the apex of the shell. However, this is determined by the length of the spire. Once median chambers have begun to form, the spire cannot be continued for more than half a turn (usually 7-9 chambers) without the typical fan shape being lost. Variation. Periembryonic spire of medium length and formed of 9 to 13 chambers. Internal diameter of the proloculus 01 33 to 01 50 mm (six measured specimens). Test up to 2-2 mm long and 0-76 mm thick; always longer than wide. Since this species evolved from M. ( M .) complanata and into M. ( M .) dehaarti, it follows that transitional forms occur and that an arbitrary specific diagnosis will not be satisfactory for the early and late representatives of the '’bantamensis' part of the lineage. Miogypsina ( Miogypsinoides ) dehaarti van der Vlerk Plate 73, figs. 12-14 1900 Heterostegina depressa d’Orbigny; Jones and Chapman, p. 257. 1924 Miogypsina dehaartii van der Vlerk, p. 429, text-figs. 1-3. 1965 Miogypsinoides dehaarti (van der Vlerk); Ludbrook, p. 293, pi. 21, figs. 9-11 and fig. 12 (part). Remarks. Little can be added to previous descriptions. In the present material the nepionic spire consists of from 7 to 10 chambers of which the last few are usually very small. The test is usually wider than long, and the lateral walls are well developed. The test ranges up to T2 mm in thickness; some specimens have a strongly pustulate appearance, others are smooth. This species forms a foraminiferal coquina in some samples (e.g. 1058, 1059, 1061, ‘G’ Traverse, and Andrews’s no. 131, Flying Fish Cove) ; it is common in the Ross Hill and Waterfall sequences. M. ( M .) dehaarti occurs in Assemblages 3 and 4. It overlaps and grades into M. ( M .) bantamensis in the lower part of Assemblage 3, and overlaps with M. ( Mio- gypsina) cf. neodispansa in Assemblage 4. Ludbrook (1965) figured an association of M. (M.) dehaarti and M. ( Miogypsina ) neodispansa. Miogypsina ( Miogypsina ) neodispansa (Jones and Chapman) Plate 71, figs. 16-18, text-fig. 11 1900 Orbitoides ( Lepidocyclina ) neodispansa Jones and Chapman, pp. 235, 240, pi. 20, figs. 3 and 4. 1926 Miogypsina neodispansa (Jones and Chapman); Nuttall, pp. 37, 38, pi. 5, fig. 4. 1965 Miogypsina neodispansa (Jones and Chapman); Ludbrook, p. 290, pi. 2, fig. 12 (part). Remarks. The original description of M. (M.) neodispansa is poor, and Nuttall’s emendation little better since he gave no information about the nepionic spire, the one part of the test considered to be of diagnostic importance by modern workers. Examination of numerous random sections has revealed the following variation in test morphology: Size. The maximum diameter of the test ranges from 2 to 4 mm, and averages about 2-5 mm. The maximum thickness (1-6 mm in the type slides) is dependent on the number (6-12) of lateral chamber layers developed. 498 PALAEONTOLOGY, VOLUME 17 Surface ornament. This varies from finely to coarsely pustulate. The maximum diameter of the pustules is about 250 pm, but many individuals are ornamented with pustules not exceeding 50 p m in diameter. Chamber shape and arrangement. The embryonic and median chambers do not always lie in the same plane, a feature which makes the preparation of oriented thin sections rather difficult. The median layer is composed of a few rhombic and many hexagonal chambers. This layer is not always flat, and wavy tests of the M. ( M .) bifida and M. (M.) poly- morpha types are quite common. Embryonic apparatus. The internal diameter of the protoconch in most specimens falls within the range 0T 2-0-20 mm. However, one individual in a sample from South Point has an initial chamber with a diameter of 0-25 mm. The deuteroconch is usually considerably larger than the protoconch. Two protoconchal spirals are clearly visible, and some individuals seem also to possess a third spire (text-fig. 11) but this is much less clear. The two primary auxiliary chambers are usually unequal in size; the larger chamber usually gives rise to a spire of three chambers while the smaller produces a shorter spire of two chambers. M. ( M .) neodispansa was one of the first miogypsinids to be described from the Indo- West Pacific region, and as such is an important species. Its association with M. ( Mio - gypsinoides) dehaarti and a fairly primitive form of Austrotrillina howchini fixes its position as Upper e, while the fact that it occurs fairly high in the succession (not very far below Flosculinella bontangensis ) probably means that it is a late Upper e form. It is still impossible to state clearly how M. ( M .) neodispansa differs from other described Indo-West Pacific species of the genus. Further data on the periembryonic and median chambers of this and other species are required, and these await the collection of better material. However, present evidence suggests that M. ( M .) neodispansa is more highly evolved than M. ( M .) thecideaeformis and M. ( M .) globulina (= M. ( M .) kotoi ) since both protoconchal spires are well developed and the median chambers appear to be markedly hexagonal. On the other hand, it is less advanced than M. ( M .) indonesiensis which has subequal primary auxiliary chambers and median chambers that are hexagonal throughout. On the basis of individual specimens it would be possible to recognize four or five ‘species’ in the Christmas Island samples. M. (M.) neodispansa occurs in great abundance in three samples from different parts of the island (Flying Fish Cove, nos. 220 and 924; South Point area, K. 129). Specimens probably referable to this species occur in three samples from the upper part of the Ross Hill Traverse. text-fig. 1 1 . Periembryonic spirals of three typical specimens of Miogypsina neodispansa. All from Andrews’s sample 220, south side of Flying Fish Cove; (u) P. 49093, (b) P. 49092, (c) P. 49090. ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 499 Family lepidocyclinidae Scheffen, 1932 Genus lepidocyclina Giimbel, 1870 Type species. Nummulites mantelli Morton 1833. Christmas Island is the type locality for six species of Lepidocyclina , five of which were erected by Jones and Chapman (1900) on the basis of a small number of random sections through fourteen samples of limestone. Nuttall ( 1 926), after having additional sections cut from the same material, recognized the following species : Lepidocyclina andrewsiana (Jones and Chapman), L. ephippioides (J. and C.), L. ( E .) Iformosa Schlumberger ( — L. murrayana J. and C), L. chapmani Nuttall, L. inaequalis J. and C., and L. insulaenatalis J. and C. Ludbrook (1965) reported, but did not figure or describe, all six species from two samples (P. 132 and P. 52). None of these authors had access to material collected in stratigraphical order, and all used test shape and size as the basis for distinguishing between species; internal characters, although sometimes mentioned, were not used in a systematic manner. The present material shows that Lepidocyclina occurs mainly in the north-west of the island. The megalospheric form of Eulepidina occurs most commonly in the lowermost 250 feet of ‘D’ Traverse (Assemblage 1). There is one doubtful occurrence (1095) in the Waterfall Traverse (Assemblage 2), and numerous inflated ‘A’ forms ( E . andrewsiana type) occur in sample 1122 (Assemblage 3, Sydney’s Dale Traverse). Individuals present in the higher part of ‘D’ Traverse (Assemblage 2) are mainly microspheric forms of the chapmani/ insulaenatalis type and show obvious signs (breakage and abrasion) of redeposition. Many are coated with calcareous algae. No recognizable megalospheric forms are seen above sample 1030, other than a single specimen in sample 1045 (from a rolled boulder). It is clear that Eulepidina is represented here by several types of test (flattened or disc-like; lenticular; saddle-shaped; and highly inflated with a flange), but so far as can be ascertained from random sections, gradation occurs between the different types. The fact that most of these can be found in two of the lowest samples (1328 and 1078) from lD’ Traverse, and that no type is restricted to a definite stratigraphical interval, probably mean that this diversity of form reflects variation within a single species. However, Andrews’s sample 827 from 3 km south of Flying Fish Cove and four of Belford’s samples from Smith Point contain a few inflated megalospheric forms which must be referred to L. ( E .) andrewsiana, pending further investigation. Unfortunately, the associated foraminifera in these five samples are undiagnostic and could represent either Assemblage 1 or Assemblage 2. The subgenus Eulepidina is in need of revision. The specific characters on which the numerous nominal species have been based need to be evaluated statistically on the basis of matrix-free material. Until this has been done it will not be possible to name specimens occurring in hard limestones satisfactorily. LEPIDOCYCLINA (EULEPIDINA) H. Oouville, 1911 Type species. Orbitoides dilatata Michelotti, 1861. Lepidocyclina ( Eulepidina ) andrewsiana (Jones and Chapman) Plate 74, figs. 7-8 1900 Orbitoides ( Lepidocyclina ) andrewsiana Jones and Chapman, p. 255, pi. 21, fig. 14. 500 PALAEONTOLOGY, VOLUME 17 Remarks. The specimens referred to this species occur in four samples from Smith Point (1257, 1258, 1262, 1263) and one (827, the type sample) from the base of a lime- stone cliff resting on basalt at 1 50 m, 3 km south of Flying Fish Cove. They appear to differ from E. ephippioides in being more inflated and in tending to develop very thick walls between the lateral chambers. It is, however, quite possible that transitional forms to E. ephippioides would be found if matrix-free material were available. Lepidocyclina ( Eulepidina ) ephippioides (Jones and Chapman) Plate 74, figs. 4-6, 9, 12, 14, text-fig. 12 1900 Orbitoides ( Lepidocyclina ) ephippioides Jones and Chapman, pp. 251-252, pi. 20, fig. 9. 1900 Orbitoides ( Lepidocyclina ) murrayana Jones and Chapman, pp. 252-253, pi. 21, fig. 10. 1902 Lepidocyclina ( Eulepidina ) formosa Schlumberger, p. 251, pi. 7, figs. 1-3. 1926 Lepidocyclina ephippioides Jones and Chapman; Nuttall, pp. 34-36, pi. 5, figs. 1, 2, 3, 8, and 10. 1926 Lepidocyclina ( Eulepidina ) Iformosa Schlumberger; Nuttall, pp. 22-30. 1965 Lepidocyclina (Eulepidina) ephippioides (Jones and Chapman); Ludbrook, pp. 290, 291. 1965 Lepidocyclina ( Eulepidina ) murrayana Jones and Chapman; Ludbrook, p. 290. Remarks. Nuttall (1926) distinguished L. ephippioides from L. andrewsiana on size (the former was, he thought, slightly larger) and on the appearance of the embryonic apparatus, which he considered to be truly eulepidine only in L. andrewsiana. How- ever, he was comparing a well-centred section of L. andrewsiana with off-centre sections of L. ephippioides , and as shown by text-fig. 12 the appearance of the em- bryonic apparatus in Eulepidina depends entirely on the plane of section. Nuttall (op. cit., p. 35) himself observed that L. ephippioides ‘except for the nucleoconch strongly resembles E. formosa', a taxon now generally regarded as synonymous with E. ephippioides. The lectotype designated by Nuttall is from Andrews’s sample 549 taken ‘at the EXPLANATION OF PLATE 74 Fig. 1. Tayamaia marianensis (Hanzawa), CPC. 13722, sample 1065, between G’ Traverse and Waterfall, slightly off-centre transverse section, x 20. Figs. 2, 10. Sorites cf. orbiculus (Forskal). 2, CPC. 13773, sample 1358, 'D' Traverse, oblique median section, x 40. 10, CPC. 13744, sample 1355, ‘G’ Traverse, highly oblique median section, x 30. Fig. 3. Flosculinella bontangensis (Rutten), CPC. 13775, sample 1237 (Ludbrook's P33 locality), near Ross Hill, off-centre axial section, x 30. Figs. 4-6, 9, 12, 14. Lepidocyclina (E.) ephippioides (Jones and Chapman), all from ‘D’ Traverse. 4-6, 9, transverse sections showing variation in shape, size, and number of lateral chambers. 4, CPC. 1 3776, sample 1045, x 20. 5, CPC. 13777, sample 1044, x 10. 6, CPC. 13778, sample 1331, x 10. 9, CPC. 13779, sample 1331, x 10. 12, 14, median sections through megalospheric forms, both x 20. 12, CPC. 13783, sample 1078. 14, CPC. 13784, sample 1328. Figs. 7-8. Lepidocyclina (E.) andrewsiana (Jones and Chapman), CPC. 1 3780 and 1 378 1 . Transverse sections showing inflated umbonal region, x 10. Both from sample 1261, Smith Point at 60 ft. Fig. 11. Marginopora vertebralis Blainville, CPC. 13782, sample 1238, off-centre transverse section, x40. Fig. 13. Lepidocyclina sp., P. 49091 . Tangential section through umbonal region of inflated form (probably L. (E.) andrewsiana) showing greatly thickened walls of lateral chambers, x 16. Thin sections cut along lines a-a or b-b would appear to show thickened umbonal pillars; Andrews’s sample 827, 2 miles south of Flying Fish Cove. PLATE 74 ADAMS and BELFORD, Oligocene Miocene foraminifera 502 PALAEONTOLOGY, VOLUME 17 text-fig. 12. Parallel median sections through the em- bryonic apparatus of a single specimen of Lepidocyclina (Eulepidina) ephippioides (sample NB 905 1 , Kinabatangan River, Sabah, Borneo). It is clear from these drawings that the degree to which the deuteroconch appears to encircle the protoconch depends entirely on the plane of section. Nuttall (1926) would have regarded a-d as typical of L. (E) ephippioides, and e-f as typical of L. (E) andrewsiana (see p. 500). base of an inland cliff at 500 feet, running south from Flying Fish Cove’. Whether from a fallen block or in situ rock was not stated. The occurrence of Eulepidina in Assemblage 1 makes it certain that L. ( E .) ephippioides came from low in the post- Eocene succession. It is usually associated with Spiroclypeus margaritatus, Miogypsina (Miogypsinoides) bantamensis , and Austrotrillina striata , and is therefore of Lower e age. It is worth noting that the type slides of L. murrayana contain a typical Assemblage 2 fauna. Lepidocyclina sp. Plate 74, fig. 13 Large inflated B’ forms are common throughout "D’ Traverse and are virtually the only specimens present in the upper part of the section, i.e. above sample 1045. They show considerable variation in shape and size. A particularly prominent feature is the tendency of inflated forms to develop thick walls between the lateral chambers in the umbonal region (PI. 74, fig. 13). Cuts through these walls can produce the appearance of coarse pustules or pillars in sagittal sections. ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 503 lepidocyclina (nephrolepidina) H. Douville, 1911 Type species. Nummulites marginal a Michelotti, 1841 . L. ( Nephrolepidina ) spp. This subgenus is surprisingly rare in the samples so far obtained from Christmas Island. It occurs sparsely in ‘D’ Traverse, Sydney’s Dale Traverse, and in a few other localities (e.g. Flying Fish Cove, Andrews’s sample 646). In the absence of oriented thin sections it is impossible to assign specific names to these individuals, some of which have hexagonal equatorial chambers. They occur in Assemblages 1-3, and it is certain that more than one species is represented. Family homotrematidae Cushman, 1927 Genus carpenteria Gray, 1858 Type species. Carpenteria balaniformis Gray, 1858. Carpenteria spp. Numerous fragments and some larger specimens of this genus occur throughout Assemblages 2-5. No attempt has been made to distinguish between different species. Family acervulinidae Schultze, 1854 Genus gypsina Carter, 1877 Type species. Polytrema planum Carter, 1876. Gypsina globula ( Reuss) 1848 Ceriopora globulus Reuss, p. 33. 1900 Gypsina globulus (Reuss); Jones and Chapman, p. 229 et seq. Remarks. G. globula occurs throughout most of the succession and it is particularly common in Assemblages 2 and 3. There is nothing to add to previous descriptions. Family planorbulinidae Schwager, 1877 Genus tayamaia Hanzawa, 1967 Type species. Gypsina marianensis Hanzawa. Tayamaia marianensis (Hanzawa) Plate 71, fig. 8; Plate 74, fig. 1 1957 Gypsina marianensis Hanzawa, p. 66, pi. 21, fig. 9; pi. 27, figs. 1-8. 1965 Gypsina marianensis Hanzawa; Ludbrook, p. 292, pi. 22, fig. 2. 1967 Tayamaia marianensis (Hanzawa); Hanzawa, p. 22, fig. 3. Remarks. A long-rangingspecies that occurs particularly frequently in Assemblages2-4. OTHER ATTACHED AND ENCRUSTING GENERA Numerous attached and encrusting forms referable to Acervulina, Borodinia , Kana- kaia , and Sporadotrema are present throughout the succession, being particularly common in Assemblages 2-5. The comparative rarity of these forms in Assemblage 1 is consistent with deposition in a fore-reef environment. 504 PALAEONTOLOGY, VOLUME 17 SUMMARY The larger Tertiary foraminifera occurring in the post-Eocene limestones of Christmas Island appear to be characteristic of the Tertiary Lower e, Upper e, and Lower / Letter Stages. Live locally significant faunal assemblages can be recognized; two of these are probably in part laterally equivalent, and brought into their present topographical positions by fault movements. The close stratigraphical juxtaposition of Miogypsina ( Miogypsinoides ) eomplanata, M. ( M .) bantamensis, and M. ( M .) dehaarti seems to indicate that the evolution of this lineage proceeded very rapidly during the late Oligocene. The early members of the M. ( Miogypsina ) kotoi-M. ( M .) indonesiensis lineage have not been observed on Christmas Island. M. ( M .) neodispansa is a fairly advanced form which may well prove to be of regional value in the recognition of late Upper e sediments. Flosculinella is poorly represented on the island. Apart from a single, specifically indeterminable specimen in Assemblage 2 (sample 69L, Batu Merah, Llying Lish Cove), the genus is not seen until well-developed specimens of F. bontangensis appear high in the Ross Hill Traverse. However, the occurrence of what appears to be a primitive form with a good Assemblage 2 fauna almost certainly means that this genus can no longer be relied on to define unequivocally the base of Upper e. Although it has not been possible to determine how many species of Lepidocyclina are represented in the Christmas Island succession, it can be stated that all those described by Jones and Chapman were from limestones of late Lower e age. Acknowledgements. We wish to thank the British Phosphate Commissioners for permitting one of us (D. J. B.) to visit Christmas Island, for providing all necessary facilities and for permitting us to quote results from the ST. 1 stratigraphic bore; Mr. K. Lourey, then Island Manager; Messrs. E. Brennan and P. J. Barrett, B.P.C. geologists at that time, for assistance with the geological work; Mr. D. A. Powell, whose knowledge of the island and of access to the sections sampled was invaluable; and Messrs. Morgan, Ingram, and Johnson of the B.P.C. Fremantle Office, who helped in many ways with travel arrangements. Our thanks are also due to Dr. N. H. Ludbrook for allowing one of us (C. G. A.) to examine her material in Adelaide, and for providing a report on material from the ST. 1 stratigraphic bore. We thank Messrs. Barrett and Powell for their assistance in determining the type locality of Miogypsina neodispansa , and for collecting in the Flying Fish Cove area. Permission to publish has been received by one of us (D. J. B.) from the Director, Bureau of Mineral Resources, Geology, and Geophysics, Canberra. REFERENCES adams, c. G. 1968. A revision of the foraminiferal genus Austrotrillina Parr. Bull. Br. Mus. nat. Hist. ( geol .) 16, 73-97, 6 pis. 1970. A reconsideration of the East Indian Letter Classification of the Tertiary. Ibid. 19, 87-137. Andrews, c. w. 1900. A Monograph of Christmas Island (Indian Ocean). London. Br. Mus. Nat. Hist. xiii + 337 pp., 21 pis. barrie, j. 1967 (unpublished). The Geology of Christmas Island. Bur. Miner. Resour. Aust. Rec. 1967/37. blainville, h. m. d. de. 1830. Mollusques, vers et zoophytes. In Dictionnaire des Sciences Naturelles, Tome 60. Paris, F. G. Levrault. carpenter, w. b. 1856. Researches on the Foraminifera Pt. 2 on the genera Orbiculina, Alveolina, Cyclo- clypeus and Heterostegina. Phil. Trans. R. Soc. 146, 547-569, pis. 30-31. carter, h. j. 1861. Further Observations on the Structure of Foraminifera and on the Larger Fossilised Forms of Sind, &c., including a new Genus and Species. J. Bombay Brch R. Asiat. Soc. 6, 31-96. ADAMS AND BELFORD: CHRISTMAS ISLAND FORAMINIFERA 505 carter, h. j. 1876. On the Polytremata (Foraminifera), especially with reference to their Mythical Hybrid Nature. Ann. Mag. not. Hist. Ser. 4 (17), 185-214, pi. 13. 1877. On a Melobesian Form of Foraminifera ( Gypsina melobesoides, mihi) ; and further Observations on Carpenteria monticularis. Ibid. 20 (117), 172-176. cole, w. s. 1965. Structure and classification of some Recent and fossil peneroplids. Bull. Amer. Paleont. 49(219), 5-37, pis. 1-10. 1969. Larger Foraminifera from deep drill holes on Midway Atoll. Prof. Pap. U.S. geol. Surv. 680-C, 1-15, 4 pis. crespin, i. 1955. The Cape Range Structure Western Australia, Part II Micropalaeontology. Bull. Bur. Miner. Resour. Geol. Geophys. Aust. 21, 49-82, pis. 7-10. CUSHMAN, J. A. 1927. An outline of a reclassification of the foraminifera. Contr. Cushman Lab. foramin. Res. 3, 1-105, pis. 1-21. douville, h. 1905. Les foraminiferes dans le Tertiaire de Borneo. Bull. Soc. geol. Fr. Ser. 4, 5, 435-464, pi. 14. 1911. Les foraminiferes dans le Tertiaire des Philippines. Philipp. J. Sci. Ser. D. 6, 53-80, pis. A-D. drooger, c. w. 1953. Some Indonesian Miogypsinae. I. Introduction and descriptions. Proc. Kon. Ned. Akad. Wet. Ser. B, 56, 104-117, pi. 1. ehrenberg, c. G. 1839-1840 (for 1838). Uber die Bildung der Kreidefelsen und des Kreidemergels durch unsichtbare Organismen. Abh. preuss. Akad , Wiss. 59-147, pis. 1-4. fichtel, L. and MOLL, J. P. c. 1798. Testacea microscopica aliaque minuta ex generibus Argonauta et Nautilus. Wien, Anton Pichler, xii + 123 pp., pis. 1-24. forskal, p. 1775. Descriptiones Animalium Avum, Amphibiorum , Piscium , Insectorum, Vermium ; quae in itinere orient ali. Hauniae. 19 + xxxiv+ 164 pp. gray, j. e. 1858. On Carpenteria and Dujardinia, two genera of a new form of Protozoa with attached multilocular shells filled with Sponge, apparently intermediate between Rhizopoda and Porifera. Proc. zool. Soc. Lond. 26, 266-271, figs. 1-4. gumbel, c. w. 1870. Beitrage zur Foraminiferenfauna der nordalpinen Eocangebilde. Abh. bayer. Akad. Wiss. 10, 581-730, pis. 1-4. hanzawa, s. 1930. Note on Foraminifera found in the Lepidocyclina- limestone from Pabeasan, Java. Sci. Rep. Tohoku Unix. Ser. 2 (Geol.), 14, 85-96, pis. 26-28. 1940. Micropalaeontological studies of drill cores from a deep well in the Kita-Daito-Zima (North Borodino Island). In Jubilee publications in commemoration of Professor H. Yabe's sixtieth birthday, pp. 755-802, pis. 39-42. 1957. Cenozoic Foraminifera of Micronesia. Mem. geol. Soc. Am. 66, 63 pp., 38 pis. 1967. Three new Tertiary foraminiferal genera from Florida, Saipan and Guam. Trans. Proc. palaeont. Soc. Japan, N.s. 65, 19-26, pis. 3-4. jones, t. r. and chapman, f. 1900. On the foraminifera of the Orbitoidal limestones and reef rocks of Christmas Island, pp. 226-264, pis. 20-21. In Andrews, c. w. A Monograph of Christmas Island ( Indian Ocean). kronen, w. F. 1931. Het Genus Spiroclypeus in het Indo-Pacifische gebied. Verb. geol. -mijnb. Genoot. Ned. 9, 77-112, pis. 1-3. ludbrook, n. h. 1965. Tertiary Fossils from Christmas Island (Indian Ocean). J. geol. Soc. Aust. 12, 285- 294, pis. 21-22. michelotti, G. 1841. Saggio storico dei Rizopodi caratteristici dei terreni sopracretacei. Memorie Soc. ital. Sci. XL 22, 253-302, pis. 1-3. 1861. Etudes Sur le Miocene inferieur de lTtalie septentrionale. Maatsch. Wetensch. Haarlem, Natuurk. Verh. Haarlem, 183 pp., 16 pis. montfort, D. de. 1808. Conchyliologie systematique et classification methodique des coquilles. Paris, F. Schoell. 1, lxxxvii + 409 pp. morton, s. G. 1833. Supplement to the ‘Synopsis of the Organic Remains of the Ferruginous Sand Forma- tion of the United States’, contained in Vols. XVII and XVIII of this Journal. Am. J. Sci. 23, 288-294, pis. 5 and 8. nuttall, w. l. f. 1926. A revision of the Orbitoides of Christmas Island. Q. J! geol. Soc. Lond. 82, 22-42, pis. 4-5. E 506 PALAEONTOLOGY, VOLUME 17 orbigny, a. d’. 1826. Tableau methodique de la classe des Cephalopodes. Ann. Sci. nat. Ser. 1, 7, 96-314, pis. 10-17. parr, w. j. 1942. New Genera of Foraminifera from the Tertiary of Victoria. Min. geol. J. 2, 361-363, figs. 1-5. reiss, z. and gvirtzman, g. 1966. Borelis from Israel. Eclog. geol. Helv. 59, 437-447, pis. 1-2. reuss, a. e. 1848. Diefossilen Polyparien des Wiener Tertiarbeckens. Naturw. Abh. Wien, 2, 1-109, pis. 1-11. rutten, l. 1913. Studien fiber Foraminiferen aus Ost-Asien. Samml. geol. Reichmus. Leiden, Ser. 1, 9, 219-224, pi. 14. sacco, f. 1893. Sur quelques Tinoporinae au Miocene de Turin. Bull. Soc. beige Geol. Paleont. Hydro!. 7, 204-207. scheffen, w. 1932. Ostindische Lepidocyclinen. Dienst. Mijnb. Wetensch Meded. Batavia, 21, 5-76, pis. 1-14. schlumberger, c. 1893. Note sur le genres Trillina et Linderina. Bull. Soc. geol. Fr. Ser. 3, 21, 1 18-123, pi. 3. 1900. Note sur le genre Miogypsina. Ibid. 28, 327-333, pis. 2-3. — 1902. Note sur un Lepidocyclina nouveau de Borneo. Samml. geol. Reichsmus. Leiden, Ser. 1, 6, 250- 253, pi. 7. Schubert, r. j. 1910. In richarz, p. s. Der geologische Bau von Kaiser Wilhelms-land nach dem heutigen Stand unseres Wissens. In boehm, g. Geologische Mitteilungen aus dem Indo-Australischen Archipel. Neues Jb. Miner. Geol. Palaont. Biel. 29, 406-536, figs. 1-10, pis. 13-14. schultze, M. s. 1854. Liber den Organismus der Polythalamien (Foraminiferen) nebst Bemerkungen iiber die Rhizopoden im allgemeinen. Leipzig, Engelmann, 1-68, pis. 1-7. tan sin hok. 1932. On the genus Cycloclypeus Carpenter. Pt. 1 and an appendix on the Heterostegines of Tjimanggoe, S. Bantam, Java. Wet. Meded. Dienst. Mijnb. Ned.-Oost Indie, 19, 3-194, pis. 1-24. 1936. Zur Kenntnis der Miogypsiniden. Ing. Ned.-Indie , 4, 45-61, pis. 1-2. todd, R. and post, R. 1954. Smaller Foraminifera from Bikini Drill Holes. Prof. Pap. U.S. geol. Surv. 260-N, 547-568, pis. 198-203. trueman, N. a. 1965. The phosphate, volcanic and carbonate rocks of Christmas Island (Indian Ocean). J. geol. Soc. Aust. 12, 261-283, pis. 18-20. vlerk, i. m. van der. 1924. Miogypsina Dehaartii nov. spec, de Larat (Moluques). Eclog. geol. Helv. 18, 429-432, figs. 1-3. yabe, h. and hanzawa, s. 1928. Tertiary Foraminiferous Rocks of Taiwan (Formosa). Proc. imp. Acad. Japan, 4, 533-536, figs. 1-3. c. g. ADAMS Department of Palaeontology British Museum (Natural History) Cromwell Road London, SW7 5DB D. J. BELFORD Bureau of Mineral Resources, Geology and Geophysics P.O. Box 378 Canberra City, A.C.T. 2601 Typescript received 14 June 1973 Australia THE ECOLOGY OF A MIDDLE JURASSIC HARDGROUND AND CREVICE FAUNA by T. J. PALMER and F. T. FURSICH Abstract. The base of the Bradford Clay (Bathonian) at Bradford-on-Avon, Wiltshire, was a hardground. Soft lime-sand below the hardground was washed out to form crevices, the walls and the floor of which also became lithified. The hardground and the crevices were colonized by encrusting and boring organisms, which polarized into : (i) an upper-surface community dominated by oysters, Apiocrinus and Nubeculinella and (ii) a crevice community, dominated by Serpula ( Cycloserpula ), encrusting ectoprocts, and Moorellina. Similar ecological distributions are known from other formations, and are attributed to differences in light and turbulence, which in turn influenced competition for space and food. Around Bradford-on-Avon, Wiltshire, the Forest Marble Formation (Upper Bathonian) is about 24 m thick (Green and Donovan 1969); the basal 3 m or so consists of well-sorted, cross-bedded oobiosparites. Above this, about 3-5 m of clay alternating with thin limestone partings, first described by William Smith (1816) as "Clay above the Upper Oolite’, soon became known as the Bradford Clay. It achieved fame on account of the rich fossil bed frequently found at its base. Ammonites of the genus Clydoniceras have been found in the Bradford Clay which allow it to be placed in the hollandi Subzone of the discus Zone (Stinton and Torrens 1968). Smith (1816) realized that the fossils in the basal shell bed of the Bradford Clay had in life been associated with the top surface of the underlying limestone, and that the clay had buried this fauna. It is the nature of this surface and its fauna which forms the subject of this paper. Several previous accounts have been published (e.g. Smith 1817; Cunnington 1859; Woodward 1894), and also Periam (Periam, C. E. 1956, The Jurassic rocks of mid-Wiltshire. Unpublished Ph.D. thesis, University of Reading), but these have tended to stress the species which became separated from their substratum after death, and are now found loose in the base of the clay. Equally important members of the community were the forms which lived firmly cemented to the sea bottom or which bored into it. Except for the conspicuous Apiocrinus, this element has been largely ignored. Good exposures in the lower part of the Bradford Clay have been very rare since the old brick pits in which it was quarried passed out of work. However, the basal shell-bed and the underlying limestone have recently been re-exposed at two localities : (i) The old Canal Quarry (ST 826600) (text-fig. 1), where about 5 m2 of the top of the limestone are exposed at the back of the quarry. The overlying clay may also be seen here. (ii) Springfield, Bradford (ST 831609), 1100 m to the NNE., where a further 5 m2 of the limestone top, as well as 2-5 m of the overlying clay, were temporarily visible in the summer of 1972. In addition, the contact between the limestone and the overlying Bradford Clay could be followed in vertical section for about 100 m towards the north. [Palaeontology, Vol. 17, Part 3, 1974, pp. 507-524, pis. 75-77.] 508 PALAEONTOLOGY, VOLUME 17 1 m text-fig. 1 . Localities and sections in the Bradford Clay at Bradford-on-Avon. The top of the limestone at both localities is an intra-formational hardground with an associated boring and encrusting fauna. The northern exposure at Spring- field showed the hardground fauna and shell-bed derived from it, dying out north- ward over 10 m. There is no sign of the hardground having been removed by erosion before the deposition of the clay, and it seems likely that it was developed only patchily over the area where it is now seen. The hardground is also undermined by crevices, which cut back beneath it for up to 25 cm. These crevices measure up to 5 cm from roof to floor, and are filled with clay and shell debris. Registration of material. Figured specimens are housed in the Oxford University Museum. THE DEVELOPMENT OF THE HABITAT The limestone beneath the clay is a light-brown, cross-bedded oobiosparite locally passing into lenses composed largely of disarticulated and broken shell material. Brachiopods, echinoids, and ectoprocts are common, and bivalves are well repre- sented by pectinids and oysters. The bed has every appearance of having been deposited in current-swept, fully marine conditions. The top of this oobiosparite is a hardground : synsedimentary calcium carbonate cementation of the lime-sand produced a rocky bottom, rather than a bottom that was merely compacted and firm. This assumption is supported by the following observations: (i) Well-developed overhangs, rigid enough not to collapse under their own weight (text-fig. 2). (ii) Encrustation on upper and lower surfaces, especially by oysters and serpulids (PI. 76, fig. 1 ; PI. 77, fig. 9). (iii) Borings cutting through sediment particles (PI. 75, fig. 2; text-fig. 2). (iv) Occasional bored and encrusted pebbles, of the same lithology as the top of the limestone, found at the base of the overlying clay. PALMER AND FURSICH: JURASSIC HARDGROUND 509 Recent hardground formation The occurrence of early diagenetic lithification of carbonate sediments in Recent environments has been reviewed by Bathurst (1971): cementation in a shallow, but permanently submerged, environment is found in parts of the Bahamas (Taft et ah 1968), and over wide areas of the Persian Gulf (Shinn 1969; Taylor and llling 1969). A similar depositional environment has already been proposed for the Upper Bathonian of the Bradford region by Green and Donovan (1969). The similarity between the Bradford hardground and those in the Persian Gulf is striking. Description of the Bradford hardground and its similarity to Persian Gulf examples (i) The upper surface of the hardground at Bradford is irregular, with a succession of flat hummocks separated by gullies up to 1 m or so across. The floors of the gullies are lithified, and their sides often recessed back under the adjoining high areas, so that crevices are formed (text-fig. 2). The roofs of these crevices are formed by thin layers of sediment 5 to 15 m thick, both top and bottom of which are encrusted and bored. It seems as if these layers are the result of cementation having occurred in thin zones, as in the Persian Gulf (Shinn 1969; Taylor and llling 1969). The soft sediment beneath the lithified layers was removed by currents or animals to form the cavities (cf. Shinn 1969, p. 124). (ii) Both top and bottom surfaces of the hard layers are pitted with irregular holes up to 3 cm across, whose sides are also encrusted. These are almost certainly the burrows of crustaceans which excavated open dwelling systems in the soft lime-sand, before lithification became too far advanced. The undersurfaces of Persian Gulf hardgrounds are also irregular, due to differential cementation around burrows. (iii) Small cracks, up to 1 mm wide and now filled with sparry calcite, cut across some of these thin cemented layers. These probably represent fractures, formed as a result of intergranular growth of cement crystals (cf. Shinn 1969, p. 128 and fig. 1 8). (iv) The frequency with which the crevices occur suggests that it was easy for either currents or vagile animals to remove soft sediment from below the lithified layers. It may be that these layers were developed only locally in the sediment (below). Alternatively, major breaks in an extensive crust may have occurred, perhaps as a result of the same intergranular crystalline growth postulated in (iii) above. In either case, removal of the underlying soft sediment would have been facilitated. (v) A patchy development of the lithified layers is supported by the observation that the hardground is locally discontinuous over the whole area studied. Alterna- tively, the lithified layer may have formed here, but become covered by a layer of soft sediment which protected it from colonization (cf. Shinn 1969, p. 115). (vi) If the gullies and crevices were formed by removal of soft sediment from between and under pieces of lithified layer, then lithification must have occurred in more than one episode, and perhaps continuously. This is so because the floors and sides of the cavities are themselves hardened and encrusted. Furthermore, the cemented sediment which forms these cavity floors contains fossils, such as Apio- crinus , which were associated with the overlying hardground. After death, these crinoids and other hardground fauna tended to accumulate in the cavities, together with sediment trapped in this far less turbulent microenvironment. Subsequent cementation of the walls of these cavities, as well as the floor deposits, provided a new 510 PALAEONTOLOGY, VOLUME 17 settlement area for the hardground fauna and, consequently, led to further intensive encrustation and boring. The history of formation of the Bradford hardground is summarized in text-fig. 2 (cf. Fursich 1971, fig. 3). THE FAUNA AND ITS ECOLOGY Faunal types and their distribution The collections made at the two exposures contain nearly all the species considered typical of the Bradford Clay by earlier workers (e.g. Woodward 1894; Periam 1956). The total fauna may be divided into three types on ecological and preservational criteria : (i) Vagile benthos including several species of micromorphic gastropods, asteroids, and the regular echinoid ‘ Cidaris" bradfordensis Wright. (ii) Bysally and pedically attached fixosessile benthos, usually attached in life to some hard substrate such as a piece of shell or the hardground. This group includes the bivalves Lima ( Plagiostoma ) sp., Oxytoma costatum (Townsend), and Radulo- pecten vagans (J. de C. Sowerby), as well as the brachiopods usually considered diagnostic of the Bradford Clay and its supposed stratigraphic equivalents (see Elliott 1973). These include Dictyothyris coaretata (Parkinson), Avonothyris bradfordensis (Davidson), Eudesia cardium (Lamark), Digonella digona (J. Sowerby), Crypto- rhynchia bradfordensis (S. Buckman), and other small rhynchonellids. Periam (1956) gives a more detailed account of the brachiopods from the Canal Quarry. Neither he nor the present authors have found E. cardium, but it has been collected extensively in the past. Also tentatively included in this group are small corals of the Anabacia type which were probably free-living in the adult stage. The fauna of these two groups may now be found loose in the cavities, or in the shell-bed above the hardground. (iii) Cemented and boring fixosessile benthos which required a hard, cemented surface. This surface was provided either by pieces of shell material, or by the exten- sive hardground whose physical and chemical properties were equivalent to those of shell material. This is the fauna with which we are most concerned here; because text-fig. 2. Diagrammatic history of cementation and colonization of the Bradford hardground: (a) Deposition of lime-sand. ( b ) Crustaceans excavated burrow systems ; discontinuous lithification of a thin layer of sediment began at or near the sediment/water interface. (c) Lithification continued ; the bottom surfaces of these lithified layers were exposed by removal of the underlying uncemented sediment. (i d ) The exposed hard surfaces were colonized by boring and encrusting animals. Periods of shell accumulation on the hardground alternated with periods of bioerosion, during which boring activity removed some encrusting shell material (PI. 75, fig. 4). The floors of the crevices started to lithify. (e) Shell material derived from the hardground accumulated within crevices. Eventually, clay deposition buried the hardground and its associated fauna. The rectangular insets on the left of the figure show the growth of cement around the grains; sediment passes from loose — > lightly cemented — > well cemented. PALMER AND FURSICH: JURASSIC HARDGROUND 51 1 512 PALAEONTOLOGY, VOLUME 17 of its attached nature, it is preserved in life position, and is therefore more suitable for palaeoecological study than faunas which became detached after death. Species distribution of the boring and encrusting fauna The encrusting and boring fauna can be further subdivided into two groups, the first of which is composed of species which preferred to live on the upper surfaces of the hardground, and the second of those forms which preferred the under surface of the lithified layers, provided by the crevice roofs. It has not been possible to collect for close examination the faunas of the crevice floors. As far as can be determined from cursory examination in the field, these resemble the crevice roof faunas, but table 1. Table showing distribution of encrusting and boring species on the Bradford hardground. Note preference of arborescent forms and bivalves for upper surface, and preference of serpulids, encrusting ectoprocts, and thecidaceans for lower surface. r = rare; o = occasional; c = common; a = abundant. GROUP SPECIES UPPER SURFACES r o c a LOWER SURFACES r o c a ENCRUSTING Foraminifera Nubeculinella sp. Porifera Limn or ia sp. Serpul idae Serpula (Cycloserpula ) sp. 1 Serpula (Cycloserpula) sp. 2 Serpula (Tetraserpula ) sp. Serpula (Dorsoserpula ) sp. 1 Serpula (Dorsoserpula) sp. 2 Bivalvia Liostrea wiltonensis (Lycett) Liostrea hebridica (Forbes) Exogyra crassa (Smith) ? Nanogyra nana (J.Sowerby) Lopha gregaria (J.Sowerby) Plicatula fistulosa (Morris & Lycett) Plicatula sp. 2 Ectoprocta Plagioecia sp. Mesenteripora sp. Stomatopora dichotoma (Lamouroux) Collapora sp. Terebellaria ramosissima (Lamouroux) Brachiopoda Moorellina sp. Cr inoidea Apiocrinus parkinsoni (Schlotheim) 1 BORING Porifera borings of Cliona sp. "Vermes" Trypanites sp. Bivalvia Lithophaga SP* Cirripedia Acrothoracican borings only in shell de sri s Phoronida Talpina ramosa (Hag. ) 1 1 1 1 1 1 r o c a r o c a PALMER AND FURSICH: JURASSIC HARDGROUND 513 pass into typical upper surface faunas as the crevice floor passes laterally into the exposed floor of the gullies (text-fig. 2). These preferences are outlined below, and summarized in text-fig. 4. Protozoa; Foraminifera; Nubeculinella sp. (PI. 76, fig. 3). This adherent calcareous foraminifera is extremely abundant on the hardground, always on the upper surface. It thrived both on the cemented sediment itself and on the shells attached to it. Porifera; Linmoria sp. (PI. 77, fig. 4). This calcisponge occurs on lower surfaces only. Specimens are usually small and poorly preserved, and seldom obvious to the naked eye. ‘ Cliona ’ sp. Irregular borings up to 2 mm diameter of the ichnogenus Entobia are abundant in the shells encrusting the upper surface of the hardground. The raised margins of Exogyra valves seem to have been particularly susceptible to attack. Similar borings also occur commonly in the loose shell debris overlying the hardground. ‘Vermes’ ; Trypanites sp. (PI. 75, fig. 1). Borings in the form of simple tubes between 1 and 3 mm diameter, inclined at various angles to the perpendicular, are assigned to this ichnogenus. They appear to be equally common on both lower and upper surfaces. Annelida ; Serpula ( Cycloserpula ) spp. Two species of this subgenus are found on lower surfaces. The first is relatively uncommon and is characterized by having started its growth in a flat spiral for 4 or 5 whorls (PI. 77, fig. 8), before taking up an irregular path. The second form is abundant and grew in an apparently random path from its earliest stages (PI. 77, fig. 9). Occasionally, a large form (2-3 mm diameter) is found growing in approximately straight lines on upper surfaces. It invariably occurs near the edge of an over- hang, and probably represents a different growth form of the species so abundant on the lower surface. Serpula ( Dorsoserpula ) spp. Two species occur; they are mutually exclusive on lower and upper sur- faces respectively. The former is common, growing up to 3 mm diameter. It is nearly circular in cross-section and the median dorsal keel is prominent (PI. 77, fig. 3). The other species is also common, and is triangular in cross-section. It usually grows to no more than 1-5 mm high and 2 cm long. It occurs both on bare hardground and on associated shell material. Serpula ( Tetraserpula ) sp. (PI. 77, fig. 2). One species of Tetraserpula is common on under surfaces, but is not found on upper surfaces. It is easily recognized by its circular cross-section (up to 2 mm across), and the three prominent longitudinal keels. On the older part of the test, these keels subtend a series of sharp projections about 1 mm long. Taken together, serpulids account for the largest proportion of the crevice roof fauna. On some roofs they appear to account for over 90% of the covered surface. In contrast, the upper surface forms are not only less common, but also smaller. Arthropods; Cirripedia. Small acrothoracian borings occur occasionally in loose Apiocrinus ossicles, but have not been recognized in the hardground itself, or in the shells attached to it. No conclusions can be drawn about their life preferences. Bivalvia; Liostrea. Two species of this genus are found attached to the hardground: L. wiltonensis, a large solitary form, is found rarely on the upper surfaces. The other species, which resembles Liostrea hebridica in overall shape, occurs occasionally on both surfaces. ‘ Exogyra ’ (PI. 76, fig. 4). The large E. crassa (Smith), which was first recorded from this horizon by Smith (1816), is everywhere abundant over the upper surface of the hardground. There are also many small forms, some of which may be referable to Nanogyra nana which Periam (1956) records from the Canal Quarry. Alternatively, they may be no more than the young stages of E. crassa. Exogyra is occasionally found on the under surfaces of overhangs, but always near the outside edges. Lopha gregaria. This species occurs commonly on the top surface, where it prefers small depressions, such as those provided by the remains of crustacean burrows. It is occasionally found on the outside edges of the under surfaces of overhangs. Plicatula. P.fistulosa occurs commonly on the upper surfaces (PI. 76, fig. 4), where it is usually found in small clusters. It is also found occasionally growing amongst the branches of arborescent bryozoans. A second species of Plicatula is found exclusively on crevice roofs (PI. 77, fig. 7). It is larger than P.fistulosa , more adpressed to the substratum, and less spinose in the attached valve. It appears to have preferred to colonize the bare hardground, rather than a surface which was already covered by a biogenic layer of calcium carbonate. 514 PALAEONTOLOGY, VOLUME 17 Lithophaga sp. (PI. 75, figs. 2, 3). This boring form is abundant on both surfaces, with a slight preference for the upper. It is bivalves, particularly ‘ Exogyra ’, which dominate the hardground’s upper surface. Bryozoa ; Ectoprocta. Three crustose ectoprocts occur on the hardground. The first two, Mesenteripora sp. (PI. 77, fig. 5) and Stomatopora dichotoma (PI. 77, fig. 1), are confined to the crevice roofs where they usually occur in close association with the serpulids and other shell material. They colonized the bare hardground less frequently. The third form, Plagioecia sp. (PI. 77, fig. 6) occurs similarly, and is also found occasionally on upper surfaces, where it displays a marked substrate preference for Apiocrinus holdfasts. Fragments of arborescent ectoprocts are common in the shell debris above the hardground. The roots of one of these ( Collapora sp.) are occasionally seen on upper surfaces (PI. 76, fig. 2). The other, Terebellaria ramosissima , is not found in situ. It seems likely, however, that it was also attached to the upper surface in a similar way. Together with the serpulids, ectoprocts are a major element of the crevice roof fauna. Phoronida. Ramose borings ( Talpina ramosa Hag.), attributed to phoronids (Voigt 1972), are very abundant in bivalve shells on both surfaces, and also in loose shell debris (PI. 75, fig. 5). Brachiopoda; Thecideacea; Moorellina sp. (PI. 77, fig. 4). Cemented brachiopods assigned to the genus Moorellina abound on the crevice roofs only. They appear to thrive equally on the bare hardground and on other shell material, but prefer the outer regions of the crevices to the inner recesses. Echinodermata ; Crinoidea; Apiocrinus parkinsoni (Schlotheim) (PI. 76, fig. 5). Black holdfasts of Apio- crinus abound on the upper surface of the hardground. Occasionally, complete specimens are found, lying flat and fully articulated along the hardground. Holdfasts are never found on undersurfaces. Community ecology The top and bottom surfaces (= crevice roofs) of the hardground each have a distinct association of species wherever they have been observed. The associations may be regarded as two distinct communities : on the upper surface of the hardground is an oyster/ Apiocrinus community, and on the roofs and, probably, floors of the crevices is an serpulid/ectoproct community. The former is characterized by the presence of Apiocrinus, Lopha, Plicatula fistulosa, Liostrea wiltonensis, Serpula ( Dorsoserpula ) sp. 1, Nubeculinella, Cliona , and the arborescent ectoprocts Collapora and Terebellaria ramosissima. The crevice community is characterized by two species of Serpula (Cyclo serpula), the crustose ectoprocts Mesenteripora, Stomatopora, Plagioecia (which also occurs occasionally on upper surfaces), the calcisponge Limnoria, a second species of Plicatula, and abundant Moorellina. Other species are represented in both communities. The two communities not only contain different members, there is also a difference EXPLANATION OF PLATE 75 Boring fauna in the Bradford hardground and associated shell material. Fig. 1. Polished section through lower surface biogenic layer, with abundant Lithophaga borings and occasional Trypanites sp. (arrowed), (J 40033), x 2-4. Fig. 2. Section through Lithophaga sp. in crypt on upper surface; truncation of grains (arrowed) shows bed was cemented before boring occurred, (J 40034), x 6-3. Fig. 3. Enlargement of section in fig. 1 through lower surface endolithic layer, showing Lithophaga borings cutting through older borings (arrowed) which indicates more than one generation of boring activity, (J 40033), x 5. Fig. 4. Polished section through upper surface biogenic layer showing Lithophaga crypt truncated and overgrown by oyster (arrowed), (J 40035), x 6-5. Fig. 5. Phoronid borings in Oxytoma costatum (Townsend), (J 40036), x 2-5. PLATE 75 PALMER and FURSICH, Middle Jurassic hardground 516 PALAEONTOLOGY, VOLUME 17 in growth forms. The cavity community is dominated by members which are closely adpressed to the substrate, as might be expected in a small space of restricted height. In contrast, the oyster/ Apiocrinus community contains forms, such as Apiocrinus itself and the arborescent ectoproct Collapora , which extended upwards to exploit food at different heights above the hardground. Altogether, five different levels at which exploitation occurred can be distinguished (text-fig. 3). This vertical stratifica- tion is analogous to that noted by Elton (1966) for terrestrial woodland. In the substrate itself is an endolithic layer, corresponding to Elton’s soil layer; on top of the substrate, in order of increasing height, come the two encrusting layers, analogous to ground and field layers, and the two silvide layers, analogous to shrub and canopy layers. p Moorellina i Exogyra & Nanogyra \tiuuK crustose ectoprocts — u“ Lithophaga \ Apiocrinus f silvide ectoprocts AT Tr^anites 1 *£20* crustose sponges text-fig. 3. Ecological stratification on hardground surfaces. Endolithic and crustose layers occur on the upper surface of the Bradford hardground. as well as in the cavities beneath. The silvide layers are confined to the upper surface. The cemented and boring forms were preserved in life position after death and their life preferences are clear. It seems probable that the other elements of the fauna which did not remain in life position after death also had similar life preferences. Many Recent asterozoa, for example, favour the protection offered by cavities and over- hangs. However, it seems likely that Cidaris with its large primary spines, was EXPLANATION OF PLATE 76 Elements of the upper surface oyster/ Apiocrinus community of the Bradford hardground. Fig. 1. Oysters heavily bored by Lithophaga sp., (J 40037), x T7. Fig. 2. Root of silvide ectoproct Collapora sp., (J 40038), x 20. Fig. 3. Encrusting foram Nubeculinella sp. on oyster, (J 40039), x 20. Fig. 4. Exogyra crassa (Smith) and Plicatula fistulosa (Morris and Lycett), (J 40032), x 1-8. Fig. 5. Apiocrinus holdfast, (J 40040), x 1 . Fig. 6. Section through hardground and overlying thick biogenic layer (junction arrowed); biogenic layer is composed of oysters (dark), which have been bored by Lithophaga (light, with rounded bottoms), (J 40041), x 2 1. PLATE 76 PALMER and FURSICH, Middle Jurassic hardground 518 PALAEONTOLOGY, VOLUME 17 unsuitable to cavity dwelling, and was active on the top surface of the hardground. Similarly, it is unlikely that the pedically attached brachiopods were confined to lower surfaces. They were too bulky for small cavities, and also occur commonly elsewhere where there is no evidence of there having been cavities of any sort (e.g. Elliott 1973). In addition to a preference for either top or bottom surfaces the crevice-living Plicatula is almost always found on the bare hardground rather than on older layers of serpulids and ectoprocts. The only biogenic material on which it grew was other Plicatula. Thus, on crevice roofs there is evidence of some sort of ecological succession, with Plicatula being the first species to colonize a newly exposed surface, but being replaced by a fauna dominated by serpulids, ectoprocts, and Moorellina. A similar succession is not obvious on the upper surface, where oysters may build up a biogenic layer, up to 2 cm in thickness (PI. 76, fig. 6). However, it seems that oyster growth at any one point was not a continuous process: sections through the oyster layer fre- quently show Lithophaga crypts, of which only the bottom quarter or so remain (PI. 75, fig. 4). The inference is that the material in which the upper part of the crypt was located, has been eroded away. At the present day, it has been estimated that bio- erosion in limestone can account for the removal of between 2 and 10 mm of sediment per year (Warme et al. 1971). A similar figure, applied to the Bradford oyster layer in a single season of spatfall, could easily account for the truncation of the Lithophaga borings. When the next heavy spatfall occurred the truncated crypts were smothered, and the thickness of the oyster layer again increased. At the time when the Bradford hardground was swamped by the first influx of the overlying clay, the oyster layer was undergoing a marked bioerosive phase. Nearly all the oysters seen are represented by their attached valves only, and most are heavily bored by ctenostomes and clionids. In places, truncated Lithophaga borings abound (PI. 76, fig. 1). The most abundant encrusting form alive at this time was the fora- minifer Nubeculinella. Although the nature of the two communities which inhabited the hardground are very different, they are similar in diversity. Seventeen species occur on each com- munity, and to each community, eight species are exclusive (Table 1). The main difference between the two is biomass. We have already noted the crustose nature of the members of the serpulid/ectoproct community, as opposed to the more arbor- escent forms on top, but the former are also considerably smaller. Consequently, EXPLANATION OF PLATE 77 Elements of the lower surface serpulid/ectoproct community of the Bradford hardground. Fig. 1. Ectoproct Stomatopora dichotoma (Lamouroux), (J 40042), x 4-7. Fig. 2. Serpula (Tetraserpula) sp., (J 40043), x 2-2. Fig. 3. Serpula ( Dorsoserpula ) sp. 1, (J 40044), x 2-5. Fig. 4. Thecideacean Moorellina sp. (left) and encrusting calcisponge (right), (J 40045), x 13. Fig. 5. Ectoproct Mesenteripora sp., (J 40046), x 14. Fig. 6. Ectoproct Plagioecia sp., (J 40047), x 4. Fig. 7. Attached valve of Plicatula sp., (J 40048), x 1-5. Fig. 8. Serpula ( Cycloserpula ) sp. 1, (J 40049), x 4-6. Fig. 9. General view of serpulids, (J 40031), x 1-8. PLATE 77 PALMER and FURSICH, Middle Jurassic hardground 520 PALAEONTOLOGY, VOLUME 17 the biogenic layer of oysters and Apiocrinus on the upper surface is always thicker than that composed of Plicatula , ectoprocts, and serpulids on the lower, in spite of the top’s greater rate of bioerosion. DISCUSSION Factors responsible for the distribution of the hardground fauna The polarization of the Bradford hardground fauna into an upper surface oyster/ Apiocrinus community and a crevice serpulid/ectoproct community is attributed to three abiotic factors : cavity size, light intensity, and degree of turbulence. Linked to these are biotic factors, such as competition for space and competition for food. Reidl (1966) has given an extensive review of the faunas found in submarine caves in the Mediterranean. He has discussed the nature of the faunas, and also the varia- tions in biotic and abiotic factors within the cave microenvironment which affect the faunal distribution. Hartman and Goreau (1966) have discussed the importance of encrusting sponges in the cryptic habitats within Jamaican reefs, and Jackson et al. (1971) discussed sponge/thecidacean brachiopod communities found on the undersides of foliaceous corals, and on the insides of caves, in pantropical reefs. Jackson et al. also mentioned the occurrence of cheilostome ectoprocts and serpulids in these habitats, and stressed the relative paucity of bivalves, which are common on other parts of the reef. Garrett et al. (1971), however, recorded that the attached valves of Spondylus americanus are extremely common in the ‘gloomy’ cavities in Bermuda patch reefs, where Chama macerophylla, Isognomon radiatus , and Lithophaga nigra are also found. Bermudan gloomy and dark cavities are also colonized by sponges, ectoprocts, and serpulids. The size of a cave determines the size of its fauna. Forms like Apiocrinus were far too big to grow in the small cavities of the Bradford hardground. The same is true to some extent of the arborescent bryozoans and some large oysters. The light intensity in crevices is substantially lower than that on upper surfaces. Within crevices the ceiling and the back form the darkest parts and therefore provide an ideal settlement area for shade-loving forms. Many larvae of the marine sessile epibenthos can distinguish between light of various intensity for settlement (Riedl 1966, p. 348). In the Bradford hardground fauna, the preference of serpulids and crustose ectoprocts for undersurfaces may partly be due to this fact. Recent ectoprocts are known to show a zonation according to light intensity in caves, and Gautier (1961) describes three species from the western Mediterranean which are particularly shade- loving. Unfortunately, no observations are available on the distribution of Recent serpulids as a function of light intensity. Turbulence is the third important abiotic factor influencing the distribution of this hardground fauna. The degree of turbulence is far lower in crevices (especially small ones) than on the sea floor, even in a high-energy environment (Riedl 1966, p. 276). Consequently, faunal elements which prefer a quiet environment are found in the crevices (cryptophilic forms), whereas other members of the fauna which prefer a high-energy environment are found on the current-swept sea floor (acrophilic forms). Most oysters of the Bradford hardground fauna belong to the latter group. PALMER AND FURSICH: JURASSIC HARDGROUND 521 Even when found in the cavities, they prefer edges and well-exposed overhangs (above). Competition for space, an important biotic factor, is correlated with light intensity. Low light intensity results in a decrease of the algal cover and the epiphyton (Riedl 1 966, p. 370). On upper surfaces with a high light intensity, fast-growing algae threaten to overgrow other sessile benthos. In crevices with a low light intensity this danger is far less pronounced, and crustose serpulids and ectoprocts suddenly become competitive and dominate the fauna. This may well have been the case in the Bradford hardground, though no evidence of an algal cover has been preserved. On upper surfaces, only relatively large crustose forms (e.g. Exogyra crassa , very large speci- mens of Cycloserpula ) or silvide forms (e.g. Apiocrinus and arborescent ectoprocts) were able to compete with algal growth. Also competitive are very small crustose forms with a very short life-cycle (Riedl 1966). The mass occurrence of the encrusting foraminifer Nubeculinella on upper surfaces is therefore well in agreement with observations in the Recent. Boring organisms, like Cliona , Lithophaga , and some polychaetes, which are common on upper surfaces of the Bradford hardground, also compete successfully and are found in great numbers even under a dense algal cover in the Recent (Riedl 1966, p. 370). This is because there is less danger of their being overgrown by encrusting forms, which are excluded by the dense algal cover. Competition for food, another biotic factor, is correlated with turbulence. In Recent caves there is a distinctive zonation of the benthonic filter-feeders according to their filtering capacity (Riedl 1966, p. 394): at the entrances passive filter-feeders predominate, but these are replaced by half-passive and finally by active filter-feeders towards the back parts of caves. The fact that all forms found on crevice roofs in the Bradford hardground are active filter-feeders is not, therefore, surprising. Turbulence is also responsible for the biomass differences (expressed by the thick- ness of shell layers) between upper and lower surfaces of the Bradford hardground. The ‘Konsumationszeit’ (the time in which the fauna filters the water content of a cave of a given size) is shorter the smaller is the cave (Riedl 1966, p. 392). The relatively small Bradford hardground cavities could only support a low biomass density. In conclusion, it becomes apparent that a combination of several biotic and abiotic factors was responsible for the polarization of the Bradford hardground fauna. It is difficult to decide which of the factors discussed above governed the distribution of a particular faunal element. We assume that in most cases several factors were responsible for distributions on the hardground. Further evidence for polarization of communities on hard surfaces The Bradford situation is not unique and a similar situation occurs in the White Limestone (Middle Bathonian) at Loss Cross quarry, Calmsden (SP 056091), where a crevice system is locally present below an extensive layer, about 3-5 cm thick, of cemented lime-sand. The lower surface is colonized by an almost identical fauna to that at Bradford, except that the calcisponge is more abundant. The upper surface, however, is encrusted only by Liostrea and Nanogyra sp. and bored only by Litho- phaga. Apiocrinus, Nubeculinella, and arborescent ectroprocts are absent so that the diversity of the upper surface fauna is lower than at Bradford. Bysally attached F 522 PALAEONTOLOGY, VOLUME 17 bivalves and pedically attached brachiopods, associated with the hardground at Bradford are also absent. The absence of the diagnostic brachiopods seems to be the general rule in the Oxfordshire-North Cotswolds province of the Middle Bathonian; they probably preferred the offshore regions, where the water was clearer and better circulated, to the nearshore regions which were more lagoonal, and contained much suspended lime-mud. The Foss Cross hardground is continuous around the whole quarry (about 150 m) but cavities are only developed locally. Elsewhere, however, the hardground is penetrated by open burrow systems. Presumably, these were excavated when the surrounding sediment was still soft, and were preserved in an open condition as cementation proceeded (see Shinn 1969). The walls of these open burrows are encrusted by Moorellina , an encrusting calcisponge, Plagioecia sp., Serpula { Cyclo - serpula) sp., and Stomatopora sp. Presumably, conditions of light and water circula- tion inside these burrow systems were similar to those in the hardground crevices. Another example of an apparent cavity fauna is from the Middle Bathonian at Tytherley Farm Lane, Wiltshire (ST 769594). A single piece of limestone (H. S. Torrens collection HT 1153), about 3 cm thick and 100 cm2 in area, is encrusted on both sides. The orientation may be determined from a geopetal fill in one of the Lithophaga crypts. The upper surface is encrusted by large oysters, Plicatula sp., Atreta sp., and small crustose Isastraea. The lower surface is covered with encrusting calcisponges. Moorellina sp., and Serpula {Cyclo serpula) sp. Lithophaga borings penetrate the hardground from both sides. Synsedimentary lithification of thin layers of carbonate sand with subsequent colonization of both the top and the bottom surfaces, has been described from the Middle Jurassic of the Paris Basin by Purser (1969). Although he does not discuss details of the respective faunas of the two surfaces, Purser does illustrate a bottom surface which is crowded by Serpula {Cyclo serpula) sp. and bored by Lithophaga. Outside the Middle Jurassic, this preference of serpulids for undersurfaces has also been noted. Hallam (1969) figures serpulids on undersurfaces from the Coinstone (Lower Lias) of Dorset, and Kennedy and Klinger (1972) note a similar preference in the Cretaceous of South Africa. Voigt (1959) in his discussion of Upper Cretaceous hardgrounds, mentions examples (e.g. from the Maastricht region) with a rich encrust- ing and boring fauna from both upper surfaces, and from burrow and crevice walls. However, he does not differentiate between the two faunas. Hardgrounds in the Corallian of central England have also been studied by one of us (F. T. F.). An example from Cothill quarry (SP 467997) again shows a clear polarization of top surface and crevice-dwelling faunas. We have also observed a similar phenomenon in Jurassic patch reefs, where the top and the bottom of the corals or coralline sponges which constitute the bioherm, support different communities. The large knolls of Isastraea in the Corallian of England and northern France support a fauna dominated by ectoprocts and serpulids on the lower surfaces, whereas the upper surfaces are colonized primarily by bivalves. Similarly, in the Bathonian sponge reefs at St. Aubin, Calvados, Atreta and arborescent ectoprocts are dominant on top of the sponges, whereas the familiar association of crustose ectoprocts, Moorellina sp. and small sponges, as well as a Spirorbis , predominate underneath. PALMER AND FURSICH: JURASSIC HARDGROUND 523 A similar polarization of faunas on upward- and downward-facing calcareous substrates is known from the Palaeozoic. Koch and Strimple (1968) describe a dis- continuity surface with crevices from the Upper Devonian of Iowa. The upper sur- faces support Spirorbis, edrioasteroids, Aulopora, stromatoporoids, together with Trypanites borings, whereas the under surfaces support cystoids, ? Aulopora and Trypanites (additional information from C. R. C. Paul, pers. comm.). We have also seen a massive stromatoporoid in Mr. J. M. Hurst’s collection from the Silurian of Gotland, where Spirorbis seems to be confined to the lower surface, whilst tabulate corals, ectoprocts, and Trypanites borings occur on the upper surface. In conclusion it would seem that the distinction between exposed surface encrusting and boring faunas, and those dwelling in cryptic crevices, has been maintained at least from the Silurian. This distinction applies to inorganic substrates (such as a hardground or a pebble), and to organically formed substrates (such as scler- actinians or foliaceous calcisponges). In the example from Bradford the upper surface fauna is an Apioerinusj oyster dominated community whilst the crevice community is dominated by serpulids and encrusting ectoprocts, with Moorellina, Plicatula , and encrusting calcisponges. Acknowledgements. The authors are indebted to Professor R. E. Garrison, Dr. R. Goldring, Dr. W. J. Kennedy, Dr. W. S. McKerrow, and Dr. D. J. Shearman, for discussion and encouragement during the preparation of this paper, and to Dr. H. S. Torrens for additional material from the Canal Quarry and for much information on all aspects of the Bradford Clay. We also wish to acknowledge the pioneer work of Dr. Alan Kendall on Cotswold Middle Jurassic hardgrounds. REFERENCES bathurst, r. G. c. 1971. Carbonate sediments and their diagenesis. Developments in Sedimentology, 12, 620 pp. Amsterdam. cunnington, w. 1859. On the Bradford Clay and its fossils. Wilts, archaeol. nat. Hist. Mag. 6, 1-10. elliott, G. f. 1973. A Palaeoecological Study of a Great Oolite Fossil-Bed (English Jurassic). Proc. Geol. Assoc. 84, 43-51. elton, c. a. 1966. The Pattern of animal communities , 432 pp. London. fursich, f. t. 1971. Hartgriinde und Kondensation im Dogger von Calvados. N. Jb. Geol. Paldont. Abh. 138, 313-342. Garrett, p., smith, d. l., wilson, a. c., and patriquin, d. 1971. Physiography, ecology and sediments of two Bermuda patch reefs. J. Geol. 79, 647-668. Gautier, Y. 1961. Recherches ecologiques sur les bryozaires cheilostomes en mediterranee Occident ale, 403 pp. Fac. Sci. Marseille. green, g. w. and donovan, d. t. 1969. The Great Oolite of the Bath area. Bull. Geol. Surv. Great Britain , 30, 1-63. hallam, a. 1959. A pyritized limestone hardground in the Lower Jurassic of Dorset (England). Sedi- mentology, 12, 231-240. hartman, w. d. and goreau, T. f. 1966. Jamaican coralline sponges: their morphology, ecology and fossil relatives. In fry, w. g. (ed.). The biology of the Porifera. Symposia Zool. Soc. Lond. 25, 205-243. jackson, j. b. c., goreau, t. f. and hartman, w. d. 1971 Recent brachiopod-coralline sponge communities and their palaeoecological significance. Science , 173, 623-625. Kennedy, w. j. and klinger, h. c. 1972. Hiatus concretions and hardground horizons in the Cretaceous of Zululand. Palaeontology , 15, 539-549, pis. 106-108. koch, d. l. and strimple, h. l. 1968. A new Upper Devonian Cystoid attached to a Discontinuity Surface. Report of Investigations 5 , Iowa Geol. Surv. 49 pp. 524 PALAEONTOLOGY, VOLUME 17 peri am, c. E. 1956. The Jurassic rocks of Mid- Wiltshire. Unpublished Ph.D. thesis, University of Reading. 264 pp. purser, b. h. 1969. Syn-sedimentary marine lithification of Middle Jurassic limestones in the Paris Basin. Sedimentology, 12, 205-230. riedl, r. 1966. Biologie der Meereshohlen , 636 pp. Hamburg and Berlin. shinn, e. a. 1969. Submarine lithification of Holocene carbonate sediments in the Persian Gulf. Sedi- mentology, 12, 109-144. smith, w. 1816. Strata identified by organized fossils, containing prints on coloured paper of the most charac- teristic specimens in each stratum, 32 pp., 19 pis. London. — 1817. Stratigraphical system of organized fossils with reference to the specimens of the original geological collection in the British Museum, explaining their state of preservation and their use in identifying the British strata, 118 pp. London. stinton, f. c. and torrens, h. s. 1968. Fish otoliths from the Bathonian of southern England. Palaeontology, 11, 246-258. TAFT, W. H., ARRINGTON, F., HAIMORITZ, A., MACDONALD, C. and WOOLHEATER, C. 1968. Lithification of modern carbonate sediments at Yellow Bank, Bahamas. Bull. Marine Sci. Gulf Caribbean, 18, 762-828. taylor, J. c. m. and illing, l. v. 1969. Holocene Intertidal calcium carbonate cementation, Qatar, Persian Gulf. Sedimentology , 12, 67-107. voigt, E. 1959. Die okologische Bedeutung der Hardgriinde (‘Hardgrounds’) 1m der Oberen Kreide. Palaont. Z. 33, 129-147, pis. 14-17. - 1972. Uber Talpina ramosa v. Hagenow 1840, ein wahrscheinlich zu den Phoronidea gehoriger Bohror- ganismus aus den Oberen Kreide. Nachrichtien der Akacl. Wiss. Gottingen Math.-Phys. Kl. 93-126. warme, J. E. , scanland, T. b. and marshall, H. f. 1971. Submarine canyon erosion ; contribution of marine rock burrowers. Science, 173, 1127. woodward, h. b. 1894. The Jurassic rocks of Britain, 4; The Lower Oolitic rocks of England (Yorkshire excepted). Mem. geol. Surv. U.K. i-xiv, 628 pp. t. j. palmer and f. t. fursich Department of Geology and Mineralogy University of Oxford Parks Road Oxford OX1 3PR Revised typescript received 24 October 1973 THE LEPIDODENDROID STOMA by B. A. THOMAS Abstract. Detailed investigations have been made of the stomata of several lepidodendroid species using cuticle preparations from compressions together with sections and peels from petrified material. The scanning electron microscope has allowed more critical observation of the compression preparations than has been previously possible. Structures are shown within the guard cells which are tentatively interpreted as the remains of lignified wall thicken- ings. This is the first time such structures have been shown in these plants. For well over a century epidermal characters have been used to help interpret the morphological and taxonomical relationships of many fossil plant organs. More recently such work has been extended to extant plants where it is also proving to be of immense value. Although the same basic ideas govern the study of epidermises from fossil and living material, the actual methods are somewhat different. Living material can be examined in a number of ways, but with fossils the type of preserva- tion controls the method of study. The best possible interpretation of a fossil epidermis is thus often a composite picture achieved by utilizing evidence from several differently preserved specimens. Arborescent lycopods are suitable for epidermal studies as they are found in large numbers and in various types of preservation. The general features of the lepido- dendroid epidermis have been already discussed (Thomas 1966) and used in re- describing certain species of Bothrodendron , Ulodendron, and Lepidodendron (Lacey 1962; Thomas 1967 et seq.). In these studies observations were made on macerated cuticles by transmitted light microscopy which were supplemented by some examina- tion of petrified material. This gave a clear picture of epidermal cell size and stomatal distribution but there were certain internal guard-cell structures which were not fully understood. Lepidodendroid stomata often possess guard cells sunk in stomatal pits, rather like those of many gymnosperms. The cushion cuticles show structures which have been described as ‘dome cells with crests’ by Bartlett (1929). This is an apt descrip- tion for their appearance until it is realized that they project inwards from the general cuticle layer, that is in the same direction as the cuticular ridges of the epidermal anti- clinal walls. They are in fact the cuticles from the sides of stomatal pits, the upper walls of the guard cells, and the outer portions of the guard-cell poral walls (PI. 78, figs. 1, 2). It should not be thought unusual that the poral wall is not continued as a cuticular flange between the fused ends of the guard cells as they appear to be absent in many plants, including some species of the extant Lycopodium sensu lato. Some lepidodendroid stomata, however, show extra structures in the form of a very con- spicuously thickened central oval area, often with a very prominent poral wall ‘crest’, and a separate oval band of thickening running around the central thickening. These outer bands may be continuous showing neither break nor change in thickness in the regions where the anticlinal walls once separated the guard cells. Cuticles from [Palaeontology, Vol. 17, Part 3, 1974, pp. 525-539, pis. 78-83.] 526 PALAEONTOLOGY, VOLUME 17 some species habitually show more stomata with these thickenings although these are not always perfectly preserved. This suggested that the maceration process itself was a major factor in their preservation, so attempts were made to observe its effects. Boulter (1970) showed that maceration affected the preservation of certain guard- cell thickenings in the Taxodiaceae and that it was the alkali clearing treatment which was particularly disruptive. The difficulty encountered in the present study was that Lepidodendralean cushion compressions are more coalified than Boulter's Tertiary Cryptomeria material. Oxidation in Schulze solution neither renders cushion com- pressions transparent nor does it separate the cuticle from the compression. Treat- ment with dilute alkali is therefore normally necessary. Occasionally, however, specimens can be found with relatively robust and uncracked cuticles which will withstand rougher treatment than usual. After oxidation in Schulze solution the compression was removed from these cuticles with a mounted needle and by ultra- sonic-wave treatment using a ‘Pulsatron’ bath. This naturally damages and destroys much of the cuticle but some usable fragments can be obtained which have thus not been treated with alkali. Very brief immersion in alkali was tried out at this point as an attempt at ‘etching’ the compression to highlight certain features such as epidermal anticlinal walls. Optical microscopy has naturally been the traditional means of studying cuticle preparations but recently the scanning electron microscope has been shown to be of immense value for this purpose. Naturally, cuticles are normally observed from their inner surfaces when using the s.e.m. as this side is the one more affected by the under- lying epidermal cells. Epidermal cell anticlinal walls are marked by cuticular ridges and their corners by slender cuticular pegs, while inwardly projecting surfaces such as pits or sunken guard cells appear raised. The cuticles of lycopod leaf cushions were therefore prepared in various ways and examined by light microscopy and by the s.e.m. in an attempt to explain the internal structure of the guard cells. The best results obtained were from specimens of Lepido- dendron veltheimii Sternberg and Lepidophloios acerosus Lindley and Hutton, so these two are described in some detail. DESCRIPTIONS Lepidodendron veltheimii Sternberg Material. No. 2411, Kidston collection. Institute of Geological Sciences, London; from immediately beneath the Orchard Limestone, New Brawden Quarry, Giffnock, Renfrewshire. Upper Limestone Group of the Carboniferous Limestone Series, Lower Namurian. EXPLANATION OF PLATE 78 Lepidodendoid stomata as seen with transmitted light. Figs. 1, 2. The cuticles prepared by Bartlett. Stomata are visible showing their 'dome and crest’ appearance as originally described, x 500. Figs. 3, 5. Lepidodendron veltheimii Sternberg. Stomata showing the dark walls of their pits, x 800. 3, obliquely compressed stoma (compare with PI. 80, figs. 1, 2). 5, horizontally compressed stoma (com- pare with PI. 79, figs. 3-6). Figs. 4, 6. Lepidophloios acerosus Lindley and Hutton. Vertically compressed stomata, x 1200. 4, stoma showing remnants of the central ‘crest’ thickening (compare with PI. 81, figs. 4, 6). 6, stoma showing remnants of the outer thickening (compare with PI. 81, figs. 2, 4). PLATE 78 THOMAS, Lepidodendroid stomata 528 PALAEONTOLOGY, VOLUME 17 This is a specimen which has previously had its cuticle prepared and figured after treatment with Schulze solution and dilute ammonia solution (Thomas 1970, pi. 33, figs. 5, 6; text-fig. 5). The guard cells were described as sunken in pits with some over- arching of the subsidiary cells; an interpretation which is supported by the use of the s.e.m. Small depressions can be just seen with the naked eye on the surface of the cushion compression and these have been interpreted as stomatal pits. Portions of com- pression were therefore removed, cleaned with hydrofluoric acid, and mounted for direct viewing with the s.e.m. Such preparations predictably revealed vast numbers of shallow depressions but also showed that these often contained what appears to be a pair of guard cells (PI. 79, figs. 1, 2). Similar evidence is clearly shown with macerated cuticles viewed from the underside. Those prepared by oxidation and clearing in alkali show the cuticles of the guard-cell outer walls as raised oval areas overlapping the underlying epidermal cell cuticle (PI. 78, figs. 3, 5; PI. 79, fig. 3); while if the guard cells are completely lost the stomatal pit wall is exposed as a layer of cuticle recurved over the general epidermal cuticle sheet (PI. 80, figs. 1, 2). The central guard-cell thickenings seen with transmitted light are usually visible here as flattened oval structures with a marked ridge orientated longitudinally to the guard cells and along the stomatal aperture (PI. 79, figs. 4, 6). Something, however, is clearly affecting these structures as some appear to be well preserved while others are damaged or virtually destroyed (PI. 79, fig. 5). The preparation process was the most likely factor involved here and, in making comparisons by varying the times of acid and alkali treatment, it was found that the longer treatments gave the greater amounts of damage. Complete, or almost complete, avoidance of the alkali stage gave rather different results. The compression is naturally left intact and must be mechanically removed with mounted needles or by immersion in an ultrasonic bath, but as such a method is difficult to duplicate variable results are obtained. Some cuticles retain a layer of cracked compression covering all features except the guard cells (PI. 80, fig. 5), while others show the remnants of isolated pairs of guard cells raised up above the remaining compression (PI. 80, fig. 6). Damage of the lower parts of the guard cells unfortunately often occurs during such preparations presumably accompanying the mechanical removal of the anatomically underlying compression. Limited immersion in alkali has no effect on some stomata (PI. 80, fig. 4), while it partially destroys others (PI. 80, fig. 3). Slightly longer alkali treatment dissolves away more of the guard cells and also more of the remaining compression revealing faint outlines of the epidermal cells. EXPLANATION OF PLATE 79 Lepidodendron veltheimii Sternberg. Figs. 1, 2. Unmacerated cushion compression in surface view. 1, showing stomatal pits; photographed at 45°, x 100. 2, single pit photographed at 25° to the horizontal, x 1000. Figs. 3, 5. Cushion cuticles prepared by oxidation with Schulze solution and clearing with dilute ammonia solution; vertical photographs. 3, epidermal cells and three stomata are visible. The guard-cell cuticles overlap the epidermal cuticle and show remnants of their central thickenings, x 500. 5, single stoma enlarged from fig. 3, x 1400. Figs. 4, 6. Stomata showing well-preserved central thickenings; vertical photographs, x 1400. PLATE 79 THOMAS, Lepidodendroid stomata 530 PALAEONTOLOGY, VOLUME 17 The central portions of the guard cells seem to be selectively dissolved away while the persistent walls show limited disjunction in their terminal regions presumably repre- senting two areas of fusion of the two cells (PI. 81, figs. 1, 3). An outer layer therefore seems to have been removed by the alkali revealing a more resistant inner layer which is presumably thickened in some way. Lepidophloios acerosus Lindley and Hutton Material. No. 764, Kidstone collection, Institute of Geological Sciences, London; from above the Kiltongue Coal, Foxley, near Glasgow, Lanarkshire; communis Zone, Westphalian A. Lepidophloios differs from Lepidodendron in having a vast majority of its leaf cushions bulging outwards and downwards in such a manner that they all partially overlap other cushions below them on the shoot. Therefore, only the cushion surfaces above the leaf scars are normally visible in compression fossils. In L. acerosus there is one further difference in that stomata are restricted to the upper cushion surfaces while the lower obscured surfaces have only elongated epidermal cells and no stomata (PI. 82, fig. 6). No details are given here of specific epidermal characters. These will be better dealt with in a comparative account including details of other Lepidophloios species. Similar methods of preparation were attempted here as were used on the specimen of L. veltheimii. Comparable results were generally obtained although in some ways the guard-cell structures revealed were more spectacular. The more fragile nature of these cuticles prevented their preparation without alkali treatment as they were always virtually destroyed during attempts to mechanically remove the adhering compression. However, limited immersion in alkali gave useful preparations in which the stomata appear entire and almost unaffected even though the epidermal cell anticlinal walls are clearly visible (PI. 82, fig. 5). In contrast, complete maceration with extensive alkali treatment entirely destroyed the guard cells except for the cuticles of their upper walls. This left the stomatal apertures as faint longitudinal ridges or as narrow slits in the guard-cell cuticles (PI. 82, fig. 6). No stomatal pit cuticles are visible so the guard cells were apparently superficial. Immersion for about one or two minutes removed most of the cell structure but left some parts of the guard cells undamaged. This treatment revealed two structures from within the guard cells. An oval plate with a central ridge (PI. 78, fig. 4) and a ridge-like structure in the vicinity of the outer edge of the guard cells (PI. 78, fig. 6). However, very few stomata were seen with both structures as variation in preservation was apparent in every cuticle fragment (PI. 81, figs. 2, 4-6; PI. 82, fig. 1). EXPLANATION OF PLATE 80 Lepidodendron veltheimii Sternberg. Figs. 1, 2. Overmacerated stomata which have lost their guard-cell cuticles. Stomatal pit cuticles can be seen overlapping the epidermal cuticles, x 700. 1, vertical photograph. 2, photographed at 45°. Figs. 3, 4. Macerated stomata after very brief alkali treatment, x 1400. 3, partially destroyed guard cells. 4, complete guard cells. Figs. 5, 6. Macerated stomata after no alkali treatment. The compression has been removed mechanically which has also partially destroyed the guard cells; vertical photographs, x 700. PLATE 80 THOMAS, Lepidodendroid stomata 532 PALAEONTOLOGY, VOLUME 17 Some of the central structures show stages of disintegration like those in L. vel- theimii (PL 82, figs. 3, 4) but in the majority of stomata they are either perfectly preserved or they are absent. They appear to be less firmly attached to the guard cells than in L. veltheimii because they nearly all appear to be undercut at their edges. This might be due to the selective removal of alkali soluble material which eventually leads to their dislodgement before they are destroyed by the macera- tion. The outer circular ridges are similarly affected so that the majority appear to be undercut. DISCUSSION These cuticle studies demonstrate the complex structure of the Lepidodendralean stomata. Their guard cells have internal thickenings which appear to be made of some other substance than cutin as they react differently to maceration and alkali treat- ment. Many other fossil and living plants have guard-cell thickenings which have been described as lignin. Such lignin has been demonstrated in the guard cells of most gymnosperms and some vascular cryptograms and angiosperms (Thomas and Bancroft 1913; Kaufman 1927; Florin 1931; Boulter 1970). Boulter has also illustrated, in his work on the Taxodiaceae, how such thickenings can be lost during cuticle preparations and in particular during immersion in alkali. If the alkali treatment was omitted the thickenings remained attached to the guard- cell cuticles and Boulter suggested that this was due to the fact that lignin is soluble in alkali as shown by Isherwood (1965). However, what is equally or probably more important is that the cellulose, or partially lignified cellulose, between the cuticle and the lignin is even more readily soluble in alkali. This removal of an intermediate soluble layer would then lead to a subsequent detachment of the lignin. Cuticle preparations are fortunately not the only means by which lepidodendroid stomata can be examined. There are numerous specimens of petrified stems which have stomata preserved in their leaf cushion epidermises. The angle of cut of the stem section is naturally important as it determines the type of stomatal section that is visible. A horizontal or radial longitudinal stem section of a Lepidodendron stem will be cut at right angles to the cushion surface giving vertical stomatal sections, while tangential longitudinal stem sections will yield nearly all horizontal stomatal sections cut at varying depths from the epidermal surface. Sections of Lepidophloios are not, however, so easily definable due to the bulging and drooping nature of the leaf EXPLANATION OF PLATE 81 Lepidodendron veltheimii Sternberg. Figs. 1, 3. Macerated cuticle after limited alkali treatment; photographed at 45°. 1, x 200. 3, x 1000. Lepidophloios acerosus Lindley and Hutton. Stomata after maceration and alkali treatment. Figs. 2, 4. Stomata showing remnants of outer bands of thickening and of their stomatal apertures, x 1 500. 2, photographed at 45°. 4, photographed at 25° to the vertical. Figs. 5, 6. One stoma showing a central thickening partially detached from the guard-cell cuticles. 5, photo- graphed at 45°, x 3000. 6, vertical photograph, x 2000. PLATE 81 THOMAS, Lepidodendroid stomata 534 PALAEONTOLOGY, VOLUME 17 cushions. In such specimens single tangential sections will cut stomata in all planes, especially if the cut is slightly oblique (text-fig. 1, PI. 83, figs. 1, 2). Section thickness is also important, but it is not simply a matter of acquiring the thinnest available. Vertical sections are in fact clearer when thin, but horizontal sections often yield more information when thicker as they allow a certain amount of focal depth through the cells. The former are thus better prepared as thin cellulose acetate peels while the latter are better as ground sections. Such observa- tions are also dependent on the perfection of mineralization as we are dealing with internal cellular features. Guard-cell thickenings are not always visible and when present may be incompletely preserved. This does not mean to imply that well-preserved guard cells are very rare, indeed in some sections the very opposite is true. Such thickenings support the evidence accumulated from the cuticle studies. They can be often clearly seen when the guard cells are sectioned horizontally (PI. 83, figs. 6-8) and sometimes when the cut is vertical (PI. 83, fig. 4). In other examples, however, they may not be so obvious. The whole cell may appear dark (PI. 83, fig. 3) possibly because the cut has been made at the ends of the cells taking in the curvature of the outer thickening, or the section may be along a guard cell thereby missing nearly all the internal thickenings (PI. 83, fig. 5). The Lepidodendralean stomata appear to have some comparable form of lignin thickening to gymnosperms, although our knowledge of their detailed structure is not so complete. The central oval structures bearing the ‘crests’ probably consist of one separate thickening from each guard cell, which became fused during the compression of the plant material. Each guard cell would therefore appear to have a band of thickening in the region of curvature from the outer to the poral walls of the cell. The ‘crest’ is apparently solid on all the specimens that I have examined with the s.e.m. text-fig. 1. Diagrammatic reconstruction of a leaf cushion of Lepidophloios. Two types of longitudinal section are shown: A being tangential, and b radial, to the stem axis. Section a cuts the epidermis vertically at y but nearly horizontally at x. EXPLANATION OF PLATE 82 Lepidophloios acerosus Lindley and Hutton. Fig. 1. Stoma showing a central thickening and a remnant of the outer thickening; photographed at 45°, x 3000. Fig. 2. Overmacerated cuticle showing stomata with no remnants of thickenings; vertical photograph, x 500. Figs. 3, 4. Stomata showing partially destroyed central thickenings; vertical photographs, x 1500. Fig. 5. Macerated cuticle with limited alkali treatment showing entire stomata; photographed at 20° to the vertical, x 550. Fig. 6. Cuticle from the lower surface of the leaf cushion showing elongated epidermal cells and no stomata ; vertical photograph, x 500. PLATE 82 THOMAS, Lepidodendroid stomata 536 PALAEONTOLOGY, VOLUME 17 although some appear to be raggedly split at their ends when seen with transmitted light. Indeed, one might always expect a narrow gap representing the stomatal aper- ture. Perhaps a more realistic suggestion would be that the ‘crest’ consists of the poral walls of the guard cells; thus being made of cuticle, cellulose, and lignin— all having become fused together during fossilization. This would then explain the undercutting of the outer edges of these inner thickenings by the removal of the intermediate cellulose wall. The outer ring-like structure can then be interpreted as separate thickenings which are adjacent to the outer walls of the guard cells, although this shape as seen in maceration preparations will have been previously altered during compression. The areas attached to the cuticle are broader than the ‘upstanding ridges’ which suggest that they were keeled curved bands similar to the type found in many gymnosperms. Compression will, however, have shortened the upstanding ridges of these thickenings. text-fig. 2. Diagrammatic reconstruction of a vertical section through a lepidodendroid stoma. Cuticle is shaded black and lignin is densely stippled. EXPLANATION OF PLATE 83 Petrified leaf cushions of Lepidophloios. Figs. 1, 2, 6-8. Lepidophloios scotii Gordon from the Lower Carboniferous, Pettycur Limestone, Fife, Scotland. No. PB 123, Institute of Geological Sciences, London. Figs. 3-5. Lepidophloios fuliginosum Williamson from the Westphalian of Shore, Lancashire. No. V 52017, British Museum (Natural History), London. 1, oblique vertical section through the outer layers of a leaf cushion. Two stomatal pits are visible opening on to the cushion surface, while a third pit is not cut through its aperture, x 400. 2, oblique horizontal section through the epidermis and sub-epidermal cells of a leaf cushion. Many guard-cell pairs are cut showing varying remnants of their internal thickenings, x 400. 3, vertical section through a stoma showing two guard cells sunken in a pit, x 800. 4, oblique vertical section through a stoma showing the sunken guard cells with their internal thickenings, x 800. 5, vertical section through a stoma showing a sunken guard cell cut longitudinally, x 800. 6, horizontal section through a pair of sunken guard cells. The outer ring-like thickening appears darker than the rest of the cell walls, x 1600. 7,8, both are horizontal sections through the guard cells showing remnants of the central thickenings, x 800. Fig. 7 appears to be cut at a lower level than Fig. 8 as it shows what appears to be a ‘crest’ in section. PLATE 83 THOMAS, Lepidodendroid stomata 538 PALAEONTOLOGY, VOLUME 17 Following the reasoning outlined above I have attempted to construct a tentative diagrammatical vertical section through a stoma as it might have appeared in life (text-fig. 2) and have given a series of sections showing the probable sequence of breakdown during maceration (text-fig. 3). a text-fig. 3. A series of diagrammatic sections showing the effects of maceration on lycopod stomata, (a) Stoma after vertical compression before maceration has begun. ( b ) Loss of the sub-epidermal tissues and the commencement of guard-cell disintegration, (c) Epidermal cell loss and breakdown of the guard- cell walls to expose the underlying lignin. ( d) Continued breakdown of the guard cells and the exposure of the epidermal cell cuticles, (e) Reduction of the lignin to remnants attached to the guard-cell cuticles. (f) Removal of all cellulose and lignin from the guard cells, (g) Loss of the guard-cell cuticles exposing the stomatal pit cuticle. The lignin would probably affect the way in which the stomatal aperture opened and closed but it is not yet perfectly clear how it did this. I would, however, tentatively suggest that the combined rigidity of the two separate lignin thickenings held the central parts of the guard cells relatively firm during turgor pressure and volume changes. Turgor pressure increase would presumably lead to an increase in volume tending to force apart the guard-cell walls but the central unfused portions would remain almost unaltered in shape. This would result in an opening of the stomatal aperture. Conversely turgor pressure decrease would result in a closure of the aper- ture. Such a mechanism is very reminiscent of that found in many monocotyledons where the dumb-bell shape of the guard cells reflects the extreme rigidity of their central portions. THOMAS: EEP1DODENDROID STOMA 539 Acknowledgements. I would like to thank the Keepers of Palaeontology of the British Museum of Natural History and the Institute of Geological Sciences for the loan of their specimens. The Scanning Electron Microscope part of the work was made possible by a grant from the Central Research Fund of the Uni- versity of London. REFERENCES bartlett, h. h. 1929. Fossils of the Carboniferous coal pebbles of the glacial drift at Ann Arbor. Pap. Mich. Acad. Sci. 9, 1 1-28. boulter, m. c. 1970. Lignified guard cell thickenings in the leaves of some modern and fossil species of Taxodiaceae (Gymnospermae). Biol. J. Linn. Soc. 2, 41-46. florin, r. 1931. Untersuchungen zur Stammesgeschichte der Coniferales und Cordaitales. K. svenska. Vetensk. Akad. Hand!. Ser. 5, 10, 1-588. isherwood, f. A. 1965. Biosynthesis of Lignin. In pridham, j. b. and swain, t. (eds.). Biosynthetic pathways in higher plants, pp. 133 146. London. kaufman, k. 1927. Anatomie und Physiologie der Spattoffungsapparate mit verholzten Schliesszell- membrane. Planta, 3, 27-59. lacey, w. s. 1962. Welsh Lower Carboniferous plants. 1. The flora of the Lower Brown Limestone in the Vale of Clwyd, North Wales. Palaeontographica, 111 B, 126-160. thomas, b. a. 1966. The cuticle of the Lepidodendroid stem. New Phytol. 65, 296-303. 1967. The cuticle of two species of Bothrodendron (Lycopsida : Lepidodendrales). J. nat. Hist. 1, 53-60. 1967A Ulodendron: Lindley and Hutton and its Cuticle. Ann. Bot. n.s. 31, 775-782. — 1968. The Carboniferous fossil lycopod Ulodendron landsburgii (Kidston) comb. nov. J. nat. Hist. 2, 425-428. 1970. Epidermal studies in the interpretation of Lepidodendron species. Palaeontology, 13, 145-173, pis. 29-34. thomas, h. h. and Bancroft, n. 1913. On the cuticles of some Recent and Fossil cycadean fronds. Trans. Linn. Soc. Lond. 8 (5), 155-204. B. A. THOMAS Department of Biological Sciences University of London Goldsmiths' College New Cross Revised typescript received 12 October 1973 London, S.E. 14. c* LOWER PERMIAN PELYCOSAURS FROM THE ENGLISH MIDLANDS by ROBERTA L. PATON Abstract. Three dentigerous bones from sandstones in the Kenilworth area are described and allocated to different genera of pelycosaurs. One of the specimens, the holotype of Oxyodon britannicus von Huene, 1908, is a sphenaco- dontine, but the generic name Oxyodon is preoccupied and the species is assigned to the genus Sphenacodon Marsh. The second specimen is designated as a new sphenacodontid species, Haptodus grandis\ and the third is identified as belonging to the genus Ophiacodon Marsh — previously known only from North American deposits. The palaeo- ecological and stratigraphical significance of the specimens is discussed; they provide additional evidence of an Autunian age for the Kenilworth Sandstones. Pelycosaurs are the dominant reptiles of the early Permian; they are also among the oldest known fossil reptiles. Several genera have been described from Upper Carboniferous (Upper Pennsylvanian) deposits in North America and western Europe. These include Clepsydrops, a primitive ophiacodont from Illinois (Cope 1875); Stereorhachis , an ophiacodont from Autun in France (Gaudry 1880); some species of Edaphosaurus from North America (Peabody 1957); Petrolacosaurus from North America, tentatively assigned by Romer (1966, p. 372) to the Edaphosauria; Macromerion, an advanced sphenacodont from the Stephanian (Upper Pennsyl- vanian) of Kuonova, Bohemia (Romer 1945); and Milosaurus, a primitive sphena- codont from Illinois (DeMar 1970). Thus members of all three suborders of pelycosaurs were present in Upper Carboniferous deposits and it is obvious that the order itself must have originated at an earlier date, possibly Middle Pennsylvanian (Westphalian A) or earlier. Carroll (1964) has described the earliest known pelycosaur, Proto- clepsydrops, an ophiacodont from near the base of the Pennsylvanian at Joggins, Nova Scotia (Westphalian A and B), although in a later paper (Carroll 1969) he has cast doubts upon its assignation to the Pelycosauria. The vast majority of pelycosaurs come from localities in North America, only a few remains having been collected from western European localities, and two genera of caseids from Russia (Olson 1962). Nothing definitely assignable to the Pelycosauria has been found in other areas. The only pelycosaur previously reported from England is Oxyodon britannicus von Huene ( 1 908), based upon part of a maxilla from probable Lower Permian rocks near Kenilworth. Von Huene (1908, 1925) thought that k Dasygnathus ’ from the Trias of Findrassie near Elgin in Scotland, might be pelyco- saurian but Walker (1964) showed that this specimen belonged to the archosaur Ornithosuchus. No other pelycosaurian bones have been described from the British Isles. In the present study, two more fragments from approximately the same horizon as Oxyodon and which can be definitely identified as pelycosaurian are also described. Their interest lies in the scarcity of pelycosaur material from England, and in showing that a range of different types of pelycosaur was in fact present in English deposits. All the specimens were collected during the nineteenth century. Unfortunately the [Palaeontology, Vol. 17, Part 3, 1974, pp. 541-552, pi. 84.] 542 PALAEONTOLOGY, VOLUME 17 localities are not precisely known but it is believed that there were two— one very close to Kenilworth and the other, one mile north-west of Coventry (grid ref. SP 327 797). Both are probably now obscured by buildings. Abbreviations preceding specimen numbers: GSM = Geological Survey Museum; Gz = Warwick County Museum. SYSTEMATIC PALAEONTOLOGY Order pelycosauria Suborder sphenacodontoidea Family sphenacodontidae Subfamily sphenacodontinae Genus sphenacodon Marsh, 1878 Sphenacodon britannicus (von Huene) Plate 84, figs. 1,2; text-fig. I 1908 Oxyodon britannicus von Huene, p. 431, fig. 1. Holotvpe. GSM 22893 (text-fig. 1b) and GSM 22894 (text-fig. 1a); part and counterpart of a left maxilla. Locality. Kenilworth. Horizon. Kenilworth Sandstone, probably Autunian, Lower Permian. Description. The two specimens, though part and counterpart, do not preserve identical portions of the left upper jaw. GSM 22894 shows an impression of the external surface of part of the left maxilla while 22893 shows the left maxilla itself, and part of the premaxilla, but with the external surface badly broken. The specimens were first noted in the 74th Annual Report of the British Association (Lomas 1904) as ‘Dinosaurian jaws, Trias? Kenilworth’. Von Huene described them in 1908 and recognized that they belonged to a pelycosaur, but he considered that they were from a form closely related to Clepsydrops or Ophiacodon. His description was very brief and several features which he did not mention are visible. Oxyodon was next dealt with by Romer and Price (1940) in their definitive review of the order and they identified it as a member of the Sphenacodontinae. The specimens are embedded in a coarse, red, loosely cemented sandstone containing pellets of red clay. The preserved portion of the maxilla measures 89 mm in length and has a maximum depth of 42 mm. The anterior 10 mm of 22893 are missing from the counterpart. 22893 shows a single small precanine tooth followed by a pair of large canine teeth, these being succeeded by nine postcanine teeth. In addition, anterior to the precanine tooth the impression of another tooth is clearly visible, about the same size as the preserved one. The bone above this impression is broken but it shows traces of the EXPLANATION OF PLATE 84 Figs. 1, 2. Sphenacodon britannicus (von Huene). Kenilworth. Left maxilla, x 1. 1, Institute of Geological Sciences, GSM 22893. 2, Silastomer cast, GSM 22894. Fig. 3. Haptodus grandis sp. nov. Kenilworth. Left maxilla, x 1. Warwick County Museum Gz 1071. Fig. 4. Ophiacodon sp. Coventry. Left lower jaw, x 1. Warwick County Museum Gz 41. PLATE 84 PATON, pelycosaurs 544 PALAEONTOLOGY, VOLUME 17 i i 1 cm text-fig. 1. Sphenacodon britannicus (von Huene). a. Silastomer cast of\ GSM 22894. b. GSM 22893. c.p.— canine pair; MAX— maxilla; max. s.— maxillary swelling; PMAX— premaxilla; prc. st.— precanine step of maxilla; r.c. 2 — root of second canine tooth. maxillary/premaxillary suture and therefore the animal had only two precanine maxillary teeth. At the anterior extremity of the specimen there appears to be an impression of a small tooth— the last premaxillary tooth— situated at a slightly higher level than the first maxillary tooth, and separated from it by a gap of about 1 1 mm. Impressions of all the teeth except the precanines can be seen on 22894. The length of the precanine tooth is approximately 8 mm, that of the canines is roughly 19 mm, and that of the postcanines, which do not show any decrease in size posteriorly, is about 10 mm. All the teeth show the typical sphenacodont features of being laterally compressed, much recurved, and with a sharp posterior cutting edge which is un- serrated. The base of the last precanine tooth is situated about 6 mm above the base PATON: PERMIAN PELYCOSAURS 545 of the first canine— this height may have been even greater as at some time the piece of matrix containing the precanines has been broken off and replaced in a slightly lower position. The bone is badly damaged in this region and this led von Huene (1908) to believe that the last precanine was just erupting and was on the same level as the other teeth. However, the length of the precanines and their height above the rest of the tooth row are sufficiently great to show that the edge of the maxilla curved upwards in the typically sphenacodontine step anterior to the first canine. The maxilla above the two canines is broken, exposing the extremely long roots of these teeth, and it can be seen from 22894 that the surface of the maxilla bulged outwards overhanging the canines in the characteristically sphenacodont swelling which accommodates the enlarged roots of these teeth. No empty sockets were present in the tooth row— two appear to be present in 22893, but in fact the teeth in them are broken off and preserved on the counterpart. The 1st, 3rd, 5th, and 7th postcanine teeth are smaller than the 2nd and 4th. No bones other than the maxilla and a very small part of the posterior end of the premaxilla are visible on the specimen. The maxilla is incomplete posteriorly and thus the tooth count is unknown. Discussion. The teeth and general structure of the maxilla are typically sphenacodont and the precanine step seen here is a feature found only in members of the Sphenaco- dontinae. Romer and Price (1940) correctly placed Oxyodon in this subfamily. As they pointed out, the size and proportions of the specimen are very similar to those in Sphenacodon ferox, a large animal with a skull length of 297 mm from the Lower Permian of New Mexico. Unfortunately, the imperfect nature of the specimen and the lack of more material prevent any further comparison with similar American forms. The only other European sphenacodontine known is Neosaurus cynodus (Gervais 1869) from the Autunian of France. This is a small, slightly aberrant form with a slight maxillary step, and as Oxyodon britannicus is from a large animal which had a pronounced maxillary step it is unlikely that the two can be placed in the same genus. The name Oxyodon proposed by von Huene (1908) is unfortunately pre- occupied by a fish of the family Acropomatidae (Brauer 1906, p. 287) although this fact has just been noticed. It was, therefore, necessary to assign the specimens to either a new, or an existing genus. As pointed out above, the size and proportions of the animal are very close to those of Sphenacodon and the canine teeth of the speci- mens are also of roughly the same proportions as in this genus (those of Dimetrodon are relatively much more prominent). It seems very likely that North America was joined to Europe in the Lower Permian so that the present geographical separation carries little weight. For these reasons it is probable that 1 Oxyodon von Huene, 1908 is a junior synonym of Sphenacodon Marsh, 1878. It was therefore considered appropriate to assign the specimens to the genus Sphenacodon while retaining them for convenience in a separate species, S. britannicus , although no diagnostic characters can be given for this species. The discovery of more complete material might make it possible to assign the material to one of the North American species of the genus. 546 PALAEONTOLOGY, VOLUME 17 Subfamily haptodontinae Genus haptodus Gaudry, 1886 Diagnosis. Primitive sphenacodontids of small size. Skull high and moderately elongated; lower edge of maxilla convex; ventral margin of cheek region strongly curved ; quadrate in low position ; quadrate region wide ; maxilla unexpanded dorsally ; lachrymal extends forward to naris; no maxillary step. Palate of sphenacodont pattern. Lower jaw deep; angular apparently notched. Dentition typically sphenaco- dontid but canines not greatly enlarged. Haptodus grandis sp. nov. Plate 84, fig. 3; text-fig. 2a Holotype. Gz 1071 ; part of the left maxilla (text-fig. 2a). Locality. Kenilworth. Horizon. Kenilworth Sandstone, probably Autunian, Lower Permian. Diagnosis. As for generic diagnosis with in addition, the following points: a large haptodont with skull length about 280 mm ; maxilla slightly swollen antero-laterally above canines; at least three precanine maxillary teeth present; canine teeth large, with poorly developed roots; postcanine teeth large and few in number. Description. This specimen has not been previously described and was identified as IMastodonsaurus. It comes from Kenilworth and is preserved in a coarse, red sand- stone. As in Sphenacodon britannicus, it shows part of the left maxilla with a pair of large canine teeth, but more of the precanine and less of the postcanine tooth row is preserved. The maxilla appears to have been partially disarticulated prior to fossilization as dorsally it bears a large area of overlap for the lachrymal bone. The length of the preserved part of the tooth row is 51 mm and the specimen has a maximum depth of 34 mm. Indications of eight teeth are preserved : three pre- canines which decrease in size anteriorly; a pair of large canine teeth (length about 16 mm), the first apparently loosening in its socket; and spaces for three postcanine teeth. The first postcanine is small and had probably just erupted; the second post- canine is missing; and the third is broken off at its base, but the maxillary surface is broken here and the root of this tooth is visible still firmly fixed to the bone— it was obviously a fairly large tooth with a length of about 10 mm. The teeth are all slightly recurved and laterally compressed with sharp, unserrated posterior edges, and the tips, where preserved, are pointed and laterally compressed, i.e. the teeth are typically sphenacodont. The lower edge of the maxilla is gently convex in side view, the first precanine thus being at a slightly higher level than the canines (about 2 mm above them) but there is no definite step such as occurs in the Sphenacodontinae, and the precanines, although smaller than the postcanines, are not greatly reduced in size. Above the canine pair the maxilla is swollen to accommodate their roots but this swelling is not as pronounced as in Sphenacodon britannicus and other sphenaco- dontines. Neither is the maxilla greatly expanded dorsally as occurs in the latter forms— its greatest depth in this specimen is 34 mm above the 2nd and 3rd post- canines; above the canines it has a depth of 26 mm. PATON: PERMIAN PELYCOSAURS 547 The surface of the maxilla is pitted with foramina for nerves and blood vessels and in many places grooves lead outwards from these foramina. Figures in the literature on pelycosaurs indicate that these foramina and grooves may possibly be more wide- spread in sphenacodonts than in other pelycosaurs. Discussion of Gz 1071. The laterally compressed, pointed teeth with sharp posterior cutting edges; the maxilla with its lateral swelling above the canine teeth; the convex maxillary margin ; the large size of the postcanine teeth ; and the much pitted maxillary surface are features which show that this specimen belonged to the family Sphenaco- dontidae. However, several features exclude it from the subfamily Sphenacodontinae : the lack of a precanine step in the dentition; the number and size of the precanine teeth (3 + ); the dorsally unexpanded maxilla which indicates (a) that the roots of the canine teeth were not exceptionally long as they are in sphenacodontines— this is confirmed by the maxillary swelling being only slight— and ( b ) that the lachrymal reached the naris. It thus seems likely that the specimen belongs to an animal of the genus Haptodus 1 cm. text-fig. 2. a. Haptodus grandis sp. nov. Warwick County Museum Gz 1071. b. Ophiacodon sp. Warwick County Museum Gz 41. a. ov. 1. — area of overlap on maxilla for the lachrymal; c.p — canine pair; DE— dentray; MAX— maxilla; max s.— maxillary swelling. 548 PALAEONTOLOGY, VOLUME 17 first described in 1886 by Gaudry and also known from several other specimens originally described as Pa/aeohatteria (Credner 1888), Callibrachion (Boule and Glangeau 1893), Datheosaurus (Schroeder 1904), and Pantelosaurus (von Huene 1925) but all now considered to be congeneric with Haptodus. Romer and Price (1940) give the characters of the subfamily Haptodontinae as primitive sphenacodontids with a high, moderately elongated skull, a moderately convex tooth row, slightly developed canines, and with the lachrymal reaching the naris. The subfamily contains only the genus Haptodus , the species contained in this being all from western European localities of Autunian age— none, however, have been previously described from England. All known haptodonts are of small size. Two characteristics thus distinguish Gz 1071 from other haptodonts: ( a ) The size suggests that it is only slightly smaller than Sphenacodon britannicus which had a skull length of approximately 300 mm. The skull length estimated from Gz 1071 is approxi- mately 280 mm. Haptodus saxonicus, which is the largest known haptodont, has a maxi- mum skull length of 180 mm. (b) All known haptodonts have only slightly developed canines. Those of Gz 1071 are fairly well developed, being only slightly smaller than those of Sphenacodon britannicus. There is not, however, a great deal of material representing the Haptodontinae and it is probably accidental that so far no large forms have been discovered. There appears to be no reason why a large species of Haptodus should not have existed. Remains of pelycosaurs in Europe suggest that Haptodus is the commonest genus and it is therefore likely that it would have developed forms of different sizes adapted to different modes of life. The relatively large size of the canines in the present speci- men is probably correlated with this larger size. It is therefore suggested that Gz 1071 represents a previously unknown species of Haptodus which, in view of its large size, is here named Haptodus grandis. Suborder ophiacodontia Family ophiacodontidae Genus ophiacodon Marsh, 1878 Ophiacodon sp. Text-fig. 2b Figured Specimen. Gz 41. Locality. One mile north-west of Coventry. Horizon. Kenilworth Breccia, probably Autunian, Lower Permian. Description. Murchison and Strickland (1840) noted the occurrence of this specimen and figured it rather poorly (pi. 28) as the maxillary bone of a fish. The matrix is a coarse, red, loosely cemented sandstone containing pellets of red clay. It shows part of a poorly preserved left dentary. The anterior part extends to the symphysis and the complete tooth row is present, but most of the ventral part is missing and the specimen ends 24 mm posterior to the last tooth. The total length is 122 mm. The impression of the missing ventral part of the jaw remains, permitting the general shape to be seen. The jaw is slender, very shallow dorso-ventrally at the front, and deepening gradually posteriorly. The dorsal edge of the dentary is moderately concave; posterior to the tooth row it becomes convex. PATON: PERMIAN PELYCOSAURS 549 The tooth row has spaces for about 34 teeth, many of which are missing— the state of preservation makes it difficult to distinguish empty sockets from broken bases, but as far as can be determined, the tooth arrangement is as follows: anterior posterior 1234 56 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 ?11 — 1 11— 1- 1 1- 1— 1— I I - 1 1 1 1 1 1 1 1 1 - Because of the curvature of the bone, tooth 1 cannot be seen, but it was probably a large tooth as are teeth 2 and 3. Tooth 3 is complete and is 1 1 mm long. Position 4 probably also contained a large tooth. Teeth 5 to 26 are all roughly equal in size with a height of 5 mm. Teeth 27 to 34 are very much smaller, diminishing in size posteriorly to a minimum of 2 mm. The teeth are not laterally compressed and do not have posterior cutting edges, and teeth 5 to 34 are conical pegs with blunt tips. The complete and enlarged tooth 3 is slightly recurved, however, and has a definite point. Discussion. The slender structure of the jaw and the nature of the teeth are typical of conditions found in the ophiacodont lower jaw whose structure is entirely different to that of sphenacodonts. The group of enlarged teeth at the anterior end of the lower jaw is characteristic of pelycosaurs. The teeth of Gz 41 are smaller and more numerous than those in sphenacodonts and there are many gaps in the tooth row. This is also a feature found in ophiacodonts, probably caused by a slow rate of tooth replacement (Romer and Price 1940, p. 91). The jaw is closest in its structure to pelycosaurs of the genus Ophiacodon. This is interesting as previously ophiacodontids, even the primitive Carboniferous forms, are known only from North American deposits. The portion of jaw preserved indi- cates a skull length of about 185 mm. This is considerably smaller than any of the known American forms. The number of teeth in the jaw is also lower than that normally occurring in the American species, although the number exceeds that which is characteristic for sphenacodonts. Apart from the smaller size and number of teeth, the jaw most closely resembles that of Ophiacodon uniformis, a form from the Wichita Beds which are generally held to be Autunian in age (Romer 1966). The specimen unfortunately does not show sufficiently diagnostic characters to enable it to be established as a separate species, or to be assigned to one of the already described species of the genus. It is quite possible that it could have belonged to a young indi- vidual and size is therefore irrelevant. General discussion Various ‘Upper Coal Measures’ deposits including red continental sediments in the Midlands Coalfields have in the past been thought to be Permian in age, but over the years there has been a general tendency to push these red beds, the upper part of which is the Keele Group, down into the Carboniferous giving them a Westphalian D age. The Keele Group is succeeded by a series of red mudstones, sandstones, and con- glomerates known as the Bowhills Group, the Calcareous Conglomerate Group, and the Gibbet Hill Group according to their locality. These beds have been assigned by various authors (Boulton 1933; Shotton 1929; Whitehead and Eastwood 1927) to either the Carboniferous or the Permian, but it is now generally considered that 550 PALAEONTOLOGY, VOLUME 17 a Stephanian age is possible although Haubold (1970, 1971, 1972) believes that, on the evidence of footprints, the beds are of early Permian age. These conglomerate- bearing strata are overlain unconformably by the Clent and Enville Breccias which have been shown (Shotton 1929) to pass laterally into the Kenilworth Breccia Group (which includes the Kenilworth Sandstone) which rests conformably on the con- glomerates of the Gibbet Hill Group. These breccias have all been tentatively assigned (Hains and Horton 1969) to the Lower Permian. They are, as a whole, virtually barren, the only other fossils found in them being the zatracheid labyrinthodont Dasyceps bucklandi (Lloyd 1850; Huxley 1859; von Huene 1910; Romer 1947) from roughly the same horizon as the pelycosaurs— this fossil (currently under study by the author) is believed to be a specialized member of a family of amphibians whose more advanced species are limited to Lower Permian sediments of Autunian age; some amphibian and reptilian footprints (Haubold 1970, 1971, 1972); and a species of the conifer Lebachia ( Walchia ) (Hains and Horton 1969) which also supports a Lower Permian age for these beds. Because of the rarity of other fossils, the pelycosaur remains described here assume some importance in the determination of the age of the Kenilworth Sandstones. The subfamily Sphenacodontinae includes specimens from North America and Europe and is mainly Autunian in age, the genus Sphenacodon , to which GSM 22893 and 22894 are here assigned, being found only in Autunian (Lower Permian) sediments (Romer and Price 1940). The subfamily Haptodontinae in which Gz 1071 is included is known only from specimens found in western European localities which are all of Autunian age (Romer and Price 1940). Specimen Gz 41 is believed to belong to the genus Ophiacodon which is known only from North American deposits of late Stephanian and early Autunian age (Romer and Price 1940). The above evidence thus points to an Autunian (Lower Permian) age for the Kenilworth Breccia Group. The beds are thought to have accumulated under arid conditions because of the presence in them of pellet rocks, footprints, raindrop impressions, and suncracks. They were produced from material eroded from the near-by Mercian Highlands and were deposited in a series of interconnecting basins, in one of which was formed the Kenilworth Breccia, the others receiving the Clent and the Enville Breccias. In view of these conditions it is hardly surprising that fossils are rare in these deposits. Those which do occur, however, suggest that the fauna of the Mercian Highlands was fairly varied. There must have been plenty of food to support two species of carnivorous pelycosaurs ( Ophiacodon is believed to have been piscivorous). A similar instance occurs in the 'Lower Keuper’ Sandstone which, near Warwick and Bromsgrove, has a relatively rich fauna of amphibians and reptiles plus inverte- brates, plants, and fish, and which contrasts with the barrenness of most of the Trias over considerable thicknesses and wide areas. Many different sorts of tracks thought to have been made by amphibians and reptiles of various sizes occur in the Permian deposits, and it is interesting to note that Haubold (1971) has suggested that the reptilian track known as Ichniotherium found in the English beds corresponds to a member of the suborder Edaphosauria. If this is true, all three suborders of pelycosaurs found in North America are represented in the English Autunian deposits, and there can be no basis for suggestions that the fauna of the latter is restricted. PATON: PERMIAN PELYCOSAURS 551 The occurrence of the hitherto solely North American genera Sphenacodon and Ophiacodon in the English Lower Permian provides additional evidence of the close land links between North America and Europe at this time. Acknowledgements. I thank Dr. A. D. Walker for reading the manuscript and for his many helpful sug- gestions throughout the work, and Mr. A. Milner for bringing to my attention the preoccupation of the name Oxyodon. I am grateful to Dr. W. Allen and Miss J. Morris of Warwick County Museum, and to Mr. D. Butler and Mr. R. V. Melville of the Geological Survey Museum for the loan of specimens. REFERENCES boule, m. and glangeau, p. 1893. Le Callibrachion, nouveau reptile du Permien d’Autun. Acad. Sci. Paris C.R. 117, 646-648. boulton, w. s. 1933. The rocks between the Carboniferous and the Trias in the Birmingham district. Q. Jl geol. Soc. Lond. 89, 53-56. brauer, a. 1906. Die Tiefsee-Fische. I: Syst. Theil. Wissenschaftl. Ergeb. der deutschen Tiefsee-Exped. 15, 1-432. carroll, r. L. 1964. The earliest reptiles. J. Linn. Soc. Zool. 45, 61-83. — 1969. A Middle Permian captorhinomorph and the interrelationships of primitive reptiles. J. Paleont. 43, 151-170. cope, e. D. 1875. On fossil remains of Reptilia and fishes from Illinois. Proc. Philad. Acad. Nat. Sci. 404-411. credner, H. 1888. Die Stegocephalen und Saurier aus dem Rothliegenden des Plauen’schen Grundes bei Dresden; vii Theil; Palaeohatteria longicaudata Crd. Zeitschr. Deutsch geol. Gesellsh. 45, 639-704. demar, r. 1970. A primitive pelycosaur from the Pennsylvanian of Illinois. J. Paleont. 44, 154-163. gaudry, A. 1880. Sur un reptile tres perfectionne trouve dans le terrain Permien. Acad. Sci. Paris C.R. 91,669-671. 1886. Sur un nouveau genre de reptile trouve dans le Permien d’Autun. Bull. geol. Soc. Fr. 14, 430-433. gervais, p. 1869. Zoologie et paleontologie generates. Nouvelles recherches sur les animaux vertebres vivant et fossiles. Paris. hains, a. and horton, a. 1969. British Regional Geology ; Central England. H.M.S.O. London. haubold, h. 1970. Versuch einer Revision der Amphibien-Fahrten des Karbon und Perm. Freiberger For sell. 260, 83-117. 1971. Die Tetrapodenfahrten aus dem Permosiles (Stefan und Rotliegendes) des Thiiringer Waldes. Abh. Ber. Mus. Nat. Gotha , 15-41. 1972. Panzerabdriicke von Tetrapoden aus dem Rotliegenden (Unterperm) des Thiiringer Waldes Sonderdruck aus Geologie. Akad-Verl. Berlin, 21, 110-115. huene, f. von. 1908. Neue und verkannte Pelycosaurier-Reste aus Europe. Centralb. fur Min. Geol. Pal. Jahrb. 431-434. 1910. liber einen echten Rhyncocephalen aus der Trias von Elgin, Brachyrhinodon taylori. Neues Jahrb. fur Min. Geol. Pal. 2, 29-62. 1925. Ein neuer Pelycosaurier aus der unteren Permformation Sachsens. Geol. Pal. Abh. Jena, 18, 215-264. HUXLEY, T. H. 1859. On Dasyceps bucklandi ( Labyrinthodon bucklandi Lloyd). Mem. Geol. Surv. U.K. 52-56. lloyd, G. 1850. On a new species of Labyrinthodon from the New Red Sandstone of Warwickshire. Rept. Brit. Assoc. Adv. Sci. 19, 56-57. lomas, J. 1904. Investigation of the fauna and flora of the Trias of the British Isles. 5th report of the com- mittee. Ibid. 77, 298-313. marsh, o. c. 1878. Notice of new fossil reptiles. Am. J. Sci. 15, 409-41 1. murchison, r. i. and Strickland, H. E. 1840. On the Upper formations of the New Red Sandstone in Gloucestershire, Worcestershire and Warwickshire. Trans, geol. Soc. Lond. 5, 331-348. olson, E. c. 1962. Late Permian terrestrial vertebrates; U.S.A. and U.S.S.R. Trans. Am. Phil. Soc. 52, 1-224. peabody, F. E. 1957. Pennsylvanian reptiles of Garnett, Kansas: Edaphosaurus. J. Paleont. 31, 947-949. romer, a. s. 1945. The late Carboniferous fauna of Kuonova (Bohemia) compared with that of the Texas Red Beds. Amer. J. Sci. 243, 417-442. 552 PALAEONTOLOGY, VOLUME 17 romer, A. s. 1947. Review of the Labyrinthodontia. Bull. Mus. comp. Zool. Harvard, 99, 1-352. — 1966. Vertebrate Paleontology (3rd edition). Chicago. — and price, l. i. 1940. Review of the Pelycosauria. Geol. Soc. Amer. Special Paper, 28, 1-538. schroeder, H. 1904. Datheosaurus macrurus nov. gen. nov. spec, aus dem Rotliegenden von Neurode. Jahrb. konig preuss geol. Landesanstalt, 25, 282-294. shotton, f. w. 1929. Geology of the country around Kenilworth. Q. Jl geol. Soc. Lond. 85, 167-222. walker, a. d. 1964. Triassic reptiles from the Elgin area: Ornithosuchus and the origin of carnosaurs. Phil. Trans. R. Soc. Lond. B , 248, 53-134. whitehead, t. h. and eastwood, t. 1927. The geology of the southern portion of the South Staffordshire Coalfield. Mem. Geol. Surv. U.K. ROBERTA L. PATON Department of Geology Royal Scottish Museum Chambers Street Edinburgh, EH1 IJF Revised typescript received 29 October 1973 THE PRODUCTION OF STR ATIGR APHIC AL RANGE-DIAGRAMS BY AUTOMATIC METHODS by IAN E. PENN Abstract. A package of computer programs has been developed on the Institute of Geological Sciences' IBM 1130 computer which, after accepting and sorting data submitted on a simple input format, draws stratigraphical range and abundance diagrams to any desired scale. The range-diagrams may be of first occurrence; last occurrence; taxonomic order; any specified order including sorting within sub-groups. An ornamented lithological plot is accurately aligned and a correctly ordered, punctuated and type-set list of fossil names may be produced from com- panion dictionaries. The resultant diagram is suitable for direct publication. It is estimated that, the programs having been installed, the computer method produces diagrams in one-tenth of the time and at one-quarter of the cost (within the Institute of Geological Sciences) of manual production. In his review of the familiar faunal range-diagram, Bursch categorized the commonly used kinds of diagram and quoted Moore’s (1948, p. 324) point that the presentation of data, often painstakingly collected, in a single tabular manner may obscure the full significance of the results (Bursch 1950, pp. 479-484). This is probably because ‘the format of the chart itself dramatizes whichever viewpoint is being used’ (McLean 1972, p. 1). Though highly desirable, the production of several diagrams from the same data has never been fully practised perhaps because of the time-consuming nature of the work which is in itself not strictly palaeontological. Automatic data- processing methods are an obvious solution, for example McLean (1972, p. 5) devised a semi-automatic system using relatively cheap card-punch and card-sorting equipment. By mid- 1972 the Institute of Geological Sciences Computer Unit had written a package of computer programs in 1130 Fortran for the display of cartographic and lithostratigraphic data on the commonly used IBM 1130 computer configuration. It was then decided to write programs to produce the conventional faunal range- diagrams and link them with part of the larger package which drew ornamented lithological sections. At the same time dictionaries were established containing full fossil names and the appropriate type-setting instructions so that a copy of the species- list for a particular diagram could be automatically punched on paper tape to be fed into a phototypesetter which can produce, if required, a correctly type-set and punctuated faunal list. This list, together with the diagram, results in the automatic production of a faunal range-diagram suitable for publication. The major features of this palaeontological program package are here out- lined (text-fig. 1) while full program listings may be obtained on request. Full details of the more generalized central part of the package are in Farmer and Johnson (in press). [Palaeontology, Vol. 17, Part 3, 1974, pp. 553-563.] H 554 PALAEONTOLOGY, VOLUME 17 text-fig. 1. Flow chart of the various program packages. L DATA, M DATA, and F DATA represent the input of lithological mineralogical and fossil data respectively. The various identifiers within rectangles represent the various program decks. CUF 1 1 and CUZ 88 produce only line-printer output; CUZ 84 and 85 may produce paper tape. DATA INPUT Data input (text-fig. 2) is short and simple. Lithological data (L DATA in text-fig. 1) consists of ; the thickness of each successive lithological unit followed by the appro- priate two-letter code corresponding to the desired lithological ornament punched on successive cards. For example, text-fig. 2a shows the entries on successive cards of a lithological sequence of (in descending order) 25T2 m of argillaceous limestone; 4-23 m of mudstone; 0T2 m of oolitic limestone; 1-56 m of clay. The whole entry occupies no more than the first nine columns of a standard 80-column punch card. At present some twenty-five ornaments are available while a zero entry results in blank space which may be ornamented manually if desired. The fossil data (F DATA in text-fig. 1) or the mineral data (M DATA) are also presented in stratigraphical succession. Thus the first card of each stratigraphical sample (text-fig. 2b and c) states the number of species or minerals in the sample together with the upper and lower depth limits of the sample; and each subsequent card corresponds to each of the species or minerals in these samples. The species or minerals are not recorded by name but are given a code number (occupying the first four columns of the punch card) which is followed, in the case of species, by a simple PENN: STRATIGRAPHICAL RANGE-DIAGRAMS 555 (1-4) abundance code corresponding to ‘present’, ‘fairly common’, ‘common’, and ‘very common’ (text-fig. 2b). A zero is entered where the occurrence is doubtful. For example, text-fig. 2b shows the entries on successive cards of the first three species (code numbers, 4, 3, and 56) of twelve species determined from a depth range of 17-52- 17-92 m, which are estimated as being present, very common and possibly occurring respectively. The abundance code is the only difference between the fossil and mineral data input formats. This is because the mineral ‘abundance’ may be in the form of percentages (text-fig. 2c) and so one extra column is required on the data card. For example, text-fig. 2c shows the entries on successive cards of the three minerals (code numbers 1, 4, and 5) of a sample at 1 1 -45 m depth which have been found by analysis to con- stitute 85%, 2%, and 10% of the rock. Percentage fossil data can, of course, be input on the M DATA format, while the replacement of abundance codes by equivalent, arbitrary percentage values would enable semiquantitative lithological and/or faunal data to be displayed alongside quantitative data. It is therefore possible to combine lithostratigraphical and biostratigraphical data on the same diagram. It will be seen then that the bulk of the data, whether in L DATA or M DATA format, occupies no more than the first seven columns of a standard data card. Dictionaries giving the full fossil name and corresponding number code are permanently stored by accession order and by taxonomic order on punch cards (text-fig. 2d) as well as on a magnetic disc. It is simple to insert a new card when a name has to be changed or added to the list or to shuffle the deck of cards when the taxo- nomic order has to be changed. A third kind of dictionary entry will be seen between the code number and the fossil name (text-fig. 2d). This is a letter code corresponding to convenient groupings, e.g. B = brachiopod, BR = rhynchonellacean brachiopod. This code enables the computer rapidly to break down large lists of species into meaningful sub-groups if desired. In addition to the code number and full fossil name, type-setting symbols are inserted (text-fig. 2d) so that the output will be in italic or roman type and have upper A lithological data input B fossil data input C mineralogical data input 12345678901234567890 1 234567890 1 234567890 1 234567890 1 23456789 25. 1 2AL 12 17.52 17.92 3 11.45 11.45 4.23MU 4 1 1 85 0. 1 20 L 3 4 4 2 1 56CL 56 0 5 1 0 D dictionary input 1234567890123456789012345678901234567890123456789012345678901234567890 224 BR£/ S%OWERBY) text-fig. 2a-d. Each row of data represents the left-hand side of one 80-column data card and shows the entries in their correct columns (as indicated by the italicized numbers) to produce the four different kinds of input. 556 PALAEONTOLOGY, VOLUME 17 MIDDLE JURASSIC MACROFOSSILS IN ACCESSION ORDER MIDDLE JURASSIC MACROFOSSILS IN MEMOIR ORDER PLANTAE 428 hSolenopora jurassica Nicholson 427 hSolenopora sp. 563 h'Solenopora' sp. 426 halgae [frags] 430lgnwood [frags] 137lgnlignite [frags] PORIFERA 423 s 'Peronidella pistilliform is 'Lamouroux 424 s ‘Cl Iona' sp. 425 ssponge [frags] 0 brKallirhynchia sp. 1 brKallirhynchia superba S.S. Buckman 2 br Rhynchonelloidella sp. 3 brRhynchonelloidella smithi (Davidson) 4 brRhynchonelloidella smithi crassa Muir-Wood 5 brRhynchonelloidella wattonensis Muir-Wood 6bts Avonothyris sp. nov. A 7btsTubithyris sp. 8btsWattonithyris sp. 9btsWattonithyris cf. fullonica Muir-Wood lObtsWattonithyris cf. midfordensis Muir-Wood MbtlOrnithella sp. 12btlOrnithella bathonica (Rollier) 13btlOrn/f/re//a bathonica bathiensis (Rollier) 14bt\Ornithella pupa Muir-Wood 15btlflug/fe/a sp. WbtIRugitela cadomensis (Deslongchamps) 1 7 at Holzbergia sp. 1 8 atHolzbergia schwandorfensis (Arkell) 1 9 atMorrisiceras sp. 20 atMorrisiceras comma (S. S. Buckman) 21 atMorrisiceras krumbecki Arkell 22 atMorrisiceras morrisi (Oppel) 23 atMorrisiceras sphaera S. S. Buckman 24 atMorrisiceras fornicatum (S.S. Buckman) 25 atMorrisiceras skipnum S. S. Buckman 26 atMorrisiceras korustes S. S. Buckman 27 az Procerites sp. 28 azProcerites progracilis Cox & Arkell 29 an Procymatoceras baberi (Morris & Lycett) 30 alTulites sp. 31 atTulites subcontractus (Morris & Lycett) ANTHOZOA 1 88 scCalamophyllia sp. 176 scChomatoseris hemisphericus (Milne Edwards & Haime) 177 scChomatoseris orbulites (Lamouroux) 1 78 scChomatoseris porpites (Wm. Smith) 1 75 scChomatoseris sp. 181 scCladophyllia babeana (d'Orbigny) 1 80 scCladophyllia sp. 1 87 scConvexastrea sp. 195 scDimorpharaea defranciana (Michelin) 199 sc Ewardsomeandra vermicularis Milne Edwards & Haime 198 scEwardsomeandra sp. 174 scMontlivaltia sp. 196 scThamnasteria neptuni (d'Orbigny) 197 scThamnasteria terquemi Milne Edwards & Haime 1 94 scThamnasteria sp. text-fig. 3. Output of fossil dictionary in accession order (to the left) and in taxonomic order (to the right) at 30.9.72. The list has been reduced in scale to the smallest possible size available. or lower-case letters. The typesetter is fed by tape produced from the dictionaries held on the magnetic disc and produces correctly type-set and punctuated lists (text- fig. 3) which it can reduce (or enlarge if desired) so that a large list can be copied at a convenient size. The input characters for typesetter instruction have been carefully chosen so that in other forms of dictionary output (e.g. the fast lineprinter output from the computer or the tape-typewriter print of the paper-tape contents) they will not be printed. These other forms of dictionary lists are then much easier to read than they would be were the redundant type-setting instructions retained. For example, text-fig. 2d shows the dictionary entries for a species (code number 224) of a rhyn- chonellacean brachiopod (BR). The typesetting instructions for upper and lower- case @ and % precede the appropriate letters and the symbols # and / precede the text to be printed in roman or italic print. The symbol £ instructs the typesetter to delete all the entries to its left if required. THE PROGRAMS AND THEIR OUTPUT The programs available within the package are labelled (text-fig. 1) by their program identifiers. CUC 90 plots a series of lithological sections side by side in specified horizontal and vertical relative positions (text-fig. 4) thus providing a skeleton correla- tion diagram. CUC 70 plots individual lithological sections in the correct vertical position relative to plots of stratigraphical range data. Both programs provide PENN: STRATIGRAPHICAL RANGE-DIAGRAMS 557 SPECIMEN DIAGRAM DEPTH IN METRES 10 - 20 - 30 - 40 - 50 - 0 2000 METRES 60 J text -fig. 4. Skeleton correlation diagram as output by CUC 90 in which the sections are correctly positioned stratigraphically and geographically. Labels and tie-lines may be added manually. a separate list of those horizons which are too thin to ornament at the chosen scale. CUZ 10 and CUF 10 read M DATA and F DATA respectively and store them on a magnetic disc for accession by the remaining programs. Thus CUF 1 1 prints out a check-list of the M DATA and F DATA stored by CUZ 10 and CUF 10, and CUZ 88 prints out, via the dictionaries, the full taxonomic name of all the different minerals or species stored by CUZ 10 and CUF 10. CUZ 12 and CUF 12 deduce the stratigraphic range of each mineral or fossil stored by CUZ 10 and CUF 10 and plot range and abundance by first occurrence or last occurrence (text-fig. 5) or by either of those within designated sub-groups (text-fig. 6). It will be seen that the horizontal mark (which is proportional in length to species possibly occurring I I clay 558 cn o o PALAEONTOLOGY, VOLUME 17 SPECIMEN DIAGRAM CO o tv> o J I o I o I 0 0 ooo c 0 0 0 0 0 o * Gervillella sp. ♦ Limatula cl helvetica (Oppe\) ' Cercomya unduiata (J. de C. Sowerby' * Camponectes sp. H Catinula knorri (Voltz) Grammatodon sp Protocardia sp '""t Hr) IT IT' Praeexogyra acuminata (J. Sowerby) Ornithella bathomca (Rollier) Ornithella sp Inoperna plicata (J. Sowerby) serpulids [indet.] Entolium corneolum (Young & Bird) Eomiodon sp. Hybodus sp. Bositra buchi (Roemer) Cucullaea nuarsensis (Cossmann) Oxytoma costatum (Townsend) Catinula matisconensis (Lissajous) Entolium sp. Bositra sp. Rhynchonelloidella smithi (Davidson) 'Corbula' sp. ( juv J Rhynchonelloidella smithi crassa Muir-Wood Rhynchonelloidella wattonensis Muir- Wood Rastellum (Arctostrea) gregareum (J. Sowerby) Rhynchonelloidella sp. Liostrea sp crinoids [indet ] Camptonectes rigidus ( J. Sowerby) I Isocyprina cf. sharpi Cox Modiolus sp Lucina'sp. Anisocardia sp Dacryomya lacryma (J.de C. Sowerby) lignite (frags) Dacryomya sp. Sowerbya sp. Anisocardia bathensis Cox Cucullaea sp. Tancredia (Isotancredia) sp. (juv.) Placunopsis socialis Morris & Lycett Kallirhynchia sp. text-fig. 5. Specimen range-diagram as output by CUZ 85 of last occurrence of Middle Jurassic macro- fossils from a borehole near Bath. DEPTH IN METRES PENN: STRATIGRAPHICAL RANGE-DIAGRAMS 559 SPECIMEN DIAGRAM CLAY MINERALS OTHER MINERALS 6 O text-fig. 6. Specimen range-diagram as output by CUZ 12 showing the grouping of minerals. Calcite (shown manually ornamented above the 50% level) was determined quantitatively by wet analysis and this was used as an internal standard for calibrating the peak heights of the X-ray diffractograms. The analysis was not fully comprehensive. abundance) enables every species occurrence to be localized. Where occurrence is from a range of depth (e.g. when fossils are recorded from ‘Bed X’) the horizontal mark becomes ‘enlarged’ to a box whose height corresponds to the range in depth and whose width corresponds to the abundance (e.g. in MK 16 on text-fig. 7). The code number of each fossil or mineral is plotted for identification but use of program CUZ 85 results in the same plots being produced together with a paper-tape list, in the correct order, of the full fossil names obtained from the appropriate dictionary corresponding to the code numbers on the plot. This paper-tape when fed through the phototypesetter produces a correctly typeset list which may then be affixed to the diagram in preparation for publication. CUZ 84 produces a paper-tape in the same way as CUZ 85 but plots the range and abundance data in fixed-order formats, e.g. 560 PALAEONTOLOGY, VOLUME 17 SPECIMEN DIAGRAM MK 16 text-fig. 7. Upper Jurassic foraminifera from two boreholes in Central England. The data of borehole MK 24, penetrating younger strata, have been arranged by last occurrence as output by CUZ 85. Data of borehole MK 16, penetrating older strata, have been arranged as output by CUZ 84 in the order dis- covered in MK 24. Ammonite subzones have been plotted instead of lithology. Species abundance key as in previous diagrams. MK 24 Ammonite Subzones MK 16 O PENN: STRAT1GRAPHICAL RANGE-DIAGRAMS 561 SPECIMEN DIAGRAM i ! t L text-fic. 7. Upper Jurassic foraminifera from two boreholes in Central England. The data of borehole • u ~4’ Penetralin8 younger strata, have been arranged by last occurrence as output by CUZ 85. Data of borehole MK 16, penetrating older strata, have been arranged as output by CUZ 84 in the order dis- covered in MK 24. Ammonite subzones have been plotted instead of lithology. Species abundance key as in previous diagrams. ■nrw 562 PALAEONTOLOGY, VOLUME 17 according to the current position of species in the taxonomic-order dictionary. In fact, by simply listing the desired sequence of code numbers, a range-diagram of species in any order at all may be drawn (text-fig. 7). PROGRAM PERFORMANCE At present the size limits of the data are 100 (species) x 100 (horizons) and the number of species at any one horizon must be limited to 50. In addition, the maximum size of any sub-groups of the data is 20 and the maximum number of sub-groups is 10. The preliminary listings given by CUF 1 1 and CUZ 88 may be used to inspect data size and to break down larger data into sub-groups of convenient size. Most projects so far handled have been found to be within the total size restrictions of the system. The largest dictionary presently held contains about 600 entries and a limit of 1000 entries is envisaged. Search times for this size of dictionary have proved to be short, so that the number of items is not considered of critical importance. A comparison of time taken by the computer method against manual methods (Table 1) is based on an example containing some 750 data entries, in this case species occurrences. In Table 1, line 1 comprises the time in entering the data on, for example, standard Fortran data sheets. The rate of card punching (line 2) is fast because of the short input format. Operator time (line 3) is considered negligible since the bulk of time is spent in automatic plotting and paper-tape punching. This figure also includes the time involved in printing the names on the phototypesetter. The estimate of time (line 4) for the manual operation is based on a fixed format plot which is quicker (by an estimated 50%) than an initial plot in the first or last occurrence. The draughtsman’s time (line 5) is again a minimum estimate and may be twice as long. Successive dia- grams prepared by hand from the same data (lines 8-10) need complete reprocessing of the data whereas by the computer method it is only prepared once. It is here that most time is saved. It will be seen (Table 1) that the manual time required to produce a diagram is of the order of ten times as long as the time taken by the computer method and increases if several diagrams are prepared from the same data. It has been calculated that this represents about three-quarters saving on costs within the Institute. The use of the package has been presented from the point of view of its main objective, i.e. the speedy and cheap production of stratigraphical range-diagrams, table 1. Comparative times of computer and manual methods. For full explanation see text. Time in hours Manual Computer Data preparation for computer 200 Data punching and checking for computer 100 Computer sorting and plotting 0-75 Total data preparation and plotting 20 3-75 Drawing office production 16 Typing of fossil names 4 Total for first diagram 40 3-75 Total for first and second diagrams 80 4-50 Total for first, second, and third diagrams 120 5-25 Total for first, second, third, and fourth diagrams 160 600 PENN: STRATIGRAPHICAL RANGE-DIAGRAMS 563 but it will be readily appreciated that the package is helpful in other respects, e.g. it has focused on errors in existing manual records; and may be used in analysis as well as in description. Thus appropriate sorting of the data cards enables species lists and plots to be generated for formations and groups or biostratigraphic zones within a region, and the ability to plot in any specified order enables diagrams to be prepared that are directly comparable to diagrams prepared from data from other areas. The dictionaries, while giving some personal nomenclatorial stability, may be output via a phototypesetter to produce simple faunal lists correctly typeset and as up to date as possible. In the long term the computer readable data may be used for input to a computerized data bank. Acknowledgements. The programs for the sorting and display of the stratigraphic range and abundance data were written by Mrs. L. Johnson (Computer Unit, IGS) and that for the display of the lithological logs by Mr. D. G. Farmer (Computer Unit, IGS) who was also responsible for integration of the computer system and the development of ancillary programs. Dr. A. W. Medd and Mrs. B. E. Coleman (Palaeontology Dept., IGS) made text -fig. 7 available, Mr. R. J. Merriman (Petrographical Dept., IGS) provided the data of text-fig. 6, and Dr. T. J. Dhonau (Editorial and Publication Section, IGS) advised on the use of the phototypesetter. The paper is published by permission of the Director of the Institute of Geological Sciences. REFERENCES bursch, J. G. 1950. The range chart as an aid to formaniferal correlation. J. Paleont. 24, 479-484. farmer, d. G. and Johnson, L. In press. Rep. Inst. geol. Sci. mclean, j. d. 1972. A philosophy of data treatment in ecology, stratigraphy and paleocology. Manual of Micropaleontological techniques, 12, 1-257. McLean Paleontological Laboratory, Alexandria, Virginia, U.S.A. moore, R. c. 1948. Stratigraphic paleontology. Bull. geol. Soc. Am. 59, 301 326. IAN E. PENN Department of Palaeontology Institute of Geological Sciences Exhibition Road London, SW7 5DE Revised typescript received 23 October 1973 THE DEVONIAN GENUS KEEGA (ALGAE) REINTERPRETED AS A STROM ATOPOROID BASAL LAYER by ROBERT RIDING Abstract. The Upper Devonian genus Keega Wray, originally described as an alga related to the crustose Coral- linaceae, is reinterpreted to be the structurally modified base, here termed basal layer, of a laminar form of the stroma- toporoid Stachyodes Bargatzky. Besides altering the status of Keega and its significance for the phylogeny of crustose coralline algae this conclusion also implies that Stachyodes adopted a laminar as well as the more common dendroid form and it supports the view that the primary internal structure of Stachyodes is cellular or microreticulate. ‘Tubu- lated’ or ‘striated’ microstructure, common in Stachyodes , is regarded as a product of diagenetic alteration. Basal layers are present in at least two other genera of laminar stromatoporoids and are interpreted to be a morphological adaptation facilitating horizontal as well as vertical growth in laminar forms. Type-material of Keega is compared with specimens from Alberta, Canada, and K. australe is emended and transferred to Stachyodes. Crustose Corallinaceae range from Jurassic to Recent but have been linked phylogenetically with algal groups originating in the Palaeozoic. Lemoine (1911) suggested that the tribes Lithothamnieae and Lithophylleae were both derived from the Solenoporaceae. This has been restated in evolutionary schemes proposed by Johnson (1960, pp. 62-63) and Wray (1970, fig. 17) which differ in detail but both derive Lithothamnium from Solenopora during the early Mesozoic and Lithophyllum and the articulated Corallinaceae from Parachaetetes by way of Upper Palaeozoic genera termed ‘ancestral corallines’ by Wray. The ‘ancestral corallines’ differ from the Solenoporaceae in their smaller size, branching habit, and varied internal struc- ture, and they constitute a much less homogeneous group. Most of them have either large-celled tissue (e.g. Cuneiphycus) or less distinctly cellular tissue with an apparently fibrous structure (e.g. Ungdarella , Foliophycus ) and lack recognizable reproductive structures. Only two Palaeozoic genera, Archaeolithophyllum Johnson from the Carboniferous and Keega Wray from the Devonian, have been reported to have the distinct tissue differentiation and reproductive structures which suggest a close affinity with crustose coralline algae. Wray (1967, p. 16) described Keega as having thick coaxial hypothallial tissue, thin perithallial tissue, and single-apertured conceptacles. He compared it with the extant genus Lithophyllum Philippi. If this interpretation is correct then Keega is not only one of the earliest ‘ancestral corallines’ but also one of the most advanced structurally. However, doubt has subsequently been cast on the algal nature of Keega by Wray and Playford (1970, p. 548) who noted that the supposed perithallial tissue ‘is com- posed of unusually large cells and has a stromatoporoid-like appearance’. They recommended further study of Keega but considered it to be a valid taxonomic entity and still tentatively assigned it to the red algae. Specimens encountered during study of calcareous algae from the Upper Devonian Ancient Wall reef complex in Alberta, Canada, necessitate substantial modification [Palaeontology, Vol. 17, Part 3, 1974, pp. 565-577, pis. 85-86.] 566 PALAEONTOLOGY, VOLUME 17 of the original interpretation of Keega. They provide evidence that Keega is a struc- tural modification, here termed basal layer, of the lower part of a laminar form of the stromatoporoid Stachyodes Bargatzky. Examination of type and comparative material of Keega from Western Australia supports this conclusion and warrants emendation of K. australe and its transfer to Stachyodes. This reinterpretation also has significant implications for Stachyodes. It suggests that Stachyodes adopted a laminar as well as the more common dendroid form and by relating Keega and Stachyodes it provides evidence to support Lecompte’s (1952, 1956) view that the primary microstructure of Stachyodes is microreticulate. Production of a basal layer by Stachyodes is a hitherto unrecognized feature of stromatoporoids which is also present in at least two other laminar genera. OCCURRENCE AND MORPHOLOGY OF KEEGA Keega was described from the Upper Devonian of Western Australia (Wray 1967, pp. 16-19) and has since been reported from the Upper Devonian of Alberta, Canada (Wray and Playford 1970; Riding 1972) and Kielce, Poland (Kazmierczak 1971, p. 21). It occurs in marine carbonate environments usually associated with stromato- poroids, corals, foraminifers, and calcareous algae. Type Material. The holotype of K. australe was not examined for this study. Material which could be borrowed included thin-sections F6162 (hypotype, Wray 1967, pi. 3, fig. 2), 2C, 50A, 290 B.H. and 3266A deposited in the palaeontological collec- tions of the Geological Survey of Western Australia at Perth. These specimens are from the Windjana Limestone (Frasnian and Famennian) of the Lennard Shelf near Derby, Western Australia (text-fig. 1 and Wray 1967, text-fig. 1). Specimen 50A is from the east end of the Windjana Gorge at co-ordinates 280, 280 E., 2809, 700 N. (Wray pers. comm.). The Windjana Limestone has been interpreted as reef facies by Playford and Lowry (1966). The thin sections show partially recrystallized bound- stone and grainstone (terminology of Dunham 1962). In the boundstone Keega is associated with tabulate corals ( Alveolites , Syringoporella) and with the algal and foraminiferal genera Epiphyton , Girvanella , Renalcis, Sphaerocodium , and Wethere- della. Keega occurs as subhorizontal layers 1 -2 mm thick and more than 30 mm long. Scattered branches, rounded in cross-section and up to 8 mm long, extend upward from the main skeleton (Wray 1967, pi. 3, fig. 1). Internally Keega consists of numerous crescentic bands of cellular tissue 80-200 p m thick which represent the hypothallus of Wray (pi. 1, fig. 1). These bands are separated by irregular arcs of clear calcite with many short lateral extensions normal to them (PI. 85, fig. 2). The calcite lacks distinct crystal boundaries. The tissue forming the bands is composed of poorly defined ovoid or rectangular cells 30 by 60 pm in size which are subcircular in cross-section. It includes small irregular patches of clear calcite similar to and often united with those defining the bands. Illustrations of the holotype (Wray 1967, pi. 3, figs. 3, 4) show ovoid areas of sparry calcite up to 500 pm across in the upper part of Keega. A short narrow tube extends from each toward the upper surface of the skeleton and Wray interpreted them to be reproductive organs. Of the specimens examined only 50A contains similar structures and they are associated with sinuous tubes, 80-200 pm RIDING: DEVONIAN STROM ATOPOROI D 567 in diameter, which pervade the skeleton (PI. 85, fig. 3). Specimen F6162 contains traces of tubes, 60-100 fxm in diameter (PI. 85, fig. 2), together with larger circular areas of sparry calcite up to 600 /uu across (PI. 85, fig. 1). Some of these are rimmed by radially fibrous or dense, finely granular calcite. The upper surface of Keega is generally a thin layer less than 400 ^.m thick of brownish calcite similar to the clear calcite patches within the tissue and defining the arcs. It is overlain by sediment or by encrusting Sphaerocodium (PI. 85, fig. 1) and corresponds to the perithallus of Wray. In specimen 3266A this layer is thicker, between 100 and 2-25 mm, and has a coarsely reticulate structure with dark ovoid areas 160-200 across (PL 85, fig. 4). The upper surface of F6162 (PI. 85, fig. 1 ) is irregular and it is possible that the specimen is eroded. Ancient Wall Material. The Ancient Wall reef complex outcrops in the folded and thrust Palaeozoic rocks of the Front Ranges of the Rocky Mountains in south- western Alberta, near the town of Jasper (text-fig. 1). The complex is 50 km across and consists of the Frasnian Cairn and Southesk Formations comprising some 400 m of calcarenites and stromatoporoid biostromes. At the south-eastern margin of the complex at Mount Haultain, a narrow zone of stromatoporoid bioherms marks the edge of the Southesk Formation at the transition to a basinal facies (Perdrix and Mount Hawk Formations). The geological setting is described by Mountjoy (1967). Keega is recognizable in thirteen samples from the Cairn, Southesk, and Mount Hawk Formations at Mount Haultain (text-fig. 2). Its apparent localization in the mega- breccias which occur in the lower part of the Mount Hawk Formation adjacent to the biohermal margin of the Southesk Formation is probably due to more intensive sampling of these units. Keega occurs in partially dolomitized stromatoporoid 568 PALAEONTOLOGY, VOLUME 17 text-fig. 2. Section showing Frasnian formations at the south-eastern margin of the Ancient Wall reef complex at Mount Haultain, Alberta. Samples marked by dots contain Keega (1, G.S.C. 33664; 2, G.S.C. 33665), x indicates IHammatostroma (3, G.S.C. 33666). Most specimens of Keega are from the megabrcccias which form tongues in the lower part of the Mount Hawk Formation marginal to the Southesk Formation. Sample 1 is not precisely located but is from the upper megabreccia. wackestones and boundstones occasionally associated with Syringoporella and in packstones and grainstones with crinoids, gastropods, and calcareous foraminifers; In form and internal structure these specimens of Keega are similar to those of the type-material except that they all show upward gradation into a 1 -0-3-0 mm thick layer of diffuse calcite crystals with a coarse columnar or reticulate structure (PI. 85, fig. 5). One specimen (PI. 86, fig. 1) appears to be a fragment with an eroded margin. The junction between the lower and upper layers is a zone of rapid transition in micro- structure from cellular to diffusely granular. Most specimens contain tubes in the axial part of the lower layer and also in the upper layer although they are less well defined there. The tubes occasionally branch dichotomously and are 100-400 fim in diameter with circular, ovoid, or irregularly rounded cross-sections. They are sub- horizontal in the lower layer and subvertical in the upper although individual tubes do not usually exceed 1 mm in length in any one section. Structures resembling the reproductive organs described by Wray occur in the axial part of the lower layer of a few specimens (PI. 86, figs. 1, 2). REINTERPRETATION In most genera of crustose coralline algae the thallus is differentiated into hypothallial and perithallial tissue (see Johnson 1961, pp. 43-45 for a general description of the skeletal structure of crustose corallines). Hypothallial tissue forms the base of flat crusts and the axial part of erect growths. It has large cells commonly arranged in curved or arcuate rows. Perithallial tissue arises from the hypothallial tissue. It has smaller cells usually growing in a regular rectilinear fashion. Flattened circular cavities (conceptacles) containing reproductive structures occur within the peri- thallial tissue. Although the cellular structure of Keega is less well defined than that of crustose Corallinaceae the gross similarity between Keega and hypothallial tissue is striking. But Keega also exhibits several differences from crustose corallines. The appearance of the upper "perithallic’ layer of Keega is, as Wray and Playford (1970, p. 548) noted, RIDING: DEVONIAN STROM ATOPOROID 569 stromatoporoid-like. Its diffuse light-coloured lineated microstructure closely resembles that of Stachyodes. The cavities scattered through the skeleton lack con- sistent size, shape, and position. The few which do resemble the conceptacles reported in the holotype appear to be fortuitous sections of the tubes which form a branching network in several specimens from both Western Australia and Alberta. Tubes (or canals) occur in most stromatoporoid genera although their significance is uncertain and comparisons with both sponge canal systems and hydrozoan zooidal tubes have been made. They have no counterpart in coralline algae. However, the lower ( Keega ) and upper ( Stachyodes ) parts of the specimens have significant structural and microstructural differences which have to be explained if they are to be regarded as parts of the same organism. Keega consists of crescentic bands of cellular microstructure and has canals which are subhorizontal in the encrusting specimens. Stachyodes commonly has a clear granular, vaguely lineated microstructure and vertical canals. These differences can be accounted for by regarding Keega as the lower structurally modified part (basal layer) of a laminar Stachyodes with relatively unaltered micro- structure. The upper, typically Stachyodes , part of the latilamina has a more regular rectilinear structure with its main elements orientated perpendicular to those of the basal layer and is less resistant to secondary alteration. The light-coloured lineated microstructure often considered typical of Stachyodes is strongly altered. Thus the junction between Keega and Stachyodes corresponds to the top of the basal layer and is a zone of relatively rapid transition in both primary megastructure and in alteration effects on the microstructure. The two aspects of this interpretation, viz. skeletal alteration and development of a basal layer, and their principal implications for Stachyodes are discussed in the following section. Algal and stromatoporoidal interpretations of the morphology of Keega are contrasted in text-fig. 3. ALGA I STROMATOPOROID perithallus hypothallus -• f-T, ' zrrr l\jj j basal layer latilamina CONCEPTACLE CANAL MICROLAMINA Keega 1 Stachyodes text-fig. 3. Diagram contrasting algal and stromatoporoidal interpretations of Keega. The original micro- reticulate microstructure is finely cross-hatched, the altered ‘tubulated’ microstructure is unshaded. I 570 PALAEONTOLOGY, VOLUME 17 Specimens which are probably worn, from both Alberta and Australia show that what appear to be laminar forms may be produced by fragmentation and erosion of columnar forms of Stachyodes by the removal of one side of the column (PI. 86, fig. 1) leaving the axial part which has a structure similar to that of the basal layer. However, this interpretation cannot be applied to most of the specimens examined. These lack any sign of erosion (PI. 85, figs. 3, 4, 5) and thus provide the strongest evidence for the presence of basal layers in indubitably laminar forms. SYSTEMATIC PALAEONTOLOGY Gallery, lamina, microlamina, and pillar are used as defined by Galloway (1957, pp. 350-360). Canal refers to the astrorhizal canal of most authors. Terms describing the tissue (solid skeleton) microstructure are defined by Stearn (1966, p. 78). The form and arrangement of the skeletal elements is termed the megastructure (Riding 1974). Specimens from the palaeontological collections of the Geological Survey of Western Australia are prefixed G.S.W.A. Figured specimens from Alberta are deposited in the palaeontological collections of the Geological Survey of Canada (G.S.C.) at Ottawa. Order stromatoporoidea Nicholson and Murie, 1878 Family stromatoporidae Nicholson, 1886 Genus stachyodes Bargatzky, 1881 Type Species. Stachyodes ramosa Bargatzky, 1881 (placed in synonymy with Stromatopora ( Caunopora ) verticillata McCoy 1851 by Nicholson 1886, p. 107). Diagnosis. Tabulate canals with numerous radiating branches which are commonly superposed. Laminae and pillars poorly differentiated; microlaminae thin, dark, intersect tissue at right angles to canals. Tissue microreticulate, commonly altered to granulate calcite with fine lineations paralleling canals. Description. The solid skeleton constitutes the major part of the coenosteum which may be dendroid, fasciculate, massive, or laminar. Internal cavities are restricted to a distinctive canal system. The canals are crossed by straight, bent, or vesiculose tabulae. The original tissue is cellular or microreticulate, formed by rectangular or subspherical ‘cells’ with dimensions of 30-55 by 60- 1 50 p m orientated with their long axes parallel to the canals. The microlaminae are distinct non-cellular layers normal to the canals. Enlargement and coalescence of calcite grains commonly destroys this microstructure to produce a mass of granular pale-brown calcite which lacks dis- tinctive structures but retains vague discontinuous lineations, poor microlaminae, and scattered speckled patches of fine-grained grey calcite. The lineations are arranged radially in cross-sections and longitudinally in axial sections. They are normal to the microlaminae and correspond to the vertical lines of the original microstructure. Discussion. Recognition of a laminar form of Stachyodes supports Lecompte’s (1956, p. FI 26) removal of the genus from the purely dendroid stromatoporoids comprising the Idiostromatidae of Nicholson (1886, p. 74). Paucity of internal spaces in Stachyodes makes distinction between laminae and pillars difficult. The galleries described by some authors are more aptly termed horizontal canals. The resulting RIDING: DEVONIAN STROM ATOPOROID 571 poorly differentiated skeletal structure has been termed reticulated (Nicholson 1886, pp. 34, 107; Lecompte 1952, p. 259; 1956, p. FI 13) or amalgamated (Galloway 1957, p. 350). Distinct thin lines within the solid skeleton were noted by Galloway (1957, p. 445) when he described Stachyodes as ‘composed of thick laminae which are separated by thin dark lines’. He went on to suppose that ‘the laminae themselves represent the interlaminar spaces of most stromatoporoids’. Whether or not inter- laminar spaces (galleries) become filled in some stromatoporoids it is confusing to then term them laminae. In contrast, Lecompte ( 1 952, p. 298 ; 1956, p. F 1 36) apparently regarded the dark lines themselves as laminae. This usage is also inappropriate since lamina in stromatoporoids usually denotes a tabular skeletal element joined by pillars and enclosing galleries, whereas the dark lines in Stachyodes are usually separated by hard tissue. Since they are comparable with the dark microlaminae present in some species of Stromatopora (Stearn 1966, p. 1 1 1 and pi. 17, fig. 7) this term is used for them here. Secondary alteration of the tissue of Stachyodes is indicated by the transition from a distinct cellular or microreticulate microstructure to a poorly defined lineated microstructure. The lineations, also termed striae by Lecompte (1956, p. FI 36), were interpreted to be tubules by Nicholson (1886, pp. 107-108) and subsequently a tubu- late or porous microstructure has commonly been regarded as diagnostic of Stachyodes (Galloway 1957, p. 440; Yavorsky 1957, p. 58). However, in well-preserved material (especially of S. caespitosa and S. paralleloporoides ) from Belgium Lecompte (1952, p. 301) noted that the fine black lines representing the tubules of Nicholson are crossed at right-angles by similar but less well-preserved ones. He compared the resulting microreticulate microstructure with that of Para/le/opora ostiolata and expressed the view that the microstructure of Stachyodes is ‘essentially microreticulate’ (1952, p. 298; 1956, p. FI 13). The tubules have since been described by Stearn (1966, p. 117) as ‘rod-like concentrations (15 ju.m in diam.) of dark specks (1 ^m across)’. Lecompte (1952, p. 301) alluded to, but did not pursue, the idea that microstructural varia- tion in Stachyodes is due to diagenetic effects. The microstructure of Stachyodes in specimens of S. costulata Lecompte, S', crebrum Stearn, S. spongiosum Stearn, and S. thomasclarki Stearn examined from several Devonian areas of western Canada is a mass of poorly developed, irregularly sized calcite grains which contrast with the clear distinct crystals occupying the canals. Stachyodes referred to as Keega provides a link between this altered microstructure and the original one identified by Lecompte by showing them in juxtaposition: the former in the upper part of the latilamina and the latter in the lower part. Preferential preservation of the lower part of the latilamina is an unexplained but characteristic feature of stromatoporoid alteration (Riding 1974) which appears to be independent of the occurrence of a basal layer. The transition between the two microstructures is effected by enlargement and coalescence of calcite grains which merge with and enclose the original microreticulate tissue until only vague patches and lines of it remain as poor microlaminae and as the aligned specks of the ‘tubules’. The degree of alteration commonly affecting Stachyodes corresponds to the advanced to extreme stages in the scheme of progressive alteration of stromatoporoids proposed by Riding (1974). Interest in skeletal alteration in stromatoporoids has focused mainly on evaluating the usefulness of microstructural criteria for taxonomy. Whether failure to recognize 572 PALAEONTOLOGY, VOLUME 17 moderate alteration has led to the erection of poorly delimited or overlapping taxa is not yet clear, but Syringostroma? confertum Stearn is an example of confusion caused by more advanced alteration. Converging alteration effects upon species belonging to several stromatoporoid genera was one of several hypotheses suggested by Stearn (1967, p. 800) to explain the nature of S'.? confertum and has been borne out subsequently (Birkhead and Murray 1970; Stearn 1972, p. 375). Thus, transferral of Stachyodes costulata Lecompte to Syringostroma ? by Fischbuch (1970, p. 1079) was based on secondary rather than original similarities in skeletal structure. Fisch- buch’s illustration (pi. 148, fig. 6) shows the central part of the column with relatively unaltered microreticulate structure grading outwards into an altered marginal zone. Stachyodes australe (Wray 1967) Plate 85, figs. 1-5 1967 Keega australe Wray, p. 18, pi. 3, figs. 1-6; text-fig. 6. Diagnosis. Thin laminar Stachyodes with basal layer exhibiting subhorizontal canals and arcuate microlaminae normal to them. Description. Laminar stromatoporoid usually less than 4 mm thick with rare rounded branches up to 2 mm in diameter and 8 mm long rising from the laminar coenosteum. Basal layer commonly well developed and constituting up to one-third of the thick- ness of the latilamina. Canals subhorizontal in basal layer, curving upward to become vertical. Tissue microreticulate or cellular. Microlaminae prominent, normal to canals; subvertically arcuate in basal layer, horizontal above. Tissue above basal layer commonly altered to indistinct grains of brownish calcite, rarely preserving the original microstructure but retaining the canals. Discussion. The original description of this species was almost entirely restricted to the basal layer. It is emended here to include the thicker upper part of the latilamina with more regular structure and common alteration. The concept of distinct structural modification of the lower part of the stromatoporoid latilamina is new. The term basal layer is introduced for this zone, defined as; the structurally modified basal part of a latilamina characterized by bending or folding, often arcuate, of the laminae EXPLANATION OF PLATE 85 Stachyodes australe (Wray), longitudinal sections except fig. 3. Figs. 1 , 2, 3, 4. Specimens from the Upper Devonian of the Lennard Shelf, Western Australia. 1 , G.S. W. A. F6162, hypotype, Windjana Limestone, Windjana Gorge; showing crescentic bands of cellular or microreticulate tissue and thin upper layer of clear calcite; Sphaerocodium encrusts upper surface at top right which may be eroded, x 20. 2, detail of 1 showing microstructure and arcs of clear calcite pene- trated by a small horizontal canal on the left, x 50. 3, G.S.W.A. 50A, Windjana Limestone, Windjana Gorge; showing sinuous canals and thin upper layer of clear calcite; the section is transverse to the microstructure which appears as a mosaic of small ‘cells’, x 20. 4, G.S.W.A. 3266A, Windjana Lime- stone, 37 km north-east of Christmas Creek; showing thick band of relatively clear altered calcite above basal layer, x 20. Fig. 5. G.S.C. 33664, Mount Flawk Formation (Frasnian), Mount Haultain, Alberta; basal layer grades upward into clear calcite with a vague reticulate structure overlain by a second latilamina lacking a recognizable basal layer; this is overlain in turn by a different stromatoporoid genus (top), x 20. PLATE 85 RIDING, Stromatoporoids 574 PALAEONTOLOGY, VOLUME 17 and microlaminae and subhorizontality of the canals; the pillars and laminae may be thickened and the lower surface is commonly irregular. In Stachyodes australe the basal layer has subhorizontal canals and arcuate microlaminae normal to them; its development is variable (PI. 85, fig. 5). Dendroid forms of Stachyodes commonly have a zone with arched microlaminae in the inner part of the column (e.g. S. costulata , Lecompte 1952, pi. 65, fig. 1 b\ Fischbuch 1970, pi. 148, fig. 7). In eroded specimens (PI. 86, fig. 1) this axial zone may be confused with the basal layer with which it is very similar and probably related. The presence of both basal layers and axial zones in Stachyodes emphasizes the broad similarity between these structural modifications of stromatoporoids and the hypo- thallial tissue of crustose Corallinaceae. Basal layers are also present in specimens of IHammatostroma and to a lesser extent in Actinostroma (see Lecompte 1956, fig. 92, 2, the specimen is apparently inverted). In IHammatostroma (PI. 86, fig. 4) the basal layer shows arcuate laminae which have pillars radiating subhorizontally from them forming large crescents whose tips protrude into the underlying cavity. Parks (1935, p. 28), arguing in favour of a foraminiferal affinity for stromato- poroids, noted ‘the occurrence in many stromatoporoids of a basal layer of chambers quite different from those of the general test. In some species this layer occurs at the base of each latilamina.’ No illustrations of the layer were given but this description corresponds well with that of the basal layer as it is defined here. A structure called the epitheca, also termed peritheca and holotheca, has been described as ‘a thin, wrinkled, basal layer, of finer and different structure than the superjacent normal structures’ occurring at the base of many stromatoporoid coenostea (Galloway and St. Jean 1957, p. 40; Galloway 1957, p. 387). It does not appear to be comparable with the basal layer described here. The origin of basal layers may be linked with a colonizing function. Their curved laminae and microlaminae would have had potential for horizontal as well as vertical growth so that they could spread laterally over new substrates and even overgrow small cavities (PI. 86, fig. 4). Secondary tissue arose perpendicularly to the basal layer and is the more regular tissue usually dealt with in systematic descriptions. Variable development of the basal layer is not readily explained, although it is probably related to differing modes of propogation in stromatoporoids about which very little is known. Some support for this idea is shown by the more common occurrence of basal layers in latilaminae which overlie sediment or cavity than in those resting upon stromatoporoids of the same species. EXPLANATION OF PLATE 86 Stromatoporoids from the Upper Devonian, Mount Haultain, Alberta; longitudinal sections. Figs. 1 , 2, 3. Stachyodes sp., G.S.C. 33665 ; Southesk Formation (Frasnian). 1 , showing basal layer or axial zone and upper altered zone with canals opening on to upper surface; fragment with probably eroded lower margin, x 20. 2, detail of basal layer or axial zone showing branching canal resembling a con- ceptacle, x 50. 3, detail of altered zone contrasting speckly microstructure with distinct calcite crystals filling canal at bottom centre; x 50. Fig. 4. IHammatostroma sp., G.S.C. 33666; Mount Flawk Formation (Frasnian); showing basal layer with crescentic laminae and sub-horizontal pillars overgrowing a cement-filled cavity ; upper part of latilamina is altered (see Riding 1974), x 20. PLATE 86 RIDING, Stromatoporoids 576 PALAEONTOLOGY, VOLUME 17 CONCLUSIONS Study of Keega highlights several aspects of the structure of stromatoporoids and the fossil record of the crustose Corallinaceae. Although Keega superficially re- sembles the hypothallial tissue of certain crustose Corallinaceae and was for this reason originally placed in the Rhodophyta it possesses canals and grades vertically into tissue characteristic of Stachyodes. Megastructural and microstructural dif- ferences between the lower and upper parts of latilaminae showing this gradation are due to both primary differences in megastructure and to subsequent alteration of the microstructure. The significance of these observations can be summarized as follows: 1. Most species of Stachyodes have a dendroid external form but at least one laminar species, S. australe (Wray), can be recognized. 2. S. australe has a structurally modified zone with arcuate microlaminae and horizontal canals at the base of the latilamina which is termed the basal layer. The upper part of the latilamina has horizontal microlaminae and vertical canals. 3. Dendroid forms of Stachyodes commonly have an axial zone which resembles the basal layer in structure but which is surrounded by ‘normal’ skeletal struc- ture with microlaminae parallel to the axis of the column. It is suggested that the axial zone and basal layer are closely related and that they are analogous with the hypothallial tissue of crustose Corallinaceae. 4. Basal layers also occur in ‘IHammato stroma and Actinostroma. They are sug- gested to be a response to the need for horizontal as well as vertical growth during colonization of a new substrate by the stromatoporoid. 5. The skeleton of Stachyodes is often diagenetically altered to a mass of brownish, poorly formed calcite grains. This ‘tubulated’ or ‘striated’ microstructure has generally been regarded as diagnostic of the genus. However, well-preserved specimens show original microreticulate microstructure especially in the basal layer and the axial zone. 6. The basal layer of S. australe, together with a thin upper layer of ‘normal’ but altered structure was originally described as Keega australe Wray and con- sidered to belong to the Rhodophyta because of its resemblance to the hypo- thallial tissue of crustose Corallinaceae. The nature of K. australe as essentially a basal layer rather than a complete latilamina necessitates emendation of the species as well as its reassignment to Stachyodes. 7. Recognition that this species is not an alga removes what was assumed to be one of the earliest and most advanced Palaeozoic ancestors of the crustose Corallinaceae. Acknowledgements. Eric W. Mountjoy and Colin W. Stearn suggested the study of Upper Devonian calcareous algae at Mount Haultain, Alberta, and provided stimulating discussion and help throughout. I especially wish to thank Colin W. Stearn for encouragement and advice which promoted that aspect of the study presented here. I am grateful to J. H. Lord for lending type material of Keega from the collections of the Geological Survey of Western Australia and to E. W. Mountjoy and C. W. Stearn for providing specimens of Stachyodes and ? Ha nun a t o strom a from western Canada. J. A. Fagerstrom, E. W. Mountjoy, J. St. Jean, C. W. Stearn, and J. L. Wray critically read the manuscript and made helpful comments. The RIDING: DEVONIAN STROM ATOPOROID 577 research was carried out during a Postdoctoral Fellowship at McGill University supported by National Research Council of Canada grants (nos. A-2128 and A-2135) to E. W. Mountjoy and C. W. Stearn, and by Chevron Standard Limited. REFERENCES birkhead, p. k. and Murray, J. w. 1970. Actinostroma papillosum (Bargatsky, 1881) a stromatoporoid from the Swan Hills Member of the Waterways Formation (Upper Devonian) of Alberta. J. Paleont. 44, 1067-1070. dunham, R. J. 1962. Classification of carbonate rocks according to depositional texture. In ham, w. e. (ed.). Classification of carbonate rocks, a symposium. Mem. Am. Ass. Petrol. Geol. 1, 108-121, 7 pis. fischbuch, n. r. 1970. Devonian reef-building stromatoporoids from western Canada. J. Palaeont. 44, 1071-1084, pis. 145-149. galloway, j. j. 1957. Structure and classification of the Stromatoporoidea. Bull. Am. Paleont. 37 (164), 341-480, pis. 31-37. — and ST. jean jr., j. 1957. Middle Devonian Stromatoporoidea of Indiana, Kentucky, and Ohio. Ibid. 37 (162), 25-308, pis. 1-23. Johnson, j. h. 1960. Paleozoic Solenoporaceae and related red algae. Colo. Sch. Min. Quart. 55 (3), 77 pp., 23 pis. 1961. Limestone-building algae and algal limestones. Boulder, Colorado. 297 pp., 139 pis. kazmierczak, J. 1971 . Morphogenesis and systematics of the Devonian Stromatoporoidea from the Holy Cross Mountains, Poland. Palaeont. polon. 26, 150 pp., 41 pis. lecompte, m. 1952. Les stromatoporoides du Devonien moyen et superieur du bassin de Dinant, 2. Mem. Inst. r. Sci. nat. Belg. 117, 216-359, pis. 36-70. 1956. Stromatoporoidea. In MOORE, r. c. (ed.), Treatise on invertebrate paleontology. Lawrence, Kansas. Coelenterata, F107-F144. lemoine, M. 1911. Structure anatomique des Melobesiees. Annls Inst, oceanogr., Monaco , 2, 2, 213 pp., 5 pis. mountjoy, E. w. 1967. Factors governing the development of the Frasnian Miette and Ancient Wall reef complexes (banks and biostromes), Alberta. In Oswald, d. h. (ed.), International symposium on the Devonian system. Calgary 2, 387-408, 4 pis. NICHOLSON, h. a. 1886. A monograph of the British stromatoporoids, 1. Palaeontogr . Soc. [Monogr.], 39, 130 pp., 1 1 pis. parks, w. a. 1935. Systematic position of the Stromatoporoidea. J. Paleont. 9, 18-29, pis. 6, 7. playford, p. E. and lowry, d. c. 1966. Devonian reef complexes of the Canning Basin, Western Australia. Bull. geol. Surv. W. Aust. 118, 150 pp. riding, r. 1972. Calcareous algae and some associated microfossils from Ancient Wall reef complex (Upper Devonian), Alberta. (Abs.). Bull. Am. Ass. Petrol. Geol. 56, 648. — 1974. Stromatoporoid diagenesis: outline of alteration effects. Geol. Mag. Ill, 143-148, 2 pis. stearn, c. w. 1966. The microstructure of stromatoporoids. Palaeontology , 9, 74-124, pis. 14-19. 1967. A preliminary study of the distribution of stromatoporoids on the southern flank of the Ancient Wall carbonate complex, Alberta. In Oswald, d. h. (ed.), International symposium on the Devonian system. Calgary 2, 797-806. 1972. The relationship of the stromatoporoids to the sclerosponges. Lethaia, 5, 369-388. wray, j. l. 1967. Upper Devonian calcareous algae from the Canning Basin, Western Australia. Colo. Sch. Mines Prof. Contr. 3, 76 pp., 1 1 pis. 1970. Algae in reefs through time. Proc. N. Am. Paleont. Conv. J, 1358-1373. and playford, p. e. 1970. Some occurrences of Devonian reef-building algae in Alberta. Bull. Can. Petrol. Geol. 18, 544-555, 2 pis. yavorsky, v. i. 1957. Stromatoporoidea Sovetskogo Soyuza, 2. Trud. Vsegei. n.s. 18, 168 pp., 43 pis. (In Russian.) ROBERT RIDING Department of Geology University of Newcastle upon Tyne Newcastle upon Tyne, NE1 7RU Manuscript received 10 May 1973 a TROPHIC GROUP AND EVOLUTION IN BIVALVE MOLLUSCS by JEFFREY S. LEVINTON Abstract. Deposit-feeding marine benthic invertebrates ingest sediments and feed principally upon bacteria, whereas suspension-feeders feed mainly upon phytoplankton. This distinction is important because the predictability of phytoplankton is less than that of within-sediment bacteria. As a result, suspension-feeding populations fluctuate more than deposit-feeding populations. Possible consequences of these differences include: (1) The evolutionary turnover of deposit-feeding groups should be less than that of suspension-feeders. (2) Being more subject to environ- mental perturbations, the longevity of suspension-feeding genera should be less than that of deposit-feeding genera, and (3) trophic structure of deposit-feeding communities should be conservative, with few changes in trophic struc- ture since the early development of the adaptive zone. Preliminary evidence from the fossil record supports these predictions, (i) Bretsky’s interpretation of Palaeozoic community evolution, as being the result of nearshore-offshore differences in environmental predictability can be shown to be strongly influenced by trophic group, (ii) If survivor- ship curves are constructed for genera of bivalve superfamilies, the following mortality rates obtain for genera : Nuculoida (deposit-feeder)— 0-8%/million years, Pectinacea (suspension-feeder)— 1-2%/my, Pteriacea (suspension- feeder)— 1-5%/my, Veneracea (suspension-feeder)— 1-5%/my. Clearly suspension-feeding bivalve genera were shorter-lived, (iii) Finally, Levinton and Bambach have shown a similarity in the ecology of Silurian and Recent deposit-feeding bivalve mollusc communities. In recent years many palaeoecologists have attempted to make evolutionary pre- dictions from ecological premises. A prediction following from ecological arguments is the statement lWe contend that the genetic-adaptive strategy employed by a species population depends in large part on the regularity, direction, and rate of change in environmental stability’ (Bretsky and Lorenz 1970, p. 2449). This and other con- tributions have speculated on the causes of major evolutionary events, such as adaptive radiations (e.g. Valentine 1968; Bretsky 1969; McAlester 1970). It is in this spirit that I present some ideas on the evolutionary consequences of different trophic adaptations. Most marine benthic invertebrates belong to one or the other of two main feeding types: deposit-feeders and suspension-feeders. Deposit-feeders are those forms that ingest sediments, whereas suspension-feeders feed by straining food out of sea-water. Many species cannot be easily classified into one feeding type or the other. For example, the mactracean bivalve, Mulinia lateralis , has a typical suspension-feeding siphon and ctenidia apparatus, but often feeds on food that is resuspended from bottom sediments. Some other species show distinct behavioural switch mechanisms from deposit-feeding to suspension-feeding (e.g. tellinacean bivalves— Brafield and Newell 1961). However, most taxa can be primarily assigned to either the deposit- feeding or suspension-feeding trophic group. In this paper, it is contended that these two trophic groups live under distinctly different regimes of food predictability. The suspension-feeding group is regarded as living with highly unpredictable food supplies, while the deposit-feeders have stable food supplies. This leads to differences in ecological interactions between species. It also implies that the evolutionary history of suspension-feeders should be more erratic than that of deposit-feeders. [Palaeontology, Vol. 17, Part 3, 1974, pp. 579-585.] 580 PALAEONTOLOGY, VOLUME 17 DIFFERENCES BETWEEN DEPOSIT-FEEDERS AND SUSPENSION-FEEDERS In a recent paper, Levinton (1972a) discussed in detail the major ecological differences between the deposit-feeding and suspension-feeding trophic groups. In summary, suspension-feeding species largely depend upon phytoplankton for their food supply. The abundance of this food supply is variable in both time and space. The phenomena of phytoplankton blooms, control of patchiness by currents and water-mixing effects, and seasonal succession of phytoplankton species, creates an essentially unpredictable food supply for benthic suspension-feeders which, being fixed upon the bottom, depend upon whatever happens to be in the immediately overlying water. As a result, suspension-feeders tend to have patchy spatial distributions, and spatially non- random source of mortality (Connell 1955, 1963). The abundance of suspension- feeders may fluctuate strongly over time (Savage 1956; Coe 1953; Levinton 1970; Trevallion, Edwards and Steele 1970). The maximum abundance of suspension- feeders is correlated with parameters related to the optimal physical characteristics of the sediment-water interface, such as its physical stability, and to the lack of bottom mobility of sedimentary grains (Rhoads and Young 1970; Sanders 1958). Suspension-feeders are probably not most abundant where potential food is greatest in abundance (Rhoads 1973). In contrast, most deposit-feeding benthic species depend upon bacteria as their proximal source of food (Fenchel 1970). Bacteria are very abundant in bottom sedi- ments, particularly in muds (Zobell 1938). Unlike phytoplankton they show relatively modest seasonal changes in abundance, as do the organic detrital particles upon which they live (Longbottom 1968; Ockelman 1958). In addition, the abundance of bacteria in bottom sediments is controlled principally by properties of the sediments themselves, as opposed to the abundance of phytoplankton, which is controlled by the overlying water. Finally, the sediment reworking activities and faecal pellet formation of deposit-feeders dramatically homogenizes the sediment, further enhancing the uniformity of the environment of deposit-feeders (Rhoads and Young 1970; Fevinton 1971; Rhoads and Stanley 1965). The mobility of deposit-feeders also permits complete choice among foods, and complete exploitation of a food source. The net result of this relatively predictable trophic network is a set of popula- tions that are usually randomly or uniformly distributed in space (Connell 1963; Gilbert 1970; Fevinton 19726; Holme 1950). Deposit-feeders show uniformity in community composition and structure, abundance being related to parameters con- cerned with food availability (Sanders 1958, 1960; Fevinton 1971 ; Newell 1965). ECOLOGICAL AND EVOLUTIONARY IMPLICATIONS The consequences of marked differences in predictability of food and energy have been discussed by Valentine (1971). Biomes with unpredictable nutrient supplies can be shown to have species with rapidly fluctuating populations, and little niche specialization. Sanders (1968) coined the terms ‘physically controlled’ and ‘bio- logically accommodated' to characterize unstable and highly stable biotic environ- ments, respectively. Although he used these terms to classify major habitat differences (i.e. shelf v. deep-sea), it is clear that within major biomes, such differences may still LEVINTON: TROPHIC GROUP 581 be observed. Suspension-feeders and deposit-feeders often live under similar regimes of temperature and salinity fluctuation. However, they operate under totally different regimes of temporal and spatial variations of food supply. Suspension-feeders should therefore not participate in communities of species whose competitive interactions for food have resulted in niche specificity. This point is not generally accepted in the literature (see Walker 1972). In contrast, deposit-feeders can be expected to show competitive interactions for food with attendant specializations in diet and living position. These competitive interactions have been observed in many studies (Seger- strale 1960, 1962, 1965; Vasallo 1969; Levinton 1969, 1971; Rhoads and Young 1970; Mangum 1964; Sanders 1960). Few studies have ever demonstrated com- petitive interactions for food among suspension-feeders (but see Sanders et al. 1962; and Bradley and Cooke 1958 for an exception). Thus, it is concluded that with respect to food supply, deposit-feeding communities are largely controlled by biological interactions, whereas suspension-feeding communities are controlled by large-scale and unpredictable fluctuations in factors unrelated to interspecific interactions. This statement applies to competition for food. The conservatism of deposit-feeding populations relative to the susceptibility to rapid change of suspension-feeding populations has some evolutionary consequences. Because deposit-feeders control their own substratum characteristics, are food- limited, and do not radically fluctuate in numbers over time, it is expected that the structure of these communities would be established very early in evolutionary time, with few subsequent basic changes. On the other hand, the variable nature of the food supply for suspension-feeders, plus the great changes in the plankton that have taken place during geologic time (Tappan 1972; Tappan and Loeblich 1971), suggest that suspension-feeding populations should have experienced many turnovers. This con- clusion is superficially at odds with the hypothesis that biotic stability is maintained by environmental instability (Bretsky and Lorenz 1969). It is possible that, for a given trophic group, the effect of trophic stability is inherently different from that of variations in the physical aspects of the environment, such as temperature and salinity. Because they depend upon optimal characteristics of the overlying water for feeding, suspension-feeders are probably more susceptible to environmental change than deposit-feeders. Some preliminary evidence suggests that the above predictions are at least con- sistent with observed patterns of evolution and extinction. The trophic structure of protobranch bivalve (deposit-feeding) communities in the Silurian of Nova Scotia is very similar to those of modern, bivalve-dominated deposit-feeding communities (Levinton and Bambach 1969; Levinton and Bambach, manuscript). Furthermore, Bretsky’s (1969) characterization of biotic stability in Palaeozoic benthic com- munities can be reinterpreted in the light of the above arguments. Bretsky notes that offshore communities, inferred to have lived under physically stable conditions, have undergone several biotic turnovers. Nearshore communities living under un- predictable regimes, changed little. However, the offshore communities are dominated almost exclusively by suspension-feeders (brachiopods, ectoprocts, etc.). In addition, at times of biotic turnover in the offshore communities, the suspension-feeding aspects (brachiopods and epifaunal bivalve molluscs) of the onshore communities change as well (Bretsky 1969, p. 56). Thus we might reinterpret the onshore-offshore 582 PALAEONTOLOGY, VOLUME 17 distinction proposed by Bretsky as being rather the result of difference in trophic stability between deposit-feeders and suspension-feeders. Probably, the truth lies somewhere intermediate between these two hypotheses. A final prediction follows from the above arguments. If we examine the fossil record, the relatively tenuous existence led by suspension-feeding taxa should result in their being shorter-lived, on the average. Thus, if we plot a survivorship curve for deposit-feeding genera, the rate of mortality should be less than that of related suspension-feeding groups. One need only have the length of life of all the individual genera. Having the distribution of life-spans, one can consider a group of genera as a cohort and plot the survivorship, as for a single species population (Levinton and Bambach 1970). Bivalve molluscs of the order Nuculoida were selected as a homogeneous deposit- feeding group. Unfortunately, no other bivalve group can be regarded as strictly deposit-feeding. The Tellinacea have both deposit-feeding and suspension-feeding representatives, sometimes within even the same genus (Pohlo 1969). A further complication is that many nuculoid bivalves come from deep-water, confounding their deposit-feeding status with factors related to physical stability. Three suspension- feeding superfamilies, Pteriacea, Pectinacea, and Veneracea, were used for contrast. The first two have representatives back into the Palaeozoic, allowing a potentially parallel history to the nuculoids. The Veneracea have a more recent origin in the Lower Cretaceous. Data were compiled from the Treatise on Invertebrate Paleon- tology (Moore and Teichert 1969). Genera with no fossil record were excluded from the analysis. The results of the survivorship analysis are shown in text-fig. 1. The nuculoids show a constant rate of mortality which is lower than the maximum rates of mortality shown by the Pteriacea and the Pectinacea. The Veneracea also display constant mortality, though higher in rate than the Nuculoida. Therefore, it is concluded that the rate of mortality of suspension-feeding taxa is higher than that of deposit-feeders. The survivorship curves for the Pteriacea and Pectinacea both show a notable break in slope, from high to low to high mortality (text-fig. la). The position of this break correlates with those taxa that are biogeographically cosmopolitan. Apparently, those pteriacean genera that originated in the Triassic produced a number of genera that were cosmopolitan and long-lived. In the case of the pectens, there were two major periods of cosmopolitan dominance: Carboniferous-Permian and Mesozoic. If we subtract those genera classified in the Treatise to be ‘cosmopolitan’, then these breaks in slope disappear from the survivorship curves almost entirely (text-fig. 1/?). The lower rate of mortality still obtains for the Nuculoida, relative to the suspension- feeding groups. It is concluded, therefore, that the ecological characteristics of different trophic groups can lead to differences in the pattern of evolution of these groups. These patterns can be observed in (1) the relatively low ‘generic mortality rates’ of deposit- feeders, relative to suspension-feeders, (2) the slower evolutionary turnover of deposit-feeding groups relative to suspension-feeding groups, and (3) the apparent tendency of some suspension-feeding groups to show periods of cosmopolitan appearances, perhaps correlated with global changes in the plankton (Tappan and Loeblich 1971). It has also been demonstrated that it is possible to partition bio- LEVINTON: TROPHIC GROUP 583 geographic phenomena from other ecological factors, through the use of survivor- ship curves. This latter conclusion may be significant in our future analyses of the major factors controlling evolution. text-fig. 1. Survivorship analysis of (la) the Veneracea (96 genera), Pteriacea (94), Pectinacea (108), and Nuculoida (56). Fig. 1 b indicates the curves obtained when cosmopolitan genera are omitted. From fig. 16, mortality rates are (first 90% of survival): Veneracea — 1-5%/million years, Pteriacea— 1-5%/million years, Pectinacea — 1-2%/million years, Nuculoida— 0-8%/million years. 584 PALAEONTOLOGY, VOLUME 17 Acknowledgements. Conversations with D. C. Rhoads, H. L. Sanders, and Peter W. Bretsky were helpful in the development of these arguments. I am also grateful to Leigh Van Valen for his provision of pre- liminary manuscripts which ingeniously attack the subject of rates of evolutionary turnover. Finally, I thank Sara S. Bretsky and Peter W. Bretsky for critically examining the manuscript. REFERENCES bradley, w. h. and cooke, p. 1958. Living and ancient populations of the clam Gemma gemma in a Maine coast tidal flat. Fish. Bull. U.S. 58, 305-334. br afield, a. E. and Newell, G. e. 1961. The behaviour of Macoma balthica (L.). J. mar. biol. Ass. U.K. 41, 81-87. bretsky, p. w. 1969. Evolution of Paleozoic benthic marine invertebrate communities. Palaeogeography, Palaeoclimatol., Palaeoecol. 6, 45-59. — and lorenz, D. M. 1970. An essay on genetic-adaptive strategies and mass extinctions. Bull. geol. Soc. Amer. 81, 2449-2456. coe, w. r. 1953. Resurgent populations of littoral marine invertebrates and their dependence on ocean currents and tidal currents. Ecology , 34, 225-229. CONNELL, J. H. 1955. Spatial distribution of two species of clams Mya arenaria L. and Petricola pholadiformis Lamarck, in an intertidal area. Eighth Kept Invest. Shell Fish. Massachusetts , 15-25 (unpublished MS.). 1963. Territorial behavior and dispersion in some marine invertebrates. Researches Popul. Ecol. Kyoto Univ. 5, 87-101. fenchel, t. 1970. Studies on the decomposition of organic detritus derived from the Turtle Grass Thalassia testudinum. Limnol. Oceanogr. 12, 443-450. gilbert, w. h. 1970. Territoriality observed in a population of Tellina agilis (Bivalvia: Mollusca). Biol. Bull. mar. biol. lab. Woods Hole , 139, 423-424. holme, n. a. 1950. Population dispersion in Tellina tenuis da Costa. J. mar. biol. Ass. U.K. 29, 267-280. levinton, J. s. 1969. Deposit-feeding associations in Quisset Harbor. Seventh Ann. Rept. Progr. Systematics- Ecology Progr. Marine Biol. Lab. Woods Hole , 34-35 (unpublished MS.). 1970. The paleoecological significance of opportunistic species. Lethaia, 3, 69-78. — 1971. The ecology of shadow water deposit feeding communities. Unpublished Ph.D. Dissertation. Yale Univ. 284 pp. — 1972u. Stability and trophic structure in deposit-feeding and suspension-feeding communities. Am. Nat. 106, 472-486. — 19726. Spatial distribution of Nucula proximo Say (Protobranchia) : an experimental approach. Biol. Bull. mar. biol. lab. Woods Hole, 143, 175-183. — and bambach, R. k. Some ecological aspects of bivalve mortality patterns. Amer. J. Sci. 268, 97-1 12. longbottom, M. R. 1968. Nutritional factors affecting the distribution of Arenicola marina L. Unpublished Ph.D. Dissertation, Univ. of London. mangum, c. p. 1964. Studies on speciation in maldinid polychaetes of the North Atlantic coast. II. Distri- bution and competitive interaction of five sympatric species. Limnol. Oceanogr. 9, 12-26. mcalester, A. l. 1970. Animal extinctions, oxygen consumption, and atmospheric history. J. Paleont. 44, 405-409. moore, r. c. and teichert, c. 1969. Treatise on Invertebrate Paleontology, Part N, v. 1 and 2 (Mollusca 6, Bivalvia). Newell, r. 1965. The role of detritus in the nutrition of two marine deposit-feeders, the prosobranch Hydrobia ulvae and the bivalve Macoma balthica. Proc. zool. Soc. Lond. 144, 25-45. ockelman, w. k. 1958. The zoology of east Greenland: marine lamellibranchiata. Medd. Gnfmland, 122, 1-256. pohlo, R. 1969. Confusion concerning deposit-feeding in the Tellinacea. Proc. malac. Soc. Lond. 38, 361-364. rhoads, D. c. 1973. The influence of deposit-feeding benthos on water turbidity and nutrient recycling. Amer. ./. Sci. 273, 1-23. — and Stanley, D. J. 1965. Biogenic graded bedding. J. sedim. Petrol. 35, 956-963. — and young, d. k. 1970. The influence of deposit-feeding organisms on sediment stability and com- munity trophic structure. J. mar. Res. 28, 150-178. LEVINTON: TROPHIC GROUP 585 Sanders, h. l. 1958. Benthic studies in Buzzards Bay. I. Animal sediment relationships. Limnol. Oceanogr. 3, 245-258. — 1960. Benthic studies in Buzzards Bay. III. The structure of the soft bottom community. Ibid. 5, 138-153. 1968. Marine benthic diversity: a comparative study. Am. Nat. 102, 243-282. — goudsmit, E. M., mills, e. l. and hampson, G. e. 1962. A study of the intertidal fauna of Barnstable Harbor, Mass. Limnol. Oceanogr. 7, 63-79. savage, r. e. 1956. The great spatfall of mussels ( Mytilus edulis L.) in the River Conway estuary in spring 1940. Fishery Invest ., Lond. ser. 2, 20, 1-22. segerstrale, s. g. 1960. Fluctuations in the abundance of benthic animals in the Balthic area. Commentat Biol. 23, 1-19. 1962. Investigations on Balthic populations of the bivalve Macoma balthica (L.). Part II. What are the reasons for the periodic failure of recruitment and the scarcity of the Macoma in the deeper waters of the inner Balthic? Ibid. 24, 1-26. 1965. Biotic factors affecting the vertical distribution and abundance of the bivalve Macoma balthica (L.) in the Baltic Sea. Proc. Fifth Mar. Biol. Symp. 1965, pp. 195-204. tappan, H. 1972. Fluctuating rates of protistan evolution, diversification and extinction. 24th Inti. Geol. Congr. (Abstracts), 245. and loeblich, a. r. 1971. Smaller protistan evidence and explanation of the Permian-Triassic crisis. Bull. Can. Petrol. Geol. 19, 363-364. trevallion, a., edwards, R. R. c. and steele, j. h. 1970. Dynamics of a benthic bivalve. In steele, j. h. (ed.). Marine Food Chains, Univ. California Press, Berkeley, 552 pp. valentine, j. w. 1968. Climatic regulation of species diversification and extinction. Bull. geol. Soc. Amer. 79, 273-276. 1971. Resource supply and species diversity patterns. Lethaia, 4, 51-61. vassalo, m. T. 1969. The ecology of Macoma inconspicua (Broderip and Sowerby, 1829) in central San Francisco Bay. Part I. The vertical distribution of the Macoma community. Veliger, 11, 223-234. walker, k. R. 1972. Trophic analysis: a method for studying the function of ancient communities. J. Palaeont. 46, 82-93. zobell, c. e. 1938. Studies on the bacterial flora of marine bottom sediments. J. sedim. Petrol. 8, 10-18. J. S. LEVINTON Department of Earth and Space Sciences State University of New York Stony Brook New York, 11794, U.S. A. Revised typescript received 6 October 1973 LEAF ANATOMY OF WEICHSELIA BASED ON FUS AINIZED MATERIAL by K. L. ALVIN Abstract. Charred leaf fragments preserved in siltstone show excellent internal structure by SEM. The anatomy presents a high degree of xeromorphism. A number of features, including sunken stomata with paired subsidiary cells, specialized cells of digitate form underlying the upper epidermis and abundant sclereids in the mesophyll, are probably unique among ferns. The significance of the anatomy is discussed in relation to the possible habitat. The common and widespread Lower Cretaceous fern Weichsella reticulata (Stokes and Webb) Fontaine possesses some unusual morphological and structural characters (Alvin 1968, 1971). Some of these, such as the small, thick pinnules the cuticle of which in well-preserved material can be prepared by oxidative maceration and the sporangia tightly packed beneath thick, sclerotic indusia crowded together into cone- like soral clusters, could be regarded as xeromorphic. Whether the plant was a xero- phyte, however, is uncertain, since xeromorphic features are commonly found among plants of wet, especially haline, soils. Gothan ( 1 923), on the basis of what he supposed to be autochthonous specimens preserved in sandstone at Quedlinburg, believed the plant to be an inhabitant of arid dunes. Daber (1968), on the other hand, working on material from the same deposits, believed it grew under conditions of high insola- tion along river banks near the sea. Its distribution, confined as it apparently is, within palaeolatitudes approximately 30° N. and 30° S. (Barnard 1973), indicates that it was tropical or subtropical Although abundant lignitized material from Belgium (Alvin 1971) and occasional petrified major axes from elsewhere, especially North Africa (e.g. Koeniguer 1966) have provided a general picture of the anatomy of the stem, petiole, and rachises, little is known of the internal structure of the vegetative pinnules. The cuticle has revealed the general structure of the epidermis, although there has been considerable variation in the interpretation of the stomata. Florin (1919) and Sukh Dev (1970) described the stoma as being sunken between a pair of subsidiary cells, whereas Reymanowna (1965) and Alvin (1971) failed to see evidence in their cuticle preparations of sunken guard cells and interpreted the pair of superficial cells as guard cells. Although Florin (1919) mentioned having prepared microtomed sections of leaves from Belgium, he did not illustrate the internal anatomy or describe it in any detail. Alvin (1971) failed to obtain worthwhile sections of Belgian leaves. MATERIAL AND METHODS In the English Wealden, Weichselia is commonly preserved as fragments of fusain (charcoal) in a sandy, often current-bedded matrix. Such material, which forms the basis of this study, was collected from a lens of siltstone at Shepherd’s Chine, half a mile north-west of Atherfield Point, Isle of Wight. [Palaeontology, Vol. 17, Part 3, 1974, pp. 587-598, pis. 87-89.] 588 PALAEONTOLOGY, VOLUME 17 The matrix is a fairly uniform, pale grey, coarse siltstone (grain size c. 0 03 mm), somewhat unevenly bedded, with occasional small pockets of finer sediment. It con- tains numerous fragments of Weichselia representing pinnules, portions of pinnae with several pinnules attached, naked rachises and larger, usually striated pieces of fusain which may represent fragments of the more massive organs of the plant (stems, petioles, or main pinna rachises). Plate 87, fig. 1 shows a typical sample of material. Nearly all of the plant fragments appear to represent Weichselia. The only other plants occasionally present are Phlebopteris dunkeri (Schenk) Schenk (sterile and fertile pinnule fragments) and ICladophlebis sp. (sterile pinnules). Bulk maceration of a small sample of matrix yielded one fragment probably representing Ruffordia goepperti (Dunker) Seward. All but very few of the plant fragments present are charred; they are extremely brittle and smear the fingers at a light touch. Occasionally present (estimated at less than 1% of the total) are non-charred or only partly charred pieces. The occurrence of such partially charred fragments provides evidence that the fusain represents true charcoal and has not been produced by a diagenetic process. Some of the non- charred (lignitized) Weichselia pinnules have been softened in alcoholic KOH solution, embedded, and sectioned. They are poorly preserved (PI. 87, fig. 3), but what structure they show assists interpretation of the scanning electron microscope pictures of fusainized material. Preparation of fusainized material for SEM is relatively simple. Since most of the pinnules and pinna fragments lie with their lamina in the bedding plane, cleavage of the matrix ruptures them in the plane of the lamina. Hardly ever does the pinnule separate them from the matrix at the surface. Freshly fractured pinnules may be mounted and viewed with no further treatment other than routine metal coating. Since fracture of pinnules usually occurs in the same plane, namely through the middle of the mesophyll and vein system, other levels of section parallel to the lamina surface (i.e. nearer to the upper or lower epidermis) have effectively been obtained by gently brushing away the charcoal with a fine camel-hair brush. Such brushed specimens are improved by immersing them subsequently in a beaker of distilled water in which an ultrasonic probe is inserted for about 10 seconds. Cross-sections of pinnules are obtained by breaking samples of matrix across the bedding plane or by grinding in this plane on fine sandpaper. Transverse sections of pinnules may be recognized by their shape (PI. 87, fig. 2). Small slabs of matrix EXPLANATION OF PLATE 87 (Figs. 4-7 are scanning electron photomicrographs.) Figs. 1 -7. Weichselia. 1 , a typical sample of material showing fragments of pinnae and pinnules with, near the top left, a piece of a fertile pinnule of Phlebopteris dunkeri , x 2. 2, a block ground at right-angles to the bedding plane showing cross-sectioned pinnules, x 4. 3, light microscope photograph of part of a microtomed section of a lignitized pinnule, x 75. 4, part of a cross-section of a pinnule. A vein is seen slightly right of centre and another near the left-hand edge, x 1 76. 5, part of a pinnule fractured parallel to the surface through the spongy mesophyll showing the rather broad vein sheaths, x 168. 6, part of a pinnule fractured parallel to the surface and brushed towards the upper side showing the regular, palisade-like cells and the vein sheaths, x 164. 7, part of a vein with tracheids and associated ruptured cells, x 1600. PLATE 87 fiEssto vS ivSSwjJi Hv9jS 2 6 ALVIN, Weichselia 590 PALAEONTOLOGY, VOLUME 17 on which such sections are exposed are stuck to the microscope stub and washed by brief ultrasonic treatment as before. The outer surface of the pinnule has been observed from specimens freed from the matrix in hydrofluoric acid. All specimens for SEM observation are cemented to the stub with ‘Durofix’ and coated with gold-palladium. RESULTS Interpretation of scanning electron micrographs of charcoal surfaces presents certain difficulties. The image obtained, which is very different from that of a sectioned tissue viewed in the light microscope, represents a view of the surface of the often very irregularly fractured cells and tissues. Some parts of this surface will represent the inside of cells, some the outside, either where natural intercellular spaces occur or perhaps where abutting cells have separated along the plane of the middle lamella, and some the fractured cell walls. Further difficulties may arise from various dis- tortions likely to have been produced (a) by charring, ( b ) by compression, and (c) by precipitation of minerals such as iron pyrites in the tissues. Harris (1957) has made some reference to the distortion in plant tissues produced by charring: the distortion in wood is unpredictable, but frequently the rays become distended into gaping holes. There is no evidence of gross distortion of this kind in the Weichselia leaves, but some of the cell walls appear to have been structurally altered. Compression seems to have resulted in a variable degree of shattering and compaction of the tissues, chiefly in the region of the spongy mesophyll. Precipitation of iron pyrites in the form of small spherical masses is frequent, especially again in the spongy mesophyll, and in some specimens this is accompanied by gross disruption of the tissues. Morphology and surface features In shape and size the pinnules show a similar range of variation to that shown by material from Belgium preserved as lignite (Alvin 1971). The upper surface is smooth and generally convex, the convexity increasing towards the thick, somewhat rolled-over margin. The midrib is sometimes marked by a shallow furrow, especially towards the base of the pinnule. The lower surface is more or less scrobiculate, with the veins represented by low ridges and the inter-vein areas by shallow depressions. A remarkably conspicuous feature of the lower surface is the large, crowded, and somewhat bulging stomatal subsidiary cells (PI. 89, fig. 1). As much as about 40% of the area of the lower surface is occupied by these cells which are 50-85 ^m in length. The cells of each pair are either in contact along their length, or, more fre- quently, separated by a narrow, slit-shaped space. At first it was thought that these cells were the guard cells, but views of the inside of the epidermis have made it clear that the true guard cells are considerably smaller and sunken below the leaf surface. The subsidiary cell pairs are oriented at random and almost uniformly scattered so that the veins are not clearly marked by their distribution. The ordinary epidermal cells are not distinguishable in the surface relief, but occasionally there occur small, round, somewhat depressed areas around which subsidiary cell pairs often tend to ALVIN: LEAF ANATOMY OF WE1CHSELIA 591 radiate (PI. 89, fig. 1, above and slightly right of centre). These areas are believed to be equivalent to the ‘papillae’ mentioned by Reymanowna (1965) and the ‘small round cells’ noted in the cuticle by Alvin (1971). Diminutive subsidiary cell pairs about half the size of normal ones also occur (PI. 89, fig. 1, bottom right-hand corner). Internal structure The leaflet fractured parallel to the surface always shows a conspicuous vein network (PI. 87, figs. 5, 6). The veins consist chiefly of thick-walled cells elongated in the direction of the vein. This fibre-like vein sheath tissue is seen in cross fracture (PI. 87, fig. 4) to extend above and below the vein to both epidermises. In sections of lignitized leaves (PI. 87, fig. 3), it is only seen as blocks of tissue beneath the epidermis, perhaps because here it was thicker walled and therefore more durable than around the veins nearer the middle of the leaf. In fractures parallel to the surface the tissue always appears broader in the spongy mesophyll region and to consist of narrower cells (PI. 87, fig. 5) than nearer the upper epidermis where there are typically only one or two rows of broader elements (PI. 87, fig. 6). In the spongy mesophyll region there is some gradation between the fibrous sheath tissue and the mesophyll. Sheath cells often show conspicuous pit-like markings under high magnification (PI. 88, fig. 7); it is possible that some of the sheath cells functioned as transfusion elements. The vascular bundle lies within the sheath where this is broadest, at or just below the middle of the leaf. It is indicated by a small core of usually shattered, thin- walled cells in which can sometimes be seen tracheids (PI. 87, fig. 7). The tracheids are scalariform, scalariform-reticulate, or sometimes annular. Very beautifully preserved scalariform tracheids have been seen in the midvein of pinnules and in pinna rachises (PI. 89, fig. 6). In some of the thin-walled elements associated with the tracheids in these stronger vascular strands, a wall sculpturing resembling sieve areas can some- times be seen (PI. 89, fig. 6, top left-hand corner). The mesophyll is clearly differentiated into two regions. Underlying the upper epidermis is a palisade-like region, well seen in cross fractures (PI. 87, fig. 4). Below this, extending to the lower epidermis and occupying rather more than half the thick- ness of the leaf is the spongy mesophyll. There is no very clear demarcation line between the two regions; they intergrade by one or two rows of rather compactly arranged rectangular cells. A section parallel to the surface through the palisade-like region (PI. 87, fig. 6; PI. 88, fig. 3) shows rather regular, cylindrical elements in cross-section with triangular or polygonal intercellular spaces. Nearer the upper epidermis (PI. 88, fig. 2, right- hand side) the intercellular spaces become smaller and eventually the cylindrical elements unite in groups of four to six to form large, lobed cells somewhat resembling epidermal cells with sinuous anticlinal walls (PI. 88, fig. 2). At first it was thought that this level did indeed represent the epidermis; however, there is an absence of any periclinal wall separating the lobed cells from their cylindrical, palisade-like exten- sions. Further, the connections between adjacent palisade-like elements can also be seen in cross fractures (PI. 88, fig. 1, top right). The true epidermis, consisting of large cells which are conspicuous in cross fractures, can also be seen in Plate 88, fig. 4, where, at the top of the photograph, the sub-epidermal layer has been completely removed by the brushing treatment. Hence, it emerges that the sub-epidermal layer 592 PALAEONTOLOGY, VOLUME 17 consists of large digitate cells each with four to six digits arising from the lobed head and oriented towards the interior of the leaf. The possible function of these cells is discussed later. The spongy mesophyll consists mainly of irregular, thin-walled cells with small pits (PI. 89, fig. 2). Also present, often abundantly, scattered in the mesophyll are sclereids or stone-cells. These are often conspicuous in sections of lignitic leaves (PI. 88, fig. 5), but only seldom are they seen in the SEM, probably because they do not easily fracture and cannot be recognized from a surface view of the cell. Two probable stone-cells are seen at the bottom of Plate 88, fig. 1, and one of these is shown enlarged in Plate 88, fig. 6. The frothy appearance of the wall may have been produced by charring. Specimens cleaved in the plane of the lamina and brushed towards the lower epidermis show spongy mesophyll between the veins and, if brushed sufficiently, the stomata and lower epidermis from the inside. Plate 89, fig. 2, shows the pair of intact guard cells of a stoma situated in a space in the spongy mesophyll presumably representing the sub-stomatal air chamber. In more vigorously brushed specimens the spongy mesophyll is often removed completely and the guard cells broken open or removed. Plate 89, fig. 3, shows a stoma in which only the outer periclinal walls of the guard cells are present; surrounding these can be seen the damaged walls of the subsidiary cells. At the centre, two pairs of lips separating narrow, slit-like open- ings are visible: the upper (inner), wider opening (c. 3 mm wide in the photograph) represents the stomatal aperture itself (i.e. between the lips of the guard cells); the lower (outer) and narrower (c. 1 mm wide in the photograph) opening represents the space between the subsidiary cells. In Plate 89, fig. 4, the specimen has been brushed so as to remove not only the spongy mesophyll and guard cells, but also the inner and, in some cases, the outer walls of the subsidiary cells; where the outer walls have been removed, the matrix can be seen. Unfortunately, no satisfactory cross-sections of stomata have been obtained, but this interpretation of the stomatal organization is in general agreement with that of Florin (1919) and Sukh Dev (1970) both of whom based their results mainly on cuticle preparations. However, both of these authors described the stomatal pit as wide. In the cuticle shown in Plate 89, fig. 5, prepared from a lignitic pinnule, the large, somewhat granular cells undoubtedly represent the subsidiary cells. If the small. EXPLANATION OF. PLATE 88 (All except fig. 5 are scanning electron photomicrographs.) Figs. 1 -7. Weichselia. 1 , detail from a section similar to that in Plate 87, fig. 4 showing the upper epidermis, the palisade-like sub-epidermal layer, part of the spongy mesophyll with two probable stone-cells; (the one near the bottom right is shown enlarged in fig. 6), x 480. 2, part of a specimen prepared in the same way as that in Plate 87, fig. 6 showing, on the right, palisade-like digits of the sub-epidermal layer and, at the centre and left, the lobed end portions of these cells, x 920. 3, detail from a similar section to that shown in Plate 87, fig. 6, x 896. 4, a low-power view of the same specimen as in fig. 2 showing three levels of section parallel to the surface : at the bottom, digits of the sub-epidermal layer in cross- section; left of centre, lobed heads of these cells (as in fig. 2); top, the epidermis. Note that in the sub- epidermal layer, fibres are seen marking the course of veins, x 224. 5, light microscope photograph from a sectioned lignitized leaf showing stone-cells in the mesophyll region, x 380. 6, probable fractured stone-cell, x 1600. 7, detail of vein sheath cells showing pit-like structures, x 1840. PLATE 88 ALVIN, Weichselia 594 PALAEONTOLOGY, VOLUME 17 clearer areas between the subsidiary cell pairs were interpreted as guard cells, this would suggest a wide stomatal pit. If, however, the clear areas represent rather the cutinized anticlinal walls of the mouth of the stomatal pit, the guard cells not being represented due to insufficient cutinization, this would accord with the SEM view. Narrowing of the stomatal pit has not been produced by charring, for the external appearance in the SEM of the stoma of a lignitic leaf is the same as that of a fusainized specimen. DISCUSSION Weichselia presents a leaf anatomy unique among known ferns. The reproductive structures as well as certain aspects of morphology and vascular anatomy suggest an affinity with Matoniaceae (Alvin 1971). The pinnule anatomy of the living Matonia pectinata R. Br. has been examined, but shows little resemblance to that of Weichselia : it shares only one character, namely, the presence of vein sheath tissue extending to the epidermis. Especially noteworthy are the stomata and the cells underlying the upper epidermis. The stoma, with its pair of specialized subsidiary cells has, according to van Cotthem’s (1970) extensive studies, no parallel among ferns. It is, however, remarkably similar to that of Equisetum , where, as in Weichselia , the subsidiary cells form a narrow pit above the sunken guard cells (Hauke 1957; Page 1972). There is also a general similarity to the Bennettitalean stoma and that of the living Welwitschia among gymnosperms. The digitate sub-epidermal cells suggest another highly specialized feature. The presence of abundant intercellular spaces between the digits and the absence of noticeably thicker walls than in the spongy mesophyll cells point perhaps to a photo- synthetic function. On the other hand, the somewhat prominent pitting might suggest a water-storage function. A sub-epidermal water-storage layer has been reported in the xerophytic fern Pyrrosia lingua (Thunb.) Farwell by Hungerbiihler (1957). The upper epidermis in Weichselia is itself quite large-celled as seen in cross fractures (PI. 88, fig. 1). One might speculate that this had a water-storage function. Apart from the sunken stomata and the possibility of the presence of a water- storage system, some of the other characters may also be regarded as xeromorphic, notably: (a) the thickness of the lamina (up to about 0-5 mm); ( b ) the presence of EXPLANATION OF PLATE 89 (All except fig. 5 are scanning electron photomicrographs.) Figs. 1-6. Weichselia. 1, part of the lower surface of the pinnule, x 160. 2, specimen fractured parallel to the surface and brushed towards the lower side showing spongy mesophyll cells and a stoma from the inside, x 420. 3, specimen brushed to lower epidermis showing a stoma the guard cells of which are ruptured to expose the outer periclinal walls; surrounding them are seen the somewhat broken walls of the subsidiary cells, x 840. 4, specimen brushed still lower showing the epidermal cells and stomatal subsidiary cells some of which have had their outer walls removed thus exposing the matrix, x 464. 5, light microscope photograph of the lower cuticle prepared from a lignitized leaf, x 380. 6, vascular tissue in a pinna rachis, x 1600. PLATE 89 ALVIN, Weichselia 596 PALAEONTOLOGY, VOLUME 17 text-fig. 1. Proposed reconstruction of the pinnule lamina as seen in transverse section. a cuticle resistant to oxidative maceration; (c) abundant fibrous tissue round the veins extending to the upper and lower epidermis; ( d ) the presence of sclereids in the mesophyll. Xeromorphic features do not occur commonly among ferns although, as Hunger- buhler (1957) has shown, certain characters comparable with some generally regarded as xeromorphic among angiosperms do occur. Thus, Ceterach officinarum DC., Cheilanthes gracillima Eaton, C. marantae (L.) Domin, and Doryopteris ornithopus (Mett.) J. Sm. have thick leaves with a double palisade. Pellaea ternifolia (Cav.) Link and Anemia millefo/ia Gardn. ex Pressl have fibrous tissue above and below the main veins. A hypodermis is reported in Elaphoglossum latifolium (Swartz) J. Sm. Pyrrosia confluens (R. Br.) Ching has sunken stomata and in a number of ferns the stomata are protected by inrolling of the margin (e.g. Cheilanthes spp. and Pellaea spp.) or by the presence of persistent scales or hairs on the lower surface (e.g. Cetarach and ALVIN: LEAF ANATOMY OF WE1CHSELIA 597 Cheilanthes spp.). Most of these ferns grow in habitats subject to periods of drought, and the xeromorphic features may be assumed to be physiologically advantageous under these conditions. Among angiosperms, xeromorphic characters may occur in plants of wet soils, and as a wet maritime habitat has been suggested for Weichselia by Daber (1968), the leaf anatomy of Acrostichum danaeifolium Langsd. et Fisch. and Blechnum indicum Burm. has been examined. Both of these ferns grow in tropical brackish water swamps often under conditions of high insolation, but neither show any notable xeromorphic characters. The combination in Weichselia of a number of xeromorphic features, some developed to an extreme degree (e.g. the cuticle, fibrous tissues, and sunken stomata) suggests strongly that the plant grew in a habitat subject to periods of extreme drought. Only comparatively rarely do other plants occur in close association with the fossil remains of Weichselia ; certain Matoniaceae are probably the most frequent associates, but are seldom present in any quantity. It may therefore be conjectured that Weichselia dominated the community in which it grew. This is consistent with the idea that it was adapted to life in an extreme environment to which few plants are likely to have been suited. Daber’s (1968) conclusion that Weichselia grew along river banks close to the sea was based on his presumed in situ specimens found just above dark humic layers in sandstones believed to have been laid down under fluvial conditions near the coast. The presumed rooted specimens, however, are almost certainly only the heads of large petioles (Alvin 1971), and therefore not indicative of in situ preservation. On the other hand, the Quedlinburg specimens of fronds and probable stems (= Stiehleria Daber) are impressively large, and it therefore seems probable that the remains had not been transported far. Thus they may have been preserved as the rivers, perhaps in times of flood, cut through stands of Weichselia growing in otherwise arid con- ditions. Large specimens of Weichselia are only comparatively rarely encountered, and the Quedlinburg conditions of preservation may therefore have been exceptional. Much more commonly the plant is preserved as small fragments, often in the charred condition. This hardly suggests that it was a plant of river banks, but rather that it grew in areas somewhat remote from sites of deposition, and was only brought down at times of flood. The frequency with which remains occur charred is consistent with the idea that the community was at times subjected to dessication when it would have been especially vulnerable to the spread of fire. Concerning proximity to the sea, Weichselia remains are nearly always found in freshwater deposits. Only rarely are fragments found in marine sediments (e.g. sand- stones in the Aptian of England). This does not suggest that the plant was especially associated with maritime conditions. Acknowledgements. I am indebted to Miss J. Fillery for her skilful assistance with the scanning electron microscopy and to Mr. D. Bebbington for photographic help. I also thank Mr. A. C. Jermy and Mr. J. A. Crabbe for helpful information about Recent ferns and for the provision of anatomical material. The Cambridge ‘Stereoscan’ microscope was obtained by Imperial College Botany Department on a grant from the Science Research Council. 598 PALAEONTOLOGY, VOLUME 17 REFERENCES alvin, K. l. 1968. The spore-bearing organs of the Cretaceous fern Weichselia Stiehler. J. Linn. Soc. (Bot.) 61, 87-92. — 1971. Weichselia reticulata (Stokes et Webb) Fontaine from the Wealden of Belgium. Mem. Inst. Roy. Sci. Nat. Belg. no. 166. Barnard, p. d. w. 1973. Mesozoic floras. In hughes, n. f. (ed.). Organisms and continents through time. Spec. Pap. Palaeont. 12, 175-187. cotthem, w. van. 1970. Comparative morphological study of the stomata in the Filicopsida. Bull. Jard. Bot. nat. Belg. 40, 81-239. daber, R. 1968. A Weichselia-Stiehleria-Matoniaceae community within the Quedlinburgh estuary of Lower Cretaceous age. J. Linn. Soc. (Bot.) 61, 75-85. florin, r. 1919. Zur Kenntnis der Weichselia reticulata (Stokes et Webb) Ward. Svensk. Bot. Tidskr. 13, 305-312. gothan, w. 1923. Ein vollstandiges Exemplar von Weichselia reticulata im Neocomsandstem von Quedlin- burg. Jb. preuss geol Landesanst. Berg-Akad. 42 (2), 772-777. Harris, t. m. 1957. A Liasso-Rhaetic flora in South Wales. Proc. Roy. Soc. B147, 289-308. hauke, r. c. 1957. The stomatal apparatus of Equisetum. Bull. Torrey Bot. Club , 84, 178-181. hungerbuhler, R. 1957. Die Xeromorphosen der Fame mit besonderer Berucksichtigung der Blattanatomie. Brunner Bodmer, Zurich. koeniguer, j.-c. 1966. Etude paleophytogeographique du continental intercalaire de l’Afrique Nord- Equatoriale. B. Sur de nouveaux echantillons du genre Paradoxopteris. Mem. Soc. geol. France, N.s. 105, 100-112. page, c. N. 1972. An assessment of interspecific relationships in Equisetum subgenus Equisetum. New Pliyt. 71, 355-369. reymanowna, m. 1965. On Weichselia reticulata and Frenelopsis hoheneggeri from the western Carpathians. Acta Palaeobot. 6 (2), 15-26. sukh dev. 1970. Some ferns from the Lower Cretaceous of Madhya Pradesh- 1. Palaeobotanist, 18, 197-206. K. L. ALVIN Department of Botany and Plant Technology Imperial College of Science and Technology London, S.W. 7 Revised typescript received 1 November 1973 OSTRACODS FROM THE DOMERI AN AND TOARCI AN OF ENGLAND by ALAN LORD Abstract. Marine ostracods from the Middle and Upper Lias are described and their lateral and vertical distribution along the strike from the Dorset to Yorkshire coasts analysed. The composition of the assemblages and the affinities and occurrence of the species are examined, and two new species are described. Ostracod diversity apparently decreased throughout the Margaritatus Zone and few ostracods have yet been found in the Spinatum Zone. However, the Toarcian transgression brought in a new ostracod fauna of Middle Jurassic aspect. The Lower Jurassic, particularly in the British Isles, is one of the most neglected periods of Mesozoic and Cenozoic time in the field of stratigraphical palaeontology of the Ostracoda. In this country only eight publications have dealt with Liassic ostracods. That pioneer student of fossil ostracods, T. R. Jones, described Lias species from Y orkshire ( 1 872) and south-west England ( 1 894), and further Y orkshire material was described by Blake (1876). In the past decade, Anderson (1964) has described Rhaetic and Lower Jurassic species particularly from the Bristol and south Wales region and has revised Jones’s (1894) work, and Field (1966) has described Hettangian Cytherellidae from the Dorset coast. The present writer has attempted to complete the revision of the older work by re-examining the Yorkshire sections from which Jones (1872) and Blake (1876) obtained their material (Lord 1971a), and has begun an analysis of Domerian and Toarcian faunas (Lord 1971 A 1972a). Also relevant to Lias ostracods are the large number of Triassic faunas described from Europe, the Soviet Union, and elsewhere in the last ten years (see review by Sohn 1968), and numerous publications concerning the Middle Jurassic of which those by R. H. Bate are of particular interest in the British context. This paper describes the lateral and vertical distribution of Domerian and Toarcian (essentially Middle and Upper Lias of British geologists) ostracods along the strike of the Lower Jurassic outcrop from the Dorset to Yorkshire coasts. The first part is an account of the sampled sections and the distribution of ostracods in them, the second discusses certain species and their affinities, while the final part reviews the material generally within the framework of the north-west European epicontinental seas. The picture obtained is necessarily incomplete because of lack of exposures, natural faunal poverty, and barren samples, particularly in the Midlands. The sample pattern is also far from ideal since what is an areal problem was examined in a linear fashion. Nevertheless, lateral control along the line of strike is good although the answer to many questions probably lies to the east, where any data from subsurface or from submarine outcrops are of especial value. Also important is the work of Curry et al. ( 1 970) in the English Channel and Dingle ( 1 97 1 ) in the North Sea. To the west of the main outcrop, the Institute of Geological Sciences borehole at Mochras, north Wales contains a remarkable thickness of Lower Jurassic sediments (Wood and Woodland 1 968) from which the ostracod faunas will prove of great value. [Palaeontology, Vol. 17, Part 3, 1974, pp. 599-622, pi. 90.] 600 PALAEONTOLOGY, VOLUME 17 OSTRACOD FAUNAS OF THE SAMPLED SECTIONS The samples were located as accurately as possible both stratigraphically and regionally, for without this precision any study of the evolution of the faunas or variation within taxa is without value. It is fortunate, therefore, that the Lower Jurassic possesses one of the more sophisticated zonal schemes and there are few sections which cannot be readily zoned with reasonable accuracy. The zonal system of Dean, Donovan and Howarth (1961, emended Howarth, 1964) for the north-west European ammonite province was employed since the present work was regional in concept. 'Domerian’ was used because it was a useful division, despite the recom- mendation of Arkell (1956, p. 8) that the term should be abandoned. The Domerian represents the life span of the amaltheid ammonites (the zones of Amaltheus mar- garitatus and Pleuroceras spinatum) and corresponds with Ober Pliensbachian or Lias delta of Hoffmann (1962) and Domerien of Mouterde (1953); it almost corre- sponds to the English usage of 'Middle Lias’. The outcrop distribution of the Lower Jurassic in England is shown in text-fig. 1, together with the location of the sampled sections. Dorset. In Dorset, beds of Domerian and Toarcian age are well exposed along the coast near Bridport. The Domerian has been described in some detail by Howarth (1957) and both Domerian and Toarcian by Wilson et al. (1958). The Domerian and Toarcian beds of the Dorset coast do not always lend them- selves to micropalaeontological investigation. Six zones and one subzone, i.e. the upper zone of the Domerian and most of the Toarcian zones, are condensed and represented by a thin bed, up to 7 feet (2 m) thick, of limestones called the Junction Bed. This thin bed is represented by over 450 feet (137 m) of sediment in Yorkshire. Jackson’s (1926) small pocket of clay (Tenuicostatum Zone) on the western side of Thorncombe Beacon preserved in a hollow in the Marlstone Rock Bed (the lower, Domerian part of the Junction Bed) could not be located. Not only is a substantial portion of the succession represented by a condensed series of hard limestones, but the top part of the Toarcian consists of the Bridport Sands, sands and sandstones in vertical cliffs both difficult to sample and barren. However, ostracods occurred throughout the lower zone of the Domerian (Margaritatus Zone) which reaches its maximum thickness at outcrop in this country of almost 4 1 0 feet ( 1 25 m) on the Dorset coast. Ostracods were also obtained from the middle unit of the Toarcian sequence, the Down Cliff Clay. The Domerian part of the Junction Bed, called the Marlstone Rock Bed after that development of the Spinatum Zone in the Midlands, has a non- sequence at the base which can be responsible for the loss of up to half the Apyrenum Subzone (Howarth 1957, p. 192). A non-sequence at the same horizon is known elsewhere and is referred to below. The Dorset coast succession yielded the most species and individuals found in this investigation and their distributions are shown in text-fig. 2. Ogmoconcha dominated the Domerian faunas numerically but Wicherella semiora semiora Lord, 1972 occurred commonly in the Stokesi Subzone and Gramannella apostolescui Gramann, 1962 was common in both Stokesi and Subnodosus Subzones. The species Ogmo- concha sp. E was a consistent faunal element in the upper part of the Stokesi Subzone, i.e. top Eype Clay and lower half of Down Cliff Sands. The latter, however, is very LORD: DOMERIAN AND TOARCIAN OSTRACODS 601 East Yorkshire Roxby' .Lincoln Mochras Borehole Napton-on-the-Hilf Byfield Robins Wood Gloucester Aston Magna Kirton- in -Lindsey Lower Jurassic sediments miles O 20 40 60 80 100 120 kilometres text-fig. 1. Lower Jurassic outcrop showing sampled sites. argillaceous and the disappearance of this species may be facies controlled. Except for Ogmoconcha contractula Triebel, 1941 and Bairdia molesta Apostolescu, 1959, which occur in both the Subnodosus and Gibbosus Subzones, no less than seven species were restricted to the Subnodosus Subzone. These species have little zonal value ; for example Polycope cerasia Blake, 1 876 ranges from Hettangian to Domerian, and Trachycythere tubulosa tubulosa Triebel and Klingler, 1959 is known from many parts of the Pliensbachian. Similarly, three of the six species found in the Gibbosus Subzone were restricted to it on the Dorset coast with Ogmoconcha sp. C by far the dominant element; this species closely resembles ostracods from the Gibbosus and L 602 PALAEONTOLOGY, VOLUME 17 text-fig. 2. Distribution of ostracod species on the Dorset coast. LORD: DOMERIAN AND TOARCIAN OSTRACODS 603 Apyrenum Subzones of the Yorkshire coast which are tentatively regarded as the same species and may again be an example of facies control. The Toarcian of the Dorset coast yielded ostracods from only the lower Down Cliff Clay, since the upper part and much of the Bridport Sands are virtually in- accessible in vertical or near vertical cliffs. The lower part of the Down Cliff Clay from the Junction Bed up was sampled in sections found on Thorncombe Beacon and Watton Cliff, and the species at the two places compared (text-fig. 2). Ostracods were not very numerous; but out of a total of ten species from Thorncombe Beacon and eight from Watton Cliff the following species: Cythere/la toarcensis Bizon, 1960, Cytherelloidea cadomensis Bizon, 1960, Kinkelinella sermoisensis Apostolescu, 1959, Cytheropteron alafastigatum Fischer, 1962, and Ektyphocythere aff. champeauae Bizon, 1960, were common to both. Since the two sets of samples are from the same subzone, same type of sediment, stratigraphically similar levels, and the localities are only 500 metres apart the faunal differences seem explicable only in terms of collection failure. Cotswolds. Northwards along the strike of the Lower Jurassic from the Dorset coast, exposures are limited to small parts of the Domerian and Toarcian. In a primary survey full or relatively large sections with good zonal control are desirable, and so small sections were omitted. The whole of the Domerian is next exposed near Gloucester, north of the Mendip Axis, where the disused brickpit of Robins Wood Hill (SO 836149) was examined. The stratigraphical palaeontology has been recently described by Palmer (1971). Only three out of twenty-four samples contained Ostra- coda. Whilst probably due to the very weathered nature of the face, natural faunal paucity cannot be excluded. The fossiliferous samples were from the upper part of the Davoei Zone ( Wicherella semiora semiora Lord, 1972 and Ogmoconcha sp. B) and from the Stokesi Subzone ( W. semiora semiora , O. sp. B, Gramannella apostolescui (Gramann 1962), Trachvcythere tubulosa Iseratina Triebel and Klingler, 1959 and two new cytherurid species). The exact influence of the Mendip Island on faunal movement is impossible to assess. Although certainly variable it was not apparently of major importance. The Gloucester evidence shows a strong connection with the Dorset Domerian fauna even to the extent of the same subspecies of Wicherella semiora being present. Howarth (1958, p. xxxvii), using evidence from Kent (1949), concluded that a few miles east of the Mendips there was free north-south access, and the evidence here supports that view. An erosion surface occurs at the base of the Spinatum Zone in the Gloucester area and as a result the Gibbosus Subzone is missing (Palmer 1971). Erosional features are known at about this level from a number of sites between East Yorkshire and Dorset and are cited below. The Toarcian in the Gloucester area was not sampled because good exposures were lacking. In the south and mid-Cotswolds the Toarcian consists of Cotswold Sands overlain by the condensed Cephalopod Bed, but northwards the facies changes to a full but poorly exposed sequence of clays. These clays could not be sampled and subsurface material is clearly necessary. In the north Cotswolds at the village of Aston Magna (SP 198356) a brick pit exposes the uppermost beds of the Lower Lias (blue-grey clays of the Davoei Zone) and the lowest beds of the Middle Lias (Domerian) which consists of poorly 604 PALAEONTOLOGY, VOLUME 17 fossiliferous red-brown sands (McKerrow and Baden-Powell 1953, p. 89). Further up the slope in the village ‘ironstone’ has been recorded (Arkell 1947, p. 17). This is the Marlstone Rock Bed, now much more ferruginous so that further north at Banbury it is actively quarried. At Aston Magna the junction between the clays and the sands was obscured so that the thicknesses present could not be determined. Although the two samples from the ferruginous Domerian sands were unfossiliferous because of decalcification or very shallow water conditions, or both, samples from the Davoei Zone clays did contain ostracods. The fauna consisted of Ogmoconcha species, Pseudohealdia , and Ostracode (513) Wicher, 1938. The latter is of particular interest having been recorded from Germany (Wicher 1938; Gramann 1962 b), from sub- marine samples from the floor of the North Sea (Dingle 1967), and the writer has found it in the Lower Pliensbachian of Lincolnshire (but not Kirton-in-Lindsey, see below), the Midlands, and the Cotswolds. The species is thought to be related in part to Wicherella. Midlands. North of Banbury and the Marlstone Rock Bed ironstone workings there is an exposure of Lower Toarcian beds in a railway cutting at Byfield (SP 512529) described by Walford (1879) and a section recorded by Beeby Thompson given by Woodward (1893, p. 275). The microfauna is rich and the foraminifera have been described by Barnard (1950). The section is largely overgrown and it was necessary to excavate the face. The zones have not been accurately delimited, but all the Lower Toarcian appears to be present. Dr. M. K. Howarth has kindly commented upon the zonal distribution of the samples. The Falciferum Subzone (Falciferum Zone) contained Kinkelinella sermoisensis (Apostolescu 1959) and Gen. indet. Ibucki Bizon, 1960; the samples from the Commune Subzone (Bifrons Zone) contained the same two species together with Cytherella toarcensis Bizon, 1960, Kinkelinella sp. I (Aposto- lescu, 1959), Trachycythere tubulosa tubulosa Triebel and Klingler, 1959, and T. verrucosa Triebel and Klingler, 1959. Only two species are common between Byfield and the two sections on the Dorset coast, viz. C. toarcensis and K. sermoisensis , but since a marked zonal difference exists this is only to be expected. Of particular interest was the occurrence of Kinkelinella sp. I and an abundance of specimens closely resembling the species originally named Procytheridea bucki Bizon, 1960 which may indicate a close faunal connection with the Paris Basin. An almost complete sequence of Domerian strata is exposed in a brick pit at Napton-on-the-Hill (SP 456613), like Robins Wood Hill an outlier separated from the main outcrop, but capped by the Marlstone Rock Bed and with no Toarcian present. The pit is at the top of the hill on the north-western side and exposes 72 feet (22 m) of Domerian sediments. The only record is the measured section given by Howarth (1958, p. xi) who proved the presence of the Stokesi and Subnodosus Sub- zones but not the Gibbosus Subzone. The latter may be represented by the clay bed (Bed 5 of Howarth) immediately below the Marlstone Rock Bed or it may have been removed by pre-Marlstone Rock Bed erosion, for which there is evidence in other sections. The base of the Margaritatus Zone was not seen, the lower part of the section differing from that of Howarth and no liparoceratid ammonites found to prove pre- Domerian strata. Of seventeen samples examined only one, from a shell band, con- tained ostracods and foraminifera (?Stokesi Subzone, 42 feet (13 m) below Marlstone LORD: DOMERI AN AND TOARCI AN OSTRACODS 605 Rock Bed). Percolating groundwater probably explains the paucity of microfossils but it is possible that ecological and geographical factors also played a part. The few ostracods found were assigned to Ogmoconcha sp. B and Ogmoconcha sp. C. Lack of microfossils at this locality and at Robins Wood Hill was critical, since these are the two most complete and best exposed inland sections south of Lincolnshire, and left little to connect the faunas of Lincolnshire with those of Dorset. As a result it was impossible to trace faunal movements or possible provincialism. The ironstone quarries in the Marlstone Rock Bed around Banbury sometimes expose lowest Toarcian clays at the top, but these proved barren, as did lowest Toarcian paper shales from Tilton-on-the-Hill to the north in the Leicestershire part of the ironstone field. The studied section at Tilton Station cutting (SK 762055) has been described by Hallam (1955). Here the Marlstone Rock Bed is composed of 94 feet (3 m) of calcareous and ferruginous sandstones overlain by 8+ feet (2-5 m) of ironstone, which is a very ferruginous oolitic limestone. A sample from the sand- rock (Bed 5 of Hallam) yielded ostracods. These specimens, together with a few from Staithes, were the only ostracods found in the Spinatum Zone. For the most part they were heavily encrusted with quartz grains and many specimens were unidentifiable. The species Tr achy cy there tubulosa tubu/osa and T. verrucosa Triebel and Klingler, 1959 occurred together with new forms which will be described later when more material is available. Lincolnshire. Immediately south of Lincoln, and 43 miles (69 km) north of Tilton- on-the-Hill is Bracebridge brick pit (SK 971671). Bracebridge has been described by Trueman (1918, pp. 103-104) and Howarth (1958, pp. xi-xii) but the worked part is now small and the lowest beds, formerly seen at the western end, are now obscured by rubbish. Trueman gives a large faunal list which includes foraminifera but no ostracods. The section sampled was about 17 feet (5 m) thick and included Howarths Beds 8, 9, and 10 which belong to the Stokesi and Subnodosus Subzones, but the upper subzone of the Margaritatus Zone, the Spinatum Zone, and the basal part of the Toarcian (Tenuicostatum Zone) were not exposed. This was unfortunate since the Spinatum Zone at Lincoln is in a clay facies. The samples yielded three species of Ogmoconcha ( O . contractula Triebel, 1941, O. sp. A, O. sp. B) and two of Pseudo- healdia (P. bispinosa and P. pseudohealdiae of Grundel 1964), but the higher samples from the Subnodosus Subzone were barren. At Kirton-in-Lindsey (SE 935005), north of Lincoln, a section of Liassic sediments has been described by Howarth and Rawson (1965). Using amaltheid ammonites to define the Domerian, these authors have proved the presence of 22 feet (7 m) of Davoei Zone clays overlain by 26 feet (8 m) of Domerian clays and Marlstone Rock Bed. In an immediately adjacent pit Toarcian shales (Tenuicostatum and Falciferum Zones) are partially exposed but have suffered prolonged weathering so that most samples were barren. A total of thirteen samples was analysed, the five lowest from the Davoei Zone and the uppermost one from the lower part of the Toarcian (Bed 21 of Howarth and Rawson, Tenuicostatum Zone). The Domerian consists of 18 feet (5-5 m) of Margaritatus Zone sediments (lower 6 feet 4 inches (1-9 m) proved to be Stokesi Subzone, clays above did not yield ammonites) and the Spinatum Zone represented by the lower part of the Marlstone Rock Bed (precise thickness as yet 606 PALAEONTOLOGY, VOLUME 17 unproved, Marlstone Rock Bed 8 feet (2-4 m) total thickness). As a result of recent work Dr. M. K. Howarth believes that the upper part of the Marlstone in the Mid- lands should be placed not in the Spinatum Zone but in the Tenuicostatum Zone on the basis of dactylioceratid ammonites found in the top part of the bed. A hard, grey, pebbly, calcareous mudstone (Bed 21) rests on top of, and appears almost con- tinuous with, the Marlstone Rock Bed and contains dactylioceratid ammonites and also ostracods including Ogmoconcha (see below) ; this bed is overlain by weathered shales and paper shales. The Domerian samples were good but Toarcian material was obtained from only one sample; three species were restricted to the Toarcian and seven to the Domerian (Margaritatus Zone), but eight species were common to both Margaritatus and topmost Davoei Zones. Samples Kl-5, Davoei Zone, Figulinum Subzone. Ogmoconcha contractula Triebel, 1941 Ogmoconcha sp. A Ogmoconcha sp. B Ogmoconcha sp. E Pseudohealdia bispinosa (Griindel 1964) Pseudohealdia aff. P. pseudohealdiae (Griindel 1964) Liasina vestibulifera Gramann, 1963 Wicherella semiora kirtonensis Lord, 1972 Samples K6-9, Margaritatus Zone. As Samples Kl-5 but with the addition of: Poly cope Ipumicosa Apostolescu, 1959 Polycope Isuborbicularis Terquem, 1885 Cytherella lindseyensis sp. nov. Pontocyprellal sp. Paracypris sp. Nanacy there (Nanacy there) simplex Herrig, 1 969 Gen. indet. sp. A Sample 13, Tenuicostatum Zone (Bed 21 of Howarth and Rawson). Ogmoconcha sp. Kinkelinella lenuicostati Martin, 1960 Trachycythere verrucosa Triebel and Klingler, 1959 Some degree of faunal change takes place between the two lower zones, but the faunal break between the Domerian and Toarcian over the sample hiatus of the Spinatum Zone is complete. The ostracod faunal change between Domerian and Toarcian has been well documented by Plumhoff (1967). Usually the genus Ogmoconcha has been thought to become extinct in the Spinatum Zone but Plumhoff has shown that the taxon survives into the oldest part of the Toarcian in a number of localities in north- west Europe. The present record of Toarcian Ogmoconcha is included in Plumhoff (1967, p. 563). A number of species are common to this locality and the Dorset coast. Wicherella semiora kirtonensis is a geographically distinct subspecies from the Dorset form W. semiora semiora (Lord 1972a), and Ogmoconcha contractula is here found in both Davoei and Margaritatus Zones but in Dorset is found only in the upper two subzones of the Margaritatus Zone. The highest Domerian samples (K 10 and K 1 1) were barren, a phenomenon attributed to leaching of the underlying clays during pre-Spinatum or early Spinatum Zone erosion. As in Dorset and Gloucester there is an erosion surface at the base of the Spinatum Zone Marlstone Rock Bed and also LORD: DOMERI AN AND TOARCI AN OSTRACODS 607 at the next locality of Roxby, near Scunthorpe, and possibly at Napton-on-the-Hill. A large ironstone working at Roxby (SE 914178), 10 miles (16 km) north of Kirton- in-Lindsey, exposes a section of the Lower Jurassic from the Frodingham Ironstone (Sinemurian, Semicostatum, and Obtusum Zones, Hallam 1963) to Toarcian clays with traces of the overlying Middle Jurassic. The Marlstone Rock Bed rests non- sequentially on Davoei Zone clays and it would appear that the Margaritatus Zone sediments, if deposited, were eroded before the deposition of the Marlstone (as at Kirton-in-Lindsey). The Toarcian is thinner at Roxby than at Kirton-in-Lindsey and probably consists only of the Tenuicostatum and Falciferum Zones capped by the Middle Jurassic. These two zones are present at Kirton-in-Lindsey but the unexposed part of the succession beneath the Middle Jurassic may contain higher zones, possibly up to the Bifrons Zone. Both Domerian and Toarcian, however, are showing a general thinning northwards towards the Market Weighton upwarp. Adams (1957) described Toarcian foraminifera from this area from borehole material. Yorkshire. The Domerian and Toarcian are poorly exposed in East Yorkshire. The Spinatum Zone is proved by a specimen of Pleuroceras spinatum from Everthorpe cutting (SE 905320), but there is no convincing record of the Margaritatus Zone (Neale 1958, p. 162). A clay sample from Everthorpe was unfossiliferous. The Toarcian is no longer exposed in East Yorkshire. The Yorkshire coast and Dorset coast provide the best exposures of the Lower Jurassic in the British Isles, but the former has not received such detailed attention. The most recent account of the Lower Lias is that of Fox-Strangways (1892) and previous to that the work of Tate and Blake (1876). Recently Howarth (1955) has redescribed the Middle Lias but modern work on the Upper Lias is incomplete and restricted to stratigraphical and palaeontological studies by Dean (1954) and Howarth (1962) and sedimentary geochemical work by Gad et al. (1969). Much of the Lias is argillaceous but highly lithified and difficult to prepare for microfossils. The results of sampling the Middle Lias (Domerian) along the coast south from Staithes (NZ 783188) have been briefly described (Lord 19716). The Margaritatus Zone fauna is comparable with Bracebridge particularly in con- sisting only of the two metacopid genera Ogmoconcha ( O . contractula Triebel, 1941, O. sp. B and O. sp. C) and Pseudohealdia (IP. cf. P. pseudohealdiae (Griindel 1964)). The Yorkshire fauna, however, is impoverished numerically, a question which has already been discussed in the light of the geochemistry of the sediments (Lord 19716). The Spinatum Zone was disappointing. Only in the Staithes section and Bracebridge do argillaceous sediments suitable for ostracod extraction techniques occur and these yielded a poor fauna at the former site and are now unexposed at the latter. The Toarcian of the Yorkshire coast has yet to be sampled properly, but reconnaissance samples from the various units contained small numbers of cytheracean ostracods. SYSTEMATIC DESCRIPTIONS In the following section well-known and illustrated species are not formally described. All specimens are deposited in the collections of the Department of Geology, Uni- versity of Hull. 608 PALAEONTOLOGY, VOLUME 17 Sample Data — full details of sample levels and localities are given in an unpublished pamphlet which has been deposited with the British Library at Boston Spa, York- shire, as Supplementary Publication No. SUP 14002 (16 pages). The sample data has been extracted from Lord (unpublished thesis, 1968). The genus Procytheridea Peterson , 1954. The use of the generic name Procytheridea in the discussion above refers to the terminology of previous authors and does not imply recognition of the genus in the Lower Jurassic. No Lower Jurassic species found so far would appear to be congeneric with Procytheridea exempla , the genotype, although this does not preclude the possibility of the presence of the genus in other parts of the European Jurassic. The genus Ogmoconcha Triebel , 1941. Assessing the degree of shape variation per- missible within one species of this genus is a recurrent difficulty and principally for this reason five ‘species’ have been left in open nomenclature. Text-fig. 3 shows shape differences between typical members of each so-called ‘species’. The taxonomy of this group is in a most unsatisfactory state, as an attempt has been made to demon- strate elsewhere (Lord 1972A). A single specimen from th egibbosus subzone in Dorset resembled Ledahia septenaria (Griindel 1964), described from the Domerian of northern Germany (see Malz 1971), but the muscle-scars were poorly preserved. Order podocopida Muller, 1894 Suborder platycopina Sars, 1866 Family cytherellidae Sars, 1866 Genus cytherella Jones, 1849 Cytherella lindseyensis sp. nov. Plate 90, figs. 1-3 Derivato nominis. Lindsey, a Lincolnshire region. Material. 8 valves, 1 carapace. Distribution. Sample K7, Margaritatus Zone, Kirton-in-Lindsey (type locality). Samples D49 and D50, Stokesi Subzone, Down Cliff. Dimensions. Length Sample K7 (mm) Holotype, Right valve, female HU. 53. J. 5 0-875 Paratypes, Left valve, female HU. 53. J. 6 0-85 Right valve, male HU. 53. J. 7 0-83 Diagnosis. A species of the genus Cytherella distinctive shape. Description. Shape sub-rectangular, elongate. Dorsal margin arched, highest point posterior of mid-length in females; ventral margin rectilinear or may be slightly convex medianly. Anterior margin rounded; posterior margin is slightly angular and Height Width (mm) (mm) 0-56 0 21 0-49 0-17 0-47 0 16 with an arched dorsal margin and LORD: DOMERIAN AND TOARCI AN OSTRACODS 609 Ogmoconcha sp. A Ogmoconcha contractula TRIEBEL .1941 O. sp.B O. sp C O. sp. D O sp.E m m . 0 1 Ledahia septenaria (GRUNDEL.1964) L.septenaria (GRUNDEL.1964) (from original illustration) text-fig. 3. Shape comparison of Ogmoconcha , Ledahia , and Pseudohealdia species. 610 PALAEONTOLOGY, VOLUME 17 with an inclined postero-ventral section. Right valve larger than left and overlaps smaller valve all round its margin. Surface of valves smooth and unornamented. Hinge simple, selvage of smaller valve fits into a peripheral groove which runs round margin of right valve. Selvage and complementary selvage groove are most pro- minently developed along the dorsal margin. Marginal zone simple. Marginal and normal pore canals simple, straight, and sparse. Females swollen posteriorly with greatest height posterior of mid-length. Males are featureless internally but females exhibit three depressions— one large depression posteriorly to accommodate eggs, one ventrally beneath the adductor muscle-scars, and a further concavity, weakly developed, in the antero-dorsal region. Males are lower posteriorly and less inflated. Muscle-scars biserial cytherellid type, composed of two rows of five or six scars. Remarks. Cytherella is not a common genus in the English Domerian and no species have been described from this sub-stage in Europe. The specimens from Dorset differ slightly from the Lincolnshire type material in their posterior shape in lateral view. This species differs from Cytherella toarcensis Bizon, 1960 (from the Upper Toarcian of Calvados) in shape and in the lack of a dorso-median sulcus. Cytherella lindseyensis is probably a direct antecedent of C. toarcensis. The depressions described in the female are not comparable to the double posterior depressions found in female Cytherelloidea and recently recorded by Bate (1972) in the Australian Cretaceous Cytherella atypica since only one depression is situated posteriorly. Genus cytherelloidea Alexander, 1929 Cytherelloidea anningi sp. nov. Plate 90, figs. 4-5 1961 Cytherelloidea sp. 24 Cousin and Apostolescu, p. 429, fig. 2. EXPLANATION OF PLATE 90 Scanning electron micrographs. Figs. 1-3. Cytherella lindseyensis sp. nov., Margaritatus Zone, Kirton-in-Lindsey, uncoated, x 66. 1, left valve, female, Paratype HU.53.J.6. 2, right valve, male, Paratype HU. 53. J. 7. 3, right valve, internal, female, Holotype HU. 53. J. 5. Figs. 4-5. Cytherelloidea anningi sp. nov. Stokesi Subzone, Down Cliff, Dorset coast, uncoated, x 66. 4, right valve, male, Holotype HU.53.J.14. 5, right valve, female, Paratype HU.53.J.15. Figs. 6-9. Kinkelinella sermoisensis (Apostolescu 1959), Lower Toarcian, Byfield, x 66. 6, left valve, HU. 55. J. 13, carbon coated. 7, right valve, HU. 55. J. 14, carbon coated. 8, right valve, aluminium coated. 9, right valve internal, aluminium coated. Fig. 10. Kinkelinella sp. I. (Apostolescu 1959), Lower Toarcian, Byfield. Right valve, HU. 55. J. 23, carbon coated, x 66. Figs. 11 -12. Ektyphocythere sp. A, Subnodosus Subzone, Thorncombe Beacon, Dorset coast, aluminium coated, x 66. 1 1, left valve, HU. 57. J.l. 12, right valve, HU. 57. J. 2. Figs. 13-15. Gen. indet. sp. A, Margaritatus Zone, Kirton-in-Lindsey, carbon coated, x 66. 13, left valve, female, HU.56.J.26. 14, right valve, female, HU. 56. J. 27. 15, carapace, male, right view, HU.56.J.28. Fig. 16. Ektyphocythere aff. champeauae Bizon, 1960. Levesquei Zone, Down Cliff Clay, Thorncombe Beacon, Dorset coast. Carapace, left view, HU.57.J.6, aluminiun coated, x 66. Figs. 17-21. Gen. indet. Ibucki Bizon, 1960. Lower Toarcian, Byfield, carbon coated, x 66. 17, left valve, male, HU.57.J.7. 18, right valve, male, HU. 57. J. 10. 19, left valve, female, HU.57.J.12. 20, left valve, interior, female. 21, right valve, female, HU.57.J.13. PLATE 90 LORD, Liassic ostracods 612 PALAEONTOLOGY, VOLUME 17 1961 Cytlierelloidea sp. 24 G. Bizon, p. 436, table 2. 1961 Cytlierelloidea sp. 24 Champeau, pp. 438, 442, and 443. 1961 Cytlierelloidea sp. 24 Oertli and Grosdidier, p. 460, table 6. 1962 Ostracod Nr. 107 Klingler, p. 101, pi. 13, fig. 31, table 7. 1963 Cytlierelloidea sp. 24 Oertli, pis. 13 (ii) and 16 (i). Derivato nominis , in honour of Mary Anning (1799-1847) of Lyme Regis, an early student of Lias fossils. Material. 5 valves. Distribution. Sample D34, D44, and D77 Stokesi and Gibbosus Subzones, Down Cliff. Dimensions. Holotype, Right valve, male (D34) HU. 53. J. 14 Length (mm) 0-64 Height (mm) 0-36 Width (mm) 010 Paratypes, Right valve, female (D34) HU. 53. J. 15 0-76 0 41 013 Right valve, male (D77) HU. 53. J. 16 0 71 0-43 012 Left valve, female (D77) HU. 53. J. 17 0-72 0-43 016 Diagnosis. A species of Cytlierelloidea distinguished by its ornamentation of a single rib which commences near the antero-dorsal margin, runs round the valve parallel to the margins, and terminates slightly below its point of commencement. Dis- tinguished also by pronounced sexual dimorphism. Description. Shape elongate, rectangular, or sub-rectangular. Sexual dimorphism marked; females larger than males and differ in outline. Dorsal and ventral margins straight but may be slightly convex or concave. Marginal rim present along the anterior and dorsal margins in females and along anterior margin in males. Females inflated posteriorly, posterior margin gently rounded in the dorsal part but more sharply curved in the ventral part. Males are more sharply rounded posteriorly than females, and slightly angular at, or just above, mid-height. Greatest length at mid- height. Dorsal and ventral margins parallel. Maximum width posteriorly in both sexes. Valve surface smooth and ornamented by a single rib which commences close to the antero-dorsal margin and runs completely round the valve parallel to the margins and terminates close to, but on the ventral side of, its starting-point— a single spiral which is almost a closed circle. At the point where it finishes the rib may incline slightly in an antero-ventral direction. This single rib is close to the valve margins but does not border the shell. In the centre of the valve a small depression is sometimes developed which corresponds to the point of attachment of the adductor muscles on the inside of the valve. Details of muscle-scars and pore canals were not visible. Flingement as in Cytherella. Female valves infilled so that the two diagnostic posterior cavities, if present, could not be seen. Remarks. Cytlierelloidea anningi differs from C. pulchella Apostolescu, 1959 in the different rib pattern and C. pulchella also possesses a reticulate surface; the LORD: DOMERI AN AND TOARCIAN OSTRACODS 613 present species differs from C. cadomensis Bizon, 1960 in its lack of a peripheral inflation. This species has been described from France and Germany by a number of authors and is notable for its restriction to the Margaritatus Zone for which it forms a useful marker fossil in north-west Europe. The only exception is Oertli and Grosdidier’s (1961, p. 460) record of throughout the Upper Sinemurian to the top of the Lower Pliensbachian in the Paris Basin. This is difficult to explain since their small drawing shows a Cytherelloidea with a distinctive coiled rib apparently synonymous with the present species. Possibly the species was restricted to the central part of the Paris Basin until the end of the Lower Pliensbachian and then extended its habitat to southern England, the borders of the Paris Basin, and north Germany (text-fig. 4). substage zone Do 'set Berne\ Paris /al 101 Basin S.E. Pari Ardennes s Basin Lorr N.Gen aine many U. Pliensbachian spinatum ? « margaritatus I i 1 a i 1 L Pliensbachian davoei wm m ? i : ibex m ? jamesoni U. Sinemurian raricostatum oxynotum d *;._)/ obtusum d * L. Sinemurian turneri text-fig. 4. Distribution of Cytherelloidea anningi sp. nov. Suborder podocopina Sars,1866 Superfamily cytheracea Baird, 1850 Genus KinkelineUa Kinkelinella sp. I (Apostolescu 1959) Plate 90, fig. 10 This species is easily distinguished from other KinkelineUa by virtue of its ornament, which is a pattern of slightly sinuous ribs which run sub-vertically from the dorsal to the ventral margin and which are slightly flexed towards the anterior in the ventral half of the shell, but lack secondary transverse ribs. KinkelineUa sp. I occurs in association with K. sermoisensis (Apostolescu 1959) (PI. 90, figs. 6-9) in the Falciferum and Bifrons Zones at Byfield. The two species are figured to demonstrate the contrast in ornamental pattern. K. sp. I was originally recorded from the Toarcian (Bifrons Zone) and Aalenian (Aalensis Zone) of the Paris Basin by Apostolescu in 1959 and appears identical with Ostracod 1081 from the Toarcian and Aalenian of south-west Germany (Buck 1954). 614 PALAEONTOLOGY, VOLUME 17 Genus ektyphocythere Ektyphoeytliere sp. A Plate 90, figs. 11-12 Material. 1 carapace, 16 valves, and fragments. Distribution. Samples D62-64— Subnodosus Subzone, Thorncombe Beacon. Dimensions. Left valve (D62) HU.56.J.1 Length (mm) 0-62 Height (mm) 0-36 Width (mm) 0-16 Left valve (D62) HU.57.J. 1 0-67 0-36 C 16 Right valve (D62) HU. 57. J. 2 (broken after photography) 0-64 + 0-33 016 Remarks. This species resembles Procytheridea rugosa Bizon, 1960 from the French Upper Toarcian in its essentially triangular ornament and in shape but differs in being somewhat more rectangular and possessing many more primary ribs. Ektyphocythere aff. champeauae Bizon, 1960 Plate 90, fig. 16 aff. 1954 Ostracod 1099a, Buck. aff. 1956 Cytlieropteron sp. 1099 (Buck 1954), Apostolescu and Bourdon, table 2. aff. 1960 (March) Procytheridea champeauae- Bizon, pp. 205, 208, and 210, pi. 1, fig. 1 a-d. PI. 2, fig. 1 a-g. aff. 1960 (June) Procytheridea arcuatocostata— Martin, pp. 142-144, pi. 11, figs. 38-39. 1962 Ostracod Nr. 85 Klingler, p. 107, pi. 14, fig. 48, table 7. Material. 4 carapaces, 7 valves, and fragments. Distribution. D80 and D82— Down Cliff Clay— Thorncombe Beacon. W1 and W6— Down Cliff Clay— Watton Cliff. Dimensions. Left valve (Wl) Length (mm) Height (mm) Width (mm) HU. 57. J. 3 Left valve (D52) 0-58 0-34 0 17 HU.57.J.4 Right valve (W6) 0-52 0-34 016 HU. 57. J. 5 Carapace (D82) 0-56 0-27 0-14 HU.57.J.6 0-54 0-30 0-25 Remarks. The specimens are identical with figures of Ostracod Nr. 85 Klingler (Klingler 1962) from the Upper Toarcian of south Germany. This species is very close to Ostracod 1099a Buck which that worker recorded from the top and bottom of the German Toarcian. In shape (laterally) and in ornament this species is close to Procytheridea champeauae Bizon, 1960 (synonymous with P. arcuatocostata Martin, 1960) and P. vitilis Apostolescu, Magne and Malmoustier, 1961 but has LORD: DOMERIAN AND TO A RCI AN OSTRACODS 615 more primary ribs and is here considered to show affinities with the former species, although the difference may prove to be of sub-specific rank. The comparison of the rib patterns is of interest (text-fig. 5). The following three taxa would appear to be closely related : (i) the present species, which may be a sub-species of (ii). (ii) Procytheridea champeauae Bizon, 1960. Fewer primary ribs than (i) but orna- ment almost same as (iii). A Procytheridea champeauau BIZON, 1960 la-left view lb-dorsal view Paratype - C B 139 Luxembourg spinatum zone P. arcuatocostata MARTI N , 1 960 Paratype -SMF Xe 2986 Sudwurttemberg tenuicostatum zone P. vitilis APOSTOLESCU et. al. 1961 3a - Paratype - left view 3b- Holotype -dorsal view. Thouars. "toarcense zone." Ektyphocythere aff champeauae BIZON, 1960 sample D 82 Dorset -Down Cliff Clay levesquei zone Sketches of original illustrations, scale lines all 0 5 mm long text-fig. 5. Morphology of Procytheridea champeauae Bizon, 1960, P. arcuatocostata Martin, 1960, P. vitilis Apostolescu et al., 1961, and Ektyphocythere all. champeauae Bizon, 1960. 616 PALAEONTOLOGY, VOLUME 17 (iii) Procytheridea vitilis Apostolescu, Magne and Malmoustier, 1961 which differs from (i) and (ii) in shape when viewed dorsally — other comparative features poorly known. The use of the generic name Procytheridea is discussed above. Genus Indeterminate Gen. indet. sp. A Plate 90, figs. 13-15 Material. 20 valves, 1 carapace. Distribution. Samples K6-9 — Margaritatus Zone. Kirton-in-Lindsey. Dimensions. Left valve ?female (K8) Length (mm) Height (mm) Width (mm) HU. 56. J. 26 Right valve ?female (K.8) 0-50 0-26 015 HU.56.J.27 Carapace ?male (K6) 0-46 0-25 014 HU.56.J.28 0-56 0-25 0-26 Remarks. This species is probably Ostracod 999 of Buck (1954) and possibly Pro- cytheridea ? sp. 22 of Oertli and Grosdidier (1961, p. 460) or the species called Procytheridea ? sp. E (Apostolescu 1959) by Oertli (1963, pi. 15). The species may be related to Procytheridea vermicu/ata Apostolescu, 1959. Gen. indet. Ibucki Bizon, 1960 Plate 90, figs. 17-21 71960 Procytheridea bucki Bizon, p. 205, pi. 1, fig. 2 a-e. 1963 Procytheridea bucki Bizon, 1960, Oertli, pi. 20. Material. 4 carapaces, 255 valves. Distribution. Samples B1 -4— Lower Toarcian. Byfield. Dimensions. Length (mm) All specimens from Sample B4 Left valve, male HU. 57. J. 7 0-79 Left valve, male HU. 57. J. 8 0-77 Right valve, male HU. 57. J. 9 0-76 Right valve, male HU. 57. J. 10 0-78 Left valve, female HU. 57. J. 11 0-66 Left valve, female HU. 57. J. 12 0-63 Height Width (mm) (mm) 0-43 0-23 0-41 0-21 0-37 0-22 0-39 0-24 0-44 0-22 0-40 0-22 LORD: DOMERI AN AND TOARCIAN OSTRACODS 617 Right valve, female Length (mm) Height (mm) Width (mm) HU. 57. J. 13 Right valve, female 066 0-38 0-23 HU. 57. J. 14 Carapace, female 0-68 0-38 0-23 HU. 57. J. 15 070 0-43 0-40 Remarks. As figured by Bizon Procytheridea bucki differs from the specimens described here which are higher and more truncated posteriorly, especially in the females; slight differences in ornament may also exist but it is difficult to be certain from Bizon’s figures. Comparison of internal structures is also unsatisfactory, the hinge of P. bucki for example having been described simply as the same as in Pro- cytheridea. A very distinctive early Toarcian species. DISCUSSION AND CONCLUSIONS Mesozoic ostracods are almost all considered to have been of benthonic habit existing as scavengers around vegetation, on the sea-floor, or between sediment particles in the uppermost layers of the substrate. Modern planktonic forms are well known, as are analogous Palaeozoic types which are interpreted as having had a similar life style, but Mesozoic planktonic ostracods are almost unknown, with the exception of a very few Cretaceous forms (Pokorny 1964; Kaye 1965). The forms described in this account thus belong to marine benthic communities, are typical elements of the ostracod fauna inhabiting the north-west European shelf sea of the time and as such are indicative of relatively shallow-water conditions. They are also influenced in their distribution by facies variation. It will be clear that only certain sediments are susceptible to micropalaeontological preparation techniques for Ostra- coda, i.e. clays, marls, sands, and that others such as limestones and well-cemented sandstones and ironstones cannot be disaggregated without destroying the calcareous microfossils. Thin-section analysis is not an appropriate method for identifying ostracods. There is thus an initial facies control over the assemblages examined in that only particular facies can be studied and it is very difficult to investigate the full range of facies control. In Margaritatus Zone times the assemblages in south-west England not unnaturally bear close comparison with those of the Paris Basin. In a similar way the species found in Lincolnshire closely resemble forms recorded from Germany and there is no reason to suppose that there was any hindrance to the movement of these benthonic organisms to the north or south of the Anglo-Belgian landmass. Faunal elements common to both France and Germany occur, but with greater affinity to the former in south-west England and to the latter in north-east England. The situation does not hold between the Cotswolds and Lincolnshire, that is immediately to the west of the landmass, where not only are exposures poor but ostracods rare, e.g. Napton-on-the-Hill. This phenomenon might be considered a feature associated with shallow-water con- ditions just off-shore (the so-called ‘Oxfordshire Shallows’), but the sediments at Napton are not particularly shallow water and ammonites occur. Evidence from further east and west would be enlightening. M 618 PALAEONTOLOGY, VOLUME 17 The progressive shallowing of the sea during Domerian times is generally thought to have reached its maximum during the Spinatum Zone. Here, shallow-water sedi- ments including oolitic limestones were deposited, local erosion took place, and the sea was shallow enough to form partially isolated water bodies thus permitting the development of provincialism in ammonites and brachiopods. Present data suggest that ostracod diversity decreased through the Margaritatus Zone, although the pattern may be disturbed at the end by pre- or early Spinatum Zone erosion. For reasons referred to earlier, ostracods from the Spinatum Zone are rare and do not permit any useful comment, but it is noteworthy that a number of species described from this zone in north-east Germany by Herrig (1969u, b ) occur in the Margaritatus Zone in England, viz. Pseudohealdia bispinosa, Ogmoconcha contractula, Trachycythere tubulosa seratina , Nanacy there (TV.) simplex, and probably Ogmoconcha adenticulata (if this is synonymous with Ogmoconcha sp. E). The deepening of the sea after Spinatum Zone time was part of an early Toarcian transgressive phase which affected most of the north-west European shelf sea and with this deepening came a different ostracod fauna distinguished by a Middle Jurassic aspect. Lower Jurassic ostracod faunas in the strict sense disappear at the end of the Domerian. Hallam (1967) has described how bottom stagnation during Falciferum Zone times resulted in extinc- tion for much of the benthonic fauna which was then in part replaced by forms with Middle Jurassic affinities; the ostracod evidence is too fragmentary at the moment to show whether the same is true for this particular group. The Whitbian (Tenui- costatum Zone) ostracods from Kirton are species described from Germany, whereas the Whitbian assemblages from Byfield contain species described from both France and Germany but with the former the first to be established in the area. For the Toarcian of south-west England from the south Cotswolds to Dorset, Davies (1969) postulated a sedimentological model in which the diachronous Middle and Upper Toarcian sands are looked on as a large sand bar which gradually migrated south- wards. Thus these youngest Lias sediments are well represented but generally in a non-calcareous condition in which ostracods are not preserved. The only Yeovillian assemblages examined, those from the Down Cliff Clay below the sand-bar facies in Dorset, contain many species from the Paris Basin but are noteworthy for the absence of Aphelocythere, an important genus in the Upper Toarcian and Aalenian of Germany and France (Plumhoff 1967). Much of the early Toarcian occurs in the form of a condensed deposit in Dorset, and around Yeovil where a clay facies is developed exposure is poor. A similar situation is found in the north Cotswolds where most of the Toarcian is clay but exposure again is poor. In the Midlands two Whitbian sections were examined (see above) but Yeovillian is absent. Thus both sediment type and exposure hamper full examination of Toarcian ostracods. Comparatively little is known about Tethyan ostracods of this period, although Barbieri (1964) has described microfossils from a Sinemurian to Aalenian sequence in Sicily wherein the ostracods are generally comparable to species from north-west Europe. The Pliensbachian-Toarcian boundary. The marked change which occurred both in fauna and sediments at the end of the Upper Pliensbachian (i.e. Domerian) is generally recognized and has been well documented, particularly by Hallam (e.g. 1967 and LORD: DOMERI AN AND TOARCI AN OSTRACODS 619 1972). Distinct changes occurred in the composition of both planktonic and benthonic faunas. The very shallow-water conditions at the end of the Pliensbachian were followed by transgressive marine conditions during which argillaceous sediments were deposited over much of north-west Europe. The change in ostracod fauna has been described by Plumhoff (1967) who emphasized the faunal turnover by referring to the Middle Jurassic aspect of Toarcian faunas. When examined in detail, as Plum- hoff has done, it is clear that the change begins to take place during the early phase of the transgression, in the Tenuicostatum Zone, where a mixed assemblage can be found. The most important change is the disappearance of metacopid ostracods, in particular the important genus Ogmoconcha. This genus occurs in the Tenui- costatum Zone (Kirton-in-Lindsey) but soon disappears and, since it is a ubiquitous and frequently dominant element in Lower and Middle Lias assemblages, its absence coupled with the appearance of new ostracods such as the cytheracean genus Kinke- linella completely alters the structure of the assemblages. The Toarcian transgression thus coincides and is intimately associated with the distribution of quite a different set of species which can be readily distinguished throughout the area of the north- west European epicontinental sea. Ostracod Zonation. No attempt has been made to provide an ostracod based zonal scheme for the Domerian and Toarcian in view of the incompleteness of the sections studied and the absence or poor representation of ostracods at certain levels, for example in the Toarcian of the Dorset coast. The best complete sequence of ostra- cods was found in the Dorset coast Domerian, the species distribution of which is given in text-fig. 2 and is discussed in the text. Individual species seem to be of value as stratigraphic indices but this may be a purely local or ecologically controlled phenomenon. With the data available it is not possible to postulate assemblage zones. The record provided in this paper is only relatively complete for the Margaritatus Zone, almost non-existent for the Spinatum Zone because of unfavourable sediment type, and fragmentary for the Toarcian for a number of reasons. To complete the investigation borehole material is essential, particularly for the Midlands. The sequence in the Llanbedr (Mochras) Borehole is of great importance since it is not only thick (Domerian 483 feet (147 m), Toarcian 858 feet (261 m)) but contains rich assemblages of foraminifera and ostracods (B. Johnson and P. L. Sherrington, personal communication) which will provide most valuable information about the history of the Lower Jurassic on the western edge of Europe. Acknowledgements. I am greatly indebted to Dr. J. W. Neale (University of Hull) and Professor D. T. Donovan (University College London) for critically reading the original manuscript. Dr. M. K. Howarth (British Museum (Natural History)) kindly provided information about his work in the Midlands and Yorkshire. Mrs. J. Irvine (University of East Anglia) prepared the scanning electron micrographs and Mrs. G. Watkin and Mr. H. Williams (University College of Wales, Aberystwyth) drafted and photographed most of the diagrams. Financial aid from the University of Hull is gratefully acknowledged. 620 PALAEONTOLOGY, VOLUME 17 REFERENCES adams, C. G. 1957. A study of the morphology and variation of some Upper Lias foraminifera. Micro- paleontology, 3, 205-226. andhrson, F. w. 1964. Rhaetic Ostracoda. Bull. geol. Surv. Gt Br. 21, 133 174. apostolescu, v. 1959. Ostracodes du Lias du bassin de Paris. Revue Inst. fr. Petrole, 14, 795-826. - and bourdon, m. In de charpel, o., vatan, a., fondeur, c., rouge, p. e., apostolescu, v. and bourdon, m. 1956. Correlation dans le Lias marneux des Causses majeurs. Ibid. 11, 439-450. — magne, j. and malmoustier, G. 1961. Quelques especes nouvelles d’Ostracodes du Toarcien de Thouars (Deux Sevres). Colloque sur le Lias Frangais. Mem. Bur. Recher ches geol. min. 4, 399-405. arkell, w. j. 1947. The Geology of Oxford. 267 pp. Oxford. 1956. Jurassic Geology of the World, xvi + 806 pp. Edinburgh. barbieri, f. 1964. Micropaleontologia del Lias e Dogger del Pozzo Raguso 1 (Sicilia). Riv. ital. Paleont. 70, 709-831. barnard, t. 1950. Foraminifera from the Upper Lias of Byfield, Northamptonshire. Q. Jl geol. Soc. Lond. 106, 1-34. bate, r. h. 1972. Upper Cretaceous Ostracoda from the Carnarvon Basin, Western Australia. Special Papers in Palaeontology, 10, 85 pp., 27 pis. bizon, G. 1961. Lorraine, region de Nancy etThionville. Colloque sur le Lias Frangais. Mem. Bur. Recher ches geol. min. 4, 433-436. bizon, j. j. 1960. Sur quelques Ostracodes du Lias du bassin parisien. Revue Micropaleont. 2, 203-211. blake, J. F. 1876. Subclass Entomostraca. In tate, r. and blake, j. f. The Yorkshire Lias, pp. 429-435. buck, E. 1954. Stratigraphisch wichtige Ostracoden aus dem Lias und Dogger von SW- Deutschland. Un- published table. champeau, H. 1961. Etude de la microfaune des niveaux marneux du Lias dans le sud-est du bassin de Paris. Colloque sur le Lias Frangais. Mem. Bur. Recherches geol. min. 4, 437-443. cousin, n. and apostolescu, v. In cousin, n., espitalier, j., sigal, j. and apostolescu, v. 1961. Ardennes, region de Mezieres (Departement des Ardennes). Ibid. 423-431. curry, d., Hamilton, d. and smith, a. j. 1970. Geological and shallow subsurface geophysical investigations in the Western Approaches to the English Channel. Report 70/3 I.G.S. H.M.S.O. davies, D. k. 1969. Shelf Sedimentation: an example from the Jurassic of Britain. J. sedim. Petrol. 39, 1344-1370. dean, w. t. 1954. Notes on part of the Upper Lias succession at Blea Wyke, Yorkshire. Proc. Yorks, geol. Soc. 29, 161-179. — donovan, d. t. and howarth, m. k. 1961. The Liassic ammonite zones and subzones of the North- west European Province. Bull. Br. Mus. nat. Hist. (Geol.), 4, 435-505. dingle, r. v. 1967. The Geology of the North Sea off the North-East coast of England. University of Hull, unpublished Ph.D. thesis. — 1971. A marine geological survey off the north-east coast of England (western North Sea). Jl geol. Soc. 127, 303-338. field, R. a. 1966. Species of the family Cytherellidae (Ostracoda) from the Lower Lias of South Dorset, England. Senckenberg. leth. 47, 87-105. fischer, w. 1962. Ostracoden der Gattungen Monoceratina Roth, 1928, Cytheropteron G. O. Sars 1865 und andere im Lias Zeta Schwabens. Neues Jb. Geol. Paldont. Abh. 114, 333-345. fox-strangways, c. 1892. The Jurassic Rocks of Britain, vol. I. Yorkshire. Mem. geol. Surv. U.K. ix + 551 pp. gad, m. a., catt, j. a. and le riche, h. h. 1969. Geochemistry of the Whitbian (Upper Lias) Sediments of the Yorkshire Coast. Proc. Yorks, geol. Soc. 37, 105-139. gramann, f. 1962a. Extremitatenfunde an liassischen Bairdien (Ostracoda). Paldont. Z. 36, 28-32. 1962A Skulptierte Ostracoden aus dem niederrheinischen Lias. Fortschr. Geol. Rheinld Westf. 6, 185-198. — 1963. Liasina n. gen. (Ostracoda) aus dem deutschen Lias. Geol. Jb. 82, 65-74. grundel, j. 1964. Zur Gattung Healdia (Ostracoda) und zu einigen verwandten Formen aus dem unteren Jura. Geologie, 13, 456-477. LORD: DOMERI AN AND TOARCIAN OSTRACODS 621 hallam, a. 1955. The Palaeontology and Stratigraphy of the Marlstone Rock-bed in Leicestershire. Trans. Leicester lit. phiL Soc. 49, 17-35. — 1963. Observations on the palaeoecology and ammonite sequence of the Frodingham Ironstone (Lower Jurassic). Palaeontology, 6, 554-574. — 1967. An environmental study of the Upper Domerian and Lower Toarcian in Great Britain. Phil. Trans. Roy. Soc. Load. B , 252, 393-445. — 1972. Diversity and density characteristics of Pliensbachian-Toarcian molluscan and brachiopod faunas of the North Atlantic margins. Lethaia, 5, 389-412. herrig, E. 1969a, b. Ostracoden aus dem Ober-Domerien von Grimmen westlich Greifswald. Geologie, 18, 446-471 and 1072-1101. hoffmann, k. 1962. Lias und Dogger im Untergrund der Niederrheinischer Bucht. Fortschr. Geol. Rheinld West f. 6, 105-184. howarth, m. k. 1955. Domerian of the Yorkshire coast. Proc. Yorks, geol. Soc. 30, 147-175. 1957. The Middle Lias of the Dorset coast. Q. Jl geol. Soc. Lond. 113, 185-203. — 1958. The Ammonites of the Liassic Family Amaltheidae in Britain. Palaeontogr. Soc. ( Monogr .), xxxvii + 53 pp. London. — 1962. The Jet Rock Series and the Alum Shale Series of the Yorkshire coast. Proc. Yorks, geol. Soc. 33, 381-422. — 1964. Whitbian and Yeovilian Substages. Colloque du Jurassique (1962), pp. 189-192. Luxembourg. — and rawson, p. f. 1965. The Liassic Succession in a Clay Pit at Kirton-in-Lmdsey, North Lincoln- shire. Geol. Mag. 102, 261-266. jackson, j. f. 1926. The Junction-Bed of the Middle and Upper Lias on the Dorset coast. Q. Jl geol. Soc. Lond. 82, 490-525. jones, t. r. 1872. On some Bivalve Entomostraca from the Lias (‘Infralias’) of Yorkshire. Ibid. 28, 146-147. — 1894. On the Rhaetic and some Liassic Ostracoda of Britain. Ibid. 50, 156-169. kaye, p. 1965. Some new British Albian Ostracoda. Bull. Br. Mus. nat. Hist. (Geol.), 11, 215-253. Kent, p. e. 1949. A structure contour map of the surface of the buried pre-Permian rocks of England and Wales. Proc. Geol. Ass. 60, 87-104. klingler, w. In simon, w. and bartenstein, h. (eds.). 1962. Leitfossilien der Mikropalaontologie. Vol. 1, viii + 432 pp. ; vol. 2, 59 plates and 22 tables. Berlin. lord, a. r. 1968. English Domerian and Toarcian Ostracoda. Unpublished thesis. University of Hull. — 1 971c/. Revision of some Lower Jurassic Ostracoda from Yorkshire. Palaeontology , 14, 642-665. — 1971A In catt, j. a., gad, m. a., le riche, H. H. and lord, a. r. Geochemistry, Micropalaeontology and Origin of the Middle Lias Ironstones in north east Yorkshire (Great Britain). Chem. Geol. 8, 61-76. — 1972a. Wicherella and Gramannella, two new genera of Lower Jurassic Ostracoda (Crustacea). Palaeontology, 15, 185-194. — 1972h. The ostracod genera Ogmoconcha and Procytheridea in the Lower Jurassic. Bull. geol. Soc. Denmark, 21, 319-336. mckerrow, w. s. and baden-powell, d. f. w. 1953. Easter Field Meeting 1952: The Jurassic rocks of Oxfordshire and their superficial deposits. Proc. Geol. Ass. 64, 88-98. malz, h. 1971. Zur Taxonomie ‘glattschaliger’ Lias-Ostracoden. Senckenberg. leth. 52, 433-455. martin, G. p. R. In hoffmann, K. and martin, G. p. R. 1960. Die Zone des Dactylioceras tenuicostatum (Toarcien, Lias) in NW- und SW-Deutschland. Palaont. Z. 34, 103-149. mouterde, r. 1953. Etudes sur le Lias et le Bajocien des Bordures Nord et Nord-est du Massif Central Fran^ais. Bull. Servs. Carte geol. Fr. 50, 63-521. neale, j. w. In de boer, G., neale, j. w. and penny, l. f. 1958. A guide to the geology of the area between Market Weighton and the Humber. Proc. Yorks, geol. Soc. 31, 157-209. oertli, h. j. 1963. Faunes dOstracodes du Mesozoique de France/ Mesozoic ostracod faunas of France. 57 pp., 90 plates. Leiden. — and grosdidier, e. 1961. Ostracodes de quelques sondages du Lias du bassin de Paris. Colloque sur le Lias Francois. Mem. Bur. Recherches geol. min. 4, 459-461. palmer, c. p. 1971. The Stratigraphy of the Stonehouse and Tuffley claypits in Gloucestershire. Proc. Bristol nat. Soc. 32, 58-68. plumhoff, f. 1967. Die Gattung Aphelocythere (Ostracoda) in NW-europaischen Jura und zur Entwicklung der Mikrofauna am Ubergang Domerium/Toarcium. Senckenberg. leth. 48, 549-577. 622 PALAEONTOLOGY, VOLUME 17 pokorny, v. 1964, Conchoecial cretacea n. sp., first fossil species of the family Halocyprididae (Ostracoda, Crustacea). Acta Univ. Carolinae Geol. 2, 175-180. sohn, i. G. 1968. Triassic ostracodes from Makhtesh Ramon, Israel. Bull. geol. Surv. Israel, 44, 1-71. terquem, o. 1885. Les Entomostraces Ostracodes du systeme oolithique de la zone a Ammonites parkinsoni de Fontoy (Moselle). Mem. Soc. geol. Fr. ( ser . 3), 4, 1-46. triebel, e. 1941. Zur Morphologie und Oekologie der fossilen Ostracoden. Mit Beschreibung einiger neuer Gattungen und Arten. Senckenbergiana, 23, 294-400. — and klingler, w. 1959. Neue Ostracoden-Gattungen aus dem deutschen Lias. Geol. Jb. 76, 335-372. trueman, a. e. 1918. The Lias of South Lincolnshire. Geol. Mag. 5, 64-73, 101-111. walford, e. a. 1879. On some Upper and Middle Lias beds, in the neighbourhood of Banbury. Proc. Warwick. Nat. Archaeol. Fid Club [for 1878], appendix pp. 1-23. wicher, c. a. 1938. Mikrofaunen aus Jura und Kreide insbesondere Nordwestdeutschlands. I. Tell: Lias alpha bis epsilon. Abh. preuss. geol. Landesanst . (n.F.) 193, 1-16. Wilson, v., welch, f. b. a., robbie, j. a. and green, G. w. 1958. Geology of the Country around Bridport and Yeovil (explanation of sheets 327 and 312). Mem. geol. Surv. U.K. xii+239 pp. wood, a. and woodland, a. w. 1968. Borehole at Mochras, West of Llanbedr, Merionethshire. Nature, 219, 1352-1354. woodward, h. b. 1893. The Jurassic rocks of Britain, vol. III. The Lias of England and Wales (Yorkshire excepted). Mem. geol. Surv. U.K. xii + 399 pp. A. R. LORD Department of Geology University College London London, WC1E6BT Revised manuscript received 21 November 1973 DINOFLAGELLATE CYSTS FROM THE APTIAN TYPE SECTIONS AT GARGAS AND LA BEDOULE, FRANCE by roger j. davey and JEAN-PIERRE verdier Abstract. Microplankton assemblages are described for the first time from the Aptian type localities at La Bedoule and Gargas, south-east France. A total of sixty-eight species and varieties of dinoflagellate cysts are recorded, a num- ber of which are discussed in detail, and four new species are erected. These are Aptea securigera , Cyclonephelium tabulatum , Meiourogonyaulax psoros , and Protoellipsodinium clavulum. Microplankton distribution charts for the sections studied and a summary chart of selected age-significant forms are included. The following species are shown to be characteristic of, and confined to, the Aptian— Aptea polymorpha , C. tabulation, M. psoros and Trichodinium sp. These species, together with longer-ranging forms, may be used to distinguish Aptian from Albian and Barremian strata. Finally, our stratigraphic results are compared with those described in studies of similarly-aged sediments. Sediments of Aptian age have been rather neglected by earlier microplankton workers; they have been studied briefly only in Germany (Eisenack 1958; Alberti 1961) , France (Millioud 1969), and Australia (Cookson and Eisenack 1958, 1960, 1962) . The present paper, the first detailed study of Aptian dinoflagellate cysts, is intended to remedy this deficiency by presenting a taxonomic and stratigraphic analysis of microplankton recovered from Aptian stratotype material. This paper is a natural continuation from our previous contributions on Albian microplankton (Davey and Verdier 1971, 1973). The sediments in the type localities of La Bedoule and Gargas, south-east France (text-fig. 1) consist predominantly of marls and clays, which usually yield rich palynologic assemblages dominated by dinoflagellate cysts, although spores, pollen grains, and woody material are generally common. The basal Bedoulian (Lower Aptian) consists only of post-Urgonian reefal limestones which proved to be almost barren of palynomorphs. All slides containing holotypes are housed in the Laboratoire de Micropaleontologie de l’Ecole Pratique des Hautes Etudes, 8, rue de Buffon, Paris 5cmc, France. STRATIGRAPHIC AND GEOGRAPHIC LOCATION OF SAMPLES The Aptian stage (text-fig. 2) was introduced in 1840 by d’Orbigny, who subsequently, in 1842 and 1850, slightly refined his original definition. As envisaged by d’Orbigny, the Aptian was a stratigraphic unit that included only what is now considered to be the Upper Aptian (Gargasian substage). His choice of the Apt region (Vaucluse) as the type area for the Aptian stage was unfortunate in that it later lead to considerable controversy over delimitation of the stage. The main reason for this controversy was that in neighbouring areas, as several geologists pointed out (Matheron 1842; Reynes 1861; Hebert 1864, 1871, 1872), calcareous sediments containing a fauna with Aptian affinities were present between the massive ‘Urgonian limestones’ and the ‘Aptian [Palaeontology, Vol. 17, Part 3, 1974, pp. 623-653, pis. 91-93.] 624 PALAEONTOLOGY, VOLUME 17 marls’. In the Apt region these latter two formations are practically in contact with only minor calcareous passage beds separating them; however, at La Bedoule (Bouches du Rhone) the passage beds are well developed. The above authors clearly recognized that the passage beds and the overlying marly sequence displayed sufficient faunal similarity to represent the same stratigraphic entity yet could be easily dis- tinguished lithologically. Thus the concept of subdividing the Aptian stage originated. In 1887 Kilian introduced the term Gargasian for the ‘Aptian marls’, and in the following year Toucas (1888) defined the Bedoulian for the underlying passage beds. The positioning of the Gargasian-Bedoulian boundary resulted in considerable controversy but, suffice it to say, these terms have survived as substages and refer respectively to the Late and Early Aptian. DAVEY AND VERDIER: APTIAN DINOFLAGELLATE CYSTS 625 STAGE SUB-STAGE AMMONITE ZONES CLANSAYESSAN Diadochoceras nodosocostatum < LATE GARGASIAN Che/oniceras suhnodosocostatum r— CL Aconeceras nisus EARLY BEDOULIAN Deshayesites deshayesi text-fig. 2. Stratigraphic subdivision of the Aptian. La Bedoule Section Sediments of latest Barremian to Early Gargasian age crop out in several active and abandoned quarries on both sides of the Aubagne-Cassis road near the village of La Bedoule. The Aptian succession (text-fig. 3) here is highly calcareous and consists of alternating beds of marls and argillaceous limestones. The latter become increas- ingly abundant towards the base of the Bedoulian and grade into the massive Urgonian Limestone of probable Barremian age. The Urgonian Limestone of the Vocontian Trough represents deposition in a relatively shallow water, quiet environment. The initiation of Aptian deposition corresponds to the beginning of a progressive deepen- ing of the sedimentary basin, to an increasing influx of argillaceous material and to less well-oxygenated bottom conditions. A total of ten samples of Bedoulian and Early Gargasian age were collected from the marl and clay beds at this locality. Gargas Section In the Apt region, the Aptian crops out in a series of exposures between the towns of Apt and Gargas. The top of the Urgonian limestones (text-fig. 4), here of Bedoulian age, is exposed near the railway bridge about two kilometres north-west of Apt. As 626 PALAEONTOLOGY, VOLUME 17 AGE GARGASIAN BEDOULIAN BARREMIAN FORMATION e Couches superieures 0 de la Carri ere Comte c Couches inferieures o de /a Carri ere Comte Ensemble . mar no - ca/caire peu compact Ensemb/e sil/ceux Ensemb/e ca/careo -marneux ( Couches de /a Carri ere a ciment) Couches de passage URGONIEN text-fig. 3. La Bedoule section showing the lithologic subdivisions and the position of samples (after Fabre-Taxy et al. 1965). the hillside is ascended towards Gargas, progressively younger sediments are en- countered, and the hill is capped by Albian marls and sands. A total of seven samples were collected and range in age from Bedoulian to Clansayesian. SYSTEMATIC DESCRIPTIONS This section is divided into two parts. The first part lists, in alphabetical order, the dinoflagellate cyst species which require no special remarks and have been previously described, with full stratigraphic annotations, by Davey and Verdier (1971, 1973). The second part deals with the cyst species recovered during the present study which were either not described in the above publications or require certain amplifying remarks. The species are arranged in alphabetical order within the Gonyaulacacean and Peridiniacean groups. DAVEY AND VERDIER: APTIAN DINOFLAGELLATE CYSTS 627 AGE FORMATION LITHOLOGY SAMPLE 7 Sab/es bar/o/es AIBIAN CLANSAYESIAN g Marnes sab /eases ^ Marries sab/euses eb -457 ^ Gres marneux -456 1 4 C 0 0 m. GARGASSAN 4 Marnes gris - bleu A tZ A _45l 3 Marno-ca/caire b/eube =i 1 r- i it-r= - 453 BEBOULIAN 2 Marnes jaunes H URGONIEN 452 text-fig. 4. Gargas section showing the lithologic subdivisions and the position of samples (after Moullade 1965). PART 1 Apteodinium granulatum Eisenack 1958. Astrocysta cretacea (Pocock 1962) Davey 1970. (PI. 93, fig. 4.) ‘ Broomea ’ micropoda Eisenack and Cookson 1960. Callaiosphaeridium asymmetricum (Deflandre and Courteville 1939) Davey and Williams 1966. Canningia colliveri Cookson and Eisenack 1960. C. minor Cookson and Elughes 1964. Cassiculosphaeridia reticulata Davey 1969. Cauca parva (Alberti 1961) Davey and Verdier 1971. Cleistosphaeridium armatum (Deflandre 1937) Davey 1969. C. huguonioti (Valensi 1955) Davey 1969. C. polypes clavulum Davey 1969. C. polypes polypes (Cookson and Eisenack 1962) Davey 1969. Cribroperidinium edwardsi (Cookson and Eisenack 1958) Davey 1969. Cyclonephelium distinctum Deflandre and Cookson 1955. 628 PALAEONTOLOGY, VOLUME 17 Exochosphaeridium arnace Davey and Verdier 1973. E. phragmites Davey, Downie, Sarjeant and Williams 1966. Florentinia laciniata Davey and Verdier 1973. F. mante/Ii (Davey and Williams 1966) Davey and Verdier 1973. Fromea amphora Cookson and Eisenack 1958. Gonyaulacysta cassidata (Eisenack and Cookson 1960) Sarjeant 1966. G. helicoidea (Eisenack and Cookson 1960) Sarjeant 1969. G. tenuiceras (Eisenack 1958) Sarjeant 1969. Hystrichodinium pulchrum Deflandre 1935. Hystrichosphaeridium arundum Eisenack and Cookson 1960. H. recurvatum (White 1842) Davey and Williams 1966. H. tubiferum (Ehrenberg 1838) Davey and Williams 1966. Kalyptea sp. (as in Davey and Verdier 1971). Microdinium crinitum Davey 1969. Odontochitina operculata (O. Wetzel 1933) Deflandre 1946. Oligosphaeridium complex (White 1842) Davey and Williams 1966. Ovoidinium scabrosum (Cookson and Hughes 1964) Davey 1970. Polysphaeridium laminaspinosum Davey and Williams 1966. Protoellipsodinium spinocristatum Davey and Verdier 1971. P. spinosum Davey and Verdier 1971. Spiniferites cingulatus cingulatus (O. Wetzel 1933) Sarjeant 1970. S. ramosus multibrevis (Davey and Williams 1966) Sarjeant 1970. S. ramosus ramosus ( Ehrenberg 1838) Sarjeant 1970. S. ramosus reticulatus (Davey and Williams 1966) Sarjeant 1970. Tanyosphaeridium variecalamum Davey and Williams 1966. Trichodinium castanea (Deflandre 1935) Clarke and Verdier 1967. PART 2 Class dinophyceae Pascher Order peridiniales Lindemann GONYAULACACEAN Group Genus achomosphaera Evitt 1963 Achomosphaera neptuni (Eisenack) Davey and Williams 1966 1958 Baltisphaeridium neptuni Eisenack, p. 51, pi. 26, figs. 7, 8. Reported occurrence. Valanginian to Hauterivian, Germany (Gocht 1959). Hauterivian, Switzerland (Millioud 1967, 1969). Barremian, England (Davey 1974 in press). Upper Aptian, Germany (Eisenack 1958). Achomosphaera cf. neptuni (Eisenack) Davey and Williams 1966 Plate 92, fig. 2 Description. This ovoidal cyst has a smooth to minutely granular, thin wall which bears many smooth to slightly fibrous processes. Each process has a wide flat base and tapers rapidly to become thin and parallel-sided. Towards the distal extremity they divide into two, rarely three, filamentous branches which occasionally can be seen to further bifurcate at their distal extremities. Some alignment of the processes is present and appears to mark the cingular margins; a short, stouter apical process is sometimes discernible. These two features allow cyst orientation which suggests that the archaeopyle is precingular in position. It is formed by the loss of a single plate and is roughly triangular in shape. The processes are apparently both gonal and sutural in position. DAVEY AND VERDIER: APTIAN DINOFLAGELLATE CYSTS 629 Dimensions. Figured specimen Range Central body diameter Length of processes 0*m) 44x51 7-15 (f*m) 41-51 7-15 Remarks. A. neptuni differs from A. cf. neptuni by being thicker walled with a coarser granulation and by having fewer and wider processes which often join proximally with neighbouring processes. The processes of A. neptuni appear to be only gonal and usually trifurcate distally to give thick spines. 1958 Chlamydophorella nyei Cookson and Eiseneck, p. 56, pi. 11, figs. 1-3. Remarks. C. nyei is distinguished from Gardodinium trabeculosum (Gocht 1959) by being more or less rounded in outline (except for the apical prominence) and by not possessing a tabulation and process alignment. Reported occurrence. Barremian, England (Davey 1974 in press, as G. cf. trabeculosum). Aptian to Lower Turonian, Australia (Cookson and Eisenack 1958, 1962fi, 1971 ). Albian to Cenomanian, England (Cookson and Flughes 1964); Canada (Manum and Cookson 1964; Davey 1970; Cox 1971; Brideaux 1971u, b\ Singh 1971). 1958 Coronifera oceanica Cookson and Eisenack, p. 45, pi. 12, figs. 5, 6. Remarks. C. oceanica occurs in most of the Aptian samples but is never abundant and, being thin-walled, is usually distorted. The range of morphological variation exhibited by this species is very great, and specimens approaching C. albertii Millioud 1969 are present. The variation can be summarized as follows: the cyst wall may be smooth, granular, or pseudo-reticulate; the processes may be simple or furcate, briefly or deeply, and distally may be acuminate, capitate, or knobbed; the apical process, when distinctive, is trifurcate and may be situated on an apical boss; the antapical process is cylindrical but varies considerably in size and denticulation of the distal margin. A precingular archaeopyle is always present. Reported occurrence. Upper Hauterivian to Upper Albian, France (Millioud 1969; Davey and Verdier 1971, 1973). Middle Barremian to basal Coniacian, England (Cookson and Flughes 1964; Clarke and Verdier 1967; Davey 1969, 1974 in press). Albian, Cenomanian, Santonian/lowest Campanian, Australia (Cookson and Eisenack 1958, 1968, 1969). 1962 Cribroperidinium sepimentum Neale and Sarjeant, p. 443, pi. 19, fig. 4. Remarks. The present specimens from the Bedoulian closely resemble the type material. However, it is difficult to discern whether the cyst wall is densely micro- perforate or microgranulate. Genus chlamydophorella Cookson and Eisenack 1958 Chlamydophorella nyei Cookson and Eisenack 1958 Genus coronifera Cookson and Eisenack 1958 Coronifera oceanica Cookson and Eisenack 1958 Genus cribroperidinium Neale and Sarjeant emend. Davey 1969 Cribroperidinium sepimentum Neale and Sarjeant 1962 Plate 91, fig. 5 630 PALAEONTOLOGY, VOLUME 17 Reported occurrence. Middle Hauterivian to Lower Barremian, England (Neale and Sarjeant 1962; Sarjeant 19666; Davey 1974 in press). Genus cyclonephelium Deflandre and Cookson emend. Cookson and Eisenack 1962 Cyclonephelium tabulatum sp. nov. Plate 92, figs. 1,4; Plate 93, fig. 6 Derivation of name. Latin, tabulatus, plated or tabulate— with reference to the distinctive tabulation. Diagnosis. The cyst is subcircular in outline and possesses a thin, smooth to lightly tuberculate wall. The processes are predominantly peritabular in position and clearly define the precingular and postcingular plates and the cingulum. The central part of the plates and the sulcal region are practically devoid of processes. The processes are short, stout, and capitate; they rarely branch distally or are joined proximally. The archaeopyle is apical and has a strongly zigzag margin and a sulcal notch; the oper- culum is typically detached. Holotype. Slide FR 443/2, La Bedoule, south-east France; Upper Aptian (Gargasian). Dimensions. Cyst height (without operculum) Cyst width Length of processes Holotype Range (pm) (pm) 57 52 (56) 60 72 59 (66) 72 4-8 2 (6) 9 Description. Since all the identified specimens possessed an apical archaeopyle and only rarely detached opercula were identified, it is difficult to precisely define the shape of the apical region. It is surmised, however, that it is of similar shape to that found in related species. That is, the cyst is rounded apically or has a reduced apical boss. The main part of the cyst is subcircular, or rarely slightly angular, in outline. The processes are neatly aligned just within the plate margins and hence clearly define the tabulation in the precingular and postcingular regions. In each case five or six plates appear to be present. The parallel lines of processes, either side of a plate boundary, EXPLANATION OF PLATE 91 Fig. 1. Aptea polymorpha Eisenack 1958. Complete specimen. Slide Gargas FR 456/1, x 450 ph. Figs. 2, 3. Aptea securigera sp. nov. 2, paratype illustrating apical archaeopyle, x 400 ph. 3, holotype, complete specimen, x 400. Figs. 4, 8. Muderongia cf. staurota Sarjeant 1966. 4, specimen illustrating broad, centrally placed antapical horn. Slide La Bedoule FR 448/2, x 700 ph. 8, specimen illustrating reduced lateral horns. Slide La Bedoule FR 448/2, x 700 ph. Fig. 5. Cribroperidinium sepimentum Neale and Sarjeant 1962. Archaeopyle to the north-west. Slide La Bedoule FR 446/1, x 600. Fig. 6. Dingodinium albertii Sarjeant 1966. Apical operculum partially detached. Slide Gargas FR 454/1 A, x 700 ph. Figs. 7, 9. Spiniferites sp. 7, ventral view illustrating sulcus and displaced cingulum. Slide Gargas FR 451/2, x650. 9, ventral view illustrating sutural spines and ‘apical’ horn. Slide Gargas FR 455/1A, x 650 ph. PLATE 91 DAVEY and VERDIER, Aptian dinoflagellate cysts 632 PALAEONTOLOGY, VOLUME 17 are 3 to 4 p.m apart. Rare processes on the cingulum may indicate the position of plate boundaries; a distinct cingular tabulation, however, is not present. Two to three antapical and four apical plates appear to be present. Remarks. The shape of the cyst and its processes are identical to that of C. distinctum Deflandre and Cookson 1955. C. tabulatum differs significantly from C. distinctum , and other members of Cyclonephelium , by the possession of peritabular processes and an encircling cingulum; the processes of C. distinctum are concentrated towards the circumferential region. The two species are clearly closely related, and although C. tabulatum does not conform exactly with the generic diagnosis of Cyclonephelium it is considered best placed in this genus at present. Genus dingodinium Cookson and Eisenack 1958 Dingodinium albertii Sarjeant 1966 Plate 91, fig. 6 1966 b Dingodinium albertii Sarjeant, p. 210, pi. 21, fig. 3; pi. 23, fig. 1. Remarks. D. albertii is distinguished from D. cervicu/um Cookson and Eisenack 1958 by its considerably smaller size. At the moment the latter species appears to be restricted to Australia. The specimens described by Brideaux 19716 as D. cerviculum do, however, approach the type material in size but are here considered still to fall within the range of D. albertii. Reported occurrence. Upper Hauterivian to Upper Barremian, France (Millioud 1969). Barremian, England (Sarjeant 19666; Davey 1974 in press). Upper Barremian, Germany (Alberti 1961). Albian, Canada (as D. cerviculum ; Brideaux 19716; Singh 1971). Genus gardodinium Alberti 1961 Gardodinium trabeculosum (Gocht) Alberti 1961 1959 Scriniodinium trabeculosum Gocht, p. 62, pi. 4, fig. 5; pi. 8, fig. 2. Reported occurrence. Lower Hauterivian to Lower Aptian, Lrance (Millioud 1967). Lower Hauterivian to Upper Aptian, Germany (Gocht 1959; Alberti 1961). Middle Hauterivian to Upper Barremian, England (Sarjeant 19666; Davey 1974 in press). Genus gonyaulacysta Deflandre emend. Sarjeant 1969 Gonyaulacysta sp. Plate 93, fig. 5 1958 Gen. et sp. indet. (ex. aff. WanaeaT) Eisenack, p. 398, pi. 25, fig. 2. Description. Two well-preserved examples of this distinctive species were found and allow a relatively complete description to be given. The cysts are flattened and orientated such that the view is apical-antapical. Equatorially the cyst is subcircular (50-57 jam diameter) in outline and is surrounded by a distinctive flange 7-12 ^.m in height which is indented at the sulcus. This flange is formed by two expansions of the periphragm along the cingulum; these run parallel to the two cingular margins with the expansion on the apical side being less than the antapical one. Distally the expansions may be irregular and may bear conical or small bifid spines (1-2 in height). The apical region is broadly conical (20 at the base) and terminates with DAVEY AND VERDIER: APTIAN DINOFLAGELLATE CYSTS 633 a short, blunt process. The cyst surface is microgranulate and bears low, smooth ridges which appear to give a Gonyaulacysta- type tabulation. The archaeopyle is formed by the loss of one or two dorsal precingular plates. Remarks. The present Gargasian specimens appear to be identical to Eisenack’s form and are assigned to the genus Gonyaulacysta. Similar forms also have been described and figured as Dinopterygium sp. A by Brideaux (19716, p. 97, pi. 28, figs. 89, 92) from the Albian of Canada. However, because of the unfavourable orientation of our specimens, they are impossible to describe completely and compare thoroughly with previously described species. The microgranulate ornamentation, the type of apical horn and sutural ridges, and the distinctive cingular periphragm expansions ade- quately differentiate this species from all known forms. Wanea Cookson and Eisenack 1958, from the Middle and Upper Jurassic, appears somewhat similar but differs significantly by the possession of an epitractal archaeopyle. Reported occurrence. Upper Aptian, Germany (Eisenack 1958). Albian?, Canada (Brideaux 1971b). Genus kleithriasphaeridium Davey 1974 Kleithriasphaeridium simplicispinum (Davey and Williams) Davey 1974 1966b Hystrichosphaeridium simplicispinum Davey and Williams, p. 59, pi. 9, fig. 3. Remarks. K. simplicispinum is extremely uncommon in the type Aptian and was only recorded a single time in each of four samples. Reported occurrence. Valanginian to Hauterivian (Aptian?), Germany (Gocht 1959). Upper Hauterivian to Upper Barremian, England (Davey and Williams 1966b; Sarjeant 1966b; Davey 1974 in press). Genus meiourogonyaulax Sarjeant 1966 Meiourogonyaulax cf. bulloidea (Cookson and Eisenack) Sarjeant 1969 Plate 92, fig. 5 1960b Gonyaulax bidloidea Cookson and Eisenack, p. 247, pi. 37, fig. 11. Dimensions. (Single specimen.) Shell length 55 /xm, shell width 60 ^m, height of crests less than 2 /xm. Remarks. The single specimen of M. cf. bulloidea found during the present study differs from M. bulloidea , from the Portlandian of Australia, by the nature of its sutural crests. In M. bulloidea these are low, entire and granular; in M. cf. bulloidea they are smooth distally and composed of low, broad spines which widen and often join distally. Very similar specimens were described as M. bulloidea by Davey (1974 in press) from the Barremian of England. Meiourogonyaulax stoveri Millioud 1969 Plate 93, figs. 2, 8 1969 Meiourogonyaulax stoveri Millioud, p. 429, pi. 3, figs. 1-3. Dimensions. Length of shell (complete specimens) 64 (72) 83 /xm, length of shell (with archaeopyle) 71- 73 /xm, width of shell 61 (73) 81 /xm, maximum height of crests 7-12 /xm. Remarks. The present Aptian specimens resemble the type material in all respects. M. stoveri is characterized by its more or less circular outline, its thick (up to 3 /xm), N 634 PALAEONTOLOGY, VOLUME 17 perforate wall, and by its membranous crests which are often perforate and most strongly developed in the antapical region. Reported occurrence. Lower Hauterivian, Switzerland; Upper Hauterivian, Barremian and Lower Aptian, France (Millioud 1969). Early and Middle Albian, France (Davey and Verdier 1971, questionable attri- bution). Meiourogonyaulax p soros sp. nov. Plate 92, figs. 8, 9 Derivation of name. Greek, psoros, scabby or mangy— with reference to the irregularly tubercled surface of the cyst wall. Diagnosis. The cyst is subcircular to slightly angular in outline. The cyst wall is rela- tively thick and is characterized by a variable surface ornamentation consisting of isolated tubercles to low irregular vermicular ridges. The sutural ridges are low and are difficult to discern except at the margins of the cyst; they are formed by the coalescence of aligned tubercles and/or broad, flat-topped processes. The cingulum, which is narrow, is displaced by approximately one cingular width along the sulcus. The latter broadens on the hypotract and has a noticeable deep, centrally placed longitudinal groove. The apical archaeopyle is angular and the operculum sometimes remains attached. Holotype. Slide FR 441/2, La Bedoule, south-east France; Lower Aptian (Bedoulian). Paratype. Slide FR 444/2, La Bedoule, south-east France; Lower Aptian (Bedoulian). Holotype Paratype Range w (f*m) (Mm) Cyst length (complete specimens) 69 59-69 Cyst length (specimens with archaeopyle) 51 51 (55) 60 Cyst width 66 64 52 (60) 68 Description. On one specimen a low (5 ^m), conical, membranous apical horn is present. The remaining complete specimens only possess slight apical prominences which are indistinguishable from the other angularities of the cyst outline. The thick (up to 3 jum) cyst wall is densely warty and rugulate. The sutural ridges, which are often discontinuous, are of similar height (up to 5 ^.m) to this ornamentation and thus are difficult to locate. Only on the cyst margin is it possible to see that the ridges EXPLANATION OF PLATE 92 Figs. 1,4. Cyclonepheliumtabulatumsp. nov. Holotype. 1, dorsal view, x600ph. 4, ventral view, x600ph. Fig. 2. Achomosphaera cf. neptuni (Eisenack 1958). Precingular archaeopyle is visible just beneath the apex. Slide La Bedoule FR 442/2, x 500 ph. Figs. 3, 6. Meiourogonyaulax sp. Ventral views. Note the deeply incised sulcal notch, the insignificant crests on the ventral surface except for the cingulum and the high lateral crests. 3, slide La Bedoule FR 447/2, x650. 6, slide La Bedoule FR 442/1, x 650. Fig. 5. Meiourogonyaulax cf. bulloidea (Cookson and Eisenack 1960). Slide Gargas FR 454/1 A, x 650. Fig. 7. Trichodinium sp. Archaeopyle view; note long, fine spines. Slide La Bedoule FR 443/1, x 1000. Figs. 8, 9. Meiourogonyaulax psoros sp. nov. 8, paratype, ventral view with apical archaeopyle developed, x 550. 9, holotype, complete specimen with operculum partially detached, x 550. PLATE 92 DAVEY and VERDIER, Aptian dinoflagellate cysts 636 PALAEONTOLOGY, VOLUME 17 are sometimes composed of long broad processes which anastomose distally. The cingulum is narrow (3-5 p m in width), does not appear to be tabulate, and sometimes forms ledges on the cyst circumference. Remarks. M. psoros sp. nov. is easily distinguished from all other species of Meiouro- gonyaulax by its distinctive rugulate ornamentation and low sutural crests. Meiourogonyaulax sp. Plate 92, figs. 3, 6 Description. The cyst is subcircular in outline and has a wall of moderate thickness (approximately 1 p m) which is smooth to lightly pitted. The tabulation is distinctive and is marked by high crests around the lateral margins of the cyst and by low crests on the ventral and dorsal surfaces. The crests are smooth, rarely perforate, with a smooth distal margin and may develop pericoels on the lateral sides of the hypo- tract. The tabulation appears to be ?4', 6", ?6c, 5-6'", lp, l"". The cingular plate crests are almost absent on the dorsal and ventral surfaces. Plate V" is narrow and poorly defined and is practically part of the sulcus. The sulcus is wedge-shaped, widening antapically, with a deeply indented central groove; sulcal plates are absent. The apical archaeopyle is slightly angular with a deeply indented sulcal notch. Figured specimens. Plate 92, fig. 3. Central body length 58 ^m, width 55 ^m, maximum height of crests 10 (im. Plate 92, fig. 6. Central body length 53 ^m, width 51 gm, maximum height of crests 9 /xm. Remarks. The present specimens, only two were observed, strongly resemble M. valensii Sarjeant 1966a. M. valensii differs from M. sp. by the possession of a more evenly punctate wall and striate crests which are sometimes finely denticulate distally. M. stoveri differs in being larger, thicker walled, and having less uniformly developed lateral crests. Genus oligosphaeridium Davey and Williams 1966 Oligosphaeridium nannum Davey 1974 1974 Oligosphaeridium nannum Davey, pi. 4, figs. 9, 10 (in press). Remarks. This species, which has only been reported from the Lower Barremian (Davey 1974 in press), was represented by a single specimen in one sample (FR 454) and may be reworked. Very rare (3) examples have been observed in the Albian of the Paris Basin and are also probably reworked. Reported occurrence. Lower Barremian, England (Davey 1974 in press). Genus prolixosphaeridium Davey, Downie, Sarjeant and Williams 1966 Remarks. In the discussion below it is considered that P. deirense Davey et al. 1966 is a junior synonym of P. parvispinum (Deflandre 1937c) Davey et al. 1969. Hence, the latter species now becomes the type species of this genus. Type species. Prolixosphaeridium parvispinum (Deflandre 1937c) Davey, Downie, Sarjeant and Williams 1969. Lower Cretaceous (Aptian), France. DAVEY AND VERDIER: APTIAN DINOFLAGELLATE CYSTS 637 Prolixosphaeridium parvispinum (Deflandre) Davey et al. 1969 1937c Hystrichosphaeridium xanthiopyxides var. parvispinum Deflandre, p. 29, pi. 16, fig. 5. 1958 Hystrichosphaeridium parvispinum Deflandre; Cookson and Eisenack, p. 45. 1960 Baltisphaeridium parvispinum (Deflandre) Klement, p. 59. 1966 Prolixosphaeridium deirense Davey el al., p. 171, pi. 3, fig. 2; text-fig. 45. 1969 Prolixosphaeridium parvispinum (Deflandre) Davey et al., p. 17. Description. The present specimens of P. parvispinum are identical with P. deirense as described by Davey et al. 1966. It should be noted, however, that firstly, the two antapical processes of this latter form are not always distinctive; and secondly, that the basal portion of the larger processes may be perforate. Remarks. P. parvispinum is undoubtedly identical to P. deirense. This relationship has been overlooked previously mainly because Deflandre’s Aptian species was recorded in a publication dealing primarily with Late Cretaceous flints. Reported occurrence. Uppermost Lower Barremian to Lower Aptian, Lrance (Millioud 1969). Middle to Upper Barremian, England (Davey et at. 1966, Davey 1974 in press). Aptian, Prance (Deflandre 1937c). Albian, Prance (Davey and Verdier 1971, 1973). Genus protoellipsodinium Davey and Verdier 1971 Protoellipsodinium clavulum sp. nov. Plate 93, fig. 7 Derivation of name. Latin, clavus , nail — with reference to the shape of the processes. Diagnosis. The cyst is elongate to ovoidal with a smooth wall. The cingulum usually lacks processes, and the hypotract is larger than epitract. The processes are fairly numerous and less than half the cyst width in length. They are typically hollow, with a restricted lumen, and distally are capitate or rarely bear two or three small spines. The archaeopyle is precingular, formed by the loss of a single plate. Holotype. Slide PR 454/2, Gargas, south-east Prance; Upper Aptian (Gargasian). Dimensions. Central body length Central body width Length of processes Holotype Range (pm) (pm) 40 40 (42) 45 23 23 (28) 32 4 11 4(10)13 Description. The cyst is thin-walled and often distorted, hence making orientation and archaeopyle identification difficult. The shape and structure of the processes are, however, very characteristic. Each process has a relatively broad base (2-3 pm wide), narrows rapidly above this, and for the more distal part of its length tapers only slightly and is more or less parallel sided (width approx. 1 p m). The processes expand at their distal extremities and are basically capitate; this expansion is usually slight (approx. 1 pm in width) but occasionally is wider and gives rise to two or three re- curved spines (up to 2-5 pm long). The processes are rigid to slightly flexuous. Remarks. The distal extremities of the processes distinguish P. clavulum sp. nov. from the two other species in this genus. The form of the processes is very similar to that 638 PALAEONTOLOGY, VOLUME 17 found in Cleistosphaeridium huguonioti (Valensi) var. pertusum Davey 1969 from the Upper Cenomanian of southern England and northern France. That species, how- ever, is spherical to subspherical in shape and has an apical archaeopyle. Genus spiniferites Mantell emend. Sarjeant 1970 Spiniferites sp. Plate 91, figs. 7, 9 Description. The specimens assigned to Spiniferites sp. are ovoidal in outline, smooth- walled, and have a distinct tabulation marked by membranous, sutural crests. The latter are highest at the plate corners where two or three crests meet and are especially noticeable in the apical region but do not appear to be best developed exactly at the apex. Between these gonal points the crests are lower (1-3 /xm) and distally may vary from being slightly irregular to bearing strong, regularly or irregularly distributed, sutural spines up to 3 in height. In certain specimens, and in varying positions, the periphragm is detached from the endophragm, producing a bladder-like expan- sion reminiscent of Lejeune-Carpentier’s (1937c and 1938a) specimens of Spiniferites ramosus (Ehrenberg 1838). Dimensions. Length of central body 41-52 ; width 31-45 ; height of crestal spines up to 3 ^m ; height of gonal elevations up to 7 p.m. Remarks. Spiniferites sp. is not an easy species to classify; with only slight morpho- logical variation, and such variations occur in the present specimens, this species can be assigned to any of three genera. The disappearance of sutural spines leads to a Leptodinium Klement or Spiniferites cingulatus (O. Wetzel 1933) assignation. The development of an apical expansion of the periphragm, or apical horn, leads to affinities with the Gonyaulacysta cretacea/helicoidea complex. The typical speci- men of Spiniferites sp., however, does appear to represent a spiny form of S. cingulatus. EXPLANATION OF PLATE 93 Fig. 1 . Pareodinia cf. aceras (Manum and Cookson 1964) comb. nov. Specimen illustrating surface reticula- tion and attached opercular plates. Slide Gargas FR 455/2, x 1000 ph. Figs. 2, 8. Meiourogonyaulax stoveri Millioud 1969. Slide Gargas FR 454/1. 2, dorsal view illustrating perforate antapical crest, x 800. 8, ventral view illustrating vacuolar wall, x 800 ph. Fig. 3. Deflandrea terrula Davey 1974 (in press). Dorsal view illustrating tabulation and intratabular granulation. Slide La Bedoule FR 441/1. x600 ph. Fig. 4. Astrocysta cretacea (Pocock 1962). Specimen illustrating quasi-tabular ridges on the dorsal hypo- tractal surface. Slide Gargas FR 454/1A, x 800 ph. Fig. 5. Gonyaulacysta sp. Apical view illustrating large precingular archaeopyle and two cingular flanges. Slide Gargas FR 455/1 A, x 650 ph. Fig. 6. Cyclonephelium tabulation sp. nov. Apical operculum illustrating peritabular processes outlining four apical plates. Slide La Bedoule FR 443/2, x 1000 ph. Fig. 7. Protoellipsodinium clavulum sp. nov. Holotype illustrating bald cingular region, x 900 ph. Fig. 9. Pareodinia sp. Partially detached opercular plates are present just beneath the cyst apex. Slide Gargas FR 457/1 B, x 800 ph. PLATE 93 DAVEY and VERDIER, Aptian dinoflagellate cysts 640 PALAEONTOLOGY, VOLUME 17 Genus systematophora Klement 1 960 Systematophora schindewolfi (Alberti) Sarjeant 1966 1 958 Hystrichosphaeridium anthophorum Cookson and Eisenack ; Eisenack, p. 402, pi. 26, figs. 1 , 2. 1961 Hystrichosphaerina schindewolfi Alberti, p. 38, pi. 10, figs. 1-3, 6, 7. 1961c Hystrichosphaeridium sp. 4 Evitt, p. 398, pi. 4, figs. 4, 5. 19666 Systematophora schindewolfi (Alberti) Sarjeant, p. 209, pi. 22, fig. 5. 1969 Perisseiasphaeridium eisenackii Davey and Williams, p. 6. 1974 Perisseiasphaeridium eisenackii Davey and Williams; Davey, pi. 6, fig. 5 (in press). Remarks. Alberti’s type specimen of S. schindewolfi comes from the Pirna borehole of Germany and was originally dated as Turonian; he also recorded specimens from the Upper Barremian of Salzgitter, Germany. Sarjeant (19666) recognized this species in the Barremian of England and validly transferred it to Systematophora. Earlier, however, Evitt (1961c) had redescribed Eisenack’s 1958 specimens of Hystricho- sphaeridium anthophorum , and it is now obvious that the latter specimens belong to S. schindewolfi. Davey and Williams (1969), however, overlooked this relationship and erected a new species, Perisseiasphaeridium eisenackii , using Eisenack’s speci- mens; this species is now considered to be a junior synonym of S. schindewolfi. An important stratigraphic point is that the dating of the Pirna borehole material has long been contested since seven of the reported fifteen species are of Aptian or older age; the oldest forms reported cannot be younger than Barremian. It thus appears probable that the type material of S. schindewolfi from the Pirna borehole is of Barremo-Aptian age rather than Turonian. Reported occurrence. Middle to Upper Barremian, England (Sarjeant 19666; Davey 1974 in press). Upper Barremian to Upper Aptian, Germany (Eisenack 1958; Alberti 1961). Genus trichodinium Eisenack and Cookson emend. Clarke and Verdier 1967 Trichodinium sp. Plate 92, fig. 7 Description. The rare specimens placed in Trichodinium sp. differ from T. castanea (Deflandre 1935a) in possessing finer and longer spines. An apical structure is not present. The cingulum and, more rarely, other sutural boundaries are marked by low thickenings of the shell wall. Dimensions. Shell length 47-61 pm, width 44-59 ,um, maximum length of spines 6-8 ^m. PERIDINIACEAN Group Genus aptea Eisenack 1958 emend. Emended diagnosis. Dorso-ventrally flattened cysts with typically a rounded triangular outline and possessing an ornamentation of membranous crests and/or processes which is better developed in the circumferential region. The apices of the triangle are situated at the apex, antapex, and a little antapically to the right cingular margin of the cyst; they are typically marked by distinctively high ornamentation, and the cyst wall may or may not have prominent rounded protuberances in these positions. A detached inner body, elongate horns, and a flat or indented (two horns) antapical DAVEY AND VERDIER: APTIAN DINOFLAGELL ATE CYSTS 641 region are never present. The crests and processes are intratabular and rarely show alignment parallel to plate boundaries. The archaeopyle is apical with a strongly zigzag margin and short breakages extending along the precingular plate boundaries. The sulcal notch is always offset from the mid-line of the cyst’s ventral surface. Finally, the operculum often remains attached. Type species. Aptea polymorpha Eisenack 1958, p. 394, pi. 22, figs. 5-12. Upper Aptian, Germany. Remarks. The morphological structure of the genus Aptea is here described in detail so as to distinguish it from morphologically similar genera. Particular stress is placed on the typical and characteristic asymmetry of the cyst which, it is considered, dis- tinguishes it from previously described and possibly related genera such as Cyclone- phelium Deflandre and Cookson 1955, Canningia Cookson and Eisenack 19606, and Tenua Eisenack 1958. These three genera have either a rounded or indented antapex and if protuberances or bulges are present in the cingular region then they are more or less symmetrically placed (text-fig. 5). Aptea probably evolved directly from Pseudoceratium Gocht 1957 by the dis- appearance of the elongate horns, while still retaining the characteristic symmetry of this genus (text-fig. 5, vm); for this reason Aptea is here considered to belong to the Pseudoceratioid branch of the Peridiniacean Group. This morphological change is apparently rapid and the first specimens assignable to Aptea appear in the topmost Barremian (Davey 1974 in press). However, this genus did not become a common constituent of the dinoflagellate cyst flora until Aptian time, only to disappear by the end of the Albian. Doidyx Sarjeant 19666 (see text-fig. 5, ix) is related to Aptea and is generically difficult to distinguish. However, D. anaphrissa Sarjeant 19666, which has a Lower to Middle Barremian range (Davey 1974 in press), is the only member of this genus and is relatively easy to distinguish from similar species. For this reason it appears unnecessary to synonymize at present Doidyx with Aptea. Aptea polymorpha Eisenack 1958 Plate 91, fig. 1 ; text-fig. 5 (v, vi) 1958 Aptea polymorpha Eisenack, p. 394, pi. 22, figs. 5-12. 1960 A. cf. polymorpha Eisenack; Eisenack and Cookson, p. 9, pi. 3, fig. 2 only. 1971 A. polymorpha Eisenack; Singh, p. 370, pi. 63, figs. 5-7; pi. 64, fig. 1. Remarks. The present Tethyan specimens differ only in detail from the Boreal type material. The most noticeable differences are that our specimens are thinner-walled, have weaker and less continuous crests, and do not stain with safranin. These charac- teristics, we consider, are primarily preservational features and are not specifically significant. The Tethyan specimens, however, are generally more angular than the type material and sometimes possess relatively strong apical and antapical pro- tuberances. Aptea eisenacki (Davey 1969) comb. nov. is more similar to the present examples of A. polymorpha than to the type material and may be distinguished only by the extremely low height of the crests. Reported occurrence. Upper Aptian, Germany (Eisenack 1958). Albian?, Canada (Singh 1971). text-fig. 5. Shape variation of certain related cysts (apex and ornamentation removed), i-iv, examples from the genera Canningia, Cyclonephelium, and Tenua. v and vi, Aptea polymorpha with ornamentation outline shown; v, from Eisenack’s 1958 type material and vi, from the present French material, vn, A. securigera. Vlli, Pseudo ceratium pelliferum. IX, Doidyx anaphrissa. Aptea securigera sp. nov. Plate 91, figs. 2, 3; text-fig. 5 (vii) Derivation of name. Latin, securiger, axe-bearing— with reference to the axe-shaped terminations of the processes. Diagnosis. The cyst central body is dorso-ventrally flattened and rounded triangular in shape. The left side is strongly, but evenly, convex; the right epitractal and hypo- tractal sides are slightly convex to straight and meet at approximately right angles in the cingular region. The right hypotractal side often has a medial convexity. The apex and, to a lesser extent, the antapex are developed as protuberances of the central DAVEY AND VERDIER: APTIAN DINOFLAGELLATE CYSTS 643 body and are rounded distally. The cyst surface bears numerous short, flattened, solid processes which are concentrated in the circumferential region. A more or less circular area in the centre of the ventral and dorsal surfaces is devoid of, or possesses only rare, processes. The processes are of variable shape but are typically discrete, expanding both distally and proximally, and are flat-topped distally; their length is more than twice their medial width. The processes are longer and more variable at the cyst apices. Very rarely the cingulum and other tabulation is marked by narrow bands devoid of processes. The archaeopyle is apical and possesses a strongly zigzag margin. The operculum is usually detached. Holotype. Slide FR 446/1, La Bedoule, south-east France; Lower Aptian (Bedoulian). Paratype. Slide FR 446/2, La Bedoule, south-east France, Lower Aptian (Bedoulian). Holotype Paratype Range Cm) Cm) Cm) Central body length 73 73-90 Central body width 69 87 69 (75) 87 Central body length (operculum detached) 73 62 (67) 73 Height of processes 2-6 2-5 2 (3) 10 Description. The processes characteristically widen both distally and proximally. Very rarely, narrow processes may bifurcate or trifurcate, or neighbouring processes may be linked medially by a crest. Rarely the bases of two or three neighbouring processes may be linked by a low ridge or thickening of the cyst wall. These thicken- ings tend to parallel the cyst sides, as do the crests in A.polymorpha. Besides the longer processes at the cyst apices, slightly longer ones may also be present on the left side in the cingular region and on the convexity or bulge, if present, of the right hypo- tractal side. These longer processes were seen to be linked distally in one specimen. Remarks. A. securigera sp. nov. is differentiated from A. polymorpha by the absence of well-developed crests and from A. eisenacki (Davey 1969) comb. nov. by the presence of numerous stout processes. A. attadalica (Cookson and Eisenack 19626) comb, nov., from the Apto-?Albian of Australia, is most similar but possesses a wide distinctive cingulum and usually a ventral furrow. Other species The following two species now fall within the emended diagnosis of Aptea and are transferred to this genus. Aptea attadalica (Cookson and Eisenack) Davey and Verdier comb. nov. = ICyclo- nephelium attadalicum Cookson and Eisenack 19626, p. 495, pi. 5, figs. 12-15. Apto- ?Albian, Australia. Aptea eisenacki (Davey) Davey and Verdier comb. nov. = Cyclonephelium eisenacki Davey 1969, pp. 170, 171 ; pi. 8, figs. 3, 4; pi. 9, fig. 4; text-fig. 17a, b. Albian, Canada. 644 PALAEONTOLOGY, VOLUME 17 Genus deflandrea Eisenack emend. Williams and Downie 1966 Deflandrea perlucida Alberti 1959 1959 Deflandrea perlucida Alberti, p. 102, pi. 9, figs. 16, 17. Reported occurrence. Middle to Upper Barremian, England (Davey 1974 in press). Upper Barremian, Germany (Alberti 1961). Albian?, Australia (as D. rotundata, Eisenack and Cookson 1960). Deflandrea terrula Davey 1974 Plate 93, fig. 3 1974 Deflandrea terrula Davey, pi. 8, figs. 4, 5 (in press). Reported occurrence. Lower to Middle Barremian, England (Davey 1974 in press). Genus muderongia Cookson and Eisenack 1958 Muder ongia cf. staurota Sarjeant 1966 Plate 91, figs. 4, 8 Remarks. Three specimens attributable to M. cf. staurota were found, one in sample FR 445 and two in sample FR 448. In each case, preservation was poor and only the antapical region was present. It is characterized by a central body of subcircular outline, a single, proximally broad antapical horn, and two (or perhaps one) very reduced lateral horns. These forms are considerably smaller than the type material of M. staurota and closely resemble specimens described by Davey (1974 in press) from the late Middle and Upper Barremian of England. Dimensions. Plate 91, fig. 8. Overall length 72 /zm, length of antapical horn 27 fim, length of lateral horns 10 and 13 p.m. Plate 91, fig. 4. Overall length 78 ^xm, length of antapical horn 33 length of lateral horn 13 fxm. Stratigraphic comments. The presence of M. cf. staurota in the Lower Aptian extends the range of the genus Muderongia upwards. It is now late Upper Kimmeridgian (see Gitmez and Sarjeant 1972) to Lower Aptian. Genus pareodinia Deflandre emend. Gocht 1970 Remarks. The present authors agree with Gocht (1970) that specimens attributable to Pareodinia may simply be poorly preserved specimens of Kalyptea Cookson and Eisenack 19606 which have lost their calyptra, the surrounding, veil-like covering. Elence, Kalyptea is considered to be a junior synonym of Pareodinia. The latter genus is characterized as being a single-walled ovoidal cyst possessing a two-plate inter- calary archaeopyle near the apex. A calyptra and apical horn are often present. Imbatodinium Vozzhennikova 1967 is distinguished from Pareodinia only with difficulty. It is characterized by an elongate body with a sulcus and cingulum, the latter being towards the antapex. The ornamentation may be coarse, and there may be an apical tentacle. At present the genus appears to be restricted to the latest Jurassic and earliest Cretaceous. Caligodinium amiculum Drugg (1970), from the Danian of the U.S. A., is very similar to the specimens herein assigned to P. cf. aceras. Caligodinium Drugg (1970) is closely related to Pareodinia but may be distinguished by the type of archaeopyle present ; DAVEY AND VERDIER: APTIAN DINOFLAGELLATE CYSTS 645 in the former the operculum consists of three parts— two dorsal intercalary plates and a larger plate composed of the three or four apical plates. If the latter large plate becomes detached only in damaged specimens, as appears likely, then Caligodinium would be a junior synonym of Pcireodinia. Pareodinia cf. aceras (Manum and Cookson 1964) comb. nov. Plate 93, fig. 1 1964 Kalyptea aceras Manum and Cookson, p. 27, pi. 6, figs. 9-11. Remarks. The rare Aptian specimens encountered are identical to the Canadian type material except that they are considerably smaller. A small apical horn may be present, and the archeopyle is formed by the displacement of two intercalary plates just beneath the apex. Dimensions. Plate 93, fig. 1 . Shell length, 48 ^m, width 37 fim. Range : shell length 48-54 ^m, width 37-40 jam, wall thickness approximately 1 p m, reticulation approximately 0-5 ^m. Reported occurrence. Early Upper Cretaceous?, Arctic Canada (Manum and Cookson 1964). Pareodinia sp. Plate 93, fig. 9 Description. The cyst is elongate-ovoidal, coarsely granular to finely reticulate, and possesses a two-plate intercalary archaeopyle just beneath the apex. A reduced calyptra is present. Polar structures and a cingulum are absent. Dimensions. Plate 93, fig. 9. Shell length 62 pm, width 42 pm. Second specimen; length 58 ^m, width 33 pm. Remarks. Only two specimens of this distinctive species were recovered. Its lack of an apical horn distinguishes it from all associated species except P. aceras , from which it may be distinguished by its more elongate shape and less strongly reticulate wall. Genus wallodinium Loeblich and Loeblich 1968 Wallodinium lima (Cookson and Eisenack 1960) comb. nov. I960 Diplotesta tuna Cookson and Eisenack, p. 10, pi. 3, fig. 21. Remarks. During the present study the procedure used by Davey (1974 in press) was followed. That is, due to continuous variation within this genus, all specimens can be assigned for practical purposes, to a single species. W. luna comb. nov. has priority. It is not our intention, at present, to synonymize these three species because their distinction in younger strata may be of importance. Reported occurrence. W. luna, W. krutzschi, and W. anglica have a combined stratigraphic range of Lower Hauterivian to Cenomanian (see Davey 1974 in press). Other species Wallodinium anglica (Cookson and Hughes) Davey and Verdier, comb. nov. = Diplotesta anglica Cookson and Hughes 1964, p. 56, pi. 11, figs. 1-5. Albian to Cenomanian, England. 646 PALAEONTOLOGY, VOLUME 17 STRATIGRAPHIC DISCUSSION The distribution of all the dinoflagellate cysts recovered from the two sections investigated is shown on text-figs. 6-7. The ranges of sixty-two species and varieties which are particularly meaningful stratigraphically are shown on the summarizing range chart (text-fig. 8). Occurrences of these taxa in older and younger strata, as reported earlier by the authors and as taken from selected European literature, are also tabulated on this chart. The stratigraphic distribution of certain species and comparisons with previously described Aptian assemblages are discussed below. The Barr emian- Aptian boundary Barremian sediments in the Provence region were not sampled during this study because of their palynologically unfavourable lithology (Urgonian Limestone). The lowest part of the Bedoulian (Couches de passage) consisted of relatively clean lime- stones and proved to be practically barren. The older fossiliferous Bedoulian samples yielded two species, Cribroperidinium sepimentum and Muderongia cf. staurota, which had been reported previously only from the Barremian (Davey 1974 in press). These were extremely rare, and the latter form appears to be characteristically confined to the uppermost Barremian and lowermost Aptian. Of particular significance is the absence from the Lower Aptian of Muderongia staurota s.s. and Pseudoceratium pelliferum. ; these species apparently became extinct during the Late Barremian. Bedoulian The following species first appear in the Bedoulian and are restricted to the Aptian— Aptea polymorpha , Cyclonephelium tabulatum, Meiourogonyaulax psoros, and Tri- chodinium sp. Deflandrea terrula and Meiourogonyaulax sp. became extinct in the Provence region during the Early Aptian, whereas other species, such as Gonyaulacysta cassidata and Protoellipsodinium spino crist atum , first occurred at that time and range into post-Aptian strata. Gargasian A few species first appear in the Upper Aptian and range into younger strata. Among these, the most restricted stratigraphically are Astrocysta cretacea , Prolixosphaeridium parvispinum, Ovoidinium scabrosum, and Cleistosphaeridium polypes clavulum. Pareodinia sp. and Gonyaulacysta sp. have limited ranges and seem, at present, to be restricted to the Gargasian. A number of species which first appear in pre- Aptian strata become extinct in the Gargasian and may be used, as may be the Aptian- restricted species, for differentiating Lower Aptian and Upper Aptian. These include Achomosphaera neptuni , Dingodinium albertii , Gardodinium trabeeulosum , Meiouro- gonyaulax st over i s.s., and Systematopliora schindewolfi. Comparison with previously described Aptian assemblages Europe. Eisenack 1958. Of the twenty-seven distinct species described by Eisenack from the Upper Aptian of Germany, eighteen have been found in the French strato- type sections. The restricted distributions of Aptea polymorpha , Florentinia laciniata (as Hystrichosphaeridium ferox), and Gonyaulacysta sp. (as Gen. and sp. indet.) in BEDOULIAN 4* 4a» & 4k 4. 4. -4 4k C n 4* •fit 4k -fa 4. 4 4 0 1 to *^4 1 00 1 05 1 cn — L- no 11 _J _! 4k _J 6AR6ASIAN AGE SAMPLE NUMBER o o o o o o o o o o o o o o o o cf. o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o ? o o o o o o o o o o o cf. o o o o o o o o o o o o o ? o o o o 6 o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o cf. o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o A . cf. neptum A . polymorphs A. securigera A . cretscea B. micropoda C. asymmebricum C. co Hi vert C. mtnor C. reticulata C. nyei C. armabum C. huguonioti huguoniobi C. polypes polypes C. ocean sea C. edwardsi C. sepimenbum C. disbin churn C. babulabum D. perfuc/da B, zerru/a D. albertii E. ph rag mites F. m ante Hi F amphora G. cassid&ta G. h el i go idea G. benuiceras H. pu/chrum K. si mp/ic/sp/n um " Ka/ypzea sp. in D/V, 197 ! M. psoros . M. stoven Meiourogonyaulax sp. M. cf.staurota 0. opercufata O. complex P. lamsnaspinosum P. parvisptnum P. clavulum P. spinosum S. cingula bus cenguiabus S. ramosus multibrevis S. ramosus ramosus S. ramosus reticu/atus S. schindewolfi T. variecalamum T. castanea Trichodinium sp. w. iuna text-fig. 6. Microplankton distribution in the samples analysed from the La Bedoule section. BEDOULIAN GARGASIAN AGE -to JS. -fc» -fck ■Cs. -U cn cn cn cn cn cn cn ro i C*j 1 1 1 Cn 1 1 ^1 1 o o cf. o o O o o o o o O o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o ? o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o cf. o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o O o o o o o o 0 o o o o o o o o o o o o o o o o o o SAMPLE NUMBER A nepbuni A cf. nepbuni A granulabum A polymorpha A crebacea B. micropoda C asymmebricum C . co! liver i C minor C. reticu/aba C. parva C nyei C. armabum C tiuguon/obi huguoniobi C po types c/avu/urn C polypes polypes C oceanica C ed ward si C. disbincbum C tabu/abum D perlucida D berrula D. a/berbii £ . arnace £ phragmibes £ /acimaba £ manbelli £. amphora G. brabecu/osum Gonyau/acysba sp G. cass/daba G. helicoidea G. tenuiceras H. pulchrum ? H arundum H recurvabum H bubiferum Ka/ypbea sp. in D/V. 1971 K simplicispinum M cf. bulloidea M. psoros M sboveri ? M crinibum 0 operculaba O. complex O. nannum O scabrosum Pareodinia cf. aceras Pareodinia sp. P. laminaspinosum P parvispmum P clavulum P spinocnstabum Spmiferibes sp. S. cingu/abus cingu tabus S ramosus mulbiorevis S. ramosus ramosus 5. ramosus rebicu/abus S. schindewo/f/ T. variecalamum T. casbanea Trichodinium sp. W. tuna text-fig. 7. Microplankton distribution in the samples analysed from the Gargas section. Co ■ 1 S' 3 g} o to *f §• ^CD !i $s N^5 ?! ^3 to O Co — Co to '-‘< Bedoulian Gargasian c^m • to C52. to>-c % If "s'6 £»> 5-?. to 05 •> 3 ■o Co M >5 vS £ .> *>. ''to ts*£ Ojto™ ^ too ^to2 ■xto3 ■vtoto §h>2. .'©X® §St toto .to cS ©1 JM Co to to" to o to A ?r>, cfl i; o| H Is ~-Qj 03 CV §" to to TAXA C. asymmetricum C. oceanica C. ed wards / C. distinctum H. pu/chrum 0. opercu/ata O. complex S. ramosus ram os us C. huguonioti huguonioti ? M. crinitum S cingu/abus cmgu/abus T. vanecalamum T. casbanea C. armabum C. co/liveri F manbelli H. bubiferum A. qranu/atum W. tuna C. reticulata C. polypes polypes E. phra g mites F. amphora Ka/yptea sp. in D/V, 1971 P laminaspinosum S. ramosus reticu/atus G. cassidaCa E. am ace F. laciniaba A . cretacea G. he/icoidea P. parvispinum C. minor O. scabrosum C. polypes clavulum B. m/cropoda G. benuiceras P. spinosum P. spino crista bum C. parva D a/bertii G. trabecu/osum C nvei C vabu/abum P. clavulum A po/ymorpha H. recurvatum Parecdinia sp. A . neptuni / A. cf. neptuni M. psoros Tnchouinium sp. S. schindewo/Fi SpiniFeribes sp Gonyau/acysta sp. P. simp/ic/'spinum M. stover/ D per/ucida M. bulloidea D. terru/a Meiourogonyaulax sp. C. sepimenbum M. cF. stauroba text-fig. 8. Range chart of selected Aptian microplankton observed in this study and as reported in certain European Cretaceous publications. 650 PALAEONTOLOGY, VOLUME 17 our samples support the Late Aptian age for the German material and also confirm the stratigraphic value of these species. Alberti 1961. Alberti describes three assemblages from the German Aptian; two are assigned to the Lower Aptian and one to the Upper Aptian. All three assemblages contain Pseudoceratium pelliferum, which has been reported otherwise only from pre- Aptian Early Cretaceous strata, and one assemblage contains Muderongia simplex , which occurs only in the pre-Barremian Early Cretaceous. Hence, it appears probable that these assemblages are of Barremian or older Early Cretaceous age. Millioud 1969. The Angles section (south-eastern France) studied by Millioud included three samples of Early Aptian age. The assemblages he reported agree well with the microplankton distribution at La Bedoule, with the exceptions that Gonyaula- cysta aptiana and Phoberocysta neocomica were not present in our samples. Their absence could be explained by the fact that the two oldest Aptian samples at La Bedoule (limestone facies) were practically barren and/or that these two species disappeared during earliest Aptian time. Australia. Cookson and Eisenack 1958, 19606, 19626. Many Aptian assemblages from various formations and localities have been described by these authors. Although the assemblages do vary somewhat in species composition basically two associations are represented. The first one, from the Roma Formation (Santos Oodnadatta No. 2 borehole), and Osborne Formation (Rakich borehole), contains Canningia colliveri, Gonyaulacysta eassidata, Spinidinium styloniferum, Carpodinium granulatum , Tri- chodinium castanea, and Gonyaulacysta tenuiceras. Although in Europe, none of these species are restricted to the Aptian, the age of this association, according to the present study, could well be Aptian. The second association, reported from the Muderong Shale, Windalia Radiolarite, Birdrong Formation (Grierson Member), Roma series and several borehole samples from unspecified formations, contains Dingodinium cerviculum , Muderongia mcwhaei, and M. tetracantha. Unless these species of Muderongia have a younger range in Australia than in Europe, their presence would indicate a pre-Aptian age. Canada. Singh 1971. Although reportedly no Aptian material was studied by Singh, some of the species he recovered strongly suggest an Aptian age for the lower part of the section. The upper part of the section does appear to be of the late Albian age, as assigned. However, the presence of Gardodinium trabeculosum (as G. eisenacki), Dingodinium albertii (as D. cerviculum), Chlamydophorella nyei, and certain of the illustrated specimens of Aptea polymorpha is certainly indicative of an Aptian age according to our recent studies. The association of some of these species with De- flandrea limpida(D. gallia Davey and Verdier 1 973 is a junior synonym) and Ovoidinium verrucosum (as Ascodinium verrucosum), which indicate a Late Albian age, strongly suggests that there is reworked Aptian in the upper part of Singh’s section. Brideaux 19716. The assemblages described by Brideaux from the Middle to Upper Albian of Canada strongly resemble those of Singh (1971) and also include Chlamy- dophorella nyei and Dingodinium albertii (as D. cerviculum ), together with Gonyaula- cysta sp. (as Dinopterygium sp. A). These species are strongly suggestive of an Aptian age, and their presence in Albian strata signifies erosion and redeposition of Aptian sediments during Albian time. DAVEY AND VERDIER: APTIAN DINOFLAGELLATE CYSTS 651 Acknowledgements. The authors would like to express their appreciation to Esso Production Research- European Laboratories, France for permission to publish this paper. In particular we would like to thank Mr. D. H. Jones, Director of Research, who facilitated its achievement and Dr. R. W. Harris, Jun. for critically reading the manuscript. REFERENCES All references to fossil microplankton not listed below are to be found in the ‘Bibliography and index of fossil dinoflagellates and acritarchs’ by Downie and Sarjeant 1964. breistroffer, m. 1947. Sur les zones d’Ammonites dans EAlbien de France et d’Angleterre. Trav. Lab. Geol. Fac. Sci. Grenoble , 26, 1-88. brideaux, w. w. 1971a. Recurrent species groupings in fossil microplankton assemblages. Palaeogeography, Palaeoclimatol., Palaeoecol. 9, 101-122, pis. 1-4. — 19716. Palynology of the Lower Colorado Group Central Alberta, Canada. I. Introductory remarks. Geology and microplankton studies. Palaeontographica , B, 135, 53-114, pis. 21-30. clarke, R. f. a. and verdier, j. p. 1967. An investigation of microplankton assemblages from the Chalk of the Isle of Wight, England. Verhandel. Koninkl. Ned. Akad. Wetenschap , Afdel. Natuurk. Sect. I. 24 (3), 1-96, pis. 1-17. davey, r. j., sarjeant, w. A. s. and verdier, j. p. 1968. A note on the nomenclature of some Upper Cretaceous and Eocene dinoflagellate taxa. Taxon , 17, 181-183. cookson, i. c. and eisenack, a. 1968. Microplankton from two samples from Gingin Brook no. 4 borehole. Western Australia. J. Roy. Soc. Western Australia, 51 (4), 110-122, figs. 1-6. 1969. Some microplankton from two bores at Balcatta, Western Australia. Ibid. 52 (1), 3-8, figs. 1, 2. 1971. Cretaceous microplankton from Eyre no. 1 bore core 20, Western Australia. Proc. Roy. Soc. Victoria , 84 (2), 217-226, pis. 7-11. and hughes, n. f. 1964. Microplankton from the Cambridge Greensand (Mid-Cretaceous). Palaeon- tology., 7, 37-59, pis. 5-11. cox, r. l. 1971. Dinoflagellate cyst structures: walls, cavities and bodies. Ibid. 14 (1), 22-33, pi. 7. davey, R. J. 1969. Non-calcareous microplankton from the Cenomanian of England, northern France and North America, pt. I. Bull. Br. Mus. nat. Hist. (Geol), 17 (3), 103-180, pis. 1-11. 1970. Non-calcareous microplankton from the Cenomanian of England, northern France and North America, pt. II. Ibid. 18 (8), 333-398, pis. 1-10. 1974. Dinoflagellate cysts from the Barremian of the Speeton Clay, England. Silver Jubilee of the Birbal Salmi Institute Volume. (In Press.) downie, c., sarjeant, w. A. s. and williams, G. l. 1966. Fossil dinoflagellate cysts attributed to Baltisphaeridium. In Studies on Mesozoic and Cainozoic dinoflagellate cysts. Bull. Br. Mus. nat. Hist. (Geol.), suppl. 3, 157-173. 1969. Generic reallocations. In Appendix to ‘Studies on Mesozoic and Cainozoic dinoflagellate cysts’. Ibid, appendix to suppl. 3, 15-17. and verdier, j. p. 1971. An investigation of microplankton assemblages from the Albian of the Paris Basin. Verhandel. Koninkl. Ned. Akad. Wetenschap , Afdel. Natuurk. Sect. I, 26 (2), 1-58, pis. 1-7. 1973. An investigation of microplankton assemblages from latest Albian (Yraconian) sediments. Revista Espahola de Micropaleonto/ogia, 5, 173-212, pis. 1-5. and williams, G. l. 1966a. The genera Ilystrichosphaera and Achomosphaera. In Studies on Mesozoic and Cainozoic dinoflagellate cysts. Bull. Br. Mus. nat. Hist. (Geol.), suppl. 3, 28-52. — 19666. The genus Hystrichosphaeridium and its allies. In Studies on Mesozoic and Cainozoic dinoflagellate cysts. Ibid. 53-106. 1969. Generic reallocations. In Appendix to ‘Studies on Mesozoic and Cainozoic dinoflagellate cysts’. Ibid, appendix to suppl. 3, 4-7. downie, c. and sarjeant, w. a. s. 1964. Bibliography and index of fossil dinoflagellates and acritarchs. Geol. Soc. Am. Mem. 94, 1-180. 652 PALAEONTOLOGY, VOLUME 17 drugg, w. s. 1970. Some new genera, species and combinations of phytoplankton from the Lower Tertiary of the Gulf Coast, U.S.A. Proc. North American Paleontological Convention , 809-843, figs. 1-19. fabre-taxy, s., moullade, m. and thomel, G. 1965. Le Bedoulien dans sa region-type, la Bedoule-Cassis (B. du R.). Mem. Bur. Rech. Geol. Minieres, 34, 173-199. flandrin, j. 1965. Rapport sur l’etage Aptien. Ibid. 227-234. gitmez, G. u. and sarjeant, w. a. s. 1972. Dinoflagellate cysts and acritarchs from the Kimmeridgian (Upper Jurassic) of England, Scotland and France. Bull. Br. Mus. nat. Hist. (Geol.), 21 (5), 175-253, pis. 1-17. gocht, H. 1970. Dinoflagellaten Zysten aus dem Bathonium des Erdolfeldes Aldorf (NW-Deutschland). Palaeontographica , B, 129, 125-165, pis. 26-35. hebert, e. 1864. Sur la craie glauconieuse du nord-ouest du Bassin de Paris. C.r. hebd. Seanc. Acad. Sci. 58, 475-479. — 1871. Le Neocomien inferieur du Midi de la France (Drome et Basses- Alpes). Bull. Soc. geol. France , 28, 137-170. 1872. Documents relatifs au terrain cretace du Midi de la France. Ibid. 29, 393. kilian, w. 1887. Systeme cretace. Extr. Ann. Geol. Univ. 3, 299-356. — 1907-1913. Lethea Geognostica. II. Des Mesozoicum, t. 3, Kreide, Unterkreide. Stuttgart, Borntrager, 398 pp. — and reboul, p. 1915. Contribution a l’etude des faunes paleocretacees du sud-est de la France. I. Fa faune de l’Aptien inferieur des environs de Montelimar (Drome), carriere de 1'Homme d’Armes. Mem. Carte Geol. France, 14, 1-221. loeblich, a. r., Jun. and loeblich, a. r.. III. 1968. Index to the genera, subgenera and sections of the Pyrrhophyta, II. J. Palaeontol. 42 (1), 210-213. manum, s. and cookson, i. c. 1964. Cretaceous microplankton in a sample from Graham Island, Arctic Canada, collected during the Second ‘Fram’-expedition (1898-1902). Skr. Norske Vidensk-Akad. mat. Nat. Kl. ( N.S. ), 17, 1-36, pis. 1-7. matheron, p. 1842. Catalogue methodique et descriptif des corps organises fossiles du departement des Bouches-du- Rhone et lieux circonvoisins. Marseille, 269 pp. millioud, m. e. 1967. Palynological study of the type localities at Valangin and Hauterive. Rev. Palaeobotan. Palynol. 5, 155-167. — 1969. Dinoflagellates and acritarchs from some western European Fower Cretaceous type localities. In bronnimann, p. and renz, h. h. (eds.). Proceedings of the First International Conference on Planktonic Microfossils, Geneva 1967, Leiden, 2, 420-434, pis. 1-3. moullade, m. 1965. Revision des stratotypes de l’Aptien: Gargas (Vaucluse). Mem. Bur. Rech. Geol. Minieres, 34, 201-214. 1966. Etude stratigraphique et micropaleontologique du Cretace inferieur de la ‘Fosse vocontienne’. Doc. Lab. Geol. Fac. Sci. Lyon, 15, 1-369, pis. 1-17. orbigny, A. d’. 1840. Paleontologie frangaise. Terrains cretaces. I. Cephalopodes. Paris, 662 pp. 1842. Paleontologie franfaise. Terrain cretace. II. Gastropodes. Paris, 807 pp. 1850. Prodrome de paleontologie stratigraphique universelle. Masson, Paris. reynes, p. 1861. Etudes sur le synchronisme et la delimitation des terrains cretaces du sud-est de la France. Mem. Soc. Emulat. Provence, 1, 5-115. sarjeant, w. a. s. 1966a. Dinoflagellate cysts with Gonyaulax- type tabulation. In Studies on Mesozoic and Cainozoic dinoflagellate cysts. Bull. Br. Mus. nat. Hist. (Geol.), suppl. 3, 107-156. — 19666. Further dinoflagellate cysts from Speeton Clay (Fower Cretaceous). In Studies on Mesozoic and Cainozoic dinoflagellate cysts. Ibid. 199-213. — 1969. Taxonomic changes. In Appendix to 'Studies on Mesozoic and Cainozoic dinoflagellate cysts’. Ibid, appendix to suppl. 3, 7-14. — 1970. The genus Spiniferites Mantell, 1850 (Dinophyceae). Grana Palynologica, 10, 74-78. singh, c. 1971. Fower Cretaceous microfloras of the Peace river area. Northwestern Alberta. Res. Council Alberta Bull. 28, 1-542, pis. 1-80. SORNAY, j. 1957. Albien. In Lexique stratigraphique international, I, fasc. 4aVI (Cretace), Paris. toucas, a. 1888. Note sur le Jurassique superieur et le Cretace inferieur de la vallee du Rhone. Bull. Soc. Geol. France, 16, 903-927. DAVEY AND VERDIER: APTIAN DINOFLAGELLATE CYSTS 653 vozzhennikova, T. F. 1967. Introduction to the study of fossilized peridinid algae. British Library, Lending Division, Boston Spa, Yorks., England (English translation by K. Syers, edited by W. A. S. Sarjeant). 233 pp., pis. 1-121. (Original in Russian.) williams, G. L. and downie, c. 1966. Further dinoflagellate cysts from the London Clay. In Studies on Mesozoic and Cainozoic dinoflagellate cysts. Bull. Br. Mus. nat. Hist. (Geol.), suppl. 3, 215-235. R. J. DAVEY and J.-P. VERDIER Esso Production Research 213 Cours Victor-Hugo 33321 Begles France Present Address R. J. DAVEY Institute of Geological Sciences Ring Road, Halton Revised typescript received 22 December 1973 Leeds CHONOPHYLLINID CORALS FROM THE SILURIAN OF NEW SOUTH WALES by R. A. MCLEAN Abstract. Ketophyllum attenuatum sp. nov. from the Rosyth Limestone (Upper Llandovery) of central N.S.W. is described. Mictocystis endophylloides Etheridge 1908, from the Quarry Creek Limestone (Upper Llandovery) in central N.S.W. is redescribed and a lectotype chosen. Paralectotype material of Yassia enormis (Etheridge 1913) from strata of Ludlow age in the Yass district, southern N.S.W. is reviewed and illustrated. The close affinities of the genera Ketophyllum , Mictocystis, and Yassia are emphasized and their taxonomic status and relationships reviewed. STRATIGRAPHY AND AGE Ketophyllum and Mictocystis occur separately in two extremely fossiliferous horizons in the area west of Orange in central N.S.W. (text-fig. 1). The Quarry Creek Lime- stone (Packham and Stevens 1955), from which Mictocystis endophylloides was described by Etheridge (1908), is best exposed in Quarry Creek (Bed ‘A’ of Siissmilch 1906). At this locality approximately 9 m of coarsely bedded, massive limestone overlie Upper Ordovician weathered, red, tuffaceous beds with marked uncon- formity. Two other horizons of massive limestone outcrop in Quarry Creek (Beds B’ and ‘C’ of Siissmilch 1906) and were considered by Packham and Stevens (1955, [Palaeontology, Vol. 17, Part 3, 1974, pp. 655-668, pis. 94-95] 656 PALAEONTOLOGY, VOLUME 17 p. 58) as strike-faulted repetitions of this same horizon. Bed ‘A’ is exposed again to the south in Spring Creek (Packham and Stevens 1955, fig. 1). The Rosyth Limestone (Walker 1959) is a unit of fossiliferous, marly limestones interbedded with shale and sandstone occurring in the Boree Creek area, south of the main Orange-Cudal road (text-fig. 1). This horizon also unconformably overlies Upper Ordovician volcanics. It reaches a thickness of about 35 m, but the most fossiliferous member, a yellow-grey nodular-weathering biomicrite occurring near the base of the sequence, is only approximately 6 m thick, and it is from this horizon that the specimens of Ketophyllum described here were obtained. The Quarry Creek Limestone is immediately overlain by beds containing a grapto- lite fauna of late Llandovery age (Packham and Stevens 1955). Walker (1959) and Packham (1969) have correlated the Quarry Creek Limestone with the Rosyth Lime- stone mainly on the basis of similarities of the coral fauna. Preliminary studies of the conodont faunas from the Quarry Creek Limestone (Bed "A’, Quarry Creek) and the coral-rich beds of the Rosyth Limestone have also suggested close age similarities between the two formations (G. Bischoff, pers. comm. 1972). The conodonts suggest an assignment no older than the celloni Zone of Walliser (1964, 1971) for these hori- zons and hence it would appear that the age of the limestones does not extend below the Upper Llandovery. The type material of Yassia reviewed here is from limestone at the ‘escarpment north-east of Boonoo Ponds Creek, Hatton’s Corner, Yass River, near Yass’ in southern N.S.W. (Etheridge 1913, p. 37). The form is represented also in the Bow- spring Limestone at Yass, considered to be an equivalent horizon by Jones (1932). The Bowspring Limestone was regarded as being of early to middle Ludlow age on the basis of conodont faunas by Link (1970) and hence a comparable age is likely for the type horizon of Yassia. SYSTEMATIC PALAEONTOLOGY Registration numbers of specimens in the University of Sydney Palaeontological Collections bear the prefix SUP, and where more than one section has been prepared from the one specimen, they have the suffix a , b , etc. Numbers of specimens from the palaeontological collections of the Australian Museum, Sydney, have the prefix AM.F and thin sections in these collections the prefix AM. Family chonophyllidae Holmes, 1887 Subfamily chonophyllinae Holmes, 1887 1927 Omphymatidae Wedekind, p. 46. 1937 Omphymatidae; Soshkina, p. 64. 1949 Chonophyllinae; Stumm, p. 48 ( nom . transl., ex Chonophyllidae Holmes 1887). 1952 Chonophyllidae (part .); Lecompte, p. 465. 1952 Ketophyllidae (part .); Lecompte, p. 467. 1956 Chonophyllinae; Hill, p. L300. 1962 Chonophyllidae (part.); Soshkina, et at., p. 300. 1962 Dokophyllidae Soshkina, et at., p. 319. 1963 Chonophyllidae (part); Ivanovskiy, p. 108. 1965 Spongophyllidae (part.); Ivanovskiy, p. 85 (non Dybowski 1873). 1965 Ketophyllidae (part.); Ivanovskiy, p. 96. McLEAN: CHONOPH YLLINID CORALS 657 Diagnosis: Coralla solitary or compound. Septa typically amplexoid, tabulae flat and commonly grouped in series. Genus ketophyllum Wedekind, 1927 1851 1854 1876 1901 1902 71919 1927 1927 1927 71937 1937 1939 1944 1945 1947 1950 1952 1956 1960 1960 1961 71962 1965 1965 1965 1971 non 1937 non 1959 non 1965 Omphyma (part.)', Edwards and Haime, p. 400. Omphyma (part.) ', Edwards and Haime, p. 287. Omphyma (part.)', Rominger, p. 1 17. Omphyma ; Lambe, p. 177. Omphyma ', Pocta, p. 137. Heterolasma Ehlers, p. 461. Ketophyllum Wedekind, p. 48. Dokophyllum Wedekind, p. 48. Omphyma (part.)', Wedekind, p. 58. Dokophyllum (part.)', Soshkina, p. 65. Omphyma (part.) ', Butler, p. 87. Omphyma ; Northrop, p. 143. Ketophyllum ; Wang, p. 26. Ketophyllum ; Smith, p. 26. Cetophyllum (7 part.)', Wang, p. 181. Ketophyllum (7 part.) ', Wang, p. 226. Dokophyllum', Bulvanker, p. 22. Ketophyllum ; Hill, p. F300. Ketophyllum', Nikolaeva in Bulvanker et al., p. 225. Dokophyllum (part.)', Zheltonogova, p. 76. Ketophyllum ; Minato, p. 29. Dentilasma (part.)', Ivanovskiy, p. 128. Ketophyllum ', Stumm, p. 47. Dokophyllum ', Strelnikov, p. 42. Dokophyllum ', Zheltonogova, p. 42. Ketophyllum', Lavrusevich, p. 90. Ketophyllum', Soshkina, p. 67. Ketophyllum ; Ivanovskiy, p. 135. Ketophyllum ; Ivanovskiy, p. 123. Type species. K. elegantulum Wedekind, 1927. Klinteberg Beds (Ludlow), Gotland. Diagnosis. Solitary corallum with fossula typically present. Septa amplexoid, often dilated peripherally, continuing over tabulae as low ridges. Dissepiments usually large, tabulae flat, mainly complete and grouped in series. Discussion. There has been some confusion in the past as to the nature of the septa of Ketophyllum. Wedekind’s original description (1927) refers to septal ieistes' or ridges on the dissepiment and tabular surfaces. Wang (1944), however, refers to both discrete and partly contiguous septal spines in his discussion of the genus. There is no mention by Wedekind (1927) of such structures in the Gotland material, nor can they be detected in his illustrations, although in some cases septa may be notched or denticulated on their upper surfaces, possibly related to the trabecular structure of the septal lamellae (e.g. K. richteri Wedekind, 1927, pi. 10, fig. 7). Such a feature can be detected in A', attenuatum sp. nov. described below. Wang later (1947, p. 181 and 1950, p. 226) described the septal structure of Ketophyllum as consisting of slender holacanthine trabeculae embedded in lamellar tissue, and he placed the genus in the Cysti- phyllidae on this basis. In his restudy of the type material of several of Wedekind’s species, Minato (1961) made no mention of septal spines in Ketophyllum, referring in fact to ‘septal lamellae’ (p. 90). The species described by Wang (1944, p. 27) as K. equitabulatum from the ?Wenlock of Yunnan, China, was considered to contain ‘typically discrete septal spines’, but there is no evidence of them in the illustrated longitudinal section (ibid., pi. 1, fig. 4b). In all other respects the form appears to be a typical example of Ketophyllum bearing short septa and is probably a representative of that genus. Ivanovskiy (1965, p. 96 and p. 124) referred to Ketophyllum as possessing septal ‘Ieistes’, but in his description of K. similis sp. nov., he mentioned the presence of septal spines, and included it in the family Ketophyllidae Lecompte. 658 PALAEONTOLOGY, VOLUME 17 Ivanovskiy also placed the spine-bearing forms Dentilasma Ivanovskiy, Nipponophyllum Sugiyama, and Spinolasma Ivanovskiy (? = Hedstroemophyllum Wedekind, see Ivanovskiy 1970a and b) in the Ketophyllidae although these genera are generally considered representative of the Cystiphyllidae. He subsequently (19706) referred K. similis to Dentilasma. There is no evidence of discrete septal spines being present in any N.S.W. representatives of the genus, nor are they apparent in specimens of Ketophyllum from Gotland in the palaeontological collections of the Geology Department, University of Sydney. Therefore, it appears best to retain the most widely accepted classification of the genus, that of Hill (1956) and others, namely with the Chonophyllidae (see Hill 1956, p. F300). In proposing the family Omphymatidae, Wedekind (1927) included in it the new genera Dokophyllum and Keto- phyllum together with Omphyma Rafinesque and Clifford, 1820. He considered Dokophyllum could be distinguished by having short septa not extending into the tabularium and not cutting any dissepiment layers vertically. Ketophyllum , on the other hand, was regarded as having septa longer than in Dokophyllum , although still being generally restricted to the dissepimentarium, and the septa being vertically continuous, cutting dissepiment layers. Ketophyllum was further distinguished from Dokophyllum by greater development of dissepiments, particularly early in ontogeny. Lastly, Omphyma was characterized by having septa similar in nature to Ketophyllum but extending into the tabularium towards the axis. Many later authors, particularly in the U.S.S.R., have followed this taxonomic scheme (e.g. Sosh- kina 1937; Bulvanker 1952; Strelnikov 1965). However, as was pointed out by Lang, Smith and Thomas (1940, p. 90) the type material of the genus Omphyma is lost and the position of the original locality and horizon is uncertain. Hence, since the specific description is not sufficient for certain identification, ‘the name Omphyma cannot be used’ (Lang, Smith and Thomas 1940, p. 91). It is evident that features ascribed by many authors to species of 'Omphyma'' are very close to those described by Wedekind (1927) for representatives of Ketophyllum , the length of the septa being a variable feature (see Hill 1956, p. F300). Hence, several forms described as belonging to Omphyma may be included in Ketophyllum, e.g. O. tenuistriata Wedekind, 1927, O. turbinata (Linnaeus, 1758), O. grande Barrande in Pocta 1902, and O. eriphyle Billings, 1862. Kaljo (1970) listed the compound species ‘Omphyma kutscheri' Wedekind, 1927 as a representative of Ketophyllum. This form shows internal characteristics typical of Ketophyllum (see Wedekind 1927, pi. 17, figs. 1, 2) differing only in its compound growth form; its taxonomic position remains uncertain for the present. The position of Dokophyllum is still debated, many Russian workers still preferring to consider the genera Doko- phyllum and Ketophyllum as distinct (e.g. Soshkina 1937 ; Bulvanker 1952; Zheltonogova 1960, 1965; Soshkina et al. 1962; Strelnikov 1965). However, other authors consider differences between the two genera to be merely gradational and that forms cannot be consistently grouped into the two categories (see Wang 1944; Smith 1945; Hill 1956; Minato 1961 ; Ivanovskiy 1962, 1965). Minato (1961), in particular, restudied the material of the type species of Dokophyllum , D. annulatum, from Gotland and stated (p. 91) that ‘there are no significant structural differences between Doko- phyllum and Ketophyllum '. Consequently, it seems more justified to consider Dokophyllum a junior synonym of Ketophyllum. Several species previously assigned to either Ketophyllum or Dokophyllum appear more likely to be representatives of other genera. The two species referred to Ketophyllum by Soshkina (1937), K. intermedium (Chernyshev, 1893) and K. amplexoidum (Chernyshev, 1893), both from the Upper Wenlock of the Urals, were considered synonymous and placed in Dentilasma by Strelnikov (1971 ). Septal ‘leistes’ were described from both forms by Soshkina (1937), those of ‘K. intermedium' occasionally showing denticulation (Soshkina 1937, p. 70). It cannot be determined from the illustrations of Soshkina (1937, pi. XIII, figs. 1, 2) or Strelnikov (1971, pi. XIX, fig. 4 and pi. XX, fig. 1) whether dis- crete septal spines are actually present. However, the weak development of septa and incomplete nature of the tabulae are atypical of Ketophyllum and the form may belong to Dentilasma. The type species of Dokophyllum , D. annulatum Wedekind, 1927, from the ?Upper Llandovery of Gotland, was not figured in longitudinal section by Wedekind. While dissepiments are lacking in early growth stages, up to five rows of large elongate dissepiments can be determined from figured transverse sections (Wedekind 1927, pi. 9, fig. 15 and Minato 1961, pi. 19, fig. 6). After restudying Wedekind’s material, Minato (1961, p. 91) included D. annulatum in Ketophyllum. Until its tabular structure is described, its inclusion in Ketophyllum is not certain but it is most probably a representative of that genus. The material from the Lower Wenlock of the Urals referred to ‘ D . annulatum' by Soshkina (1937) appears to have short spinose septa mainly confined to the periphery and lacks typical ketophyllid tabular and dissepiment structure. It does not appear to be conspecific with Wedekind's species and appears to show greater similarities to Dentilasma. Dokophyllum sociale Soshkina, 1937, also from the Lower Wenlock of the Urals, appears to lack dissepiments and is a fasciculate form. It does not seem to be a representative of Ketophyllum. Dokophyllum tabulatum Bulvanker, 1952, from the Upper Silurian (Skala Horizon) of Podolia, has almost no development of dissepiments and the septa are strongly trabeculate, in some cases almost consisting of isolated spines. The tabulae, however, are typically ketophyllid and so the taxonomic position of the species is uncertain. It may per- haps be a representative of Dentilasma as suggested by Ivanovskiy (1962, p. 128). Of the original twenty species and varieties from Gotland included in the genera Dokophyllum and Ketophyllum by Wedekind (1927), it is evident that many are synonymous. However, revision of these forms must await detailed study of the type material. The genus Heterolasma Ehlers, 1919 (type species H.foerstei, Ehlers 1919, Manistique Formation, Upper Llan- dovery-Lower Wenlock, Michigan) has been generally included as a synonym of Ketophyllum (Hill, 1956). Un- fortunately, the genus has not been studied in thin section and the internal morphology is unclear. Ehlers (1919, McLEAN: CHONOPH YLLINID CORALS 659 p. 466) stated that true dissepiments probably do not occur although the tabulae tend to become incomplete peri- pherally. Until the genus is studied in thin section its taxonomic position cannot be clarified. The genera Chonophyllum Edwards and Haime, 1850, Pilophy/lum Wedekind, 1927, and Lindstroemophyllum Wang, 1947 show similarities to Ketophyllum. Chonophyllum differs in possessing very strongly dilated septa in the peripheral region. Hill (1956, p. F300) referred the species of ‘Omphyma' described by Wedekind (1927) to this genus and ‘O.’ | flabellata Wedekind is certainly a representative of Chonophyllum , having a characteristic narrow tabularium. ‘0.’ tenuistriata Wedekind, however, has far less strongly dilated septa and appears to be closer to Ketophyllum, in which it is included here. It seems that the extent to which the septa are peripherally dilated is a very variable feature, ranging from the very heavily dilated forms in Chonophyllum to very thin and undilated types in some species of Ketophyllum such as K. pseudoannulatum Wedekind. As there seems to be a much greater preponderance of forms in Ketophyllum bearing generally weakly or only moderately dilated septa, it seems useful to distinguish Chonophyllum for forms bearing heavily dilated septa. Pilophy/lum Wedekind, 1927 also has generally strongly dilated septa at the periphery but forms showing less dilation (e.g. P. progressum Wedekind, 1927) may be distinguished from Ketophyllum by having an arched series of incomplete tabulae, as compared to the series of flat complete tabulae in Ketophyllum. Pilophyllum also tends to show a more convolute arrangement of the septa axially. Lindstroemophyllum Wang, 1947 has a septal structure comparable to Ketophyllum but the septa are strongly con- volute axially. There are also no true dissepiments in Lindstroemophyllum. Although Ivanovskiy (1965, p. 96) included it as a synonym of Ketophyllum , the lack of dissepiments, apparently throughout ontogeny, seems sufficient to dis- tinguish it from that genus. Range. Upper Llandovery of Britain, ?north-east U.S.S.R., ?Michigan, N.S.W.; Wenlock of Britain, Gotland, ?Estonia, ?Podolia, ?north-east U.S.S.R., Vaygach I., Tadzhikistan, China, Indiana-Kentucky, Quebec; Upper Silurian of Gotland, Czechoslovakia, Kazakhstan, south-west Siberia. Ketophyllum attenuatum sp. nov. Plate 94, figs. 1, 2, 5-8; Plate 95, fig. 3 Derivation of name. Latin attenuatus = reduced, referring to marked thinning of septa in tabularium. Material. Holotype SUP 46190; paratypes SUP 46191-46193, 46195-46200. Addi- tional material doubtfully referred to this species— SUP 46194, 46201-46203. Distribution. Rosyth Limestone, Upper Llandovery. Diagnosis. Trochoid-turbinate Ketophyllum having septa moderately dilated peri- pherally, the septa extending as very low ridges over tabulae to corallite axis where they may be somewhat twisted. Tabulae complete and incomplete, grouped in series to varying degrees. Dissepiments large, elongate, steeply inclined. Description. Corallum solitary, trochoid to turbinate, with epitheca showing rather weakly developed septal grooves and transverse wrinkling. Most specimens show corrosion with epitheca not preserved. Corallite diameter ranges from approximately 30-55 mm with one specimen (SUP 46203), doubtfully referred to this species, reaching diameter of at least 85 mm. Corallite height ranges up to at least 70 mm, although all specimens are incomplete. Large specimen (SUP 46203) has height of at least 150 mm. Calice shallow. Prominent tabular fossula developed. Septa long, typically reaching axis, where they may show some twisting. Septa thin in tabularium, occurring as low ridges on tabular floors. Vertically, septa may be seen to be peripherally based on dissepiment crests, rarely cutting dissepiment layer. Upper surface of septum may be notched (see PI. 94, fig. 8) reflecting trabeculae in septum. State of preservation is insufficient to determine nature of trabeculae, but bundles of fibres based vertically on dissepiment crests may be discerned. Trabeculae 660 PALAEONTOLOGY, VOLUME 17 closely spaced and wrapped in indeterminate tissue to give appearance of complete lamellar septum. Septal number difficult to determine owing to irregularity of septal development, but 75-80 septa present in holotype (SUP 46190), ranging to 100-120 septa in largest specimen (SUP 46203), doubtfully referred to this species. Major and minor septa not generally clearly differentiated, although they can be detected in SUP 46191 (PI. 95, fig. 3). Tabulae mainly flat or slightly concave, with downturned edges, complete and incomplete, typically grouped in closely spaced series of approximately 2-5 tabulae. Tabular spacing ranges generally from 0-2 to 2 mm, with average of 0-4-0-8 mm. Tabularium diameter varies from about 17 to 22 mm, i.e. approximately 0-4-0-5 of corallite diameter. Dissepiments large, strongly elongate, very steeply inclined towards axis. Average dimensions 5-10 mm wide and 1-1-5 mm high. Approximately six rows of dissepi- ments developed in distal part of corallum. Dissepiments developed at all growth stages preserved, but proximal portions not found intact and corrosion has typically removed outer dissepiment layers. Remarks. Several specimens are rather atypical of the form described above. The large specimen (SUP 46203) has been mentioned and it is doubtfully placed in K. attenuatum. It differs in having more dilated septa in the tabularium and a generally narrower tabularium as well as its greater size (PI. 94, figs. 3, 4). Further material would be necessary to confirm its inclusion in K. attenuatum. Several rather smaller specimens are also only doubtfully placed in K. attenuatum. They differ from the typical K. attenuatum in having reduced septa which are weakly dilated peripherally and apparently extend just a short distance into the tabularium. Only one small specimen (SUP 46194) shows these features in section (PI. 95, figs. 1, 2), together with two fragmentary etched silicified specimens (SUP 46201, 46202). Whether the reduced septa represent merely less silicification in these specimens, giving the appearance of weaker dilation of the septa, or whether the septa are genuinely thinner, is not certain and until further better-preserved material can be found, these specimens are tentatively included in K. attenuatum. Of the described species of Ketophyllum , none is very close in structure to K. attenuatum. Most species have relatively short septa and the only forms with septa approaching the length typical of K. attenuatum are K. tenuistriatum (Wedekind), K. hoegbomi (Wedekind), and K. intertrium (Hall). K. tenuistriatum (Wedekind, 1927) from the Mulde Beds (Upper Wenlock) of Got- land may be distinguished by having separation of the tabulae into groups and clearer distinction of major and minor septa. K. hoegbomi ( Wedekind, 1927) from the Wenlock of Gotland differs in having fewer rows of dissepiments and more widely spaced EXPLANATION OF PLATE 94 Figs. 1, 2, 5-8. Ketophyllum attenuatum sp. nov., x 2, Rosyth Limestone. 1, SUP 46190a, holotype, trans- verse section. 2, SUP 46192, paratype, longitudinal section. 5, SUP 46190b, holotype, longitudinal section. 6, SUP 46191a, paratype, transverse section. 7, SUP 46193a, paratype, transverse section. 8, SUP 46193b, paratype, longitudinal section. Figs. 3, 4. Ketophyllum attenuatum ? sp. nov., x 1-5, Rosyth Limestone. 3, SUP 46203b, longitudinal section. 4, SUP 46203a, transverse section. PLATE 94 McLEAN, Chonophyllinid corals 662 PALAEONTOLOGY, VOLUME 17 tabulae. Finally, K. intertrium (Hall 1882) from the Upper Llandovery — Wenlock of Michigan (Manistique Dolomite) and Upper Wenlock— ?Lower Ludlow of Indiana-Kentucky (Louisville Limestone) has generally weaker dilation of the septa and a greater number of dissepiments, which are generally smaller and less elongate than in K. attenuatum (see Stumm 1965). Genus mictocystis Etheridge, 1908 1908 Mictocystis Etheridge, p. 18. 1956 1 Mictocystis; Hill, p. F300. 1962 ? Mictocystis; Soshkina et al., p. 311. 1965 ? Mictocystis; Ivanovskiy, p. 87. Type species. M. endophylloides Etheridge, 1908. Quarry Creek Limestone, Spring Creek, N.S.W. Upper Llandovery. Diagnosis. Aphroid corallum with large, widely spaced tabularia set in lonsdaleoid dissepimentarium. Septa confined to tabularium and surface of dissepiments adjoin- ing tabularium. Tabulae mainly flat, complete, often with downturned edges. Dissepi- ments very large, mainly elongate. Discussion. The genus Mictocystis Etheridge was originally grouped with the cysti- phyllids by Etheridge (1908) on the basis of its having abundant dissepimental tissue typical of members of that family. Hill (1956), however, included the genus doubt- fully in the subfamily Chonophyllinae Holmes. Re-examination of the syntype material and additional specimens of Mictocystis has shown that the septal structure consists of lamellae based on the dissepiments adjacent to the tabularia and on the tabulae themselves. In the tabularia the septa may or may not pierce the overlying tabula in the peripheral region. Such a septal structure is typical in the subfamily Chonophyllinae. Ivanovskiy (1965) has suggested that the genus Evenkiella Soshkina, 1955, may be a synonym of Mictocystis. However, Evenkiella is a cerioid form in which the septa extend much further into the dissepimentarium and are apparently not of amplexoid type (see Soshkina 1955 and Ivanovskiy 1963). Evenkiella is possibly a synonym of the genus Strombodes Schweigger. In septal development within the tabularium and in the nature of tabulae and dissepiments, Mictocystis is closely similar to Ketophyllum Wedekind, discussed above. Ketophyllum is, however, a solitary form and has more strongly developed septa in the dissepimentarium. EXPLANATION OF PLATE 95 Figs. 1, 2. Ketophyllum attenuatum ? sp. nov., x 2, Rosyth Limestone. 1, SUP 46194b, longitudinal section. 2, SUP 46194a, transverse section. Fig. 3. Ketophyllum attenuatum sp. nov., Rosyth Limestone. SUP 46191b, paratype, longitudinal section, x 2. Figs. 4, 5. Mictocystis endophylloides Etheridge, x 1, Quarry Creek Limestone. 4, AM.F 13616, lectotype, distal surface. 5, AM.F 13616, lectotype, weathered surface showing portion of a tabularium and adjacent dissepimentarium. Figs. 6-8. Yassia enormis (Etheridge), x 1-5, limestone at escarpment north-east of Boonoo Ponds Creek, Hatton’s Corner, Yass River. 6, AM 847, paralectotype, longitudinal section. 7, AM 869, paralecto- type, transverse section. 8, AM 674, paralectotype, transverse section. PLATE 95 McLEAN, Chonophyllinid corals , r * ' 664 PALAEONTOLOGY, VOLUME 17 Mictocystis endophylloides Etheridge, 1908 Plate 95, figs. 4, 5; text-fig. 2 a-c 1908 Mictocystis endophylloides Etheridge, p. 20, pi. 4, figs. 1-4; pi. 5. 1956 Mictocystis endophylloides'. Hill, p. F300, fig. 204, 1. Material. Lectotype (here designated) AM.F 13616; paralectotype AM.F 13617. Additional material SUP 27154, 63256, 63272. Distribution. Fectotype and paralectotype. Quarry Creek Limestone (Bed ‘A’), Spring Creek. Additional material, Quarry Creek Limestone (Bed ‘C’), Quarry Creek. Upper Llandovery. Description. Corallum aphroid, large, with incomplete width 130 mm and height 80 mm. Spacing of tabularia in dissepimentarium ranges from 25 to 30 mm in small specimen (lectotype AM.F 13616, Etheridge 1908, pi. 4, fig. 1) to more commonly 30-50 mm, although this value varies greatly within the one colony. Calicular pits generally deep (10-15 mm, see text-fig. 2c). External morphological features are illustrated by Etheridge (1908, pi. 4) and herein (PI. 95, fig. 4). Septa mainly confined b a c text-fig. 2. Mictocystis endophylloides Etheridge, a , SUP 63272a, transverse section of a tabu- larium and part of dissepimentarium, x 3; 6, SUP 27154, longitudinal section of a tabularium and adjacent dissepimentarium, x2; c, AM.F 13616, lectotype, tangential longitudinal section of a tabularium and adjacent dissepimentarium. drawn from a weathered surface (note deep calice), x 2. The surface illustrated is shown also in Plate 95, fig. 5. McLEAN: CHONOPH YLLINID CORALS 665 to tabularia, occurring as broad, low ridges on surface of tabulae, generally not piercing overlying tabulae except near periphery of tabularium. Septa extend on to dissepiments adjacent to tabularia in some cases and are apparently lamellar, although material is mainly silicified and recrystallized and detail of septal structure is obscured. Septal number approximately 44-50, with major septa almost reaching axis and minor septa very short, approximately 0-2 of length of major septa. Septa decrease slightly in width towards axis. Tabularium diameter ranges from approximately 8 to 12 mm in lectotype (AM.F 13616, text-fig. 2c) to more commonly 12-15 mm. Tabulae mainly complete, flat or slightly arched, with downturned edges and average spacing 0-8- 1 -5 mm. They may be weakly grouped in series (text-fig. 2b). Dissepi- ments large, elongate, in gently arched series between tabularia, becoming strongly downturned at margins of tabularia. Dissepiment size variable with average of 7-12 mm in width and 2-4 mm in height. Thin coating of ?lamellar sclerenchyme present on some dissepiment surfaces although recrystallized calcite obscures this in most cases. Remarks. No type specimens of Mictocystis were designated by Etheridge. From the syntype material now housed in the Australian Museum, Sydney, the specimen of Mictocystis endophylloides figured by Etheridge (1908, pi. 4, figs. 1-4, AM.F 13616) is here chosen as the lectotype and the other specimen illustrated by Etheridge (1908, pi. 5, AM.F 13617) is designated paralectotype. Etheridge’s syntype material is almost entirely silicified and is unsuitable for thin-section study. However, additional material collected is somewhat less silicified although extensively recrystallized, and it was possible to prepare drawings from the thin sections obtained. These are illustrated in text-fig. 2a , b. The genus is at present known only from the Quarry Creek Limestone of N.S.W. Genus yassia Jones, 1930 1913 Spongophyllum ; Etheridge, p. 35 {non Edwards and Haime 1851). 1930 Yassia Jones, p. 36. 1932 Crinophyllum Jones, p. 61. 1940 Yassia - Hill, p. 409. 1963 Yassia; Ivanovskiy, p. 111. 1965 Yassia; Lavrusevich and Ivanovskiy in Ivanovskiy, p. 1 19. 1970 Yassia; Ivanovskiy, p. 15. 1971 Yassia; Lavrusevich, p. 92. 71972 Klamathastraea Merriam, p. 40. Type species. Spongophyllum enorme Etheridge, 1913. Limestone at escarpment north-east of Boonoo Ponds Creek. Hatton’s Corner, Yass River, N.S.W. Ludlow. Diagnosis. (Modified after Hill 1940, p. 409.) Cerioid or fasciculate Rugosa with septa developed weakly as low crests on dissepiments and tabulae. Tabulae typically complete, flat or sagging; dissepiments large, strongly elongate, steeply inclined towards corallite axis. Discussion. The genus Klamathastraea Merriam, 1972 was described from the Gazelle Formation (Unit 2), (?Ludlow) of the Klamath Mountains, California and was con- sidered to differ from Yassia by the latter having narrower tabulae and a lack of septa (Merriam 1972, p. 40). However, Yassia may show well-developed septa at the p 666 PALAEONTOLOGY, VOLUME 17 tabularium-dissepimentarium boundary, as can be seen from the type species, Y. enormis (Etheridge 1913, pi. VII, figs. 2, 3, and herein, PI. 95, figs. 7, 8). The septa continue as very low ridges over the dissepiments, as illustrated in external view by Etheridge (1913, pi. V), but in this region are rarely apparent in thin section. Hence there do not appear to be any significant differences between Klamathastraea and Yassia, and the two genera are most probably synonymous. Similarities between Yassia and Ketophyllum were noted by Jones (1932). However, Yassia differs in its colonial growth form and weaker septal development, particularly in the tabularium. Mictocystis shows closer similarities to Yassia, but may be dis- tinguished by its lack of corallite walls and stronger septal developments in the tabularium. It is possible that a form such as Yassia could have arisen from Micto- cystis by development of corallite walls and reduction of septa, or have been derived from Ketophyllum by formation of colonial coralla together with reduction of septa. Range. Upper Llandovery of the Siberian Platform; Lower Wenlock of the Siberian Platform and Tadzhikistan; Ludlow of N.S.W. and ?California. Yassia enormis (Etheridge, 1913) Plate 95, figs. 6-8 1913 Spongophyllum enorme Etheridge, p. 35, pis. IV-VII. 1930 Yassia enormis Jones, p. 36. 1932 Crinophyllum enorme Jones, p. 61, pi. IV, figs. 2, 3. 1940 Yassia enormis ; Hill, p. 409, pi. XIII, fig. 6a, b. 1963 Yassia enormis ; Ivanovskiy, p. Ill, pi. XXXII, fig. 2. Material. Lectotype (chosen Hill 1940) AM.F 8572. Paralectotypes AM.F 8769, 8770; thin sections AM 674, 847, 869. Distribution. Limestone equivalent to Bowspring Limestone, escarpment north-east of Boonoo Ponds Creek, Hatton’s Corner, Yass River, N.S.W. Ludlow. Diagnosis. Cerioid Yassia, with septa occurring as very low ridges on dissepiments, but only weakly developed in tabularium. Tabulae mainly flat and complete. Description. See Etheridge (1913) and Jones (1932). Remarks. The two other described species of Yassia appear closely similar to Y. enormis. Y. fasciculata Lavrusevich and Ivanovskiy in Ivanovskiy 1965, occurs in the Upper Llandovery of the Siberian Platform and Lower Wenlock (Horizon K) of Tadzhikistan. It differs mainly in its fasciculate growth form and the tabulae appear generally to be rather more incomplete (Ivanovskiy 1965, pi. XXXI, fig. 1 and Lavrusevich 1971, pi. XXIV, fig. 2b). Y. cystifera Ivanovskiy, 1965, from the Lower Wenlock of the Siberian Platform, was subsequently considered a variety of Y. enormis by Ivanovskiy (19706). It appears to have a narrower dissepimentarium and more incomplete tabulae than Y. enormis (Ivanovskiy 1965, pi. XXX, fig. 2). Lurther- more, Ivanovskiy (1965, p. 120) considered septa to be entirely lacking in Y. cystifera. k Klamathastraea ’ dilleri Merriam, 1972, appears to differ from Y. enormis in having smaller corallites and more strongly developed septa in the tabularium, although the McLEAN: CHONOPH YLLINID CORALS 667 septa do not extend more than one-third the radius of the tabularium (Merriam 1972, pi. 5, figs. 1-5). Y. enormis has been described also by Ivanovskiy (1963) from the Wenlock of the Siberian Platform. Acknowledgements. I am grateful to Dr. B. D. Webby for helpful discussions and critical review of the manuscript. Dr. G. Bischoff kindly made available unpublished information about conodont faunas from central N.S.W. The project was partly supported by a Commonwealth Post-graduate Research Studentship. REFERENCES bulvanker, E. z. 1952. Korally rugosa silura Podolii. Trudy vses. nauchno-issled. geol. Inst. 1-33. et al. 1960. Podklass Tetracoralla (Rugosa). In markovskiy, b. p. (ed.). Novye vidy drevnikh rasteniy i bespozvonochnykh SSSR ch. 7, Moscow. Gosgeoltekhizdat. (Vses nauchno-issled. geol. Inst.) pp. 220- 254. butler, a. j. 1937. A new species of Omphyma and some remarks on the Py enact is Phaulactis Group of Silurian Corals. Ann. Mag. nat. Hist. ser. 10, 19, 87-96. edwards, h. m. and haime, j. 1851. Monographic des polypiers fossiles des terrains palaeozoiques. Archs Mus. natn. Hist. nat. Paris, 5, 1-502. 1854. A monograph of the British Fossil Corals. Pt. 5. Corals from the Silurian Formation. Palaeontogr. Soc. (Monogr.), 245-322. ehlers, G. m. 1919. Heterolasma foerstei, a new genus and species of Tetracoralla from the Niagaran of Michigan. Am. J. Sci. 48, 461-467. etheridge, r., Jun. 1908. An undescribed Australian cystiphylhd— Mictocystis—ivom the Upper Silurian rocks of the Mount Canoblas district. Rec. Aust. Mus. 7, 18-20. 1913. A very remarkable species of Spongophyllum from the Upper Silurian rocks of New South Wales. Ibid. 10, 35-37. hill, d. 1940. The Silurian Rugosa of the Yass-Bowning District, N.S.W. Proc. Linn. Soc. N.S.W. 65, 388-420. 1956. Rugosa. In moore, r. c. (ed.). Treatise on Invertebrate Paleontology. Part F. Coelenterata. Lawrence, Kansas, pp. F233-F324. ivanovskiy, A. b. 1959. O nekotorikh kolonnalnikh rugosa s. r. Sukhaya Tunguska. Trudy sib. nauchno- issled. Inst. Geol. Geofiz. miner. Syr. 2, 135-143. 1962. Dva novikh roda siluriyskikh rugoz. Ibid. 23, 126-133. — 1963. Rugozy ordovika i silura Sibirskoy Platformi. 160 pp. Moscow. Nauka. (Akad. Nauk SSSR, Sib. Otd. Inst. Geol. Geofiz.) — 1965. Drevneyshie rugozy. 152 pp. Moscow. Nauka. (Akad. Nauk SSSR, Sib. Otd. Inst. Geol. Geofiz.) — 1970a. O sistematicheskom polozhenii nekotori rugoz ordovika i silura. Geol. & Geofiz. 1970, 2, 120-122. 1970A Stratigraficheskie i paleogeograficheskie compleksi rugoz na Sibirskoy Platforme. Ibid. 7, 17-18. jones, o. a. 1932. A Revision of the Australian Species of the Coral Genera Spongophyllum E. & H. and Endophyllum E. & H. with a Note on Aphrophyllum Smith. Proc. R. Soc. Qld. 44, 50-63. kaljo, d. l. (ed.). 1970. Silur Estonii. 343 pp. Tallin. Valgus. (Eesti NSV Tead. Akad. geol. Inst.) lambe, L. m. 1901. A Revision of the Canadian Palaeozoic Corals; the Madreporaria aporosa and the Madreporaria rugosa. Contr. Can. Palaeont. 5, 97-197. lang, w. D., smith, s. and THOMAS, H. D. 1940. Index of Palaeozoic Coral Genera. 231 pp. London. Brit. Mus. nat. Hist. lavrusevich, a. i. 1971. Rugozy rannego silura Zeravshano-Gissarskoy gornoy oblasti. In lavrusevich, a. i. (ed.). Paleontologiya i stratigrafiya. 3. Trudy Uprav. geol. sov. Minist. Tadzhik. SSR. pp. 38-136. lecompte, m. 1952. Madreporaires paleozoiques. In piveteau, j. (ed.). Trade de Paleontologie. Tom. I. Paris, pp. 419-538. link, a. g. 1970. Age and correlations of the Siluro-Devonian strata in the Yass Basin, New South Wales. J. geol. Soc. Aust. 16, 71 1-722. 668 PALAEONTOLOGY, VOLUME 17 merriam, c. w. 1972. Silurian Rugose Corals of the Klamath Mountains Region, California. Prof. Pap. U.S. geol. Surv. 738, 1-50. minato, m. 1961. Ontogenetic Study of some Silurian Corals of Gotland. Stockh. Contr. Geol. 8, 38-100. Northrop, s. a. 1939. Paleontology and Stratigraphy of the Silurian Rocks of the Port Daniels-Black Cape Region, Gaspe. Spec. Pap. geol. Soc. Am. 21, 1-302. packham, G. h. (ed.). 1969. The Geology of New South Wales. J. geol. Soc. Aust. 16, 1-654. — and stevens, n. c. 1955. The Palaeozoic Stratigraphy of Spring and Quarry Creeks, West of Orange, N.S.W. J. Proc. R. Soc. N.S.W. 86, 55-60. pocta, p. 1902. Anthozoaires et Alcyonaires. In barrende, j. Systeme Silurien du Centre de la Boheme, 8 (2), 1-347. rominger, c. 1876. Lower Peninsular 1873-1876. Palaeontology. Fossil Corals. Geol. Surv. Mich. 3 (2), 1-161. smith, s. 1945. Upper Devonian Corals of the Mackenzie River Region, Canada. Spec. Pap. geol. Soc. Am. 59, 1-126. soshkina, E. d. 1937. Korally verkhnego silura i nizhnego devona vostochnogo i zapadnogo sklonov Urala. Trudy paleozool. Inst. 6 (4), 1-112. — 1955. Fauna ordovika i gotlandiya nizhnego techeniya r. Podkamennaya Tunguska, ee ekologiya i stratigraficheskoe znachenie. Trudy paleont. Inst. 56, 93-196. — Dobrolubova, t. a. and KABAKOVICH, n. v. 1962. Podklass Tetrakoralla. In sokolov, b. s. (ed.). Osnovy Paleontologii. 2. Gubki , Arkheosiaty, Kishechnopolostnye, Chervi. Moscow. (Akad. Nauk SSSR), pp. 286-356. strelnikov, s. i. 1965. Ordovikskie i siluriyskie rugozy ostrovov Vaygacha i Dolgogo. Uchen. Zap. nauchno- issled. Inst. Geol. Arkt. (Paleont. Biostratigr.), 8, 24-57. — 1971. Znachenie rugoz dlya stratigrafii siluriyskikh otlozheniy Pripolyarnogo Urala i Gryady Chern- sheva. In Ivanovskiy, a. b. (ed.). Rugozy i stromatoporoidei paleozoya SSSR. Trudy II Vsesoyuznogo simpoziuma po izucheniyu iskopaemykh korallov SSSR. 2. Moscow. Nauka. (Akad. Nauk SSSR, Sib. Otd. Inst. Geol. Geofiz.), pp. 71-88. stumm, e. c. 1949. Revision of the Families and Genera of the Devonian Tetracorals. Mem. geol. Soc. Am. 40, 1-92. — 1965. Silurian and Devonian Corals of the Falls of the Ohio. Ibid. 93, 1-184. sussmilch, c. a. 1906. Note on the Silurian and Devonian Rocks Occurring to the West of the Canoblas Mountains near Orange, N.S.W. J. Proc. R. Soc. N.S.W. 40, 130-141. walker, d. b. 1959. Palaeozoic stratigraphy of the Area to the West of Borenore, N.S.W. Ibid. 93, 39-46. walliser, o. h. 1964. Conodonten des Silurs. Abh. hess. Landesamt. Bodenforsch. 41, 1-106. — 1971. Conodont Biostratigraphy of the Silurian of Europe. In sweet, w. c. and Bergstrom, s. m. (eds.). Symposium on Conodont Biostratigraphy. Mem. geol. Soc. Am. 127, 195-206. wang, h. c. 1944. Silurian Rugose Corals from Eastern and Northern Yunnan. Bull. geol. Soc. China , 24, 21-32. 1947. New Material of Silurian Rugose Corals from Yunnan. Ibid. 27, 171-192. — 1950. A revision of the Zoantharia Rugosa in the light of their minute skeletal structures. Phil. Trans. R. Soc. (B), 234 (611), 175-246. wedekind, R. 1927. Die Zoantharia Rugosa von Gotland (bes Nordgotland). Sver. geol. Unders. Afh. ser. Ca, no. 19, 1-94. zheltonogova, v. a. 1960. Siluriyskaya Sistema. Podklass Tetracoralla (Rugosa). Tetrakorally. In khalfin, l. l. (ed.). Biostratigrafiya paleozoya Sayano-Altayskoy gornoy oblasti. Tom. 2. Sredniy Paleozoy. Trudy sib. nauchno-issled. Inst. Geol. Geofiz. miner. Syr. 20, 74-86. — 1965. Znachenie rugoz dlya stratigrafi silura gornogo Altaya i Salaira. In sokolov, b. s. and Ivanov- skiy, a. b. (eds.). Rugozy paleozoya SSSR. Trudy I Vsesoyuznogo simpoziuma po izucheniya iskopaemykh korallov. 2. Moscow. Nauka. (Akad. Nauk SSSR, Sib. Otd. Inst. Geol. Geofiz.), pp. 33-44. R. A. MCLEAN Department of Geology and Geophysics University of Sydney Sydney 2006, N.S.W. Manuscript received 4 September 1973 Australia A NEW QUATERNARY OSTRACOD GENUS FROM ARGENTINA by ROBIN C. WHATLEY and TERESA DEL CARMEN CHOLICH Abstract. A new genus, Pampacythere (Limnocytheridae), is described from the so-called marine transgressions of Quaternary age in the cordon-litoral between Buenos Aires and Mar del Plata. The genus is based upon two species, P. multiperforata and P. solum. The relationships between Pampacythere and other members of the Limnocytheridae are discussed. It is concluded from an analysis of their faunal associations, with both ostracods and other inverte- brates, that the environment in which the two species lived was of brackish-water, probably oligohaline to meiomeso- haline. The Ostracoda herein described were encountered during a study of the micro- palaeontology of the Quaternary deposits of the cordon-litoral which occur along the coastal periphery of the Province of Buenos Aires, between the cities of Buenos Aires and Mar del Plata (text-fig. 1). The age and stratigraphy of the so-called marine deposits of the north-eastern and eastern parts of the Province of Buenos Aires remains uncertain, because there is no consensus of opinion about the position of the Plio-Pleistocene and the Pleistocene- Holocene boundaries (Pasqual and Fidalgo 1972). The problem is aggravated by the largely indigenous nature of both the invertebrate and vertebrate faunas which these deposits contain. Pasqual and Fidalgo (1972, table 1, pp. 212-213) give the different schemes outlined by Ameghino 1906, 1908, 1910; Rovereto 1914a, 19146; Roth 1921 ; Kraglievich 1934; Simpson 1940; Frenguelli 1957, and also present their own views (table 2, p. 220 and table 3, p. 23 1 ). We do not wish to enter into this controversy but believe that the majority of the deposits are Holocene, or certainly not older than late Pleistocene. Since D’Orbigny (1835-1847) and Darwin (1846), these deposits, which consist mainly of silts, sands, and shell banks exhibiting varying degrees of consolidation, have been accepted as being of marine origin, based principally upon their content of molluscan and diatomacean fossils. Various authors, however, including Frenguelli (1950, p. 63) have indicated that at least some of the deposits are of estuarine origin. From our studies of the ostracoda, and to a lesser extent the foraminifera, using the methods of Whatley and Kaye (1971, p. 311), it has become evident that the majority of the biocoenoses we have been able to recon- struct pertain to oligohaline-meiomesohaline brackish-water conditions. At certain levels and in certain areas, especially in the north, freshwater biocoenoses pre- dominate and biocoenoses indicating strongly brackish-water, approaching marine conditions (polyhaline), are also encountered. The levels in which these latter occur are almost always thin. They do not seem to occur north of about Canal No. 15, but increase in thickness and frequency towards Laguna Mar Chiquita (text-fig. 1). Over most of the area, the substantial number of marine ostracods present represent a thanatocoenosis because they do not show the population age structure indicative of a biocoenosis. The other invertebrates and diatoms were used by earlier workers [Palaeontology, Vol. 17, Part 3, 1974, pp. 669-684, pis. 96-97.] WHATLEY AND CHOLICH: PAMPACYTHERE 671 to label many of these deposits as marine. However, the great majority of these whose ecology is known from studies on living forms, are able to tolerate, or indeed flourish under brackish-water conditions. Localization of the type material. The type locality for Pampacythere is centred on Laguna Salada Grande, some 15 km to the north-east of the town of General Madariaga (text-fig. 1). Here, from the banks and from sediment cores taken through the bed of the large brackish-water lake of this name, five samples were processed by Lie. N. V. Dangaus. The nature and level of the samples is as follows: Sample Position in relation to water level of the lake Lithology PA/A + 0-30 metres Clayey silt with molluscan remains S.G. 7 — 0-30 metres Clayey silt with molluscan remains S.G. 5 — 1 -20 metres Clayey silt S.G. 1 — 2-50 metres Sandy silt S.G. 2 — 3-40 metres Sandy silt Pampacythere multiperforata and P. solum were both found in all of these samples except S.G. 1. These deposits are considered by Dangaus (verb. comm. 1973) to belong to the Piso Querandinense (Querandinian Stage) which is considered by most authors, including ourselves, to be of early Holocene age. Some authors, however, refer this stage to the upper part of the Pleistocene. One or both species have also been found in various other localities at more or less the same stratigraphical level and in similar sediments, especially in the area around Canal No. 15 and in cores taken through the bed of Laguna de Chascomous (text-fig. 1). DISCUSSION OF THE LIMNOCYTHERIDAE Including Pampacythere there are sixteen described genera of Limnocytheridae known to the authors. Bisulcocypris (Pinto and Sanguinetti 1958) being here con- sidered synonymous with Theriosynoecum Branson (1936) and the subgenus Limno- cy there ( Denticulocythere ) of Carbonnel and Ritzkowski (1969) being considered as warranting generic status. In addition there are a number of other taxa within the family which probably warrant reconsideration as new genera, such as the material described as "Gomphocy there' from the Neocomian of Argentina (Musacchio 1970). The other fifteen genera are : 1. Limnocythere Brady 1868. Tertiary (?01igocene) to Recent. Cosmopolitan. Fresh and slightly brackish-water. 2. Afrocythere Klie 1935. Recent. Africa. Freshwater. 3. Cytheridella Daday 1905. Tertiary (?01igocene) to Recent. South America, Caribbean (Van den Bold 1971), ?Europe (Carbonnel and Ritzkowski 1969). Freshwater to brackish-water (Van den Bold 1971). 4. Elpidium F. Muller 1880. Recent. South America. Freshwater. 5. Cordocythere Danielopol 1965. Oligocene (Carbonnel and Ritzkowski 1969) to Recent. Europe. Freshwater. 672 PALAEONTOLOGY, VOLUME 17 6. Gomphocythere Sars 1924. Pleistocene to Recent. Europe (Wicher 1957), Africa, South America, New Zealand, Australia. Freshwater. 7. Leucocythere Kaufmann 1892. Recent. Europe. Freshwater. 8. Metacypris Brady and Robertson 1870. Pleistocene and Recent. Cosmopolitan. Freshwater. 9. Neolimnocy there Delachaux 1928. Recent. South America. Freshwater. 10. Paracythereis Delachaux 1928. Recent. South America. Freshwater. 11. Pseudolimnocythere Klie 1938. Recent. Europe. Freshwater (Subterranean). 12. Theriosynoecum Branson 1936. Triassic (Gerry and Oertli 1967) to Lower Cretaceous. Cosmo- politan. Freshwater. 13. Paralimnocythere Carbonnel 1965. Miocene to Recent. Europe. Freshwater. 14. Denticulocythere Carbonnel and Ritzkowski 1969. Oligocene. Europe. Freshwater. 15. Cladarocy there Keen 1972. Eocene to Oligocene. Europe. Brackish-water. Only Theriosynoecum and the last two genera are extinct and only Limnocythere, Cytheridella, Cordacythere, and the last three genera have a fossil history extending back beyond the Pleistocene. If Theriosynoecum is indeed a member of the same family as the younger genera, one must expect eventually to encounter taxa bridging the gap between the Lower Cretaceous and the first appearance of species which can be assigned with reasonable certainty to living genera in the Eocene and Oligocene. Certain forms from the Upper Jurassic and Cretaceous, such as, for example Limno- cythere fragilis Martin (1940), bear a certain resemblance to later members of Limnocythere s.s. and require further examination to ascertain their true affinities. Also Keen (1972, p. 287) mentions the fact that Limnocythere sp. A (Bate 1965) from the Middle Jurassic (Bathonian) of Oxfordshire, may perhaps belong in Cladaro- cythere. It is notable that of the 16 genera in the family, 4 are Cosmopolitan, or virtually so ; 6 are confined to Europe and the remaining 6 to the Southern continents. Of these, Afrocythere is known only from Africa, and Cytheridella, Elpidium, Paracythereis, Neolimnocy there, and Pampacythere are confined to South America. Cytheridella in fact is also recorded from the Caribbean (Van den Bold 1971) and there is also a somewhat doubtful record of this genus from the Oligocene of France (Carbonnel and Ritzkowski 1969). The importance of South America as an area of considerable divergence within the Limnocytheridae is evident in these figures. Pampacythere seems to be most closely allied to the Cladarocythere, Gompho- cythere, Cytheridella, Elpidium group than to the other limnocytherids. Pinto and Sanguinetti (1958, 1962), Pinto and Purper (1970), and Danielopol (1969) have dis- cussed the nature of the generic relationships within the family. Detailed com- parisons between Pampacythere and other genera of the Limnocytheridae are difficult to make since modern studies, particularly of the pores, are lacking for the other genera. ECOLOGY OF PAMPACYTHERE At the type locality, the genus occurs in association with the following ostracods which exhibit a population age structure indicative of a biocoenosis. These are in approximate order of abundance: Cyprideis multidentata Hartmann 1955, Callisto- cy there nucleoperiscum Whatley and Moguilevsky (in press), Perissocytheridea sp. cf. P. krommellbeini Pinto and Ornellas 1970, Limnocythere neotropica Klie 1934, Limnocythere sp., and Cyprideis saetosa Hartmann 1955. WHATLEY AND CHOLICH: PAM PACYTHERE 673 All these species are typical of brackish-water environments today, although the presence of the two Limnocythere species would seem to indicate a substantial degree of freshwater influence. Cyprideis multidentata (which the authors consider to be synonymous with C. riograndensis Pinto and Ornellas 1965) was first described from essentially brackish-water environments along the coast of Brazil, and later by Pinto and Ornellas (1965) from a brackish-water environment in southern Brazil with a salinity range of 6- 1 0-29- 1 1 °/00- Pinto and Ornellas (1970) record Perissocytheridea krommelbeini from the same locality and salinity range, and one of the present authors (R. C. W.) has recovered this species from a number of brackish-water localities in the Argentine. Callistocythere nucleoperiscum , whilst common in marine littoral environments, also penetrates into brackish-waters in substantial numbers, in the Argentine at the present day. Also occurring with the above fauna are rare freshwater elements such as Ilyocypris gibba (Ramdohr 1808), Chlamydotheca alegrensis Tressler 1949, Cyprinotus similis Wierzejski 1892, Cyprinotus incongruens (Ramdohr 1808), and Cypridopsis assimilis Sars 1901 (Ramirez 1967), all of which present a population age structure indicative of a thanatocoenosis rather than a biocoenosis. Marine elements include species of the following genera which also, since they occur rarely and without more than one, or at most two, growth stages are considered as being of an introduced nature: Xestoleberis , Loxoconcha , P at agonacy there, Procythereis , Leptocythere, Cytheretta , ITriginglymus, Semieytherura , Microcytheridea , Hemicytherura , Cytheropteron , Cushmanidea, Hu/ingsina , Munseyella, Parakrithe/la , Paradoxostoma, and Pel/u- cistoma. Foraminifera include, in approximate order of abundance, the following species : Elphidium discoidale (d’Orbigny) 1840, Ammonia becarii parkinsoniana (d’Orbigny) 1840, Elphidium gunteri Cole 1931, E. galverstonensis Kornfield 1931, and E. advenum (Cushman 1922). In some samples Buce/la frigida Cushman 1921 also occurs abundantly. This association of Foraminifera is thought to represent brackish-water. The molluscs, which at the type locality occur mainly as embryos or as fragments of larger shells are also, according to Ageitos de Castellanos (1967, verb. comm. 1973), in the association in which they occur, indicative of brackish-water. They are: Littoridina parchappii (d’Orbigny) 1835, L. australis (d’Orbigny) 1835, Mactra isabelleana d’Orbigny 1846, Tagelus plebeius (Solander) 1786, and Eurodona mac- troides Daudin 1802. Taken together, all this evidence is overwhelmingly indicative of a brackish-water environment, probably within the oligo-meiomesohaline range. Detailed reconstructions of the various palaeoenvironments we have encountered will be published by the Ministry of Public Works of the Province of Buenos Aires (LEMIT). Here we can postulate that during the deposition of the sediments from which we have recovered Pampacythere, in the area of Laguna Salada Grande, a lagoonal environment prevailed. Our reconstruction implies the existence to the north of an estuary, probably very similar to that of the present-day River Plate, and in this area being of polyhaline salinity, separated from a shallow lagoon by sand or shell banks. It is evident that the lagoon had freshwater draining into it from the south and west and was periodically subject to the influence of more highly saline waters from the closely adjacent estuary. From an analysis of the Argentine Recent marine Ostracoda being carried out by 674 PALAEONTOLOGY, VOLUME 17 one of the authors (R. C. W.), we conclude, by comparison with those which occur in these deposits, that slightly warmer waters prevailed than at the present day. Most of the marine species recovered, today live more commonly along the southern coast of Brazil than along the north-eastern coast of the Argentine. SYSTEMATIC DESCRIPTIONS Subclass ostracoda Latreille, 1806 Order podocopida Muller, 1894 Suborder podocopina Sars, 1 866 Superfamily cytheracea Baird, 1850 Family limnocytheridae Klie, 1938 Genus pampacythere gen. nov. Type species. Pampacythere multiperforata sp. nov. Diagnosis. Thin shelled. Dimorphic, with males longer and proportionally less high and less tumid than females. Left valve larger than right. Anterior margin rounded. Posterior margin bluntly pointed at or below mid-height. Dorsal margin straight, or posteriorly arched in females. Not strongly flattened ventrally. With two weak antero-ventrally directed sulci on the antero-lateral surface, behind the anterior cardinal angle and just in front of mid-length respectively. Surface smooth to ‘wrinkled’. Normal pores large, numerous and comprising 3 or 4 variants of simple open and sieve-type pores, occurring in chain-like groups of 2 or 3 pores subparallel to the margins of the carapace. Hinge modified lophodont with a posteriorly lobate median element in the left valve. Adductor muscle scars comprise a vertical or slightly oblique line of 4 scars with a more or less heart-shaped frontal scar. Two well- developed mandibular scars occur en echelon below and anterior to the adductors and a number of dorsal scars above. Discussion. From Gomphocy there, Pampacythere differs in outline and in lacking a well-defined flattening of the ventral surface. From Cytheridella it differs in being more posteriorly acuminate and more rounded anteriorly, in the nature of its hinge- ment and in lacking an interior vestibule and an internal elevation in the region of the muscle scars. From Elpidium it differs in shape and outline, in lacking a strongly flattened ventral surface and substantially in terms of hingement. From Cladarocy- there it differs in lacking reticulate ornament, crenulate terminal, and a completely crenulate median hinge element and in possessing an undivided frontal scar. In overall shape, the nature of the sexual dimorphism, the nature of the inner lamella and in their ecology, the two genera are very similar. From all these genera, and apparently from all other members of the family, Pampacythere differs in possessing numerous very large normal pore canals, both of simple and sieve type, arranged in ‘chains’ of groups of two or three pores forming subparallel rows. It also differs in exhibiting a hinge with a strongly lobate postero-median element in the left valve and in being generally acuminate posteriorly. Apart from Cladarocy there the only other members of the family which bear any resemblance to Pampacythere in terms of shape and outline, are certain species of Limnocythere, such as L. trapeziformis Staplin 1963 and L. ornata wabashensis Staplin 1963 from the Pleistocene of Illinois. This WHATLEY AND CHOLICH: PAMPACYTHERE 675 resemblance, however, is thought to be totally superficial and coincidental. Pampa- cythere bears a strong, but quite superficial, resemblance to certain species of Para- cyprideis Klie (1930), such as P. rarefistulosa Linenklaus from the Oligocene of Belgium. This resemblance is in shape and also in the similarity, size, and disposition of the normal pores. However, Pampacythere does not exhibit the large vestibule characteristic of Paracyprideis, and also differs in hingement and musculature. Considering the fact that very few Quaternary invertebrates are extinct, Pampa- cythere is very probably a living genus. The authors, however, despite collecting from numerous localities with similar environmental conditions to those under which we suppose the fossil material to have lived, have failed to encounter it. Pampacythere multiperforata sp. nov. Plate 96, figs. 1-21 Holotype. Female right valve, MLP 11954. Derivato nominis. From the large number of normal pores which perforate the carapace of this species. Type level and locality. Sample S.G. 7, Laguna Salada Grange, Province of Buenos Aires, Argentina. Material. Some 200 specimens, mainly open valves. Dimensions. Length Height Width (mm) (mm) (mm) Holotype 0-77 0-38 014 Paratypes (from the same sample as the Holotype) Female right valve. MLP 1 1955 0-77 0-38 015 Male left valve. MLP 1 1956 0-89 0-41 0-20 Female left valve. MLP 1 1958 0-73 0-36 0-17 Male right valve. MLP 1 1959 0-75 0-37 014 Female right valve. MLP 1 1960 0-69 0-35 016 Male right valve. MLP 11962 0-72 0-35 015 Diagnosis. Pampacythere with carapace strongly perforated by four different types of normal pore canals arranged in groups of two or three pores to form chain-like rows subparallel to the margins of the valve. Dorsal margin straight in both sexes; acuminate postero-ventrally. Hinge with strongly crenulate to lobate postero-median element in the left valve, above which is a shelf-like accommodation groove. Description. Large. Subrectangular. Males longer and proportionally less high than females. Left valve larger than right. Overlap (only seen in juveniles since adult carapaces have not been found), occurs postero-ventrally and along the anterior part of the dorsal margin. Anterior margin broadly rounded with extremity at about mid- height; posterior margin strongly asymmetrical, the ventral part being narrowly rounded whilst the postero-dorsal slope is long and straight or slightly concave; extremity below mid-height. Dorsal margin long and straight or with slight concavity behind mid-length. Ventral margin biconvex with median concavity overhung by the slight ventral tumidity of the valve. Posterior cardinal angle more strongly marked than anterior. Greatest length below mid-height ; greatest height usually at the anterior cardinal angle, although this is only a little greater than that at the posterior and in 676 PALAEONTOLOGY, VOLUME 17 some specimens the height is the same at each angle. Surface strongly punctate, par- ticularly posteriorly where the undulations between the punctae impart to the valve surface a ‘wrinkled’ appearance, although the surface between the punctae is generally smooth. A weak ridge extends from the anterior cardinal angle to an antero-ventral position, paralleling the anterior margin. The immediate antero-marginal border is strongly laterally compressed. Two feeble oblique sulci occur, behind the anterior cardinal angle and mid-dorsally respectively, which both extend antero-ventrally for a short distance. Normal pores comprise four types: 1, single open pores without lip (PI. 96, fig. 6); 2, semicircular sieve plate with small well-spaced circular openings (PI. 96, fig. 3); 3, semicircular to irregular sieve plates of very closely spaced large polygonal holes resembling honeycomb; 4, sieve plate of type 2 but with a single larger open pore perforating the plate at one edge (PI. 96, fig. 8), or more or less centrally (PI. 96, fig. 2). These four pore types correspond approximately to types A, B, B, and C of Puri and Dickau (1969) respectively. Types 2 and 3 seem to be rarer than the others and also differ in not occurring in a depression as do the others. Pores of types 2-4 occur flush with the outer lateral surface of the valve, and are open internally (PI. 96, fig. 7). In transparent specimens the pores can be seen to occur in groups of two or three, and sometimes more, in chain-like rows which, especially antero-ventrally and ventrally, are subparallel to the valve margin. All pores are relatively large and occur in greatest density in the posterior half of the valve. Approximately 120 pores occur in each valve and their position is very constant between different specimens. Eye spot absent. In some specimens there is a star-shaped depression on the outer lateral surface in the region of the adductor muscle scars (PI. 96, fig. 1). Inner lamella relatively narrow, and only slightly wider postero- ventrally than elsewhere. Line of concresence and inner lamella coincide throughout. Radial pore canals thin and regularly spaced. Anteriorly there are between 14 and EXPLANATION OF PLATE 96 Figs. 1-21. Pampacythere multiperforata gen. et. sp. nov. 1-3, holotype MLP 11954, female R.V. Stereo- scan micrographs, external views. 1 , outer lateral view, x 120. 2, sieve plate of type 4 (a large open pore perforating subcentrally a sieve plate of type 2), x 6000. 3, type 2 sieve plate, x 120. 4, paratype MLP 1 1955, female R.V. Stereoscan micrograph, internal view, x 120. 5, 6, paratype MLP 1 1956, male L.V. Stereoscan micrographs, external views. 5, outer lateral view, x 120. 6, type 1 open pore, x 2000. 7, 11, 19, paratype MLP 1 1957, female R.V. Stereoscan micrographs, internal views. 7, sieve type pore, x 6000. 1 1, internal view, x 120. 19, adductor and frontal scars, x 600. 8, paratype MLP 1 1958, female L.V. Stereoscan micrograph, external view. Type 4 sieve plate (sieve plate of type 2 with open pore to one side), x 12 000. 9, paratype MLP 11959, male R.V. Stereoscan micrograph, outer lateral view, x 120. 10, paratype MLP 1 1960, female R.V. Stereoscan micrograph, outer lateral view, x 120. 12, 13, para- type MLP 11961, female L.V. Stereoscan micrographs, internal views. 12, internal view, x 120. 13, posterior terminal and postero-median hinge elements, x 12 000. 14, paratype MLP 1 1962, male R.V. Stereoscan micrograph, internal view, x 120. 15, 16, paratype MLP 11963, female R.V. Transmitted light photo-micrographs. 15, external view to show distribution of the normal pore canals, x240. 16, internal view of anterior margin, x240. 17, paratype MLP 11964, male L.V. Transmitted light photo-micrograph, internal view of posterior margin, x240. 18, paratype MLP 11965, male R.V. Transmitted light photo-micrograph, external view of anterior margin, x240. 20, paratype MLP 11966, female L.V. Stereoscan micrograph, outer lateral view, x 120. 21, paratype MLP 11967, male L.V. Transmitted light photo-micrograph, internal view, x 120. PLATE 96 WHATLEY and CHOLICH, Pamparythere 678 PALAEONTOLOGY, VOLUME 17 17, together with 2 or 3 false canals which sometimes cross the true canals. Between 12 and 15 canals occur posteriorly of which those in the postero-ventral part are paired or bunched (PL 96, fig. 17), and diverge in the form of an inverted ‘V’. Selvage feeble and sub-peripheral. Ventral locking structures poorly developed. Hinge modified lophodont. In the right valve the anterior terminal element is a weak smooth bar, apparently formed by the edge of the valve being bent inwards. The posterior terminal element is similar, but shorter and weaker. The median element is a narrow smooth groove, well marked antero-medianly but becoming weaker posteriorly until immediately behind mid-length, where it widens and deepens and becomes loculate. This groove is somewhat overhung dorsally but is largely open to the interior ventrally . The dorsal edge of the valve is obliquely crenulate. In the left valve the terminal elements are completely open to the interior. The median element is, in its anterior part, a thin smooth bar which widens and becomes divided into four or five large lobes posteriorly (PI. 96, fig. 13). Above the median element is a narrow shelf-like accommodation groove. Adductor muscle scars comprise a slightly oblique line of four scars of which the two central ones are more elongate than the dorsal and ventral scars. Anteriorly and more dorsal to the most dorsal adductor, is a ‘heart’-shaped frontal scar. Below this frontal scar and more ventral than the most ventral adductor, occur two mandibular scars arranged en echelon. Dorsal and antero-dorsal to the adductors occur four or five dorsal scars. Ontogeny. (Specimens Instar —1 1 1 right valves 8 left valves Instar — 2 3 right valves 4 left valves Instar —3 1 left valve sample S.G. 7.) Length Range 0-610-68 Mean 0-64 Range 0-65-0-69 Mean 0-66 Range 0-58-0-60 Mean 0-59 Range 0-57 0-60 Mean 0-59 0-51 Height Range 0-305-0-35 Mean 0-32 Range 0-30-0-33 Mean 0-3 1 Range 0-29 0-305 Mean 0-30 Range 0-28-0-32 Mean 0-30 0-27 Pampacythere solum sp. nov. Plate 97, figs. 1-16 Holotype. Female right valve, MLP 11968. Derivato nominis. From the resemblance of the female of this species to the sole of a shoe. Type level and locality. Sample S.G. 7, Laguna Salada Grande, Province of Buenos Aires, Argentina. Material. Some 150 specimens, mostly valves. Dimensions. Length Height Width (mm) (mm) (mm) Holotype 0-75 0-36 0-14 Paratypes (from the same sample as the Holotype) Male left valve. MLP 1 1969 0-72 0-35 0-15 Male left valve. MLP 11971 0-69 0-35 0 12 Female left valve. MLP 11976 0-76 0-37 0-155 WHATLEY AND CHOLICH: PAMPACYTHERE 679 Diagnosis. Pampacythere with strong dimorphism expressed in the degree of arching of the dorsal margin. In the female, the posterior part of the dorsal margin is strongly arched; in the male it is straight or virtually so. Surface smooth to ‘wrinkled’ and not strongly punctate. Normal pores of three types and arranged in subparallel chain- like rows formed of groups of normally three pores. Left valve without accommoda- tion groove. Frontal scar ‘heart’-shaped in the horizontal rather than the vertical sense. Description. Medium to large. Males more elongate and less high than females. Left valve larger than right. Anterior margin broadly rounded with extremity at about mid-height. Posterior margin rounded with a convex postero-ventral and a straight or slightly concave postero-dorsal slope; apex at or just below mid-height. Dorsal margin long and straight in left valves, especially males; in right valves, especially females, it is strongly arched in its posterior part. Ventral margin biconvex about a concavity which is antero-median in right valves and median in left valves. Greatest length at about mid-height; greatest height at the posterior cardinal angle; greatest width in the posterior third of the valve. Anterior cardinal angle subrounded especially in right valves, posterior cardinal angle strongly marked . Surface smooth to ‘wrinkled’ . Eye spot absent. A poorly defined rib extends from the anterior cardinal angle sub- parallel to the anterior margin and separates the strongly laterally compressed anterior marginal border from the more elevated lateral surface. The anterior marginal border bears 8-10 weakly developed short tubercules which extend on to the flange. Two short and poorly defined oblique sulci extend antero-ventrally from just behind the anterior cardinal angle and mid-dorsally respectively. The remainder of the shell surface is smooth or ‘wrinkled’-up between the pore depressions. Normal pores occur in groups of normally three closely spaced pores of which the centre one is often the smallest. In some parts of the shell they appear to be fused and present a ‘dumb- bell’-like appearance (PI. 97, fig. 8). Some pores are single, but all are orientated in more or less distinct chain-like rows which ventrally are sub-parallel to the ventral margin and which postero-medianly are sub-vertical. The following types of normal pores have been observed: 1, single open pores without lips (PI. 97, fig. 12); 2, semi- circular to oval sieve plates perforated by mostly circular well-spaced holes, the size of which often diminishes towards the periphery (PI. 97, fig. 12); 3, semi-circular, oval or ‘butterfly’-shaped sieve plates bearing a single, sub-central or peripheral, larger open pore (PI. 97, figs. 9, 12). These pores correspond approximately to types A, B. and C of Puri and Dickau (1969) respectively. Pore type 3 always occurs in a depression, whereas the remainder are flush with the outer lateral surface of the valve. Type 2 pores seem to occur frequently in association with the ‘butterfly’-shaped pores of type 3 (PI. 97, fig. 12) and this figure also shows the frequent association of type 2 with type 1 pores (type D of Puri and Dickau 1969). All the sieve type pores have the sieve plate on the outer lateral surface and are open internally (PI. 97, fig. 16). Some specimens exhibit an irregular star-shaped depression on the outer lateral sur- face in the position of the adductor muscle scars. Inner lamella relatively narrow; widest postero-ventrally. Line of concrescence and inner lamella coincident through- out. Radial pore canals thin and evenly spaced. There are between fifteen and seventeen anteriorly which are almost all slightly concave upwards, and with at least 680 PALAEONTOLOGY, VOLUME 17 one canal in all individuals bifurcate. Seven canals occur posteriorly and a similar number postero-ventrally. Selvage sub-peripheral, becoming peripheral mid- ventrally. Hinge modified lophodont. In the right valve the terminal elements are smooth, weak elongate ridges formed by the infolding of the edge of the valve, the posterior being the more strongly developed. The median element is a smooth narrow bar, deepest and widest at its extremities especially posteriorly. In the left valve the terminal elements are smooth sockets, the anterior of which is open internally, the posterior being more strongly developed and partially enclosed to the interior. The median element is a smooth or weakly and irregularly crenulate bar. It is slightly expanded antero-distally and posteriorly it becomes strongly lobate. Accommoda- tion groove absent. Muscle scars similar to those of P. multidentata except that the frontal scar is ‘heart’-shaped in the horizontal rather than the vertical sense, in that the indentation occurs anteriorly and not dorsally. Ontogeny. (Specimens from sample S.G. 2.) Adult 9 right valves 10 left valves Instar —1 13 right valves 1 0 left valves Instar —2 1 right valve 2 left valves Length (mm) Range 0-68-0-85 Mean 0-78 Range 0-69-0-95 Mean 0-74 Range 0-62-0-685 Mean 0-66 Range 0-62-0-69 Mean 0-66 0-56 Range 0-57-0-59 Mean 0-58 Height (mm) Range 0-32-0-45 Mean 0-38 Range 0-32-0-45 Mean 0-39 Range 0-29-0-345 Mean 0-32 Range 0-30-0-345 Mean 0-32 0-28 Range 0-26-0-285 Mean 0-27 Remarks. Pampacythere solum differs from P. multiperforata in the following major characteristics: it is smoother and less strongly punctate; it is arched dorsally in the female; the left valve does not have an accommodation groove; the frontal scar is ‘heart’-shaped in the horizontal rather than the vertical sense. Deposition of material. Holotypes and paratypes are deposited in the collections of EXPLANATION OF PLATE 97 Figs. 1-16. Pampacythere solum gen. et. sp. nov. 1, paratype MLP 11969, male L.V. Stereoscan micro- graph, outer lateral view, x 120. 2, 9, 12, holotype MLP 11968, female R.V. Stereoscan micrographs, external views. 2, outer lateral view, x 120. 9, type 3 sieve plate, x 6000. 12, normal pores of types 1, 2, and 3, x 2000. 3, 5, 6, and 10, paratype MLP 11970, male L.V. Stereoscan micrographs, internal views. 3, internal view, x 120. 5, adductor, frontal and mandibular scars, x 600. 6, anterior part of valve, x400. 10, anterior part of hinge, x600. 4, 13, paratype MLP 11971, male L.V. Stereoscan micrographs, internal views. 4, internal view, x 120. 13, posterior terminal and postero-median hinge elements, x 1200. 7, paratype MLP 11972, female R.V. Stereoscan micrographs, outer lateral view, x 120. 8, 15, paratype MLP 11973, male L.V. Transmitted light photo-micrographs, external views. 8, outer lateral view to show distribution of the normal pores, x 120. 15, posterior margin, x240. 11, paratype MLP 11974, female R.V. Transmitted light photo-micrograph, external view of anterior margin, x 240. 14, paratype MLP 1 1975, female R.V. Transmitted light photo-micrograph of anterior part of valve, external view, x240. 16, paratype MLP 11976, female L.V. Stereoscan micrograph, internal view of two sieve type pores and one open normal pore, x 3000. PLATE 97 WHATLEY and CHOLICH, Pampacythere 682 PALAEONTOLOGY, VOLUME 17 the Catedra de Micropaleontologia, Division de Paleozoologia, Museo de La Plata, to which the numbers prefixed MLP apply. Topotype collections have been sent to the British Museum (Natural History) and the United States National Museum, Washington. Acknowledgements. The authors acknowledge the help of the following from the Facultad de Ciencias Naturales y Museo, Universidad Nacional de La Plata: Lie. Nauris Dangaus, Lie. Mirta Lagreca, Dra. Zulma Ageitos de Castellanos, Lie. Celia Gaillard de Dangaus, Lie. Martha Ferrario, and Lie. Hugo Valicenti. We also thank the Consejo Nacional de Investigaciones Cientificas y Tecnicas of the Argentine Republic for permission to use their Scanning Electron Microscope and for financial support for one of the authors (R. C. W.), and the Laboratorio de Ensayo de Materiales e Investigaciones Tecnologicas, of the Ministry of Public Works of the Province of Buenos Aires, also for financial assistance. We are grateful to Drs. Arturo Amos and Alberto Riccardi for their constructive criticisms, and to Dra. Irma Lucia Costa for assistance in many matters. REFERENCES ageitos de Castellanos, z. J. 1967. Catalogo de los Moluscos Marinos Bonaerenses. Anales de la Comision de Investigacion de la Provincia de Buenos Aires, 8, 1-365. ameghino, F. 1875. Ensayos para servir de base a un estudio de la formacion Pampeana. Mercedes, Argentina. 1881. La formacion Pampeana. Paris. — 1906. Les formations sedimentairs du Cretace Superieur de Patagonia avec un parallel entre leurs faunes mamalogiques et celles de l’ancien continent. An. Mus. nac. B. Aires (Ser. 3u), 8, 1 568. — 1908. Las formaciones sedimentarios de la region litoral de Mar del Plata y Chapadmalal. An. Mus. nac. B. Aires, 10, 343-428. — 1910. Geologia, Paleografica, Paleontologia y Antropologia de la Republica Argentina. Diario ‘ La Nacion' , Publ. del Centenario, 174-180. Buenos Aires. bate, r. h. 1965. Freshwater ostracods from the Bathonian of Oxfordshire. Palaeontology, 8, 109-1 1 1 . bordas, a. f. 1957. Argumentos paleontologicos y climaticos para establecer relaciones estratigraficas del Pleistoceno-Holoceno de Argentina. Ameghiniana. 1, 15-79. brady, G. s. 1868. A monograph of the Recent British Ostracoda. Trans. Linn. Soc. Lond. 26, 353-459. — and robertson, d. 1870. The Ostracoda and Foraminifera of Tidal Rivers. With an analysis and description of the Foraminifera by h. b. brady. Ann. Mag. nat. Hist. (Ser. 4), 31, 1-33. branson, c. c. 1936. New Name for a Morrison Ostracode Genus. J. Paleont. 10, 323. brehm, v. 1924. Uber Sudamerikanische Ostracoden des Zoologischen Staatsinstitutes u Museum in Hamburg. Zool. Anz. 108, 74-84. 1937. Weitere Mitteilungen fiber die Sfisswasser fauna Urugauys. Ibid. 120, 305-306. 1939. Zur Systematischen und Tiergeographischen der als Limnicythere problematica beschriebenen Gomphocy there. Ibid. 125, 190-195. carbonnel, g. 1965. Sur un nouveau genre (Paralimnocy there) et une nouvelle espece (P. boueigensis ) d’ostracodes du Tortonien. Arch. Sc. Geneve. 18, 146-152. — and ritzkowski, s. 1969. Ostracodes Lacustres de l’Oligocene (Melanienton) de la Hesse— (Alle- magne). Ibid. 22, 55-82. cole, w. s. 1931. The Pliocene and Pleistocene Foraminifera of Florida. Bull. Lla. geol. Surv. 6. cortelezzi, c. r. and lerman, j. c. 1971. Estudio de las Formaciones de la costa Atlantica de la Provincia de Buenos Aires. Lab. de Ensayo de Materiales e Investigaciones Tecnicas, Ministerio de Obras Publicas de la Provincia de Buenos Aires (Ser. 2), 178, 135-164. Cushman, J. a. 1918-1931. The Foraminifera of the Atlantic Ocean. Parts. 1-8. Washington. daday, e. 1905. Untersuchungen fiber die Sfisswasser-Mikrofauna Paraguays. Zoology, 44, 234-270. danielopol, d. l. 1965. Nouvelles donnees sur les Ostracodes d'eau douce de Roumanie: Cordocythere phreaticola n.g., n.sp., Limnocytherini et Matacyprini nouvelles tribus de la sous famille Limnocytherinae, Sars 1925. Ann. Limn. 1, 443-468. 1969. Notes sur la morphologie et la systematique de la sous famille Limnocytherinae Sars (Crustacea, Ostracoda). Ann. Speleol. 24, 131-142. WHATLEY AND CHOLICH: PAMPACYTHERE 683 darwin, c. 1846. Geological observations on South America. 279 pp. London. delachaux, T. 1928. Faune invertebree d’eau douce de la hauts plateaux du Perou. Ostracodes. Bull. Soc. neuchatel. Sci. nat. (Ser. 1), 45-77. d’orbigny, a. 1835-1847. Voyage dans LAmerique Meridionale. Paris and Strasbourg. ferguson, e. 1967. Three new species of freshwater Ostracods (Crustacea) from Argentina. Notul. nat. 405, 1-7. frenguelli, j. 1945. El Piso Platense. Rev. Mus. La Plata (n.s.), 2, 287-31 1 . — 1950. Rasgos generales de la morfologia y la geologia de la Provincia de Buenos Aires. Lab. de Ensayo de Materiales e Investigaciones Tecnicas, Ministerio de Obras Publicas de la Provincia de Buenos Aires (Ser. 2), 33, 1-72. 1957. Neozoico. In Geografia de la Repubhca Argentina. Gaea , 2, 1-219. gerry, e. and oertli, h. j. 1967. Bisulcocvpris Itriassica n.sp. (Crust., Ostrac.) from Israel. Bull. Centre Rech. Pau-SNPA, 1, 375-381. hartmann, G. 1955. Neue marine Ostracoden de Familie Cypridae und der Subfamilie Cytherideinae der Familie Cytheridae aus Brasilien. Zool. Anz. 154, 109-127. — 1959. Ostracodos de aguas salobres de America latina y su significacion para la investigation paleon- tologica. Ados y trabajos del ler. Congreso Sudamericano de Zoologia Museo de La Plata, 2, 169-176. kaufmann, a. 1892. Uber die Gattung Acanthopus Vernet und einer neue Siisswassercytheride. Zool. Anz. 15, 393-395. keen, m. c. 1972. The Sannoisian and some other Upper Palaeogene Ostracoda from North-West Europe. Palaeontology, 15, 267-325. klie, w. 1930. Ostracoden aus dem paraguayischen Teile des Gran Chaco. Arch.f Hydrob. 22, 221-258. — 1934. Zur kenntnis der Ostracoden-Gattung Lymnocythere. Arch. Naturgesch. n.f. 3, 534-544. 1935. Siisswasser Ostracoden aus Uruguay. Arch. Hydrob. 29, 282-295. — 1938. Ostracoda, Muschelkrebse. In dahl, f. (ed.). Die Tierwelt Deutschlands und der angrenzenden Meeresteile nach ihren Merkmalen und nach ihrer Lebensweise, 34 (3), 1-230. kraglievich, l. 1934. La antigiiedad pliocena de las faunas de Monte Hermoso y Chapadmalal deducidos de su comparacion con las que procedieron y sucedieron. El Siglo Illustrado. Montivideo. martin, g. p. R. 1940. Ostracoden des norddeutschen Purbeckund Wealden. Senckenberg. leth. 22, 275-361. mckenzie, k. g. 1968. Relevance of a Freshwater Cytnerid (Crustacea, Ostracoda) to the Continental Drift Hypothesis. Nature, Lond. 220, 806-808. muller, f. 1880. Wasserthiere in Baumwipfeln Elpidium bromelarium Kosmos, 6, 386-388. — 1881. Descripcao de Elpidium bromelarium Crustaceo da Familia dos Cytherideos. Arch. Mus. nac. Rio de J. 4, 27-34. musacchio, e. a. 1970. Ostracodos de las Superfamilias Cytheracea y Darwinulacea de la Formation La Amarga (Cretacio Inferior) en la Provincia de Neuquen Republica Argentina. Rev. Mus. La Plata, 7, 301-316. olivier, s. 1955. A few aspects of the regional Limnology of the Province of Buenos Aires. Proc. Int. Assoc. Limnol. 12, 296-301. parodiz, j. j. 1962. Los Moluscos marinos del Pleistoceno Rioplatense. Comm. Soc. Malacogica del Uruguay, 1, 29-46. pasqual, r. et al. 1966. Las edades del cenozoico mamalifero de la Argentina, con especial atencion a aquellas del territorio bonaerense. Ann. Comm. Invest. Prov. Buenos Aires, 6, 165-195. — and fidalgo, f. 1972. The problem of the Plio-Pleistocene boundary in Argentine (South America). Int. Coll, on the problem : ‘ The Boundary between Neogene and Quaternary'. Collection of Papers, 2, 205- 262. Moscow. petrovski, t. k. 1969. Zwei neue Limnocythere- Arten aus Mazedonien (Crust. -Ost.). Acta. mus. maced. Sci. nat. 12, 1-8. pinto, i. d. and ornellas, l. 1965. A new brackishwater ostracode Cyprideis riograndensis Pinto et Ornellas, sp. nov. from Southern Brazil and its ontogenetic carapace development. Publo. Esp. Esc. Geol. P. Alegre. 8, 1-80. — 1970. A new brackishwater ostracode, Perissocytheridea krommelbeini Pinto and Ornellas sp. nov. from Southern Brazil. Ibid. 20, 1-19. — and purper, i. 1970. A neotype for Elpidium bromelarium Muller 1880 (type species for the genus) and a revision of the genus Elpidium. (Ostracoda). Ibid. 19, 1-23. 684 PALAEONTOLOGY, VOLUME 17 pinto, i. d. and sanguinetti, y. t. 1958. Bisulcocypris , a new Mesozoic Genus and Preliminary Note about its Relation with Metacypris and allied forms. Bol. Soc. Brasil, geol. 7, 75-90. — 1962. A complete revision of the genera Bisulcocypris and Theriosynoecum (Ostracoda) with a world geographical and stratigraphical distribution (including Metacypris , Elpidium , Gomphocy there, and Cytheridella). Publo. Esp. Esc. Geol. P. Alegre. 4, 1-165. puri, h. s. and dickau, b. e. 1969. Use of normal pores in taxonomy of Ostracoda. Trans. Gulf Coast Assn. Geol. Soc. 19, 353-367. ramdohr, F. a. 1808. Uber die Gattung Cypris Muller und drei zu derselben gehorige neue Arten. Ges. Naturforsch. Freunde Berlin, Magazin nuesten Entdeckungen gesammte Naturkunde. 2, 83-93. ramirez, f. c. 1967. Ostracodos de Lagunas de la Provincia de Buenos Aires. Rev. Mus. La Plata, 10, 5-54. Richards, H. G. 1962. Studies on the Marine Pleistocene. Trans. Amer. Philos. Soc. 52, 1141. — and craig, J. 1963. Pleistocene molluscs from the Continental Shelf off Argentina. In fray, c. Pleisto- cene Sedimentation and Fauna of the Argentine Shelf. Proc. Acad. Nat. Sci. Phila. 115, 113-152. roth, s. 1921. Investigaciones geologicos en la llanura pampeana. Rev. Mus. La. Plata, 25, 135-342. rovereto, G. 1914a. Studi di geomorfologia Argentina, 4. La Pampa. Boll. Soc. geol. Ital. 33, 75-128. — 19146. Los estratos araucanos y sus fosiles. An. Mus. nac. B. Aires , 25, 1-224. rusconi, c. 1937. Contribution al conocimiento de la geologia de la ciudad de Buenos Aires y sus alrede- dores y referencia de su fauna. Act. Acad. Cienc. Cordoba, 10 (3). sars, G. o. 1901. Contribution to the knowledge of the freshwater Entomostraca of South America, as shown from artificial hatching from dried material. Arch. Math. Naturv. 24, 1-52. — 1924. The Fresh-water Entomostraca of the Cape Province. Ann. S. Afr. Mus. 20, 105-193. simpson, G. G. 1940. Review of the Mammal-Bearing Tertiary of South America. Proc. Amer. phil. Soc. 83, 649-709. staplin, F. l. 1963. Pleistocene Ostracoda from Illinois. Part 2. Subfamilies Cyclocyprinae, Cypridopsinae, Ilyocyprinae; families Darwinulidae and Cytheridae. Stratigraphic ranges and assemblage patterns. J. Paleont. 37, 1164-1203. tressler, w. 1949. Freshwater ostracods from Brasil. Proc. U.S. nat. Mus. 100, 61-83. van den bold, w. A. 1971. Ostracode associations, salinity and depth of deposition in the Neogene of the Caribbean Region. Bull. Centre Rech. Pau-SNPA. 5 suppl. 449-460. vavra, v. 1898. Siisswasser Ostracoden. Hamburger Magalhaenische Sammelreise. 2, 1-26. whatley, R. c. and kaye, p. 1971. The Palaeoecology of Eemian (Last Interglacial) Ostracoda from Selsey, Sussex. Bull. Centre Rech. Pau-SNPA. 5 suppl. 311-330. — and moguilevsku, a. In press. The Family Leptocytheridae in Argentine Waters. Symposium on the Biology and Palaeobiology of Ostracoda, Newark, Delaware 1972. wicher, c. a. 1957. Die Gattung Gomphocythere in Nordwestdeutschland und das problem der brackischen Ostracoden. Micropal. 3, 269-275. wierzejski, a. 1892. Skorupiaki i wrotki slodkowodne zebrane w Argentinie. 3. Ostracoda. Rezpr. Wydz. mat.-przyr. Akad. Um. 24, 239-242. ROBIN CHARLES WHATLEY Department of Geology University College of Wales Aberystwyth TERESA DEL CARMEN CHOLICH Facultad de Ciencies Naturales y Museo Universidad Nacional de La Plata La Plata, Argentina Manuscript received 4 September 1973 THE AFFINITIES OF THE TRIFOBITE GENUS SC H ARY I A, WITH A DESCRIPTION OF TWO NEW SPECIES by R. M. OWENS Abstract. New discoveries of the trilobite Scharyia have led to a critical reappraisal of the genus, in particular the nature of the cedariiform facial suture. Scharyia has been considered by a number of authors to belong to the Proeti- dae, but it is argued here that it has closer affinities with the Otarionidae. Two new species are described, S. siceripotrix from the Silurian (Ludlow) of the British Isles and S. heothina from the Ordovician (Ashgill) of Sweden, the latter being the first record of the genus in the Ordovician. During the course of work on proetid trilobites from Britain and Scandinavia specimens were examined of the genus Scharyia , which has commonly been classified with proetids. In 1967 Dr. J. S. W. Penn and Miss J. Vinnicombe kindly put at the author’s disposal proetid and otarionid trilobites which they had collected from the Silurian of the Malvern Hills, England. Among these, from one locality in the Lower Elton Beds (Ludlow Series), were numerous well-preserved specimens of an un- described species of Scharyia , which included small growth stages. In 1969, on a field excursion to the Lake Siljan district, Sweden, led by Dr. V. Jaanusson, three speci- mens of Scharyia were found by the author in the Boda Limestone (Ashgill). Examina- tion and preparation of this new British and Swedish material necessitated comparison with described Scharyia species. This led to a critical reappraisal of the detailed morphology of the genus, in particular the cedariiform facial suture, of which new interpretations are presented below. In attempting to work out possible phyletic lineages in the earlier proetids and otarionids, it has not been possible to find any close links between Scharyia and proetids, but there does appear to be a relationship with the otarionid Panarchaeogortus, reasons for which are discussed below. Terminology follows that of Harrington et al. in Moore (ed.) (1959, pp. 0117- 0126) and Owens (1973, p. 4, text-fig. 1a), except for the following modification and additions: the term cedariiform suture (see Harrington et al. op. cit., p. 0119) is slightly modified thus: ‘opisthoparian suture in which the posterior section runs outwards (abaxially) and on to, or across the lateral border before curving backwards and then turning adaxially to the posterior margin’ ; ‘0’ refers to the angle between the section /3-S of the anterior branch of the facial suture and an exsagittal line drawn through 8; ‘/3-yS’ is the transverse distance between these points on the anterior branches of the facial sutures. REMARKS on morphology Hawle and Corda (1847, p. 78) first described Proetus micropygus (the type species of Scharyia) from a single pygidium. Barrande (1852, pi. 15, figs. 37-38; 1872, pi. 14, figs. 20-21) later figured several cephala as P. micropygus , and complete specimens [Palaeontology, Vol. 17, Part 3, 1974, pp. 685-697, pis. 98-99.] 686 PALAEONTOLOGY, VOLUME 17 (1852, pi. 17, figs. 16-17) belonging to this species as growth stages of Proetus [= Decoroproetus ] decorus Barrande, but he was evidently unaware of the nature of the posterior branch of the facial suture, as it is not included in any of his figures, nor was it mentioned in the text. Pribyl (1946a, p. 5, fig. 19) was the first to illustrate the cedariiform suture of Scharyia, and his reconstructions (see also Pribyl 19466, pi. 2, fig. 9; 1967, p. 289, text-fig. 1, fig. 1) show the abaxial part of the posterior branch running along or just inside the posterior end of the lateral border furrow. Examination of material of S. micropyga from the type and other localities in the Prague district, Czechoslovakia [in the collections of the National Museum, Prague and the British Museum (Natural History)], and of other species of Scharyia , has shown that two types of course of the posterior branch of the facial suture are developed, which are here referred to as ‘type A’ and ‘type B’. Type A. The suture crosses the lateral border furrow and runs backwards, parallel to the lateral margin of the cephalon on to the anterior part of the genal spine, and then turns abruptly adaxially across the base of the genal spine (text-fig. 1a-c). Type B. The suture does not cross the lateral border furrow, but runs backwards along or inside it, and turns adaxially across the base of the genal spine (text-fig. Id). Apart from the details of the posterior branch of the facial suture the cephalic morphology of Scharyia remains remarkably stable over a long period of time (late Ordovician to Middle Devonian). Small differences are apparent in the degree of divergence of the anterior branches of the facial sutures, in the glabellar outline and proportions, and in the presence or absence of lateral glabellar furrows. text-fig. 1. Comparison of facial sutures in three Scharyia species, a-c, type ‘A’, running outside lateral border furrow. A, c, S. siceripotrix, B, S. micropyga. Note difference in angles at ‘X’. c shows area of doublure of free cheek over- lain by fixed cheek (cf. PI. 98, fig. 9 a-b). d, type ‘B\ running inside or along lateral border furrow (the position of the lateral border furrow is assumed, as the free cheek is un- known), as in S. heothina. OWENS: SCHARYIA 687 The most variable feature of the pygidium is the border, which, broadly speaking, is absent or only poorly developed in earlier species and well developed in later ones. A feature found in all known Scharyia species (which is not known in any other genus) is a small granule on the adaxial end of each posterior pleural band (e.g. PI. 98, fig. 5), both on the pygidium and on the thorax, which enables dissociated parts to be readily identified. ORIGIN AND AFFINITIES OF SCHARYIA Pribyl (1946a) erected the genus Scharyia , which has subsequently been incorporated in various proetid subfamilies— e.g. the Tropidocoryphinae (Pribyl 19466, p. 23) and the Eodrevermanniinae (Hupe 1953, p. 217), but Richter, Richter and Struve in Moore (ed.) (1959, p. 0414) placed it in ‘subfamily uncertain’ within the Proetacea. Osmolska (1957, p. 61), however, proposed a new subfamily, the Scharyiinae to accommodate it. Pribyl (1967, p. 289) later raised the Scharyiinae to family status, and discussed the relationships of Scharyia. He suggested that it might be a distant descendant of the Upper Cambrian cedariids, on account of the similar facial suture and the small number of thoracic segments. This relationship is considered to be unlikely, as there is an enormous time span, embracing most of the Ordovician, between the last known cedariids and the earliest Scharyia , and it is believed that the similar morphology of Scharyia and the cedariids is due more probably to homoeo- morphy than to phyletic relationship. Erben (1961, p. 88) suggested that the cedariiform suture of Scharyia was simply an aberrant development of the normal type of opisthoparian suture found in proetids. If one follows Erben’s hypothesis, the ancestors of Scharyia presumably had a similar morphology to it, but with a normal opisthoparian suture. Species of Panarchaeogonus Opik, 1937 (e.g. P. whittardi (Begg), from the Ashgill of the Gir- van district (text-fig. 2b)) have such a morphology. Features shared by P. whittardi and Scharyia species include the triangulate glabella, large palpebral lobe, position of the eye, and shape of occipital ring (cf. text-fig. 2a and b). The principal differences, apart from the suture, are the prominent basal glabellar lobes and the larger number (nine) of thoracic segments of P. whittardi. Because of the striking similarities between the two genera, however, it is considered that Scharyia might be derived from Panarchaeogonus. The systematic position of the latter has been somewhat confused, but on investigation of species of this genus for an intended revision, it is considered that it is an otarionid. There seems little evidence for considering Scharyia to be a proetid. The combination of characters seen in Scharyia— the triangulate glabella, eye well out from the axial furrow, the convex preglabellar field, the shape of the occipital ring, the small number of thoracic segments, and the pygidial morphology, as well as the cedariiform suture— are not known in any proetid. Therefore, the sub- family Scharyiinae is tentatively included in the Otarionidae. Scharyia is the only known genus in the Scharyiinae. Its diagnosis has been emended because of the new morphological interpretations and familial affinities proposed here. The identification of the genus in the Boda Limestone places its origins at least as far back as the late Ordovician. The Boda Limestone is in the ‘Remopleuridid Province’ of Whittington and Hughes (1972) and thus it is agreed PALAEONTOLOGY, VOLUME 17 text-fig. 2. a, Scharyia micropyga (Hawle and Corda, 1847) wenlockiana Pribyl, 1967 (BM 42384), dorsal view, xll. Silurian, Wenlock Series, Liten Formation, Lodenice, Prague district, Czechoslovakia, b, Panarchaeogonus whittardi (Begg, 1939) (HM A1083), holotype internal mould, dorsal view, x 12. Ordovician, Ashgill Series, Rawtheyan Stage, Upper Drummuck Group, Starfish Bed no. 2, Ladyburn, Girvan, Ayrshire. Original of Begg 1939, pi. 6, fig. 3. with Schrank (1972, p. 32) that Scharyia did not originate in the ‘Mediterranean Zooprovince’ (‘Selenopeltis Province’ of Whittington and Hughes) as Pribyl (1967, p. 286) has claimed. SYSTEMATIC PALAEONTOLOGY ?Family otarionidae Richter and Richter, 1926 Subfamily scharyiinae Osmolska, 1957 Genus scharyia Pribyl, 1946u Type species. (Original designation) Proetus micropygus Hawle and Corda, 1847. Diagnosis. Glabella triangulate, with or without shallow lateral furrows; occipital ring of almost constant width (sag. and exsag.), without lateral lobes; preglabellar field weakly convex in longitudinal section; facial suture cedariiform; thorax of 6 segments; pygidium with or without border, margin entire or crenulate; axis conical with 5-9 rings, pleural areas with 4-6 pairs of ribs; small granule on adaxial end of each thoracic and pygidial posterior pleural band; sculpture granular or surface smooth. Known Scharyia species and their distribution , arranged stratigraphically . Scharyia heothina sp. nov., S. sp. 1 : Ordovician, Ashgill Series, Boda Limestone, Lake Siljan district, Sweden (described herein). Scharyia sp. and S.l sp. : Silurian, Wenlock Series (or slightly older— see Norford 1973, p. 20), NE. Green- land (Lane 1972, p. 352, pi. 61, figs. 1 1-12). Scharyia micropyga (Hawle and Corda, 1847) wenlockiana , Pribyl 1967: Wenlock Series, M. flexilis-M. OWENS: SCHARYIA 689 testis Zones, Prague district, Czechoslovakia (Pribyl 1967, p. 295, pi. 1, figs. 1-3 and text -fig. 1, fig. 1 and herein PI. 98, fig. 12 and text-fig. 2a); Upper Wenlock erratics, Hiddensee Island, German Demo- cratic Republic (Schrank 1972, p. 31, pi. 10, fig. 4); S. micropyga possibly referable to this subspecies: Wenlock Series, G. nassa Zone, Holy Cross Mountains, Poland (Tomczykowa 1957, pp. 104-105, text- fig. 10a-c); Wenlock, Roquemailliere, Montagne Noire, southern France (Chaubet 1937, p. 202, pi. 7, fig. \la-b). Scharyia micropyga micropyga (Hawle and Corda, 1 847) : Ludlow— Pridoli, Prague district, Czechoslovakia (Pribyl 1967, p. 293, pi. 1, figs. 4-6; pi. 2, fig. 6 and text-fig. 1, fig. 2 a-b). Scharyia micropyga (Hawle and Corda, 1847) meridiana Alberti, 1970: Ludlow Series, Ostracodenkalk, NW. Morocco (Alberti 1970, p. 71, pi. 9, figs. 5, 23). Scharyia siceripotrix sp. nov. : Ludlow Series, Lower Elton Beds, Malvern Hills and Wenlock Edge (described herein). Scharyia yolkini Pribyl, 1970: Ludlow Series, Kamyschenka River, north Altai region, south central U.S.S.R. (Yolkin 1965, p. 153, text-fig. 1 b and Pribyl 1970, p. 109, text-fig. 2c). Scharyia sp. (undescribed): Ludlow Series, Carnic Alps, Austria (Pribyl 1967, p. 285). Scharyia sp.: Ludlow Series, Bowning, New South Wales, Australia (Etheridge and Mitchell 1892, p. 317, pi. 25, fig. 2d [figured as ‘larval form’ of ‘ Proetus ’ rattei Etheridge and Mitchell 1892] and refigured herein). Scharyia nympha Chlupac, 1971: Pridoli Formation, Prague district, Czechoslovakia (Chlupac 1971, p. 172, pi. 23, figs. 1-6 and text-fig. 5). Scharyia angusta Pribyl, 1966: Devonian, Gedinnian, Lochkov Limestone, Prague district, Czechoslovakia (Pribyl 1966, p. 52, pi. 1, fig. 8; 1967, p. 296, pi. 2, figs. 1-2, text-fig. 1, fig. 4 a-b). Scharyia vesca Pribyl, 1966: Devonian, Pragian, Slivenec Limestone, Prague district, Czechoslovakia (Pribyl 1966, p. 53, figs. 6-7; 1967, p. 297, pi. 2, figs. 3-4 and text-fig. 1, fig- 3 a-b). Scharyia brevispinosa Pribyl, 1967: Devonian, Zlichovian, Chynice Limestone, Prague district, Czecho- slovakia (Pribyl 1967, p. 299, pi. 2, fig. 5, text-fig. 1, fig. 5). Scharyia sp. : Devonian, probably late Emsian, New South Wales, Australia (Chatterton 1971, pi. 19, figs. 25-28 [figured as ‘otarionid? sp.'] and Schrank 1972, p. 32). Scharyia tafilaltensis Alberti, 1970: Devonian, high upper Emsian, southern Morocco (Alberti 1970, p. 73, pi. 9, fig. 6). Scharyia maura Alberti, 1970: Devonian, basal Eifelian, NW. Morocco (Alberti 1970, p. 72, pi. 9, figs. 1-4). Scharyia couviniana Osmolska, 1957: Devonian, Lower Eifelian, Wydryszow, Holy Cross Mountains, Poland (Osmolska 1957, p. 62, pi. 2, figs. 1, 2 and text-fig. 2, p. 63). Scharyia siceripotrix sp. nov. Plate 98, figs. 1-9; text-figs. 1a, c, 3 Derivation of name. From Latin sicera, cider and potrix, drinker; the type locality is in a well-known cider- producing region. Holotype. NMW 71.6G.488, external mould of a cranidium (PI. 98, fig. 1) from Silurian, Ludlow Series, Lower Elton Beds, exposure in lane near Oldcastle Farm, 800 m SE. of Colwall Green, Herefordshire (SO 756405). Material. Numerous detached exoskeletal parts (NMW 71 .6G.268-278, 486-487, 489-493, 529), including meraspides and late protaspides from the type locality ; a pygidium (NMW 72. 1 8G. 153) from same horizon, road cutting 680 m at 95° from Ledbury station, Herefordshire (SO 71553857) and a few pygidia (BM It8860) from same horizon, exposure by ford opposite farm, Middlehope, Wenlock Edge, Shropshire (SO 49748828). Diagnosis. Glabellar furrows lacking in larger specimens, present in smaller ones; angle at ‘X’ (see text-fig. 1) about 70°; 6 35°-55°; pygidial border poorly developed in adult, but distinct in transitory pygidia; dorsal surface of exoskeleton smooth. Description. Cephalon with narrow border, defined by shallow, distinct lateral and anterior border furrows. Palpebral width ranges from being equal to sagittal length 690 PALAEONTOLOGY, VOLUME 17 in small specimens to being 72% of it in largest ones. Glabella of triangulate outline, of approximately equal length (sag.) and maximum width (trans.). No lateral furrows seen on large specimens, but two weak pairs seen on some smaller ones (See PI. 98, fig. 3). Occipital furrow nearly as wide as and somewhat shallower than axial furrows. Occipital ring wider (sag.) than anterior border, and a little wider (trans.) than glabella. Preglabellar field 50-75% length (sag.) of glabella in larger and smaller specimens respectively, and is weakly convex in longitudinal section. text-fig. 3. Reconstruction of cephalon of Scharyia siceripotrix sp. nov. (based on PI. 98, figs. 1 and 9 a-b), x 25 approx. EXPLANATION OF PLATE 98 Figs. 1-9. Scharyia siceripotrix sp. nov. Ludlow Series, Lower Elton Beds, exposure in lane near Oldcastle Farm, 800 m SE. of Colwall Green, Herefordshire (SO 756405). 1, NMW 71.6G.488, hOlotype cranidium, silicone rubber cast of external mould, dorsal view, x 20. 2, NMW 71.6G.490b (right) and 491b (left), pygidia, silicone rubber cast of external moulds, dorsal view. Compare development of border on two specimens, x 15. 3, NMW 71.6G.493a, small cranidium, internal mould, dorsal view, x20. 4, NMW 71.6G.492a, transitory pygidium, internal mould, dorsal view. Note well-developed border, x20. 5, NMW 71. 6G. 278a, pygidium, internal mould, dorsal view, x 16. 6, NMW 71. 6G.489, thoracic segment, internal mould, dorsal view, x 20. 7, NMW 71.6G.486, late protaspis, internal mould, dorsal view, x20. 8, NMW 71.6G.529a, late protaspis, internal mould, dorsal view, x22. Note external mould of eye showing facets to right of protaspis. 9 a-b, NMW 71.6G.487a-b, internal and counterpart external mould of free cheek. Note course of facial suture (cf. text-fig. lc), x 20. Figs. 10, 11. Scharyia sp. 10, cranidium, dorsal view, x 15. 11, complete specimen, dorsal view, original of Etheridge and Mitchell 1892, pi. 25, fig. 2d, x 14. Silicone rubber casts of specimens (both on same piece of rock) in Australian Museum, Sydney from Ludlow Series, Lower Trilobite Bed, Bowning Creek, Bowning, New South Wales, Australia. Fig. 12. Scharyia micropvga (Hawle and Corda, 1847) wenlockiana Pribyl 1967. NMP-Akz. Kat. 2929/1 893. Cranidium, dorsal view, silicone rubber cast of original of Pribyl 1967, pi. 1, fig. 2. Wenlock Series, Liten Formation, M. flexilis Zone, Lodenice, Prague district, Czechoslovakia, x 14. Figs. 13-14. Scharyia angusta Pribyl, 1966. 13, NMP IT 185, cranidium, dorsal view, silicone rubber cast of original of Barrande 1852, pi. 15, figs. 37-38, and Pribyl 1967, pi. 2, fig. I, x 15. 14, NMP IT 187, holotype pygidium, dorsal view, silicone rubber cast of original of Barrande 1852, pi. 15, figs. 39-40, and Pribyl 1967, pi. 2, fig. 2, x 13. Both specimens from Lower Devonian, Lochkov Limestone, Lochkov, Prague district, Czechoslovakia. PLATE 98 ' '-'V'S OWENS, Scharyia 692 PALAEONTOLOGY, VOLUME 17 Anterior branches of facial sutures strongly divergent, with j8 an acute angle and 6 35°-55°. Posterior branches cedariiform, following course ‘type A’ (see text-fig. 1a), angle at ‘X’ about 70°. Palpebral lobe crescentic, about 75% sagittal length of glabella, and placed at about its own sagittal length from it. Shallow palpebral furrow runs more or less exsagittally from y to e. Abaxial part of fixed cheek elevated almost to height of glabella. Eye large, crescentic, its individual facets seen on some specimens (e.g. PI. 98, fig. 8). Field of free cheek weakly convex. Genal spine blade-like, without median groove. Posterior border furrow broader and wider than anterior and lateral, truncating the latter at base of genal spine. No complete examples of this species are known, but number of thoracic segments assumed to be six, as in other species. Axis divided into annulus and articulating half ring of approximately equal width (sag.), articulating furrow deep. Pleura with broad, shallow pleural furrow, dividing it into slightly wider anterior band and narrower posterior band. Small granule on adaxial end of posterior band. Pygidium parabolic, margin entire. Border distinct in transitory pygidia (e.g. PI. 98, fig. 4) but indistinct in larger specimens (e.g. PI. 98, fig. 2 [left-hand specimen]). Axis conical, anteriorly 30-35% pygidial width (trans.), not reaching posterior margin, and consisting of seven rings and a terminal piece. Ring furrows shallow. Pleural areas with six pairs of ribs, curving gently backwards and widening slightly abaxially. Pleural and interpleural furrows of more or less equal depth, deepening abaxially. On adaxial end of each posterior pleural band is a small granule. Pygidial doublure weakly ventrally convex with fine, subparallel terrace lines. Exoskeleton smooth. Measurements of figured specimens (in mm) Cranidia A A, A2 a3 a4 K 8-S JM 0 NMW 71.6G.488 (E) 1 65 0-70 0-45 0-20 0-30 0-70 (L50) 1-80 35° NMW 71.6G.493b (I) 0-80 0-40 0-25 0-05 110 0-40 (0-80) 100 50° Pygidia Z Y W X NMW 71. 6G. 278a (I) 2-50 1-85 3-25 115 NMW 71 ,6G.491a (I) 1 95 1 50 2-75 0-95 NMW 71. 6G. 490a (I) 1-50 110 2-60 0-75 NMW 71.6G.492a (I) 0-90 0-70 1-45 0-45 Sagittal lengths of late protaspides NMW 71.6G.486 (I) 0-85 NMW 71. 6G. 529a (I) 0-75 Discussion. The species most similar to S. siceripotrix is S. angusta Pribyl, 1966 (see PI. 98, figs. 13-14) from the Lower Devonian Lochkov Limestone of the Prague district, Czechoslovakia. The former differs from the latter in the proportionately narrower glabella lacking furrows in the adult stage, the wider (trans.) pygidial axis, and the poorly developed pygidial border in adult specimens (cf. PI. 98, figs. 1, 2, 5, and figs. 13-14). Both species have an angle of about 70° at ‘X’ (see text-fig. 1), con- trasting with that of about 55° in S. micropyga. The increase of this angle, the increase in divergence of the anterior branches of the facial sutures, and the development of a pygidial border appear to be evolutionary trends in Scharyia. A species very similar to S. siceripotrix occurs in the Ludlow of New South Wales, Australia, but a detailed OWENS: SCHARYIA 693 comparison is not possible here, as the author has seen only silicone rubber casts of rather poorly preserved material (see PI. 98, figs. 10-11). The material of S. siceripotrix is of particular interest as it includes a number of small growth stages. The smallest specimens are late protaspides (PI. 98, figs. 7-8). As these are preserved in rather coarse sediment, it is not possible to see fine details; however, a pygidial border is present (seen on PI. 98, fig. 8), although detail of the facial suture is unclear. There also seem to be traces of weak eye ridges. Transitory pygidia (PI. 98, fig. 4) have a distinct border, and their proportions are not unlike those of the adult S. angusta (PI. 98, fig. 14). Small cranidia (PI. 98, fig. 3) are also similar to those of adult S. angusta (PI. 98, fig. 13), especially in glabellar proportions and in the presence of lp glabellar furrows. Scharyia heothina sp. nov. Plate 99, figs. 1-6, 8 1925 ‘ Cyphaspis ' sp. ind. a Warburg, p. 205, pi. 5, figs. 59-60. 1925 ‘ Cyphaspis ' sp. ind. b Warburg, p. 208, pi. 5, fig. 61. Derivation of name. From Greek heothinos, meaning early. Holotype. A cranidium (RM Ar47554) (PI. 99, fig. 1 a-d) from Ordovician, Ashgill Series, Boda Limestone, Kallholn, Lake Siljan district, Dalarne, Sweden. Material. Cranidia UM D75-76. RM Ar47490 and pygidia, UM D77, RM Arl0805, Ar 1 08 1 0- 1 08 1 1 , Ar47496, Ar47499, Ar47555 from the Boda Limestone of Kallholn, Boda and Gryssen, Lake Siljan district, Dalarne, Sweden. Diagnosis. Glabella with two very weak pairs of lateral furrows; broad anterior border as wide (sag.) as preglabellar field; prominent occipital tubercle; facial suture follows ‘type B’ course; pygidium without border, narrow interpleural furrows reaching margin; broader, shallow pleural furrows falling short of it; sculpture of fine granules. Description. Cranidium with sagittal length slightly greater than palpebral width. Glabella about as wide (trans.) as long, with two pairs of very weak backwardly directed furrows. Occipital furrow narrower, and medially shallower than axial furrows. Occipital ring as wide (sag.) as anterior border, a little wider (trans.) than glabella and with a prominent median tubercle. Preglabellar field as wide (sag.) as anterior border, weakly convex in longitudinal section. Anterior border furrow narrow, distinct. Anterior border gently convex, slightly more convex than preglabellar field. Anterior branches of facial suture divergent, d 22°. Palpebral lobe about 75% sagittal length of glabella, crescentic and with a faint smooth band running parallel with its margin. Weak eye ridge seen on some speci- mens (e.g. PI. 99, fig. 8). Posterior branch of facial suture cedariiform, following course ‘type B’. Posterior border furrow broad and deep, shallowing at extreme abaxial end. Free cheek and thorax unknown. Pygidium subparabolic in outline. Axis conical with six or seven shallow ring furrows arched forwards sagittally, first widened sagittally. Pleural areas with four or five pairs of ribs which widen abaxially. Pleural and interpleural furrows more or less parallel, pleural wider and dying out before reaching margin, interpleural 694 PALAEONTOLOGY, VOLUME 17 narrower and reaching margin. Both sets of furrows very weak on some specimens (e.g. PL 99, fig. 4). No pygidial border. Sculpture of fine granules. Measurements of figured specimens (in mm) Cranidia A Ax a2 a3 a4 K 8-8 |8-/3 e UM D75 (E) 2-15 L20 0-35 0-30 0-30 100 (1-95) (2-00) 29° UM D76 (E) 1 10 (0-25) 105 ( 1 -80) (1'60) RM Ar47554 (E) 1 -80 105 0-25 0-25 0-25 LOO (1-70) (1-60) 22° Pygidia Z Y W X UM D77 (E) 2-50 200 3-30 1-15 RM Ar47555 (E) 2-05 1-65 2-85 0-95 RM Ar47496 (E/I) 1-55 3-05 110 RM Arl0805 (E) 1 -75 1-35 2-65 0-80 Discussion. Warburg (1925) figured and described cranidia as ‘ Cyphaspis' sp. ind. a (p. 205, pi. 5, figs. 59-60), and pygidia as ‘C.’ sp. ind. b (p. 208, pi. 5, fig. 61), and compared them with the type species of Scharyia. Both of these cranidia (see PI. 99, figs. 6, 8) are badly preserved, but one (PI. 99, fig. 8) shows the characteristic cedarii- form suture on the left-hand side. Additional material (PI. 99, fig. 1 a-d) from the same horizon and locality is better preserved and confirms that the cranidia can be confidently assigned to Scharyia. The pygidia are, as Warburg (1925, p. 209) has pointed out, quite similar to that of 5. micropyga, although they lack the border, but they do have the characteristic granule on the adaxial end of each pleural band (e.g. PI. 99, fig. 3). Scharyia sp. 1 Plate 99, fig. 7 1925 ‘Cyphaspis'l sp. ind. d Warburg, p. 210, pi. 5, fig. 62. Material. A pygidium, UM D78 from the Boda Limestone, Boda, Lake Siljan district, Sweden. Original of Warburg 1925, pi. 5, fig. 62. Description. Pygidium subtriangular in outline, about 70% as long (sag.) as wide (trans.) with narrow border which widens slightly towards posterior. Anteriorly EXPLANATION OF PLATE 99 Figs. 1-6, 8. Scharyia heothina sp. nov. 1 a-d, RM Ar47554, holotype cranidium, with external surface preserved; la, dorsal view; lb, anterior oblique view; lc, anterior view; Id, lateral view. Ordovician, Ashgill Series, Boda Limestone, Kallholn, Lake Siljan district, Sweden. All x 25. 2, UM D77, pygidium with external surface preserved, dorsal view. Original of Warburg 1925, pi. 5, fig. 61. Boda Limestone, Boda, Lake Siljan district, Sweden, x 15. 3, RM Ar47496, pygidium, partially exfoliated, dorsal view. Horizon and locality as fig. 1, x 16. 4 a-c, RM Ar47555, pygidium, with exoskeleton preserved; 4 a, dorsal view; 4 b, posterior view; 4c, lateral view. Horizon and locality as fig. 1, x20. 5, RM Arl0805, pygidium with exoskeleton preserved, dorsal view. Boda Limestone, Gryssen, Ostbjorka, Lake Siljan district, Sweden, x 20. 6, UM D75, incomplete cranidium, dorsal view. Original of Warburg 1925, pi. 5, fig. 59. Horizon and locality as fig. 1, x 18. 8, UM D76, incomplete cranidium, dorsal view. Original of Warburg 1925, pi. 5, fig. 60. Horizon and locality as fig. 1, x 20. Fig. 7. Scharyia sp. 1. UM D78. Pygidium with exoskeleton preserved, dorsal view. Original of Warburg 1925, pi. 5, fig. 62. Horizon and locality as fig. 2, x 14. PLATE 99 OWENS, Scharyia mm. 696 PALAEONTOLOGY, VOLUME 17 narrow axis about 30% of pygidial width (trans.), and tapers gently backwards, not reaching border. It is composed of nine rings, defined by shallow ring furrows which become progressively shallower and narrower (sag.) towards posterior. Articulating half ring about 70% width (sag.) of first axial ring, and separated from it by deep, narrow articulating furrow. Pleural areas with six pairs of ribs, pleural and inter- pleural furrows nearly parallel, latter a little narrower and shallower than former. Both curve gently backwards abaxially and are truncated at inner edge of border. Small granule on adaxial end of each posterior pleural band. Sculpture granulose. Measurements (in mm) z y w x UM D78 (E) 2-30 2 00 3-30 100 Discussion. As with the pygidia described above as S. heothina , Warburg (1925, p. 211) compared this specimen with S. micropyga. Scharyia sp. 1 differs from S'. heothina in outline, in having a narrower axis with more rings (nine compared with six or seven), a narrow border, much deeper and more distinct interpleural furrows, and in having the pleural and interpleural furrows curved backwards abaxially. Repositories. The following abbreviations are used herein: NMW— National Museum of Wales, Cardiff; BM— British Museum (Natural History); UM— Museum of Palaeontological Institute, Uppsala; RM— Naturhistoriska Riksmuseet, Stockholm; NMP — National Museum, Prague; HM — Hunterian Museum, Glasgow. Acknowledgements. I thank Dr. M. G. Bassett, Dr. D. L. Bruton, and Dr. J. H. McD. Whitaker for critically reading through the manuscript at various stages of its preparation and for suggesting useful amendments. I am also grateful to Professor Dr. H. K. Erben for beneficial discussion at early stages of the work, and for showing me casts of Australian material. Dr. V. Jaanusson (RM), Mr. S. F. Morris (BM), Dr. S. Steunes (UM), Dr. J. K. Ingham (HM), and Dr. V. Zazvorka and Mr. F. Bastl (NMP) kindly allowed me to examine and borrow specimens in their care. REFERENCES alberti, G. k. b. 1970. Trilobiten des jungeren Siluriums sowie des Unter-und Mitteldevons. II. Abh. senckenb. naturforsch. Ges. 525, 1-233, pis. 1-20. barrande, J. 1852. Systeme Silurien du centre de la Boheme. premiere partie. Recherches paleontologiques , Vol. 1. Crustaces, Trilobites. xxx + 935 pp., 51 pis. Prague and Paris. — 1872. Systeme Silurien du centre de la Boheme. premiere partie. Recherches paleontologiques. Supple- ment au vol. 1. xxx+647 pp., 35 pis. Prague and Paris. begg, J. l. 1939. Some new species of Proetidae and Otarionidae from the Ashgillian of Girvan. Geol. Mag. 76, 372-382, pi. 6. chatterton, b. d. E. 1971. Taxonomy and ontogeny of Siluro-Devonian trilobites from near Yass, New South Wales. Palaeontographica (A) 137, 1-108, pis. 1 24. chaubet, m. c. 1937. Contribution a l’etude geologique du Gothlandien du versant meridional de la Montagne Noire. Lab. Geol. Univ. Montpelier , Fac. Sci. Mem. 1, Montpelier. chlupac, i. 1971. Some trilobites from the Silurian/Devonian boundary beds of Czechoslovakia. Palaeon- tology, 14, 159-177, pis. 19-24. erben, h. k. 1961. Blinding and extinction of certain Proetidae (Tril.). Palaeont. Soc. India, 3 (Birbal Sahni Memorial Number, 1958), 82-104. etheridge, r. and Mitchell, j. 1892. The Silurian trilobites of New South Wales, with reference to those of other parts of Australia. Part 1. Proc. Linn. Soc. New South Wales, 6, 311-320, pi. 25. hawle, i. and corda, a. j. c. 1847. Prodrom einer Monographic der bohmischen Trilobiten. 176 pp., 7 pis. Prague. hupe, p. 1953. Classe de trilobites. In piveteau, j. (ed.). Traite de paleontologie, 3, 44-246, 140 text-figs. Paris. OWENS: SCHARYIA 697 lane, p. d. 1972. New trilobites from the Silurian of north-east Greenland, with a note on trilobite faunas in pure limestones. Palaeontology , 15, 336-364, pis. 59-64. MOORE, R. c. (ed.). 1959. Treatise on Invertebrate Paleontology , Part O , Arthropoda 1. xix+ 560 pp., 415 figs. Geol. Soc. Amer. and Univ. Kansas Press (Lawrence). norford, b. s. 1973. Lower Silurian species of the trilobite Scotoharpes from Canada and northwestern Greenland. Bull. Geol. Surv. Canada , 222, 9-25, pis. 1-4. osmolska, h. 1957. Trilobites from the Couvinian of Wydryszow (Holy Cross Mountains, Poland). Acta Palaeont. Polonica , 2, 53-77, 3 pis. owens, R. m. 1973. British Ordovician and Silurian Proetidae (Trilobita). Palaeontogr. Soc. [Monogr.] 1-98, 15 pis. pribyl, a. 1946a. O nekolika novych trilobitovych rodech z ceskeho siluru a devonu. Pfiroda , Brno. 38 (5-6), 7 pp., 1 1 text-figs. 19466. Prispevek k poznani ceskych Proetidu. Rozpr. II tr desk Akad. 55 (10), 1-37. Prague. (Notes on the recognition of the Bohemian Proetidae. Bull. int. Akad. tcheque Sci. 1945, 46, 1-41, pis. 1-4. Prague.) 1966. Proetidni trilobiti z novych sberu v ceskem siluru a devonu. (Proetiden aus neueren Auf- sammlungen im bohmischen Silur und Devon (Trilobitae) II.) Casop. narodn Muzea, Odd. prirod. 135, 49-54, pi. 4. [Czech and German texts.] 1967. Die Gattung Scharyia Pribyl, 1946 (Trilobita) und ihre Vertreter aus dem bohmischen Silur und Devon. Spis. bulg. geol. Druzh. 28, 285-301, 2 pis. — 1970. O nekolika ceskych a asijskych zastupcich proetidnich trilobitu. (Uber einige bohmische und asiatische Vertreter von Proetiden (Trilobita).) Cas Miner. Geol. 15, 101-111, 1 pi. [In German, with Czech summary.] schrank, e. 1972. Proetacea, Encrinuridae und Phacopina (Trilobita) aus silurischen Geschieben. Geologie, No. 76, 117 pp., 21 pis. tomczykowa, e. 1957. Trilobity z hipkow graptolitowych Wenloku i dolnego Ludlowu Gor Swit^tokrzy- skich (Trilobites from the Wenlock and Lower Ludlow graptolitic shales of the Swi?ty Krzyz Moun- tains). Biul. Inst. Geol. 122, 83-143, 4 pis. [In Polish, with English summary.] warburg, e. 1925. The trilobites of the Leptaena Limestone in Dalarne. Bull. geol. Instn. Univ. Uppsala , 17, vi + 446 pp., 1 1 pis. Whittington, h. b. and hughes, c. p. 1972. Ordovician geography and faunal provinces deduced from trilobite distribution. Phil. Trans. R. Soc. (B), 263, 235-278. yolkin, e. a. 1965. A new genus of trilobite (Proetidae) from the Silurian of the Altai. Paleont. Zh. 1965, l, 152-154. [In Russian.] Typescript received 8 October 1973 r. m. OWENS Department of Geology National Museum of Wales Cardiff, CF1 3NP R A NEW PROBOSCIDEAN FROM THE LATE MIOCENE OF KENYA by VINCENT J. MAGLIO Abstract. A new species of proboscidean, Gomphotherium ngorora, is described from the Ngorora Formation which has been K/Ar dated provisionally between 9 and 12 m.y. BP. This new material helps to bridge the morphological gap between the earlier Miocene gomphotheres of Africa and early Pliocene elephants. It confirms earlier views on the origin of the family Elephantidae. T he late Miocene epoch, between 14 and 6 m.y. ago, was perhaps the most important period in the development of the modern African fauna. Yet, until recently it has remained the poorest known period. Since 1965 a detailed programme of mapping in the Lake Baringo region of the Gregory Rift, central Kenya, has been undertaken by the East African Geological Research Unit based at Bedford College, Uni- versity of London. The programme has led to the discovery and further investigation of a series of fossiliferous sedimentary units, and potassium-argon ages have been established for associated lavas (Bishop et al. 1971 ; Bishop 1972). The Lake Baringo sequence consists of twelve sedimentary units ranging in age from 12 to 14 m.y. for the Alengerr beds to less than 0-23 m.y. for the Kapthurin Formation. Large assemblages of mammals occur only in four of these units and a significant gap still remains between 10 and 6 m.y. (text-fig. 1). Nevertheless, the 0-, text-fig. 1. Suggested phyletic relationships between Miocene gomphotheres of Africa and the earliest elephantids Stegotetra- belodon and Primelephas. The genus Anancus is a late specializa- tion of the Gomphotheriidae. Olduvai Omo- East Rudolf Kanapoi Lothagam Elephas Loxodonta Anancus Primelephas Stegotetrabelodon s' Ngorora Alengerr beds Fort Teman Gomphotherium sp. Maboko G. cf angustidens Rusinga Gomphotherium ngorora cf Gomphotherium [Palaeontology, Vol. 17, Part 3, 1974, pp. 699-705, pi. 100.] 700 PALAEONTOLOGY, VOLUME 17 importance of the Baringo sequence in narrowing this void, and its potential for further closing it, is now being amply demonstrated. One of the most important of the Baringo units is the Ngorora Formation which was first described by Bishop and Chapman (1970). The formation consists of five members ranging from clays and bedded tuffs at the base (A), through ferruginous channel-conglomerates and grits (B), to tuffaceous silts (C), another series of channel- conglomerates (D), and a capping unit of diatomites and grits (E). Vertebrate fossils are usually concentrated in the channel-conglomerates, but chelonian and fish remains occur abundantly in other units as well. Preliminary potassium-argon determinations for phonolite lavas occurring above and below the Ngorora Forma- tion suggest an age for the deposit of somewhat less than 12 m.y., but greater than 9 m.y. A preliminary faunal list has been recorded by Bishop et al. (1971), but much new material has been recovered recently and significant additions to the list will be published shortly. Among the new materials is a proboscidean collected in 1972 by M. Pickford that provides clear confirmation of a gomphotheriid origin for the later elephants. Mr. Pickford has kindly provided the following description of the locality: The specimen was found on the surface having been derived from a clay horizon the upper part of which graded into a calcrete. The horizon is in a sequence of rhythmic units with alternating coarse pumiceous beds and fine-grained suncracked and frequently calcified lacustrine sediments containing numerous comminuted fish remains. Stratigraphically, it lies in Member E of the Ngorora Formation. DESCRIPTION Order proboscidea Family gomphotheriidae Genus gomphotherium Gomphotherium ngorora sp. nov. Plate 100, figs. 1-3 Holotype. KNM-BN 571 and 577, right maxillary fragment with complete M2-M3, posterior part of left M3; KNM-BN 567 and 584B, proximal one-fourth of the left radius and ulna; fragmentary pieces of vertebrae, ribs, carpal bones, and maxillary tusks, all specimens from a single individual. Housed in the National Museum of Kenya. Type horizon and locality. Upper Miocene of the Ngorora Formation, site 2/49A, Baringo District, central Kenya. Diagnosis. A species of Gomphotherium bearing three cone-pairs on M2 and four on M3; pre- and post-trite cones transversely elongated and antero-posteriorly com- pressed forming very elephant-like plate structures; each cone subdivided into two or three parts by deep anterior and posterior vertical grooves, forming a transverse series of adjacently fused pillars; intravalley columns prominent, isolated apically but fused into cones basally; median clefts persistent between all cone-pairs but very EXPLANATION OF PLATE 100 Figs. 1-3. Gomphotherium ngorora sp. nov., KNM-BN 571 and 577, holotype, right M2-M3. 1, buccal view. 2, occlusal view. 3, lingual view. PLATE 100 MAGLIO, Gomphotherium 702 PALAEONTOLOGY, VOLUME 17 tightly compressed at apex; enamel thick and smooth except toward base of crown where it becomes coarsely folded ; cement present in valley troughs. Description. Gomphotherium ngorora is in many ways a typical gomphothere in which the molar teeth are subdivided into three or four cone-pairs (PI. 100). Each cone-pair is separated from others by a broad transverse valley. Also typical of the Gom- photheriidae are the isolated enamel columns standing in the transverse valleys and rising to the height of the principal cones. Similar features are seen in East African Miocene gomphotheres (below) and are characteristic of the family. More advanced elephant-like trends are also seen here which suggest that we are sampling a population that has already diverged from the main gomphotheriid stem. In the upper molars there are three cone-pairs on M2 and four pairs on M3. These are drawn out transversely with the median cleft tightly compressed between buccal and lingual components. This antero-posterior compression results in a plate-like arrangement of enamel folds not seen elsewhere in the family to this degree. The buccal cone of each pair is swollen basally and subdivided into three parts by vertical grooves on the anterior and posterior surfaces. These subdivisions form free pillars apically, but these fuse into a single transversely elongated hemi-plate as wear proceeds to about one-third of the crown height. The buccal-most division in each case is the largest. In all other known Miocene gomphotheres from Africa, including the Fort Ternan species, the buccal cones are only vaguely grooved and slightly com- pressed in the fore and aft direction, and the general structure remains that of a cone, not a flattened plate. The lingual cone of each pair is subdivided into two unequal parts on the first several pairs, but may remain as a single rounded cone on the last pair, as it is on all lingual cones of the Fort Ternan, Rusinga, and Maboko species. Toward the crown base both lingual and buccal cones arise from a common enamel fold so that in heavy wear even the median cleft would be obliterated and a single flattened enamel loop would remain on the occlusal surface. A posterior fold with several small enamel pillars represents the early development of a fifth cone-pair. These structural features are even more pronounced on the M3 in which wear has progressed further; the lingual and buccal cones have joined or nearly joined in the mid-line. By the time wear has proceeded to the lower quarter of the crown height a continuous enamel plate remains on the occlusal surface. The result of these trends is the establishment of a transverse arrangement of enamel and dentine bands with cutting surfaces deployed for a more fore and aft masticatory function. This was the principal feature of the adaptive shift that accom- panied the rise of the Elephantidae (Maglio 1972). It is seen in no other proboscidean group except the stegodonts of Asia. However, the Ngorora specimen retains pro- minent intravalley columns which remain free along their apical half, fusing into the plate-like cone-pairs below. Such structures are always seen on the anterior half of early elephant molars, but never in stegodonts. Of the post-cranial remains only the radius and ulna fragments are worthy of special mention. The ulna is relatively massive with a very deep interarticular sulcus which separates the capitular and trochlear facets nearly to the base of the short olecranon process. The latter is broad antero-posteriorly and terminates at a large MAGLIO: MIOCENE PROBOSCIDEAN 703 bulbous rugosity for insertion of the triceps muscle. An oval, flat facet for articulation with the radius is found on the medial side of the capitular facet only; the radius did not fill the interarticular sulcus as it does generally in elephants. These bones resemble more closely the comparable elements of Gomphotherium than they do any of the elephants. However, little post-cranial material of any fossil proboscideans has been properly described, and virtually none is available with positively identified dental remains from the African Miocene. The present material thus adds little to our know- ledge of the Ngorora species other than to confirm its gomphotheriid affinities. Although the Ngorora specimen shows a number of elephant-like features, it remains close to other Miocene gomphotheres in basic structure. A number of generic names have been applied to these forms, but until a revision of this group is undertaken it is best to refer the present specimen to the genus Gomphotherium. Dimensions (in mm) of holotype dentition of Gomphotherium ngorora , KNM-BN 571 and 577. M3 (577) M2 (571) Cone-pairs 4X 3X Overall length 1474 98-5 Maximum width 854 75-8 Maximum height Enamel thickness 613 40-4-9 DISCUSSION The African Miocene Proboscidea were last reviewed by Maclnnes (1942) when he referred material from Rusinga and Maboko to the European species Gomphotherium angustidens. He proposed the subspecific name G. angustidens kisumuensis for this material from East Africa. In 1945 Arambourg recognized in Maclnnes’s 1942 figured hypodigm what he considered to be a distinct morphotype in which the cone-pairs were tending toward alternation of lingual and buccal cones. This condition is seen in exaggerated form in the Pliocene and Pleistocene genus Anancus. Arambourg proposed the new name Protanancus macinnesi for this material, designating specific figures from Maclnnes’s monograph as his hypodigm. An examination of the original collection by the present author does not support Arambourg’s contention. Contrary to Maclnnes’s original analysis, there are two distinct species involved, but their distribution and diagnostic features are not as proposed by Arambourg. This is not the place to go into a review of the Miocene gomphotheres of Africa as the group has not yet been adequately studied, and most of the material is yet to be described. It is clear, however, that the Rusinga species was already specialized in a direction away from the later anancine gomphotheres and elephants, and paralleling the kinds of dental characteristics seen in the Mam- mutidae. The latter family, long believed to have been excluded from sub-Saharan Africa, is now known to have been present there in early Miocene times, as evidenced by specimens from Napak (W. W. Bishop pers. comm.) and from Moroto 1 and Lothidok (C. Madden pers. comm.). It is possible that the Rusinga material may prove to have been related to early differentiation of this family. The Maboko proboscidean assemblage, which derives from a somewhat younger 704 PALAEONTOLOGY, VOLUME 17 horizon than that from Rusinga (Maclnnes 1942; Bishop 1967), retains its distinctive gomphothere dental structures and could easily have been ancestral to the form represented in the younger Fort Ternan deposits. The Ngorora specimen, younger than the Fort Ternan species, departs from the gomphothere pattern in the development of those features mentioned above, strongly indicative of an approach to the grade seen later in the Elephantidae. The latter family includes the living Asiatic and African elephants, the Pleistocene mammoths, and their extinct relatives. Until recent years the group’s ancestry was poorly known and it was believed to have derived from Asiatic stegodonts. On the basis of rather incom- plete fossil evidence from Africa, Aguirre (1969) proposed a gomphotheriid ancestry for elephants in Africa via the stegolophodonts. This general picture was later sup- ported with better material from East Africa but it was argued that Stegolophodon could not have provided a transitional stage (Maglio 1970). The latter genus is here considered to represent an early stage of the Stegodon group. The origin of elephants, like so many other groups in Africa, has been difficult to interpret because of the 8 m.y. gap in the late Miocene African record. Prior to the gap a series of gomphotheres can be traced through fragmentary evidence from early Miocene deposits at Mfwanganu, Karunga, Maboko and others, to the middle Miocene at Fort Ternan. During this period of some 7 or 8 m.y. a considerable degree of evolution is seen in crown height and molar complexity. After a gap of about 8 m.y. the earliest elephantids of the genera Primelephas and Stegotetrabelodon are recorded at Lothagam (Maglio 1970), Sahabi (Petrocchi 1941), and at several other localities (Bishop 1972). This material is clearly indicative of a gomphotheriid ancestry (Maglio 1973), but no intermediates were known until now. The Ngorora proboscidean provides this transitional stage (text-figs. 1, 2). From this morphological grade the elephant molar could have been derived merely by an increase in number of cone-pairs, the further obliteration of the median cleft, and text-fig. 2. Diagrammatic crown views of proboscidean molar teeth showing major evo- lutionary trends in thinning of enamel, reduction of intravalley columns, and consolidation of cone-pairs into plates, a, Gomphotherium angustidens ; b, Gomphotherium ngorora sp. nov. ; c, Stegotetrabelodon orbus. MAGLIO: MIOCENE PROBOSCIDEAN 705 loss of intravalley columns so that only one remains behind each plate. All of these trends are well under way in the Ngorora specimen. Because of its geological and morphological position it seems reasonable to suggest that Gomphotherium ngorora may be close to a true intermediate stage in the evolution of the Elephantidae in Africa. The swollen cone-pair bases and abrupt apical tapering more closely resemble the plate structure in Primelephas than in Stegotetrabelodon. However, it is too early to be certain whether G. ngorora gave rise to the Elephantidae directly or via the Stegotetrabelodontinae as a transitional group, if indeed it is itself on the main-line phyletic lineage. Acknowledgements. I am indebted to Dr. W. W. Bishop for permission to examine the specimen and for comments on the manuscript. The study was conducted while the author was in Kenya under grants from the National Geographic Society and the Wenner-Gren Foundation for Anthropological Research. To these organizations I extend my gratitude. Mr. Cary Madden offered helpful discussion. REFERENCES aguirre, e. 1969. Evolutionary history of the elephant. Science , 164, 1366-1376. arambourg, c. 1945. Anancus osiris, un nouveau Mastondonte du Pliocene inferieur d’Egypte. Bull. Soc. Geol. Fr. ser. 5, 15, 479-495. bishop, w. w. 1967. The late Tertiary in East Africa— volcanics, sediments and faunal inventory. In bishop, w. w. and clark, j. d. (eds.). Background to evolution in Africa, pp. 31-56. Chicago Univ. Press. 1972. Stratigraphic succession ‘versus’ calibration in East Africa. In bishop, w. w. and miller, j. a. (eds.). Calibration of hominoid evolution, pp. 219-246. Edinburgh, Scottish-Academic Press. and chapman, G. r. 1970. Early Pliocene sediments and fossils from the northern Kenya rift valley. Nature, Lond. 226, 914-918. hill, A. and miller, J. a. 1971 . Succession of Cainozoic vertebrate assemblages from the northern Kenya rift valley. Ibid. 233, 389-394. macinnes, d. g. 1942. Miocene and post-Miocene Proboscidea from East Africa. Trans. Zool. Soc. Lond. 25, 33-106. maguo, v. j. 1970. Four new species of Elephantidae from the Plio-Pleistocene of northwestern Kenya. Breviora , no. 341, 1 -43. 1972. Evolution of mastication in the Elephantidae. Evolution, 26 (4), 638-658. 1973. Origin and evolution of the Elephantidae. Trans. Amer. Phil. Soc. n.s. 63 (3), 1-149. osborn, H. f. 1936. Proboscidea, 1, 802 pp. New York, Amer. Mus. Nat. Hist. Press. petrocchi, c. 1941. I giacimento fossilifero di Sahabi. Boll. Soc. Geol. Ital. 60 (1), 107-1 14. v. J. MAGLIO Department of Geological and Geophysical Sciences Princeton University Revised typescript received 14 January 1974 Princeton, New Jersey 08540, U.S.A. NEW MICROFOSSILS FROM THE SILURIAN (LLANDOVERY, STAGE 6) OF THE OSLO REGION, NORWAY by 0RNULF L AURITZEN Abstract. Two new microfossils of problematic affinities are described from the Llandovery of the Oslo Region, viz. Sandvikina brachiata gen. et sp. nov. and Regnellia camera gen. et sp. nov. Both occur in biomicritic limestones in the Oslo region, while the latter is also found in a biosparite from Dalarna, Sweden. During a current study of the sediments and fauna of stage 6 in the Lower Silurian of the Oslo region (Lauritzen and Worsley in prep.), two undescribed types of micro- fossils have been found in biomicritic limestones. Both types were found in samples collected in Sandvika (Oslo- Asker district, see Stunner 1953, p. 53), which is located 12 km WSW. of Oslo centre. The fossils can only be detected in thin section. Samples collected from the same beds at other localities in the central part of the Oslo region (at Spirodden 16 km SW. of Oslo centre and on the island of Malm0ya 5 km S. of Oslo centre) contained neither of these microfossils. At Sandvika both types occur together with the blue-green alga Girvanella sp., suggesting a shallow-water deposi- tional environment for these beds in the area (Lauritzen and Worsley 1974); this alga is also found at Spirodden, but not on Malm^ya. Neither of these microfossils have been previously described or recorded from the Oslo region, but Regnellia camera has been found in Dalarna, Sweden (Regnell 1947). Regnell did not suggest any possible affinities for these structures, and this question remains open. As these microfossils seem to be of stratigraphic value, a new description is given in the hope that it may stimulate information on their geographical distribution, stratigraphical range, and possible affinities. DESCRIPTIONS PROBLEMATICA Genus sandvikina gen. nov. Type species. Sandvikina brachiata sp. nov. Derivation of name. After the village of Sandvika 12 km WSW. of Oslo centre. Diagnosis. In section a 4-75 mm almost trapezoidal-shaped microfossil; hollow, with complex wall-structure. Sandvikina brachiata sp. nov. Plate 101 and Plate 102, fig. 1 ; text-figs. 1 and 2 Holotype. The specimen in thin section figured on Plate 101, PMO 93905. [Palaeontology, Vol. 17, Part 3, 1974, pp. 707-714, pis. 101-102.] 708 PALAEONTOLOGY, VOLUME 17 Type horizon and type locality. Llandovery, stage 6c in a roadsection by Ringeriksveien, Sandvika, Oslo- Asker district, Norway. Derivation of name. The species occurs with arm-like structures in some sections (PI. 101, fig. 2). Material. In addition to the holotype, four other specimens have been found. One has all the elements shown on text-fig. 1 and Plate 101, fig. 1. In two others the outgrowth and two sides of the main part are preserved (text-fig. 2 and PI. 101, fig. 2), while in the last the outgrowth only is preserved (PI. 102, fig. 1). Description. The holotype (text-figs. 1 and 2 and PL 101) has a total length of approxi- mately 4-75 mm (in section). It has an almost trapezoidal shape, and an outgrowth is seen in one corner. The terms ‘outgrowth’ and ‘main part’ are purely descriptive, and are not intended to have any genetic meaning. The outgrowth (text-fig. 2, PI. 101 and PI. 102, fig. 1) is about 1-75 mm long, and consists of an outer wall surrounding a central chamber. Septa point into the chamber from the inside of the wall. The septa are somewhat irregularly orientated, but tend text-fig. 1. An interpretation of the holotype of Sandvikina brachiata (see PI. 101, fig. 1). The outgrowth and the two sides of the main part attached to it have good contact with each other. The three other corners of the trapezoidal structure are not preserved, but the position of the remaining sides and structures in the sediment points to the configuration shown above. EXPLANATION OF PLATE 101 Figs. 1, 2. Sandvikina brachiata gen. et sp. nov., holotype. PMO 93905. 1, having the almost trapezoidal shape, with the outgrowth in the lower left-hand corner. 2, details from the outgrowth and the two sides of the main part attached to it. Both the septa of the outgrowth and the complexity of the walls can be seen. PLATE 101 LAURITZEN, Silurian microfossils 710 PALAEONTOLOGY, VOLUME 17 to point towards the apex of the outgrowth. One septum completely separates the outgrowth from the main part of the organism, and no aperture is evident. The wall is imperforate and consists mostly of one layer, but in one part a double wall appears. The main part (text-figs. 1 and 2, and PI. 101) of the organism is almost trapezoidal in shape in the section seen here, and is about 3 mm in diameter. The two sides adjacent to the outgrowth are the best preserved, their lengths being about 2-7 mm (see text-fig. 2 and PI. 101, fig. 2). text-fig. 2. The diagram shows the best-preserved parts of Sandvikina brachiata and the names used here for the different organic structures described. The wall of the main body appears to consist of at least two layers of organic material, a relatively thick (approx. 0- 1 mm) outer layer with a composite structure, and a simple membraneous inner layer; these are separated by a calcite layer varying in thickness between 0-05 and 0-25 mm. In the best-preserved portion of the section, the thicker outer layer appears to consist of two membranes with an intervening space irregularly filled with wall material. In one portion of the section the arrangement of the latter is reminiscent of septa. The calcite layer consists of crystalline calcite with crystals of varied size. EXPLANATION OF PLATE 102 Fig. 1. Sandvikina brachiata gen. et sp. nov., PMO 93903. An incomplete outgrowth without other parts of the organism preserved. Figs. 2, 3. Regnellia camera gen. et sp. nov., holotype. PMO 93904. 2, clearly divided into chambers, and thicker in the middle than at the ends. 3, detail from the holotype. The chamber walls are all double, but no outer wall can be seen. PLATE 102 0-0 0.5 mm Li i i i J LAURITZEN, Silurian microfossils 712 PALAEONTOLOGY, VOLUME 17 A grained structure is apparent (Horowitz and Potter 1971, p. 63) with irregular optic orientation under crossed nicols. All walls are imperforate and brownish in colour. Distribution. The specimens described here are found in biomicritic limestones from stage 6c at Sandvika. Samples with this species have been found with a stratigraphical separation of 50 metres. Genus regnellia gen. nov. Type species. Regnellia camera sp. nov. Derivation of name. This genus is named in honour of Professor G. Regnell, who first recorded and illus- trated, but did not name, this microfossil from Dalarna, Sweden (Regnell 1947). Diagnosis. In section a 3-25 mm long sub-cylindrical and chambered microfossil, with no recognizable wall structure. Regnellia camera sp. nov. Plate 102, figs. 2 and 3 Holotype. The specimen in thin section figured on Plate 102, figs. 2 and 3, PMO 93904. Type horizon and type locality. Llandovery, stage 6c in a roadsection by Ringeriksveien, Sandvika, Oslo- Asker district, Norway. Derivation of name. The species shows chambers in section. Material. Three thin sections from stage 6c at Sandvika, and one thin section from a crackfilling in the Boda limestone of Kallholn, Dalarna, Sweden (The University of Lund, LO 3434-3437 Slide 4772). Description. The holotype (PI. 102, fig. 2) has a total length of 3-25 mm, while other specimens in thin section are all fragments. The fossil is sub-cylindrical with a medial diameter of about 0-35 mm, tapering to 0-2 mm at both ends. It has well-defined chambers, and the chamber walls are all double (PI. 102, fig. 3). The walls are imperforate and no wall-structures are visible. There are 24-26 chamber walls per mm length in the Sandvika specimens, while there are 13 per mm in the Swedish specimens. No outer walls can be seen in the Norwegian material, but some of the Swedish specimens show a poorly defined wall-like structure (Regnell 1947, p. 4, fig. 3). Distribution. Regnellia camera is found in biomicritic limestones of stage 6c at Sand- vika, Oslo region, Norway, and in crackfilling in the Boda limestone of Kallholn, Dalarna, Sweden. DISCUSSION The affinities of these microfossils are not clear, and it is also uncertain whether they represent fragments of larger organisms or are complete as seen in the thin sections figured here. The completeness of Sandvikina brachiata is highly uncertain as there is no natural ending seen in three of the corners of its trapezoidal structure. The only complete corner is that with an outgrowth (text-fig. 1 and PI. 101, fig. 1) so that there LAURITZEN: SILURIAN MICROFOSSILS 713 could originally have been equivalent structures in the three other corners, although this is unlikely. Regnellia camera appears to be more complete, with a natural termination at both ends. The Norwegian specimens are thought to belong to the same species as those found in Sweden in spite of the different proportions of the chambers in the two localities. R. camera occurs in quite different sediment types in Norway and Sweden. The Swedish specimens are found in a biosparite (suggesting a rather turbulent depositional environment), while the Norwegian material is found in biomicrites indicative of quieter conditions. Since the bottom conditions were probably so different, a pelagic way of life may be indicated. Some specimens of S', brachiata are obviously incomplete and lack the main part, and this partial preservation might indicate some post-mortem transport, but says nothing about the mode of life of the organism. Both the Norwegian and the Swedish material of these microfossils appear together with fragments of other fossils. Regnell ( 1 947) stated that the only macrofossils present in hand specimens were brachiopods and sparse columnals of pelmatozoans, but in the thin section I have also found whole ostracods and fragments of ostracods, trilobites, and bryozoans. The faunal elements seen in thin sections of the Norwegian material are somewhat similar; fragments of brachiopods, bryozoans, corals, echinoderms, ostracods, trilobites, and colonies of the blue-green alga Girvanella sp. The sediment in which the Swedish specimens of R. camera are found occur as crackfillings in the Upper Ordovician (Harjuan) Boda limestone (Regnell 1947). The fissures are filled with graptolitic shales belonging to the Middle Llandovery (Thors- lund 1960). These shales sometimes contain concretionary horizons of dark lime- stone, and the biosparite from Dalarna with R. camera probably comes from such limestone nodules. In the Oslo region the same microfossils are found in the uppermost part of stage 6. Bassett and Rickards (1971) correlated stage 6c of the Oslo district with the Middle and Upper Llandovery (B3-C2/3) beds of the Llandovery district. If the Swedish and the Norwegian specimens of R. camera are of the same age, then this could sup- port the view of Thorslund (1960) that in Dalarna the Lower Llandovery is missing above the Boda limestone. Acknowledgements. I am indebted to Professor Gerhard Regnell, University of Lund, Sweden, for letting me borrow the Swedish specimens from Dalarna. I am grateful to Dr. Svein B. Manum for his help in preparing parts of this manuscript. 1 would also like to thank Drs. David Bruton and David Worsley for their criticism and help. REFERENCES bassett, m. g. and rickards, R. b. 1971. Notes on Silurian stratigraphy and correlation in the Oslo district. Norsk. Geol. Tidsskr. 51, 247-260. horowitz, A. s. and potter, p. e. 1971. Introductory Petrography of Fossils. Springer-Verlag. Berlin- Eleidelberg-New York. lauritzen, 0. and worsley, d. 1974. Algae as depth indicators in the Silurian of the Oslo region. Lethaia , 7, 157-161. regnell, G. 1947. Some problematic micro-fossils from the Silurian of Dalarna. Kungl. Fysiografiska Salsk. i Lunds Forhandl. bd. 17, nr. 5. S 714 PALAEONTOLOGY, VOLUME 17 st0rmer, l. 1953. The Middle Ordovician of the Oslo region, Norway. 1. Introduction to Stratigraphy. Norsk Geol. Tidsskr. 31, 37-141. thorslund, p. 1960. Notes on the Geology and Stratigraphy of Dalarna. Intern. Geol. Congr. 21st Session, Norden. Swedish geol. guidebooks. Typescript received 2 November 1973 0RNULF LAURITZEN Geological Institute University of Oslo Postbox 1047, Blindern Oslo 3 Norway THE AUSTRALIAN TABULATE CORAL GENUS HATTONIA by J. w. pickett and j. s. jell Abstract. Hattonia Jones, reinterpreted on type and topotype material of Hattonia etheridgei Jones, the type species, is a Silurian and Devonian favositid. It is characterized by distant groups of tabulae developed at the same level throughout the corallum and by pores which are confined to these levels. The genus is endemic to eastern Australia. Two new species, H. fascitabulata from the lower Gedinnian of New South Wales and H. spinosa from the Emsian of north Queensland, are referred to it. Favositid-like tabulate corals with tabulae developed at the one level throughout their corallum have a global distribution and are referred to a variety of genera. Many of these genera are in urgent need of revision before their taxonomic and stratigraphic significance can be assessed. One such genus is Hattonia , originally referred to the family Chaetetidae by Jones when he first described it in 1927. Subsequently it has been variously referred to the Chaetetidae, Favositidae, or Lichenariidae by authors revising the genus without access to the type material. Re-examination of the holo- type, further collections of the type species from the type locality and elsewhere, and recent discovery of other species referable to the genus have allowed more precise definition of this genus and a better understanding of its taxonomic position and distribution. All measurements of corallite diameter are made diagonally angle to angle, median dark line to median dark line (as seen in transmitted light). Wall thickness is given as the total thickness of the dividing wall at the mid-point of the side. Specimens prefixed UQF are housed in the Palaeontological collections of the Department of Geology and Mineralogy, University of Queensland, Brisbane; those prefixed MMF are stored at the Geological and Mining Museum, Sydney, New South Wales. Permission to publish (for J. W. Pickett) was granted by the Under Secretary, New South Wales Department of Mines. Family favositidae Dana, 1846 Genus hattonia Jones, 1927 Hattonia Jones 1927, p. 438; Lang, Smith and Thomas 1940, p. 65; Bassler 1944. p. 48; Sokolov 1947, p. \lbl , pars', 1949, p. 83 , pars', Lecompte 1952, p. 517; Sokolov 1955, p. 154, pars'. Hill and Stumm 1956, p. F455; Mironova 1957, p. 88, pars', Sokolov 1962, p. 222, pars', Scharkova in Bogdanov 1963, p. 149, pars'. Hill, Playford and Woods 1967, p. d6. Type species (by monotypy): Hattonia etheridgei Jones, 1927, from the ‘Barrandella Shale’ (upper part of the Silverdale Formation, Link 1970), Hatton’s Corner, Yass, New South Wales; Middle Ludlow (Jaeger 1967; Link 1970). Revised diagnosis. Massive favositid corals with small corallites; tabulae in groups of 2 to 4 (rarely singly or in 5 or 6), at similar levels in adjacent corallites and indeed throughout the corallum; mural pores large, always associated with the groups of tabulae; septal apparatus absent or developed as several rows of short spines. [Palaeontology, Vol. 17, Part 3, 1974, pp. 715-726, pis. 103-105.] 716 PALAEONTOLOGY, VOLUME 17 Discussion. Hattonia was originally referred to the family Chaetetidae (rather than the Favositidae) as Jones (1927) considered mural pores and septal structures to be lacking. Hill and Stumm (1956) followed Jones without comment. However, as early as 1944 Bassler remarked in a consideration of various poorly known tabulates that he ‘suspected that a tangential section . . . will show favositoid characters’ (p. 48) and included the genus in the Favositidae. This he did, however, without ever having seen material belonging to the genus. A second species, Hattonia marinae Sokolov, 1947, p. 1766, figs. 1, 2, from the Upper Silurian, Southern Fergana, U.S.S.R., was referred to the genus by Sokolov in 1947. (Precise details of locality and stratigraphic horizon are lacking. The age is variously given as Wenlock, Wenlock-Ludlow, Upper Silurian.) In so doing, Sokolov amended the definition of the genus so that forms referred to it should bear mural pores in a single row. He concluded that Hattonia could not belong to the Chaetetidae because of the occurrence of intermural budding in the type species, although the two features emphasized by Jones— absence of pores and lack of septal structures— both indicate affinity with that family. Sokolov considered that mural pores are likely in H. etheridgei , but had escaped Jones’s notice because of the inferior quality of his thin sections. The figures given by Jones, according to Sokolov, even suggest the presence of pores. Sokolov concluded that Hattonia belonged to the Favositidae, to which he continued to refer it (1955, 1962). In 1952 Lecompte placed Hattonia in the family Lichenariidae, believing mural pores and septal apparatus to be lacking. Sokolov (1955, p. 236) disagreed with this, pointing out that the group became extinct before Upper Ordovician times. A re-examination of the holotype of H. etheridgei substantiates the remarks made by both Bassler and Sokolov on the structure and family position of the genus. Mural pores are definitely present, mostly occurring between the pairs of tabulae. In all species now referred to the genus this same association of groups of tabulae and mural pores is seen. Structures which might be termed ‘lateral tabulae’ by analogy with the lateral dissepiments of some rugosans, in that they do not cross the lumen, but are blister-like in longitudinal section (see text-fig. 1 a, b, e), are reminiscent of syringoporoids, particularly when these are opposite one another (text-fig. 1 a and PI. 104, figs. 1, 3), and give the impression of a syrinx. These ‘lateral tabulae’, when they occur, usually lie between adjacent mural pores, like a syrinx communicating with that of the adjacent corallite. However, such corallites have only been seen to communicate with the lumen of a corallite with normal tabulae. Struc- tures which present a syrinx-like appearance in transverse section prove to be sections of domed or saucered tabulae, with no vertical continuity. Further, a ‘lateral tabula' may occasionally span a mural pore (text-fig. 1 a, b ), thus disrupting communica- tion between syrinx-like structures of adjacent corallites. These are favositoid rather than syringoporoid features, according to Hill and Jell (1970) who summarized the important features which may be used in distinguishing the superfamilies Favo- sitoidea and Syringoporoidea. The usually flat tabulae, occasional septal spines, and direct communication between the lumina through the mural pores are strongly indicative of favositoid affinities. Hattonia is distinguished from other favositids by the grouped tabulae, occurring at similar levels through the colony and the associa- tion of mural pores with these levels. PICKETT AND JELL: TABULATE CORALS 717 Grouping of tabulae in tabulate corals has been discussed by Tong-Zyui (1965, p. 44), who ascribes it to periodic growth, pointing out that there is often a thickening of skeletal elements corresponding to the halts in growth. This last is not exclusive to tabulates, as it also occurs in rugose corals. Although the growth of Hattonia may have been periodic, the grouping of the tabulae and their association with the mural pores is a genotypic feature, and not in any way imposed by external factors such as climate or seasons. In H. fascitabulata thickening of the walls does seem to be associated with rate of growth, but grouping of the tabulae is quite independent of this thickening. Tong-Zyui also points out that ‘level’ tabulae, i.e. those occurring at similar levels in adjacent corallites, have been reported in many genera of tabulates ; Dictyofavosites, Hattonia , Dania , Laceripora , Paleofavosites, Mesosolenia , Favosites, Squameofavosites, Sapporipora , Pachyfavo sites , Parastriatopora, Eehyropora , and Caliapora. The occurrence of this feature in such a variety of genera obviously does not imply consanguinity. On the other hand, those genera which for other reasons are held to be closely related ( Favosites , P achy favosites, Squameofavosites , Dictyo- favosites) seem, with the exception of Dictyofavosites , to indicate that ‘level’ tabulae may have arisen independently. Dictyofavosites Chernyshev, 1951 with type species D. salairieus Chernyshev, 1951, p. 37, pi. 9, figs. 1-2 from the Lower Devonian (given as Upper Silurian by Cherny- shev), above the mouth of the rivulet Khvoshchevki, River Pavlova, Salair, Kuznets Basin, U.S.S.R., is characterized by distant tabulae developed at the same level in adjacent corallites throughout the corallum and by mural pores occurring in regularly spaced series along the length of the corallite without relation to the tabular levels. Thus we do not consider it a synonym of Hattonia as did Sokolov (1947, 1955) when he referred H. marinae Sokolov, 1947, p. 1766, figs. 1, 2 to it. This latter species is more closely related to D. salairieus than to H. etheridgei. Similarly, the two species referred tentatively to Hattonia by Scharkova (1963), H. elegans and H. parvula, from the Lower Ludlow of the River Kulun-Bulak and the River Ayagus respectively, of the southern slopes region of Tarbagatay, U.S.S.R., do not show the grouping of tabulae and the associated mural pores which we consider characteristic of Hattonia , although the tabulae do occur at similar levels through the colony, as in H. marinae. These three Russian species seem to be closely related to each other, but they are not congeneric with H. etheridgei. The longitudinal section of the holotype of Favosites ( Salairia ) peetzi figured by Chernyshev (1951, p. 38, pi. 9, figs. 5, 6) from the Lower Devonian of Salair, Kuznets Basin, U.S.S.R., and type species of his subgenus Salairia , suggests the presence of paired tabulae. However, one of us (J. S. J.) has examined the original slide and con- siders this appearance is due to the recrystallization of the specimen. The preservation is too poor to determine if mural pores are associated with these levels. It seems probable that this species is not related to Hattonia ; Sokolov (1962) continued to consider it a subgenus of Favosites. Species referred to Hattonia : Hattonia etheridgei Jones, 1927. Hattonia fascitabulata sp. nov. Hattonia spinosa sp. nov. 718 PALAEONTOLOGY, VOLUME 17 Species not referred to Hattonia : Hattonia marinae Sokolov, 1947, p. 1766, figs. 1, 2; 1949, p. ?, figs. ?; 1955, p. 154, figs. 21a, b; 1962, p. 222, fig. 18. Hattonia elegans Scharkova in Bogdanov, 1963, p. 149, pi. 22, figs. 3-6. Hattonia parvula Scharkova in Bogdanov, 1963, p. 150, pi. 23, figs. 1, 2. Distribution '. Wenlock to Emsian of eastern Australia. Hattonia etheridgei Jones, 1927 Plate 103, figs. 1-5; Plate 104, fig. 1 ; Plate 105, fig. 1 1927 Hattonia etheridgei Jones, p. 438, pi. 12, figs. 1-3. 1944 Hattonia etheridgei Jones; Bassler, p. 48, figs. 12, 13. 1952 Hattonia etheridgei Jones; Lecompte, p. 517, fig. 35. 1956 Hattonia etheridgei Jones; Hill and Stumm, p. F455, fig. 345 (3 a, b ). Holotype. Specimen and three slides UQF 7200 (No. A15 of Jones’s Collection) from the ‘Barrandella Shale’ (upper part of the Silverdale Formation of Link, 1970), Hatton’s Corner, Yass, New South Wales; Middle Ludlow (Jaeger 1967; Link 1970). Diagnosis. Hattonia with slender corallites (0-8 to 10 mm occasionally up to 1-4 mm in diameter), dividing walls thin (0-5 to 01 mm); tabulae mostly grouped in distant pairs with associated large circular mural pores; septal spines absent. Description of holotype. The holotype is a fragment 7x6x5 cm of a large massive favositid colony. The weathered surface shows a distinct and regular lamellation transverse to the length of the corallites (PI. 103, fig. 3). The lamellae correspond to the development of the tabulae at the one level throughout the corallum. The corallites are slender and polygonal (four- to seven-sided but more commonly six-sided), the sides are nearly always straight. The corallites range from 0-8 to 10 (mean 0-92) mm in diameter. The wall is quite thin, 0 05 to 0T mm thick in the centre of the faces and widening slightly towards the angles. In longitudinal section the corallite walls are parallel and straight or slightly curved. The tabulae tend to be grouped in pairs 0-4 to 0-5 mm apart with the pairs 1-6 to 1-9 mm apart, and are developed at the same level in adjacent corallites. In places the pair may not be developed, or only the top or bottom tabula of the pair is present. In others, another plate may be developed between the pair and may be based on only one wall, based on the wall on one side and the tabula beneath, or based on two tabulae. Mural pores are developed in a single series in the centre of the faces between the paired tabulae. They are circular, 0-25 to 0-3 mm in diameter and in transverse section occasionally show a median diaphragm. Recrystallization has obscured the microstructure of the wall. In places small dark patches which do not project into the lumen are seen in tangential sections of the wall ; these may represent the bases of septal spines. EXPLANATION OF PLATE 103 Figs. 1-3. Hattonia etheridgei Jones, holotype, UQF7200 from the ‘Barrandella Shale’ (Upper part of Silverdale Formation), Hatton’s Corner, Yass, New South Wales, Middle Ludlow. 1, transverse thin section, x 4. 2, longitudinal thin section. Mural pores can be observed in several places, e.g. at the left between the tabulae of the top pair, between the first and second and the second and third corallites (arrow), x4. 3, weathered surface, xl. Fig. 4. Hattonia etheridgei Jones, paratype, UQF35754 from probably the ‘Hume Limestone’ (top of Silverdale Formation), Limestone Creek, a tributary of Derringullen Creek, near Yass, New South Wales, Middle Ludlow. Longitudinal thin section showing less regular arrangement of tabulae, x 4. Fig. 5. Hattonia etheridgei Jones, MMF15779 from the Mirrabooka Formation, Cheesemans Creek, west of Orange, New South Wales, Upper Wenlock. Transverse thin section showing a number of average corallites, x4. PLATE 103 PICKETT and JELL, Hattonia 720 PALAEONTOLOGY, VOLUME 17 Supplementary material. Five paratypes from Jones’s collection, now sectioned; these are UQF3755 and Bristol University, Department of Geology Collection 6771 from the type locality, UQF3756 and UQF6471 from Derringullen Creek, near Yass, New South Wales, probably from the ‘Flume Limestone’ (top of Silverdale Formation), and UQF3754 from Limestone Creek, a tributary of Derringullen Creek, probably also from the same horizon. MMF15745 from the ‘Hume Limestone’, Derringullen Creek, and MMF15763, MMF15766, MMF15768, MMF15774, MMF15775, MMF15779, and MMF15785 from Limestone Member D, Mirrabooka Formation, Cheesemans Creek, west of Orange, New South Wales, of probable Wenlock age (Sherwin 1971). Variation in supplementary material. The largest specimen from the Yass area, MMF15745, measures 12x10x13 cm, but is part of a much larger corallum. The variation in features such as size of corallites, thickness of wall and diameter of mural pores is similar to the holotype. The disposition of the tabulae in other specimens (UQF3755, MMF15745) is less regular than in the holotype. In these, although paired tabulae are frequent, they occur more commonly in groups of three or four, and the number of tabulae of irregular shape is much greater (see text-fig. 1). The Mirrabooka Formation specimens show some text-fig. 1. Arrangement of tabulae in Hattonia spp., x2-5 approx.: (a) H. spinosa sp. nov., UQF63993, Emsian, Pandanus Creek, north Queensland, (b) H. fascitabulata sp. nov., MMF14761, Gedinnian, west of Yass, New South Wales, (c, d, e) H. etheridgei Jones, MMF15774, MMF15766, MMF15768, Mirra- booka Formation, Cheesemans Creek, New South Wales. (/) H. etheridgei Jones, MMF15745, ‘Hume Limestone’, Derringullen Creek, west of Yass, New South Wales, (g) H. etheridgei Jones, UQF3755, Formation uncertain, Hatton’s Corner, Yass, New South Wales. PICKETT AND JELL: TABULATE CORALS 721 differences from those from Yass, though these are too slight to warrant recognition even as a subspecies. They are : (i) The average distance between the groups of tabulae is greater, usually over 2 mm and occasionally exceeding 3 mm. (ii) Rarely, an isolated corallite conspicuously larger than its fellows may occur. The normal diameter is very close to 1 mm, and the size is remarkably constant, so that a larger corallite is quite noticeable. The largest observed has a diameter of 1-4 mm (MMF15763). (iii) Occasionally, one side wall may fail to develop, so that two adjacent corallites together present an hour-glass shape in transverse section, reminiscent of Multisolenia. This degree of communication between corallites is much greater than that through the mural pores. In longitudinal section this is expressed by the absence of the wall for 5 mm or so (see PI. 104, fig. 1). The largest specimen also comes from this locality, an incomplete colony measuring 42 x 29 x 12 cm (MMF15785). The specimens from the Mirrabooka Formation are the oldest known representatives of the genus and here they form the dominant element of the fauna, in both number and size of colonies. The age (Sherwin 1971) is Upper Wenlock. Certainly the beds are overlain by sediments containing halysitids (Limestone Member E), whereas these do not occur as high as the Silverdale Formation in the Yass sequence. Other corals occurring in Limestone Member D are not helpful in establishing a correlation, on the basis of our present knowledge of the faunas. Other species present include Phaulactis sp. nov., Tryplasma cf. derrengullenensis Etheridge Jr., Favosites spp. (a large and a small species, neither closely comparable with any described from Australia so far). Alveolites sp., Heliolites cf. daintreei Nicholson and Etheridge (a form not closely comparable with any of the ‘groups’ of Jones and Hill 1940), Propora conferta Edwards and Haime and stromatoporoids. Increase. All specimens of H. etheridgei examined are fragments of mature colonies, in which division of the corallites is rare. Serial sections made at intervals of 0-2 mm through specimens from Cheesemans Creek (MMF15779) and Derringullen Creek (MMF 15745) revealed only one corallite which appears to be dividing (see text-fig. 2). Even a distance of 0-2 mm between sections was insufficient to disclose all details of division, which is accomplished in this interval, i.e. less than one-fifth of the corallite diameter. The parent corallite in the second section figured gives no indication (not even greater size) that the division is about to occur. text-fig. 2. Division in Hattonia etheridgei Jones: MMF15745 from Derringullen Creek, west of Yass, New South Wales. The sections are separated by an interval of 0-2 mm. The separation of the new corallite from the parent is entirely accomplished within this short interval. Hattonia fascitabulata sp. nov. Plate 105, figs. 2-3 Holotype. MMF14761 from the Lower Gedinnian Elmside Formation (Link 1970), 1-2 km south of the Hume Highway, 0-8 km west of the junction with the Boorowa road, west of Yass, New South Wales (GR1 85701, Goulburn 1 : 250,000 sheet). Only a single specimen is known. Diagnosis. Hattonia with mostly six-sided corallites 0-5 to 0-7 mm in diameter ; thick walls (up to 0-20 mm) ; mural pores almost as wide as the sides of the corallites (0-2 to 0-22 mm in diameter) ; tabulae occurring in groups of three to six; septal apparatus virtually absent. 722 PALAEONTOLOGY, VOLUME 17 Description. Details of the appearance of the corallum are not known. The holotype is an eroded, rounded fragment 1 1 x 10 x 7-5 cm. The corallites are slightly sinuous ; this and zones of thickening and pigmentation indicate that the surface of the corallum was undulose. The corallites are four- to six-sided, regular in size, being 0-5 to 0-7 mm in maximum diameter, rarely reaching as much as 0-8 mm. The walls are generally thickened, more so in the angles than on the sides, so that the lumen varies from polygonal in the rarer unthickened areas to sub-rounded in areas of considerable thickening. In unthickened areas the wall is just under 0- 1 mm thick at the centre of the face, in thickened areas it may reach 0-2 mm. The walls, especially where thickened, show a central light line in transmitted light, somewhat wider than the usual dark line in other favositids : this is, however, a feature of the preservation rather than a result of a structural difference. Between this central clear line and the lumen the wall is fibrous with the fibres normal to the surface of the wall. The boundary between the clear central line and the outer fibrous area is not distinct. Growth lines normal to the fibres parallel the inner edge of the lumen and at low magnification give the wall a lamellar appearance. Mural pores are rounded or slightly elongated vertically, 0-20 to 0-22 mm in diameter. They occur always in a zone of tabulae, and at the same level throughout the colony. There is only one per side of a corallite, as the pores are almost as wide as the sides. Pores occur on several faces of one corallite at the same level, resulting in a localized apparent meandroid pattern in transverse section (see PI. 105, fig. 2). No pore plates have been recognized. Septal structures are poorly developed, being reduced to rare septal spines, which occur in vertical rows. These have a patchy distribution through the colony. They are short, scarcely reaching 0-2 mm, and less than 0- 1 mm in diameter, and so inconspicuous that their presence was overlooked until the larger spines of H. spinosa were seen. Tabulae occur in groups of up to six, the most common number being four. The distance apart both of tabulae within a group and of adjacent groups varies considerably. In the former case the tabulae are usually less than 0-5 mm apart or as close as 0- 1 mm, but even up to 0-7 mm. The latter measured most accurately as the distance apart of the mural pores, is 2 to 4-5 mm. Tabulae are very thin and their shape is variable in the extreme. They may be concave, convex, may bear a marked angular down- ward deflection, or even sit blister-like on the wall. In this latter case the upper and lower points of attach- ment, as observed in longitudinal section, frequently lie above and below a mural pore. Remarks. Where the walls are thickened at the angles the transverse sections of H. fascitabulata are similar to those of Pachyfavosites Sokolov but where unthickened the polygonal corallites resemble the thicker- walled corallites of H. etheridgei. This rounding of the lumen by the thickening of the wall especially at the angles is considered of no more than specific significance. The slender corallites and generally thick walls as well as the more irregular grouping of commonly four tabulae at various levels distinguish H. fascitabulata from the type species of the genus. table 1 . Comparative values of various parameters for three species of Hattonia. Species Corallite Wall Mural Interval Interval diameter thickness pore between between diameter grouped tabular tabulae groups //. etheridgei 0-8-1-0 005-0- 1 0-25-0-3 0-4-0-5 1 -6-1-9 H. fascitabulata 0-5-0-7 007-0-2 0-2-0-22 01-0-5 2-0-4-5 H. spinosa 0-9-1-4 0-1-0-15 0-25-0-3 0-4-0-5 1 -5-2-5 EXPLANATION OF PLATE 104 Fig. 1. Hattonia etheridgei Jones, MMF15774 from the Mirrabooka Formation, Cheesemans Creek, west of Orange, New South Wales, Upper Wenlock. Longitudinal thin section showing increased number of tabulae in some groups, ‘lateral’ tabulae, and missing sections of wall, x 4. Figs. 2, 3. Hattonia spinosa sp. nov., Holotype, UQF50894, Martins Well Limestone Member, Broken River Formation, at Martins Well, 8 km north-west of Pandanus Creek homestead, north Queensland, Emsian. 2, transverse thin section, x 4. 3, longitudinal thin section, x 4. Figs. 4, 5. Hattonia spinosa sp. nov., paratype UQF63992, same loc. as figs. 2, 3. 4, transverse thin section, x 4. 5, longitudinal thin section, x 4. PLATE 104 PICKETT and JELL, Hattonia 724 PALAEONTOLOGY, VOLUME 17 Hattonia spinosa sp. nov. Plate 104, figs. 2-5 1967 Hattonia sp. nov. Hill, Playford and Woods, p. d6, pi. D3, fig. 5a, b. Holotype. UQF50894 from the Emsian (Jell 1968; Strusz 1972) Martins Well Limestone Member, Broken River Formation, at Martins Well, 8 km north-east of Pandanus Creek homestead, north Queensland (GR61 1847 Clarke River 1 : 250,000 sheet). Other material. Three specimens UQF63992, UQF63993, and UQF63994, from the same locality. Diagnosis. Hattonia with numerous short horizontal septal spines; mural pores oval, in a single vertical series in the centre of each face; tabulae tending to be grouped in pairs with incomplete or sometimes com- plete tabulae between pairs. Description. The holotype is an abraded part (10x10x15 cm) of a tall corallum with corallites radiating upwards and outwards from its base. The distal surface of the colony was probably domed. The other specimens are fragments of probably similar coralla. The corallites are polygonal, four- to six-sided, the sides in some being a little curved between the angles, and are 0-9 to 14 (mean 102) mm in diameter. The walls vary in thickness from 01 to 0T5 mm in the centre of the faces and thicken towards the angles, which in some become rounded. In transverse section the wall consists of a thin median dark line (as seen in transmitted light) which is the same width throughout but discontinuous and may be represented by a line of closely spaced dark spots, and a wider zone of lighter-coloured material on either side which thickens towards the angles. In most places the microstructure of the wall is not obvious; however, in parts where the central dark line is represented by dark points, lighter tissue on either side seems to consist of fibres radiating outwards from these points. In longitudinal section, tangential sections of the wall show a fibrous structure with the fibres directed upwards and inwards from the angles towards the centre of the faces. Whether the fibres are grouped into trabecular bundles or are all parallel is not known. This structure resembles that seen in the Australian species of Squameofavosites. Juvenile corallites develop by the insertion of a new partition across the angle of an adult corallite. Septa are represented by numerous short, blunt, nearly horizontal septal spines arranged in two or three subopposite to alternating vertical series on each face. Each spine seems to have a median dark line and is possibly trabeculate. The mural pores are commonly oval 0-25 to 0-3 mm in width and 0-3 to 0-35 mm in height, or less commonly circular, 0-3 mm in diameter. They are arranged up to 2-5 mm apart in a single vertical series in the centre of each face and seem to be developed at the same level as the tabulae. The tabulae are grouped in pairs and tend to be at or nearly at the same level in adjacent corallites. However, there is only a general regularity in their arrangement and local disparities are common. The lower tabula of a pair is usually saucered and the upper horizontal or arched; they are usually 0-4 to 0-5 mm apart. Only occasionally are there any tabulae developed between those of a pair. The pairs are 1-5 to 2-5 mm apart and sometimes incomplete tabulae are developed between them, which are commonly convex in longitudinal section and are usually based above and below on the one wall. Remarks. Commensal worm tubes similar to those in Favosites duni (Etheridge Jr.) are seen occasionally in the wall at or near the angles of the corallites. This species differs from the type species in its slightly larger corallites, more numerous septal spines, although their absence in the holotype of the type species may be due to the recrystallized nature of the specimen, less regularity in the arrangement of the tabulae and the more common incomplete tabulae between tabular pairs. EXPLANATION OF PLATE 105 Fig. 1. Hattonia etheridgei Jones. MMF 15766 from the Mirrabooka Formation, Cheesemans Creek, west of Orange, New South Wales, Upper Wenlock. Longitudinal thin section showing regular disposition of tabulae, x 4. Figs. 2, 3. Hattonia fascitabulata sp. nov., holotype, MMF14761 from the Elmside Formation, 1-2 km south of the Hume Highway, 0-8 km west of the junction with the Boorowa road, west of Yass, New South Wales, Gedinnian. 2, transverse thin section, x 4. 3, longitudinal thin section, x 4. PLATE 105 PICKETT and JELL, Hattonia 726 PALAEONTOLOGY, VOLUME 17 REFERENCES bassler, R. s. 1944, Parafavosites and similar tabulate corals. J. Paleont. 18, 42-49, text-figs. 1-29. bogdanov, A. A. (ed.). 1963. Stratigrafiya i fauna paleozoyskikh otlozheniy Khrebta Tarbagatay ( Ordovik , Silur, Devon , nizhney Karbon). 1-471, pis. 1-67. Gosgeoltekhizdat, Moscow, Vses. aerogeol. trest minist. geol. okhrany SSSR. Chernyshev, B. B. 1951. Siluriyskie i devonskie Tabulata i Heliolitida okrain Kuznetskogo uglenosnogo basseyna. 1-160, pis. 1-26. Gosgeolizdat, Moscow. Vses. nauchno-issled. geol. Inst. hill, d. and jell, j. s. 1970. The tabulate coral families Syringolitidae Hinde, Roemeriidae Pocta, Neo- roemeriidae Radugin and Chonostegitidae Lecompte, and Australian species of Roemeripora Kraicz. Proc. R. Soc. Viet. 83, 171-190, pis. 16-20. — playford, G. and woods, j. t. (eds.). 1967. Devonian fossils of Queensland, dl-d32, pis. D1-D15. Qd Palaeontogr. Soc., Brisbane. — and stumm, e. c. 1956. Tabulata. In moore, r. c. (ed.). 1956. Treatise of invertebrate paleontology. Part F, Coelenterata. F444-F447. Geol. Soc. Amer. and Univ. Kansas Press, Lawrence, Kansas. jaeger, h. 1967. Preliminary stratigraphical results from graptolite studies in the Upper Silurian and Lower Devonian of south eastern Australia. J. geol. Soc. Aust. 14, 281-286. jell, j. s. 1968. New Devonian rock units of the Broken River embayment, north Queensland. Qd Govt Min. J. 69, 6-8. jones, o. a. 1927. A new genus of tabulate corals from New South Wales. Geol. Mag. 64, 438-440, pi. 12. — and hill, d. 1940. The Heliolitidae of Australia, with a discussion of the morphology and systematic position of the family. Proc. R. Soc. Qd, 51, 183-215, pis. 6-11. lang, w. d., smith, s. and thomas, h. d. 1940. Index of Palaeozoic coral genera. 1-231. British Museum (Natural History). lecompte, m. 1952. Madreporaires paleozoiques. In piveteau, j. (ed.). 1962. Trade de Paleontologie. Vol. I, Generalites, Protistes, Spongiaires, Coelenteres, Bryozoaires. 419-538, Paris. link, a. g. 1970. Age and correlations of the Siluro-Devonian strata in the Yass Basin, New South Wales. J. geol. Soc. Aust. 16, 71 1-722. mironova, n. v. 1957. O favozitidakh devona tsentral’nogo Salaira. Vest, zap.-sib. geol. upr. 1, 85-89. sherwin, L. 1971. Stratigraphy of the Cheesemans Creek district, New South Wales. Rec. geol. Surv. N.S. W. 13, 199-237, pis. 1-3. sokolov, b. s. 1947. Rod Hattonia Jones i ego sistematicheskoe polozhenie. Dokl. Akad. Nauk SSSR. 58, 1765-1768. — 1949. Tabulata i Heliolitida Siluria SSSR. In Atlas rukovodyashchikh form iskopaemykh faun SSSR. 2, 75-98, pis. 6-10. Gosgeolizdat, Moscow. — 1955. Tabulyaty paleozoya Evropeyskoy chasti SSSR. Vvedenie. Obshchie voprosy sistematiki i istorii razvitiya tabulyat (s kharakteristikoy morfologicheski blizkikh grupp). Trudy vses. neft. nauchno- issled. geologorazv. Inst, (n.s.) 85, 1-527, pis. 1-90. — 1962. Klass Anthozoa, korallovye polipy, podklass Tabulata, tabulyaty (Aseptata, Trichocorallia). In ORLOV, Y. A. (ed.) 1962. Osnovy paleontologii, 2, Gubki, Arkheotsiaty, Kishechnopolostnye chervi. 192-265, pis. 1-18. Akad. Nauk SSSR, Moscow. strusz, d. l. 1972. Correlation of the Lower Devonian rocks of Australia. J. geol. Soc. Aust. 18, 427-455. tong-zyui tkhan’ (tong-dzuy than) 1965. O raspolozhenii dnishch u tabulyatomorfnykh korallov. Paleont. Zh. 1965, 44-47, pi. 2. J. W. PICKETT Geological and Mining Museum Geological Survey of New South Wales Department of Mines George Street North Sydney, N.S.W. 2000 J. S. JELL Department of Geology and Mineralogy University of Queensland St. Lucia, Queensland, 4067 Revised typescript received 15 July 1973 SHORT COMMUNICATION THE LOWER CRETACEOUS AMMONITE GENERA PROPOSED BY C. JACOB IN 1907 by M. K. HOWARTH Abstract. It is confirmed that the Lower Cretaceous ammonite genera Jauberticeras , Kossmatella , Latidorsella, Leymeriella , and Uhligella were proposed by Jacob in a paper first published in June 1907 giving them priority over their publication by Kilian in October 1907. During revision for the ammonoid volume of the Treatise of Invertebrate Paleon- tology investigations were made into the five Lower Cretaceous ammonite genera Jauberticeras , Kossmatella , Latidorsella , Leymeriella , and Uhligella said to have been proposed by Jacob in 1907. The traditional view (e.g. Spath 1925, pp. 75, 83; Casey 1957, p. 30; Casey 1961, p. 160) that they were first published in 1907 by Jacob in his paper in Trav. Lab. Geol. Univ. Grenoble , vol. 8, fasc. 2, pp. 280-590, was found to be erroneous because that paper was published in 1908, as is indicated by the date printed on the cover of fasc. 2. All five generic names appeared in Kilian (1907), the publication date being 26 October 1907, as printed on the cover and confirmed by the date received stamp of 9 November 1907 on the copy in the library of the Palaeon- tology Department of the British Museum (Natural History). Thus the date of Kilian’s work is not in doubt and clearly antedates Jacob’s paper, so further investi- gations were made into the publication of Jacob’s work. This was Jacob’s thesis and was entitled ‘Etudes paleontologiques et strati- graphiques sur la partie moyenne des terrains cretaces dans les alpes frangaises et les regions voisines’. It was set in type by Allier Freres of Grenoble and appeared in four different versions (all have 6 plates and about 314 pages, and the same typographical details in the three published versions show that they were all printed-off from the same type) : 1. Some copies were distributed as his thesis, which he submitted in Paris on 1'3 June 1907 (Casteras and Laffitte 1964, p. 691). This version is not a publication for purposes of zoological nomenclature, but a copy was seen by the compilers of the Zoological Record (vol. 45 for 1908, Mollusca p. 124) and led to quotations of Jacob’s generic names with wrong page numbers, e.g. ^Jauberticeras— Ann. Univ. Grenoble, 19, p. 65’. 2. Jacob (1907): Annls Univ. Grenoble , vol. 19, fasc. 2, pp. 221-534, 6 pis. The only copy of this volume in the British Isles was lent to me by Trinity College, Dublin, but the original wrapper for fascicule 2 is missing and there is no evidence for its exact date of publication. However, Dr. J.-P. Thieuloy of Grenoble University has kindly informed me that there is evidence at Grenoble that fascicule 2 was published in June 1907. This settles the priority of Jacob’s publication of the generic names over their publication by Kilian in October 1907. [Palaeontology, Vol. 17, Part 3, 1974, pp. 727-728.] 728 PALAEONTOLOGY, VOLUME 17 3. Jacob (1908a) : Trav. Lab. Geol. Univ. Grenoble , vol. 8, fasc. 2, 1908, pp. 280- 590, 6 pis. The date 1908 is printed on the cover of fascicule 2. This is the version of Jacob’s paper always quoted hitherto, usually wrongly quoted as published in 1907. 4. Jacob (19086): Bull. Soc. Statist. Sci. nat. Arts ind. Dep. Isere , Grenoble, ser. 4, vol. 10, pp. 201-514, 6 pis. Dated 1908 on the cover, and was probably late in 1908 because the volume contains a report of a meeting held on 27 January 1908. Longer descriptions and figures of the five genera appeared in Jacob (1908c). Modern taxonomic details of the five new ammonite genera which appeared in the first version (Jacob 1 907a) of Jacob’s paper are as follows: 1. Jauberticeras Jacob, 1907, p. 285: type species Ammonites jaubertianus d’Orbigny, 1850, p. 200, by original designation. 2. Kossmatella Jacob, 1907, p. 285; type species Ammonites agassizianus Pictet, 1847, p. 303, subsequently designated by Roman, 1938, p. 43. 3. Uhligella Jacob, 1907, p. 293; type species Desmoceras clansayense Jacob, 1905, p. 403, subsequently designated by Kilian, 1907, p. 63. 4. Latidorsella Jacob, 1907, p. 295; type species Ammonites latidorsatus Michelin, 1838, p. 101, by tautonomy. 5. Leymeriella Jacob, 1907, p. 311; type species Ammonites tardefurcatus d’Or- bigny, 1841, p. 248, subsequently designated by Spath, 1925, p. 75. REFERENCES casey, r. 1957. The Cretaceous ammonite genus Leymeriella , with a systematic account of its British occurrences. Palaeontology , 1, 28-59, pis. 7-10. — 1961. A monograph of the Ammonoidea of the Lower Greensand. Part 3, pp. 119-216, pis. 26-35. Palaeontogr. Soc. ( Monogr .) casteras, m. and laffitte, R. 1964. Charles Jacob. Bull. Soc. geol. Fr. (7) 5, 663-694. Jacob, c. 1905. Etudes sur les ammonites et sur Fhorizon stratigraphique du gisement de Clansayes. Bull. Soc. geol. Fr. (4) 5, 399-432, pis. 12, 13. — 1907 (June). Etudes paleontologiques et stratigraphiques sur la partie moyenne des terrains cretaces dans les Alpes franqaises et les regions voisines. Annls Univ. Grenoble , 19 (2), 221-534, 6 pis. 1908o (? early). Same paper. Trav. Lab. Geol. Univ. Grenoble, 8 (2), 280-590, 6 pis. — 1908ft (late). Same paper. Bull. Soc. Statist. Sci. nat. Arts ind. Dep. Isere (Grenoble), (4) 10, 201- 514, 6 pis. — 1908c. Etudes sur quelques ammonites du cretace moyen. Mem. Soc. geol. Fr. 15 (3, 4) (mem. 38), 63 pp., 19 pis. kilian, w. 1907 (26 October). Lethaea geognostica. II Teil, Das Mesozoicum; 3 Band, Kreide; I Abt., Unter kreide, 1 Lief., pp. 1-168. Stuttgart. michelin, h. 1838. Note sur une argile dependant du Gault. Mem. Soc. geol. Fr. (1) 3, 97-103, pi. 12. orbigny, a. d'. 1840-1842. Paleontologie Frangaise. Terrains Cretaces. 1, Cephalopodes. 662 pp., 148 pis. Paris. — 1850. Notes sur quelques nouvelles especes remarquables d’ammonites des etages Neocomien et Aptien de la France. J. Conch. Paris, 1, 196-201, pi. 8. pictet, f.-j. 1847. Description des mollusques fossiles qui se trouvent dans les Gres Verts des environs de Geneve. Mem. Soc. Phys. Hist. nat. Geneve, 11 (2), 257-412, 15 pis. roman, f. 1938. Les ammonites jurassiques et cretacees. 554 pp., 53 pis. Paris. spath, l. f. 1925. A monograph of the Ammonoidea of the Gault. Part 2, pp. 73- 1 10, pis. 5-8. Palaeontogr. Soc. {Monogr.) m. K. howarth Department of Palaeontology British Museum (Nat. Hist.) Cromwell Road London, SW7 5BD Typescript received 20 December 1973 THE PALAEONTOLOGICAL ASSOCIATION The Association was founded in 1957 to further the study of palaeontology. It holds meetings and demonstrations as well as publishing Palaeontology and Special Papers in Palaeontology. Membership is open to individuals and to institutions on payment of the appropriate annual subscription : Institutional membership .... £10 00 (U.S. $26.00) Ordinary membership .... £5-00 (U.S. $13.00) Student membership .... £3-00 (U.S. $8.00) There is no admission fee. Institutional membership is only available by direct application, not through agents. Student members are persons receiving full-time instruction at educational institutions recognized by the Council. On first applying for membership, an application form should be obtained from the Membership Treasurer. Subscriptions cover one calendar year and are due each January; they should be sent to the Membership Treasurer, Dr. E. P. F. Rose, Department of Geology, Bedford College, Regent’s Park, London, NW1 4NS, England. PALAEONTOLOGY All members who join for 1974 will receive Volume 17, Parts 1-4. All back numbers are still in print and may be ordered from B. H. Blackwell, Broad Street, Oxford, OX1 3BQ, England, at £5 per part (post free). A complete set. Volumes 1-16, consists of 63 parts and costs £315. SPECIAL PAPERS IN PALAEONTOLOGY The subscription rate is £8 (U.S. $22.00) for Institutional Members and £4 (U.S. $11.00) for Ordinary and Student Members. Subscriptions should be placed through the Membership Treasurer, Dr. E. P. F. Rose, Department of Geology, Bedford College, Regent’s Park, London, NW1 4NS, England. Ordinary and Student members only may obtain individual Special Papers from Dr. Rose at reduced rates. Non-members may obtain them at the stated prices from B. H. Blackwell, Broad Street, Oxford, OX1 3BQ, England. COUNCIL 1974-1975 President : Professor C. H. Holland, Department of Geology, Trinity College, Dublin, 2, Ireland Vice-Presidents : Dr. R. Goldring, Department of Geology, The University, Reading, RG6 2AB Dr. W. D. I. Rolfe, The Hunterian Museum, The University, Glasgow, G12 8QQ Treasurer-. Dr. J. M. Hancock, Department of Geology, King’s College, Strand, London, WC2R 2LS Membership Treasurer. Dr. E. P. F. Rose, Department of Geology, Bedford College, Regent’s Park, London, NW1 4NS Secretary. Dr. C. T. Scrutton, Department of Geology, The University, Newcastle upon Tyne, NE1 7RU Editors Dr. J. D. Hudson, Department of Geology, The University, Leicester, LEI 7RH Dr. L. R. M. Cocks, Department of Palaeontology, British Museum (Natural History), Cromwell Road, London, SW7 5BD Dr. C. P. Hughes, Department of Geology, Sedgwick Museum, Downing Street, Cambridge, CB2 3EQ Dr. J. W. Murray, Department of Geology, The University, Bristol, BS8 1TR Other Members of Council Dr. M. G. Bassett, Cardiff Dr. M. C. Boulter, London Professor D. L. Dineley, Bristol Dr. J. K. Ingham, Glasgow Dr. J. E. Pollard, Manchester Dr. A. W. A. Rushton, London Dr. P. Wallace, London Dr. D. D. Bayliss, Llandudno Dr. C. H. C. Brunton, London Dr. J. A. E. B. Hubbard, London Dr. C. R. C. Paul, Liverpool Dr. P. F. Rawson, London Professor D. Skevington, Galway Overseas Representatives Australia '. Professor Dorothy Hill, Department of Geology, University of Queensland, Brisbane Canada '. Dr. B. S. Norford, Institute of Sedimentary and Petroleum Geology, 3303-33rd Street NW., Calgary, Alberta 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 30368, Lower Hutt West Indies and Central America '. Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad, Inc., Pointe-a-Pierre, Trinidad, West Indies Western U.S. A. . Professor J. Wyatt Durham, Department of Paleontology, University of California, Berkeley 4, California Eastern U.S. A. . Professor J. W. Wells, Department of Geology, Cornell University, Ithaca, New York Palaeontology VOLUME 17 PART 3 OCTOBER CONTENTS The Upper Palaeozoic tetracoral genera Lophophyllidium and Timorphyllum JERZY FEDOROWSKI 441 Foraminiferal biostratigraphy of the Oligocene-Miocene limestones of Christmas Island (Indian Ocean) C. G. ADAMS and D. J. BELFORD 475 The ecology of a Middle Jurassic hardground and crevice fauna T. J. PALMER and F. T. FURSICH 507 The Lepidodendroid stoma B. A. THOMAS 525 Lower Permian Pelycosaurs from the English Midlands ROBERTA L. PATON 541 The production of stratigraphical range-diagrams by automatic methods IAN E. PENN 553 The Devonian genus Keega (Algae) reinterpreted as a stromatoporoid basal layer ROBERT RIDING 565 Trophic group and evolution in bivalve molluscs JEFFREY S. LEVINTON 579 Leaf anatomy of Weichselia based on fusainized material K. L. ALVIN 587 Ostracods from the Domerian and Toarcian of England ALAN LORD 599 Dinoflagellate cysts from the Aptian type sections at Gargas and La Bedoule, France ROGER J. DAVEY and JEAN-PIERRE VERDIER 623 Chonophyllinid corals from the Silurian of New South Wales R. A. MCLEAN 655 A new Quaternary ostracod genus from Argentina ROBIN C. WHATLEY and TERESA DEL CARMEN CHOLICH 669 The affinities of the trilobite genus Scharyia, with a description of two new species R. M. OWENS 685 A new proboscidean from the late Miocene of Kenya VINCENT J. MAGLIO 699 New microfossils from the Silurian (Llandovery, Stage 6) of the Oslo region, Norway 0RNULF LAURITZEN 707 The Australian tabulate coral genus Hattonia J. w. PICKETT and J. s. jell 715 Short communication : The Lower Cretaceous ammonite genera proposed by C. Jabob in 1907 M. K. HOWARTH 727 Printed in Great Britain at the University Press , Oxford by Vivian Ridler, Printer to the University Paiaeontoio VOLUME 17 PART 4 NOVEMBER 1974 Published by The Palaeontological Association London Price € 5 THE PALAEONTOLOGICAL ASSOCIATION The Association publishes Palaeontology and Special Papers in Palaeontology. Details of member- ship and subscription rates may be found inside the back cover. PALAEONTOLOGY The journal Palaeontology is devoted to the publication of papers on all aspects of palaeontology. Review articles are particularly welcome, and short papers can often be published rapidly. A high standard of illustration is a feature of the journal. Four parts are published each year and are sent free to all members of the Association. SPECIAL PAPERS IN PALAEONTOLOGY This is a series of substantial separate works. Members may subscribe to the Series; alternatively, Ordinary and Student members only may obtain individual copies at reduced rates. The following Special Papers are available : 1. (for 1967): Miospores in the Coal Seams of the Carboniferous of Great Britain, by a. h. v. smith and m. a. butterworth. 324 pp., 72 text-figs., 27 plates. Price £8 (U.S. $22.00), post free. 2. (for 1968): Evolution of the Shell Structure of Articulate Brachiopods, by a. williams. 55 pp., 27 text- figs., 24 plates. Price £5 (U.S. $13.00). 3. (for 1968): Upper Maestrichtian Radiolaria of California, by Helen p. foreman. 82 pp., 8 plates. Price £3 (U.S. £8.00). 4. (for 1969): Lower Turonian Ammonites from Israel, by r. freund and m. raab. 83 pp., 15 text-figs., 10 plates. Price £3 (U.S. $8.00). 5. (for 1969): Chitinozoa from the Ordovician Viola and Fernvale Limestones of the Arbuckle Moun- tains, Oklahoma, by w. a. m. jenkins. 44 pp., 10 text-figs., 9 plates. Price £2 (U.S. $5.00). 6. (for 1969): Ammonoidea from the Mata Series (Santonian-Maastrichtian) of New Zealand, by R. a. Henderson. 82 pp., 13 text-figs., 15 plates. Price £3 (U.S. $8.00). 7. (for 1970): Shell Structure of the Craniacea and other Calcareous Inarticulate Brachiopoda, by a. williams and a. d. wright. 51 pp., 17 text-figs., 15 plates. Price £1-50 (U.S. $4.00). 8. (for 1970): Cenomanian Ammonites from Southern England, by w. j. Kennedy. 272 pp., 5 tables, 64 plates. Pnce £8 (U.S. $22.00). 9. (for 1971): Fish from the Freshwater Lower Cretaceous of Victoria, Australia, with Comments on the Palaeo-environment, by M. waldman. 130 pp., 37 text-figs., 18 plates. Price £5 (U.S. $13.00). 10. (for 1971): Upper Cretaceous Ostracoda from the Carnarvon Basin, Western Australia, by r. h. bate. 148 pp., 43 text-figs., 27 plates. Price £5 (U.S. $13.00). 11. (for 1972): Stromatolites and the Biostratigraphy of the Australian Precambrian and Cambrian, by M. R. Walter. 268 pp.. 55 text-figs., 34 plates. Price £10 (U.S. $26.00). 12. (for 1973): Organisms and Continents through Time. A Symposium Volume of 23 papers edited by n. f. hughes. 340 pp., 132 text-figs. Price £10 (U.S. $26.00) (published with the Systematics Asso- ciation). 13. (for 1974): Graptolite studies in honour of O. M. B. Bulman. Edited by r. b. rickards, d. e. jackson, and c. p. hughes. 261 pp., 26 plates. Price £10 (U.S. $26.00). 14. (for 1974): Palaeogene foraminiferida and palaeoecology, Hampshire and Paris Basins and the English Channel, by J. w. Murray and c. a. wright. 171 pp., 45 text-figs., 20 plates. Price £8 (U.S. $22.00). SUBMISSION OF PAPERS Typescripts 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 The Secretary, P.A. Publications Committee, Department of Geology, Sedgwick Museum, Downing Street, Cambridge, CB2 3EQ, England, who will supply detailed instructions for authors on request (these are published in Palaeontology, 15, pp. 676-681). © The Palaeontological Association , 1974 Cover: Dactylioceras commune (J. Sowerby), Upper Lias, Jurassic, Whitby, Yorks. In collection of Professor J. E. Hemingway. The ‘head’ was carved by a Whitby jet worker to illustrate the traditional story of St. Hilda of Whitby (614-80), who turned snakes into stones by the power of prayer. THE CONODONT APPARATUS AS A FOOD-GATHERING MECHANISM by MAURITS LINDSTROM The Seventeenth Annual Address, delivered March 1974 Abstract. The possible functioning, rather than the zoological affinity, of the conodonts is discussed. Symmetry considerations, as well as arrangement of conodonts in apparatuses found on bedding planes, are best compatible with arrangement of the elements about the mouth. Homology between the elements indicates that certain con- clusions may apply generally. Many kinds of element could scarcely have functioned if turned with denticles toward the inside, hence the hypothetical animal is figured with denticles toward the outside. The apparatus was covered by soft tissue that is likely to have belonged to a lophophore. This functioned in the same way throughout the post- larval life of the animal (isometric growth of elements indicated by germ denticles). The denticulate aspect of the animal would have had a deterrent effect on predators, and denticles broken on encounters with predators could be regenerated. Long sharp teeth and subapical barbs on certain teeth served as passive defence. When attacked the animal might have contracted within the expandable, spinose lophophore part. Conodont remains ingested by conodontochordates suggest the upper size limit of the condodont animals. They probably fed on microscopic particulate, as well as on dissolved, nutrients. The conodont animals had a skeleton that consisted of several kinds of individual conodonts. As the animals died, the skeleton usually fell apart. Conodont faunas therefore consist of mixed and sometimes incomplete assemblages of different- looking elements from which the skeletons, or apparatuses, have to be reconstructed. Fortunately, more or less complete apparatuses have been found in certain lithologies (Schmidt 1934; Scott 1934; Rhodes 1952; Schmidt and Muller 1964; Lange 1968; Mashkova 1972). Using these as a template it has been possible to reconstruct group- ings of homologous apparatuses and follow their evolution in the Ordovician (Bergstrom and Sweet 1966; Schopf 1966; Webers 1966; Kohut 1969; Lindstrom 1971 ; Sweet and Bergstrom 1972), Silurian (Walliser 1964; Jeppsson 1969), Devonian (Klapper and Philip 1971), Carboniferous (von Bitter 1972), and Triassic (Sweet 1970). As one might expect, the establishment of skeletal evolution for whole apparatuses makes it easier to understand the evolution and suprageneric taxonomy of the conodonts (Lindstrom 1970). Unfortunately, knowledge of conodont apparatuses has not solved the enigma of the zoological nature of the conodonts. One cannot even say that we know what function the conodonts had in the animal, and how they fulfilled this function. I have assumed that they served as part of a feeding apparatus. In the past they have been taken to be gastropod radular teeth (Loomis 1936), polychaete jaws (Zittel and Rohon 1886; Du Bois 1943; Rhodes 1954), fish teeth (Pander 1856; Ulrich and Bassler 1926), lophophore support (Lindstrom 1964, 1973), and jaws of chaetognath type (Rietschel 1973), although other authors have considered the conodonts as possible vertebrate parts without emphasizing their function as teeth (Schmidt 1934; Gross 1954; Schmidt and Muller 1964). Inner structure, outer morphology, shape, symmetry, and homology must be the principal clues to the possible mode of life of the conodonts. The conodonts consist [Palaeontology, Vol. 17, Part 4, 1974, pp. 729-744.] A 730 PALAEONTOLOGY, VOLUME 17 of calcium carbonate fluorapatite (Pietzner et al. 1968), but the chemistry apparently does not help much to explain their functioning and nature, so it will not be further discussed. Palaeoecological aspects are considered briefly, but they have not yet been found to be decisive. STRUCTURE AND GROWTH Complete conodont elements consist of conodont and basal filling. The basal filling may be funnel- or plate-shaped ; it is inserted into a conical cavity at the widened base of the conodont. The conodont may consist of hyaline and white matter, but whereas the hyaline parts of well-preserved conodonts are amber coloured and translucent, the white matter is opaque to transmitted light. Some conodonts are entirely hyaline. Those with white matter may be referred to as albid (Lindstrom and Ziegler 1971). The white matter is concentrated in the denticles, particularly in the main denticle, or cusp, that is situated above the tip of the basal cavity. Conodonts have growth lamellae, added in outwards sequence like the annual growth layers of wood, but in the white matter the lamellae are destroyed. This destruction occurred during growth, for the outermost, and hence youngest, layer is nearly always hyaline. The white matter may be recrystallized (Lindstrom and Ziegler 1971; Barnes et al. 1973, 1973a) and contains small vesicles that may be radially arranged, or irregular, thin canals (Lindstrom 1964; Pietzner et al. 1968; Lindstrom and Ziegler 1971 ; Muller 1972). Muller and Nogami (1971) consider the interlamellar spaces as a special kind of white matter. However, this is a different thing from white matter as originally defined (Lindstrom 1964). Rays of white matter may form a V-shaped pattern that opens toward the apex and transects the conodont lamellae (Lindstrom 1964; Muller 1972). The other configurations of white matter described by Muller and Nogami (1971) and Muller (1972) can be regarded as effects due to the plane of sectioning. There is much evidence that the conodonts fractured easily across the white matter even during the life of the animal. Indeed, the white matter, with frequent cross-cutting planes formed by recrystallization, as well as increased porosity, appears to be designed to somewhat weaken the conodont structure. Because denticles contain a great deal of white matter, the denticles of early growth stages appear through a hyaline overgrowth. The occurrence of overgrown ‘germ denticles’ (Branson and Mehl 1933-1934) proves that conodont growth was centri- fugal. However, it also demonstrates the strong tendency toward isometric growth, that is, retention of shape during successive growth stages. If the denticles had not become overgrown, the outer shape of the elements would have had to change. That this did not happen is important. A further point is provided by a group of compound conodonts occurring in Carboniferous and younger rocks. In these the main denticle, or cusp, has concentrated white matter in its initial stages. Later it becomes more hyaline so that one can see the shape of the initial cusp, which is reclined at an angle to the growth axis of the mature cusp. The initial cusp thus could have the same angle of recurvature as the fully-grown one. When a conodont denticle broke off it was replaced by a new denticle growing from the stump. The latter, with a broken edge, may be seen in transmitted light, and this provided one of the first clues to the mode of growth of the conodonts (Hass M. LINDSTROM: CONODONT RECONSTRUCTIONS 731 1941). It is far from rare; in some faunas it is even the rule (Miller 1969). What happened to the part that was broken off is uncertain, in most cases it probably left the conodont organism. Muller (1972) has suggested that it was resorbed. However, despite the claims of Muller and Nogami (1971) and Muller (1972), no convincing proof of resorption has been presented (the apparent disappearance of lamellae on certain conodont platforms can equally well be interpreted as due to extremely low rates of secretion, with thin to absent lamellae as a result). Thus the regenerated denticles are a further indication that conodonts grew outward. They must have been surrounded by soft secreting tissue, were subepidermal (Hass 1941), or perhaps mesodermal. The conodont lamellae end at the base. Inside the basal cavity, and surrounding it on the lower face of the base, the edges of successive lamellae appear as concen- tric lines surrounding the oldest lamella, which is wrapped about the tip of the basal cavity. In most complete conodonts there is a basal filling that occupies the basal cavity. However, complete conodonts in this sense are relatively rare, since the basal filling very readily falls off. It is a relatively weak structure consisting of small, disorderly-arranged phosphate crystals (Pietzner et al. 1968; Lindstrom and Ziegler 1971 ). There is evidence that it consisted of pliable matter during the life of the animal and in early diagenesis (Lindstrom and Ziegler 1971). Its inner structure is con- centrically lamellar like that of the conodont. Thus the whole conodont was sur- rounded by soft tissue and there is no evidence that this condition changed during the life of the animal. Since special modes of growth were devised in order to maintain the outer shape essentially unaltered during growth, the elements evidently had the same function all the time, whatever this function was. Thus, conodonts cannot have had the biting, chewing, or rasping functions of polychaete jaws or radular teeth. Although the similarity between a conodont apparatus and the jaw apparatus of a polychaete may be striking, annelid jaws are epidermal structures formed to a certain size and functioning only as originally formed. Furthermore, they do not have any organ similar to the basal filling of conodonts. SHAPE OF CONODONTS The main groups of post-Cambrian conodonts can be derived from simple forms with a single, very long and very slender, more or less recurved denticle (that was apt to break). This is important to remember, because more evolved shapes might give the impression of possible functions which then turn out to be impossible to derive from any function that might be reasonably ascribed to the primitive conodonts. In the Distacodontacea the basic shape was retained. The recurvature of the cusp may be very strong, so strong in some species that the apex is directed at 180° opposite to the basal part. In many species there are longitudinal ridges that regularly face toward the concave side. Very early in the Ordovician certain simple forms evolved into the two principal kinds of conodont element found in younger faunas, the platform type and the rami- form. In both kinds it is the basal part that was the most strongly modified, expanding into branches, lobes, and ledges that carry denticulate patterns on their upper faces. 732 PALAEONTOLOGY, VOLUME 17 In platform conodonts occurring in the Lower Ordovician there are three branches or processes, each with a row of denticles that begins at one of the edges of the cusp. In these forms the cusp is still very high, sharp, and slender. In more evolved conodont elements of platform type the cusp is relatively low, and the denticles of the anterior process are fused to form a blade. The posterior process may either form a continua- tion of the blade or expand into a platform with ridge and denticle patterns. The text-fig. 1 . Hypothetical orientation of certain platform- type elements with respect to the mouth. Arrows — direc- tion of main food current; p.p.— posterior process; l.p.— lateral process; M— mouth. The elements are seen in ‘oral’ view. Top \ pair of prioniodiforms of Oepikodus evae (Lindstrom 1955). Below, platform elements of Palmatolepis. third, or lateral, process is reduced in many stocks but turns up again and again in the course of later evolution. This is the homeomorphy which is a recurrent feature of conodont evolution. The basic pattern appears in the Lower Ordovician and recurs until the Upper Triassic, and this indicates a strong correlation between shape and function. The ramiform elements can have one to four denticle rows beginning at the base M. LINDSTROM: CONODONT RECONSTRUCTIONS 733 of the cusp. In fully-evolved forms at least one denticle row continues as a long, slender process. The cusp is thin, high, and pointed, and several of the denticles may also be quite prominent. One process, most commonly the posterior one, is as a rule much longer than the cusp and other denticles. The denticles may rise thin and erect at 90° from this bar, or be reclined at various angles. In ramiforms with more than one prominent process, the arch formed by each pair of processes may be quite sharp, and the denticles of different processes point in different directions. The space between the processes may be occupied by the basal filling, when present. This applies to the earliest conodonts of platform type as well as for all ramiform ones. One reason why the denticles are not always in one plane is that there may be more than two denticulate processes. These diverge at different angles from the base of the cusp. Even where there are only two processes, the denticle row of the anterior process may be directed toward one side at about right-angles to the posterior process, or the latter may form a crescent with the denticles inclined toward the convex side. A Lower Carboniferous element, Hindeodella segaformis Bischoff, 1957, has parti- cularly long denticles interspersed between smaller ones at regular intervals along the bar. The long denticles are inclined to the left and the right in regular alternation. Not all denticles are situated on processes and platform lobes constructed about the basal part of the cusp. In a few Ordovician genera the apical part of the long, sharp cusp carries barb-like denticles, which are always inclined toward the base. In the preceding section it was argued that because of their structure the conodont elements must have functioned as subepidermal organs throughout their growth. This excludes the possibility that they could have dealt mechanically with food particles. The shape of the elements strengthens this point. There are few, if any, conodonts that were ideally shaped for seizing and holding food. Teeth used for these purposes ideally taper rapidly from a broad base. The slender, recurved conodonts are the opposite to this ideal shape. Furthermore, as we have seen, their structural design allowed them to snap with some ease. Rietschel (1973) pointed to the similarity between certain simple conodonts and the pinching jaws of chaetognaths, and also showed how such elements might have functioned mechanically. However, the lever mechanism suggested by Rietschel would be practical only with some elements and not with more strongly recurved, homologous elements belonging to the same conodont genera. As Rietschel himself points out, it would be ineffective in the case of ramiform elements that are direct homologues of the simple conodonts. The presence of a basal filling modifies the shape and possible mechanical functioning of a conodont element. This is ignored by Rietschel’s model. The very long and slender cusps were useful neither for seizing nor for holding food. If they had been used for such purposes the bases would have had to be very far apart, which would make the basal denticulation ineffective. Apical barbs, if penetrating into the food, would certainly have held it, but would also have prevented it from being passed on toward the pharynx. There is no way of orientating a great number of simple and ramiform elements so that they would efficiently seize, hold, or chew the food. The shape of certain platform elements suggests that these might have been used in chewing or grinding food particles. Jeppsson (1971) has drawn a hypothetical section of an opposed pair of Idiognathodus platforms that match one another so as to leave very little room between. However, if such matches do occur, they are very 734 PALAEONTOLOGY, VOLUME 17 rare, and most platforms would not function well in grinding. This is certainly true for all of those forms in which the denticulated portions would have been kept apart by the long cusps, had they been arranged in opposed positions in the animal. The persistency of conodont gross shape from the Lower Ordovician to the Upper Triassic suggests that shape and function were closely dependent on one another. The inner structure and growth show that conodont elements supported an organic tissue. It is plausible that the outer shapes of the conodont elements and the supported organ were interdependent (Lindstrom 1964, 1973). If this was the case, the structure supported by the conodont skeleton was frilled or tentaculate. The evolution of the conodonts indicates that it was advantageous that this structure had a great spread and surface. The functions that require great surface and spread are breathing, excretion, and the uptake of nutrients. It is only the gathering of particulate nutrients, which have to be forwarded to the pharynx, that poses specific requirements on the shape. MORPHOLOGY In addition to processes, denticles, and tubercles that can take the place of denticles, many conodont elements have thin, isolated ribs, or costae, along the sides of the main denticle. If the base is drawn out into processes, such costae continue along the pro- cesses as their denticulate edges. If the processes supported tentaculate frills, the costae are likely to have done so too. Many conodont elements are smooth, but some carry longitudinal striations on the denticles, and platform surfaces commonly have a network of thin ridges defining numerous polygonal pits. As the denticle bases are approached, the meshes are drawn out toward the denticle axes, and the ridges con- tinue up the denticles as longitudinal striations. This pattern has a distinct polarity in the direction at right-angles to the platform and parallel to the denticles. To ascribe any function to the pattern of striation and reticulae, one has to consider that the earliest forms in which it appears (Arenigian species of Prioniodus and Oepikodus ) had platform ledges that must have been far removed from one another owing to the size of the cusp and other denticles. I have suggested that the pattern served as attachment for muscles (Lindstrom 1973). These would have functioned as retractors of tentacles that were otherwise kept extended by turgor. It is improbable that these muscles moved the conodonts relative to one another since in that case the operating surface of the conodont elements would have been a closed system, one element facing another with only muscles between, and it would be very difficult to imagine any shape-related function. Symmetry. Most conodont elements are asymmetrical, as is apparent from the description of the platform elements. These may have three processes, one anterior, one posterior, and one lateral. In most cases there are right and left versions of such elements, and these are mirror images of one another. Thus it is generally assumed that most conodont animals were bilaterally symmetrical (Lindstrom 1964; Lane 1968). Certain elements, called trichonodelliform or hibbardelliform after the form genera Trichonodella and Hibbardella, are bilaterally symmetrical in themselves. They have two denticulate processes, one to each side, anteriorly and eventually a third process to the posterior. Such elements would have been situated in the M. LINDSTROM: CONODONT RECONSTRUCTIONS 735 sagittal plane of the animal (Jeppsson 1971 ; Lindstrom 1964, 1973). This particular kind of conodont element is associated with others with identical morphology except that they show various degrees of asymmetry and reduction of processes. Such associations have been called symmetry transitions (Lindstrom 1964), and they provide a key to the composition and arrangement of conodont apparatuses. Not all conodont apparatuses are bilaterally symmetrical. Asymmetry can appear in different manners in the platform component. In some cases only left- or right-sided elements are known (Lindstrom 1959; Lane 1968). For instance, only right-sided platforms of Cavusgnathus s.s. have so far been found. In others, the left element differs from its right counterpart by not being its mirror image ( Pseudopolygnathus primus Group of Voges 1959, Spathognathodus costatus Group of Ziegler 1962, Amorphognathus Bergstrom, 1971). Such cases can be identified with relative ease when one of the alternative shapes is always left-sided and the other right-sided. However, it is possible that the distinguishing features may not be constantly tied to right or left. Then one will find specimens with two different sets of shape charac- teristics, left and right elements occurring with either set of features. This is normally interpreted as two different species, each of which had a bilaterally symmetrical apparatus. If the constant association of the two kinds of elements were established on a statistical basis, this might be erroneously inferred to be due to sexual dimorphism. Associations of this kind might, in the author’s opinion, perhaps occur in American Carboniferous faunas (compare Merrill 1974). Owing to the wide, in some cases global, distribution of conodont species in many kinds of marine sediments, including pelagic ones, conodonts are usually recon- structed as pelagic swimmers and the bilateral symmetry agrees with this. The absence of bilateral symmetry in some species has led to the suggestion that certain conodonts were floating, perhaps colonial forms (Lindstrom 1964). However, as remarked by Lane (1968), this latter interpretation is not compelling if the conodont apparatus was internal, for internal organs can be asymmetrical without affecting swimming capability. Now that there are reasons to believe that the conodont apparatus was external rather than internal, the question still remains whether conodonts must have been active swimmers. Urbanek (1966) demonstrated that lophophores of colonial organisms can show a marked asymmetry. Apparatuses. In most conodont faunas of Silurian and younger age at least two kinds of complete apparatuses are represented. One of these consists of a pair of platform elements, a pair of platform-like (ozarkodiniform) elements, a pair of pick-like elements with straight and proclined, knife-like cusp (neoprioniodiform or synprionio- diniform), and a number of very long and almost straight, bar-like elements with many very small denticles regularly alternating with fewer somewhat larger ones (hindeodelliform elements). The hindeodelliforms mostly show a well-developed symmetry transition at the anterior end (that end at which the initial denticle or cusp is situated). There may be further ramiform elements, a possible instance is the Silurian to Devonian Ozarkodina. According to Jeppsson (1969) this genus has trichonodelliform and plectospathodiform elements in addition to the hindeo- delliform ones. However, a complete Ozarkodina apparatus found by Mashkova (1972) does not clearly show the latter kinds of element. 736 PALAEONTOLOGY, VOLUME 17 The other common type of apparatus consists chiefly of a symmetry transition of ramiform elements each of which carries several long denticles. This apparatus can be readily homologized with the first one. The place of the platform elements might be occupied by ramiform elements with similarities in plan to Ordovician prionio- diform elements (form genera Metalonchodina , Enantiognathus, etc.). In those cases text-fig. 2. Hypothetical functioning of muscles in conodont animals, seen in longitudinal section. The posterior direction is to the right. The outer sur- face of soft tissue is shown by broken line. Muscles are stippled. The basal filling is indicated by diagonal line pattern. Above : platform element with long retractile tentacles. Specimen’s muscle attachment is indicated by reticulate and pitted pattern on platform, strongly striate pattern on denticles. Muscles are likely to have been present on the blade (left) as well as on the platform; however, the former are likely to have been very short. Below, muscles attached to base of conodont serve to contract animal and erect the denticles in defence position. in which the apparatuses are found with the elements obviously more or less in situ , the hindeodelliforms are aligned parallel to one another with the denticles pointing apparently in random directions (this includes the occasional specimen in which they all point in one direction). The neoprioniodiforms, as pointed out by Jeppsson (1971) are mostly outside of the sheaf of hindeodelliforms. The paired platforms, mostly lying together, can occupy almost any position relative to the ramiform elements, M. LINDSTROM: CONODONT RECONSTRUCTIONS 737 and they may be turned toward or away from one another. They are as a rule anteriorly situated with respect to the similarly arranged ozarkodiniforms. (This circumstance was neglected in the reconstruction of Lindstrom 1973.) Arrangement of the Conodont Elements. Fortunately enough data are available to put severe limitations on any reconstruction of the conodont apparatus. The homology of the initial portion of the elements, that part carrying the cusp, requires that all elements are turned with the posterior face of the cusp in the same direction. The hindeodelliforms must be parallel and close together as one group (if Jeppsson’s reconstruction of Ozarkodina is correct, this genus must have had two batteries of hindeodelliforms as shown in the sketch by Lindstrom 1973). The bilaterally symmetri- cal ramiform must be at the mid-line. The platform elements must occupy neighbour- ing positions on each side of the mid-line. The same is true for the ozarkodiniforms, and is corroborated by the discovery of secondarily fused pairs of spathognatho- diforms as well as similarly fused ozarkodiniforms (Rexroad and Nicoll 1964; Pollock 1969; and interpretations in Lindstrom 1973). Reconstructions satisfying these data can be made according to two principles. Either the elements are arranged in groups after one another along the mid-line of the animal (Schmidt 1934; Rhodes 1952; Lindstrom 1964; Jeppsson 1971); or they encircled the mouth (or the pharynx if inside the mouth) (Lindstrom 1973), with the bilaterally symmetrical ramiform and the pairs of platforms and ozarkodiniforms at the sagittal plane but on opposite sides of the opening. In either case the platforms and ozarkodiniforms must be in a row, with the former in front of the latter ones. The second reconstruction was chosen because it allows the conodont elements to support a lophophore in close proximity to the mouth but situated outside of it. It also explains the disposition of the platforms and platform-like ozarkodiniforms at different orientations relative to the ramiform elements in apparatuses found on shale surfaces. According to this reconstruction the apices of the condont denticles are turned away from the mouth rather than toward its interior. There is probably no other direction in which they could effectively have been turned. A ramiform with a long sharp cusp could not have worked against soft tissue within the mouth. The long posterior bar would have prevented it from functioning as a mobile seizing organ projecting from the mouth. We have already seen that the conodont elements could not have worked against one another as seizing or masticating organs, and that they are more likely to have been the support for a spreading lophophore. As such, they would have been most efficiently deployed along a loop surrounding the mouth, with the denticulation pointing outward. There remains the difficult question of which side was fore and which was aft on the conodont elements? Did the denticles curve backwards or forwards? The recon- struction of Lindstrom (1973) shows them to curve backward on the assumption that a basket formed by the posterior processes of certain platforms might have strained the food most efficiently with this orientation. At least one further argument speaks in favour of this orientation. If the concave side of the initial denticle, or cusp, was back- ward, the anterior part, carrying the cusp and the processes, including the one com- monly referred to as posterior, would be close to the mouth. With the opposite orientation the growth of the long posterior process would push the cusp with its 738 PALAEONTOLOGY, VOLUME 17 plexus of processes and denticles a considerable distance away from the mouth. The orientation of the lateral processes would then have required that a sharp offset was present at the far end of the apparatus in order to leave the denticles free to carry tentacles. This offset must have moved away from the mouth during growth. Such an arrangement would be detrimental to efficiency because it separates the most strongly denticulated and spreading part of the apparatus from the mouth. In the reconstruction by Lindstrom (1973) the planes of the hindeodelliform elements are radially arranged round the mouth. This interpretation is improbable in the case of ramiforms, including many hindeodelliforms, with symmetry transition. The symmetry transition involves a rotation of two anterior lateral processes with respect to the cusp and posterior process, so that in the end member of the series one process becomes anterior and one postero-lateral. If homology is to be preserved, with the posterior processes parallel and all denticulate edges near the surface, the anterior ends of elements belonging to a symmetry transition have to be arranged along an arch, with the planes of the posterior processes essentially parallel. This agrees with the observation that the denticles of hindeodelliform bars found parallel in bundles on shale surfaces can be randomly turned to one side or the other, as if they had leaned over at random from an original orientation normal to the plane of sedimentation. CONODONTS AND THE CONODONTOCHORDATES In 1 969 William Melton discovered an animal with conodonts inside it in the Namurian Bear Gulch Limestone east of Lewistown, Montana. Four other specimens with identical shape and organization have since been discovered and described by Melton and Scott (1973) and Scott (1973), who referred the animals to the Conodonto- chordata, a new subphylum. The Bear Gulch Limestone is a fine-grained laminated limestone, apparently similar to the famous Upper Jurassic Solnhofen Limestone that has yielded numerous fossils with the flattened forms of soft parts preserved. The associated fauna is dominated by fish and shrimps. There are no burrowing forms except for a number of Lingula shells that might have been brought in from clastic nearshore sediments in the vicinity. Scott (1973) also reports two specimens of conodont apparatus that occur isolated from the conodontochordate animals. These apparatuses are of the same type as specimens previously found on shale surfaces. One of them appears to be double, since two pairs of platforms and more than one pair of neoprioniodiforms are visible. Scott interprets these specimens as fragments of the conodontochordate animal. As orientated by Melton and Scott the animal is 60-70 mm long and 13-15 mm high, with the ends rounded, and the specimens apparently compressed from the sides. The anatomical features of the fossils are constant. There is a sac-like body to which the above measurements apply, although at one end the margin is poorly defined (this applies particularly for the side regarded as dorsal), and the details in this part of the animal might be somewhat arbitrary. At mid-length the fossil has an elliptical body of darker material, one-third as long as the surrounding sac, which Melton and Scott call the ‘deltaenteron’. It is attached at one end, regarded as posterior to that margin of the surrounding sac that is regarded as dorsal. By this M. LINDSTROM: CONODONT RECONSTRUCTIONS 739 text-fig. 3. Hypothetical arrangement of conodont elements in prioniodontid (above) and polygnathid animal (below). M— mouth; ambal.— ambalo- diform; amorph.— amorphognathiform; hind. — hindeodelliform; neopr.— neoprioniodiform ; oi. — oistodiform ; oz.— ozarkodiniform ; platf. platform; ram. — ramiform type of element. Lophophore and branches stippled. Cono- dont elements stylized, with cusp shown by empty triangle, and processes shown by denticulate lines. orientation, there is ventrally to the deltaenteron a small, rounded, ferrugineous body called the Terrodiscus’. The so-called dorsal margin is simple, and the margin regarded as ventral is double and seems to consist of a narrow fold. Opposite to the attached end of the deltaenteron the margin is cut by a slit that was regarded as the anal pore. At the sharply defined end regarded as posterior, there is a grid-like structure in three of the specimens. Three new species were based on this material. Conodonts were found in the deltaenteron of four of the specimens, but for the specimen in which no conodonts were found it was stated (Scott 1973) that the mould of the deltaenteron had fallen out. The conodonts are mostly disordered and do not occur in any particular area within the deltaenteron, and they may be more scattered than in other natural assemblages, including those from the Bear Gulch. Their sizes differ greatly; the hindeodelliforms found in one specimen are about 0-5 mm long, those found in another specimen about T5 mm. Only one deltaenteron (that of Scottognathus elisabethae) contains the remains of a nearly complete apparatus; 740 PALAEONTOLOGY, VOLUME 17 however, in this case the platform set occurs twice. In the other specimens several important constituents of a conodont apparatus are missing. For these several reasons I regard the conodontochordates as conodont eaters, rather than conodont animals. Even if this is the case, the discovery made by Melton is very important because it associates the conodonts with another animal. If the conodonts were eaten by conodontochordates, this gives us an idea of the maximum size of the conodont animal. Even an animal that carried 1-5 mm long hindeodelliforms was small enough to pass into the deltaenteron that in its present state of preservation is about 20 mm long. This probably means that the conodont animal was, at the most, five or six times as long as the hindeodelliform bars. With this limitation on size, the animal is more likely to have been oblong or even barrel-shaped rather than long and worm- shaped. CONODONT ELEMENTS AS PASSIVE PROTECTION How could the conodont denticles have been broken and regenerated in the living animal? Since the environment of most conodonts appears to have been free of purely mechanical clashes against hard objects— high-energy shore deposits with conodonts are rare— it is likely that breakage was usually the consequence of contacts with predators, which implies that the conodont apparatus served as a defence mechanism. In particular the function for defence of the very long and sharp denticles with barbs near the apex (referred to in a previous section) appears obvious. According to a well- known synecological rule, the defence must have been displayed to be effective. This means once more that the outer shape suggested the presence of the denticulation. The denticles must have been directed outwards. If attacked, the animal might have been able to contract within the denticulated zone surrounding the mouth; this zone could be radially expanded because the conodont elements were not fused with one another. If the conodont animals were oblong rather than elongated in their outer shape and carried singly arranged or aligned denticles as their most obvious ornamentation, their aspect might, at least from some directions, have so much resembled certain palaeocope ostracodes that representatives of the two groups may have mimicked one another. The habitats of conodonts and palaeocope ostracodes are known to have overlapped to a great extent, and their stratigraphic ranges are very similar. Any criticism that this theory is extremely speculative is endorsed with conviction, how- ever the idea of a defensive mimicry between the Palaeocopa and the Conodonta might perhaps be worth recording for future scrutiny. PALAEOECOLOGY Conodonts occur in many different kinds of marine sedimentary rock, but not all conodonts occur in all lithologies. In particular several authors have remarked on differences in conodont faunas between shallow-water and pelagic facies. To explain such differences Seddon and Sweet (1971) compared the conodonts with the chaeto- gnaths, proposing similarities between the two groups, and suggesting that conodonts, like chaetognaths, were depth stratified. Accordingly the inhabitants of deeper water could not reach the shallow zones. Species living near the surface would occur near shore as well as in the uppermost part of the oceanic water column and would thus M. LINDSTROM: CONODONT RECONSTRUCTIONS 741 have a wider distribution than deep-water forms. This agrees with some but not all experience (Barnes et al. 1973a; Druce 1973). There is, however, general agreement that the conodonts would have belonged to widely distributed, plankton-feeding text-fig. 4. Hypothetical reconstruction of conodont animal of the Superfamily Polygnathacea, with emphasis on the aspect of the lophophore. Length (sa.) of lophophore about 3-6 mm. THE PLAN OF THE CONODONT LOPHOPHORE The recurrence of branching according to a limited set of plans, in conodont elements of all stratigraphic ages, suggests that the lophophore consisted of tentaculate frills that branched according to a certain pattern. The idea underlying this interpretation is that the conodont soft tissue was to a variable extent supported by conodont elements, and that the soft structure retained a similar plan even where it was not supported by conodont elements. Evolution in different conodont stocks included differing degrees of support of the lophophore by skeletal matter. Conodont elements preferentially grew from the junctions between branches of the lophophore, and some junctions might have lacked the corresponding conodont elements in certain species, genera, or families. In such cases the skeletal apparatus was incomplete, compared with taxa in which all loci of potential skeleton formation were occupied by conodont elements. Apparatuses with presumably only one pair, or a few pairs, of elements are known from all the major groups of conodonts (Sweet 1970; Sweet and Bergstrom 1972). Of all kinds of elements the platforms have the strongest tendency to branch and form accessory denticles. This is particularly evident in the Ordovician Amorpho- gnathus, the Silurian Kockelella, the Devonian Pedavis and Ancyrodella, and the Carboniferous Staurognathus. Accessory denticle rows, probably corresponding with tentaculate frills, are arranged parallel to the main denticle row, as in the rostrum of Siphonodella and several other polygnathids, including species of Palmatolepis\ or repeatedly at right angles to it, as on Polygnathus linguiformis and many gnathodids. The tentacles of these forms might, as suggested by the micromorphology, have been 742 PALAEONTOLOGY, VOLUME 17 provided with retractor muscles. It is suggested that these elements were close to the mouth and thus formed the terminus of the lophophore loop. The relative position of the platform elements and the platform-like ozarkodini- forms is not without problems. Silurian clusters (Rexroad and Nicoll 1964; Pollock 1969) contain fused ozarkodiniforms as well as fused spatognathodiforms, so that neither type of element could have flanked the other laterally. However, well- preserved assemblages may be found on shale surfaces with the posterior processes of spathognathodiforms or other elements of platform type overlapping on the anterior process of ozarkodiniforms, although admittedly this could be due to tele- scoping by contraction of the animal. However, in the Ordovician homologues of ozarkodiniforms and platforms, referred to respectively as ambalodiforms and amorphognathiforms, the ambalodiforms are so strongly arched that it is almost impossible to figure them as aligned behind the amorphognathiforms without much of the denticulation being deeply hidden in soft tissue. This problem could perhaps be solved by assuming that the amorphognathiforms were flanked by ambalodiforms (that are intermediate in shape between the amorphognathiforms and the ramiforms making up the rest of the apparatus) and that in the course of evolution the platforms migrated forwards so as to allow the evolved homologues of the ambalodiforms, i.e. the ozarkodiniforms, to align along the food groove at the mid-line. This is sug- gested by the illustrated reconstruction. It remains to be seen how the platform elements of each pair might have been arranged relative to one another. Possible arrangements of Oepikodus evae (text- fig. 1) illustrate this problem. If the arguments presented above are correct, the cusps could not have been opposed, nor were their points turned in opposite direc- tions, nor were the anterior parts of the elements facing one another. Oepikodus evae, like several other elements with the same function, has a terminal twist (Lindstrom 1973) of the posterior process. If food passed forwards between the elements, the con- cave side of the twist would have collected the food current, provided that the concave sides were turned toward one another. But the concave side faces in the direction of the lateral process that is customarily referred to as outer; hence in this case there would be a discrepancy between the (arbitrary) terminology and the actual orientation of the elements. The suggested orientation could mean that the concave side of the blade of palmatolepids faces toward the mouth, which appears to be a reasonable interpretation. If we assume that the platform elements supported that part of the lophophore which was the least suited for defensive display, their position might have been ventral. In a preceding section it was argued that the main ramiform elements were concentrated on the opposite side of the animal, so that by this orientation they were dorsal. To judge from the disposition of processes in the ramiform elements, the lophophore loop in this sector had several parallel branches in the posterior direction as well as a variable number of shorter branches in the direction of the mouth. FEEDING AND LIVING HABITS OF THE CONODONTS The conodont animal thus sketched was not necessarily very mobile. It might even have been a passive floater, relying on its battery of unpalatable denticles for its pro- M. LINDSTROM: CONODONT RECONSTRUCTIONS 743 tection. Some conodont animals might have formed colonies. Dispersion took place by ocean currents. This might be a reason why the conodont fauna can differ so strongly between pelagic geosynclinal regions and shelf environments, for example in the Ordovician of North America (Barnes, Rexroad and Miller 1973; Bergstrom 1973). The food gathered by conodonts might have been both microscopic particulate matter and dissolved material. If this is true we might expect the conodonts to occur most plentifully in areas where such nutrients are abundant, as for instance in environ- ments characterized by upwelling deeper ocean water. Such environments may occur on the margins of oceanic troughs or along submarine rises. The occurrence of conodonts in certain fossil sediments (black muds and trough-rise limestone facies) appears to agree with this prediction. REFERENCES barnes, c. R. and poplawski, m. l. s. 1973. Lower and Middle Ordovician conodonts from the Mystic Formation, Quebec, Canada. J. Paleont. 47, 760-790. — rexroad, c. b. and miller, ). f. 1973. Lower Palaeozoic Conodont provincialism. Geol. Soc. Am. Spec. Pap. 141, 157-190. sass, d. b. and monroe, e. a. 1973. Ultrastructure of some Ordovician Conodonts. Geol. Soc. Am. Spec. Pap. 141, 1 30. — and poplawski, m. l. s. 1973a. Conodont Ultrastructure: the Family Panderodontidae. Roy. Ontario Mus. Life Sci. Contr. 90, 1-36. bergstrom, s. m. 1971. Conodont biostratigraphy of the Middle and Upper Ordovician of Europe and Eastern North America. Geol. Soc. Am. Mem. 127, 83-162. — 1973. Ordovician conodonts. In hallam, a. (ed.). Atlas of Palaeobiogeography . Amsterdam (Elsevier), 47-58. and sweet, w. c. 1966. Conodonts from the Lexington Limestone (Middle Ordovician) of Kentucky and its lateral equivalents in Ohio and Indiana. Bull. Am. Paleont. 50 (229), 271-441. bischoff, G. 1957. Die Conodonten-Stratigraphie des rheno-herzynischen Unterkarbons mit Beriick- sichtigung der Wocklumeria-Stufe und der Devon-Karbon-Grenze. Abhandl. Hess. Landesamt. Bodenf. 19, 1-64. bitter, p. h. von. 1972. Environmental control of conodont Distribution in the Shawnee Group (Upper Pennsylvanian) of eastern Kansas. Univ. Kansas Paleontol. Contr. 59, 1-105. branson, E. b. and mehl, M. G. 1933-1934. Conodont Studies 1-4. Univ. Missouri Studies , 8, 1-349. druce, e. c. 1973. Upper Paleozoic and Triassic conodont Distribution and the recognition of biofacies. Geol. Soc. Am. Spec. Pap. 141, 191 -237. du bois, e. p. 1943. Evidence on the nature of conodonts. J. Paleont. 17, 155-159. GROSS, w. 1954. Zur Conodonten-Frage. Senckenberg. Leth. 35, 73-85. hass, w. h. 1941. Morphology of conodonts. J. Paleont. 15, 71-81. jeppsson, l. 1969. Notes on some Upper Silurian multielement conodonts. Geol. For. Stockholm Forh. 91, 12-27. — 1971. Element arrangement in conodont apparatuses of Hindeodella type and in similar forms. Lethaia, 4, 101-123. klapper, G. and philip, G. m. 1971. Devonian conodont apparatuses and their vicarious skeletal elements. Ibid. 429-452. kohut, j. j. 1969. Determination, statistical analysis, and interpretation of recurrent conodont groups in Middle and Upper Ordovician Strata of the Cincinnati region (Ohio, Kentucky, and Indiana). J. Paleont. 43, 392-412. lane, h. r. 1968. Symmetry in conodont element-pairs. Ibid. 42, 1258-1263. lange, f. G. 1968. Conodonten — Gruppenfunde aus Kalken des tieferen Oberdevon. Geolog. et Palaeontol. 2, 37-57. lindstrom, m. 1959. Conodonts from the Crug Limestone (Ordovician, Wales). Micropaleontology , 5, 427-452. 744 PALAEONTOLOGY, VOLUME 17 lindstrom, m. 1964. Conodonts. Amsterdam (Elsevier). 196 pp. 1971. Lower Ordovician conodonts of Europe. Geol. Soc. Am. Mem. 127, 21-61. 1973. On the affinities of conodonts. Geol. Soc. Am. Spec. pap. 141, 85-102. and ziegler, w. 1971. Feinstrukturelle Untersuchungen an Conodonten, 1 . Die Uberfamilie Pandero- dontacea. Geolog. et Palaeontol. 5, 9-33. loomis, f. b. 1936. Are conodonts gastropods? J. Paleont. 10, 663-664. mashkova, t. c. 1972. Ozarkodina steinhornensis (Ziegler) apparatus, its conodonts and biozone. Geolog. et Palaeontol. SB1, 81-90. melton, w. and scott, h. w. 1973. Conodont-bearing animals from the Bear Gulch Limestone, Montana. Geol. Soc. Am. Spec. pap. 141, 31-65. Merrill, G. 1974 (in press). Geolog. et Palaeontol. 8. miller, i. f. 1969. Conodont fauna of the Notch Peak Limestone (Cambro-Ordovician), House Range, Utah. J. Paleont. 43, 413-439. muller, k. j. 1972. Growth and function of conodonts. Internal. Geol. Congr. Rep. 24th Sess. Montreal, 7, 20-27. and nogami, Y. 1971. Uber den Feinbau der Conodonten. Mem. Sci. Fac. Kyoto Univ., Ser. Geol. Mineral. 38, 1-87. pander, c. h. 1856. Monographic der fossilen Fische des Silurischen Systems der russisch— baltischen Gouvernements. St. Peter sb. Konigl. Akad. Wiss. 1-91. pietzner, h., vahl, j., werner, h. and ziegler, w. 1968. Zur chemischen Zusammensetzung und Mikro- morphologie der Conodonten. Palaeontographica, Abt. A. 128, 115-152. pollock, c. A. 1969. Fused Silurian conodont clusters from Indiana. J. Paleont. 43, 929-935. rexroad, c. b. and nicoll, r. 1964. A Silurian conodont with tetanus? Ibid. 38, 771-773. Rhodes, f. h. t. 1952. A classification of Pennsylvanian conodont assemblages. Ibid. 26, 886-901. 1954. The zoological affinities of the conodonts. Biol. Rev. 29, 419-452. rietschel, s. 1973. Zur Deutung der Conodonten. Natur und Museum, 103, 409-440. schmidt, h. 1934. Conodonten— Funde in urspriinglichem Zusammenhang. Palaont. Z. 16, 76-85. — and muller, k. j. 1964. Weitere Funde von Conodonten— Gruppen aus dem oberen Karbon des Sauerlandes. Ibid. 38, 105-135. schopf, t. j. m. 1966. Conodonts of the Trenton Group (Ordovician) in New York, southern Ontario, and Quebec. N.Y. State Mus. Sci. Serv. Bull. 405, 1-105. scott, h. w. 1934. The Zoological relationship of the Conodonts. J. Paleont. 8, 448-455. 1973. New Conodontochordata from the Bear Gulch Limestone (Namurian, Montana). Publ. Mus. Mich. State Univ. Palaeont. Ser. 1 (2), 85-99. seddon, G. and sweet, w. c. 1971 . An ecologic model for conodonts. J. Paleont. 45, 869-880. sweet, w. c. 1970. Permian and Triassic conodonts from a section at Guryul Ravine, Vihi district, Kashmir. Univ. Kansas Paleont. Contr. 49, 1-10. — and bergstrom, s. m. 1972. Multielement taxonomy and Ordovician conodonts. Geolog. et Palaeontol. SB1, 29-42. ulrich, e. o. and bassler, r. s. 1926. A classification of the toothlike fossils, conodonts, with descriptions of American Devonian and Mississippian Species. U.S. Natnl Mus. Proc. 68 (12), 1-63. urbanek, a. 1966. On the morphology and evolution of the Cucullograptinae (Monograptidae, Grapto- lithina). Acta Geol. Polonica. 11, 291-544. voges, a. 1959. Conodonten aus dem Unterkarbon I und II ( Gattendorfia-und Pericyclus- Stufe) des Sauer- landes. Palaont. Z. 33, 266-314. walliser, o. H. 1964. Conodonten des Silurs. Abhandl. Hess. Landesamt . Bodenf. 41, 1-106. webers, G. f. 1966. The Middle and Upper Ordovician conodont faunas of Minnesota. Minnesota Geol. Surv. Spec. Publ. 4, 1-123. ziegler, w. 1962. Taxonomie und Phylogenie oberdevonischer Conodonten und ihre stratigraphische Bedeutung. Abhandl. Hess. Landesamt. Bodenf . 38, 1-166. zittel, K. a. and rohon, j. v. 1886. Uber Conodonten. Sitzber. Mathem-Phys. Classe Bayer. Akad. IViss. Munchen. 16, 108-136. m. lindstrom Geologie-Palaontologie Fachbereich Geowissenschaften der Philipps-Universitat D-3550 Marburg/Lahn, West Germany Typescript received 13 March 1974 REVIEW OF THE STRATIGRAPHY OF THE WENLOCK SERIES IN THE WELSH BORDERLAND AND SOUTH WALES by MICHAEL G. BASSETT Abstract. The stratigraphy and correlation of the Wenlock Series (Silurian) in its type area of the Welsh Borderland and also in South Wales are reviewed in the light of new faunal information. Two correlation charts are presented to relate the various lithostratigraphical sequences to one another and to the standard graptolite zones. The Wenlock Limestone is shown to be diachronous from the lundgreni to the ludensis Zone between Dudley and Ludlow. Palaeo- geographical maps are constructed for early and late Wenlock times. T he recent Special Report of the Geological Society of London on Silurian correlation in the British Isles (Cocks et al. 1971) outlined both the present state of our knowledge and the unresolved problems. The authors pointed out (1971, p. 104) that ‘since the Second World War, the Wenlock Series has received less attention than the Llandovery and Ludlow' and that ‘the time is not yet ripe for the formal erection of stages within the Wenlock Series’. These comments are especially relevant to the strata of the shelly facies, in which correlation is still based largely on the lithological divisions erected by Murchison (1833, 1834, 1835, 1839). As a result of the Special Report, the Stratigraphy Committee of the Geological Society set up a Working Group to investigate the problems of erecting stages within the Wenlock Series; this Group now consists of Professor C. H. Holland, Dr. L. R. M. Cocks, Dr. R. B. Rickards, Dr. P. T. Warren, and the present author. Immediate agreement was reached that the main problem lay in correlating the shelly sequence of the type area of Wenlock Edge in Shropshire with the succession of graptolite zones in the basin facies (Elies 1900), which have been revised recently by Rickards (1967, 1969), working in the north of England, and Warren (1971), working in North Wales. The Working Group thus decided to carry out extensive fieldwork along Wenlock Edge in an attempt to establish graptolite control prior to the erection of formal stages; this work is now completed and the results in press elsewhere. This review is intended to provide a background to the stratigraphy of the Wenlock Series throughout the Welsh Borderland and South Wales (text-fig. 1) in order to set the type Wenlock area in its regional context. The term Wenlock was first used by Murchison ( 1 833, 1 834, 1 835) for the shales and limestones in Shropshire between the top of his Caradoc Sandstone (i.e. top of the Llandovery Series of modern nomenclature) and the Ludlow rocks; he later (1839, p. 409) grouped the Wenlock rocks as a formation, and following modern strati- graphical practice the beds are now regarded as being of Series rank (see Evans in Whittard 1961, p. 253). In 1880 Lapworth (p. 48) introduced the term Salopian for strata of Wenlock and lower Ludlow age, and the name has subsequently been used by numerous authors, especially for rocks in graptolitic facies. Jones (in Evans and Stubblefield 1929, p. 89) suggested that if Salopian is to be retained it should embrace [Palaeontology, Vol. 17, Part 4, 1974, pp. 745-777.] B 746 PALAEONTOLOGY, VOLUME 17 text-fig. 1 . Outcrop and location map of the Wenlock rocks of Wales and the Welsh Borderland. The areas discussed in this paper are shaded in solid black. BASSETT: STRATIGRAPHY OF THE WENLOCK SERIES IN WALES 747 the whole of the Wenlock and Ludlow, but 1 agree with Cocks et al. (1971, p. 104) that the term has no place in modern stratigraphical classification and should not be used. In Shropshire the Wenlock Series comprises two broad lithological divisions, the Wenlock Shale and the Wenlock Limestone, while in parts of the southern half of the Welsh Borderland a third unit, the Woolhope Limestone, is present at the base of the Series as a lateral equivalent of part of the Wenlock Shale of Shropshire. The Woolhope Limestone was initially included by Murchison (1839, p. 429) in the top of his ‘Caradoc Sandstone’ but, following De la Beche (1846, pp. 38-39) and Phillips (1848, pp. 75, 167), was later placed by him (1854, p. 106) at the base of the Wenlock, an interpretation which is now accepted in the light of subsequent correlation. In South Wales the limestone and shale facies of the Welsh Borderland are replaced by clastic sequences consisting mainly of calcareous and arenaceous mudstones and silt- stones. Variations in the thickness of the Wenlock Series and the correlation of its constituent lithostratigraphical units are illustrated in text-figs. 2 and 4. Following the outline of the graptolite succession given immediately below, the nature and age of the contacts of the Wenlock with Llandovery and older rocks, and with Ludlow and younger rocks, are then discussed in order to define the limits of the Series throughout the shelf area, and to stress the pulsatory pattern of Lower Palaeozoic earth movements which partially controlled Wenlock sedimentation and palaeogeography (see George 1963; Ziegler 1970). The stratigraphy of individual areas shown in text-fig. 1 is then described, followed by a synthesis of Wenlock palaeogeography. THE WENLOCK GRAPTOLITE SUCCESSION Although graptolites have been recorded comparatively rarely from the Wenlock shelf facies, some valuable specimens have been mentioned by authors from different areas. Since some of the following discussion on stratigraphy is based on graptolite evidence, the zonal scheme recognized in the basin facies is given below (see also text- figs. 2 and 3). This is the succession established at Builth (Elies 1900; Elies and Wood 1918), with modifications in nomenclature based on Rickards ( 1 965, p. 248, text-fig. 1 ; 1969), Warren et al. (1966, p. 466), Holland et al. (1969, p. 676, text-fig. 4), Warren (1971), and Cocks et al. (1971, text-fig. 2). Monograptus ludensis Cyrtograptus lundgreni Cyrtograptus ellesae Cyrtograptus linnarssoni Cyrtograptus rigidus Monograptus riccartonensis Cyrtograptus murchisoni Cyrtograptus centrifugus It should be emphasized that this range of zones is Wenlock by correlation as opposed to being Wenlock by definition, since the limits of the type Wenlock Edge succession are not yet known with certainty in graptolite terms. Subdivisions of these zones, and regional variations in nomenclature in the basin successions in different (= vulgaris auct.) ( = rigidus auct.) (= symmetricus auct.) 748 PALAEONTOLOGY, VOLUME 17 parts of Britain, are given by Rickards (1965, text-fig. 1 ; 1967, text-fig. 8; 1969, text- fig. 2) and Warren (1971, text-figs. 1 and 2). WENLOCK/PRE-WENLOCK CONTACTS In the southern half of the Welsh Borderland the junction between the uppermost Llandovery and lowest Wenlock rocks is exposed only in the Tortworth, May Hill, and Woolhope Inliers and to the west of the Malvern Hills, and in all cases the relation- ship is one of conformity. Thus at Woolhope, May Hill, and in the Malverns the Woolhope Limestone succeeds deposits of latest Upper Llandovery (C6) age (Upper Haugh Wood Beds, Yartleton Beds, and Wych Beds respectively) with no apparent break, and in the Tortworth Inlier an unnamed limestone unit at the base of the Wenlock Brinkmarsh Beds similarly rests on uppermost Tortworth Beds of C6 age. In the western English Midlands, around Walsall and Dudley, the Llandovery/ Wenlock contact is not exposed at the surface, but a borehole at Walsall (Butler 1937, p. 249) shows a conformable transition between the two Series. In parts of the northern half of the Welsh Borderland and in south Central Wales there are stratigraphical breaks of varying magnitude between Llandovery and Wenlock strata, although previous descriptions of breaks within some parts of this area require further comment. At the north-east end of Wenlock Edge the base of the Wenlock is taken at the base of the Buildwas Beds (Wenlock Shale) at their contact with the Purple Shales. This horizon has generally been accepted as the base of the Series since its definition by Salter and Aveline (1854, p. 63), following the original description of the area by Murchison; however, the correlation of this horizon with the graptolite sequence has been the subject of some discussion. Whittard ( 1 952, p. 1 69) suggested that graptolites recorded by earlier authors, notably Whittard ( 1 928), Das Gupta (1932), and Pocock et al. ( 1 938), from low Wenlock Shale horizons, ‘do not warrant correlation with any zone older than linnarssonf , implying that the centrifugus , murchisoni, riccartonensis , and rigidus zones are absent. In addition, Whittard found no evidence from his own earlier work (1928, 1932) for the Upper Llandovery griestoniensis and crenulata zones, and he suggested that there is a con- siderable stratigraphical break below the Buildwas Beds. More recently Cocks and Rickards (1969, pp. 224-227) reassessed all previous records of graptolites from the uppermost Llandovery and lowest Wenlock of Shropshire and concluded (pp. 228- 229) that within the type area there is no large sedimentary break at the Purple Shales/Buildwas Beds junction which ‘approximately coincides with . . . the base of the centrifugus Zone’. Two of the zones (, griestoniensis and rigidus) thought by Whittard to be missing were shown by Cocks and Rickards to be present, thus reducing the previously postulated break, and although the crenulata , centrifugus , murchisoni , and riccartonensis zones remained unproven by graptolites, some 60 m of rock without diagnostic graptolite species could accommodate some or all of them. From this evidence the correlation of the base of the Buildwas Beds with that of the centrifugus Zone is a reasonable approximation; the current work of the Geological Society Working Group is aimed at producing positive evidence to resolve the problem. At the south-west extremity of Wenlock Edge the Buildwas Beds are overlapped by higher horizons of the Wenlock Shale (Coalbrookdale Beds) which eventually BASSETT: STRATIGRAPHY OF THE WENLOCK SERIES IN WALES 749 overstep on to the Purple Shales. West of the Church Stretton fault the extent of the overstep is known in some detail from the faunal evidence in boreholes (Cocks and Rickards 1969) where the centrifugus, murchisoni , and all or part of the riccartonensis zones are cut out; to the south of Church Stretton the Wenlock Shale finally over- steps on to the Ordovician (Dean 1964). At Bishop's Castle, some 15 miles to the south-west of Wenlock Edge (see text-fig. 1) the same infra-Wenlock unconformity is detected, with the local basal Wenlock beds, belonging to the lundgreni Zone (Allender in Allender et al. 1960, p. 223), resting directly on the Upper Llandovery. At Nash Scar, near Presteigne in Radnorshire, the base of the Wenlock succession is repre- sented by the Nash Scar Limestone, which rests on the Polly Sandstone of early Upper Llandovery (Cw) age (Ziegler et al. 1968, p. 750); there is thus a possibility that part of the Nash Scar Limestone is as old as C3, but the total fauna collected by the present author (p. 759) is more suggestive of a Wenlock age, and this, together with the evidence of a disconformable relationship with the Polly Sandstone on the north side of the Presteigne Inlier (Ziegler et al. 1968), strongly points to an uncon- formable Polly Sandstone/Nash Scar Limestone contact throughout the area. At Dolyhir, near Old Radnor, some 7 km south-west of Nash Scar, the local basal Wenlock correlative of the Nash Scar Limestone is the Dolyhir Limestone, which rests with gross unconformity on Pre-Cambrian rocks. In the Long Mountain syncline of Montgomeryshire and north Shropshire Das Gupta (1932, pp. 326-327) suggested that there was also an unconformity at the base of the Wenlock throughout the area, with riccartonensis Zone or younger strata resting directly on the Upper Llandovery Buttington [Purple] Shales (see also Wade 1911, p. 436; Elies 1900, pp. 386-397). However, a preliminary revision of the stratigraphy by Palmer (1970), with revised nomenclature, indicates conformity between the Llandovery (Buttington Mudstone Pormation) and the Wenlock (Trewern Brook Mudstone Pormation), and this is supported by faunal evidence from Buttington brickworks in the north of the syncline (SJ 3266 3100) where a basal Wenlock graptolite sequence through the centrifugus , murchisoni , and riccartonensis zones is present (Ziegler et al. 1968, p. 766; Cocks and Rickards 1969, pp. 226-227 ; Cocks et al. 1971, p. 109). Throughout the Builth district of Radnorshire, where Elies (1900) first demon- strated the succession of Wenlock graptolite zones, basal Wenlock Beds are largely obscured by overlap by the higher zones (Jones 1947). In the sections north and east of Builth this overlap is most marked and increases in intensity to overstep on to Llandovery and then Ordovician rocks of the Carneddau range. The Llandovery Trecoed Beds (Ziegler et al. 1968, p. 769; Cocks et al. 1971, p. 108, text-fig. 2 column D) of this area crop out in a thin, discontinuous strip from Park Wells to near Maesgwynne (Jones 1947, pi. 1) and contain a mid Upper Llandovery (C3 4) fauna. Highest Upper Llandovery Beds are not known below the Wenlock but the detailed extent of the sub-Wenlock unconformity is uncertain since the graptolite faunas of the lowest Wenlock Beds need revision. The lowest ( murchisoni ) Zone as recognized by Elies (1900) is now generally subdivided into a centrifugus Zone below and a murchisoni Zone above (e.g. see Rickards 1967, 1969), and the centrifugus Zone has not yet been proved at Builth. Elles’s faunal lists (1900, table 1, p. 378) are of little help, but her record of Retiolites geinitzianus Barrande from Pencerig may indicate that the centrifugus Zone is represented at that locality, since the species is rare or 750 PALAEONTOLOGY, VOLUME 17 absent in many areas at higher horizons; it has, however, been recorded from the murchisoni Zone of North Wales (Warren 1971) and Poland (Teller 1969). West of Builth, and south-westwards towards the Garth area (Breconshire), basal Wenlock Beds are again cut out both by overlap and overstep on to Llandovery horizons, but around Garth itself the two Series reappear for a short distance in conformable sequence (Andrew 1925, p. 399; Andrew and Jones 1925, p. 412); as at Builth there is no modern record of the centrifugus Zone, but since mudstones of the highest Llandovery crenulata Zone pass transitionally into sediments containing C. murchisoni, the centrifugus Zone is probably represented. Within a mile to the south-west of Garth overlap within the Wenlock again takes place, followed by very rapid overstep across the Llandovery strata and on to the Ordovician, a relationship which is continued for some 16 km in the same direction towards the town of Llandovery (Carmarthenshire). Around Llandovery itself, and south-westwards to Maes-y-fallen, some 3-2 km south-east of Llandeilo, Llandovery rocks reappear between the Wenlock and the Ordovician. In the northern part of the Llandovery district, the uppermost Llandovery (Pale Mudstone Group) is either faulted against or followed conformably by basal Wenlock rocks (Jones 1949, p. 58, pi. 3), but in the southern district a south-westerly overstepping relationship between the two Series takes place once more (Jones 1925, p. 376, pi. 21). The increasing intensity of this overstep south-westwards to Maes-y- fallen has been well documented by Williams (1953, pp. 198-200, pi. 9) and beyond this point Wenlock beds again transgress on to the Ordovician. Some modification in the geographical extent of the overstep described by Williams is suggested by a re-examination of the section in the Sawdde gorge south of Llangadog; here Williams recorded (1953, p. 199) Llandovery strata of age overlain by Wenlock beds provisionally assigned to the riccartonensis Zone, but recent collecting close to the boundary, both by the author and independently by Mr. N. J. Hancock, has revealed brachiopod faunas of probable C6 age, with an apparently conformable transition upwards into the Wenlock, although firm graptolite control is still lacking (see p. 766); however, Williams’s reconstruction south-westwards from the Sawdde is confirmed by his faunal evidence and rapid thinning of both the Llandovery and lower Wenlock sediments. Between Maes-y-fallen and Llanarthney, beds of late Wenlock age rest directly on Llandeilo and Llanvirn rocks, and to the west of Llanarthney the Wenlock is itself finally overstepped by the Old Red Sandstone (Strahan et a/. 1907). Westwards through the remainder of Carmarthenshire and across central and north Pembrokeshire there are no outcrops of Wenlock rocks, but along the south coast of Pembrokeshire deposits of this age are present in tectonically isolated sections (text-figs. 1 and 6). The base of the Series is exposed only at Marloes Sands and at Renney Slip, although its exact level has not yet been determined ; in both sections beds low in the Coralliferous ‘Series’ contain distinctive highest Llandovery (C6) faunas which grade upwards into the Wenlock (Cantrill et al 1916; Ziegler et al. 1969, pp. 429-435; Bassett 1971, pp. 207-216). South of Milford Haven, at Freshwater East and Freshwater West, beds of probable Wenlock age (see p. 769) rest directly on Llanvirn graptolitic shales. BASSETT: STRATIGRAPHY OF THE WENLOCK SERIES IN WALES 751 WENLOCK/POST-WENLOCK CONTACTS In all but six of the areas included in this account conformable contacts are exposed between highest Wenlock beds and the overlying Ludlow Series. The exceptions are: (1) the Eastern Mendips Inlier, where the Wenlock is followed unconformably by Old Red Sandstone sediments (Reynolds 1907); (2) the Tortworth Inlier where the uppermost Brinkmarsh Beds, of late Wenlock age, are succeeded with gross uncon- formity by Upper Old Red Sandstone and Mesozoic strata (Curtis and Cave 1964, p. 438; Curtis 1972, p. 3, fig. 3); (3) Gorsley where beds of Lower Leintwardinian (mid-Ludlow) age rest on an eroded surface of the local equivalent of the Wenlock Limestone (Lawson 1954, p. 231, text-fig. 2); (4) around Walsall where late Carboni- ferous strata transgress across the Wenlock (e.g. Cantrill 1919; Whitehead and Eastwood 1927); (5) at Kendal End where the Wenlock is now seen faulted against Pre-Cambrian and late Carboniferous; and (6) at Winsle in Pembrokeshire where some marine Silurian beds appear to be cut out beneath overstepping Old Red Sandstone (Cantrill et al. 1916, p. 86, see also p. 769). Throughout the Welsh Border- land the Wenlock/Ludlow boundary is drawn at the base of the Lower Elton Beds, at their contact with the Wenlock Limestone. This is the horizon established originally by Murchison (1839, p. 209) and defined by Holland et al. (1963, pp. 139-141, text- fig. 1 1) at a standard section at Pitch Coppice, Ludlow. The base of the Ludlow Series in the graptolite sequence has, until recently, been generally accepted as at the base of Wood’s (1900, p. 422) zone of Monograptus vulgaris (= ludensis : see Warren et al. 1966). In the Ludlow anticline, however, there is now evidence for the ludensis Zone both below and within the Wenlock Limestone (Warren et al. 1966, p. 466; Holland et al. 1969) and most or all of this Zone belongs in the Wenlock; this is supported by the probable presence of the same Zone below the Wenlock Limestone of Wenlock Edge (Cantrill 1927, p. 43; Pocock et al. 1938, pp. 101, 1 13). From the reassessment of the level of the ludensis Zone, Warren et al. and Holland et al. indicated that the base of the Ludlow in the shelf facies (i.e. the base of the Lower Elton Beds) correlates most closely with the base of the nilssoni Zone, the latter authors pointing out at the same time ( 1 969, p. 26 1 ) that the correlation is still not precise, since diagnostic graptolites are absent in the Lower Elton Beds and highest Wenlock Limestone. Bassett and Shergold (1967, p. 395) and Shergold and Bassett (1970, p. 135) also commented on the degree of uncertainty in the cor- relation and suggested that along Wenlock Edge the ludensis Zone extends into the Lower Elton Beds, based on Das Gupta’s records (1932, pp. 351-352; 1933, p. 113 and map, p. 1 1 1 ) of ludensis ( sic vulgaris) from that horizon, together with the identifica- tion of lower nilssoni Zone faunas by Shergold and Shirley (1968, text-fig. 1) from the basal Middle Elton Beds. Holland et al. (1969, pp. 673-674) also recorded M. ludensis from within the lower nilssoni Zone and so this species in itself does not prove the ludensis Zone. Das Gupta’s evidence is thus not necessarily at variance with that of Holland et al., though final clarification must await further graptolites from the Lower Elton Beds at Ludlow and Wenlock Edge. The stratigraphical consequences of slightly different interpretations of the position of the base of the Lower Elton Beds in the graptolite sequence have been outlined by Holland et al. (1969, p. 681) and discussed by Lawson (1971, pp. 304-306, fig. 1). 752 PALAEONTOLOGY, VOLUME 17 One consideration, not mentioned by these authors, is that the base of the Lower Elton Beds may be diachronous, being at or close to the ludensis/nilssoni boundary at Ludlow but becoming progressively older in a general easterly direction. This sug- gestion is supported by the evidence of graptolites from the Wenlock Limestone of Wren’s Nest, Dudley; collections in Birmingham University Museum contain a number of specimens of Monograptus flemingii (Salter) from Dudley, and although not all of them are accurately localized within the Wenlock Limestone, the lithology of the matrix is undoubtedly from that horizon, and one specimen (BU511 and counterpart 51 1 A) is known to be from 24 m above the base of the Limestone on the east side of Wren’s Nest hill. M. flemingii is not known above the lundgreni Zone, providing a firm upper age limit for at least the lower part, and probably all, of the Wenlock Limestone at Dudley (see p. 757). Since the lower part of the Wenlock Limestone at Ludlow is known to be in the ludensis Zone (see above), the base of the formation must be diachronous between the two areas, and thus possibly also from Ludlow north-eastwards along Wenlock Edge. A corollary to this is that the base of the overlying lithological unit (Lower Elton Beds at Ludlow/Lower Ludlow Shales at Dudley) is probably also diachronous. This picture is in keeping with the general palaeogeographical reconstruction of continuity of deposition across the shelf area during late Wenlock/early Ludlow times, when limestone deposition would have commenced earlier in the shallower water to the east than along the shelf/basin margins to the west, followed by shallowing and seaward spreading of limestone deposition with time (Scoffin 1971, p. 212 and fig. 27). In the north and west of the Welsh Borderland and through central and south Wales the Wenlock Limestone is not developed, but, apart from at Rumney (Cardiff) and in Pembrokeshire, the Wenlock/Ludlow boundary can be correlated satisfactorily by means of graptolites. Around Builth and in the Long Mountain, the argillaceous sequence of this level is well documented (Jones 1947; Elies 1900; Wood 1900; Das Gupta 1932; and Palmer 1970). South-westwards from Builth towards Llandovery and Llandeilo the Wenlock/Ludlow succession passes laterally into a conformable sequence of calcareous mudstones and sandstones with a mixed shelly and graptolite fauna. The sequence was formerly mapped as a single unit (Murchison 1839; Strahan et al. 1907) but recently Potter and Price (1965) subdivided the Ludlow through much of the area and correlated the base of the Series with the base of the Tresglen Beds, mainly on shelly faunal criteria; this correlation is supported by Price’s earlier (in Lawson et al. 1956, p. 568) correlation of the lower Tresglen Beds with the nilssoni shales of Builth, that is immediately above the ludensis Zone. At Rumney (Cardiff) there is apparent conformity between beds of Wenlock and Ludlow age, through a very poorly fossiliferous succession of coarse sandstones, sandy mudstones, and sandy limestones (Sollas 1879, p. 488, fig. 4), but although Sollas correlated part of the sequence with the Wenlock Limestone the sections are now partly obscured and the base of the Wenlock cannot now be fixed accurately; the writer is remapping the inlier and also excavating temporary sections. There are also difficulties in recognizing Wenlock/Ludlow contacts in south Pembrokeshire; north of Milford Haven, where the Coralliferous ’Series’ passes conformably upwards into the Sandstone ‘Series’ at Pittingales Point on the south-east side of Deadman’s Bay, along Marloes Sands, and in the Lindsway Bay sections near St. Ishmaels, BASSETT: STRATIGRAPHY OF THE WENLOCK SERIES IN WALES 753 a number of authors have suggested that the limited faunal evidence indicates a Wenlock/Ludlow transition low in the Sandstone Series, but this evidence is still far from conclusive (Cantrill et al. 1916, p. 55; Bassett 1971, p. 207; Sanzen-Baker 1972, p. 144); south of Milford Haven Dixon (1921, pp. 12-13) described an uncon- formable Wenlock/Ludlow contact at Freshwater East, but other accounts suggest conformable relationships throughout the sections, though it is not yet clear if the beds are of Wenlock or Ludlow age (Bassett 1971, p. 220; Walmsley in Owen et al. 1 97 1 , pp. 46-47 ; Sanzen-Baker 1972, p. 1 47). The problems of the age of the Pembroke- shire sections are discussed further below (p. 767). LOCAL STRATIGRAPHY WENLOCK EDGE Text-fig. 2, col. 1 Murchison (1833, p. 475) first divided the Wenlock sequence of Wenlock Edge into an uppermost Wenlock Limestone division, underlain by ‘Lower Ludlow Rock or Die Earth’, but later (1834, p. 14 and table opposite p. 13) referred the latter to the Wenlock Shale and consistently employed this terminology in subsequent accounts (1835, 1839, 1854-1872). Salter and Aveline (1854, pp. 63, 71) clarified the separation of the Wenlock Shale from the remainder of Murchison’s ‘Caradoc Sandstone’ and confined the Shale to the unit overlying the Purple Shales. Davidson and Maw (1881, pp. 102-104; in Davidson 1882, pp. 67-70) further subdivided the Wenlock Shale into Basement Beds, Buildwas Beds, Coalbrookdale Beds, and Tickwood Beds, and also included shales above the Wenlock Limestone within the Wenlock ; this classifica- tion was followed in full by Lapworth and Watts (1894, p. 325) and in part by Watts (1925, p. 346), but Whittard (1928, p. 752) subsequently pointed out that the Basement Beds belong to the Llandovery Purple Shales, while most authors have followed Murchison’s original definition (1839) in including the shales above the Wenlock Limestone within the Ludlow. The lithological divisions now recognized in the type area are : Wenlock Limestone 21-28-5 m f Tickwood Beds 24-27 m Wenlock Shale < Coalbrookdale Beds 190-250 m (. Buildwas Beds 25-5-27 m In the south-west half of Wenlock Edge the Wenlock Shale has been described recently by Greig et al. (1968) as a single lithological unit. The Buildwas Beds are well exposed only in the north-east of the area, being covered largely by drift south-westwards from Ticklerton and overlapped by the Coalbrookdale Beds close to the Onny River; they have been described by Davidson and Maw (1881) and Pocock et al. (1938). As discussed earlier (p. 748) Whittard considered that the Buildwas Beds belong either to the rigidus or linnarssoni Zone, but Cocks and Rickards have suggested that their base may approximate to the base of the centrifugus Zone. Aldridge (1972, p. 141) has recently recorded an amorphognathoides Zone conodont fauna from the Buildwas Beds near Ticklerton (SO 481 901); correlatives of this zone elsewhere are known to span the Llandovery/ Wenlock boundary (in graptolite terms), lending support to a centrifugus Zone age for at least part of the Buildwas Beds. Graptolites recorded by Pocock et al. ( 1 938) from the lower part of the overlying Coalbrook- dale Beds suggest a correlation with the linnarssoni Zone, and the higher faunas from this unit suggest the ellesae, lundgreni, and possibly part of the ludensis zones. The record of abundant specimens of Gotho- graptus nassa in the Tickwood Beds indicates that they fall within either the upper part of the lundgreni Zone or the ludensis Zone (Pocock et al. 1938, p. 101; Das Gupta 1935, p. 110), and the overlying Wenlock Limestone is probably wholly within the latter zone. The nature and variation in lithology of the Wenlock Limestone have been described by Crosfield and Johnson (1914), Hill et al. (1936), Pocock et al. (1938), Greig et al. (1968), Shergold and Bassett (1970), and Scoffin (1971). 754 PALAEONTOLOGY, VOLUME 17 1 2 3 4 5 WENLOCK LUDLOW WALSALL & KENDAL LONG EDGE DUDLEY END MOUNTAIN Carboniferous basal Ludlow basal Ludlow Measures top faulted out basal Ludlow text-fig. 2. Correlation chart of Wenlock sequences in south-central Wales and the Welsh Borderland north of the Bristol Channel. BASSETT: STRATIGRAPHY OF THE WENLOCK SERIES IN WALES 755 6 7 8 9 10 11 12 DOLYHIR & ABBERLEY MALVERN WOOLHOPE GORSLEY MAY HILL USK NASH SCAR HILLS HILLS basal Ludlow shales and siltstones faulted contact faulted contact Dolyhir and Nash Scar Limestone 24-60m taM Pre-Cambrian and late Llandovery ( C, -2 ) basal Ludlow Wenlock Limestone 36m base not seen basal Ludlow Wenlock Limestone 60- 150m 180 -270m Wool hope Limestone 15-60 m uppermost Llandovery Wenlock Limestone 45-51m Woolhope Limestone 36m uppermost Llandovery middle Ludlow WH] Gorsley Limestone 5 5m base not seen basal Ludlow Wenlock Limestone 30 -105m Woolhope Limestone 21-60 m uppermost Llandovery basal Ludlow Wenlock Limestone < 1 - 1 3-5m sandstone beds 4-5m base not seen 756 PALAEONTOLOGY, VOLUME 17 THE LUDLOW ANTICLINE Text-fig. 2, col. 2 The base of the Wenlock is not exposed in the Ludlow district, where the core of the anticline is occupied by at least 300 m of Wenlock Shale (Holland et al. 1963). Although this thickness is approximately equal to the full Wenlock sequence at Wenlock Edge, the beds at Ludlow fall wholly within the upper Wenlock, since graptolites from the lower half of the succession suggest a lundgreni Zone age (Holland et al. 1969, p. 676). At the top of this zone a thin horizon with a restricted graptolite fauna of Gothograptus nassa and Pristiograptus dubius correlates with the ‘ nassa/dubius Interregnum’ of Jaeger (1959). Immediately succeed- ing beds, some 100 m below the base of the Wenlock Limestone, contain Monograptus ludensis, marking the base of that zone, which is now known to extend through the upper half of the Wenlock Shale and at least the lower part of the Wenlock Limestone (Holland et al. 1969, pp. 676-678). The Wenlock Shale at Ludlow has not been subdivided lithologically as at Wenlock Edge, and Lawson (1971, p. 306 and fig. 1, col. E) has pointed out that the lower, graptolite-bearing part of the Wenlock Limestone as mapped by Holland et al. (1963) is lithologically similar to, and may correlate with, the Tickwood Beds. Thus in graptolite terms the age of the highest Wenlock Limestone at Ludlow remains equivocal, but it probably falls within the upper part of the ludensis Zone, since the base of the overlying Elton Beds is at or close to the ludensis /nilssoni boundary (Holland et al. 1969, text -fig. 4, pp. 679-681). The Wenlock Lime- stone thins from west to east across the Ludlow anticline from 135 m to 60 m; the lower, graptolitic beds consist of alternations of hard ribs of flaggy, silty limestone and grey silty mudstones and shales, while the highest 1 5 m are hard grey to buff nodular limestones. The base of the Ludlow Series, Eltonian Stage, and Lower Elton Beds is defined at a standard section in the old quarry in Pitch Coppice (SO 4726 7301), where some 4-5 m of uppermost Wenlock Limestone is overlain by Lower Elton siltstones (Holland et al. 1963, pp. 139-141 and fig. 11). WALSALL AND DUDLEY Text-fig. 2, col. 3 The Walsall borehole section (Butler 1937a) and neighbouring surface exposures indicate that there is a full Wenlock succession within the English Midlands. I agree with Ziegler et al. (1968, p. 763) that in the borehole the Llandovery/Wenlock boundary is best correlated with the base of Butler’s division M at a depth of 261 m where there is a lithological change from grey mudstones to purple and grey-green shales, corresponding to that at the Buildwas Beds/Purple Shales junction in Shropshire. Above this Butler’s divisions M to F inclusive correspond with the Wenlock Shale and comprise 237 m of grey and grey-green mudstones and shales, with thin limestones and bands of calcareous nodules, and twenty-four bentonitic clays. Cyrtograptus murchisoni and Monograptus priodon are recorded from division M of the borehole at a depth of 279-5 m (Butler 1937a, p. 246), indicating the presence of the murchisoni Zone, and although not recorded as such the basal Wenlock centrifugus Zone is probably represented in the underlying grey sedi- ments of the same division, since these pass downwards without a sedimentary break into the purple shales which contain an uppermost Llandovery crenulata Zone fauna. A prominent 10-m thick calcareous horizon known as the Barr Limestone occurs 22 m above the base of the Wenlock in the Walsall borehole. Some previous accounts (Cantrill 1919; Whitehead and Eastwood 1927) have equated this unit with the Woolhope Limestone at the base of the Wenlock in the southern part of the Welsh Borderland, but the presence of murchisoni Zone faunas below the Barr Limestone indicates that it is younger than the Woolhope Limestone (see text-fig. 2), and that it probably correlates with all or part of the riccartonensis Zone. The Barr Limestone is poorly exposed but can be examined in its type area between Hay Head and Great Barr, some 2 km east of Walsall, where it crops out in the line of old quarries between Cuckoo’s Nook (SP 05309900) and Daisy Bank (SP 04109765). Only the top 4-5 m remain exposed here, consisting of grey and olive, calcareous, blocky and flaggy-bedded shales and silt- stones, with lines of limestone nodules. There are also three bentonites within these sections, the lower two of which are 25 mm thick, while the upper one is 1 52 mm. It is not possible to correlate these bentonites with any of the four recorded by Butler (1937a, p. 254) from within the Barr Limestone of the Walsall borehole, since none of the thicknesses of bentonites or intervening sediments are comparable. From a limestone sample from the stream bed at a small waterfall near the north end of the old quarries (SP 04849879) within the top 3 m of the Barr Limestone, Dr. R. J. Aldridge has identified the following BASSETT: STRATIGRAPHY OF THE WENLOCK SERIES IN WALES 757 conodont form species: Drepanodus aduncus Nicoll and Rexroad, Hindeodella equidentata Rhodes, Ligonodina sp., Lonchodina sp., Neoprioniodus excavatus (Branson and Mehl), Ozarkodina media Walliser, Ozarkodina sp., Paltodus costalatus Rexroad, Panderodus simplex (Branson and Mehl), P. unicostatus (Branson and Mehl), Plectospathodus extensus Rhodes, Spathognathodus inclinatus (Rhodes), Spatho- gnathodus n. sp. Walliser 1964, and Trichonodella excavata (Branson and Mehl). Dr. Aldridge comments that the only specimens of possible stratigraphical value are the two identified as Spathognathodus n. sp. Walliser, recorded by Walliser (1964) only from the sagitta conodont Zone, but this is a tenuous correlation with the conodont sequence. Limestones and siltstones associated with the conodont-bearing sample yielded fragments of the trilobite Bumastus barriensis Murchison and the brachiopods Dicoelosia biloba (Linnaeus) and Leangella segmentum (Lindstrom). The beds immediately above the Barr Limestone are not exposed, but at least 63 m of the upper part of the Wenlock Shale crop out below the Wenlock Limestone in the railway cutting at Daw End, Walsall (SK 0360 0030). Comparison with the Walsall borehole suggests that the lowest beds in the railway cutting are within Butler’s division H. The topmost few metres of the Wenlock Shale are also exposed at Wren’s Nest, Dudley. Butler (1939) divided the Wenlock Limestone of Dudley into five units, totalling 66 m. Passage Beds 2-66 m Upper Quarried Limestone 7-49 m Nodular Beds 37-41 m Lower Quarried Limestone 9-78 m Basement Beds 2-94 m These were formerly accessible in the en-echelon periclines of Dudley Castle Hill, Hurst Hill, and Wren’s Nest Hill, but in the first two they are now obscured by buildings. At Wren’s Nest, however, the full sequence can be seen, although the Lower and Upper Quarried Limestones have been almost removed by quarrying. Reef structures are well developed at Wren’s Nest (Butler 1939). As mentioned earlier, the Wenlock Lime- stone of Dudley has yielded specimens of M.flemingii. Butler (1939, p. 55) also recorded flemingii from both immediately below and above the Limestone, indicating that the whole sequence, including the local ‘Lower Ludlow Shale’, is no younger than the lundgreni Zone. The Basement Beds and the Lower Quarried Limestone also crop out in the Daw End cutting at Walsall (Butler 1939, pp. 54-55), and were penetrated by the Walsall borehole (Butler 1937a), but higher beds have been removed by erosion. Wenlock strata are known to occur at depth below Upper Carboniferous and younger rocks over a wide area of the west Midlands. Details of these sub-surface sequences, and of numerous surface outcrops which are no longer accessible are well documented in the relevant memoirs of the Geo- logical Survey (Jukes 1853, 1859; Cantrill 1919; Eastwood et al. 1925; Whitehead and Eastwood 1927; Whitehead and Pocock 1947; Whitehead et al. 1928; Pocock et al. 1938). KENDAL END Text-fig. 2, col. 4 Hardie (1954) has described the small area of Wenlock rocks at Kendal End Farm (SP 004 745) near Barnt Green, Worcestershire, at the southern end of the Cambrian ridge of the Lickey Hills. The outcrop was first recorded by Murchison (1839, p. 493), but by 1898 Lapworth (p. 350) reported that there were no visible exposures and today the section remains overgrown. However, by trenching and augering Hardie was able to map a wedge-shaped outcrop some 225 m long with a maximum width of about 70 m, faulted on all sides against Pre-Cambrian, together with minor outcrops of limestone some 1 35 m to the west, faulted between Pre-Cambrian and Carboniferous. The total thickness is not known, but from the descrip- tions by Murchison and Hardie it appears that at least 4-5 m of steeply dipping nodular shales and lime- stones, with a bed of massive limestone about 1 m thick, were once exposed in the old quarry near the northern end of the outcrop. On lithological grounds both Murchison and Lapworth correlated the Kendal End rocks with the Woolhope Limestone, supported by Hardie from the evidence of faunas collected from the old quarry waste, with particular emphasis on varieties of Atrypa reticularis as described by Alexander (1949). Through the courtesy of Dr. I. Strachan I have examined the material collected by Hardie, now in the Geological Museum of Birmingham University, and identified the following brachiopods : Atrypa reticularis (Linnaeus), Antirhynchonella linguifera (J. de C. Sowerby), Dicoelosia biloba (Linnaeus), Dolerorthis rustica (J. de C. Sowerby), Leptaena depressa (J. de C. Sowerby), Orbiculoidea forbesi (Davidson), Resserel/a canalis 758 PALAEONTOLOGY, VOLUME 17 (J. de C. Sowerby), Dalejina hybrida (J. de C. Sowerby), Eoplectodonta duvalii (Davidson), Streptis grayii (Davidson), IWhitfieldella sp., Meristina obtusa (J. Sowerby), and Isorthis amplificata Walmsley. This list is a revision of the species listed by Hardie (1954, p. 14), with the addition of M. obtusa and I. amplificata. The assemblage is clearly of upper Wenlock age, precluding correlation with the Woolhope Limestone. In particular, D. rustica , M. obtusa , and I. amplificata occur only in the upper half of the Wenlock elsewhere throughout the Welsh Borderland. The lithologies described from Kendal End compare most closely with the uppermost, nodular Wenlock Shale, or Wenlock Limestone of the Walsall and Dudley sequence. Hardie’s specimens of A. reticularis are too poorly preserved and limited in number to allow confident comparisons with Alexander’s (1949) varieties, but my own studies throughout the British Wenlock suggest that her varieties do not show consistent variation of stratigraphical value. The remainder of the fauna listed by Hardie is consistent with the late Wenlock age indicated by the brachiopods, although the stratigraphical range of them all is not yet known. I have, however, confirmed the identification of the coral Ketophyllum duplex (Butler), which has been recorded elsewhere only from an horizon high in the Wenlock Shale at Dudley (Butler 19376, p. 82). On the northern margin of the Lickey Hills at Rubery, some 3 km NNW. of Kendal End, Murchison (1839, p. 493), Lapworth (1898, p. 358), Eastwood et al. (1925, pp. 14-15), and Wills et al. (1925, p. 67), recorded the presence of grey shales, thin sandstones, and limestones considered to be of early Wenlock (Woolhope Limestone) age, conformably succeeding the Llandovery. I have not examined these beds (Rubery Shale ‘Series’) which are reported to crop out in Callow Brook in the grounds of Rubery Hill Asylum (SO 992 778), but as noted by Wills (1938, p. 177) and Ziegler et al. (1968, p. 764) both the grapto- lites and brachiopods recorded indicate that the sequence is entirely of late Upper Llandovery age, possibly no younger than C5. THE LONG MOUNTAIN Text-fig. 2, col. 5 Palmer’s (1970) revised stratigraphy of the Long Mountain syncline includes a complete Wenlock sequence within the Trewern Brook Mudstone Formation. This unit comprises 450-610 m of grey mudstones with occasional interbedded nodular horizons and calcareous shelly mudstones. Earlier accounts (Elies 1900; Das Gupta 1932) of the sequence of graptolite zones in the area suggested that some of the lower Wenlock horizons are cut out by overlap. Palmer does not comment on this problem, but as noted earlier (see p. 749) a conformable faunal succession from the Llandovery can be demonstrated around Buttington. In the upper part of the Trewern Brook Mudstone Formation a persistent, lenticular development of shelly mudstones, within the laminated graptolitic mudstones, is referred by Palmer to the Glyn Member, vary- ing in thickness from 0 to 76 m. Dr. Palmer has previously informed the author (see Bassett 1972, p. 57) that the Glyn Member includes horizons which correlate with the late lundgreni Zone and ‘ nassa-dubius Inter- regnum’. The uppermost beds of the Trewern Brook Mudstone Formation grade upwards into the Long Mountain Siltstone Formation of Ludlow age. BUILTH Elies (1900, p. 371) regarded the Builth district as the type area for the Wenlock graptolite zones, a sequence later confirmed by Jones (1947, p. 3). Only six zones from murchisoni to lundgreni were included initially in the scheme, but at the top of the sequence, the ludensis Zone, now known to be of Wenlock age, was shown by Wood (1900, pp. 432-438, as vulgaris) to be present in the area. A separate centrifugus Zone has not yet been confirmed at the base of the sequence, but as at Pencerig (see p. 749), it may be represented in the River Ithon where Jones (1947, p. 9) described a gradation from the uppermost Llandovery (crenulata Zone) to the murchisoni Zone. Jones ( 1 947, p. 4) divided the Wenlock at Builth into two lithological divisions, from murchisoni to rigidus (of current nomenclature) and linnarssoni to lundgreni zones respectively; with the addition of a thickness of 30 m for the ludensis Zone (Wood 1900, pp. 433, 435) the total thickness varies from 630 m in undisturbed bedded sediments, to 900 m in areas where slumped beds are developed. The lower division, some 155 m thick, consists of dark blue-grey, flaggy-bedded, silty mudstones, thin striped shales, and thin horizons of shelly siltstones, with an irregular calcareous horizon (‘ Acidaspis limestone'), less than 1 m thick, developed locally at the base. In the upper division the bedded sediments are again dark, striped silty mudstones and shales, but this part of the sequence contains three major slump BASSETT: STRATIGRAPHY OF THE WENLOCK SERIES IN WALES 759 sheets, one within each of the linnarssoni, ellesae , and lundgreni zones. The slumped beds consist of hard, massive, dark siltstones with pockets of shelly fossils, and form prominent positive topographical features; they vary in thickness across the area, the lower sheet from 0 to 75 m, the middle from 0 to 105 m, and the upper from 10 to 150 m. The Builth sections need revision to provide modern faunal lists for the type sequence of graptolite zones, and for correlation with the type Wenlock of Shropshire; the Builth area is in any case important in its intermediate position between the shelf and basin successions and in having thin shelly horizons within the graptolite beds. DOLYHIR AND NASH SCAR Text-fig. 2, col. 6 In east Radnorshire, around Dolyhir near Old Radnor and Nash Scar near Presteigne, the lower part of the Wenlock succession comprises a thick development of massive, dark grey to white, crystalline algal and bryozoan limestone. The minimum average thickness of the limestone is about 24 m, while at Nash Scar Quarry (SO 3025 6225) up to 60 m may be present, although this may be an over-estimate due to repetition by faulting. Algal colonies make up a high proportion of the limestone, with Solenopora gracilis Garwood and Goodyear, Rothpletzella gotlandica (Rothpletz), Girvanella pusilla Johnson, and G. prob- lematica Nicolson and Etheridge dominant. Garwood and Goodyear (1919) have described the limestone, together with the structural complexity of this area, close to the Church Stretton fault belt. The basal metre or so of the Dolyhir Limestone consists of conglomerates and breccias which rest directly on shattered Pre-Cambrian (Longmyndian) grits, sandstones, mudstones, and dolerites. The Nash Scar Limestone sits disconformably on the Folly Sandstone of early Upper Llandovery (Cu) age (Ziegler etal. 1968, p. 750). Although Garwood and Goodyear (1919, pp. 17, 19) recorded large faunas from around Dolyhir, well-preserved fossils are not common, but the specimens collected by the author indicate that the limestone is of the same age in both areas. However, both the lower and upper age limits are difficult to assess in detail. The brachiopod fauna includes Antirhynchonella linguifera (J. de C. Sowerby), Streptis grayii (Davidson), Atrypa sp., Leptaena sp., Megastrophia ( Protomegastrophia ) sp., lAncillotoechia sp., Whitfieldella sp., and Plectatrypa sp., and Dr. R. M. Owens has kindly identified the following trilobites from Dolyhir: Cornuproetus peraticus Owens, Bumastus sp., IPrantlia sp., a lichid, a scutellid, and an odontopleurid. As pointed out by Ziegler et al. (1968, p. 750), the evidence of the Folly Sandstone indicates that the Nash Scar Limestone could possibly be as old as C3 on stratigraphical grounds, but the fauna listed above contains no diagnostic Llandovery elements, and both Cornuproetus and Bumastus are known only from post-Llandovery strata. A Wenlock age is supported by the record of Rhipidium in the Garwood Collection at the Institute of Geological Sciences (Ziegler et al. 1968), but I have not been able to trace the specimens. The specimens of Whitfieldella and Plectatrypa are close to, and possibly conspecific with, material collected by the author from the Woolhope Limestone in the Welsh Borderland, suggesting an early Wenlock age for at least part of the Dolyhir and Nash Scar Limestones, and supporting a general correlation with the Woolhope Limestone as assumed by Garwood and Goodyear (1919) and earlier authors. Kirk ( 195 la, p. 56) regarded the Dolyhir Limestone and shelly mudstones of Han ter Hill as contemporaneous with the murchisoni and riccartonensis zones. From both the Nash Scar and Dolyhir Limestones Dr. R. J. Aldridge (pers. comm.) has recovered conodont faunas of the sagitta Zone (Walliser 1964), with Spatho- gnathodus sagitta rhenanus Walliser dominant. The earliest known sagitta Zone faunas are in the Hogklint Beds of Gotland (Fahreus 1969, p. 9), which contain riccartonensis Zone graptolites (Bassett and Cocks 1974, p. 5); thus the Dolyhir and Nash Scar Limestones appear to extend upwards into the riccartonensis Zone or younger horizons. Both at Dolyhir and Nash Scar the overlying shales and siltstones are faulted against the limestones, precluding a firm stratigraphical assessment of the original age relationships. Kirk (1951a, p. 56) briefly stated that The limestone is overlain by mudstones with Cyrtograptus symmetries ( sic rigidus). Within the main limestone mass at Dolyhir Quarries (SO 2412 5808), the small, faulted patch of shale referred to by Garwood and Goodyear (1919, p. 19 and pi. 7), has yielded Monograptus flemingii (Salter), indicating an age within the rigidus to lundgreni zones. The shales immediately above the Nash Scar Limestone at the north-east end of Nash Scar Quarry (SO 3045 6245), have yielded to Mr. N. J. Hancock and the author a fauna definitely referable to the lundgreni Zone. The thickness of Wenlock beds above the limestones is probably at least 90 m, and includes part of the shelly olive mudstones which Kirk ( 1 95 1 A, p. 72) reported to span the lundgreni, ludensis , and nilssoni zones. 760 PALAEONTOLOGY, VOLUME 17 THE ABBERLEY HILLS Text-fig. 2, col. 7 The complex structural setting of the Silurian of the Abberley Hills (Herefordshire and Worcestershire) has long been known (Murchison 1839; Phillips 1848; Groom e.g. 1900), and the stratigraphy has been revised by Mitchell et al. (1962) and Phipps and Reeve (1967). Lowest Wenlock rocks are nowhere exposed, being cut out by thrusting or overlain unconformably by Triassic strata, and younger Wenlock Shale is exposed only in a few small, isolated patches (for details see Mitchell et al. 1962, pp. 26-27). In the largest of these outcrops, between Hillside (SO 754612) and Kingswood Common (SO 747 601) to the north- west of Martley, up to 360 m of grey calcareous shales and siltstones with thin nodular limestones may be present, but part of the succession may be repeated by faulting. The Wenlock Limestone, 36 m thick, is well exposed in the southern half of the area, where steeply dipping and overturned beds crop out in a line of quarries for about 3 km northwards from the large quarry at Penny Hill (SO 7515 6145). The blue-grey bedded and nodular limestones contain numerous thin partings of grey shale, and a number of cream- coloured bentonitic clays up to about 200 mm thick. THE MALVERN HILLS Text-fig. 2, col. 8 The thrusting which affects the Silurian of the Abberley Hills is not developed to the same extent along the southerly continuation of the structural line into the Malvern Hills (Herefordshire), and so in the latter area a full Wenlock succession is present along most of the outcrop. Phipps and Reeve (1967) have revised the stratigraphy and given an outline of previous work, and Penn and French (1971) have also commented on the succession. The basal lithological unit of the Wenlock is the Woolhope Limestone, which is present here in its most northerly development within the inkers of the Welsh Borderland. It has an average thick- ness of 15-21 m, but may thicken to a maximum of 60 m near North Malvern (Phipps and Reeve 1967, p. 343). It mainly consists of olive-grey, rubbly, calcareous siltstones and argillaceous limestones, which separate an upper and lower development of flaggy bedded, silty limestones ; southwards through the area the calcareous beds become more dominant at the expense of the siltstones. Eocoelia cf. sulcata (Prouty) occurs in this unit, indicating faunal continuity from the underlying Wych Beds, which are of latest Upper Llandovery (C6) age in their upper part (Ziegler et al. 1968, p. 757). Olive-grey siltstones and shales with lines of calcareous nodules comprise the bulk of the Wenlock Shale, which varies from 180 to 270 m thick; the nodule horizons become more common in the top 10 m of the sequence, immediately below the transition into the Wenlock Limestone. Within the latter division Phipps and Reeve (1967, p. 345) described five principal lithofacies types: calcareous mudstones, nodular limestones, bioclastic (coquinal) limestones, pisolitic limestones, and bioherms. These may occur in any order or association within the succession, which varies in thickness from 60 to 150 m; however, the pisolitic limestones are often developed close to the base of the Wenlock Limestone and usually close to bioherms. Penn (1971) has described the develop- ment of bioherms in the Malvern area. THE WOOLHOPE INLIER Text-fig. 2, col. 9 The Woolhope Limestone of the type area is 36 m thick (Squirrell and Tucker 1960) ; apart from a develop- ment of thickly bedded shelly limestone in the middle of the sequence, it consists largely of rubbly calcareous siltstones and nodular argillaceous limestones similar to the Malvern Hills. Costistricklandia lirata lirata (J. de C. Sowerby) and Eocoelia cf. sulcata (Prouty) occur rarely in this unit, confirming the evidence from the Malverns of faunal continuity from the Upper Llandovery. There is no graptolite evidence to confirm the basal Wenlock age of the Woolhope Limestone, which on the evidence of C. lirata lirata could, in part, be as old as latest Upper Llandovery (C6), but the base of the Woolhope Limestone is some metres strati- graphically above typical late C6 associations of Costistricklandia and Palaeocyclus porpita (Linnaeus), while its total shelly fauna and lithologies compare closely with those of the Buildwas Beds of Shropshire and contain brachiopod elements such as Eoplectodonta duvalii (Davidson) and Anastrophia deflexa (J. de C. Sowerby), which are very rare or absent in the British Upper Llandovery. Dr. P. D. Lane informs me that the Woolhope Limestone also contains the trilobite genus Bumastus which is not known outside the W enlock BASSETT: STRATIGRAPHY OF THE WENLOCK SERIES IN WALES 761 elsewhere in Britain. I have argued previously (in discussion of Cocks and Rickards 1969, p. 236) that the Woolhope Limestone is slightly younger than the Buildwas Beds, but in the light of subsequent studies I now consider that minor differences in the two faunas are more likely to be environmentally, rather than stratigraphically, controlled. The uppermost Woolhope Limestone grades into the siltstones and thm argillaceous limestones of the overlying Wenlock Shale, which varies in thickness from 300 to 360 m and which in turn is overlain by 45-51 m of Wenlock Limestone. Shales with limestone nodules, similar to the Tickwood Beds of Wenlock Edge, are developed at the top of the Wenlock Shale, and nodular limestones occur both at the base and top of the irregularly bedded Wenlock Limestone. GORSLEY Text-fig. 2, col. 10 Between the Woolhope and May Hill Inliers, Silurian rocks are exposed around the village of Gorsley, Herefordshire, where the succession has been described by Lawson (1954). The oldest Silurian beds belong to the Gorsley Limestone, whose base is not seen but which is exposed to a thickness of 3-6 m in Linton Quarry (SO 67702574) and 5-5 m in Hartley’s Quarry (SO 67702616) (Lawson 1954, pp. 229-231). Through the courtesy of Dr. I. Strachan I have examined some of Lawson’s specimens from the Gorsley Limestone, now deposited in Birmingham University; these faunas confirm Lawson’s assignment of the Gorsley Lime- stone to the Wenlock rather than the Ludlow (Aymestry Limestone) as previously supposed (Lawson 1954, p. 227). In particular, the common occurrence of Meristina obtusa (J. Sowerby) is a clear indicator of an age within the upper half of the Wenlock, and combined with the distinctive blue-grey massive limestone lithology there can be no doubt that the Gorsley Limestone is a correlative of the Wenlock Limestone. The top of the Gorsley Limestone is an irregular erosion surface on which early Leintwardian (Ludlow) silt- stones rest unconformably. THE MAY HILL INLIER Text-fig. 2, col. 1 1 In the May Hill Inlier the Woolhope Limestone, similar to the same horizon at Woolhope and the Malverns, appears to thicken southwards from 21 m at May Hill itself to over 60 m at Little London (Lawson 1955); as at Woolhope these beds contain C. lirata lirata. The Wenlock Shale is 210-240 m thick and the succeed- ing Wenlock Limestone reaches a maximum thickness of 105 m. Lawson (1955, pp. 89-90) described three units within the Wenlock Limestone. The lowest division is some 18 m of thinly bedded grey limestone in which reef structures are developed, with bands of pisolitic limestone similar to those low in the Wenlock Limestone of the Malverns ; these are succeeded by thinly bedded nodular limestones and shales, in turn overlain by thinly bedded nodular limestones passing laterally southwards into ferruginous and oolitic beds. In the extreme south of the inlier the Wenlock Limestone thins to 30 m and the upper unit cannot be recognized separately. THE USK INLIER Text-fig. 2, col. 12 The Wenlock Shale, with a minimum thickness of 240 m (Walmsley 1959), occupies the core of the Usk inlier. At a trackside section (SO 3435 0395) above the right bank of the River Usk 100 m south-east of the pumping station on Craig y Pandy, beds close to the axis of the anticline, and hence close to the oldest exposed, have yielded a number of Monograptus flemingii flemingii (Salter). These are the first graptolites recorded from the Wenlock of Usk, and since flemingii first occurs with certainty in the rigidus Zone, the whole of the exposed Usk Silurian succession can be dated as post -riccartonensis Zone in age. The bulk of the Wenlock Shale consists of grey-green and buff shales and calcareous siltstones with occasional bands of nodular limestone, but the top 4-5 m comprise rust brown calcareous and micaceous sandstones indica- tive of a period of shallowing. The overlying Wenlock Limestone has a maximum thickness of 13-5 m in the west of the area, where a lower, massive division and an upper, nodular division may be recognized, but it thins rapidly eastwards to less than 1 m (Walmsley 1959, p. 487). In most exposures the highest lime- stones are succeeded by 1-2 m of buff-coloured calcareous siltstones with a typically Wenlock fauna and C 762 PALAEONTOLOGY, VOLUME 17 grouped by Walmsley (p. 487) within the Wenlock Limestone; more recently Squirrell and Downing (1969, pp. 11 12) mapped these beds as basal Elton Beds within the Ludlow. In the northern half of the inlier limestone beds are rarely exposed, but at Weir Wood, Trostrey (SO 36020420) about 6 nr of nodular beds are seen to overlie some 12 m of green and buff siltstones (Walmsley 1959, pp. 487-488, fig. 2). A thin bed of crinoidal limestone within the siltstones has yielded a conodont fauna (Austin and Bassett 1967) referable to the sagitta Zone. In 1964 a large trench was excavated across the central area of the Usk anticline, in a roughly east-west direction, in connection with a drainage scheme from the Llandegveth reservoir. The trench, which was about 4 km long, 3 m wide, and 2-4 m deep, provided some excellent exposures in the Wenlock Shale at its western end (text-fig. 3), but unfortunately over 3 km of the section to the east passed through glacial and alluvial drift. The oldest beds in the trench, at Grid Reference SO 3435 0245, were some 247 nr strati- graphically below the base of the Wenlock Limestone, which must be close to the oldest beds brought to the surface in the inlier. The beds there consist of fossiliferous grey-green silty shales with occasional muddy grey calcareous nodules. In the debris of material excavated from these oldest beds, one nodule yielded a single graptolite identified as Monograptus flemingii flemingii, confirming the post -riccartonensis Zone age of the sequence indicated by the fauna from Craig y Pandy (see above). Between the above-mentioned exposure and the railway line (see text-fig. 3) at Grid Ref. SO 337 1 02 1 5 the trench was cut through deep drift deposits, but the remaining 562 m of the section to a point near Cwm (SO 3335 0173) revealed continuous exposure through a thickness of some 135 m of Wenlock Shale. The beds consist mostly of a series of monotonous grey and green blocky mudstones with subordinate bands of pale olive laminated shales. Fossils were found to be rare throughout the bulk of the succession although good collections of typical late Wenlock brachiopods were made from the horizons indicated in text-fig. 3. Of especial interest was the presence of two thin (100 mm) beds of coarse-grained, buff, calcareous, ripple- rnarked sandstone 45 m below the base of the Wenlock Limestone, containing many brachiopods and bivalves and lithologically similar to the topmost 4-5 m of the Wenlock Shale of the inlier. Another thin sandstone band 37-5 m below the base of the Wenlock Limestone includes a number of grey mudstone pellets of Wenlock Shale sediment. The youngest beds exposed in the trench were 4-5 m below the base of the Wenlock Limestone and immediately below the topmost sandy beds of the Wenlock Shale which out- crop in the adjacent laneside section at Grid Ref. 34809990 (Walmsley 1959, p. 487, loc. 3). THE TORTWORTH INLIER Text-fig. 4, col. 1 The Wenlock rocks in the Tortworth Inlier, Gloucestershire, have been assigned by Curtis (1972) to a single stratigraphical unit, the Brinkmarsh Beds, which comprise some 244 m of shales, mudstones, siltstones, and calcareous sandstones, with a number of impersistent limestones. The sequence is best exposed in the south- west of the inlier, to the south of Whitfield, where three prominent limestones occur, at the base and near the middle and top of the succession respectively (Curtis 1972, pp. 20-21 and fig. 3). The basal limestone, which passes laterally into calcareous sandstones, is a correlative of the lower part of the Woolhope Lime- stone, and as at May Hill, Woolhope, and in the Malverns it contains rare remnant Llandovery braclnopod species such as Leptostrophia compressa (J. de C. Sowerby). Immediately above the limestone there is a dis- tinctive horizon, the Pyaiactis Band, consisting of about 3 m of maroon-red shales and siltstones with a rich coral/brachiopod fauna, of which Pycnactis mitratus (Schlotheim), Phaulactis glevensis (Ryder), Resserella whitfieldensis Bassett, and Whitfieldella sp. are the most common species. The middle part of the sequence is poorly exposed, but the uppermost 30 m, immediately below the unconformable base of the Upper Old Red Sandstone, are very well exposed in the A38 road near Buckover (SO 6668 907 1 -SO 6677 9078). The late Wenlock age suggested by Curtis ( 1 972, p. 26) and Curtis and Cave ( 1 964, p. 43 1 ) for these uppermost beds is confirmed by abundant specimens of Meristina obtusa (J. Sowerby), together with species such as Trigonirhynchia stricklandi (J. de C. Sowerby), Cordatomyonia edgelliana (Davidson), Leptaena depressa (J. de C. Sowerby), Amphistrophia funiculata (McCoy), and Protochonetes minimus (J. de C. Sowerby), which together indicate a correlation with an horizon high in the Wenlock Shale or the Wenlock Limestone. Curtis ( 1 972, pp. 25-26) has commented on the similarity of the arenaceous deposits at Tortworth with parts of the succession in the Usk and Rumney (Cardiff) Inliers. Early work in both the Tortworth and Eastern Mendips Inliers (see below) has been summarized by Curtis (1955). Wenlock rocks also crop out in three very small inliers in the bed of the Little Avon river near Wickwar, Key text-fig. 3. Map and vertical section of temporary (1964) exposure in the Wenlock Shale of the Usk Inlier between Cwm and Monkswood. TORTWORTH EASTERN RUMNEY LLANDOVERY FRESHWATER MARLOES- WINSLE MENDIPS (CARDIFF) - LLANDEILO EAST-WEST WOOLTACK- ST. ISHMAEL'S >1 3 O 1 N 3 M 6 D o U 00 _o o o P-1 pp OX) G o § > o G Q g .G cd zd < “i cd OX) ° 'S £ c B * < OX) c a cd C/D g O T3 G Q N® o \ O O -C £ C/D c o g cd T3 G 3 X) cd "O ^ 53 g OX) > cd cd G -G cj ^ oj ^2 ex' C/D g -5 s & C/D Oh flj C/D L> x> 75 Oh s cl "O o ° r?J 3 3 3 3 3 c 3 .g s O ^ w a. & CL c O O ^ G H £ ■■d c/d -O _aj cd £> a Oh -G C/D >> ■G G > c 73 £ s £ 2 O G >> o G G a- hG OX) 3 Oh G -G £ Oh C/D C/D -G C/D 3 .3 s § 1 i o ° Z U o. o \o c/d Ph OhI o o Z o c C4-H o cd O H o o ° rs cj O <“• 1 o o CL ^ cd 3 S L> y s; S §• o> C/D n ^ y II _S CL « 3 >> 13 > O a> 3 5 i-t cr 3 S « 5^ -Ci - o + 2? <11 a'S 3 3 3 75 H Z tu -J < > w 3 0- >< H z p s s o X U H hJ J < w Q Q O £ S z w O o s o X Q Q > H z p s s o u tu O V CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 783 THE COMMUNITIES The five major communities are described by means of the quantitative information in Tables 2-11. The tables are based solely on the brachiopod fraction of the fauna for two reasons: (1) brachiopods generally make up at least 90% of the total fauna, (2) the taxonomic uncertainty is less with brachiopods than with most other groups. Additional information about the community, including its non-brachiopod fauna, follows each table. table 2. Composition of the Wenlock Salopina community. WENLOCK SALOPINA COMMUNITY PREVALENT SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 1. Salopina 94-1 1981-0 2. ‘ Camaroloechia ’ nucula 88-2 2236-2 3. A. ( Pembrostrophia ) 29-4 545-0 4. Lingula 29-4 93-9 5. Rhynchotreta cuneata 23-5 173-4 6. Craniops 23-5 137-1 7. Leptostrophia filosa 23-5 49-2 OTHER SPECIES 8. A try pa 23-5 43-9 9. Meristina 23-5 32-4 10. Orbiculoidea 23-5 28 1 11. ‘ Camaroloechia' tripartita 17-6 211-1 12. Protochonetes sp. 17-6 180-6 13. Homoeospira 17-6 121-2 14. Eocoelia angelini 17-6 89-0 15. Howellella spp. 17-6 88-7 16. Sphaerirhynchia wilsoni 17-6 38 6 17. Amphistrophia spp. 17-6 37-7 18. Athyrids 17-6 36-9 19. Whitfieldella 17-6 12-7 20. Marklandella 1 1-8 241-7 21. Strophochonetes 1 1-8 140-1 22. Coolinia 11-8 23-0 23. Rhynchonellids 118 15-9 24. Mclearnites 5-9 257-0 25. Sphaerirhynchia davidsoni 5-9 173-3 26. ‘ Camaroloechia ’ llandoveriana 5-9 26-5 27. Shagamella 5-9 21-2 28. Schizotreta 5-9 12 9 29. Cordatomyonia edgelliana 5-9 9-8 30. Striispirifer 5-9 8-8 31. Leptaena spp. 5-9 5-9 32. Gypidula 5-9 4-7 33. Hyattidina 5-9 4-5 34. Eospirifer 5-9 2-9 Number of collections studied = 17 Species density = 6-5 Homogeneity = 47-33% Distinctness coefficient = 571% 784 PALAEONTOLOGY, VOLUME 17 1. Salopina community ( Tables 2 and 3) Besides the brachiopod content listed in the tables, bivalves, especially Pteronitella and Palaeopecten, and the distinctive genus Nuculites, frequently occur. Gastropods are often present, though they are rarely abundant. Tentaculites is common, especially in the Wenlock. A common feature of the Salopina community is the numerical dominance of a few species. For example, collection SG-15 (Ludlow of the Sawdde Gorge) contains 54% Salopina lunata, 24% Howellella elegans, and 9% Sphaerirhynchia wilsoni, a total of 87% for just three species. The most important difference between Tables 2 and 3 is the enormous increase from the Wenlock to the Ludlow in Protochonetes (excluding P. minimus which is confined to deeper-water communities). The lower homogeneity in the Wenlock is attributable to locally distributed species: for example, Amphistrophia ( Pembro - strophia) freslmaterensis appears to be confined to Pembrokeshire (Bassett 1971, pp. 325-327), and the rhynchonellid we have called '’Camarotoechm tripartita has been recorded only from the East Mendips. table 3. Composition of the Ludlow Salopina community. LUDLOW SALOPINA COMMUNITY PREVALENT SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 1. Protochonetes ludloviensis 1000 3185-1 2. ‘ Camarotoechia ' nucula 100-0 1274-8 3. Salopina 83-3 2456-8 4. Howellella spp. 75-0 498-9 5. Sphaerirhynchia wilsoni 58-3 185-4 6. Orbiculoidea 50-0 48-5 7. Dayia 41-7 254-4 OTHER SPECIES 8. Lingula 417 47-5 9. Whitfieldella 25-0 183-7 10. Leptostrophia filosa 25-0 121-7 11. Atrypa 25-0 514 12. C rani ops 16-7 162-0 13. Isorthis 16-7 70-2 14. Strophochonetes 16 7 63-9 15. Hyatt idina 8-3 30-3 16. Schizotreta 8-3 9-6 17. Schizocrania 8-3 9-6 18. ‘ Camarotoechia' tripartita 8-3 5-0 19. Leptaena spp. 8-3 4-7 20. Slialeria 8-3 3-5 Number of collections studied = 12 Homogeneity = 69-93% Species density = 7-2 Distinctness coefficient = 571% CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 785 2. Homoeospira or Sphaerirhynchia communities ( Tables 4 and 5) The faunas of the Homoeospira and Sphaerirhynchia communities are very similar. Because of this the two communities may be considered variants of one, living at different times. During the Wenlock the brachiopod Homoeospira is one of the dominant species. Unpublished studies by Mr. J. M. Hurst of Oxford University table 4. Composition of the Wenlock Homoeospira community. WENLOCK HOMOEOSPIRA COMMUNITY PREVALENT SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 1. Howellella spp. 1000 1477-1 2. ‘ Camarotoechia' nucula 90-9 628-7 3. Salopina 81-8 444-2 4. Homoeospira 72-7 1030-5 5. Atrypa 63-6 1375-9 6. Leptostrophia filosa 45-5 197-5 7. Amphistrophia spp. 45-5 177-7 8. Meristina 45-5 138-5 9. Protochonetes sp. 36-4 176-3 10. Craniops 36-4 117-5 OTHER SPECIES 1 1 . Leptaena spp. 36-4 108-1 12. Sphaerirhynchia wilsoni 27-3 69-6 13. Gypidula 27-3 57-3 14. Isorthis 27-3 52-7 15. Dalejina 27-3 32-7 16. Protochonetes minimus 18-2 203-1 17. Marklandella 18-2 156-3 18. Rhynchotreta cuneata 18-2 86-4 19. Resserella canalis 18-2 78-2 20. Eospirifer 18-2 72-7 21. Eocoelia angelini 18-2 37-3 22. Protomegastrophia 18-2 26-4 23. Nucleospira 18-2 12-7 24. A. (Pembrostrophia) 9-1 410-9 25. Hyattidina 9-1 67-8 26. Mclearnites 9-1 62-7 27. Brachyprion 9-1 32-0 28. Striispirifer 9-1 27-5 29. Aegiria grayi 9-1 21-8 30. Whitfieldella 9-1 21-8 31. Coolinia 9-1 16 4 32. Strophochonetes 9-1 12-7 33. Sphaerirhynchia davidsoni 9-1 12-5 34. Athyrids 9-1 4-5 35. Strophonella 9-1 3-6 Number of collections studied = 1 1 Homogeneity = 60-21% Species density = 10-2 Distinctness coefficient = 40 0% 786 PALAEONTOLOGY, VOLUME 17 indicate that in the Wenlock Limestone (late Wenlock) Sphaerirhynchia has become equally important. By the Ludlow Sphaerirhynchia has become a dominant species with Homoeospira of minor significance. Text-fig. 3 shows the shift in importance. Changes in community composition with time can have various causes, such as disappearance of a species followed by the spread of a new species into the vacant ecological niche, or the result of facies changes. The shift from Homoeospira to Sphaerirhynchia can probably be related to facies changes, Sphaerirhynchia pre- ferring carbonates. In the Ludlow of Wales, Sphaerirhynchia predominates in both carbonates and elastics. Pterioid bivalves are frequently present and so are the locally abundant Fuchsella amygdalina and Paracyclas sp., especially in the Ludlow. The name Homoeospira community has been selected for the Wenlock despite the greater abundance of Howellella (Table 4) because of the wide distribution of Howellella (text-fig. 3) through the communities. Other abundant species (‘'Camaro- toechia' nucula and Salopina) are even more abundant in the Salopina community. table 5. Composition of the Ludlow Sphaerirhynchia community. LUDLOW SPHAERIRHYNCHIA COMMUNITY PREVALENT SPECIES 1 . Sphaerirhynchia wilsoni 2. ‘ Camarotoechia ’ nucula 3. Salopina 4. Protochonetes ludloviensis 5. Whitfieldella 6. Isorthis 7. Howellella spp. 8. Dayia 9. Leptostrophia filosa 10. Lingula OTHER SPECIES 11. Atrypa 12. Craniops 13. Mesopholidostrophia spp. 14. Homoeospira 15. Orbiculoidea 16. Shagamella 17. Trig on irhyn ch ia 18. Gypidula 19. Leptaena spp. 20. Protochonetes minimus 21. Strophonella 22. Amphistrophia spp. 23. Nucleospira 24. Strophochonetes Number of collections studied = 9 Species density = 9-7 PRESENCE PERCENTAGE FREQUENCY PRESENCE 1000 1442-0 1000 1099-0 88-9 1106-5 88-9 848-7 77-8 779-1 66-7 979-5 55-6 857-6 55 6 309-2 44-4 452-0 44-4 133-8 44-4 131-7 33-3 70-9 22-2 72-4 22-2 45-7 22-2 25-7 111 140-0 111 49-8 11-1 18-7 111 12 4 111 12-2 111 6-2 111 6-2 111 6-2 111 4-3 Homogeneity = 7417% Distinctness coefficient = 40 0% CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 787 3. Isorthis community ( Tables 6 and 7) Bivalves and gastropods are uncommon in this community, and there are no characteristic genera. The high species density in the Isorthis community is part table 6. Composition of the Wenlock Isorthis community. WENLOCK ISORTHIS COMMUNITY PREVALENT SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 1. Isorthis 1000 957-0 2. Atrypa 1000 597-4 3. Howellella spp. 87-5 388-6 4. Resserella canalis 75-0 638-6 5. Eospirifer 75-0 386-5 6. Eoplectodonta spp. 75-0 366-0 7. Amphistrophia spp. 75-0 186-6 8. ‘ Camarotoechia nucula 62-5 226-3 9. Coolinia 500 180-6 10. Craniops 37-5 638-9 11. Striispirifer 37-5 319-0 12. Anastrophia 37-5 140-7 13. Protochonetes sp. 37-5 130-4 14. Leptaena spp. 37-5 122-6 15. Salopina 37-5 76-6 16. Gypidula 37-5 48-1 OTHER SPECIES 17. Orthids 37-5 36-6 18. Atrypina 37-5 27-9 19. Dinobolus 25-0 543-8 20. Protochonetes minimus 250 180-4 21. Meristina 25-0 156-3 22. Dalejina 250 131-2 23. Whitfieldella 25-0 106-6 24. Skenidioides 25-0 85-0 25. Leptostrophia filosa 25-0 82-5 26. Mesopholidostrophia spp. 250 63-7 27. Athyrids 25-0 50-0 28. Leptaena aff. purpurea 250 36-9 29. Nucleospira 25-0 30-7 30. Clorinda sp. 250 18-7 31. Strophonella 25-0 11-2 32. Cordatomyonia edgelliana 12-5 104-1 33. Trigonirhynchia 12-5 92-5 34. Cyrtia 12-5 512 35. Sphaerirhynchia wilsoni 12-5 40-0 36. Leangella 12-5 23-7 37. Homoeospira 12-5 23-7 38. Orbiculoidea 12-5 21-2 39. Eocoelia angelini 12-5 21-2 40. Strophochonetes 12-5 19 1 41. Katastrophomena 12-5 13-8 42. Spirigerina 12-5 12-5 43. Shagamella 12-5 7-5 788 PALAEONTOLOGY, VOLUME 17 TABLE 6 (coni.) : OTHER SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 44. Brachyprion 12-5 7-5 45. Protomegastrophia 12-5 6-2 46. Leptostrophia compressa 12-5 6-2 47. Dictyonella 12-5 5-0 48. Lingula 12-5 5-0 Number of collections studied = 8 Homogeneity = 60-83% Species density = 15-7 Distinctness coefficient = 50 0% table 7. Composition of the Ludlow Isorthis community. LUDLOW ISORTHIS COMMUNITY PREVALENT SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 1 . Isorthis 1000 2567-8 2. Mesopholidostrophia spp. 71-4 668-1 3. Sphaerirhynchia wilsoni 64-3 354-5 4. Atrypa 57-1 559-0 5. Dalejina 50-0 1137-8 6. ‘ Camarotoechia ’ nucula 50-0 191-1 7. Craniops 50-0 154-8 8. Amphistrophia spp. 50-0 141-4 9. Shagamella 42-9 761-6 10. Leptostrophia filosa 42-9 321-2 OTHER SPECIES 11. Howellella spp. 42-9 232-7 12. Homoeospira 42-9 62 1 13. Salopina 35-7 126-1 14. Leptaena spp. 35-7 114-1 15. Coolinia 28-6 77-8 16. Protochonetes minimus 28-6 21-2 17. Shaleria 21-4 430-6 18. Strophonella 21-4 106-6 19. Aegiria grayi 21-4 40-3 20. Lingula 214 37-1 21. Whitfieldella 214 18-9 22. Dayia 14 3 146-0 23. Glassia 14 3 119-3 24. Skenidioides 14 3 116-7 25. Strophochonetes 14 3 68-2 26. Gypidula 14-3 54-4 27. Meristina 14-3 19 6 28. Conchidium 7-1 221-6 29. Protochonetes ludloviensis 7-1 57-8 30. Lissatrypa 7-1 17-4 3 1 . Dicoelosia biloba 7-1 12-3 32. Eospirifer 7-1 12-3 33. Athyrids 7-1 9-3 34. Schizotreta 7-1 3-4 Number of collections studied = 14 Homogeneity = 55-61% Species density = 10-4 Distinctness coefficient = 40-0% CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 789 of a continuous gradient from the Salopina community increasing through the Homoeospira/ Sphaerirhynchia and Isorthis communities, and reaching the highest values in the Dicoelosia community. In the Ludlow, the Isorthis community is more highly dominated by Isorthis spp. than in the Wenlock. In the earlier part of the Wenlock the fauna is usually dominated by Eospirifer radiatus or Striispirifer plicatellus, while Isorthis is often rare. The Ludlow Isorthis community still maintains its ecological position between the Sphaerirhynchia and Dicoelosia communities, and so it occupies the same range of environment as in the Wenlock, despite the increase in Isorthis itself. 4. Dicoelosia community ( Tables 8 and 9) Bivalves are uncommon in the Dicoelosia community, but trilobites are frequently found, and orthocone nautiloids and graptolites, while not members of the living benthonic community, are sometimes present in the thanatocoenose. table 8. Composition of the Wenlock Dicoelosia community. WENLOCK DICOELOSIA COMMUNITY PREVALENT SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 1. Leangella 1000 764-6 2. Dicoelosia biloba 77-8 1765-3 3. Eospirifer 77-8 451-3 4. Skenidioides 77-8 372-8 5. Dalejina 77-8 283-2 6. Eoplectodonta spp. 77-8 262-6 7. A try pa 66-7 253-0 8. Resserella canalis 66-7 216-8 9. Atrypina 66-7 152-8 10. Howellella spp. 55-6 328-4 1 1 . Isorthis 55-6 262-1 12. Craniops 55-6 248-6 13. Orthids 55-6 63-2 14. Resserella sabrinae 44-4 451-8 15. Mesopholidostrophia spp. 44-4 209-9 16. Glassia 44-4 63-3 17. Strep tis 44-4 40-8 OTHER SPECIES 18. Protochonetes minimus 33-3 945-0 19. Lissatrypa 33-3 322-2 20. Nucleospira 33-3 37-7 21. Leptaena all. purpurea 33-3 37-3 22. Lingula 33-3 33-7 23. Leptostrophia filosa 33-3 29 1 24. Cordatomyonia edge 1 liana 22-2 318-4 25. Anastrophia 22-2 135-8 26. cf. Visbyella trewerna 22-2 105-8 27. Shagamella 22-2 91-7 28. Striispirifer 22-2 67-4 29. Whitfieldella 22-2 52-2 30. Cyrtia 22-2 36-4 790 PALAEONTOLOGY, VOLUME 17 table 8 ( cont .): OTHER SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 31. Hyattidina 22-2 20-7 32. Dolerorthis spp. 22-2 20-4 33. Gypidula 22-2 19 6 34. Salopina 22-2 111 35. Trigonirhynchia 111 61 8 36. Amphistrophia spp. 111 58-9 37. Orbiculoidea 111 50-9 38. Coolinia 111 37-8 39. Protochonetes sp. 111 30-9 40. Mesounia 111 28-7 41. Sphaerirhynchia wilsoni 111 25-4 42. Dictyonella 111 20-4 43. Meristina 111 16-7 44. Clorinda sp. 111 111 45. lEridorthis 111 111 46. Plectatrypa 111 9-6 47. ‘ Camarotoechia nucula 111 5-6 Number of collections studied = 9 Homogeneity = 64-31% Species density = 16-8 Distinctness coefficient = 52-9% table 9. Composition of the Ludlow Dicoelosia community. LUDLOW DICOELOSIA COMMUNITY PREVALENT SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 1. Isorthis 88-9 1171-1 2. Dalejina 88-9 872-7 3. Howellella spp. 88-9 645-8 4. Protochonetes minimus 88-9 604-3 5. Skenidioides 88-9 521-5 6. Atrypa 88-9 484-1 7. Dicoelosia biloba 77-8 745-9 8. Leptostrophia filosa 77-8 242-9 9. Mesopholidostrophia spp. 55 6 476-9 10. Shagamella 55-6 455-0 1 1 . Nucleospira 55 6 370-7 12. Aegiria grayi 55-6 251-1 13. Amphistrophia spp. 44-4 272-6 14. Craniops 44.4 185-0 15. Cyrtia 44.4 129-7 16. Lingula 44.4 100-4 OTHER SPECIES 17. Glassia 444 97-9 18. Leptaena aff. purpurea 33-3 175-8 19. Leangella 33-3 166-7 20. Gypidula 33-3 106-2 21. Eospirifer 33-3 59-7 22. Strophonella 22-2 68-5 23. Resserella canalis 22-2 67-3 CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 791 table 9 (i cont .): OTHER SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 24. Hyattidina 22-2 62 1 25. Whitfieldella 22-2 58-2 26. ‘ Camarotoechia' nucula 22-2 55-2 27. Sphaerirhynchia wilsoni 22-2 38-4 28. Orthids 22-2 37-2 29. Salopina 22-2 29-6 30. Trigonirhynchia 22-2 16 2 31. Nanospira 111 56-2 32. Leptaena spp. 111 41 1 33. cf. Visbyella trewerna 111 40-6 34. Rhynchotreta cuneata 111 25-8 35. Athyrids 111 25-6 36. Katastrophomena 111 21-8 37. Coolinia 111 8-6 38. Striispirifer Number of collections studied = 9 111 Homogeneity = 69-30% 7-8 Species density = 15-6 Distinctness coefficient = 43-7% The community is characterized by brachiopod species with a small adult size, such as Skenidioides lewisii , Dicoelosia biloba, Streptis grayii, Protochonetes minimus, and plectambonitid species. Another characteristic feature is the lack of dominant species in contrast to the Salopina community. The Wenlock and Ludlow Dicoelosia communities differ slightly in species com- position. Isorthis is much more common in the Ludlow than the Wenlock, while Leangella segmentum , Eoplectodonta duvalii, and Streptis are important in the Wenlock, but rare in the Ludlow; indeed Eoplectodonta and Streptis have not been recorded in the bulk collections, and are therefore absent from Table 9. 5. Visbyella community ( Table 10 for Wenlock only ) Certain species are virtually restricted to the Visbyella community (Hancock, Hurst andFiirsich 1974). In this class are Visbyella trewerna and another tiny resserellid very similar to Visbyella (Bassett 1970-1972), ‘ Clorinda ' dormitzeri, Bracteoleptaena, Mesounia, and the bivalve Cardiola interrupta. Ostracods are often abundant, and pelagic groups, notably orthocones and graptolites, are commonly preserved with the benthonic assemblage. table 10. Composition of the Wenlock Visbyella community. WENLOCK VISBYELLA COMMUNITY PREVALENT SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 1 . cf. Visbyella trewerna 100-0 2268-6 2. Protochonetes minimus 100-0 615-8 3. Glassia 80-0 367-2 4. Leangella 60-0 780-2 5. Lingula 400 285-6 6. Hyattidina 400 184-4 7. ‘ Clorinda ’ dormitzeri 40-0 49-8 792 PALAEONTOLOGY, VOLUME 17 table 10 ( cont .): OTHER SPECIES PRESENCE PERCENTAGE FREQUENCY PRESENCE 8. Bracteoleptaena 200 142-8 9. Strophochonetes 20-0 107-0 10. Aegiria grayi 20-0 83-2 11. Cyrtia 20-0 53-4 12. Nucleospira 20-0 35-6 13. Mesounia 20-0 35-6 14. Craniops 200 35-6 15. Orthids 20-0 32-0 16. Eospirifer 20-0 32-0 17. Dalejina 20-0 25-0 18. Isorthis 20-0 25-0 19. Leptaena affi purpurea 20-0 17-8 Number of collections studied = 5 Homogeneity = 63-89% Species density = 7-2 Distinctness coefficient = 85-7% Coming beyond the Dicoelosia community, where the highest species density (diversity) values are found, the Visbyella community contrasts strongly in having a species density as low as that of the Salopina community. Individual collections generally contain one to three rare species, and only collection DB-C-1 is anomalous, with seven. Distribution of individual species The ecological distribution of each species in the complete brachiopod fauna is summarized in Table 1 1 (see also text-fig. 3). The fidelity columns show that most species occur in two or more communities. However, even the most tolerant species (those occurring in four or five communities) are usually very rare in one of their communities of occurrence and can be looked on as occurring there accidentally. table 11. Ecological distribution of individual species. WENLOCK LUDLOW SPECIES COMMUNITY IN FIDELITY COMMUNITY IN FIDELITY WHICH MODAL WHICH MODAL Lingula Visbyella 4 Sphaerirhynchia 4 Craniops Isorthis 5 Dicoelosia 4 Dinobolus Isorthis 1 Schizocrania Salopina 1 Orbiculoidea Dicoelosia 3 Salopina 2 Schizotreta Salopina 1 Salopina 2 Dolerorthis spp. Dicoelosia 1 lEridorthis Dicoelosia 1 Orthids Dicoelosia 3 Dicoelosia 1 Skenidioides Dicoelosia 2 Dicoelosia 2 Salopina spp. Salopina 4 Salopina 4 Isorthis spp. Isorthis 4 Isorthis 4 Resserella canalis Isorthis 3 Dicoelosia 1 Resserella sabrinae Dicoelosia 1 cf. Visbyella trewerna Visbyella 2 Visbyella 2 Dicoelosia biloba Dicoelosia 1 Dicoelosia 2 Dalejina Dicoelosia 4 Isorthis 2 CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 793 TABLE 11 ( cont .): SPECIES Marklandella Cordatomyonia edgelliana Streptis Dictyonella Mesounia Leangella Eoplectodonta spp. Aegiria grayi Katastrophomena Leptaena spp. Leptaena aff. purpurea Bracteoleptaena Leptostrophia compressa Leptostrophia filosa Brachyprion Protomegastrophia Mclearnites Amphistrophia spp. A. ( Pembrostrophia ) Mesopholidostrophia spp. Strophonella Shaleria Coolinia Strophochonetes Protochonetes spp. Protochonetes minimus Shagamella Anastrophia Conchidium Gypidula Clorinda sp. lClorinda ’ dormitzeri ‘ Camarotoechm llandoveriana ‘ Camarotoechm nucula 'Camarotoechm tripartita Trigonirhynchia Rhynchotreta cuneata Sphaerirhynchia wilsoni Sphaerirhynchia davidsoni Rhynchonellids Atrypina Plectatrypa Spirigerina Atrypa Lissatrypa Glassia Nanospira Dayia Eocoelia angelini Homoeospira E WENLOCK COMMUNITY IN FIDELITY WHICH MODAL Salopina 2 Dicoelosia 3 Dicoelosia 1 Dicoelosia 2 Visbyella 2 Visbyella 3 Isorthis 2 Visbyella 2 Isorthis I Isorthis 3 Dicoelosia 3 Visbyella 1 Isorthis 1 Homoeospira 4 Homoeospira 2 Homoeospira 2 Salopina 2 Isorthis 4 Salopina 2 Dicoelosia 2 Isorthis 2 Isorthis 4 Salopina 4 Salopina 4 Dicoelosia 4 Dicoelosia 3 Isorthis 2 Homoeospira 4 Isorthis 2 Visbyella 1 Salopina 1 Salopina 4 Salopina 1 Isorthis 2 Salopina 2 Homoeospira 4 Salopina 2 Salopina 1 Dicoelosia 2 Dicoelosia 1 Isorthis 1 Homoeospira 4 Dicoelosia 1 Visbyella 2 Salopina 3 Homoeospira 3 LUDLOW COMMUNITY IN FIDELITY WHICH MODAL Dicoelosia 1 Visbyella 3 Dicoelosia 1 Isorthis 4 Dicoelosia 1 Sphaerirhynchia 4 Dicoelosia 3 Isorthis 3 Isorthis 3 Isorthis 2 Isorthis 2 Isorthis 3 Salopina 3 Dicoelosia 4 Isorthis 3 Isorthis 1 Dicoelosia 3 Visbyella 1 Salopina 4 Salopina I Sphaerirhynchia 2 Dicoelosia 1 Sphaerirhynchia 4 Isorthis 4 Isorthis 1 Isorthis 3 Dicoelosia 1 Sphaerirhynchia 3 Isorthis 2 794 PALAEONTOLOGY, VOLUME 17 table 11 ( cont .): WENLOCK LUDLOW SPECIES COMMUNITY IN WHICH MODAL FIDELITY COMMUNITY IN WHICH MODAL FIDELITY Meristina Isorthis 4 Isorthis 1 Hyattidina Visbyella 4 Visbyella 3 Whitfieldella Isorthis 4 Sphaerirhynchia 4 Nucleospira Dicoelosia 4 Dicoelosia 2 Athyrids Isorthis 3 Dicoelosia 2 Cyrtia Visbyella 3 Dicoelosia 1 Eospirifer Dicoelosia 5 Dicoelosia 2 Striispirifer Isorthis 4 Dicoelosia 1 Howellella spp. Homoeospira 4 Sphaerirhynchia 4 EXPLANATION OF PLATE 106 All specimens are decalcified internal moulds, treated with ammonium chloride. Grid references are in the Appendix, or given below. Ligs. 1, 2. Salopina lunata (J. de C. Sowerby). Wood Green Railway Cutting, May Hill (Grid ref. SO 6943.1664). 1, pedicle valve, x 2. 2, brachial valve, x 2. Lig. 3. Protochonetes ludloviensis Muir-Wood. Pedicle valve, x2, locality as Pig. 1. Pig. 4. ‘ Camarotoechia nucula (J. de C. Sowerby). Brachial valve, x 2\, locality as Fig. 1. Figs. 5, 6. Homoeospira cf. H. baylei (Davidson). Coll. LD-S-4. 5, pedicle valve, BC 33721, x 2\. 6, brachial valve, BC 33748, x 2\. Fig. 7. Howellella sp. Brachial valve BC 33852, x 2\, coll. LD-S-4. Fig. 8. Dayia navicula (J. de C. Sowerby). Pedicle valve, x 2, Downton, near Ludlow (Grid ref. SO 43 1 1 .7314). Fig. 9. Bivalves and gastropods in the Salopina community, x 1, coll. FE-3. Pigs. 10, 11. Sphaerirhynchia wilsoni (J . Sowerby). 10, pedicle valve, BC 2 1268, x 2, coll. SG-21. 11, brachial valve, x 2, Sawdde Gorge (Grid ref. SN 7250.2482). Fig. 12. Leptostrophia filosa (J. de C. Sowerby). Pedicle valve, x 1^, Sawdde Gorge (Grid ref. SN 7250.2482). Fig. 13. Atrypa reticularis (Linnaeus). Pedicle valve, BC 31840, x 1, coll. LD-S-2. Figs. 14, 15. Isorthis orbicularis (I . deC. Sowerby). 14, pedicle valve, BC 25776, x 2, coll. Usk4. 15, brachial valve, BC 25209, x 2, coll. Usk 5. Fig. 16. Mesopholidostrophia sp. Pedicle valve, x2, Elton Beds, Ludlow. Fig. 17. Eospirifer radiatus (J. de C. Sowerby). Pedicle valve, BC 30618, x 1^, coll. LD-S-16. Fig. 18. Dalejina hybrida (J. de C. Sowerby). Pedicle valve, BC 20684, x 1^, coll. SG-16. Fig. 19. Dicoelosia biloba (Linnaeus). Pedicle valve, x 3, Elton Beds, Ludlow. Fig. 20. Streptis grayii (Davidson). Pedicle valve, x 2, Kilbride Peninsula, Co. Mayo, Ireland. Fig. 21. Eoplectodonta duvalii (Davidson). Brachial valve, BC 30416, x H, coll. LD-S-1 1. Fig. 22. Gypidula galeata (Dalman). Pedicle valve, BC 28106, x 2, coll. Lud 10. Fig. 23. Amphistrophia funiculata (M’Coy). Pedicle valve, BC 30995, x 2, coll. LD-S-17. Fig. 24. Lingula sp. BC 42065, x2, coll. 69-A. Fig. 25. Dolerorthis sp. Brachial valve, x H, Kilbride Peninsula, Co. Mayo, Ireland. Fig. 26. Skenidioides lewisii (Davidson). Brachial valve, x 3, Elton Beds, Ludlow. Fig. 27. Leangella segmentum (Lindstrom). Pedicle valve, BC 30263, x 2}, coll. LD-S-14. Fig. 28. Nucleospira pisum (J. de C. Sowerby). Pedicle valve, x2^, coll. W-N-2. Figs. 29, 30. Cf. Visbyella trewerna Bassett. 29, pedicle valve, BC 36015, x 3, coll. PS-N-1. 30, brachial valve, x 3, coll. B.L.II. Fig. 31. Protochonetes minimus (J. de C. Sowerby). Pedicle valve, BC 24649, x 4, coll. Usk 2. Fig. 32. ‘ Clorinda ’ dormitzeri (Barrande). Brachial valve, GSM DT6061, x 3, North Wales. Negative supplied by Dr. M. G. Bassett. Fig. 33. Glassia sp. Brachial valve, BC 30450, x 2-f coll. LD-S-11. PLATE 106 CALEF and HANCOCK, Silurian brachiopods 796 PALAEONTOLOGY, VOLUME 17 Those species which are confined to one community are usually exceedingly rare, being represented in the total collections by only a few specimens. For example, there are only two specimens of Schizocrania out of a total fauna of more than 20,000 specimens. Other associations A few collections do not easily fit into any of the main communities. They all have a high dominance of one or two species, ones which are otherwise usually quite rare. They may be highly distinctive assemblages living in atypical and rare environments; alternatively they could represent clusters of rare species, or population explosions of opportunistic species (Levinton 1970). A single collection from the basal Downtonian of May Hill contains 98% Lingula (MH-1 1, see Appendix), and faunas with abundant Lingula occur in the Platyschisma Beds low in the Downtonian at Knighton (Holland 1959) and Ludlow (Holland 1962). These rare assemblages resemble the Llandovery Lingula community (Ziegler et al. 1968) and particularly the ‘restricted’ Lingula community (Ziegler et al. 1969) which contains Lingula alone. Assemblages dominated by bivalves have been obtained at several localities in the Wenlock (see Appendix) and about a metre above the Ludlow Bone Bed at Ludlow. Characteristic genera for this association are Aetinodonta , Nuculites, large pterioids, and modiomorphaceans. Nucula and Grammysia also occur. Brachiopods charac- teristic of the Salopina community may be present, particularly ‘ Camarotoechia ' nucula , Salopina , and Lingula. In contrast to the main communities described above, the Bivalve association is predominantly infaunal, with free-burrowing species and endobyssate forms (Stanley 1972). It, too, resembles some developments of the Llandovery Lingula community. Three Wenlock collections have been grouped as the Resserella association which is dominated by R. whitfieldensis (Bassett 1972, p. 50). The other species in the collections suggest that this association is most closely related to the Homoeospira community. In the Ludlow a Dayia- dominated association shares most of its genera with the Sphaerirhynchia community. Virtually monospecific assemblages of Dayia navicula occur through large sedimentary thicknesses at Builth Wells, which suggests either that this species could colonize areas of the sea-floor inimical to other brachiopods of the Sphaerirhynchia community, or that its presence in dense concentrations excluded other species populations. Thus the Dayia association is identified by having a limited number of species rather than by species peculiar to the association. The Gypidula/ Atrypa association is dominated equally by Gypidula galeata and Atrypa reticularis. The assemblage is found more frequently in limestones than clastic sediments. The group of species associated with Gypidula and Atrypa varies, so in different cases the assemblage most closely resembles Isorthis, Homoeospira/ Sphaerirhynchia, or Salopina communities. Relations between communities The individualistic hypothesis of the community proposed by Gleason (1926) holds that communities are combinations of organisms which, in responding to CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 797 similar ecological requirements, happen to occur together. The theory stresses the individual response of the organisms to the environment. Their interaction is of secondary importance. The applicability of this idea to the Silurian is indicated by the variability (low homogeneity) of the communities and also by their continuously intergrading nature. The primary influence of the environment in determining the structure of an association leads to a corollary of the individualistic hypothesis: the concept of the continuum. Along an environmental gradient species composition changes until, by degrees, one community is replaced by another. The rate of compositional change is proportional to the steepness of the gradient. The compositional and geographical boundaries dividing communities are arbitrary since no natural hiatus exists. The continuum is demonstrated in text-fig. 3 where a selection of species is plotted to show their frequency presence values through the full community spectrum or environ- mental gradient for both the Wenlock and the Ludlow. With one exception, all the curves are unimodal and show the gradual changes from one community to the next. Another striking feature of text-fig. 3 is the similarity of the Wenlock and Ludlow curves for many species. The continuum is also expressed by the similarity coefficient (see Table 1) between communities. Text-fig. 4 shows the value of the similarity coefficient between each community and the Salopina community in the Wenlock and the Ludlow. The pro- gressive fall in value towards the Visbyella community reflects the decreasing number of species in common, and the diverging values of frequency presence of those species which remain. ENVIRONMENTAL INTERPRETATION Studies on Recent communities (Thorson 1957) have repeatedly shown a relation between depth of water and community. Studies in Lower Palaeozoic fossil com- munities point to the same conclusion (Ziegler 1965; Bayer 1967; Seilacher 1967; Bretsky 1969). In this section we present evidence indicating that Wenlock and Ludlow communities occupied progressively deeper marine environments from the Salopina community the shallowest, through Homoeospiraj Sphaerirhynchia and Isorthis, to the deeper Dicoelosia community. The ecology of the Visbyella com- munity is discussed briefly in a later section. Apart from the normal correlation of lithology with depth (i.e. little coarse sedi- ment in deep water), no good correlation has been seen between sediment type and community within the clastic facies covered by this paper. Temperature control has been put forward by Berry and Boucot (1967), but temperatures were regarded by them as normally being depth related. There is no independent evidence to enable the determination of palaeo-temperatures. Analogy with previously studied communities Ziegler, Cocks and Bambach (1968) defined five depth-related communities in the Llandovery, in order of increasing depth : (1) Lingula community, (2) Eocoelia com- munity, (3) Pentamerus community, (4) Stricklandia community, and (5) Clorinda community. A still deeper (6) "Marginal’ Clorinda community has been described by Frequency Presence 2000 1000 0 2000 1000 0 600 300 0 3000 1500 0 2000 1000 0 2000 1000 0 2000 1000 0 Sa Ho/Sp Is Di Sa Ho/Sp Is Di V Sa Ho/Sp Is Di V Salopina \r Leptostrophia Amphistrophia |\ 2000 Isorthis 1000 • 0 2000 r 1000 ■ ✓ |\ 0L 600 ■ Mesopholidostrophia\ 300 0 200 Dalejina |\ Protochonetes sp "C." nuc ula Atrypa ' ' / 7 Howelle lla 100 0 2000r 1000 - 0 2000 r 1000 • o 500 250 0L N Gypidula Dicoelosia 2000 1000 0 2000 1000 - 0 - 600 300 1000 cf Visbyella l\\ I I \ Coolima /N 7 ' 500 - 0L 500 Protochonetes minimus/ Sphaerirhynchia Homoeospira Js. Striispirifer / \ \ i 250 • 0 200 100 0 L 500 250 Cyrtia /i\ Eospiriter I Sa Ho/Sp Is Di V Sa Ho/Sp Is Di V Sa Ho/Sp Is Di V Community text-fig. 3. Abundance of selected genera in the communities for the Wenlock (dashed lines) and Ludlow (solid lines). Sa — Salopina community, Ho/Sp—Homoeospira/Sphaerirhynchia community, Is — Isorthis community, Di— Dicoelosia community, V —Visbyella community. CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 799 text-fig. 4. Similarity of each community to the Salopina community as calculated with the Similarity Coefficient (Table 1). Dashed line for the Wenlock; solid line for the Ludlow. Cocks and Rickards ( 1 969). Many genera of brachiopod living during the Llandovery survived into the Wenlock and Ludlow and thus form links between Llandovery and later Silurian communities. Table 12, which lists the modal communities of some of these genera in the upper Llandovery, Wenlock, and Ludlow, shows that the Dicoelosia community is the approximate later Silurian equivalent of the Clorinda community, Isorthis of the Stricklandia, Homoeospira/ Sphaerirhynchia of the Pentamerus, and Salopina of the Eocoelia community. If we assume that similar ecological controls applied in the Wenlock and Ludlow to those in the Llandovery, then it follows that Wenlock and Ludlow communities were also depth related. Sedimentary and stratigraphic relationships At the Sawdde Gorge, the Llandovery exhibits a deepening from the Eocoelia com- munity to the Costistricklandia community. The Wenlock (text-fig. 5) begins with the Dicoelosia community and then shows the following sequence of communities: Dicoelosia-Isorthis-Homoeospira/ Sphaerirhynchia- Salopina. The Salopina com- munity occurs in sediments showing flaser-bedding, herring-bone cross-bedding, and other extremely shallow-water features, and these are interbedded with silts contain- ing the Bivalve association. The Wenlock succession thus records a progressive shallowing. An abrupt change occurs from the topmost Wenlock Salopina community 800 PALAEONTOLOGY, VOLUME 17 table 12. Modal communities of brachiopod genera in the upper Llandovery, Wenlock, and Ludlow. In the upper Llandovery where two communities are given, the first applies to CUj, the second to C4_6. Abbrevia- tions as follows: E—Eocoelia community, E— Pent aments community, St —Stricklandia community, C —Clorinda community, Sa —Salopina community, H —Homoeospira community, Sp —Sphaerirhynchia community, I — Isorthis community, D — Dicoelosia community, V Visbyella community. Upper Genus Llandovery Wenlock Ludlow Salopina E Sa Sa Protochonetes E Sa Sa ‘ Camarotoechia ' E Sa Sa Howellella E H Sp Leptostrophia E H Sp Atrypa P-St H I Coolinia P-C I I Mesopholidostrophia St D I Isorthis St I I Strophonella c I I Dicoelosia c D D Resserella c D D Skenidioides c D D Leangella c V D Cyrtia c V D Aegiria c V V to the basal Ludlow Dicoelosia community, accompanied by an equally sharp change from shallow-water sediments to deep, quiet-water silts and muds. The Ludlow sequence repeats the communities in the same order as in the Wenlock, suggesting a second regression (text-fig. 5). The Salopina community is developed in the highest marine beds prior to the continental Trichrug Beds, the precursor of the Old Red Sandstone (Potter and Price 1965). Similar relations can be seen at Usk (text-fig. 5), with two regressive sequences, the lower one culminating in the wave- and current-rippled sandstone directly below the Wenlock Limestone and the upper one continuing through the Ludlow into the Old Red Sandstone. A single collection (U-C-l) immediately above the Wenlock Lime- stone shows the Isorthis community, and records an intermediate stage in the post- Wenlock transgression. The community succession and the available sedimentary and stratigraphic evidence show the same double regression throughout the Welsh Borderland in each strati- graphic section which has been examined. We have found no evidence of widespread cyclic transgressions and regressions such as those postulated by Phipps and Reeve (1967, fig. 6) for the Malvern Hills area. Shallowing at the end of the Wenlock is further shown by the green algae in the Wenlock Limestone, which are believed by Scotfin (1971) to represent water less than 30 m deep. The Ludlow commences every- where with the much deeper quiet-water muds of the Eltonian, containing the Dicoelosia community. The sequence of communities during the Ludlow shallowing is modified at Ludlow, where the Sphaerirhynchia community is represented by the Dayia association. At the top of the Ludlow and in the basal Downton, the Bivalve association, along with Lingula, occurs frequently between the Salopina community and the CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 801 SAWDDE GORGE USK OLD RED SANDSTONE O -J Q ID USK I USK 4 USK 5 USK 6 USK 3 USK 2 UC I WENLOCK LST ^ UP I y MB 024(18) =J MB 02409) 5 MB 024(30) ^ MB 024(32) Abbreviations SAL Salopina Community SPH Sphaerirhynchia Community HOM Homoeospira Community ISO Isorthis Community DIC Dicoelosia Community BIV Bivalve Association GYP Gypidula /Atrypa Association B+L Bivalves 8t Lingula text-fig. 5. Stratigraphic sequence of collections and their community designations in the Sawdde and Usk sections. The base of the Wenlock Shale is not exposed at Usk, and while there are Ludlow rocks above the Trichrug Beds at Sawdde, no collections were made from them. fluviatile environments of the Old Red Sandstone. Bivalves thus dominated areas shoreward of the Salopina community, where sedimentary structures indicate a similar, extremely shallow, depth of water (Allen and Tarlo 1963; Sanzen-Baker 1972). The double regression cannot be seen in Pembrokeshire, because the Old Red Sandstone appears to have arrived early at Marloes, perhaps even in Wenlock times 802 PALAEONTOLOGY, VOLUME 17 (Sanzen-Baker 1972), and the Ludlow shallowing is thus not recorded there. The upper Ludlow in Denbighshire has been removed by erosion, though evidence for a Ludlow shallowing comes from the Salopina community preserved in pebbles in the basal Carboniferous (Strahan and Walker 1879). Possible causes for the depth correlation Stability and predictability. In modern benthonic communities species diversity increases with depth of water as shown by soft-bottom samples collected along a tran- sect between Gayhead and Bermuda (Sanders and Hessler 1 969). The diversity gradient is attributable to the increasing stability and predictability of the environment (Slo- bodkin and Sanders 1969). A similar gradient is present in our communities. Sanders’s (1968) method for diversity comparisons has been used, treating similar habitats, namely soft, clastic, usually fine-grained, level bottoms of varying depth, and using the brachiopod-bivalve fraction of the collections. The rarefaction tech- nique of Sanders (1968) has been used to avoid the problem of the dependence of the number of species in a collection on the collection size. Rarefaction is used to derive ‘expected’ values of species diversity at a variety of reduced collection sizes, retain- ing the relative proportions between species in the original collection. The values form a ‘rarefaction curve’ (e.g. text-fig. 6). The diversity of different-sized collections is obtained from rarefaction values at a collection size common to each. Text-fig. 6 shows that most curves reach the sixty individual point so this rarefied size has been used for comparisons. Table 13 lists the diversity values of the collections available at the time of rare- faction computation. The average diversity values for each community form a pro- gression from the low diversity Salopina community to the very diverse Dicoelosia community. This feature of the depth gradient probably reflects a gradual increase in environmental stability and predictability. Food supply. We have observed a rough gradient in fossil density in our collections. Fossils are very sparse in the deep-water Dicoelosia community collections despite the concentrating effect of a continuously slow sedimentation rate in that text-fig. 6. Rarefaction curves for May Hill Ludlow collections. Curve numbers refer to MH collections. CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 803 table 13. Diversity of collections grouped into communities, using the 60-individual size derived from the rarefaction method. Values marked * represent collections of less than 60 individuals and have been obtained by extrapolation. Salopina Homoeospira / Isorthis Dicoelosia community Sphaerirhynchia community community community Lud 5 5-043 MH-3 16-818 Lud 1 7-513 69-A 17-609 Lud 6 7-546 MH-4 9-364 Lud 2 14-000* MH-1 24-613 Lud 11 5-579 MH-8 12-511 Lud 7 11-075 MH-9 13-753 MH-5 14-636 Usk 5 11-509 Lud 8 10-502 Usk 2 1 1 -667 MH-6 11-889 Usk 4 13-358 EH-1 6-278 LD-S-15 15-352 MH-10 10-615 LD-S-4 7-326 EH-2 17-330 LD-S-13 13-651 BW-4 8-380 LD-S-3 11-304 Usk 3 16-308 LD-S-14 13-000* Usk 1 15-263 LD-S-5 20-497 LD-S-16 20-633 LD-S-21 12-579 LD-S-2 10-774 LD-S-8 1 1 -692 LD-S-18 9-034 SG-8 13-000* SG-5 16-286 SG-15 6-348 SG-21 6-628 SG-16 11-333 SG-19 7-586 SG-11 7-316 SG-20 8-000* SG-10 6-623 SG-17 11-134 SG-22 11-091 SG-7 6-200 FE-1 8-478 FE-2 6-000* FE-3 14-500* FW-I 5-937 FW-3 6-700* WT-X 8-975 WT-A5 3-000* Mean 8-848 11-700 12-020 15-664 Standard deviation 3-416 4-037 4-392 4-373 environment. By contrast the Salopina community is notable for its rich though undiverse fauna. Similar density gradients are observed in modern oceans: Sanders and Hessler (1969) found the abundance of benthonic animals to diminish from 13 000-23 000/m2 on the shelf edge to about 500/m2 at the base of the continental slope. This decrease in density reflects the decrease in food supply with increasing depth. The deeper the water column, the less food eventually reaches the bottom owing to scavenging and bacterial degradation during settling, so that deep-sea areas are impoverished in available nutrients (Marshall 1954). We believe food supply to have been the most important single controlling factor of upper Silurian brachiopod distribution. This hypothesis can be examined by measuring the size of Isorthis populations, which tolerated a wide depth range. At any one locality we find gradations between all the Isorthis present. Given the low food supply in deep water, individuals should have been unable to reach the same size as those of the same species inhabiting shallow, nutrient-rich waters. The widths of all Isorthis brachial valves in all collections were measured. The average width for each collection was then calculated, the collec- tions were grouped into communities and the mean width for each community was 804 PALAEONTOLOGY, VOLUME 17 WIDTH (mm) text-fig. 7. Average size of Isorthis in Ludlow communities. Vertical line = average of collection means for each community. Horizontal line = observed range of means. Only one collection in the Salopina community. determined. There is a correlation between community and size of Isorthis , the smaller sizes occurring in deep-water communities, and the larger in shallow water (text-fig. 7). The presence in the Dicoelosia community of many extremely tiny taxa further supports the food hypothesis. Low food requirements, and hence small adult size, may be strongly selected for in deep environments. Limited food supply affects the biomass of organisms in the community as a whole. Biomass is determined by both density and size, and these diminish together into the Dicoelosia and Visbyella communities; size may be particularly important because the volume decreases much faster (as the cube) of the width. The food supply contrast between shallow and deep communities may have been considerable. Other controlling variables While food may determine the lower depth range of a species, the limitations on the upper range are likely to involve the instability of the near-shore region. Shallow- water species are adapted to wide temperature, salinity, and turbidity fluctuations which characterize the shallow, near-shore region. A deep-water species penetrating into the shallows lacks such adaptations and is hence unlikely to survive. CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 805 The faunas of this study are benthonic and therefore the substrate is a variable with potentially powerful ecological effects. This is commonly the case in modern faunas (Purdy 1964), but the effects are less profound in lower Palaeozoic brachiopod communities. With the exception of their rocky bottom community, Ziegler, Cocks and Bambach (1968) found little correlation between sediment type and community, and the same is true of this study. The probable reason for this independence is the epifaunal nature of almost all elements of the Silurian communities, as contrasted with the dominantly infaunal modern level bottom associations. As epifaunal filter- feeders, brachiopods are much less dependent on the substrate than animals which live and feed in the sediment. Ecology of the Visbyella community The Visbyella community has not been treated in the preceding discussion: it is not yet well known from the Ludlow, and it is discussed by Hancock, Hurst and Fursich (1974). Some ecological conclusions, however, are given here. The Visbyella community probably lived seaward of the Dicoelosia community because it appears prior to the Dicoelosia community in some shallowing sequences. In addition, graptolites and other pelagic groups are often preserved with the community. The very low population density is consistent with a deep-water environment and follows logically from the Dicoelosia community. Extrapolations of rarefaction curves sug- gest the Visbyella community may be much less diverse than the Dicoelosia com- munity. Low species diversity might reflect stress conditions on the bottom, such as low oxygen concentrations (Sanders 1969; Rhoads and Morse 1971), but probably means that only a few species were adapted to the conditions of the Visbyella com- munity. At the present time, significant diversity reductions take place in lophophorate groups at considerable depth (Jorgensen 1966; Ryland 1970). The Visbyella com- munity has some resemblance to the depleted ‘Marginal’ Clorinda community in the Llandovery (Cocks and Rickards 1969), but differs in having its own characteristic species. CONCLUSIONS Using a statistical technique which has taken into account all our collections, this paper has described five major Silurian communities. These are closely comparable to the communities already described in the early Silurian, our Salopina community paralleling the Eocoelia community, and so on through to the Visbyella community, which is equivalent in position to the ‘Marginal’ Clorinda community. All our evidence indicates that these communities are correlated with depth. Their increasing diversity and decreasing density from Salopina to Dicoelosia agree with the depth- dependent gradients found in modern benthonic assemblages. The Visbyella com- munity, with its sparse fauna, probably represents the limit of Silurian benthonic life. The communities have been shown to reflect an environmental continuum with no natural breaks, and are of the type described by Johnson (1964, p. 107) as ‘asso- ciations of largely independent species . . . with similar responses to the physical environment’. This is not surprising since the communities are dominated by brachiopods whose ecological requirements as suspension feeders make them more 806 PALAEONTOLOGY, VOLUME 17 or less independent of other living animals. Food supply is believed to have been the most important factor controlling their distribution. In addition to the major communities, we have described several other associations represented by only a few collections. Some of these may result from population explosions of opportunistic species, while others may be assemblages which lived in atypical environments. It is likely that more of these faunas remain to be discovered, but they will be quantitatively much less significant than the main marine communities. Acknowledgements. We thank Dr. W. S. McKerrow for his advice and for supervising our projects in the Department of Geology and Mineralogy, Oxford, on which this paper is based, and Drs. R. K. Bambach and A. M. Ziegler for providing helpful suggestions. We are grateful to Dr. M. G. Bassett, Dr. A. M. Ziegler and Mr. J. M. Hurst for making available unpublished collection data and specimens, and to Dr. D. E. White of the Institute of Geological Sciences for access to material in his care. Thanks are also due to the Institute of Geological Sciences for providing locality information for North Wales; to the technical staff of the Department of Geology for much assistance with labelling the collections and to Robertson Research Laboratories for typing facilities. Hancock also acknowledges financial support from the Natural Environment Research Council and St. John’s College, Oxford. REFERENCES allen, J. R. l. and tarlo, l. b. 1963. The Downtonian and Dittonian facies of the Welsh Borderland. Geo l. Mag. 100, 129-155. bassett, m. G. 1970-1972. The articulate brachiopods from the Wenlock Series of the Welsh Borderland and South Wales. Palaeontogr. Soc. [Monogr.] (1-2), 1-78. — 1971. Wenlock Stropheodontidae (Silurian Brachiopoda) from the Welsh Borderland and south Wales. Palaeontology , 14, 303-337. bayer, T. N. 1967. Repetitive benthonic community in the Maquoketa Formation (Ordovician) of Minne- sota. J. Paleont. 41, 417-422. berry, w. b. n. and boucot, a. j. 1967. Pelecypod-graptolite association in the Old World Silurian. Bull, geol. Soc. Am. 78, 1515-1522. bretsky, p. w. 1969. Central Appalachian Late Ordovician communities. Ibid. 80, 193-212. cocks, l. r. m. and rickards, r. b. 1969 (for 1968). Five boreholes in Shropshire and the relationships of shelly and graptolitic facies in the Lower Silurian. Q. Jl geol. Soc. Lond. 124, 213-238. craig, G. Y. 1954. The palaeoecology of the Top Hosie Shale (Lower Carboniferous) at a locality near Kilsyth. Ibid. 110, 103 119. curtis, j. t. 1959. The Vegetation of Wisconsin. University of Wisconsin Press, Madison. 657 pp. gleason, H. A. 1926. The individualistic concept of the plant association. Bull. Torrey Bot. Club , 53, 7-26. Hancock, n. j., hurst, J. m. and fursich, f. t. 1974. The depths inhabited by Silurian brachiopod com- munities. Jl geol. Soc. Lond. 130, 151-156. Holland, c. h. 1959. The Ludlovian and Downtonian rocks of the Knighton district, Radnorshire. Q. Jl geol. Soc. Lond. 114, 449-482. — 1962. The Ludlovian-Downtonian succession in Central Wales and the Central Welsh borderland. Symposiums-band der 2. internationalen Arbeitstagung iiber die Silur/Devon-Grenze und die Stratigraphie von Silur und Devon. Bonn- Bruxelles I960. Stuttgart, 87-94. Johnson, r. g. 1962. Interspecific associations in Pennsylvanian fossil assemblages. J. Geol. 70, 32-55. — 1964. The community approach to palaeoecology. In imbrie, j. and Newell, n. d. (eds.). Approaches to Paleoeco/ogy, 107-134. John Wiley and Sons Inc., New York, London, Sydney. J0RGENSEN, c. b. 1966. Biology of suspension feeding. Pergamon, Oxford. levinton, j. s. 1970. The palaeoecological significance of opportunistic species. Lethaia, 3, 69-78. marshall, n. b. 1954. Aspects of Deep Sea Biology. Hutchinson, London. 380 pp. phipps, c. b. and reeve, f. a. e. 1967. Stratigraphy and geological history of the Malvern, Abberley and Ledbury Hills. Geol. Jnl, 5, 339-368. CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 807 potter, j. F. and price, J. H. 1965. Comparative sections through rocks of Ludlovian Downtonian age in the Llandovery and Llandeilo Districts. Proc. Geol. Assoc. Lond. 76, 379-401. purdy, e. G. 1964. Sediments as substrates. In imbrie, j. and Newell, n. d. (eds.). Approaches to Paleoecology, 238-271. John Wiley and Sons Inc., New York, London, Sydney. rhoads, d. c. and morse, j. w. 1971. Evolutionary and ecologic significance of oxygen-deficient marine basins. Lethaia , 4, 413-428. ryland, j. s. 1970. Bryozoans. Hutchinson University Library, London. Sanders, H. L. 1968. Benthic marine diversity: a comparative study. Amer. Naturalist, 102, 243-282. — 1969. Benthic marine diversity and the stability-time hypothesis. In woodwell, g. m. and smith, h. h. (eds.). Diversity and Stability in Ecological Systems, 71-81. Brookhaven Sympos. Biol. 22, Upton, New York. — and hessler, r. r. 1969. Ecology of the deep-sea benthos. Science, 163, 1419-1424. sanzen-baker, i. 1972. Stratigraphical relationships and sedimentary environments of the Silurian-Old Red Sandstone of Pembrokeshire. Proc. Geol. Assoc. Lond. 83, 139-164. scoffin, t. p. 1971. The conditions of growth of the Wenlock reefs of Shropshire (England). Sedimentology , 17, 173-219. seilacher, a. 1967. Bathymetry of trace fossils. Marine Geol. 5, 413-428. slobodkin, l. b. and sanders, H. l. 1969. On the contribution of environmental predictability to species diversity. In woodwell, g. m. and smith, h. h. (eds.). Diversity and Stability in Ecological Systems, 82-95. Brookhaven Sympos. Biol. 22, Upton, New York. Stanley, s. m. 1972. Functional morphology and evolution of byssally attached bivalve mollusks. J. Paleont. 46, 165-212. strahan, a. and walker, a. o. 1 879. On the occurrence of Upper Ludlow fossils in the Lower Carboniferous conglomerates of North Wales. Q. J! geol. Soc. Lond. 35, 268-274. thorson, G. 1957. Bottom communities (sublittoral or shallow shelf). In hedgpeth, j. w. fed.). Treatise on Marine Ecology and Paleoecology. Mem. geol. Soc. Am. 67, 461-534. walker, k. r. and bambach, r. k. 1971. The significance of fossil assemblages from fine-grained sediments: time averaged communities. Abstr. of meetings, Geol. Soc. Amer. 3, 783-784. ziegler, a. m. 1965. Silurian marine communities and their environmental significance. Nature Lond. 207, 270-272. — cocks, l. r. m. and bambach, r. k. 1968. The composition and structure of Lower Silurian marine communities. Lethaia, 1, 1-27. — mckerrow, w. s., burne, r. v. and baker, p. e. 1969. Correlation and Environmental Setting of the Skomer Volcanic Group, Pembrokeshire. Proc. Geol. Assoc. Lond. 80, 409-439. c. E. CALEF Brookhaven National Laboratories Associated Universities Inc. Upton Long Island New York 1 1973 U.S.A. N. J. HANCOCK Robertson Research Laboratories Ty'n-y-coed Llanrhos Llandudno North Wales Typescript received 16 May 1973 Revised typescript received 23 January 1974 808 PALAEONTOLOGY, VOLUME 17 APPENDIX Localities of collections Area Collection Grid reference Formation WENLOCK Salopina community Pembrokeshire FE-3 SS 0165.9753 Silurian of Freshwater East FW-3 SS 8840.9940 Silurian of Freshwater West FW-1 SS 8840.9940 Silurian of Freshwater West 3-doors SM 7622.0827 Sandstone ‘Series’ WT-X SM 7670.0790 Sandstone ‘Series’ WT-A5 SM 7626.0828 Sandstone ‘Series’ P-SP-3 SM 8628.0876 Winsle ‘Series’ P-W-2 SM 8212.0918 Winsle ‘Series’ Sawdde Gorge, Carmarthenshire LD-S-18 SN 7221.2507 ‘Upper Wenlock’ LD-S-21 SN 7219.2509 ‘Upper Wenlock’ Usk U-P-l ST 3480.9990 Wenlock Shale *MB 66 (18) SO 3355.0096 Wenlock Shale East Mendips ND-M-3 ST 6638.4573 Wenlock Shale ND-M-4 ST 6632.4576 Wenlock Shale fND-T-1 ST 6754.4514 Wenlock Shale tND-AD-1 ST 6768.4585 Wenlock Shale Tortworth JT-W-B ST 6947.9376 Brinkmarsh Beds Homoeospira community Pembrokeshire FE-1 SS 0165.9753 Silurian of Freshwater East P-SP-1 SM 8626.0874 Winsle ‘Series’ P-W-l SM 8212.0918 Winsle ‘Series’ P-WD-1 SM 8332.0931 Winsle ‘Series’ Sawdde Gorge, Carmarthenshire LD-S-5 SN 7217.2511 ‘Upper Wenlock’ LD-S-4 SN 7218.2510 ‘Upper Wenlock’ LD-S-3 SN 7218.2510 ‘Upper Wenlock’ LD-S-2 SN 7218.2510 ‘Upper Wenlock’ Usk *MB 66(19) SO 3355.0097 Wenlock Shale Tortworth T-BK-1 ST 6672.9068 Brinkmarsh Beds JT-Z-A ST 6672.9068 Brinkmarsh Beds Isorthis community Sawdde Gorge, Carmarthenshire LD-S-16 SN 7200.2533 ‘Upper Wenlock’ LD-S-17 SN 7204.2527 ‘Upper Wenlock’ LD-S-8 SN 7209.2522 ‘Upper Wenlock' LD-S-12 SN 7213.2520 ‘Upper Wenlock’ Usk U-C-l SO 3331.0160 Basal Elton Beds *MB 66 (30) SO 3367.0111 Wenlock Shale East Mendips ND-M-1 ST 6646.4570 Wenlock Shale tND-RL-2 ST 6647.4569 Wenlock Shale Dicoelosia community Sawdde Gorge, Carmarthenshire LD-S-14 SN 7173.2563 ‘Fower Wenlock’ LD-S-1 1 SN 7182.2553 ‘Fower Wenlock’ LD-S-15 SN 7192.2542 ‘Upper Wenlock’ Usk *MB 66 (32) SO 3372.0116 Wenlock Shale CALEF AND HANCOCK: WENLOCK AND LUDLOW COMMUNITIES 809 Area Wenlock Edge Woolhope Collection §P.R.C.I WE-H-1 *MB 4 W-B-l W-N-l Grid reference SO 5805.9740 SJ 5924.0045 SJ 6435.0445 SO 6180.3568 SO 5817.3525 Formation Tickwood Beds Buildwas Beds Buildwas Beds Wenlock Shale Wenlock Shale Visbyella community Presteigne Ludlow Denbighshire Resserella association May Hill Tortworth PS-N-1 SO 3045.6245 Wenlock mudstones PS-D-1 SO 2439.5782 Wenlock mudstones §B.L.II SO 4425.7253 Wenlock Limestone §B.L.I SO 4425.7253 Wenlock Shale DB-C-1 SH 8177.6174 Upper Mottled Mudstone JM-G-B SO 7055.2103 Woolhope Limestone +M-O-A SO 6869.2244 Woolhope Limestone T-BR-3 ST 6736.9130 Pycnactis Band Bivalve association Pembrokeshire WT-3 WT-5 Sawdde Gorge, Carmarthenshire LD-S-19 Tortworth T-BR-1 SM 7621.0834 SM 7649.0806 SN 7220.2508 ST 6735.9131 Sandstone ‘Series’ Sandstone ‘Series’ ‘Upper Wenlock’ Brinkmarsh Beds LUDLOW Salopina community Sawdde Gorge, Carmarthenshire SG-19 SN 7263.2477 Black Cock Beds SG-15 SN 7266.2480 Black Cock Beds SG-22 SN 7262.2483 Black Cock Beds SG-10 SN 7253.2482 Black Cock Beds Builth Wells BW-4 SO 0875.4367 Whitcliffian Usk Usk 1 ST 3681.9826 Leintwardinian May Hill MH-10 SO 6930.1866 Upper Longhope Beds MH-6 SO 6943.1664 Lower Longhope Beds MH-5 SO 6943.1664 Lower Longhope Beds Ludlow Lud 5 SO 4377.7358 Whitcliffe Beds Lud 6 SO 4442.7425 Whitcliffe Beds Lud 1 1 SO 4975.7244 Whitcliffe Beds Sphaerirhynchia community Sawdde Gorge, Carmarthenshire SG-ll SN 7254.2481 Black Cock Beds SG-21 SN 7248.2484 Black Cock Beds SG-8 SN 7245.2485 Black Cock Beds Usk Usk 5 SO 3749.0017 Lower Llanbadoc Beds Usk 4 SO 3757.0007 Upper Llanbadoc Beds * Collection in the National Museum of Wales, Cardiff, examined by kind permission of Dr. M. G. Bassett. Locality numbers refer to Bassett’s monograph (1970, pp. 7-11). t Collection (made by S. H. Reynolds) housed in the Geological Survey Museum, kindly made available by Dr. D. E. White, of the Institute of Geological Sciences. { Collection data provided by Dr. A. M. Ziegler. § Collection made by Mr. J. M. Hurst, Oxford. F 810 PALAEONTOLOGY, VOLUME 17 Area Collection Grid reference Formation Woolhope WH-1 SO 5950.3987 Lower Perton Beds May Hill MH-8 SO 6944.1664 Blaisdon Beds MH-3 SO 6944.1663 Blaisdon Beds MH-4 SO 6944.1663 Blaisdon Beds Isorthis community Sawdde Gorge, Carmarthenshire SG-16 SN 7243.2489 Black Cock Beds SG-7 SN 7242.2489 Black Cock Beds SG-17 SN 7240.2490 Black Cock Beds SG-20 SN 7234.2494 Black Cock Beds SG-4 SN 7232.2496 Black Cock Beds Usk Usk 3 ST 3517.9777 Upper Forest Beds Ludlow Lud 1 SO 4289.7296 Bringewood Beds Lud 2 SO 4296.7312 Leintwardine Beds Lud 7 SO 4953.7255 Leintwardine Beds Lud 8 SO 4968.7245 Leintwardine Beds Wenlock Edge EH-1 SO 5705.9500 Lower Elton Beds EH-2 SO 5737.9425 Lower Bringewood Beds EH-3 SO 5737.9425 Upper Bringewood Beds Denbighshire GB-9 SH 8625.6226 Elwy Group Dicoelosia community Sawdde Gorge, Carmarthenshire SG-1 SN 7228.2499 Tresglen Beds SG-5 SN 7237.2492 Tresglen Beds Usk Usk 2 ST 3650.9827 Lower Forest Beds May Hill MH-1 SO 6950.1649 Lower Flaxley Beds MH-9 SO 6944.1859 Upper Flaxley Beds Malverns §C.W.III SO 7605.4413 Elton Beds Ludlow 69-A SO 4389.7278 Lower Elton Beds LEB 3rd SO 4348.7262 Lower Elton Beds Woolhope W-N-2 SO 5815.3516 Lower Wooton Beds Visbyella community Ludlow 3-NFG SO 4337.7263 Middle Elton Beds Dayia association Builth Wells BW-1 SO 0550.4890 Oriostoma Beds BW-2 SO 0930.4670 Lingula lata Beds BW-3 SO 0677.4776 Lingula lata Beds Ludlow Lud 3 SO 4311.7314 Leintwardine Beds Lud 4 SO 4318.7314 Leintwardine Beds Gypidula Atrypa association Sawdde Gorge, Carmarthenshire LD-S-1 SN 7225.2503 ‘Upper Wenlock’ Usk Usk 6 SO 3747.0019 Lower Llanbadoc Beds Ludlow Lud 10 SO 4934.7263 Upper Bringewood Beds Lingula fauna May Hill MH-1 1 SO 6908.1907 Clifford’s Mesne Beds § Collection made by Mr. J. M. Hurst, Oxford. A LOWER CARBONIFEROUS BRACHIOPOD FAUNA FROM THE MANIFOLD VALLEY, STAFFORDSHIRE by c. h. c. brunton and C. CHAMPION Abstract. A silicified brachiopod fauna from North Staffordshire is described and thought to be early Visean. It includes the new taxa Lambdarina manifoldensis and Crurithyris nastus, and other species showing North American Mississippian affinities. The fauna may indicate a relatively shallow-water environment with a low rate of sedimentation. During the course of many years of collecting in the Staffordshire/Derbyshire border area one of us (C. C.) discovered the fauna here described in silicified lime- stones which crop out in the Manifold Valley, close to the village of Wetton (text- fig. 1). The fauna contained several species of unusual aspect as well as a totally new form and the material was sent to C. H. C. B. for comment. The distinctive new species proved to be most closely related to Cardiarina, an Upper Carboniferous genus known only from North America, but differs sufficiently to be separated as a new genus, here called Lambdarina, a name proposed in unpublished work by Dr. P. G. Morris for congeneric specimens he collected a few miles to the south of our localities at Waterhouses, on the Hamps river. (See addendum, p. 840.) The significance of this new species in British Lower Carboniferous rocks and the relationships of some of the other species, including the recognition of stenoscismataceans, add to the importance of this fauna. Classification and nomenclature are, unless otherwise stated, as in the Treatise volumes (1965). The figured specimens are all housed in the British Museum (Natural History). GEOLOGICAL SETTING Bedded limestones, with variable dips, overlaid by obscurely bedded ‘reefal’ lime- stones are exposed on the south-west slope of Wetton Hill, above Leek Road, where it descends to the Manifold river, one-half mile west of Wetton village. Weathering of the lower group has revealed the presence, at certain levels, of well-preserved silicified fossils. These fossils are seen to have restricted vertical ranges, not being found in the ‘reefal’ limestones and only rarely 30 m or more below the ‘reefal’ limestones, where chert and authigenic quartz are frequently found in small amounts. The fossils described below were extracted from the silicified limestones at the localities on the map (text-fig. 1, table 1). Specimens from A1 were from loose frag- ments at locality A and those from localities B and C came from loose blocks close by the river. Unfortunately most of the specimens of the new species of Crurithyris and Lambdarina came from the loose blocks rather than the in situ Leek Road localities. [Palaeontology, Vol. 17, Part 4, 1974, pp. 811-840, pis. 107-111.] 812 PALAEONTOLOGY, VOLUME 17 N A A - SAMPLING SITE SCALE 100 metres BASED ON OS PLANS O0SS AND 0955 VISEAN BEDDED LIMESTONE REEF LIMESTONE 097 096 099 IOO text-fig. 1 . Geological sketch map of the area from which collections were made (marked a-f) approximately 1 km due west of Wetton village. Grid reference numbers, within the area sk, are given. Partially silicified limestones occur widely in the Lower Carboniferous of Derby- shire and Staffordshire, but are not well documented. Thomas and Ford (1963) have described a silicified tabulate coral from Bradbourne, and Morris (1970) has noted silicified microfossils in the Hamps Valley, from where he has collected specimens of a new Lambdarina species and another related small brachiopod genus from rocks of high Tournaisian age. The age of the bedded limestones from which the fauna was collected is in question. Prentice (1951) mapped the whole of Wetton Hill and this part of the Manifold Valley as, what Ludford (1951) and he called, the Waterhouses Limestone; he assigned this to the D Zone of the Visean. However, Parkinson and Ludford (1964) believe that the Waterhouses Limestone, forming the east summit of Wetton Hill, is faulted BRUNTON AND CHAMPION: CARBONIFEROUS BRACHIOPODS 813 ~~ LOCALITIES SPECIES ' A Al B c D E F Acanthocrania cf 1 aevi s (Keyes) 2 Schuchertella sp. 3 Avonia (Quasiavonia ) aculeata (J Sowerby) 1 2 ?x 'Stegacanthia' sp. X X Pleuropugnoides pleurodon (Phillips) 2 Lambdarina manifoldensis sp.n. 1 15 72 ca.40 Coledium seminulum (Phillips) ca.40 23 ca.6 ca. 7 1 3 Hustedia cf radialis (Phillips) 223 41 ca.25 15 Hustedia ulothrix (de Koninck) 16 1 1 X Cleiothy rid ina fimbriata (Phillips) ca-4 ca.5 ca.4 X C leiothyridina deroissyi (Leveiile) 1 1 Crurithyris nastus sp.n. 1 63 ca.200 3 Cyrtina cf burlingtonensis Rowley 1 1 Fusella rhomboidea (Phillips) X Spi r i ferel 1 ina perpl icata (North) ca.20 ca.20 ca.13 ca- 3 Girtyella saccula (j de C Sowerby) 5 5 X ?x Rhipidomel la fragments X 2 1 table 1 . The brachiopod fauna and localities from which the indicated numbers of specimens were collected. Samples from localities Al, B and C were not in situ , x indicates few fragments of the species; approximate numbers ( ca .) are used for species in which complete specimens are augmented by some fragmentary material. against C2 or C2S, limestones forming the west flank of the hill and extending into the Hamps and Manifold Valleys. Thus, according to Parkinson and Ludford, our silicified rocks are pre-D Zone in age and could perhaps be assigned to the Manifold Limestone-with-Shales. Prentice's 'conglomerate' (1951, p. 182) is topographically and stratigraphically below our collecting localities but does not crop out within the area of text-fig. 1. He interpreted this ‘conglomerate’ as separating rocks of Q age from those of Dt age. Thin sections of this rock show that the limestone fragments are angular and in some instances fit one to another. The calcite matrix appears to have crystallized within the interspaces and to have assisted in splitting some of the limestone fragments, which are fossiliferous. We believe, therefore, that this rock represents a true ‘reef’ breccia derived from the near-by Thor’s Cave ‘reef’ of C age and that, although it may represent a break, sedimentation of pre-D Zone age con- tinued. Our view, therefore, is to favour Parkinson and Ludford’s suggestion that these rocks are of C2Sj age and the brachiopods here described tend to support this in that Lambdarina is known from supposed Cj limestones in Co. Dublin, Ireland, as well as the type area, but is unknown from D Zone rocks. Crurithyris nastus sp. nov. could be considered as ancestral to C. urei (Fleming), a species typical of the D Zone. In addition the small coral Cladochonus, especially C. crassus M’Coy is reported as common in C2 and S Zone rocks (Thomas and Ford 1963; Parkinson and Ludford 1964). A species of Cladochonus like the young of C. crassus or C. bacil- larius M’Coy is common in our collections. 814 PALAEONTOLOGY, VOLUME 17 THE FAUNA Limestone samples were taken from several points west of Wetton village (text-fig. 1). After treatment in dilute hydrochloric or acetic acids the fauna was found to be com- posed predominantly of brachiopods together with abundant corallites of Clado- chonus , a few gastropods, crinoid ossicles and plates, echinoid spines, and small bryozoan fragments. Silicification of the brachiopods was almost complete but replacement of the shell substance is by rather coarse silica, which in places has developed beekite rings. Some of the smaller specimens are silica-filled. Rarely small crystals of chalcopyrite impregnate the silica. The fauna is enigmatical in that whilst it contains little of diagnostic stratigraphical importance (no microfossils have been found despite the use of acetic acid on some samples), it does yield some intriguing brachiopod species, e.g. the StegacanthiaAike species, the new Lambdarina species, the abundance of Crurithyris nastus, and the records of Fusel/a and Cyrtina , poorly known genera in our Lower Carboniferous rocks. Thus age determination of these rocks remains equivocal. Some species, such as Spiriferellina perplicata and Pleuropugnoides pleurodon , are more typical of D strata. Most of the other species are wide ranging, whilst the two species compared with North American forms (viz. Acanthocrania cf. laevis and Cyrtina cf. burlingtonensis ), as well as Stegacanthia s.s., are all commonly found in the Kinderhook and Osage, equivalent to the Tournaisian. The brachiopod fauna is characterized by the preponderance of Spiriferida and by the small sizes of its constituent species, some of which are represented only by young specimens, e.g. the Avonia sp. and Hustedia ulothrix. The other species are commonly of small size, reaching less than 10 mm in length. The best preserved and most prolific species tend to have a highly biconvex profile and there is a high pro- portion of articulated shells of Hustedia, Crurithyris, Lambdarina, and Coledium. Almost all the larger Spiriferellina shells are disarticulated and, allowing for imper- fections of silicification, external ornamentation is reasonably well preserved. Crush- ing of some shells occurred prior to silicification, particularly in Hustedia and to a lesser extent in Crurithyris. This crushing probably took place during sediment compaction and might have preferentially damaged larger specimens originally present in the fauna. However, silicified fragments of larger specimens are rare. An incompletely silicified, and consequently fragmentary productacean has spines up to 30 mm long and it seems unlikely that this shell could have been moved any great distance after death. On the other hand, the general small size of the specimens and their biconvexity might suggest that they were winnowed from some near-by region. With the exception of rare productaceans, which were passively sessile after the brephic stage, the other species were probably all attached to the substrate by pedicles. It seems unlikely that these species all had ‘rooting’ pedicles specialized for gripping in soft sediment and if not then there must have been sufficient firm substrate, either the bottom itself or skeletal debris, upon which their pedicles attached. If this were so a high rate of sedimentation is unlikely and it may well be that the area was one subject to slight current activity. BRUNTON AND CHAMPION: CARBONIFEROUS BRACHIOPODS 815 Ramsbottom (1970), in discussing British Namurian facies, and Samtleben (1971), in describing Bolivian Lower Permian brachiopods, suggest that Crurithyris indicates a near-shore or relatively shallow-water environment. In view of the ‘reefal’ deposits both below and above the rocks from which the fauna was collected and the con- glomeratic breccia near by, a shallow-water environment seems likely. The small sizes of the individual species and apparent pedicle attachment of the brachiopods may support the view that they lived in conditions of shallow agitated water. SYSTEMATIC DESCRIPTIONS Class inarticulata Huxley 1869 Order acrotretida Kuhn 1 949 Superfamily craniacea Menke 1828 Genus acanthocrania Williams 1943 Type species. Crania spiculata Rowley 1908, by original designation of Williams (1943, p. 71). Type specimen from the Louisiana Limestone, Missouri, U.S.A. Acanthocrania cf. laevis (Keyes) Plate 107, figs. 1-5 1894 Crania laevis Keyes, p. 40. 1914 Crania laevis Keyes; Weller, p. 47, pi. 1, fig. 33. 1968 Acanthocrania cf. laevis (Keyes); Brunton, p. 7, pi. 1, figs. 10-14. Two specimens believed to be dorsal valves of A. laevis but differing in outline and profile come from locality Al. Silicification is not perfect so the external ornamenta- tion is poorly preserved and the interiors are obscured by siliceous residues. Both specimens are small, being no more than 2 mm at their widest. One is 1-2 mm high (PI. 107, fig. 1), the other only 0-6 mm high (PI. 107, fig. 3). Although differing in shape the external ornamentation, where preserved, is identical and we believe the two to be conspecific. The external ornamentation can be seen to consist of radially arranged minute spines or papillae which tend to give the impression of a delicate radial ribbing. This ornamentation is characteristic of the genus. The species A. laevis was fully described by Weller (1914) who first drew attention to the spinose exterior. It was first recorded and figured from Upper Palaeozoic rocks of Britain by Brunton (1968) in describing silicified specimens from the Visean of Ireland. It is probably closely related to A. spiculata (Rowley), but without good type material with which to make comparison the two species must remain. Class articulata Huxley 1869 Order strophomenida Opik 1934 Superfamily davidsoniacea King 1850 Family schuchertellidae Williams 1953 Genus schuchertella Girty 1904 Three very small specimens of Schuchertella- like form have been collected from locality A ; two are almost complete shells (PI. 107, figs. 6-7) and one is an incomplete 816 PALAEONTOLOGY, VOLUME 17 dorsal valve (PI. 107, figs. 8, 9). None is more than 3 mm wide. The unfigured specimen is broken at the umbo revealing neither dental plates nor a median septum. The cardinalia (PI. 107, fig. 8) is Schuchertella- like in character and also similar to that of Serratocrista Brunton. What appears on the figure to be a median anterior node on the cardinal process is actually the broken edge of the valve floor. The external ornamentation is more like Schuchertella than Serratocrista so we suggest that the former genus was present in the fauna. Superfamily productacea Gray 1840 Family overtoniidae Muir-Wood and Cooper 1960 Genus avonia (quasiavonia) Brunton 1966 Type species. Productus aculeatus J. Sowerby 1814, p. 156, pi. 68, fig. 4. Avonia ( Quasiavonia ) aculeata (J. Sowerby) Plate 107, figs. 12-13 1814 Productus aculeatus J. Sowerby, p. 156, pi. 68, fig. 4. 1966 Avonia (Quasiavonia) aculeata (i . Sowerby); Brunton,p. 218, pi. 10, figs. 8-1 7 ; pi . 11, figs. 1-21. Lectotype. Specimen from the Sowerby Collection, chosen by Muir-Wood (1951, p. 101) and housed in the British Museum (Nat. Hist.), B 60992. This species was fully described, externally and internally, by Brunton (1966), using silicified material from Ireland. The Manifold silicified fauna has yielded a few young specimens of this species which reach almost 6 0 mm in length. Their rounded- quadrate to subcircular outline, lamellose ornament on the ventral valve, and suberect scattered spine bases are characteristics of this species. The spine bases of the juvenile clasping spines can be distinguished upon the ventral umbo, similar to those figured by Brunton (1966, pi. 11, fig. 9), but on none of the three complete specimens is a pedicle sheath preserved. The concave dorsal valve has an ornamentation of well- developed growth lines but appears to be devoid of spine bases. Normally the adults of this species have spinose dorsal valves, but the first-formed spines are either not preserved or did not grow until the valve was at least 5 0 mm long. The species is rare, but has been collected from locality C (a loose block) and locality A, in situ. EXPLANATION OF PLATE 107 Figs. 1-5. Acanthocrania cf. laevis (Keyes). 1-2, lateral and dorsal views of the more conical specimen. Locality Al. SEM x25. BB 60241. 3-5, lateral and dorsal views (SEM x 25) and detail of external ornamentation (SEM x 160). Locality AL BB 60240. Figs. 6-9. Schuchertella sp. 6-7, dorsal and postero-dorsal views of a young shell from locality A. SEM x 23. BB 60242. 8-9, posterior and interior views of a young dorsal valve. SEM x 21. BB 60243. Figs. 10-12. Cf. Stegacanthia sp. 10, incomplete dorsal valve interior from locality D, x 2. BB 60246. 11, general view of the ventral valve fragment from locality Al. The mid-line is marked by an arrow, x 1-7. BB 60245. 12, detail of mid-region of fig. 11 showing the line of large more erect spine bases, x 4-3. Figs. 13-14. Avonia ( Quasiavonia ) aculeata (J. Sowerby). Dorsal and ventral views of the young specimen from locality A. x 5. BB 60244. PLATE 107 '-tAVs W3 14 13 12 BRUNTON and CHAMPION, Carboniferous brachiopods 818 PALAEONTOLOGY, VOLUME 17 Family echinoconchidae Stehli 1954 Genus stegacanthia Muir- Wood and Cooper 1960 Type species. S. bowsheri Muir-Wood and Cooper 1960, p. 199, pi. 48, figs. 1-12, from the Lake Valley Formation, early Mississippian, Lake Valley, New Mexico, U.S.A. Muir-Wood and Cooper (1960) and Muir-Wood (in Williams et al. 1965) placed this genus with the Overtoniidae but it would seem to us more closely allied to the genus Pustula and other genera placed within the Echinoconchidae. cf. Stegacanthia sp. indet. Plate 107, figs. 10-12 Fragments of a lamellose and highly spinose productacean have been recovered from limestone blocks at locality Al and probably also from locality D. The external ornamentation of these fragments is of growth lamellae, which are well preserved peripherally, from which extend a single series of recumbent spines. Each spine is accompanied by a spine ridge posteriorly which may cause a fold in the anterior margin of the adjacent lamella (PI. 107, fig. 12). This style of spinose ornamentation is very similar to that seen on Stegacanthia and were it not for the addition of one feature on our figured specimen it would certainly be placed within this genus. However, along the mid-line of the ventral valve there is a line of spine bases, one per lamina, which indicate that there was a line of rather larger suberect spines. This is not a feature of Stegacanthia , nor, so far as is known, a feature upon other highly spinose genera. Fragments of what would seem to be a second individual include a piece approximately 1 cm square which probably shows this same feature on two adjacent laminae. If there were, indeed, two specimens with a median line of suberect spines it is unlikely that the feature is a freak. The pos- terior regions of the valves are missing from locality Al, but the piece of shell from locality D includes the mid-region of the dorsal valve (PI. 107, fig. 10). The sessile cardinal process supported by a wide but tapering low median septum and strong lateral ridges are charac- teristics seen also in the type species of Stegacanthia. Preservation of the valve exterior is poor, but from what can be seen it is almost certainly the same species as that from locality Al. In 1966, while commenting upon a few Irish silicified specimens he attributed to Pustula cf. pyxidiformis (de Koninck), Brunton indicated several similarities between his material and Stegacanthia. However, a restudy of the Irish material shows no sign of the median row of suberect spines and it cannot be considered as conspecific with our Manifold specimens. text-fig. 2. Camera-lucida sketch of the Stegacanthia- like specimen BB 60246. a, detail of the cardinal process exterior, b, internal view of the dorsal valve fragment with part of the ventral valve exterior preserved at the anterior margin. (See PI. 107, fig. 10.) BRUNTON AND CHAMPION: CARBONIFEROUS BRACHIOPODS 819 Until more material of this species is available we can say no more than that this is probably a new species closely allied to Stegacanthia bowsheri. At a width estimated at about 50 mm, this was the largest species within the fauna. Order rhynchonellida Kuhn 1949 Superfamily rhynchonellacea Gray 1848 Family wellerellidae Likharev in Rzhonsnitskaya 1956 Genus pleuropugnoides Ferguson 1966 Type species. Terebratula pleurodon Phillips 1836, p. 222, pi. 12, figs. 25-28. Ferguson (1966) first described this genus from type and other material in the British Museum (Nat. Hist.) and compared the type species with his new species P. greenleightonensis. He included within his new genus the species Terebratula flexistria Phillips and T. proava Phillips. Pleuropugnoides pleurodon ( Phillips) Plate 108, figs. 1 4 1836 Terebratula pleurodon Phillips, p. 222, pi. 12, figs. 25-28. 1861 Rhynchonella pleurodon (Phillips); Davidson, p. 101 (pars), pi. 23, figs. 1-6. 1966 Pleuropugnoides pleurodon (Phillips); Ferguson, p. 355, pi. 23, figs. 1-6. Type specimen. Lectotype chosen by Ferguson (1966), specimen illustrated by Phillips (1836), pi. 12, figs. 25, 26. Gilbertson Collection, British Museum (Nat. Hist.), B 361. Diagnosis. Adults transversely elliptical rhynchonelliform shape, deep bodied with strongly uniplicate anterior commissure. Entirely costate, commonly with four ribs in the ventral sulcus. Dorsal median septum extending about one-third valve length, fused posteriorly to deep ventrally curved crura connected to short sockets by narrow outer hinge plates. Discussion. This species is represented in our Manifold collections by only two specimens from locality E. Both are young and the better-preserved individual (BB 60247) is illustrated here (PI. 108, figs. 1-4). The outline, being almost circular, is atypical for adult shells. However, from studies of conspecific material in which various growth stages are represented it can be seen that our specimen is of normal shape for its stage of development. Internal details are lacking in our material and use has been made of Irish silicified specimens to describe the cardinalia. Family cardiarinidae Cooper 1956 Subfamily lambdarininae new subfamily Genus lambdarina new genus Type species. L. manifoldensis sp. n. from Lower Visean strata about one-half mile (1 km) west of Wetton, north Staffordshire. Diagnosis. Cardiarinidae lacking a parathyridium, ventral umbo elongate and strongly bilobed body, each lobe extending antero-laterally at approximately 45° from mid- line. Discussion. At an early stage in the preparation of this paper we became aware of work by Dr. P. G. Morris which included the description of new brachiopods 820 PALAEONTOLOGY, VOLUME 17 congeneric with L. manifoldensis for which he intended the generic name Lambdarina. We gratefully acknowledge the opportunity of seeing his specimens and prepared script enabling us to compare his species from Brownend Quarry with L. mani- foldensis. We retain the name Lambdarina both in recognition of Dr. Morris’s work and because of its suitability in describing the outline of these species. The classification of the Cardiarinidae has always been uncertain. We retain the family within the Rhynchonellacea as there is no sign of endopunctation in our material although endopunctate shells found with Lambdarina specimens retain this feature. Lambdarina manifoldensis sp. nov. Plate 108, figs. 5-11 Type specimens. From rocks believed to be of c. C2S! age. Figured specimens, British Museum (Nat. Hist.), BB 60248 to BB 60252 and BB 60264, from locality B. Holotype, BB 60248. Diagnosis. Lambdarina with rounded lobes about one-quarter of the shell’s length, separated on ventral valve by low median ridge. Lateral profile plano-convex with ventral umbo slightly upcurved. Teeth and sockets well developed, low notothyrial platform posteriorly, and no differentiated crural processes. Description. The largest of these small shells reaches 2-5 mm in length. The over-all lateral profile is piano to biconvex, the ventral valve umbo tending to curve dorsal of the commissural plane, but individually the lobes of both valves are strongly convex. External ornamentation is lacking, save for fine growth lines. The pedicle foramen is at the postero-dorsal extremity of the tube-like ventral valve umbo and the delthyrium is closed by a flat symphytium (PI. 108, figs. 5, 11). The dorsal valve umbo is prominent medianly but somewhat flattened laterally. From a length of about 0-5 mm the dorsal valve is sulcate whereas in the correspond- ing position on the ventral valve there is a low ridge within the sulcus which divides the lobes of the shell (PI. 108, fig. 6). The teeth are rounded in lateral profile and project above the valve margins (PI. 108, fig. 11); they are narrow and extend antero- dorsally and slightly medianly to fit tightly behind the strong inner socket ridges of the dorsal valve. The teeth are supported by thin dental plates partially fused to the inner walls of the valve (text-fig. 3). Muscle scars have not been distinguished but the dorsal diductor muscle attach- ments were probably on the posterior ridge connecting the inner socket ridges and enclosing the notothyrial platform. The species is named after the Manifold river. text-fig. 3. Postero-ventral view of the pos- terior portion of the ventral valve of Lambdarina manifoldensis sp. nov. showing the disposition of the teeth (t) and dental plates (dp), the symphytium (s), and pedicle aperture (pa). (See PI. 108, fig. 11.) BRUNTON AND CHAMPION: CARBONIFEROUS BRACHIOPODS 821 Discussion. A different species of Lambdarina is known from Brownend Quarry, Waterhouses, about 3 miles south of our localities, as a result of the acid develop- ment of limestones near the Tournaisian/Visean boundary by Dr. Morris, during his investigations of conodont faunas. This older species from Brownend Quarry is similar in shape and dimensions to L. manifoldensis, but differs in having relatively longer lobes and in lacking the low median ridge on the ventral valve of our species. From a brief study of Morris’s material it can be seen that there are only slight internal differences; the inner socket ridges of L. manifoldensis are more strongly developed and do not recurve to the valve margins as appears to be the situation in Morris’s specimens. Internally L. manifoldensis is similar to Cardiarina cordata Cooper, described from Pennsylvanian rocks of New Mexico, U.S.A. Cardiarina is furnished with a parathyridium (Cooper 1956, text-fig. 1) and it is this structure which most clearly differentiates the genus from Lambdarina. If there is a direct phylogenetic relationship between these two genera then it might be that the reduction of bilobation seen between the Brownend specimens and L. manifoldensis con- tinued with the evolution of C. cordata. The development of the parathyridium might be associated with a reduction in the anterior growth of the inner socket ridges and teeth, both of which seem less well developed in the American species. During early ontogeny, up to a length of from 0-8 mm to 10 mm, the outline of L. manifoldensis was triangular, the length being nearly twice the width (PI. 108, fig. 7). Bilobation of the brachial cavity then developed but the mid-point of the anterior commissure continued to grow anteriorly, although less rapidly than the antero-lateral margins. Thus, once an individual had grown to about 1 -6 mm long its relative shape, including the proportions of its lobes had reached that of the largest specimens collected, i.e. 2-5 mm long. The ventral umbo of juvenile specimens extended about 01 mm beyond the dorsal umbo and the pedicle aperture seems to have been connected to a small open delthyrium. During ontogeny the ventral umbo grew as a posteriorly tapering cone, on the dorsal side of which is a very narrow flat interarea and flat symphytium. The terminal aperture seems to have remained func- tional throughout life. As yet this is the only published description of a species assigned to Lambdarina. However, in addition to the species from Brownend Quarry Dr. G. Sevastopulo of Trinity College, Dublin, has presented to the British Museum (Natural History) a few specimens (probably conspecific with the Brownend species) from uppermost Tournaisian shales of the Feltrim ‘reef’ in Co. Dublin, Ireland. No specimens that could be placed in the Cardiarinidae have been collected from the Visean, D Zone, silicified faunas of Co. Fermanagh, Ireland, described by C. H. C. Brunton. Thus, as yet, our knowledge of the geological and geographical ranges of this and related species is very limited. Habitat. Not having seen the relationship of any of these shells to the enclosing sedi- ment it is impossible to do more than speculate upon the possible habitat of this species. In common with most other elements of the fauna the pedicle appears to have remained functional throughout life. In view of the small size of the species it seems likely that attachment was to a hard surface rather than by ‘rooting’ into relatively soft sediment. The morphology of the umbones and articulatory surfaces indicates 822 PALAEONTOLOGY, VOLUME 17 that the shell could not gape widely. It is likely, therefore, to have been able to adjust its position in free water using the pedicle, in contrast to non pediculate sessile and cemented species of brachiopod which tend to have wide gapes allowing the full exposure of the lophophore. In view of the small size of the specimens it might be that in life they attached to plants or animals and may have been epiplanktonic. Like the shells of Cardiarina cordata and the Brownend specimens, L. manifoldensis was attacked by boring organisms. In a sample of 54 specimens, 7 are bored of which only one is bored on the dorsal valve. All the holes are about 0- 1 mm in diameter, positioned near the mid-line and within the mid one-third of the shell’s length. There is some evidence for slight bevelling of the outer edges of the bored hole. The size corresponds to the minimum size of supposed gastropod borings on brachiopods described by Brunton (1966), but is less than that usually seen in shells and less than that expected from gastropod attack. If, as suggested above, L. manifoldensis was not sessile on a soft substrate it precludes the possibility of attack by burrowing organisms, such as the recent naticids, which approach their prey from below the sediment surface. Superfamily stenoscismatacea Oelert 1887 (1883) Family stenoscismatidae Oelert 1887 (1883) Genus coledium Grant 1965 Type species. Coledium erugatum Grant (1965, p. 96) from the Moorefield Formation (Upper Mississippian) of Oklahoma, U.S.A. Stenoscismataceans, as such, are rarely recorded from British or European Lower Carboniferous rocks. However, utilizing specimens in the U.S. National Museum, Washington identified as Terebratula globulina Phillips and T. rhomboidea Phillips, Grant (1965) placed these British species within Coledium. Grant followed Davidson (1861) in placing T. seminula Phillips from the Lower Carboniferous of Bolland into synonymy with T. globulina, a Permian species originally described by illustration only. Inspection of the type material of T. seminula and specimens of T. globulina studied by Phillips, all housed in the British Museum (Nat. Hist.), leads to the con- clusion that these species are distinctive. Despite the possibility of these three species names being used in British Upper Palaeozoic faunal lists, commonly linked with the generic names Rhynchonella, Athyris, or Camarophoria, records are few and this description is the first, with photographs, of a well-authenticated stenoscismatacean species from the British Lower Carboniferous. EXPLANATION OF PLATE 108 Figs. 1-4. Pleuropugnoides pleurodon (Phillips). Dorsal, ventral, anterior, and lateral views of a specimen from locality E. x 3. BB 60247. Figs. 5-11. Lambdarina manifoldensis gen. et sp. nov., from locality B. 5, dorsal view of an adult shell show- ing the symphytium. SEM x 22. BB 60248. 6, ventral view of an adult shell showing the low median ridge. Incomplete silicification replicates the growth lines externally and some of the fibrous shell postero- laterally. SEM x 20. BB 60264. 7, dorsal view of a young shell. SEM x 40. BB 60249. 8, oblique view of an incomplete young shell showing the teeth. SEM x 30. BB 60250. 9-10, oblique (SEM x 85) and posterior (SEM x 1 00) views of a dorsal valve showing the sockets (that on the right is filled with a broken- off tooth), and postero-median knob-like cardinal process. BB 60252. 1 1 , postero-lateral view of an adult ventral valve showing the symphytium and strong teeth. (See also text-fig. 3.) SEM x 43. BB 60251 . PLATE 108 BRUNTON and CHAMPION, Carboniferous brachiopods 824 PALAEONTOLOGY, VOLUME 17 Coledium seminulum (Phillips) Plate 109, figs. 1 -9 1836 Terebratula seminula Phillips, p. 222, pi. 12, figs. 21-23. 1861 Camarophoria globulina (Phillips); Davidson, p. 115, pi. 24, figs. 13-16. Type specimen. T. seminula Phillips 1836, pi. 12, fig. 23, from Bolland, Yorkshire. Gilbertson Collection, British Museum (Nat. Hist.), B 355. Here selected as lectotype. Diagnosis. Small (about 5 mm long) Coledium , slightly wider than long. Uniplicate anterior commissure modified normally by three ribs on dorsal fold with origins about 3 0 mm from dorsal umbo. Spondylium sessile posteriorly and elevated on low median septum anteriorly. Camarophorium narrow, concave postero-ventrally, vertically disposed anteriorly and supported by high median septum which does not extend anteriorly on valve floor. Discussion. The complete specimen figured (PI. 109, figs. 1-4) is slightly larger than the lectotype (which measures 4-2 mm long and 4-5 mm wide) and consequently has more fully developed lateral ribs, but in other external details the two appear to be identical. Unfortunately the interiors of type or topotypic material are unknown, but there seems little doubt that the present specimens should be assigned to C. seminulum. Davidson (1861) con- sidered this species to represent the young of C. rhomboidea (Phillips) but inspection of the type material and sections of a specimen believed to be rhomboidea shows that these two species are distinc- tive both externally and internally; C. rhomboidea (Phillips) has a spondylium supported along its length by a well-developed median septum whilst that of C. seminulum is sessile save for its anterior end. This species is relatively common at the main collecting sites (table 1) and the complete collection contains about sixty-five more or less whole speci- mens; few, however, show the internal structures. Within the dorsal valve the camarophorium extends antero-ventrally as a trough-like process facing directly into the cavity of the spondylium (PI. 109, fig. 9). The sockets and hinge plates appear to be short and give rise to small diverging and antero- ventrally directed crura. An intercamarophorial plate, as described by Grant (1965, p. 99) for the type species of Coledium , has not been observed in our material. Being relatively common and represented by specimens of various sizes, it is possible to say a little about the ontogeny of this species. Shells down to a size of about 2 mm in length can be assigned to the species with some degree of certainty but there are many other individuals down to a length of about 1 mm for which a specific deter- mination is doubtful. From the Manifold material it seems that the shell of C. semi- text-fig. 4. Drawings of an incomplete shell viewed postero-laterally (a) and of a ventral valve interior (b) of Coledium seminulum (Phillips) showing the spondylium (s) and camarophorium (c). Specimen a is illustrated in Plate 109, fig. 9. BRUNTON AND CHAMPION: CARBONIFEROUS BRACHIOPODS 825 nulum grew in a regular, smooth subcircular to broadly ovate form, with neither folding nor ribbing, up to a length of 3 0 mm to 3-5 mm. Beyond this length the rounded ribs and plication of the commissure began to develop; normally it is the ribs within the fold and sulcus region which developed first. During this growth period, and to maturity, shell accretion at the anterior commissure became increas- ingly opposed so as to increase considerably the depth of the shell cavity. Clearly, it was during this period too that the camarophorium changed its direction of growth from essentially anterior to ventral. Thus, the normal ontogenetic situation seems to have been that these shells were gently biconvex, narrow-bodied brachiopods up to a length of about 3-5 mm after which opposed growth at the anterior margins led to an increase in shell thickness (depth of body cavity) and a slight relative increase in shell width. Grant (1965) pointed out that late Devonian and early Mississippian species of Coledium have an open pedicle foramen, which presumably remained functional during life. This condition would appear to have been true for C. seminulum and it is assumed that during life these shells were attached to the substrate by their pedicles. All the pedicle apertures are small (about 0T mm in diameter) and show no sign of abrasion at their margins. It is possible, therefore, that attachment was to a relatively soft substrate and/or that the habitat was one of quiet conditions. Many of the specimens are sufficiently finely silicified for the fibrous nature of the secondary shell to be replicated and distinguishable at magnifications as low as x 10. At this magnification the shell fibres of the unsilicified type specimen can be seen. These fibres are distinctly larger than those of endopunctate species, such as Hustedia and Spiriferellina species, and slightly larger than those commonly seen in impunctate spire-bearing brachiopods. Order spiriferida Waagen 1883 Superfamily retziacea Waagen 1883 Family retziidae Waagen 1883 Genus hustedia Hall and Clarke 1893 Type species. Terebratula mormoni Marcou 1858, p. 57, pi. 6, fig. 1 1, from near Salt Lake City, Utah, U.S.A. Type specimen not in Marcou Collection at British Museum (Nat. Hist.). Hustedia cf. radialis (Phillips) Plate 109, figs. 10-18 1836 Terebratula radialis Phillips, p. 223, pi. 12, figs. 40, 41. 1861 Retzia radialis (Phillips); Davidson, p. 87, pi. 17, figs. 19-21. 1887 Retzia multicostata de Koninck, p. 95, pi. 22, figs. 20-24. Type specimen. The specimen figured by Phillips (1836, pi. 12, figs. 40-41) from Bolland, Yorkshire. Gilbertson Collection, British Museum (Nat. Hist.), B 328. Diagnosis. Small subcircular to broadly ovate Hustedia with slight dorsal sulcation and emarginate anteriorly. Entirely costate, dorsal valve with 15 to 17 rounded ribs. Discussion. This species is one of the most prolific within the faunas collected from the area. Despite this we are unable to assign the name radialis with certainty because of G 826 PALAEONTOLOGY, VOLUME 17 the confused situation still surrounding the specimens called radialis within the Gilbertson Collection used by Phillips for his 1836 publication. His collection includes specimens closely comparable to our material as well as specimens much larger (9-6 mm long), one of which should be considered as the type specimen on the grounds that it is probably the specimen figured by Phillips in 1836. Thus, there is some doubt as to whether both the large and small specimens should be called radialis. There are differences in the normal numbers of ribs on the dorsal valves of the type specimens and those from the Manifold Valley (19 to 21 and 15 to 17 respec- tively), but details of internal characteristics are unknown for the Gilbertson specimens. For the present, therefore, we compare our common Hustedia species to radialis, but recognize that it may prove to be distinctive. Recommendations upon the problem of H. radialis s.s. and a description of the species are beyond the scope of this paper but are being worked upon for publication elsewhere. In Lower Carboniferous, Mississippian rocks of North America Hustedia pygmaea Rowley and H. texana Girty would appear to be similar to our species. Rowley (1900) described his species from ‘the soft white cherts of the Lower Burlington limestone’ of Missouri. It is a strongly biconvex species reaching about 6-5 mm in length and c P.a c.p l.p text-fig. 5. The internal umbonal region of a young shell of Huslediac f. radialis (Phillips) looking postero-ventrally and showing the cardinalia. The ventral valve (w) is uppermost and the position of the pedicle aperture (pa) is indicated. The cardinal process (cp) and median ligulate process (lp) recurve postero-ventrally, but are incom- pletely preserved. The crura (c) diverge antero-ventrally. EXPLANATION OF PLATE 109 Figs. 1-9. Coledium seminulum (Phillips). 1-4, dorsal, ventral, anterior, and lateral views of a specimen from locality B. x4. BB 60253. 5-8, dorsal, ventral, anterior, and lateral views of a specimen from locality A. (Note the prominent halt in shell deposition at about three-quarters of the specimen length.) x 5. BB 60255. 9, oblique view into a broken shell from locality C showing the high camarophorium. (See also text-fig. 4.) x 7. BB 60254. Figs. 10-18. Hustedia cf. radialis (Phillips). 10-12, dorsal, ventral, and lateral views of one of the larger specimens from locality A. x 7. BB 60256. 13-15, dorsal, ventral, and lateral views of a specimen from locality C. x 6. BB 60257. 16-18, lateral, dorsal, and ventral views of a young shell from locality C. x 10. BB 60258. Figs. 19-21. Hustedia ulothrix (de Koninck). Lateral, dorsal, and ventral views of a young shell from locality C. x 7. BB 60259. Figs. 22-23. Cleiothyridina fimbriata( Phillips). Ventral view of a young slightly crushed shell from locality A ( x 6), and detail of the external ornamentation antero-laterally ( x 22). BB 60260. PLATE 109 BRUNTON and CHAMPION, Carboniferous brachiopods 828 PALAEONTOLOGY, VOLUME 17 differs from our specimens in lacking a dorsal valve sulcus and in having about 15 ribs on that valve. H. texana Girty (1926) is similar to the Manifold specimens in having a dorsal sulcus and occasionally an emarginate anterior commissure, but with about 19 ribs on the dorsal valve it is slightly more finely ribbed. H. texana appears to reach about 7 mm long but Girty’s types are barely 5 mm long. Carter (1967) has redescribed this species and presents serial sections showing the internal structures. From these, and his plate figures, it seems that this mid Tournaisian species in the U.S.A. is closely related to the small forms called H. radialis and H. multicostata (de Koninck) in Britain and western Europe. Most of the Manifold specimens are complete, silica-filled or slightly crushed so internal features are rarely seen. However, sufficient material is available to show the typical Hustedia cardinalia with its posteriorly recurved cardinal process extending into the ventral valve and the antero-ventrally projecting crural processes (text- fig. 5). The brachidium consists of laterally directed spiralia, each having at least three coils. Hustedia ulothrix (de Koninck) Plate 109, figs. 19-21 1843 Terebratula ulotrix de Koninck, pi. 19, fig. 5 (plate explanation only). 1861 Retzia ulotrix (de Koninck); Davidson, p. 88, pi. 18, figs. 14, 15. 1887 Retzia ulothrix (de Koninck); de Koninck, p. 92, pi. 22, figs. 1-4. Type specimen. That figured by de Koninck in 1887 (pi. 22, figs. 1 -4) from Tournai, Belgium, housed in the Brussels Museum. This specimen is accepted as type although it cannot be proved to be the specimen figured in 1843. The label with this specimen, in de Koninck’s handwriting, uses the spelling ulothrix, as do his publications after 1843, and this spelling is here followed in the belief that ulotrix , used only on a plate explanation, was a printer’s error. Diagnosis. Hustedia with few (7 to 9 costae on dorsal valve) strong ribs and with width commonly exceeding length. Cardinal process lacking median ligulate process. Discussion. The species is less common than H. radialis, accounting for only about 16% of all the Hustedia specimens. This order of abundance is seen elsewhere in localities from which full faunas have been collected by acid extraction. In museum collections H. ulothrix is normally over-represented as it is a much more obvious species to collect by traditional methods than is H. radialis. The Manifold specimens do not attain the size of the type specimens or those figured by Davidson (1861, pi. 18; 1863, pi. 54), collected from near Wetton from rocks probably of Dj age. Superfamily athyridacea M’Coy 1844 Family athyrididae M’Coy 1844 Subfamily athyridinae M’Coy 1844 Genus cleiothyridina Buckman 1906 1850 Cleiothyris King, p. 137. Type species. Atrypa pectinifera J. de C. Sowerby 1840, by original designation of King 1850. An application to the International Commission on Zoological Nomenclature by BRUNTON AND CHAMPION: CARBONIFEROUS BRACHIOPODS 829 Brunton (1972) seeks the validation of A. pectinifera as type species of Cleiothyridina. This is on the grounds that Buckman’s generic name was a substitute for Cleiothyris King, for which King had designated as type A. pectinifera. The type species of Cleiothyridina has commonly been quoted as C. deroissyi (Leveille), but this designa- tion was based upon a misconception of Leveille's original description. C. deroissyi (Leveille) (1835, pi. 2, figs. 18-20) is in reality a wider and more strongly folded species than C. fimbriata (Phillips), the species with which it has been confused for over one hundred years (Davidson 1861). Cleiothyridina fimbriata (Phillips) Plate 109, figs. 22, 23 1836 Spirifera fimbriata Phillips, p. 220 (no figure). 1861 Athyris Royssii L’Eveille; Davidson, p. 84 (pars), pi. 18, figs. 8-11 (fig. 11 from Phillips’s original specimen). Type specimen. From Visean strata near Florence Court, Co. Fermanagh, Ireland. Phillips Collection, Oxford University Museum, E 1093. The Manifold Valley material is assigned to this species with some doubt as it consists only of one young specimen (PI. 109, fig. 22) and several very young specimens, a few millimetres long, together with fragments of larger specimens estimated to be 8 mm to 10 mm long. These larger fragments are ornamented by the spine-like frilly con- centric outgrowths typical of C. fimbriata. It would appear, therefore, that this species was represented in the fauna by both adults and juveniles, possibly forming a breed- ing population. Cleiothyridina deroissyi (Leveille) Plate 1 10, figs. 1-5 1835 Spirifer De Roissyi Leveille, p. 39, pi. 2, figs. 18-20. 71836 Spirifera globularis Phillips, p. 220, pi. 10, fig. 22. 1843 Terebratula royssii (Leveille); de Koninck (pars), p. 300, pi. 21, fig. la?, b-d. 1887 Athyris Roissyi (Leveille); de Koninck, p. 85, pi. 19, figs. 28, 29. 71887 Athyris squamigera (de Koninck), p. 82, pi. 20, figs. 16-22. 71914 Cleiothyridina prouti (Swallow); Weller, p. 474, pi. 79, figs. 13-16. Type specimen. The original specimens of Leveille are unknown, but probably came from the Tournai district. Diagnosis. Outline subcircular when young to transversely elliptical when adult, strongly biconvex shell. Persistent fold and sulcus forming parasulcate anterior com- missure. External ornamentation of closely spaced growth lines from which extend radially arranged finely spinose lamellae. Discussion. A single complete young specimen was recovered from a block, not in situ, at locality A 1 . It measures 6-3 mm long, 6-5 mm wide, and 4-2 mm thick and the ventral sulcus starts at a length of about 3-5 mm. The sulcus, external ornamentation, and relatively greater thickness of this shell distinguish it from C. fimbriata. In C. fimbriata the lamellae are longer before splitting up into flat spatula-like spines, whilst the lamellae of C. deroissyi soon divide into needle-like spinose clusters. 830 PALAEONTOLOGY, VOLUME 17 Especially on young shells this needle-like ornamentation is radially arranged. Although these two species are readily distinguishable they have been confused since 1861 when Davidson figured fimbriata Phillips but placed it in synonymy with deroissyi Leveille. C. deroissyi is less common than C. fimbriata within the British Isles and probably also from Europe. C. prouti, from mid Tournaisian strata of the U.S.A., appears from the literature to be conspecific with C. deroissyi and to form a minor constituent of the North American fauna. Superfamily cyrtiacea Frederiks 1919/1924 Family ambocoeliidae George 1931 Genus crurithyris George 1931 Type species. Spirifer urei Fleming 1828, pi. 14, fig. 12. Lectotype selected by George (1931) from the Ure Collection, Hunterian Museum, Glasgow (L 1790), from high Visean strata of Strathaven, Lanarkshire, Scotland. Crurithyris nastus sp. nov. Plate 110, figs. 6-16 Type specimens. Figured specimens are BB 60263 to BB 60268 (excluding BB 60264), from Visean strata at locality B. In addition there are about 300 specimens from localities indicated on table 1. Diagnosis. Crurithyris with longitudinally striated, tending to bifid, cardinal process. Inner socket ridges at about 45° to hinge line strongly fused to valve floor and dif- ferentiated from crural bases diverging from notothyrial platform. Ventral valve interior with low median ridge extending about three-quarters of the valve length. Description. The shape and main internal features of these shells are figured (PI. 1 10, figs. 11,15 and text-fig. 6). The anterior commissure is rectimarginate to broadly and gently sulcate, the dorsal sulcus being slightly deeper than that on the ventral valve. Muscle scars are not clearly differentiated but the dorsal adductor scars are narrowly obovate in outline, raised on a low median ridge and extend anteriorly to about one- third of the valve’s length (PI. 110, fig. 11). The ventral muscle scars are separated EXPLANATION OF PLATE 110 Figs. 1-5. Cleiothyridina deroissyi (Leveille). 1-3, dorsal, ventral, and lateral views of a small specimen from locality A. x 4. BB 60261 . 4-5, dorsal and anterior views of a young shell showing some external orna- mentation. Locality Al. x 3. BB 60262. Figs. 6-16. Crurithyris nastus sp. nov., from locality B. 6-10, dorsal, ventral, lateral, anterior, and posterior views of a complete shell, x 5. BB 60263. 11, dorsal valve interior of a mature specimen showing the adductor muscle scars. SEM x 14. BB 60267. 12, ventral valve exterior. (See also text-fig. 6.) x6. BB 60265. 13, ventral view of an incomplete shell showing part of the spiralia on the left side. x5. BB 60266. 14, stereoscopic pair of the last specimen illustrating the crura (that on the right arrowed, a), part of the spiralia, the tooth ridges bordering the delthyrium internally (arrowed, b), and the articulation, x 10. BB 60266. 15-16, dorsal valve interior (SEM x 10) and detail of the cardinalia (SEM x 34). BB 60268. Figs. 17-19. Cyrtina cf. burlingtonensis Rowley. Posterior, dorsal, and ventral views of the specimen from locality C. x 3. BB 60269. “ PLATE 110 17 16 18 BRUNTON and CHAMPION, Carboniferous brachiopods 19 832 PALAEONTOLOGY, VOLUME 17 medianly by a low ridge; the pair of adductor scars are postero-medianly placed, sur- rounded by shell thickening posteriorly and are otherwise similar to the dorsal scars. The diductor scars are less well differentiated but appear to be widely spreading on the valve floor antero-laterally of the adductor scars. The specific name is derived from nastos (Greek), meaning close-pressed or solid and refers to the characteristic socket ridges. Discussion. The delicate spinose external ornamentation common to Crurithyris species was discussed by George (1931) who concluded, it is believed correctly, that the lack of this feature resulted from preservation failure. This ornamentation has not been seen on the present material and is thought to have been lost in the process of silicification. More delicately silicified specimens from elsewhere display the spinose ornamentation clearly. However, internal morphology is well replicated by the silica and enables a ready comparison to be made with similarly silicified Crurithyris specimens available from Ireland and attributed to C. urei (Fleming). Externally C. nastus is similar to C. urei, the clearest difference being that the dorsal valve of C. urei is gently and regularly convex whilst that of the Manifold species is convex umbonally but becomes relatively flat anteriorly. Internally the differences are distinct. The cardinalia of C. nastus is illustrated (PI. 1 10, fig. 16) and contrasts with that of C. urei in which the cardinal process is indistinct, the sockets are at a low angle from the hinge line (20°-25°) and terminate anteriorly free of the valve floor. The inner socket ridges of both species are strong, but those of C. urei are much more widely separated (105° as compared to c. 85° in C. nastus) and the crura extend anteriorly, subparallel to each other, directly from the socket bases and are free of the valve floor. In the Manifold specimens crural bases are differentiated from the inner socket ridges and diverge at about 60° from the base of the cardinal process; in older specimens they fuse to the valve floor posteriorly. British Carboniferous Crurithyris species are most commonly recovered from D Zone limestones. C. nastus comes from rocks probably of basal Visean age and as such is intermediate in age between the younger species, such as C. urei, and the 1 mm text-fig. 6. The ventral valve interior of Crurithyris nastus sp. nov. viewed postero-ventrally (a) and ventrally (b) showing the low median ridge and faint positions of the diductor muscle scars. BRUNTON AND CHAMPION: CARBONIFEROUS BRACHIOPODS 833 Upper Devonian species C. unguiculus (J. de C. Sowerby). The morphology of the dorsal valve of our species is such that it could phylogenetically link C. unguiculus to C. urei. If this were so then we have evidence for a reduction in the development of crural bases and an increased separation of the anterior ends of the sockets from one another and from the floor of the valve. The cardinal process became less distinct, losing its bifid nature to become a tuberculate or ridged region between the apex of the inner socket ridges. Rarely, in partially broken shells, the first coil of the laterally directed spiralia can be seen. This seems to be much the same as that of C. urei (as seen in Irish silicified specimens) and C. planoconvexa (Shumard), from the Permian of Bolivia, as illus- trated by Samtleben (1972, pi. 8, fig. 3 a, b ). From about one-half of the dorsal valve length the lophophore support becomes a vertically disposed ribbon which twists ventro-laterally and then ventrally into the first coil. Evidence from other specimens indicates up to three full coils on each spiralia. Superfamily suessiacea Waagen 1883 Family cyrtinidae Frederiks 1912 Genus cyrtina Davidson 1858 Type species. Calceola heteroclita Defrance 1828, p. 306. Cyrtina cf. burlingtonensis Rowley Plate 1 1 0, figs. 17-19; Plate 111, figs. 1 -2 1893 Cyrtina burlingtonensis Rowley, p. 308, pi. 14, figs. 15-17. 1967 Cyrtina burlingtonensis Rowley; Carter, p. 354, pi. 34, figs. 1-8. 1968 Cyrtina burlingtonensis Rowley; Rodriguez and Gutschick, p. 1030, pi. 128, figs. 10-16. The type species is not recorded by either Carter (1967), who presumes it to be in the University of Illinois collections, or Rodriguez and Gutschick (1968). However, these authors fully describe the specimens they ascribe to the species, the latter pair utilizing good silicified material from Montana, U.S.A. True Cyrtina species are poorly known from British Lower Carboniferous rocks, species such as septosa Phillips, dorsata M’Coy, and carbonaria M’Coy now being placed in the genus Davidsonina Schuchert and Le Vene 1929. A specimen figured by Davidson (1863, pi. 52, fig. 15) and called Spiriferina insculpta (Phillips) should be identified as a Cyrtina , possibly conspecific with the Manifold specimen. Within the British Museum (Nat. Hist.) collections there are a few specimens, believed to be conspecific, from Visean rocks at Treak Cliff and Parkhouse Hill, Derbyshire, and from Wetton, Staffordshire. Brunton has a large collection of similar silicified specimens from Co. Fermanagh, Ireland, which are being described elsewhere. Unfortunately the present study has yielded only one complete specimen, a poorly preserved dorsal valve and a few minute specimens possibly belonging to this species (PI. Ill, figs. 3-5). For this reason the specific designation must be in doubt. How- ever, externally the complete specimen seems identical to those described by Carter or by Rodriguez and Gutschick. The specimen is 6-7 mm in over-all length, 8 0 mm wide, and 5-0 mm high, dimensions which fall very close to those given by Carter (1967, p. 357). Internal morphology is known only from the one incomplete dorsal 834 PALAEONTOLOGY, VOLUME 17 valve (PL 1 1 1, fig. 2). However, from what can be seen it is evident that the cardinalia resembles that illustrated by Rodriguez and Gutschick (1968). The inner socket ridges closely confine the triangular, highly sculptured and ventrally directed myophore face of the cardinal process. The crural bases are poorly differentiated from the inner socket ridges and are joined to the valve floor at the base of the cardinal process. There is a low, indistinct median ridge. This small Cyrtina species does seem to be rare in British (exclusive of Ireland) Lower Carboniferous rocks; even in regions of prolific brachiopods, collected over many years, the number of specimens recorded or seen in collections is small. Superfamily spiriferacea King 1 846 Family spiriferidae King 1846 Genus fusella M’Coy 1844 Type species. Spirifera fusiformis Phillips 1836, p. 210, pi. 9, figs. 10, 1 1, from Holland, Yorkshire. Specimen in the Gilbertson Collection of the British Museum (Nat. Hist.), B 249. Diagnosis. Small (up to approx'. 30 mm wide), fusiform (transversely narrowly rhombic) spiriferides with sharply pointed cardinal extremities and finely denticulate high ventral interarea. Ventral sulcus bordered by a pair of prominent ribs plus seven to ten additional ribs. Dental plates short, subparallel and fused umbonally by shell thickening which fills delthyrial apex. Dorsal fold may be weak with indistinct median rib. Discussion. The characteristics of this genus have been ill-defined owing to the rarity of the type species and the name has been used in the past for a variety of differing spiriferide forms. Conversely in recent years other genera have been described which are closely related to Fusella , such as Amesopleura Carter 1967 (? = Voiseyella Roberts 1964). The genus Mirifusella Carter 1971 is said to be ‘most similar to Fusella M’Coy’ (Carter 1971, p. 250). In fact the two differ considerably in their external morphology, Mirifusella tending to be longer than wide and having rounded cardinal extremities. A recent restudy of the type specimen plus silicified material from Ireland allows a more complete description of the genus and this is to be pub- lished elsewhere by C. H. C. B. The genus is represented in the present fauna by a single incomplete specimen, plus some fragments, believed to be F. rhomboidea (Phillips), a species closely related to F. fusiformis (Phillips). EXPLANATION OF PLATE 111 Figs. 1 -2. Cyrtina cf. burlingtonensis Rowley. Detail of the dorsal cardinalia (SEM x 50) of an incompletely preserved valve (SEM x 15) from locality A. BB 60270. Figs. 3-5. ICyrtina sp. Lateral, posterior (SEM x 30), and ventral (SEM x 20) views of a young shell from locality Al. BB 60279. Figs. 6-7. Fusella rhomboidea (Phillips). Dorsal and ventral views of the incomplete shell from locality B. x 3. BB 60271. Figs. 8-13. Spiriferellina perplicata (North). 8-9, exterior and interior of a ventral valve from locality C. x 3. BB 60274. 10-11, ventral and posterior views of a slightly crushed specimen from locality A. x 3. BB 60272. 12, dorsal view of an incomplete juvenile shell from locality B. x 3. BB 60275. 13, fragment of a dorsal valve interior showing the ridges bounding the adductor muscle scars, x 4. BB 60273. Figs. 14-16. Girtyella saccula (J. de C. Sowerby). 14, dorsal view of a slightly crushed shell from locality A. x 3. BB 60278. 15-16, lateral and dorsal view of a specimen from locality Al. x 3. BB 60277. PLATE 111 BRUNTON and CHAMPION, Carboniferous brachiopods 836 PALAEONTOLOGY, VOLUME 17 Fusella rhomboidea (Phillips) Plate 111, figs. 6-7 1836 Spirifera rhomboidea Phillips, p. 217, pi. 9, fig. 8. 1858 Spirifera convoluta var. rhomboidea Phillips; Davidson, p. 35, pi. 5, fig. 2. Type specimen. Spirifera rhomboidea Phillips from Bolland, Yorkshire. Gilbertson Collection, British Museum (Nat. Hist.), B 236. Diagnosis. Relatively large Fusella with prominent ribs, fold, and sulcus. Dorsal fold with low median rib and first pair of bordering ribs bifurcating close to umbo. Ventral interarea high and strongly curved, delthyrium narrow and teeth supported by short, narrowly divergent and vertically subparallel dental plates which converge slightly medianly. Discussion. Details of the external ornamentation are lacking from the type specimen but can be seen on the silicified Manifold specimen to consist simply of variously developed growth-lines producing a slight imbrication. Otherwise the exterior is smooth. Incomplete silicification has resulted in the loss of most internal features on this specimen. All that can be seen is part of one dental plate and signs of the umbonal shell thickening. Superfamily spiriferinacea Davidson 1884 Family spiriferinidae Davidson 1884 Genus spiriferellina Frederiks 1919 (1924) Type species. Terebratulites cristatus von Schlotheim 1816, p. 28. Lectotype chosen by Campbell (1959, p. 350) and housed in the Geologisch-Palaontologisches Institut und Museum, Berlin. Spiriferellina perplicata (North) Plate 111, figs. 8-13 1920 Spiriferina perplicata North, p. 219, pi. 13, figs, la-c, 10. Type specimen. That figured by North (1920) on plate 13 as fig. la-c , from D[ Zone strata of Treak Cliff, Castleton, Derbyshire. J. W. Jackson Collection, Manchester Museum, L 11601. North (1920, p. 219) gives Peaks Hill as the locality but J. W. Jackson (1952) has pointed out that this is an error and that the true locality is Treak Cliff, as stated on the box containing the specimen. Diagnosis. Transverse Spiriferellina with greatest width at hinge-line. Adult shells with six strongly developed ribs on each side of mid-line on ventral valve. Concentric lamellae most prominent on rib crests. Anterior edges of dental plates diverge slightly to valve floor. Ventral median septum high anteriorly and extending about one-half of length of shell. Cardinal process low, from base of which extends pair of ridges laterally bordering adductor scars. Description. That given by North (1920, p. 219) adequately describes the exterior of this species. The only internal details that he gave indicated the presence of dental plates and a ventral median septum. The morphology of the interiors of both dorsal and ventral umbones are illustrated here (PI. Ill, figs. 9, 13). Discussion. The type species of Spiriferellina was redescribed in 1959 by Campbell BRUNTON AND CHAMPION: CARBONIFEROUS BRACHIOPODS 837 and later Logan (1964) reviewed the history of the classification of Spiriferina d’Orbigny and concluded that Spiriferellina should include S. insculpta (Phillips), S. octoplicata (J. de C. Sowerby), and S. perplicata (North), all Carboniferous species believed to be closely related to the type species, S. cristata , of Permian age. Thus in the British Lower Carboniferous we have two endopunctate spiriferinide genera, Spiriferellina and Punctospirifer North, based upon the species P. scabricosta North; the latter being distinguished externally by its greater number of less angular and lower ribs and relatively wide rounded ventral sulcus. Spiriferellina perplicata is most closely related to S. insculpta (Phillips), but the two differ in that the latter species is more strongly biconvex, owing to having a strongly convex dorsal valve, and the number of ribs on each side of the ventral mid-line of adult specimens is normally only four. On particularly well-preserved specimens of S. insculpta there is a fine spinose ornamentation, a feature which has not been observed on S. perplicata. Internally the two species are similar. Order terebratulida Waagen 1883 Superfamily dielasmatacea Schuchert 1913 Family dielasmatidae Schuchert 1913 Genus girtyella Weller 1911 Type species. Harttina indianensis Girty 1908, p. 293, pi. 19, figs. 6-15. Pella beds, Pella, Iowa, U.S.A. Mississippian. Weller first described his genus Girtyella in 1911 (p. 442), not in 1914 as stated by Muir-Wood (1951) and the Treatise (1965). It was separated from Dielasma on account of its inner hinge plates being supported on a median septum. These plates should be depressed medianly rather than forming a flat cardinal plate as shown in the Treatise (1965: H 754), and in some species it may be that the median septum is low so that the septalium is virtually sessile anteriorly. However, the inner hinge plates do not separate dorso-medianly to expose the valve floor, as in Dielasma. Girtyella saccula (J. de C. Sowerby) Plate 111, figs. 14-16 1824 Terebratula sacculus J. de C. Sowerby, p. 65, pi. 446, fig 1. 1951 IGirtyeHa sacculus (J. de C. Sowerby); Muir-Wood, p. 113, pi. 5, fig. 1 a-c. Type specimen. Lectotype selected by Muir-Wood (1951) from the Sowerby Collection, Lower Carboni- ferous of Derbyshire. British Museum (Nat. Hist.), B 61653. Diagnosis. Small (up to about 15 mm long) strongly biconvex shell with emarginate anterior in adults resulting from opposite ligate to narrowly uniplicate folding. Commonly with marked growth lines anteriorly. Dorsally curved epithyridid ventral umbo. Discussion. This species is rare within the Manifold faunas, being represented by about ten reasonably preserved specimens measuring up to 8 mm long. Two broken specimens show some of the internal structures and enable the assignment of this species to Girtyella. The ventral valve has well-developed dental plates and within the dorsal valve the crural or inner hinge plates are V-shaped and supported medianly on 838 PALAEONTOLOGY, VOLUME 17 a low median septum. Differentiation between this median region and the flanking outer hinge plates and the inner socket ridges is weak, but is indicated by a change in angle at the point from which the crura extend anteriorly (text-fig. 7). The extent of the crura and complete form of the loop have not been observed. The paucity of this species in the Manifold lime- stones is not surprising since normally it is found abundantly only in ‘reefal’ limestones. However, specimens assigned to the species are found widely and vary greatly in external form. The lectotype measures 16-3 mm long, 14-8 mm wide, and 10-8 mm thick. This compares closely with specimens of G. indianensis (Girty), the type species, and falls about midway within the range of sizes seen in museum collections, where specimens reach at least 30 mm long. Sowerby’s figures (1824, pi. 446, fig. 1) illustrate the variation seen within specimens attri- buted to G. saccula; the top-left specimen (B 61654) being small (11 mm long) and ventrally sulcate from a point about 5 mm in front of the umbo. The central figure, the lectotype, is larger and specimens like this tend to develop a ventral sulcus only after a length of at least 10 mm, and the anterior commissure is less strongly folded than that of the small specimens. The top-right illustration is of a younger specimen (6-6 mm long) lacking any prominent sulcation. Specimens in museum collections, said to be from single localities, are indicative of continuous variation between these two forms, but whether we are dealing with intraspecific variation or not is beyond the scope of the present study. As there is no strongly developed ventral sulcus on the few specimens extracted from the Manifold lime- stones they are probably the young of the larger form of G. saccula. v v dp cp c text -fig. 7. An oblique view of the dorsal valve interior plus umbonal region of the ventral valve, of Girtyella saccula (J. de C. Sowerby). The illus- tration is composite from two partly preserved specimens. Ventral valve (vv), dental plate (dp), crural plate (cp), and a broken portion of a crus (c). ADDITIONAL MATERIAL In addition to those specimens already described there are many minute individuals, less than 3 mm long, which are difficult to classify. Some are almost certainly the young of Coledium and are mentioned under that genus. Of the others from locality A (in situ) four specimens may be the young of Cyrtina; the silicified shell looks as if it may be a replacement for endopunctate shell, and the ventral valve is quite strongly convex in profile (PI. Ill, figs. 3-4). Other problematical small specimens appear to have been impunctate, they are ventribiconvex in profile, have a ventral median sulcus, and the narrow ventral interarea has an open delthyrium. The only adult species in the fauna with which these may be related is Crurithyris nastus. Fragments of a Rhipidomella species occur at locality A. Otherwise this genus has been recovered from locality D, where it occurs with some minute indeterminate brachiopod specimens and fragments of the gastropod Anomplialus, and also at locality F. BRUNTON AND CHAMPION: CARBONIFEROUS BRACHIOPODS 839 REFERENCES brunton, c. H. c. 1966. Silicified productoids from the Visean of County Fermanagh. Bull. Br. Mus. nat. Hist. (Geo!.), 12, 173-243, pis. 119. — 1968. Silicified brachiopods from the Visean of County Fermanagh (11). Ibid. 16, 1-70, pis. 1-9. — 1972. Cleiothyridina Buckman, 1906 (Brachiopoda): Proposed validation under the plenary powers. Bull. Zool. Nomencl. 29, 142-144. Campbell, k. s. w. 1959. The type species of three Upper Palaeozoic punctate spiriferoids. Palaeontology, 1, 251-363, pis. 58-60. carter, j. l. 1967. Mississippian brachiopods from the Chappel Limestone of central Texas. Bull. Amer. Paleont. 53 (238), 253-488, pis. 13-45. — 1 97 1 . New early Mississippian silicified brachiopods from Central Iowa. Smithson. Contribn. PaleobioL 3, 245-255, pis. 1-2. cooper, g. a. 1956. New Pennsylvanian brachiopods. J. Paleont. 30, 521-530, pi. 61. davidson, T. 1861-1863. British Carboniferous Brachiopoda. Palaeontogr. Soc. (Monogr.), 2, (5), 81-280, pis. 17-60. ferguson, J. 1966. Variation in two species of the Carboniferous brachiopod Pleuropugnoides. Proc. Yorks. Geol. Soc. 35, 353-374, pi. 23. GEORGE, T. n. 1931. Ambocoelia Hall and certain similar British Spiriferidae. Quart. J. geol. Soc. Loud. 87, 30-61, pis. 3-5. girty, G. h. 1926. Mississippian formations of San Saba County, Texas, III. The macrofauna of the lime- stone of Boone age. U.S. geol. Surv. Prof. Paper , 146, 24-43, pis. 5-6. grant, r. e. 1965. The brachiopod superfamily Stenoscismatacea. Smiths. Misc. Coll. 148, 1-192, pis. 1-24. jackson, J. w. 1952. Catalogue of type and figured specimens in the Geological Department of the Manchester Museum. Mus. Publ. 6, 170 pp. Manchester. koninck, L. G. DE. 1843. Descriptions des animaux fossiles , pp. 241-480. Liege. — 1887. Faune du calcaire Carbonifere de la Belgique. Brachiopodes. Ann Is Mus. r. Hist. nat. Belg. 14(6), i-ix, 1-154, pis. 1-37. leveille, c. 1835. Aperpu geologique de quelques localites tres riches en coquilles sur les frontieres de France et de Belgique. Mem. Soc. geol. France , 2, 29-40, pi. 2. logan, A. 1964. An Indo-pacific spiriferinid from the Triassic of North-eastern British Columbia. Bull. Canadian Petrol, geol. 12, 692-718, pis. 1, 2. ludford, a. 1951 . The stratigraphy of the Carboniferous rocks of the Weaver Hills district. North Stafford- shire. Quart. J. geol. Soc. Lond. 106, 211-230, pi. 16. morris, p. G. 1970. Holothurian spicules from the Lower Carboniferous near Waterhouses, North Stafford- shire. Mercian Geol. 3, 353-359. muir-wood, h. m. 1951. The Brachiopoda of Martin’s ‘Petrificata Derbiensia’. Ann. Mag. nat. Hist. 4, 97-118, pis. 3-6. — and cooper, g. a. 1960. Morphology, classification and life habits of the productoidea (Brachiopoda). Mem. geol. Soc. Amer. 81, 1-447, pis. 1 135. north, f. j. 1920. On Syringothyris Winchell, and certain Carboniferous Brachiopoda referred to Spiriferina D’Orbigny. Quart. J. geol. Soc. Lond. 76, 162-227, pis. 11-13. Parkinson, d. and ludford, a. 1964. The Carboniferous Limestone of the Blore-with-Swinscoe district, northeast Staffordshire, with revisions to the stratigraphy of neighbouring areas. Geol. J. 4, 167-176, pi. 8. Phillips, J. 1836. Illustrations of the geology of Yorkshire. Part II. The Mountain Limestone District. 253 pp., 25 pis. John Murray, London. prentice, j. e. 1951. The Carboniferous Limestone of the Manifold Valley region. North Staffordshire. Quart. J. geol. Soc. Lond. 106, 171-209, pi. 15. ramsbottom, w. h. c. 1970. Carboniferous faunas and palaeogeography of the South West of England region. Proc. Ussher. Soc. 2, 144-157. rodriguez, j. and gutschick, r. c. 1968. Productina, Cyrtina, and Dielasma (Brachiopoda), from the Lodgepole Limestone (Mississippian) of southwestern Montana. J. Paleont. 42, 1027-1032, pi. 128. 840 PALAEONTOLOGY, VOLUME 17 rowley, r. r. 1900. Descriptions of new species of fossils from the Devonian and subcarboniferous rocks of Missouri. Amer. Geol. 25, 261-272, pi. 5. sowerby, j. de c. 1824. The Mineral Conchology of Great Britain, vol. 5, 63-138, pis. 444-485. samtleben, c. 1971. Zur kenntnis der produktiden und spiriferiden des Bolivianischen Unterperms. Beih. geol.Jb. Ill, 1-163, pis. 1-11. 1972. Feinbau und wachstum von spiriferiden-armgerusten. Palaont. Z. 46, 20-33, pis. 5-8. thomas, H. D. and ford, T. D. 1963. A new tabulate coral from the Visean of Derbyshire. Proc. Yorks, geol. Soc. 34, 45-50. weller, s. 1911. Genera of Mississippian loop-bearing Brachiopoda. J. Geol. 14, 439-448. — 1914. The Mississippian Brachiopoda of the Mississippi Valley Basin. Illinois State geol. Surv. Mon. 1, 508 pp., 83 pis. Urbana. williams, a. 1965. In moore, r. c. (ed.). Treatise on Invertebrate Paleontology. Pt. H , Brachiopoda , 927 pp., 746 figs. Kansas. williams, j. s. 1943. Stratigraphy and fauna of the Louisiana Limestone of Missouri. U.S. geol. Sur. Spec. Paper , 203, 1-131, pis. 1-9. Manuscript received 28 September 1973 C. H. C. BRUNTON Department of Palaeontology British Museum (Natural History) London, SW7 5BD C. CHAMPION 4 Chatsworth Drive Little Eaton, Derby, DE2 5AP ADDENDUM (added August 1974) It has proved impossible to contact Dr. P. G. Morris for the purpose of discovering the reference to his paper describing Lambdarina. It seems likely that our paper will now be published before that of Dr. Morris, so in order to reduce taxonomic confusion we have emended our text at this late date to avoid referring to an unpublished paper. At the beginning of 1972 C. H. C. B. had arranged to discuss with Dr. Morris the relationships of our faunas and prospects of joint authorship. However, circum- stances did not allow this meeting until January 1973 when C. H. C. B. saw Dr. Morris, his material and completed script awaiting submission for publication. At that time this paper was in the earliest stages of drafting and we continued in the belief we would be comparing our material to that already published by Dr. Morris. In retaining the name Lambdarina , which we anticipated as by now being published, we acknowledge Dr. Morris’s recognition of these unusual brachiopods in the Waterhouses area. TRILOBITES FROM THE SHOLESHOOK LIMESTONE (ASHGILL) OF SOUTH WALES by DAVID PRICE Abstract. Eight trilobite species based on type material from the Sholeshook Limestone are redescribed and two new species erected. For the first time a topotype pygidium is illustrated for Stenopareia bowmanni (Salter). Tretaspis moeldenensis Cave is retained as a distinct species ; a topotype pygidium of this form is also figured for the first time and histograms given of selected fringe characters. The previously undescribed pygidium of Lehua ? princeps (Reed), is very similar to that of the type species L. vincula (Barrande) and supports the assignment of the species to Lehua made previously on the basis of cranidia only. Topotype material of Pseudosphaerexochus (Pseudosphaerexochus) juvenis (Salter) is distinguished from P. ( P .) octolobatus (M’Coy) and a pygidium tentatively referred to the species. It is not evident that the two syntypes of Sphaerexochusl hoops Salter are specimens of the same species; that from Sholeshook is not a sufficient basis for the recognition of a distinct species. Flexicalymene cavei sp. nov. and Kloucekia extensa sp. nov. are distinguished from other Ashgill species of those genera. Lectotypes are selected for K. robertsi (Reed) and ‘ Acidaspis ' turnbulli Reed. The latter species is tentatively referred to Diacanthaspis; it shows great similarity to D. decacantha (Angelin) of which it may ultimately prove a junior synonym. I n common with many other Upper Ordovician Limestones in Wales, the Sholeshook Limestone of South Wales early attracted the attention of palaeontologists on account of its rich shelly faunas. Trilobites from the formation were among those described and figured by J. W. Salter in an appendix to the Geological Survey Memoir of 1848 and some of these were subsequently redescribed and refigured by Salter in various publications between that date and 1867. Later, between 1904 and 1908, F. R. C. Reed described and figured some of the trilobites from the Sholeshook Limestone col- lected by V. M. Turnbull and then recently presented to the Sedgwick Museum, Cambridge. Most of these older species have not been redescribed; some of them are strati- graphically important. It is the aim of this paper to deal with species erected on type specimens from the Sholeshook Limestone, both from the original type material where this needs redescription and from topotype and other material which has subsequently become available, including material collected by the author. Apart from those included among the early forms described by Salter and Reed, the only other species to have been so far erected on type material from the Sholeshook Limestone is Tretaspis moeldenensis Cave, 1960. The redescription of this species is made desirable both by the recently much-increased knowledge of the genus Tretaspis (Ingham 1970, pp. 39-45) and by the stratigraphical importance attached to it by the author in recent discussion of the age and correlation of the Sholeshook Limestone (Price 1973, pp. 238-239). In addition, new species of Flexicalymene and Kloucekia from the Sholeshook Limestone are erected herein. The stratigraphical terminology used throughout is that of Price (1973). That is to say, the term Sholeshook Limestone is used to include the argillaceous-calcareous or arenaceous-calcareous horizons developed between the Mydrim Shales and the Slade and Redhill Mudstones at Sholeshook (Flaverfordwest, Pembrokeshire) and in the area around Llandowror (Carmarthenshire) and that developed between the type [Palaeontology, Vol. 17, Part 4, 1974, pp. 841-868, pis. 112-116.] H 842 PALAEONTOLOGY, VOLUME 17 Robeston Wathen Limestone and the Slade and Redhill Mudstones at Robeston Wathen. Thus defined, the formation has a diachronous base and overlies the Mydrim Shales unconformably. The base of the normally succeeding Slade and Redhill Mudstones is also diachronous and to the north and west of Haverfordwest this formation contains strata laterally equivalent to the Sholeshook Limestone. The Sholeshook Limestone is thought to range in age from the upper part of Zone 1 to probably Zone 3 of the Cautleyan Stage of the Ashgill Series. These descriptions of type trilobites from the formation are intended as a pre- liminary work to description of the Sholeshook trilobite fauna as a whole. The material upon which this paper is based is housed in the following museums, the prefixes for whose specimen-numbers are indicated in brackets: British Museum (Natural History) (BM), Hunterian Museum, Glasgow (HM), Institute of Geological Sciences (GSM), and the Sedgwick Museum, Cambridge (SM). The following descriptions are based on all available material in these collections. Maps showing detailed localities for specimens collected by the author and for other well-localized material have been given else- where (Price 1973, text-figs. 1-5) and it is to these that the locality numbers cited in the text refer. SYSTEMATIC PALAEONTOLOGY Family illaenidae Hawle and Corda, 1847 Subfamily illaeninae Hawle and Corda, 1847 Genus stenopareia Holm, 1886 Type species. (Original designation) Ulaenus linnarssonii Holm, 1882. Stenopareia bowmanni (Salter, 1 848) Plate 112, figs. 1-8, ?9 1848 Illaenus bowmanni Salter (pars), p. 339, pi. 8, fig. 1, la; non figs. 2, 3. 1851 Dysplanus centrotusl ( Dal.); M’Coy, pi. 1e, figs. 19, 19a. 1866 Illaenus bowmanni, Salter; Salter (pars), pi. 18, fig. 8a; non fig. 8. 1867 Illaenus (Dysplanus) bowmanni Salter; Salter (pars), p. 185, pi. 28, figs. 10, 12; non fig. 7. 1885 Illaenus bowmanni Salt.; Marr and Roberts, pp. 480, 481. 1909 Illaenus bowmanni Salt. ; Strahan et al., table p. 58. 1914 Illaenus bowmanni Salt. ; Strahan et al., table p. 63. 1933 Illaenus bowmanni Salter; Reed, pp. 121-135. 1938 Illaenus ( Dysplanus ) bowmanni Salter; Stubblefield, p. 33. 1961 Stenopareia bowmanni (Salter); Whittard, p. 217, pi. 31, figs. 1, 2. 1970 Stenopareia cf. bowmanni (Salter); Ingham, pp. 23-25, pi. 4, figs. 1-7, ?8. 1973 Stenopareia bowmanni (Salter); Price, tables 1-4. Lectotype. Subsequently designated and refigured by Whittard (1961), GSM 24553, original of Salter 1848, pi. 8, fig. 1, la; from the Sholeshook Limestone, probably at Sholeshook. (The locality may possibly have been Llandowror; Stubblefield (1938, p. 33, footnote) refers to some slight doubt. In the present author’s opinion, the lithology of the matrix would strongly indicate Sholeshook as the locality.) Horizons and localities. Around Llandowror the species is known from about 14 m above the base of the Sholeshook Limestone in the Mylet Road section (locality 24a of Price 1973) and similar horizons at Craig-y-deilo quarry (localities 18c, d) through to horizons near the top of the formation at Laynor (locality 20). At Sholeshook it ranges from about the middle of the Sholeshook Limestone through to the highest part of the formation and the basal part of the overlying Slade and Redhill Mudstones at Prendergast (localities 8a, b). An incomplete pygidium (PI. 112, fig. 9) from an outcrop of ‘Bala Limestone’ at Lron, near Whitland, Carmarthenshire (see Strahan et al. 1914, fig. 5) appears also to belong with this species. PRICE: ASHGILL TRILOBITES 843 Description. Cranidium semi-elliptical to sub-parabolic in dorsal outline, generally about nine-tenths as long (sag.) as wide (tr.), moderately convex transversely; in longitudinal profile the convexity increases as the cranidium is declined more steeply anteriorly (PI. 1 12, figs. 2, 3). Anterior margin strongly and evenly rounded. Glabella with only slight independent convexity (tr.) occupies about one-half total cranidial width (tr.) at posterior margin. Axial furrows deep and rounded in profile posteriorly, shallowing anteriorly and extending to between one-third and one-half the cranidial length; sub-parallel over most of length but sigmoidally curving at first adaxially and then outwards before dying out (PI. 1 12, figs. 4-6). On internal moulds, a sharp furrow across the posterior end of the glabella represents the inner margin of the posterior doublure (PI. 1 12, figs. 1, 4). Immediately anterior to this is a broad (sag. and exsag.) and very shallow (?occipital) furrow, only clearly seen near the axial furrows; deep pits, probably representing articulating processes, are developed where it crosses these (PI. 112, fig. 4). Posterior border furrows broad (exsag.) and prominent; narrowest adaxially, broadening outwards; behind them are narrow (exsag.), convex borders. Palpebral lobes small, situated far back at about their own length (exsag.) from the posterior margin. Short posterior branches of facial suture apparently directed straight back from palpebral lobes; anterior branches running directly forwards at first, then gradually converging in smooth curves (PI. 112, fig. 5) to become confluent antero-ventrally. Free cheeks sub-triangular, almost twice as long (exsag.) as wide (tr.), convex (tr. and exsag.), broadly rounded at genal angles. Eyes short (exsag.) and strongly convex abaxially (PI. 1 12, fig. 5), standing rather high and dropping steeply to the general level of the free cheeks. Doublure broad and flexed close to ventral surface of exoskeleton. Surface of cranidium usually smooth or, in some internal moulds, finely granular. Thorax, hypostoma, and rostral plate not known from Sholeshook material. Pygidium transversely sub-parabolic in dorsal outline, about three-quarters as long (sag.) as wide (tr.), only gently convex transversely; in longitudinal profile (PI. 1 12, fig. 7) only gently convex over anterior two-thirds of length, then dropping steeply posteriorly before flattening slightly near the posterior margin. Axis occupies about one-third of total width (tr.) anteriorly. Axial furrows broad (tr.) and shallow at anterior margin; very ill-defined behind but appearing to converge rapidly. Fulcrum about half-way between furrows and lateral margins. Beyond this, anterior margins of pleural lobes deflected back through about 45° and then gradually curving into the lateral margins. Facets narrow and very steeply declined. Antero-lateral corners of pleural lobes partly crossed by broad, shallow furrows extending obliquely back from anterior ends of axial furrows. Doublure broad, occupying about one-half total pygidial length along saggital line where partially crossed by faint shallow groove (on ventral mould) running from posterior margin; with fine, closely-spaced, sub- parallel terrace-lines. Inner margin incompletely seen but scalloped, apparently with three broad (tr.) lobes separated by distinct cusps (PI. 1 12, fig. 8). Discussion. In his description of topotype material of ‘ Illaenus ’ bowmanni, Reed (1933, pp. 124-125) described in some detail two pygidia from the Geological Survey collections which he presumed were those used by Salter for his description in the 1867 Monograph. Only one of these, the smaller, GSM 24555 (PI. 112, figs. 7, 8 of 844 PALAEONTOLOGY, VOLUME 17 this paper) is here accepted as belonging to S. bowmanni. The other, GSM 24557 (a partial external mould), is considered to belong to the species listed by the author (Price 1973, tables 1 -4) as Illaenus ( Parillaenus ) cf .fallax Holm, where the form of the pygidium is known from a complete articulated specimen. Cranidia described and figured by Ingham (1970, pp. 23-25, pi. 4, figs. 1-3, 6, 7) from Zones 2 and 3 of the Cautley Mudstones are closely similar to those described above, though no pygidia are known from South Wales which show traces of trans- verse grooves on the pygidial axis as does the original of Ingham’s plate 4, fig. 4. This pygidium, however, resembles those figured here in general form and pro- portions and Ingham’s material appears acceptable to the present author as S. bowmanni. Two complete specimens, SM A41852a, b and A41853, from the Ashgill Series of Llanwddyn in the Berwyn Hills, originals of M’Coy (1851, pi. 1e, figs. 19, 19 a) one of which was refigured by Whittard (1961, pi. 31, fig. 2), are also accepted as belong- ing here. Family trinucleidae Hawle and Corda, 1847 Subfamily tretaspinae Whittington, 1941 Genus tretaspis M’Coy, 1849 Type species. Asaplms seticornis Hisinger, 1840. Tretaspis moeldenensis Cave, 1960 Plate 112, figs. 10-12; Plate 113, figs. 1-4; text-fig. 1 1909 Trinucleus fimbriatus Murch. ; Strahan et al., p. 56. 1960 Tretaspis moeldenensis Cave, pp. 334-337, pi. 10, figs. 1-7. 1961a Tretaspis kiaeri Stprmer radialis Lamont; Dean (pars), pp. 122-125. 1962 Tretaspis kiaeri St^rmer radialis Lamont; Dean, p. 86, pi. 9, figs. 2-4. 71970 Tretaspis cf. moeldenensis Cave; Ingham, pp. 54-55, pi. 8, figs. 21-26; pi. 9, figs. 1-7; text- figs. 14c, 19. 1973 Tretaspis moeldenensis Cave; Price, tables 3 and 4. Holotype. Figured by Cave 1960, pi. 10, figs. 1 and 3, SM A50668, from the basal Sholeshook Limestone of Moldin, near Llandowror (locality 25). EXPLANATION OF PLATE 112 Figs. 1-8. Stenopareia bowmanni (Salter). 1, SM A77827a, internal mould of cranidium, low Sholeshook Limestone of Llandowror (locality 18c), dorsal view, x 1. 2, SM A31496, internal mould of cranidium, highest Sholeshook Limestone of Prendergast Place (locality 8b), Haverfordwest, lateral view, x I. 3, SM A77905a, internal mould of cranidium, Sholeshook Limestone, Sholeshook, lateral view, x 1. 4, SM A77573, internal mould of cranidium, horizon and locality as for fig. 2, dorsal view, x 1. 5, BM 1.16446, internal mould of cranidium and right free cheek, Sholeshook Limestone of Sholeshook cutting, dorsal view, x 1 . 6, BM 1. 1 6433, internal mould of cranidium, horizon and locality as for fig. 5, dorsal view, x 1 . 7 and 8, GSM 24555, internal mould of incomplete pygidium, horizon and locality as for fig. 3, lateral and dorsal views, x 2. Fig. 9. ? Stenopareia bowmanni (Salter), SM A31642, internal mould of incomplete pygidium, ‘Bala Lime- stone’, quarry at Fron, near Whitland (Carmarthenshire), dorsal view, x 1. Figs. 10 12. Tretaspis moeldenensis Cave, SM A35500a, internal mould of incomplete cephalon, basal Sholeshook Limestone of Moldin (locality 25), near Llandowror, lateral, dorsal, and anterior views, x 4. PLATE 112 PRICE, Ashgill trilobites 846 PALAEONTOLOGY, VOLUME 17 Horizons and localities. The species is restricted to a thin basal horizon of the Sholeshook Limestone in its development around Llandowror. Apart from Moldin, it is known also from road sections near Mylet and Pentre-howell (localities 24a and 17). Description. Cephalon sub-semicircular in dorsal outline. Pseudofrontal lobe of glabella occupies almost two-thirds total cephalic length (sag.); surface smooth; convex (tr. and sag.) but never approaching sub-spherical, rather longer (sag.) than broad (tr.) in dorsal view, bearing small, apically situated median tubercle. Occipital ring narrow and strongly convex (sag. and exsag.), oriented postero-dorsally, curving forwards abaxially. Occipital furrow broad (sag. and exsag.) and shallow mesially, abaxially containing deep, ovoid apodemal pits. Ip lateral glabellar lobes short (tr.), gently convex (exsag.), abaxially rounded. Ip lateral furrows in form of strongly oblique shallow slots, diverging posteriorly. 2p lobes only gently convex (exsag.), set transversely, narrowest adaxially, broadening outwards. 2p furrows in form of large, shallow, ovoid, anteriorly diverging depressions which constrict the posterior margin of the pseudofrontal lobe. 3p lateral furrows not seen. Axial furrows broad (tr.), particularly posteriorly; anteriorly containing small, deep fossulae. Genal lobes sub-quadrant shaped, smooth, moderately convex (tr. and exsag.), not overhanging fringe; bearing lateral tubercles, rather larger than the median tubercle, at about the level of the 2p lateral furrows; dropping steeply to broad (exsag.) posterior border furrows which abaxially contain large posterior fossulae. Posterior borders set trans- versely, narrow (exsag.) and only gently convex adaxially; becoming broader and more prominent outwards where deflected postero-laterally. Fringe broad, com- prising steeply inclined, convex genal roll and well-developed, gently concave brim; there are long genal prolongations and slender genal spines. Internal moulds show a broad, deep girder with strong, closely spaced terrace-lines. In front of the axial furrows seven pit-arcs are present, E1_2, Ix_4, In (in the notation of Ingham 1970, p. 40). On one specimen (PI. 113, fig. 3) I4 is also present in front of the glabella but on all other specimens up to three I4 pits (half-fringe) are absent frontally (text-fig. 1). The I4 arc persists laterally to around R13 where it merges with the In arc. All pits pits in E | I4 extends to (R no.) 5 — | n=io pits in I4 missing frontally pits in posterior row text-fig. 1. Histograms of selected fringe characters in Tretaspis moeldenensis Cave, topotype material including that utilized by Cave (1960). n is the sample number for each character chosen. All histograms show half- fringe data. (Where columns on the horizontal scale are numbered alternately, this is to allow for half-pits in the count, i.e. a pit on the centre-line.) PRICE: ASHGILL TRILOBITES 847 arranged in strict radial alignment until genal prolongations are reached; on the upper lamella, pits of arcs E2, E1? and Ix are contained in deep radial sulci. On the genal prolongations there is a tendency for these sulci to contain only the E4 and E2 pits and the general radial alignment of the fringe breaks down due to the inter- calation of extra pits between the Ix and I2 arcs; the fringe, thus expanded, may have from 1 1 to 14 pits in the posterior row. The Ex arc contains from 29 to 30 pits in the half-fringe. In the posterior row, and occasionally in the posterior-most two rows, the E2 pit is absent. Generally concentric ridges (lists) are not developed, or are only very weakly developed, between the internal pit-arcs on the upper lamella. On the lower lamella the internal pits are arranged in strong radial sulci (PI. 112, figs. 10-12) which, on the genal prolongations, become disrupted by the additional pits inserted between the I4 and I2 arcs. Thorax unknown. Pygidium (PI. 113, fig. 4) sub-triangular in outline, broad (tr.), the sagittal length only about one-third of the maximum width; postero-lateral margins moderately convex; bluntly rounded posteriorly. Axis narrow, only about one-fifth maximum pygidial width (tr.), tapers gradually posteriorly at about 25°, only moderately convex (tr.). Ring furrows shallow, gently arched forward; each contains a pair of deep apodemal pits a short distance from the axial furrows; axis bears eight such pairs of pits in all, the posterior-most pair usually only weakly developed. Axial furrows shallow. Pleural lobes flat, usually with four faintly defined, broad (exsag.) pleural ribs visible. Discussion. Dean (1961a, 1962, see synonymy) described a form from the Pusgillian Stage of Cross Fell which he identified as T. kiaeri radialis Lamont and synonymized with T. moeldenensis. Ingham ( 1970, pp. 55-57) has indicated, however, that T. radialis may differ from T. moeldenensis and be more likely to prove identical with a closely related but distinct and stratigraphically younger form present in both the North of England and South Wales, the T. cf. radialis of Ingham (1970, pp. 55-57, pi. 9, figs. 8- 20; text-figs. 14 d, 19) and Price (1973, tables 1-4). Dean’s material, which has a non- reticulate, not highly inflated pseudofrontal glabellar lobe, only six pit-arcs in front of the glabella and apparently similar pit counts, is extremely like that described here and Dean’s synonymy with T. moeldenensis seems acceptable to the present author, this name being best retained pending clarification of the relationship with T. radialis. Ingham (1970, see synonymy) has described a closely similar form from the highest Pusgillian and Cautleyan Zone 1 of the Cautley Mudstones. This form differs primarily in that the I4 pit-arc is complete frontally in well over half the sample (Ingham 1970, p. 55; text-fig. 19). In the author’s opinion, this form and that from Cross Fell discussed above are probably both to be regarded as belonging to local populations of the geographically widespread species T. moeldenensis. Two other related forms are T. colliquia Ingham (1970, pp. 53-54, pi. 8, figs. 8-20; text-fig. 146) from the lower part of the Pusgillian Stage at Cautley and the form described by Dean (1961, p. 125, pi. 8, figs. 2, 6-8) as T. kiaeri duftonensis from the Pusgillian Stage of Cross Fell. The former differs in possessing a narrower fringe and a faintly reticulated glabella and in lacking an extensively developed I4 arc; the latter has a much higher peripheral pit count and, again, the pseudofrontal glabellar lobe is reticulated. 848 PALAEONTOLOGY, VOLUME 17 Family cheiruridae Hawle and Corda, 1847 Subfamily cheirurinae Hawle and Corda, 1847 Genus lehua Barton, 1916 Type species. Cheirurus vinculum Barrande, 1872. Lehual princeps (Reed, 1908) Plate 113, fig. 12 1908 Typhloniscus princeps Reed, p. 433, pi. 14, figs. 1-3. 1916 Lehua princeps (Reed); Barton, p. 129. 1958 Typhloniscus princeps Reed; Whittard, p. 115. 1971 Lehual princeps (Reed); Lane, p. 36, pi. 7, fig. 17a, b. 1973 Lehual princeps (Reed); Price, table 2. Holotype. (By monotypy) SM A3 161 7a, b, counterpart moulds of cranidium, from the Sholeshook Lime- stone of Sholeshook railway cutting. Discussion. Lane (1971) has refigured the holotype and redescribed this and another cranidium from the same locality. Contrary to Reed (1908) and Whittard (1958), he has considered this species to possess visual organs, the palpebral lobes being placed far forwards, opposite the frontal lobe, and adjacent to the axial furrows. The present author has examined the two cranidia in question and agrees with Lane’s interpreta- tion. No further material of the cranidium has become available but an incomplete pygidium (PI. 1 13, fig. 12) from the Sholeshook Limestone of Sholeshook railway cutting appears to belong with this species. The axis is moderately convex (tr.), wide (tr.) anteriorly, and tapers back at about 40° ; it comprises three axial rings and a terminal piece. Rings narrowest (sag. and exsag.) and most convex over mesial region which is transverse; abaxially they widen considerably, are less convex, and are slightly deflected posteriorly. Convex (tr.) terminal piece sub-parabolic in out- line; reaches posterior margin. Ring-furrows over most of course broad and deep, abaxially narrowing and deflected posteriorly. Axial furrows broad (tr.) and deep adjacent to axial rings, much shallower over region of terminal piece. Pleural regions narrow; three broad ribs continue beyond them as broad, flat pleural spines. Inter- pleural furrows deep, curving outwards and backwards and widening (exsag.) abaxially. Pleural furrows short (tr.) and deep, commencing at inner anterior corners of ribs and curving outwards convex posteriorly. Surface ornamentation of large, scattered granules absent in furrows. This pygidium is very similar in general form to that of the type species L. vincula (Barrande) described by Prantl and Pribyl (1947, p. 19, pi. 3, figs. 1, 2) from the Dobrotiva Formation (Llandeilo) of Bohemia. It differs in the axial region in lacking paired tubercles on the axial rings and in having a larger and less-pointed terminal piece. The incompleteness of the pleural spines on the Sholeshook specimen prevents more detailed comparison but the author feels that the otherwise strong similarity with the pygidium of the Bohemian form supports the assignment of the species to the genus Lehua made by previous workers on the basis of cranidia only. PRICE: ASHGILL TRILOBITES 849 Subfamily eccoptochilinae Lane, 1971 Genus pseudosphaerexochus Schmidt, 1881 Subgenus pseudosphaerexochus Schmidt, 1881 Type species. Sphaerexochus hemicranium Kutorga, 1854. Pseudosphaerexochus ( Pseudosphaerexochus ) juvenis (Salter, 1848) Plate 113, figs. 5-7, 8?, 9? 1848 Sphaerexochus juvenis Salter (pars), pp. 344-345, pi. 7, figs. 1, 1 a, 2, 3, 3a; non fig. 3 b. 1852 Cheirurus clavifrons Dalman; Salter, p. 3. 1864 Cheirurus ( Actinopeltis ) juvenis Salter; Salter (pars), pp. 67-69, pi. 5, figs. 10, 11 ; non figs. 9, 12. 1866 Cheirurus juvenis Salter; Salter, p. 323, pi. 18, figs. 1, 2. 1881 Pseudosphaerexochus juvenis (Salter); Schmidt, p. 152. 1881 Pseudosphaerexochus juvenis Salter; Salter, p. 521, pi. 18, figs. 1, 2. 1885 Cheirurus juvenis Salt. ; Marr and Roberts, lists pp. 480, 481. 1914 Cheirurus juvenis Salt. ; Strahan et al„ table, p. 63. 1938 Pseudosphaerexochus juvenis (Salter); Stubblefield, p. 32. 1965 Pseudosphaerexochus juvenis (Salter); Whittington, pp. 40-41, pi. 12, figs. 2-5, 8, 15. 71968 Pseudosphaerexochus sp. ind.; Whittington, pp. 102-104, pi. 31, figs. 13, 17. 1971 Pseudosphaerexochus (Pseudosphaerexochus) juvenis (Salter); Lane, p. 47. 1973 Pseudosphaerexochus (Pseudosphaerexochus) juvenis (Salter); Price, tables 1-4. Lectotype. Subsequently designated by Whittington 1965, p. 40, GSM 24534, internal mould of cranidium from the Sholeshook Limestone of Sholeshook; figured Whittington 1965, pi. 12, figs. 2, 4, 8. Horizons and localities. Cranidia of the type regarded below as belonging to this species range from about the middle of the Sholeshook Limestone at Sholeshook (locality 9e and railway cutting) to the high Sholes- hook Limestone at Prendergast (locality 8c) and also occur in low horizons of the Sholeshook Limestone around Llandowror (localities 18b, c). Pygidia tentatively referred to this form are known from the railway cutting at Sholeshook and the low Sholeshook Limestone of Craig-y-deilo quarry (locality 18c). Discussion. The range in morphology of cranidia of Pseudosphaerexochus found in the Sholeshook Limestone is considerable. Some clearly belong to P. ( P .) octolobatus (M'Coy) and have the characteristic sub-circular, longitudinally convex glabella seen in material described by Whittington ( 1 965, pp. 37-40, pi. 10, figs. 14, 15, 1 7-20 ; pi. 11, figs. 1-13) from the Rhiwlas Limestone and by Lane (1971, pp. 48-50, pi. 8, figs. 1 -8) from the Starfish Bed of Girvan. At the other extreme of the morphological range are cranidia in which the glabella is much longer (sag.) than wide (tr.), frontally sub-parabolic in outline and in lateral profile does not overhang the anterior border (PI. 113, figs. 5-7). Cranidia with these characteristics the present author has con- sidered to belong to P. (P.) juvenis. The distinction of this form from P. (P.) octo- lobatus and from other species on the basis of the cranidium, however, is complicated by the poor preservation of the lectotype and the great range of variation shown by Sholeshook specimens (see also comments by Whittington 1965, p. 41). Many of the cranidia are to some extent distorted (PI. 113, fig. 6), but others with an elongate, frontally parabolic glabella not overhanging the anterior border show no signs of dis- tortion (PI. 113, fig. 5). There are also cranidia which show glabellar characters inter- mediate between those just described and those of P. (P.) octolobatus (see PI. 113, fig. 8 here and Whittington 1965, pi. 12, figs. 3, 5, 15). The problem is further complicated by the presence in the Sholeshook Limestone 850 PALAEONTOLOGY, VOLUME 17 (as noted by Whittington 1968, p. 104, pi. 31, fig. 17) of pygidia apparently indis- tinguishable from that of a form from the Lower Tre-wylan Beds of the Llansantffraid- ym-Mechain district ( Pseudosphaerexochus sp. ind. of Whittington 1968, pp. 102- 104, pi. 31, fig. 13). There is a strong possibility that these are the pygidia of P. (P.) juvenis, particularly as there is, as yet, no other cheirurid cranidium known from the Sholeshook Limestone to which a pygidium has not been assigned. Because, however, no entire exoskeleton is known from Sholeshook and because of the difficulty of comparing the poorly preserved cranidium of the Llansantffraid specimen with the cranidia here referred to P. ( P .) juvenis , the present author is inclined to follow Whittington’s (1968, p. 104) note of caution and to regard the pygidium only tenta- tively as that of P. (P.) juvenis. A further example of this pygidium is figured here (PI. 113, fig. 9). Further collecting may clarify the status of the Llansantffraid specimen. The cranidium of the form described by Whittington (1965, p. 39, pi. 11, fig. 14; pi. 12, figs. 1, 6, 7) from the Dolhir Mudstones is not like those here thought to belong to P. (P.) juvenis and the form of the pygidium of this specimen is not clearly seen; it appears possible to the present author that this and the Llansantffraid specimen may represent different species. Pseudosphaerexochus ? hoops (Salter, 1864) Plate 113, figs, 10, 11 1851 Ceraurus clavifrons (Dal. Sp.); M’Coy (pars), p. 154, pi. If, fig. 12. 1864 Sphaerexochus! boops Salter, p. 79, pi. 6, figs. 27, 28. 1873 Sphaerexochus boops Salter; Salter, p. 50. 1891 Sphaerexochus boops J. W. Salter; Woods, p. 151. Discussion. Sphaerexochus ? boops was founded by Salter (1864) on two syntypes which are refigured here. One of these, GSM 24560 (PI. 113, fig. 10) is the poorly EXPLANATION OF PLATE 113 Figs. 1-4, Tretaspis moeldenensis Cave. 1, HM A9803a, incomplete cephalon with upper lamella of fringe preserved, basal Sholeshook Limestone of Moldin (locality 25), near Llandowror, dorsal view, x 4. 2, HM A9804a, two superimposed partial cephala with portions of fringe upper lamellae preserved, horizon and locality as for fig. 1 , antero-lateral oblique view, x 4. 3, SM A50670a, partial cephalon with portion of fringe upper lamella preserved showing frontally complete I4 pit-arc, horizon and locality as for fig. 1, anterior view, x 4. 4, SM A77717, internal mould of pygidium, horizon and locality as for fig. 1, dorsal view, x4. Figs. 5-7. Pseudosphaerexochus ( Pseudosphaerexochus ) juvenis (Salter). 5, SM A77569, incomplete internal mould of cranidium, middle part of Sholeshook Limestone (locality 9e), Sholeshook, dorsal view, x 2. 6 and 7, SM A3 141 8, internal mould of cranidium, Sholeshook Limestone of Sholeshook railway cutting, dorsal and lateral views, x 2. Figs. 8-9. ? Pseudosphaerexochus ( Pseudosphaerexochus ) juvenis (Salter). 8, BM It. 9226, internal mould of cranidium, horizon and locality as for fig. 5, dorsal view, x 2. 9, SM A3 1405b, cast from external mould of incomplete pygidium, horizon and locality as for figs. 6-7, dorsal view, x 3. Figs. 10-11. Pseudo sphaerexochus j boops (Salter). 10, GSM 24560, partial internal mould of cranidium, syntype, Sholeshook Limestone of Sholeshook, dorsal view, x3. Original of Salter 1864, pi. 6, fig. 27. 11, SM A41905, crushed internal mould of cranidium, syntype, Applethwaite Beds of Applethwaite Common, dorsal view, x 3. Original of M’Coy 1851, pi. If, fig. 12 and Salter 1864, pi. 6, fig. 28. Fig. 12. Lehual princeps (Reed), SM A31451, internal mould of incomplete pygidium, horizon and locality as for figs. 6-7, dorsal view, x 4. PLATE 113 PRICE, Ashgill trilobites 852 PALAEONTOLOGY, VOLUME 17 preserved, partial internal mould of a cranidium from the Sholeshook Limestone of Sholeshook. The incompleteness of the specimen makes comparisons difficult but the present author considers that it could possibly be a partial cranidium of P. ( P .) octolobatus (M’Coy). Lane (1971, ph 8, fig. 1 u, b) has recently figured a specimen of that species in which the basal lateral glabellar lobes are prominent, large and rounded in dorsal view and in which the occipital furrow is arched forward mesially though not as strongly as in the Sholeshook specimen. The Sholeshook specimen, however, does show clear signs of distortion. This specimen cannot serve as a suffi- cient basis for the recognition of a distinct species. It is far from evident that the other syntype, SM A41905 (PI. 113, fig. 1 1), a more complete but badly crushed cranidium from the Ashgill of Applethwaite Common, which also appears to belong to Pseudosphaerexochus , is conspecific with the Sholes- hook specimen. The glabella of this specimen is oval in dorsal outline; the basal lateral glabellar lobes are relatively smaller than in the Sholeshook specimen, occupy- ing well under one-third of the glabellar width (tr.) on that level, and angular antero- laterally. The lp lateral furrows are wider and more distinct abaxially, shallowing as they curve back and reaching the occipital furrow as broad (tr.), indistinct depressions. The 2p lateral lobes appear to be slightly shorter (exsag.) than the lp and the 3p lateral lobes about two-thirds the length (exsag.) of the 2p. The 2p and 3p furrows are shallow; other details of lobation are not clear. The glabella is ornamented, except in the furrows, with closely and evenly spaced tubercles of about 0-1 mm. Clarification of the affinities of this specimen, and so of the status of Sphaerexochusl boops Salter, must await the collection and description of more complete and better- preserved material from Applethwaite Common. Family calymenidae Edwards, 1840 Genus flexicalymene Shirley, 1936 Type species. (Original designation) Calymene caractaci Salter, 1865. Flexicalymene cavei sp. nov. Plate 114, figs. 1-15 1914 Calymene sp. ; Strahan et al., table p. 63, list p. 67, and table p. 70 (pars). 1960 Flexicalymene sp.; Cave, p. 334. 1973 Flexicalymene sp. nov.; Price, tables 1-4. Holotype. SM A57050, internal mould of cranidium (PI. 114, figs. 1, 2) from the basal Sholeshook Lime- stone of Moldin, near Llandowror (locality 25). Horizons and localities. Not yet known from the basal few metres of the Sholeshook Limestone at Sholeshook or from the basal Slade and Redhill Mudstones to the north and west of Haverfordwest (localities 1-6). Otherwise ranges through the Sholeshook Limestone from the basal Moldin horizon (locality 25) through to the highest horizons at Prendergast (locality 8b), Robeston Wathen ( 10a), and Faynor (22) and into the overlying Slade and Redhill Mudstones of Prendergast (8a) and Glog-y-fran (locality 15). Diagnosis. Glabella (excluding occipital ring) rather broader (tr.) than long, extend- ing slightly further forwards than convex parts of fixed cheeks, lp lateral lobes sub- quadrate in outline, with acutely angular antero-lateral corners; lp furrows not clearly bifurcating. 2p lobes elongated antero-laterally; 3p lobes small, transverse; PRICE: ASHGILL TRILOBITES 853 3p furrows very faint. Broad, steeply upturned anterior border. Eyes rather far for- ward, mid-length of palpebral lobes opposite antero-lateral corners of 2p lateral lobes; thorax of 13 segments. Pygidium with 5 axial rings and 4 well-marked pleural bands. Description. Cephalon sub-semi-elliptical in dorsal outline, almost 2\ times as wide (tr.) as long (sag.). Glabella, excluding occipital ring, occupying about two-thirds of cephalic length (sag.); sub-parabolic in dorsal outline, rather broader (tr.) than long (sag.); strongly convex transversely; in lateral profile dropping anteriorly in increas- ingly steep convex curve (PI. 114, figs. 2, 6). Occipital ring widest (sag. and exsag.) mesially where arched gently forward; abaxially, posterior to the lp lateral glabellar lobes, it curves forwards narrowing slightly. Occipital furrow broad (sag. and exsag.) and distinct mesially, a shallow U in profile; abaxially it curves around the posterior margins of the lp lateral lobes, narrowing as it nears the axial furrows, lp lateral glabellar lobes large, sub-quadrate in form with acutely angular antero-lateral corners ; total length (exsag.) about one-third that of glabella (excluding occipital ring). The lobes have an independent convexity and drop steeply outwards to the axial furrows; this slope becomes less steep antero-laterally. lp lateral furrows strongly curved posteriorly, deepest abaxially, shallowing inwards and continuing back as broad, shallow furrows to separate the basal lateral lobes from the median lobe of the glabella; not clearly bifurcating though the outline of the 2p lateral lobes is slightly constricted at their inner posterior margins. 2p lateral lobes only half the length (exsag.) of the lp lobes; highest part of lobe has a sub-circular outline and is separated from the median lobe by a slight depression; from this area there is a moderately steep slope antero-laterally which has the effect of giving the lobe an elongate- oblique outline (PI. 114, figs. 1, 3, 4). 2p lateral furrows posteriorly convergent, shallowing inwardly. 3p lobes small and set transversely, not distinctly separated from median lobe by a depression. 3p lateral furrows very faint. Frontal lobe moderately convex (tr. and sag.), much broader (tr.) than long (sag.), extending slightly further forwards than the convex parts of the fixed cheeks; the width (tr.) across its base is only slightly less than across the 3p lateral lobes. Behind, the median lobe narrows slightly posteriorly. Axial furrows deep, wide (tr.) and distinct, gradually converging anteriorly and with gentle, antero-laterally convex curvatures; ending in distinct antero-lateral pits situated slightly nearer to the 3p lateral furrows than to the anterior margin of the frontal lobe. Pre-glabellar area composed of short (sag. and exsag.), gently concave anterior border furrow and steeply upturned border (PI. 1 14, figs. 2, 6). Anterior parts of fixed cheeks moderately convex (tr.) but standing much lower than the level of the glabella (PI. 1 14, fig. 7). The slope down from the cheeks to the axial furrows becomes increasingly steep posteriorly and at the sharply angular inner posterior corner of the cheek it meets a similar steep slope down to the posterior border furrow; this latter slope becomes less steep abaxially. Posterior border furrows deep and distinct adaxially, shallowing outwards. Posterior borders narrow (exsag.) and strongly convex adaxially, broadening, flattening, and curving gently posteriorly outwards. Palpebral lobes rather far forward, their mid-lengths on the level of the antero-lateral ends of the 2p lateral lobes. Suture gonatoparian; anterior branches run forward from the eye in gradually converging gentle curves; posterior 854 PALAEONTOLOGY, VOLUME 17 branches run outwards for a short distance in posteriorly convex curves then are deflected to curve strongly postero-laterally. Free cheeks sub-quadrant shaped, gently convex (tr. and exsag.) with broad (tr.), moderately convex lateral border and broad, shallow border furrow. Flypostoma (PI. 1 14, fig. 8) with elongate (sag.), strongly convex (tr.) median body bearing a prominent pair of large maculae at about the mid-length; the anterior section is sub-parallel sided, behind the maculae the sides are strongly tapered posteriorly. Anterior margin transverse. Well-developed anterior wings. Lateral border furrow broad (tr.); broad, flat lateral border widest (tr.) posteriorly, dying out in front of maculae. Bifurcated posterior margin with prominent median notch. Thorax of thirteen segments. Axis strongly convex (tr.), occupying one-third of total width (tr.) anteriorly, tapering only gradually posteriorly. Axial rings strongly convex (sag. and exsag.), arched forward mesially and curving again forwards abaxially to end in sub-quadrilateral axial lobes. Articulating furrows broad (sag. and exsag.), shallowest mesially, abaxially becoming deep slots between the axial lobes. Axial furrows broad (tr.) and rather shallow. Pleurae horizontal over inner portion; distally deflected strongly downwards and curved gently forwards (PI. 114, fig. 9). Broad (exsag.), distinct pleural furrows commence near the axial furrows and run roughly parallel to the posterior margin dividing each pleura into a broad (exsag.), adaxially convex posterior band and a narrow anterior pleural band. Anterior bands bear, at the fulcrum, prominent articulating bosses (PI. 114, fig. 9) which fit corresponding sockets in the posterior band of the next anterior pleura. Pleural furrows die out before reaching smooth, flattened, distally rounded pleural tips (PI. 114, fig. 10). Pygidium sub-triangular in dorsal view, convex (tr.), widest (tr.) on level of fourth or fifth axial ring, tapering strongly back to form an obtuse angle at the posterior margin. Axis strongly convex (tr.), tapering gradually posteriorly, formed of five axial rings and a sub-triangular terminal piece. Rings convex (sag. and exsag.), their dorsal surfaces rather flat in lateral view (PL 114, fig. 14); ring furrows broad and prominent. There are four well-defined pleural ribs, strongly curved posteriorly, narrowest adaxially, and broadening slightly outwards. Narrow (exsag.) pleural EXPLANATION OF PLATE 114 Figs. 115. Flexicalymene cavei sp. nov. 1 and 2, SM A57050, holotype, internal mould of cranidium, basal Sholeshook Limestone of Moldin (locality 25), near Llandowror, dorsal and right-lateral views, x 2. 3 and 4, HM A9686a, b, internal mould and cast from external mould of almost entire but slightly crushed cephalon, low Sholeshook Limestone, track south of Craig-y-deilo quarry, Llandowror (locality 18d), dorsal views, x 2. 5-7, SM A3 1390, internal mould of cranidium, Sholeshook Limestone of Sholeshook railway cutting, dorsal, left-lateral and anterior views, x2. 8, BM It. 9256b, cast from external mould of hypostoma, middle section of Sholeshook Limestone (locality 9e), Sholeshook, dorsal view, x 4. 9-11, SM A77824b, cast from external mould of almost complete articulated exoskeleton, Sholeshook Limestone horizon at Robeston Wathen (locality 10a), dorso-lateral oblique, right-lateral oblique, and posterior-oblique views, x 2. 12, SM A57052, internal mould of pygidium, horizon and locality as for tigs. 1-2, dorsal view, x 2. 13 and 14, SM A31393, internal mould of pygidium, horizon and locality as for figs. 5-7, dorsal and right-lateral views, x 2. 15, detail of specimen illustrated in figs. 9-1 1 to show ornamentation, anterior axial rings, x 10. PLATE 114 PRICE, Ashgill trilobites 856 PALAEONTOLOGY, VOLUME 17 furrows run along the ribs in a position rather nearer to the posterior margin; these are strong abaxially but much weaker near the axial furrows. Anteriorly the pleural regions bear narrow, convex articulating bands with articulating bosses similar to those on the thoraxic segments. A broad (tr.), gently convex, sub-parallel sided post- axial ridge extends forward to the anterior margin of the terminal piece. The dorsal exoskeleton, with the exception of the major furrows, is uniformly ornamented with closely packed, irregularly spaced small granules, the largest about 0T25 mm. Discussion. F. cavei sp. nov. is readily distinguished from two other British Ashgillian species F. brevicapitata (Portlock 1843) and IF. quadrata (King 1923). The former species from the Killey Bridge Beds(?) of Desertcreat (see Shirley 1931, pp. 28-31, pi. 2, figs. 9, 10), has a shorter (sag.) glabella which does not extend as far forward as the fixed cheeks, transverse 2p lateral lobes, only slightly oblique lp and 2p lateral furrows of which the lp furrow bifurcates strongly, rather less prominent 3p lateral lobes, and a frontal glabellar lobe which rises very steeply from the pre-glabellar field; the 3p lateral furrows are stated to be absent. F.l quadrata (King 1923, pp. 504- 505, pi. 26, figs. 1, 2) from the Ashgill Series of the south-west Berwyns differs in having a glabella which is frontally more quadrate in dorsal view, more strongly convex in lateral profile, and with almost equi-sized 2p and 3p lateral lobes and in possessing only twelve thoracic segments. As Dean (1962, p. 113) has commented, the systematic position and affinities of this form are rather uncertain. Although the position of the eyes is rather far forward in F. cavei, it is not so far forward as in forms from the Maysville and Richmond Stages of the Cincinnati area, Ohio, usually assigned to F. meeki (Foerste 1910, p. 84, pi. 3, fig. 18; 1919, pi. 18, fig. 3) or related species such as F. retrorsa (Foerste 1910, p. 85, pi. 3, fig. 19; 1919, pi. 18, fig. 2); that is, opposite the 3p lateral lobes. Other differences also, parti- cularly the less anteriorly convergent glabellar outline and the relatively longer frontal glabellar lobe, suggest that F. cavei is not closely related to these North American forms. Family encrinuridae Angelin, 1854 Subfamily encrinurinae Angelin, 1854 Genus encrinuroides Reed, 1931 Type species. (Original designation) Cybele sexcostata Salter, 1848. Encrinuroides sexcostatus (Salter, 1848) Plate 1 1 5, figs. 1-8 1848 Cybele sexcostata Salter, p. 343, pi. 8, fig. 10; non fig. 9. 1852 Encrinurus sexcostatus Salter, p. 4, pi. 1g, figs. 6, 7. 1853 Encrinurus sexcostatus Salter, pp. 1-5, pi. 4, figs. 1-12. 1866 Encrinurus sexcostatus Salter; Salter, p. 324, pi. 19, figs. 5, 6. 1909 Encrinurus sexcostatus (Salt.); Strahan et al., table p. 58. 1914 Encrinurus sexcostatus (Salt.); Strahan et al., table p. 63. 1938 Encrinurus sexcostatus Salter; Stubblefield, p. 35. 1950 Encrinuroides sexcostata (Salter); Whittington, pp. 535-538, pi. 68, figs. 7 16; text-fig. 2. 1960 Paracybeloides cf. loveni (Linnarsson); Cave (pars), faunal list p. 334. PRICE: ASHGILL TRILOBITES 857 1965 Encrinuroides sexcostatus (Salter); Whittington, pp. 42-43, pi. 12, figs. 9-14; pi. 13, figs. 1,2. 1973 Encrinuroides sexcostatus (Salter); Price, tables 1-4, 7. Neotype. SM A30695a, b, counterpart moulds of entire specimen, selected and figured by Whittington 1950, p. 535, pi. 68, figs. 7, 9, 10. Horizons and localities. The species is abundant throughout most of the Sholeshook Limestone from the basal Moldin horizon of the Llandowror area upwards but has not yet been found in the highest horizons of the formation such as those at Prendergast (localities 8b, c) or Robeston Wathen (locality 10a) or in the overlying Slade and Redhill Mudstones. It does occur, however, in the basal Slade and Redhill Mudstones of localities (2, 3, 4) to the north and west of Haverfordwest. Discussion. Material collected by the present author is similar to that described by Whittington (1950, 1965) from both the Sholeshook and the Rhiwlas Limestones. Cranidia show the pre-glabellar and longitudinal furrows (PL 115, figs. 2, 3) and free cheeks the anterior furrows (PI. 115, figs. 4, 5) described by Whittington in 1950 (his fig. 2). The eye-stalks are distally expanded with visual surfaces composed of closely packed hexagonal lenses and surrounded by a prominent furrow (PI. 115, figs. 5, 6). All the larger and well-preserved pygidia have seven pleural ribs and external moulds show that these and the axial rings are finely granulated. Family dalmanitidae Vogdes, 1890 Genus kloucekia Delo, 1935 Type species. (Original designation) Phacops phillipsii Barrande, 1846. Kloucekia robertsi (Reed, 1904) Plate 115, figs. 9-14; Plate 116, figs. 1, 2 1904 Phacops robertsi Reed, pp. 106-109, pi. 5, figs. 1-7. 1904a Phacops ( Dalmanites ) robertsi ; Reed, p. 384. 1905 Dalmanites robertsi ; Reed, p. 177. 1914 Phacops robertsi Reed; Strahan et al., tables pp. 63, 75. 1973 Kloucekia robertsi (Reed); Price, tables 1, 2; p. 242. Lectotype. (Here selected from among Reed’s syntypes), SM A30706, internal mould of incomplete cephalon, original of Reed 1904, pi. 5, fig. 1, from the basal Redhill Mudstones of Prendergast Place (locality 8a), Haverfordwest. Horizons and localities. The species is restricted to the uppermost part of the Sholeshook Limestone and the base of the overlying Slade and Redhill Mudstones at Prendergast (localities 8a, b) and the Slade and Redhill Mudstones of a similar horizon at Redhill quarry (locality 7). Description. Cephalon sub-semicircular in dorsal outline. Glabella sub-pentagonal in outline, widest (tr.) across frontal lobe where maximum width is about equal to sagittal length (including occipital ring); only gently convex (tr.). Occipital ring convex (sag. and exsag.), in lateral profile stands above rest of glabella (PI. 115, fig. 9), broadest (sag. and exsag.) mesially where posterior margin is gently concave, abaxially curves slightly forwards. Occipital furrow shallow, broad (sag. and exsag.), and gently arched forward mesially; abaxial portions in form of deep, broad (exsag.) apodemal slots set slightly oblique, posteriorly divergent, with their distal-most parts rather shallower. Ip lateral glabellar lobes sub-quadrilateral in outline, broadening (exsag.) abaxially, anterior margins gently convex. Ip lateral furrows strong, set i 858 PALAEONTOLOGY, VOLUME 17 slightly oblique, anteriorly divergent, except for short (tr.) adaxial portions which are deflected to become anteriorly convergent; main part of furrow a straight, deep apodemal slot with the adaxial-most portion shallower. 2p lateral lobes sub-parallel sided, both anterior and posterior margins slightly convex in such a way that the maximum width (exsag.) occurs near the adaxial end. 2p furrows shallow, transverse, gently curved, convex anteriorly; slightly deeper adaxially. 3p furrows shallow, sub- linear, strongly oblique, anteriorly divergent, so that 3p lobes rapidly broaden (exsag.) abaxially. All three pairs of furrows end adaxially close to the saggital line, leaving a narrow (tr.), sub-parallel sided median lobe indistinctly separated from the lateral lobes. Median lobe flat in lateral profile. Frontal lobe about twice as broad (tr.) as long (sag.), drops anteriorly in steep convex slope (PI. 115, fig. 9) to very narrow (sag. and exsag.) but distinct anterior border furrow. Anterior border narrow (sag. and exsag.), sloping less steeply forward than frontal glabellar lobe. Axial furrows broad (tr.), moderately deep, diverging more strongly forward in front of the lp lateral lobes. Cheeks sub-triangular in outline, dropping steeply to shallow lateral border furrows which merge with the less steep lateral borders. Cheeks drop near vertically to posterior border furrows which are deep and broad (exsag.) adaxially where they are set very slightly oblique, anteriorly divergent, but die out as they curve gently towards the genal angles. Genal angles bearing short (exsag.), rapidly tapering spines (PI. 115, fig. 14; see also Reed 1904, pi. 5, fig. 4). Eye-lobes large, anterior edges in line with antero-lateral corners of 3p glabellar lobes, posterior edges in line with adaxial ends of lp furrows. Palpebral lobes sub-semicircular; indistinct furrows separate broad (tr.), prominent palpebral rims which stand on about same level as glabella (PI. 115, fig. 11). Visual surfaces sloping steeply, bearing circular, convex facets. Postero-laterally to the eye-lobes are broad (exsag.) semi-crescentic depres- sions (PL 115, figs. 10, 12) which die out as they cross the posterior branches of the facial suture. These latter run outwards and slightly forwards from the eye, sub- parallel to the posterior cheek margin. Anterior branches run forwards and slightly EXPLANATION OF PLATE 115 Figs. 1-8. Encrinuroides sexcostatus (Salter). 1, SM A31484, internal mould of cranidium, Sholeshook Limestone of Sholeshook railway cutting, dorsal view, x 2. 2, SM A77562, internal mould of partial cranidium, middle part of Sholeshook Limestone (locality 9e), Sholeshook, antero-lateral oblique view, x 2. 3, SM A77560, internal mould of cranidium, horizon and locality as for fig. 2, anterior view, x 2. 4, HM A9735, internal mould of right free cheek, low Sholeshook Limestone of Llandowror (locality 18d), dorsal view, x 3. 5 and 6, HM A9713, internal mould of right free cheek, horizon and locality as for fig. 4, anterior view, x 4 and detail x 10 to show form of visual surface of eye. 7, SM A77834b, cast from external mould of pygidium, low Sholeshook Limestone of Craig-y-deilo quarry, Llandowror (locality 18b), dorsal view, x 3. 8, SM A77922, part of cast from external mould of pygidium, horizon and locality as for fig. 2, dorsal view, x 6 to show ornamentation. Figs. 9 14. Kloucekia robertsi (Reed). 9-11, SM A30706, lectotype, internal mould of incomplete cephalon, basal Slade and Redhill Mudstones of Prendergast Place (locality 8a), Haverfordwest, lateral, dorsal, and anterior views, x 2. Original of Reed 1904, pi. 5, fig. 1. 12, SM A77928, internal mould of cephalon, horizon and locality as for figs. 9-11, dorsal view, x 2. 13, internal mould of eye-lobe, horizon and locality as for figs. 9-1 1, oblique view, x 4 showing details of visual surface. 14, SM A77933, part of internal mould of right fixed cheek showing form of genal spine, horizon and locality as for figs. 9-11, dorsal view, x 2. PLATE 115 12 11 13 PRICE, Ashgill trilobites 860 PALAEONTOLOGY, VOLUME 17 outwards and then around the frontal glabellar lobe, joining anteriorly in a smooth curve coincident with the anterior border furrow. Cephalic doublure about as broad anteriorly as anterior border, narrowing rather posteriorly. Thorax of eleven segments; tapering only gradually posteriorly in over-all width (tr.). Axis moderately convex (tr.), occupies about one-third of total width anteriorly. Dorsal surface of axial rings rather flat in lateral profile. Rings gently arched forward mesially, where narrowest (sag. and exsag.), broadening abaxially to form prominent axial lobes which are sub-quadrilateral in dorsal outline. Articulating furrows deep, broad (sag. and exsag.), rounded basally; shallowest mesially, deepening to form apodemal slots abaxially. Axial furrows broad (tr.). Inner parts of pleurae horizontal ; outer parts deflected steeply ventrally and curving gently forwards, the tips smooth and abaxially rounded. Broad (exsag.), deep, basally rounded pleural furrows com- mence at the axial furrows, separating narrow posterior from wider (exsag.), adaxially convex anterior pleural bands, then run parallel to posterior margins of pleurae over most of their length, curving slightly forwards and dying out before reaching the pleural tips. Pygidiuin sub-semicircular in dorsal outline, moderately convex (tr.); maximum width (tr.) on level of fourth axial ring. Axis moderately convex (tr.), anteriorly rather over one-quarter the maximum width (tr.) of the pygidium, strongly tapered posteriorly, extending almost as far back as the anterior margin of the doublure. At least nine axial rings present on figured specimen (PI. 116, fig. 2); these strongly convex (sag. and exsag.), separated by broad ring furrows which are shallowest over the mesial third but become much deeper and slot-like abaxially; both become pro- gressively less prominent posteriorly. Axial furrows broad (tr.), converge posteriorly at about 25°. Five broad (exsag.) pleural ribs are visible with broad, shallow rib furrows running along the mid-line and separated by broad, deep pleural furrows. Both sets of furrows die out laterally to leave a broad (tr.), smooth border. Doublure broad, widest anteriorly, narrowing slightly posteriorly (PI. 1 16, fig. 2). Discussion. K. robertsi differs from the type species K. phiUipsii (Barrande) (see Whittington 1 962, pp. 7-9 ; text-fig. 2) in having a relatively shorter (sag.) and broader (tr.) cephalon with a narrower (sag. and exsag.) occipital ring, a broader anterior border and longer (exsag.) palpebral lobes with their posterior ends relatively nearer to the axial furrows. The pygidium is similar in outline and convexity but differs in having a greater number of axial rings. It differs from the two British Caradocian EXPLANATION OF PLATE 116 Figs. 1-2. Kloucekia robertsi (Reed). 1, SM A56041, internal mould of complete articulated exoskeleton, highest Sholeshook Limestone of Prendergast Place (locality 8b), Haverfordwest, dorsal view, x 2. 2, SM A30709, internal mould of pygidium, horizon and locality as for fig. 1, dorsal view, x 2. Original of Reed 1904, pi. 5, fig. 7. Figs. 3-5. Diacanthaspisl turnbulli (Reed). 3, SM A30722, lectotype, incomplete internal mould of arti- culated exoskeleton, Sholeshook Limestone of Sholeshook, dorsal view, x 3. Original of Reed 1905, pi. 4, fig. 4. 4, GSM Pg.277, cast from external mould of incomplete thorax and anterior part of pygidial axis, Sholeshook Limestone, middle section of Sholeshook railway cutting, dorsal view, x 4. 5, SM A3 1 363, internal mould of incomplete cranidium and external mould of partial thorax, Sholeshook Limestone of Sholeshook, dorsal view, x 4. PLATE 116 PRICE, Ashgill trilobites 862 PALAEONTOLOGY, VOLUME 17 species K. harnagensis (Bancroft) (see Dean 1961, pp. 321-324, pi. 49, figs. 9, 14; pi. 50, figs. 1-5) and K. apiculata (M’Coy) (see Whittington 1962, pp. 9-12, pi. 1; pi. 2, figs. 1-3, 5-7, 9-12) in that the cephalon is again relatively broader and shorter and has a more rounded anterior glabellar margin. The pygidium is more bluntly rounded posteriorly with a less convex and posteriorly less well-defined axis and lacks the caudal spine. Kloucekia extensa sp. nov. Text-fig. 2 a-k 1973 Kloucekia sp. nov.; Price, tables 3, 4. Holotype. HM A9682, external mould of incomplete cranidium (latex cast figured as text-fig. 2a) from the low Sholeshook Limestone of Craig-y-deilo quarry, Llandowror (locality 18c). Other material. Nine incomplete cranidia and four pygidia, all from the low Sholeshook Limestone of either Craig-y-deilo quarry (localities 18b, c, d) or the Mylet road section (localities 24a, b), near Llandowror. Diagnosis. Glabella attaining maximum width at level of 3p lateral lobes; 2p and 3p lateral lobes partly coalesced, short 2p lateral furrows not reaching axial furrows; 3p lateral furrows with moderately strong, even, posteriorly convex curvature; frontal lobe with strongly rounded anterior margin. Axial furrows narrow (tr.), with distinct change of course at level of lp lateral furrows. Pygidium sub-semi-elliptical, gently convex (tr.) with flatter border, at least eight axial rings and six pleural ribs. Axis well defined posteriorly. Description. Glabella moderately convex (tr.), longer (sag.) than wide (tr.), widest at level of mid-length of 3p lateral glabellar lobes. Occipital ring standing rather higher than rest of glabella in lateral profile; wide (sag. and exsag.) mesially, where the anterior margin is arched forward; abaxially narrowing and curving strongly forwards. Occipital furrow broad and shallow mesially, abaxially containing deep apodemal slots which are present in the internal mould as ovoid pits. 1 p lateral glabellar lobes short (exsag.), sub-quadrilateral in dorsal outline, narrowest adaxially, broaden- ing outwards, lp lateral furrows in form of broad (exsag.), distinct slots set slightly oblique, anteriorly divergent, except for short (tr.) adaxial sections which are posteriorly divergent. 2p furrows narrow (exsag.), set transversely or very slightly oblique, posteriorly divergent; deepening abaxially, not reaching axial furrows. 2p and 3p lateral glabellar lobes thus fused abaxially. 2p lobes broadest (exsag.) near adaxial ends, narrowing slightly outwards. 3p furrows shallow and indistinct adaxi- ally, broadening slightly outwards; set oblique, anteriorly divergent, with moderately strong, posteriorly convex curvatures. 3p lobes thus broadening abaxially. Frontal lobe of glabella sub-semi-elliptical in outline, sagittal length slightly less than three- quarters of maximum width which is attained on level of abaxial ends of 3p lateral furrows; anterior margin strongly rounded; in lateral profile dropping only gently anteriorly and flattening out near the anterior margin (text-fig. 2b). Central lobe of glabella narrow (tr.), flat in lateral profile. Axial furrows narrow (tr.) and shallow but quite distinct. Posteriorly they are curved around the abaxial ends of the lp lateral glabellar lobes then, on the level of the lp furrows, are deflected to run forwards and strongly outwards, at first linearly but opposite the anterior region of the 3p lateral lobes curving gradually adaxially. Palpebral lobes sub-semicircular with PRICE: ASHGILL TRILOBITES 863 text-fig. 2a- k. Kloucekia extensa sp. nov. a-b , HM A9682, holotype, cast from external mould of incom- plete cranidium, low Sholeshook Limestone of Craig-y-deilo quarry, Llandowror (locality 18c), dorsal and right-lateral views, x 4. c, SM A78010, internal mould of partial cranidium, 9^-10 m above base of Sholeshook Limestone in Mylet road section (locality 24a), near Llandowror, dorsal view, x4. d , HM A9703, cast from external mould of partial cranidium, low Sholeshook Limestone, track south of Craig-y-deilo quarry, Llandowror (locality 18d), dorsal view, x 4. e, HM A9698, internal mould of incom- plete cranidium, horizon and locality as for 2d, dorsal view, x 4. /, BM In. 54701, cast from external mould of partial cephalon, low Sholeshook Limestone of Craig-y-deilo quarry (locality 18b or c), Llandowror, dorsal view, x 4. g , SM A77967, internal mould of incomplete cranidium, 14 m above base of Sholeshook Limestone in Mylet road section (locality 24a), near Llandowror, dorsal view, x 4. h, HM A9605, internal mould of small pygidium, higher part of Mylet road section (locality 24b), Llandowror, dorsal view, x 4. i, HM A9612a, internal mould of pygidium, horizon and locality as for 2 h, dorsal view, x 4. j, HM A9606, internal mould of pygidium, horizon and locality as for 2 h, dorsal view, x4. k, HM A9612b, cast from external mould of incomplete pygidium, horizon and locality as for 2 h, dorsal view, x 4. 864 PALAEONTOLOGY, VOLUME 17 narrow but distinct furrows separating long (exsag.), broad (tr.) palpebral rims. Anterior margins of eyes opposite antero-lateral corners of 3p lateral lobes, posterior margins opposite anterior parts of lp lobes. Free cheeks drop steeply antero-laterally to broad, shallow, and indistinct border furrow and flat, near horizontal border. Anterior border narrow (sag. and exsag.). Posterior border furrow narrow (exsag.) and slot-like adaxially. Pygidium sub-semi-elliptical in dorsal outline, sagittal length about three-quarters of maximum width which occurs at about level of fourth axial ring. Axis moderately convex (tr.) ; anteriorly occupies about one-third of total pygidial width. Tapers posteriorly at about 25°, apparently extending to the posterior margin. Axial rings strongly convex (sag. and exsag.), narrowest mesially, broadening outwards ; separated by broad and mesially shallow ring furrows which deepen and become slot-like abaxially. At least eight such rings are present followed by an apparently smooth, convex (tr.) terminal piece. Axial furrows broad (tr.). Pleural lobes gently convex (tr.), with flatter border; crossed by six broad (exsag.) pleural ribs with broad, shallow inter-pleural furrows running roughly along the mid-line and separated by deeper pleural furrows. Both sets of furrows die out before reaching the lateral margin to leave a broad, smooth border. Generally both internal and external moulds are smooth but one cranidium (text- fig. 2c) retains on the occipital ring and basal lateral glabellar lobe scattered tubercles of about 0-1 mm diameter. Discussion. Several features of the glabellar region of K. extensa sp. nov.— the strong divergence of the axial furrows in front of the lp lateral furrows, the level at which the maximum width is attained, the posteriorly convex, even curvature of the 3p lateral furrows and the relatively long (sag.) and anteriorly strongly rounded frontal lobe— are characteristic and serve to distinguish it from other species of the genus. It differs in all these respects from K. poueyti Destombes (1972, pp. 57-58, pi. 15, figs. 1-3, text-fig. 20) from the Upper Ashgill of Morocco and differs from this form also in that the pygidial axis is much better defined posteriorly and contains two or three more axial rings. The pygidium is similar in over-all form to those of K. phillipsii (Whittington 1962, text-fig. 2/) and K. robertsi (PI. 116, fig. 2 of this paper), differing mainly in being relatively longer (sag.) and, again, with the axis better defined at the posterior margin ; that of K. phillipsii bears only five axial rings. Family odontopleuridae Burmeister, 1843 Subfamily odontopleurinae Burmeister, 1843 Genus diacanthaspis Whittington, 1941a Type species. (Original designation) Diacanthaspis cooperi Whittington, 1941a. Diacanthaspis ? turnbulli ( Reed, 1905) Plate 1 16, figs. 3-5 1905 Acidaspis ( Ceratocephala ) turnbulli Reed, pp. 99-100, pi. 4, figs. 4-7. 1973 Diacanthaspis ? turnbulli (Reed); Price, tables 1, 2. PRICE: ASHGILL TRILOBITES 865 Remarks. Reed (1905) appears to have described Aeidaspis turnbulli on the basis of two specimens, that figured on his pi. 4, fig. 4 and (p. 99) ‘another specimen exhibiting some of the missing features’. These appear to be the only specimens of Reed’s species in the Sedgwick Museum collections (one further specimen has been added by the author), other material from the Slade and Redhill Mudstones to which he refers on p. 100 having been referred to the form listed by the author (Price 1973, table 4) as Primaspis a IT. semievoluta (Reed). Lectotype. (Here selected from Reed’s two syntypes), SM A30722, internal mould of incomplete, arti- culated exoskeleton (PI. 1 16, fig. 3) from the Sholeshook Limestone of Sholeshook; original of Reed 1905, pi. 4, fig. 4. Other material. SM A3 1363, internal mould of incomplete cranidium and external mould of partial thorax from the Sholeshook Limestone of Sholeshook, second specimen mentioned by Reed (1905, p. 99, see above); GSM Pg. 277, external mould of incomplete thorax from middle section of Sholeshook railway cutting; SM A77944, very poor internal mould of partial cranidium from Sholeshook (locality 9e). Description. Cranidium broadest (tr.) posteriorly, moderately strongly convex (tr.). Sub-parallel sided median lobe of glabella defined by broad longitudinal furrows which are shallowest opposite the 2p lateral lobes; expanding (tr.) forwards into frontal lobe. Occipital furrow and occipital ring broad (sag. and exsag.), form of latter not clearly seen. Ip lateral glabellar lobes convex (tr.) with elongate-ovoid out- lines, set slightly anteriorly divergent. Ip lateral furrows only faintly developed, anteriorly divergent; at their inner ends, where they meet the longitudinal furrows, shallow pits are developed. 2p lateral lobes with elongate-ovoid outlines ; aligned sub- parallel. Axial furrows deep posteriorly where strongly anteriorly divergent, in front of mid-lengths of 1 p lobes curving gradually adaxially ; only faintly developed opposite the 2p lateral lobes. Fixed cheeks sub-crescentic in outline, convex (tr.), narrow (tr.) anteriorly, broadening, and becoming more strongly convex posteriorly, reaching a maximum width opposite the posterior parts of the lp lateral lobes and then narrow- ing and curving strongly adaxially towards the occipital ring. Cheeks drop steeply postero-laterally to the posterior border furrows. Posterior borders narrow (exsag.) and strongly convex adaxially, curving gently forwards and very rapidly broadening (exsag.) outwards. Narrow, convex eye-ridges, separated from the cheeks by broad furrows, curve backwards to the palpebral lobes which are small and situated opposite the posterior parts of the Ip lateral lobes. Anterior branches of facial suture appear to run close to eye-ridge over most of its length; posterior branches curve outwards and strongly backwards to cross the posterior border near its adaxial end (part of the librigenal posterior border is present on the lectotype, PI. 116, fig. 3). There appears to be a broad (sag. and exsag.) anterior border furrow. Thorax of nine segments. Axis strongly convex (tr.), occupies less than one-third of total width (tr.) anteriorly and tapers only very gradually backwards. Axial rings broad (sag. and exsag.) and arched slightly forwards mesially, abaxially they curve forwards to form sub-quadrilateral axial lobes separated by deep slots which are the abaxial continuations of the mesially shallow and broad (sag. and exsag.) articulating furrows. Axial furrows broad (tr.) and shallow. Inner portions of pleurae straight and horizontal, composed of narrow (exsag.), convex anterior and broad, strongly convex posterior pleural bands separated by broad pleural furrows. The posterior pleural 866 PALAEONTOLOGY, VOLUME 17 bands bear two prominent tubercles, one a short distance out from the axial furrow, the other at the fulcrum and give rise to long, gradually tapering, outwardly and back- wardly directed pleural spines (PI. 1 1 6, fig. 4). These pleural spines are rather broadened at their bases, adjacent to the fulcrum. Pygidium incompletely known. First and second axial rings well defined (PI. 1 16, fig. 3), appearing each to bear paired tubercles. Five pairs of posterior border spines are clearly visible on the lectotype (PI. 116, fig. 3) and there is room for a sixth pair posteriorly. Discussion. The form described above is extremely similar in many respects to Diacanthaspis decacantha (Angelin) described by Kielan (1960, pp. 103-106, pi. 15, figs. 1-3; pi. 16, figs. 2, 3; pi. 17, figs. 7, 8; text-fig. 27) from the Upper Ordovician of Sweden and Poland (see also Bruton, 1966, pp. 11, 12, pi. 2, figs. 7, 8) and by Whitting- ton (1962, pp. 23-24, pi. 5, figs. 9, 16, 17, 20; 1965, pp. 33-34, pi. 9, figs. 1-10; 1968, pp. 99-100, pi. 30, fig. 24) from the Rhiwlas Limestone of North Wales. Thecranidium is similar in all visible features (though Kielan’s material does not show the eye- ridge) except that the longitudinal furrow is much more pronounced in the lectotype of D. ? turnbulli , very distinctly separating the 2p lateral lobe from the median lobe of the glabella. In this respect, however, the lectotype differs from the two other cranidia known, which are more like D. decacantha , and the furrow has probably been emphasized by distortion. The thorax is strikingly similar and clearly shows the thickening of the pleural spines near the fulcrum (see moulds of ventral surfaces of spines in PI. 116, fig. 3) which has been regarded (Whittington 1968, p. 99) as a feature characteristic of D. decacantha. It seems probable that when more and better-preserved material is available (including the free cheek and pygidium), Acidaspis turnbulli Reed will prove to be a junior synonym of D. decacantha (Angelin). The likelihood of this is increased by the similarity of many of its associates in the Sholeshook Limestone to forms associated with D. decacantha in both Poland and North Wales (see Price 1973, table 2). Acknowledgements. Most of the work on which this paper is based was undertaken during the tenure of a N.E.R.C. Research Studentship in the Department of Geology, University College, London. The author wishes to acknowledge the help of the following in arranging the loan of museum material : Drs. C. L. Forbes and R. B. Rickards (SM), Dr. J. K. Ingham (HM), Mr. S. F. Morris (BM), and Dr. A. W. A. Rushton (GSM). Dr. Rushton prepared the pygidium figured here as Plate 1 12, fig. 8 to show the inner margin of the doublure and kindly drew the author’s attention to this feature of the specimen. Professor H. B. Whitting- ton kindly read the manuscript and suggested useful modifications. REFERENCES angelin, N. p. 1854. Palaeontologia Scandinavian I: Crustacea formationis transitionis. Fasc. 2, 21-92, pis. 25-41. Lund. barrande, J. 1846. Notice preliminaire sur le Systeme Silurien et les Trilobites de Boheme. Leipzig. — 1872. Systeme Silurien de centre du centre de la Boheme. lere partie. Recherches paleontologiques. Supplement au vol. 1. barton, D. c. 1916. A revision of the Cheirurinae, with notes on their evolution. Wash. Univ. Stud, scient. Ser. 3 (4), 101-152, 1 chart. bruton, d. l. 1966. A revision of the Swedish Ordovician Odontopleuridae (Trilobita). Bull. Geol. Instn Univ. Uppsala, 43, 1-40, pis. 1-6. PRICE: ASHGILL TRILOBITES 867 burmeister, H. 1843. Die Organisation der Trilobiten aus ihren lebenden Verwandten entwickelt ; nebst einer systematischen Uebersicht aller zeither beschriebenen Arten. Berlin. cave, r. 1960. A new species of Tretaspis from South Wales. Geol. Mag. 97, 334-337, pi. 10. dean, w. T. 1961. The Ordovician trilobite faunas of South Shropshire, 2. Bull. Br. Mus. nat. Hist. ( Geol .), 5 (8), 313-358, pis. 49-55. — 1961a. Trmucleid trilobites from the higher Dufton Shales of the Caradoc Series in the Cross Fell Inlier. Proc. Yorks, geol. Soc. 33, 119 134, pis. 7-9. 1962. The trilobites of the Caradoc Series in the Cross Fell Inlier of northern England. Bull. Br. Mus. nat. Hist. (Geol.). 7 (3), 67-134, pis. 6-18. delo, D. M. 1935. A revision of the Phacopid trilobites. J. Paleont. 9, 402-420, 45 figs. destombes, j. 1972. Les Trilobites du sous-ordre des Phacopina de l’Ordovicien de l’Anti-Atlas (Maroc). Notes Mem. Serv. geol. du Maroc, No. 240. Edwards, H. M. 1 840. Histoire naturelle des crustaces comprenant Vanatomie, la physiologie et la classification de ces animaux. Paris, t. 3, 1-638, pis. 1-42. foerste, a. f. 1910. Preliminary notes on Cincinnatian and Lexington fossils of Ohio, Indiana, Kentucky and Tennessee. Bull, scient. Labs Denison Univ. 16 (2), 17-87, pis. 1-6. 1919. Notes on Isotelus, Acrolichas, Calymene, and Encrinurus. Ibid. 19, 65-81, pis. 14-18. hawle, i. and corda, a. j. c. 1847. Prodrom einer Monographic der bohemischen Trilobiten. Prague. hisinger, w. 1840. Lethaea Svecica seu Petrificata Sveciae, iconibus et characteribus illustrata. Suppl. secundum. Holmiae. holm, G. 1882. De svenska artena af Trilobitslagtet Illaenus (Dalman). Bill. K. svenska Vetensk-Akad. Handl. 7 (3), 1-148, pis. 1-6. 1886. Revision der ostbaltischen silurischen Trilobiten. Abth. 3. Die ostbalischen Illaemden. Mem. Acad. imp. sci. St.-Petersbourg, 33, 1-173, pis. 1-12. ingham, j. k. 1970. A monograph of the upper Ordovician trilobites from the Cautley and Dent districts of Westmorland and Yorkshire. Palaeontogr. Soc. (Monogr.) (1), 1-58, pis. 1-9. Kiel an, z. 1960. Upper Ordovician trilobites from Poland and some related forms from Bohemia and Scandinavia. Palaeont. pol. 11, vi + 198 pp., pis. 1-36. king, w. r. b. 1923. The Upper Ordovician rocks of the south-western Berwyn Hills. Q. Jl geol. Soc. Lond. 79, 487-507, pi. 26. kutorga, s. 1 854. Einige Sphaerexochus und Cheirurus aus den silurischen Kalksteinschichten des Gouverne- ments von St. Petersburg. Zap. imp. miner. Obshch. 13, 105-126, pis. 1-3. lane, p. D. 1971. British Cheiruridae(Trilobita). Palaeontogr. Soc. (Monogr.), 1-95, pis. 1-16, text-figs. 1-13. marr, j. e. and Roberts, t. 1885. The Lower Palaeozoic Rocks of the Neighbourhood of Haverfordwest. Q. Jl Geol. Soc. Lond. 41, 476-491, pi. 15. m’coy, f. 1849. On the classification of some British fossil Crustacea, with notices of new forms in the University Collection at Cambridge. Ann. Mag. nat. Hist. (2), 4, 161-179, 330-375, 392-414, 15 figs. — 1851. In Sedgwick, A. and m’coy, f. A synopsis of the classification of the British Palaeozoic rocks, with a systematic description of the British Palaeozoic fossils in the geological museum of the University of Cambridge. 184 pp. Cambridge and London. portlock, j. e. 1843. Report on the geology of the county of Londonderry and parts of Tyrone and Fermanagh. Dublin and London. prantl, f. and pribyl, a. 1947. Classification of some Bohemian Cheiruridae (Trilobita). Sb. ndr. Mus. Praze, 3, Geol. (Palaeont.) 1, 1-44, pis. 1-6. price, d. 1973. The Age and Stratigraphy of the Sholeshook Limestone of South-West Wales. Geol. J. 8 (2), 225-246. reed, f. r. c. 1904. Sedgwick Museum Notes. New Fossils from the Haverfordwest District. Geol. Mag. 5 (1), 106-109, pi. 5. — 1904a. Sedgwick Museum Notes. New Fossils from the Haverfordwest District. 2. Ibid. 383-388, pi. 12. 1905. The Classification of the Phacopidae. Ibid. (2), 172-178. — 1908. Sedgwick Museum Notes. New Fossils from the Haverfordwest District. 8. Ibid. (5), 433-436, pi. 14. — 1931. The Lower Palaeozoic trilobites of Girvan. Supplement no. 2. Palaeontogr. Soc. (Monogr.). 30 pp. 868 PALAEONTOLOGY, VOLUME 17 reed, f. r. c. 1933. Notes on the species Illaenus bowmanni Salter. Geol. Mag. 70, 121-135. salter, j. w. 1848. In Phillips, j. and salter, j. w. Palaeontological appendix to Professor John Phillips’ Memoir on the Malvern Hills, compared with the Palaeozoic districts of Abberley etc. Mem. geol. Surv. U.K. 2 (1), viii-xiv-t- 33 1-386, pis. 4-30. — 1852. Appendix A in Sedgwick, a. and m’coy, f. A synopsis of the classification of the British Palaeozoic rocks, with a systematic description of the British Palaeozoic fossils in the geological museum of the Uni- versity of Cambridge. London and Cambridge. — 1853. Figures and Descriptions illustrative of British Organic Remains. Mem. geol. Surv. U.K. Dec. 7. 12 pp., 2 pis. — 1864, 1865, 1867. A monograph of the British trilobites from the Cambrian, Silurian and Devonian formations. Palaeontogr. Soc. (Monogr.) (1), 1864, 1-80, pis. 1 6; (2), 1865, 81-128, pis. 7 14; (4), 1867, 177-214, pis. 25*-30. — 1866. Appendix on the Fossils in ramsay, a. c. The Geology of North Wales. Mem. geol. Surv. U.K. 3, pp. 239-363, 372-381, 26 pis. — 1873. A catalogue of the collection of Cambrian and Silurian fossils contained in the Geological Museum of the University of Cambridge. Cambridge. 1881. In salter, j. w. and etheridge, r. On the fossils of North Wales. Appendix in ramsay, a. c. The Geology of North Wales. Mem. geol. Surv. U.K. 3, 2nd ed., pp. 333-611, 26 pis. schmidt, f. 1881. Revision der ostbalischen silurischen Trilobiten nebst geognostischer Ubersicht des ostbaltischen Silurgebiets. Abth. 1. Phacopiden, Cheiruriden und Encrinuriden. Mem. Acad. imp. sci. St.-Petersbourg, (7), 30, 1-237, pis. 1-16. shirley, j. 1931. A redescription of the known British Ordovician species of Calymene (s.l.). Mem. Proc. Manchester Lit. Phil. Soc. 75, 1 33, pis. 1, 2. 1936. Some British Trilobites of the Family Calymenidae. Q. Jlgeol. Soc. Loud. 92, 384-422, pis. 29-31 . strahan, a., cantrill, T. c. and dixon, E. E. L. 1914. The geology of the South Wales coalfield. Pt. 11. The country around Haverfordwest. Mem. geol. Surv. U.K. Sheet 228. — and thomas, h. h. 1909. The geology of the South Wales coalfield. Pt. 10. The country around Carmarthen. Ibid. Sheet 229. Stubblefield, c. J. 1938. The types and figured specimens in Phillips and Salter’s palaeontological appendix to John Phillips’ Memoir on ‘The Malvern Hills compared with the Palaeozoic districts of Abberley, etc.’ Mem. geol. Surv. Summ. Prog, (for 1936), 2, 27-51. vogdes, a. w. 1890. A bibliography of Palaeozoic Crustacea from 1698 to 1889, including a list of North American species and a systematic arrangement of the genera. Bull. U.S. Geol. Surv. 63, 1-177. whittard, w. f. 1958, 1961. The Ordovician trilobites of the Shelve Inlier, West Shropshire. Palaeontogr. Soc. {Monogr.) (3), 1958, 71-116, pis. 10-15; (6), 1961, 197-228, pis. 26-33. Whittington, h. b. 1941. The Trinucleidae— with special reference to North American genera and species. J. Paleont. 15, 21-41, pis. 5, 6. 1941a. Silicified Trenton trilobites. Ibid. 492-522, pis. 72-75, 13 figs. 1950. Sixteen Ordovician Genotype Trilobites. Ibid. 24, 531-565, pis. 68-75, 9 figs. 1962, 1965, 1968. The Ordovician trilobites of the Bala area, Merioneth. Palaeontogr. Soc. (Monogr.) (1), 1962, 1-32, pis. 1-8; (2), 1965, 33-62, pis. 9-18; (4), 1968, 93-138, pis. 29-32. woods, H. 1891. Catalogue of the type fossils in the Woodwardian Museum , Cambridge. Cambridge. DAVID PRICE Department of Geology Sedgwick Museum Downing Street Cambridge, CB2 3EQ Manuscript received 6 June 1973 A NEW SPIDER-CRAB FROM THE MIOCENE OF NEW ZEALAND by R. J. F. JENKINS Abstract. Actinotocarcinus chidgeyi gen. et sp. nov. from the Waiauan or Middle Miocene of North Canterbury, New Zealand, is a long-spined, spider-crab belonging to the family Majidae Samouelle, 1819. While it shows a mosaic of individual features present within other subfamilies of the Majidae it differs from all in its total combination of characters, and a new subfamily Actinotocarcininae subfam. nov. is erected to contain it. In 1967 a loose concretion containing three specimens of an extraordinary, long- spined spider crab was found by Mr. S. A. Chidgey at Glenafric Beach, about 50 km NNE. of Christchurch, South Island of New Zealand, and subsequently sent to Professor M. F. Glaessner, University of Adelaide, South Australia, who suggested the present study. A further dozen or so specimens of the spider-crab, all in loose concretions, have since been collected at the same locality by Mr. S. A. Chidgey and his son Mr. S. J. Chidgey. The taxonomy of the modern decapod crustacean fauna of Australasia is becoming increasingly complete (Griffin and Yaldwyn 1968; Dell 1968; Yaldwyn 1971). A monographic treatment of the fossil decapods of New Zealand was presented by Glaessner (1960) and latterly I described an interesting new fossil lobster from the Pliocene of North Canterbury (R. J. F. Jenkins 1972a). The Cretaceous fossil decapods of Australia are relatively well known (Woods 1953, 1957, and several later authors) and I have recently studied most of the more common Tertiary forms (R. J. F. Jenkins \912b). The new spider-crab described here is unlike any form with which I am familiar, and Professor Glaessner is of the same opinion. Thus the new crab is described as a new genus and species. It can be confidently referred to the family Majidae Samouelle, 1819, of the Oxyrhyncha, but shows a combination of features significantly different from all of the subfamilies at present recognized within this family. The present subdivision of the Majidae is unsatisfactory in so far as it is based partly on gradational differences and lacks definitive phylogenetic control from fossils; however, the new crab is so distinctive that a new subfamily is erected to accommodate it. The rocks exposed in the immediate vicinity of Glenafric Beach (the sea coast near the mouth of Dovedale Stream) are shown by Wilson (1963) to be part of the Mount Brown Beds and are dated by Fleming (1963) as of Waiauan age, which, according to D. G. Jenkins (1971) and Berggren (1972) is equivalent to the later part of the Middle Miocene. The repositories of materials and abbreviations used to indicate them are: C.M. Canterbury Museum, Christchurch, New Zealand ; specimen numbers pertain to the ‘Canterbury Museum register of fossil Arthropodab S.A.M. South Australian Museum, Adelaide, South Australia; Palaeontological collection. [Palaeontology, Vol. 17, Part 4, 1974, pp. 869-877, pi. 117.] 870 PALAEONTOLOGY, VOLUME 17 SYSTEMATIC PALAEONTOLOGY The present crab shows features which clearly place it in the Oxyrhyncha ; notably the narrowed anterior portion of the carapace and inflated branchial regions, the general form of the other regions of the carapace, the wide epistome and nearly square buccal frame, and the longitudinally elongate antennular fossae (text-fig. 1a, b). Its well-calcified carapace with fine hair pores on the dorsal surface, its incom- plete spinose orbits and the fusion of the basal antennal articles to the epistome and text-fig. 1. Reconstruction of carapace of Actinotocarcinus chidgeyi gen. et sp. nov. A, dorsal view of carapace, x 1|; carapace regions notated as follows : f, frontal ; o, supra-orbital ; pg, protogastric ; h, hepatic ; mg, metagastric ; m, mesogastric ; u, urogastric ; c, cardiac ; i, intestinal ; eb, epibranchial ; mb, mesobranchial ; mt, metabranchial; other lettering: ba, basal antennal article; so, supra-orbital eave; e, eye-stalk; ant, antorbital spine; int, intercalated spine; lo, lateral-orbital process, b, ventral view of anterior parts of carapace, x 3: a, antenna; other lettering as above. The endostome is not known and has been omitted. R. J. F. JENKINS: MIOCENE SPIDER-CRAB 871 rostrum, and its apparently small and slender chelae enable it to be confidently identified as belonging to the family Majidae. Recent reviews of the major subdivisions within the Majidae have been presented by Garth (1958) and Griffin (1966a, 19666). These authors recognize seven extant subfamilies belonging to this family, the Oregoniinae Garth, 1958; Inarchinae McLeay, 1838; Ophthalmiinae Balss, 1929; Acanthonychinae Stimpson, 1870; Pisinae Dana, 1852; Majinae Samouelle, 1819; and Mithracinae Balss, 1929. The characteristic features of each of these subfamilies are summarized by Griffin (1966a). The new crab described here shows a mosaic of individual characters present within these divisions, but in its total combination of characters differs from all. The single rostral spine and relatively slender basal antennal articles of the new crab superficially resemble the same features in the single-horned section of the Inarchinae. It is distinguished from the Inarchinae, however, by its more elaborately developed orbits. Its orbits equally well distinguish it from the circum-Arctic Oregoniinae, which are similar to the Inarchinae in most features. The Oregoniinae have a two-horned rostrum. The slender form of its rostrum is sharply distinct from the huge, broad beak-like or double rostrum of the Acanthonychinae; also its orbits are larger and more complex than in this division. The remaining subfamilies all have a two-spined or bifid rostrum (sometimes with the rostral spines more or less fused or forming a broad lamella) and are primarily distinguished by the features of their more or less complex orbits. In the present crab the orbits are formed by a supra-orbital hood or eave, an intercalated spine, and a large post- or lateral-orbital process. The orbits are formed in essentially the same way in a section of the Pisinae, the Majinae, and part of the Mithracinae. In general the orbits are not greatly developed or are ‘incomplete’ in the Pisinae, rather better developed in the Majinae, and strongly developed and tubular or ‘complete’ in the Mithracinae. The basal antennal articles of the new crab are tapered and rather narrow, somewhat as in the Pisinae, but the supra-orbital and particularly the lateral- orbital structures are much more strongly developed than in this subfamily. The same features more closely approximate the condition in the Majinae, but the orbits face forwards rather than laterally, the basal antennal articles are more slender and lack the usual conspicuous terminal spinules, and the lateral-orbital processes are much more massive basally and also differ in bearing a long, slender spine. The orbital features differ sharply from those of the Mithracinae in not being tubular with the basal antennal articles expanded laterally to form a floor to the orbits. A feature in common with members of the Mithracinae is the general expansion of the orbits and reduction of the hepatic regions to form a narrowed ‘neck’ to the carapace. In the Ophthalmiinae the orbits are formed by an expanded supra-orbital eave or an equivalent elongate spine, and a short post-orbital spine, an arrangement distinctly different from that in the present crab. Reference to Garth (1958) and Glaessner (1969) shows that five of the above- mentioned subfamilies are represented by fossil remains, the Majinae (?Upper Cretaceous, ?Eocene, and from Miocene to Recent), Inarchinae (Upper Eocene to Recent), Oregoniinae (Miocene to Recent), Pisinae (Pliocene to Recent), and Acanthonychinae (Pliocene to Recent). The Upper Cretaceous record of the Majinae 872 PALAEONTOLOGY, VOLUME 17 and an Upper Eocene record are based on fragmentary claws of questionable identity (Glaessner 1969). The only other Eocene record of the Majinae is of a highly aberrant genus Periacanthus Bittner, 1875; the majinid genus Leptomithrax Miers, 1876, occurs as early as the Oligocene in Australia (R. J. F. Jenkins 19726). An additional fossil subfamily of the Majidae is recognized, the Micromaiinae Beurlen, 1930, of Middle Eocene to Lower Oligocene age. This subfamily is primitive in that the second antennal segment (equivalent to the basal antennal article) is free; the orbits are relatively incomplete and the rostrum is bifid. From the above discussion it is evident that the new crab almost certainly belongs to the Majidae but that it cannot be referred to any of the presently recognized sub- families of this taxon. A new subfamily is proposed to accommodate it. Order decapoda Suborder pleocyemata Infraorder brachyura Section oxyrhyncha Family majidae Samouelle, 1819 Subfamily actinotocarcininae subfam. nov. Type genus. Actinotocarcinus gen. nov. Diagnosis. Majid with a single, slender rostral spine; forwardly directed orbits formed above by a supra-orbital eave, an intercalated spine, and a massive lateral- orbital process bearing a slender spine ; and tapering, relatively narrow, basal antennal articles fused to rostrum and epistome, and lacking conspicuous terminal spinules. Genus actinotocarcinus gen. nov. Type species. Actinotocarcinus chidgeyi sp. nov. Plate 117, figs. 1-4 a, b; text-fig. 1a, b Derivation of name. Actinoto, from the Greek adjective actinotds, rayed, in reference to the rayed pattern of the spines of the carapace ; carcinus, from the Greek carcinos, crab. Diagnosis. As for type species. Actinotocarcinus chidgeyi sp. nov. Plate 117, figs. 1 -4a, b ; text-fig. 1a, b Derivation of name. Named for Messrs. S. A. Chidgey and S. J. Chidgey. EXPLANATION OF PLATE 117 Actinotocarcinus chidgeyi gen. et sp. nov. Fig. 1 , holotype, C. M . zfc 1 7 1 A, x 1(; carapace with branchial spines damaged and central part of posterior margin and right postero-lateral aspect eroded away. Fig. 2, paratype, CM. zfc 171B, x 1^-; carapace with posterior part missing and associated chela. Fig. 3, holotype (lower right) and two paratypes in concretion after preparation, x 1. Fig. 4 a, b, paratype C.M. zfc 171C, x 3 ; 4 a, dorsal aspect of anterior part of carapace showing eye-stalk (e), distal tip of basal antennal article (ba), and basal joint of antennal peduncle (bp); 4 b, ventral aspect of anterior portion of carapace with eye-stalk (e), second joint of antennal peduncle (sp), and mandibular gnathobases (mg) visible. PLATE 117 JENKINS, Miocene spider-crab 874 PALAEONTOLOGY, VOLUME 17 Diagnosis. Carapace flask-shaped, constricted at level of hepatic regions, 1-2 times as long as wide (excluding spines); orbits large; rostrum and spine on lateral-orbital process long and slender; branchial margins each with two long spines and a shorter spine behind; surface ornamented by granules and low tubercles between which it is smooth. Material. Three specimens of the carapace (C.M. zfc 171 A, B, C) in a single con- cretion and another (S.A.M. P.15916) in a similar concretion. The holotype, C.M. zfc 171 A, has the more posterior branchial spines eroded or broken and only a small part of the posterior margin remaining. Paratype C.M. zfc 171 B has remains of the chelae associated with it. The concretions containing the remains consist of highly indurated grey calcareous siltstone. Occurrence. The crab occurs in loose concretions found at Glenafric Beach, North Canterbury, New Zealand. The concretions are eroded from a Waiauan (Middle Miocene) interval of the Mount Brown Beds which stand as high cliff's backing a nar- row beach zone. Concretions with the crab occur on the wave-cut platform and storm beach between about 300 and 800 m east of the mouth of Dovedale Stream; they are most common at the eastern end of the occurrence. Other concretions containing the fossil crab Tumidocarcinus giganteus Glaessner, 1960, occur commonly in associa- tion with those containing the spider-crab and are found between 1-4 km south-west and 800 m east of the mouth of Dovedale Stream. Rarely, concretions containing this crab are found in situ (cf. Glaessner 1960; Fleming 1963). Description. Carapace flask-shaped in outline, with large forwardly directed orbits, a single long, slender rostral spine, a similar slender spine arising from a massive lateral-orbital process, and two long spines and a shorter spine behind on each branchial margin. Dorsal aspect of the carapace somewhat flattened, but still moder- ately convex longitudinally and transversely; a distinct, obtuse, longitudinal ridge in the mid-line; regions moderately distinct, delimited by shallow grooves; surface ornamented by granules and low tubercles between which it is smooth and polished ; median ridge with a line of relatively conspicuous low tubercles; much of the surface with exceedingly fine pores which may have borne slender hairs in life. Rostral spine nearly straight, equal to about two-fifths the length of the carapace, directed forwards and somewhat deflexed, rounded-triangular in section, the two dorso-lateral margins bearing occasional granules. Orbits consisting above of a supra-orbital eave, an intercalated spine, and a lateral- orbital process, the three separated by deep fissures; supra-orbital eave narrow, granulate, bearing a prominent antorbital spine; intercalated spine small; massive basal part of the lateral-orbital process with its upper margin forming a blunt crest bearing a line of granules and terminating in a rectangular prominence; inner surface of the process shallowly hollowed; outer distal part of the process with a long, taper- ing, spine; spine narrow in dorsal view, but relatively deep in section, its rather sharply rounded upper margin bearing a line of granules. Proximal portion of the eye-stalks slender. Protogastric regions relatively small. Mesogastric region limited posteriorly by deep grooves; two slightly more conspicuous low tubercles on its mid-line. A short, R. J. F. JENKINS: MIOCENE SPIDER-CRAB 875 deeply impressed portion of the cervical groove curves forward and outward from the posterior part of the mesogastric region. Hepatic regions small and hardly defined. Metagastric, urogastric, cardiac, and intestinal regions distinguished longi- tudinally only by slight swellings and the gentle curvature of the lateral cardiac grooves; the latter show the fine crenulation indicative of underlying muscle attach- ments. A conspicuous tubercle on the median, anterior part of the intestinal region. Cardiac region with its surface relatively smooth, two low tubercles occur on the mid- line of the anterior portion. Epi- and mesobranchial regions with coarse ornament, and each bearing a long, deep, lateral spine. Metabranchial regions separated from the epibranchial regions by a deep, pitted groove and with their surface coarsely pitted ; a short postero-lateral spine present. Basal antennal articles fused to the epistome and the rostrum, tapered towards their distal end and mostly rather slender; two rounded lobes on the ventral part of their distal extremity; foramen for the peduncle opening upward and forward. Open- ings of the antennal glands elongate transversely. Antennular fossae narrow, incom- pletely separated by a narrow median septum supported by the base of the rostrum. Epistome wide and smooth, with a high posterior lip. Narrow, oblique sub- hepatic lobes each with a strong, flattened posterior tubercle. Pterygostomial elements small. Buccal frame a little wider than long, with its lateral margins subparallel. Mandibular gnathobases with their nearly straight inner margin smooth. Chelae relatively small, the right more robust than the left. Palm narrow; slender tapered fingers equal to about three-fifths the length of the palm; prehensile margins of fingers seemingly smooth. Dimensions. Measurements of the materials studied are given in Table 1. TABLE 1 Holotype Paratype C.M. zfc 1 7 1 A C.M. zfc 1 7 1 B Measurements (mm) Rostrum 16 — Carapace, length 36 (impf.) — Carapace, width 30 31 Lateral-orbital process, length basal portion 4-7 4-8 Lateral-orbital process, length outer spine 20 21 Epibranchial spine, length — 13 (impf.) Mesobranchial spine, length — 8 (impf.) Cheliped palm (left), length — 6 Movable finger of chela (left), length — 3-6 Remarks. The long spines of A. chidgeyi are presumably an adaptation to prevent it being readily swallowed by fishes. Other majid species such as Maja arambougi Van Straelen, 1936, from the Pliocene of Algeria, and the living. Eastern Asiatic Maja spinigera de Haan, 1839, have evolved a comparable armament. The predominantly siltstone beds in which A. chidgeyi occurs, the Mount Brown Beds, are considered by Fleming (1963) to be ‘typical of middle depths on the continental shelf’. In general, modern long-spined crabs seem to be most frequent on the outer part of the continental shelf and bathyal regions. The distinctive combination of rostral and orbital features in A. chidgeyi suggests 876 PALAEONTOLOGY, VOLUME 17 that it represents a stock which underwent a lengthy and separate evolutionary history from the other recognized majid divisions. The structure of the orbits most closely resembles that in the section of the family embracing the subfamilies Pisinae, Majinae, and Mithracinae and thus it seems reasonable to infer that its ancestry lies within this section. A remarkable aspect of the preservation of this crab is the association of several individuals in a single concretion (PI. 117, fig. 3). This is evidently not unusual for the species as Mr. S. A. Chidgey has several other concretions containing more than one individual in his collection. At least one of the specimens in the association studied (specimen C.M. zfc 1 7 1C) is a dead animal, not a moult, as the pterygostomial elements are still in place (PI. 117, fig. 4b), and it seems likely that the other specimens are also remains of dead individuals. The remains cannot have been moved far after death as parts of delicate appendages still remain and the long carapace spines are unbroken. Chance current transportation thus seems an unlikely explanation for the associations. The associations do not apparently reflect an originally dense popula- tion as the crab is a rare fossil, much less common than specimens of Tumidocarcinus occurring at the same locality. The explanation for the associations favoured here is that the crabs tended to group together in small numbers during life and died while grouped, perhaps smothered under a sudden influx of sediment. Some small post- mortem transport evidently occurred as the ventral skeleton of at least one specimen is missing. The crab Trichopalt avion gveggi Dell, 1969, which occurs in similar con- cretions weathering out of the Pliocene at Motunau Beach, North Canterbury, New Zealand, also occurs in associations of several individuals to a concretion (material in Chidgey collection). Here the ventral skeleton and legs are also present. Acknowledgements. Emeritus Professor M. F. Glaessner, University of Adelaide, suggested this work and offered valuable discussion at several stages. Dr. D. R. Gregg, Director of the Tasmanian Museum and Art Gallery, Hobart, Tasmania, arranged the loan of material when he was Keeper of Fossils at the Canterbury Museum, Christchurch. I am also grateful to Mr. S. A. Chidgey and Mr. S. J. Chidgey for their donation of material and for Mr. S. J. Chidgey’s guidance in the field. REFERENCES berggren, w. a. 1972. A Cenozoic time-scale— some implications for regional geology and paleobio- geography. Lethaia, 5, 195-215. dell, r. k. 1968. Composition and distribution of the New Zealand brachyuran fauna. Trans. R. Soc. N.Z. Zool. 10, 225-240. Fleming, c. a. 1963. Macrofossils. In wilson, d. d. Geology of Waipara Subdivision (Amerley and Motunau Sheets S68 and S69). Bull. geol. Surv. New Zealand , 64, 52-57 . garth, j. s. 1958. Brachyura of the Pacific Coast of America Oxyrhyncha. Allan Hancock Pacif. Exped. 21, i-xii-1- 1-854, pis. 1-55. glaessner, m. f. 1960. The fossil decapod Crustacea of New Zealand and the evolution of the order Decapoda. Palaeont. Bull. Wellington, 31, 1-63, pis. 1-7. 1969. Decapoda. In moore, r. c. (ed.). Treatise on invertebrate paleontology. Pt. R. Arthropoda 4, pp. 399-533, 552-566. Univ. Kansas Press and Geol. Soc. Amer. griffin, d. j. G. 1966a. A review of the Australian majid spider crabs (Crustacea, Brachyura). Aust. Zool. 13 (3), 259-298, text-figs. 1-3, pis. 15-17. — 19666. The marine fauna of New Zealand: spider crabs, family Majidae (Crustacea, Brachyura). Bull. N.Z. Dept, scient. ind. Res. 172, 1-111, text-figs. 1-23, pis. 1-4. R. J. F. JENKINS: MIOCENE SPIDER-CRAB 877 griffin, d. j. G. and yaldwyn, j. c. 1968. The constitution, distribution and relationships of the Australian decapod Crustacea. Proc. Linn. Soc. N.S.W. 93, 164-183. jenkins, D. G. 1971. New Zealand Cenozoic planktonic Foraminifera. Palaeont. Bull. Wellington , 42, 1-278, pis. 1-23. jenkins, R. J. F. 1972a. Metanephrops, a new genus of late Pliocene to Recent lobsters (Decapoda, Nephropidae). Crustaceana , 22, 161-177, pis. 1-2. — \912b. Australian fossil decapod Crustacea: faunal and environmental changes. Unpublished Ph.D. thesis, University of Adelaide. wilson, d. d. 1963. Geology of Waipara Subdivision (Amberley and Motunau Sheets S68 and S69). Bull. geol. Surv. New Zealand , 64, 1-122. woods, J. t. 1953. Brachyura from the Cretaceous of central Queensland. Mem. Qd Mus. 13, 50-57, pi. 2. 1957. Macrurous decapods from the Cretaceous of Queensland. Ibid. 155-175, pis. 4-6. yaldwyn, j. c. 1971. Preliminary descriptions of a new genus and twelve new species of natant decapod Crustacea from New Zealand. Rec. Dom. Mus. Wellington, 7, 85-94. Manuscript received 9 July 1973 r. J. F. jenkins Department of Geology University of Adelaide Adelaide South Australia 5001 ENVIRONMENTAL FACTORS DETERMINING THE DISTRIBUTION OF BRACHIOPODS by f. t. fursich and j. m. hurst Abstract. The distribution of depth-related Silurian brachiopod communities is shown to be correlated with a diminishing food supply towards deeper offshore water. On a palaeoslope, the distribution of brachiopod species is determined by their ability to collect adequate food ; this in turn is related to the surface area of the lophophore and morphological adaptations which assist the feeding. It is argued that the Spiriferida and Pentameridina, which pre- dominate in the deeper water, had the most complex lophophores : the Orthida, Strophomenida, and Rhynchonellida, which are common in shallower water, had less complex lophophores. Morphological adaptations either increase the feeding capacity of the lophophore (e.g. development of plicae, large sulcus, and wings), or are a response to the physical environment (e.g. strong or weak pedicle, thick or thin shell, large resting area). These special adaptations enabled some brachiopod species to colonize niches not normally occupied by the other members of the same order. When forms with less complex lophophores inhabit deeper water there is a decrease in size of the individuals. Complementary evidence from the Devonian, Jurassic, and Recent, as well as evidence from other lophophorate groups, shows that these principles are a basic pattern applicable throughout time, possibly to all suspension- feeders. Over the past ten years there has been a large increase in the amount of data pub- lished on fossil brachiopod communities, especially those in Palaeozoic rocks. Bretsky (1969, 1970) has described late Ordovician benthonic marine communities from the central Appalachians and north-central New York. Stevens (1966) and Sutton et a/. (1966) have described communities in the Permian and Devonian respec- tively. However, the pioneer work was carried out by Ziegler ( 1 965) on the Llandovery (Lower Silurian) of Wales and the Welsh Borderland. Knowledge of Llandovery marine brachiopod communities has been supplemented by several further studies (Cocks 1967; Ziegler et al. 1968 a). The purpose of this paper is to assess the factors responsible for the areal distribu- tion of brachiopods. The evidence is drawn from studies on fossil and present-day brachiopod communities, but mainly from the Silurian. For the most part brachiopods will be treated at ordinal level; only where necessary will reference be made to individual species. From both carbonate and clastic rocks of the Wenlock and Ludlow, about 40000 fossils have been collected by Calef, Hancock, and Hurst. This represents over 200 collections, which have been grouped into several communities. The percentage occurrence of individual genera in the community spectrum has been calculated. The Llandovery data have been obtained from Ziegler et al. (1968a) and from the unpublished theses of L. R. M. Cocks and A. M. Ziegler. This added another 17 000 fossils to the list of those already considered. Statistical data from the Devonian (Winter 1971) and the Lower Jurassic (Tchou- matchenco 1972) were also considered. [Palaeontology, Vol. 17, Part 4, 1974, pp. 879-900.] 880 PALAEONTOLOGY, VOLUME 17 SILURIAN BRACHIOPOD DISTRIBUTION The Silurian brachiopod-dominated communities are related to water depth (Ziegler et al. 1968a; Calef and Hancock 1974). Evidence for this is derived from the distribu- tion of the communities which successively border the land areas (Ziegler et al. 19686). In the Llandovery, the full community spectrum from shallowest to deepest is repre- sented by Lingula , Eocoelia , Pentamerus , Costistricklandia, and Clorinda. An equivalent succession has been found in the Wenlock and Ludlow clastic and carbonate rocks and is represented by Lingula , Salopina , Homoeospira/ Spltaerirhyncliia, Isorthis, Dicoelosia , and Visbyella communities, the latter exceeding the depth range of the Llandovery Clorinda community (Hancock et al. 1974). Calef and Bambach (1973) go a step further and suggest that this distribution is linked to a diminishing food supply in deeper water. This food (mainly particulate organic matter) decreases in abundance up to fifty times from nearshore shallow water towards offshore deep water (Jorgensen 1966). Some Recent brachiopods have evolved to colonize regions of the sea where only small amounts of food are available, by increasing the surface area of their filter- feeding system, the lophophore. In the Llandovery, a basic distribution pattern of brachiopods is present (Ziegler et al. 1968a, 1 9686) : the Spiriferida and Pentameridina preferred the deeper quieter water, whilst the Rhynchonellida were in the shallower more turbulent water. The Orthida and Strophomenida were scattered throughout the depth spectrum (text- fig. 1). In the Wenlock clastic rocks there is a large increase in the number of Spiriferida, and a decrease in the common Llandovery Pentameridina. The Rhynchonellida still expressed a strong preference for the shallower water. However, the remaining orders do not conform with the distribution pattern observed in the Llandovery, Wenlock Limestone, and Ludlow (text-fig. 1). An added factor which appears at the beginning of the Wenlock is a totally new community, deeper than the Dicoelosia one. This is the Visbyella community. It represents the establishment of a completely new ecological habitat (Hancock et al. 1974), and there is no direct Llandovery equivalent, other than the marginal Clorinda association (Cocks and Rickards 1969) which may be its approximate equivalent in depth, though not in content. In the Wenlock carbonate rocks there appears to be a fairly rigid brachiopod distribution pattern (text-fig. 1). The majority of the Strophomenida and the Rhynchonellida are most abundant in the shallow water as inferred from sedi- mentary associations, and size, density, and biomass of brachiopods (Hancock et al. 1974). The Orthida occupy an intermediate position, and the Pentameridina and Spiriferida are more abundant in the deeper water. In the Ludlow elastics the Spiriferida and Pentameridina are still far more abundant in the deeper water shelf facies, and the Rhynchonellida in the shallower water elastics. The Orthida and Strophomenida are again most abundant in the intermediate depths (text-fig. 1). Hancock et al. (1974) give a full interpretation of the depths involved. They suggest that the Silurian communities inhabited depths from shallow sub-tidal, to possibly C n 70 Llandovery A;: Wenlock Sh . jiiWen lock L Lud low Sh brachiopod L E PSt Dale jina Isorthi s Resserella Saloplna Visbyella STreni rH text-fig. 4. Size decrease of Silurian rhynchonellids, strophomenides, and dalmanellids from the shallow into the deep water; the result of a relatively inefficient lophophore. Key : vertical line = average size of specimens in that community; horizontal line = observed size range of specimens in that community. FURSICH AND HURST: BRACHIOPOD DISTRIBUTION 891 Gypidula text-fig. 5. Spiriferides and the penta- merid Gypidula do not decrease in size towards deeper water; the result of their highly efficient lophophore. 892 PALAEONTOLOGY, VOLUME 17 the filter-feeding (e.g. wings, deep sulcus) and, therefore, one would expect these groups to have predominated in shallower water where there was more food. However, the brachiopods with the simplest lophophores, i.e. Orthida and Stropho- menida are not most abundant in the shallow water, but at intermediate depths. This is explained in the following ways: (a) Most of the Orthida in the Silurian probably had a weak pedicle not suitable for life in a very turbulent environment ; in most Strophomenida the pedicle had atrophied (Muir-Wood and Williams 1965). ( b ) Most Silurian Orthida and Strophomenida had a small mantle cavity. To feed, they probably had to open their valves wide so as to give maximum display to the BASIC MORPHOLOGICAL ADAPTATIONS OF SILURIAN BRACHIOPODS * o o All in millimetres Of the three fossil trionychid skulls examined, two ( T . tritor Hay and T. levalensis Dollo) appear to resemble the majority of Recent species in the three main characters listed above. The third, Eurycephalochelys fowleri Moody and Walker 1970, prob- ably had an orbit that could be derived from that of T. silvestris, but it differs in many other major characters and cannot really be related. Conclusions. T. silvestris without doubt belongs to the genus Trionyx, but any dis- cussion of its relationships must await a complete analysis of all living and fossil trionychids. Acknowledgements. We thank Messrs. James and George for presenting the specimen to the B.M. (N.H.); Dr. A. J. Charig for criticizing the manuscript; and Mr. T. Parmenter for taking the photographs. Dr. R. T. J. Moody thanks both the N.E.R.C. and the University of London Central Grant Committee for their support during this research programme. WALKER AND MOODY: EOCENE TURTLE 907 REFERENCES dollo, L. 1909. The fossil vertebrates of Belgium. Ann. N. Y. Acad. Sci. 19, 99 1 19. hay, o. p. 1908. The fossil turtles of North America. Publ. Carneg. Inst. 75, 529-532, figs. 687-689. loveridge, a. and williams, e. e. 1957. Revision of the African Tortoises and Turtles of the Suborder Cryptodira. Bull. Mus. Comp. zool. Harv. 115, 164-577, pis. 15-18. moody, r. t. j. and walker, c. a. 1970. A new trionychid turtle from the British Lower Eocene. Palaeontology, 13, 503-510, pi. 102, text-figs. 1-5. wrigley, a. g. 1931. The Lower Eocene Mollusca of Abbey Wood and of High Halstow (Kent). In The Vertebrate faunas of the English Eocene , ed. white, E. i. B.M. (N.H.), London, 1, 110-112. C. A. WALKER Department of Palaeontology British Museum (Natural History) Cromwell Road London, SW7 5BD R. T. J. MOODY Department of Geology Kingston Polytechnic Penrhyn Road Kingston upon Thames Final typescript received 19 February 1974 Surrey LOWER CARBONIFEROUS CONODONT BIOSTRATIGRAPHY OF NEW SOUTH WALES by T. B. H. JENKINS Abstract. Based on collections totalling some thousands of specimens, the distribution of conodont elements in three main and several other subsidiary sections in the Carboniferous of New South Wales is summarized in terms of seven successive conodont faunas, six of which are regarded as indicating biostratigraphic zones. In upward succes- sion these are individually characterized by (a) Siphonodella spp., ( b ) Gnathodus punctatus , (c) Gnathodus semiglaber, (d) Gnathodus sp. A, (e) Scaliognathus anchoralis, (/) Pseudopolygnathus cf. nodomarginatus, and (g) Patrognathus ? cf. capricornis. Correlations with other Australian areas and with North American and European sections are briefly discussed on the basis of conodont distributions and comparison is made with previous intercontinental correlations of the New South Wales Carboniferous based on ammonoids. A conflict emerges between the conodont and ammonoid evidence on Visean correlations. Although the conodont biostratigraphy of the Carboniferous generally has reached a considerable degree of refinement little has been published of such studies for the system in New South Wales. More northerly Australian areas have received more attention in this connection, there being several recent publications dealing with Carboniferous conodonts from Queensland (Palmieri 1967, 1969; Druce 1970) and the Bonaparte Gulf Basin (Druce 1969), the latter greatly amplifying the initial reports by McWhae et al. (1958, p. 49) and Glenister (1960). Two recent publications refer briefly to Carboniferous conodonts from N.S.W. Firstly, Rhodes et al. (1969, p. 60) commented on a fauna with pseudopolygnathids, Gnathodus cf. punctatus and Bactrognathus ‘from the Berwick Formation of Australia’. (The formation name is undefined; the specimens came from the Carellan section, see below.) Secondly, Branagan et al. (1970, p. 129) recorded species of Gnathodus , Polygnathus , Pseudopolygnathus, Neoprioniodus, Spathognathodus, and Hindeodella from both of two conodont horizons in the Carboniferous sequence at Glenbawn dam in the Scone district of N.S.W. The lower horizon is recorded as carrying also species of Siphonodella and Patrognathus and the upper level as having forms belong- ing to Dollymae and Ozarkodina. The foregoing N.S.W. conodont records are based on collections assembled over the last decade by the present writer and which now total some thousands of specimens. This communication summarizes the vertical distribution of conodont elements which emerges from a study of these collections and discusses these results in the light of what is known of coeval conodont distributions elsewhere. No attempt is made here to provide the fully detailed systematic analyses of the collections upon which the conclusions are based. Accounts of particular conodont faunas are in preparation for publication. Major stratigraphical sections relevant to the present study are shown in text- fig. 1 ; a general account of the Carboniferous System in N.S.W. is to be found in Campbell et al. (1969). It may be noted in passing that limestones constitute only [Palaeontology, Vol. 17, Part 4, 1974, pp. 909-924, pi. 119.] 910 PALAEONTOLOGY, VOLUME 17 a very small part of the Carboniferous as it is developed in N.S.W. Clastics and volcanics form the bulk of the system, with glacial sediments appearing in the upper- most beds. Total thickness locally exceeds 6100 m, folding is moderately intense, and induration is such as effectively resists the disaggregation techniques usually employed for non-calcareous sediments. Consequently, conodont investigations in the N.S.W. Carboniferous have been limited by the distribution of calcareous rocks which seem to be confined to the lower part of the system. This paper employs the nomenclature of disjunct conodonts, which has been used for many years in biostratigraphic work on the Carboniferous to the virtual exclusion of apparatus-based nomenclature. Many of the disjunct conodonts from the Carboni- ferous of N.S.W. are conspecific with the coeval forms elsewhere. Furthermore, their order of appearance and extinction in the investigated sections allows close com- parisons with certain overseas sequences and the recognition of some conodont zones from northern continents. SUCCESSION OF CONODONT FAUNAS As has been pointed out in connection with the general sequence of invertebrate faunas (Campbell and Roberts 1969, p. 261), no single section or district exhibits the whole marine Carboniferous sequence of N.S.W. For the lower, intermittently calcareous, part of the sequence it is possible to propose a general succession of conodont faunas from detailed study of three main sections in the main Carboniferous belt of N.S.W. : (i) The western limb of the Belvue Syncline about 1 km north-west of Carellan homestead (see text-fig. 1), where an unusually thick calcareous facies is developed low in the local Carboniferous section. This was reasonably assumed by Campbell and Engel (1963, p. 59) to be an expanded development of the persistent Rangari Limestone, but it has not proved possible yet to demonstrate the suggested equivalence. Two other relations are possible, {a) the Rangari is equivalent to the lower, more continuously calcareous part of the Carellan beds, and ( b ) the Rangari wholly underlies at an unknown depth the lowest of the beds exposed in the Carellan section. The latter possi- bility seems to the writer to be the most likely one at present, and the Keepit column in text-fig. 1 is drawn accordingly. (ii) The Glenbawn dam area, where the sequence has been outlined by Branagan et al. (1970) and which has been included in a wider mapping project by Roberts and Oversby (in press) for the Bureau of Mineral Resources. (iii) The Brownmore fault block, about 8 km north-west of Dungog, recently investigated by Dr. J. Roberts and his students at the University of New South Wales. Supplementing these main sections are many shorter ones ; two such minor sections, the first 1-5 km east of Gloucester (at the Stock Reserve) and the other about 5 km west of the town on the Walcha road, provided the first specimens of the important Scaliognathus anchoralis fauna, which has not been found equally well developed in any of the three main sections. — — ✓ V V V/ V I I — — lv> V vl shale and volcanics limestone mudstone conglomerate sandstone text-fig. 1. Locality map with outline of Carboniferous outcrops; conodont horizons in three principal stratigraphic columns for the main Carboniferous belt of New South Wales; conodont fauna (a) and pro- posed zones (b-g) indicated by letters keyed to text. Recent work by Roberts (pers. comm.) shows that near Brownmore all the strata between the Bonmngton Siltstone and the unnamed conglomeratic sandstone belong to the Flagstaff Formation, which is 1800 m thick, i.e. about half the thickness depicted here. 912 PALAEONTOLOGY, VOLUME 17 Successive conodont faunas from these sections are characterized by the following species : (top) (, g ) Patrognathusl cf. capricornis (Druce) (/) Pseudopolygnathus cf. nodomarginatus (E. R. Branson) (e) Scaliognathus anchoralis Branson and Mehl (d) Gnathodus sp. A (c) Gnathodus semiglaber Bischoff (b) Gnathodus punctatus (Cooper) (a) Siphonodella spp. (base) For a number of reasons it is proposed at present to regard the successive conodont faunas ( b ) to (g) as indicating informal biostratigraphic zones. The lowest fauna listed, characterized by Siphonodella spp., is distinct but insufficiently well-founded to be accorded zonal status; its known limits are determined by lithology. The next three faunas, (b) to (d), are both distinct in their platform elements and have accurately definable mutual boundaries in sections lacking detected signs of hiatus. The upper- most zone, with Pa.l cf. capricornis, is based on material from a few thin limestones disposed through about 600 m of section, within which is a narrower zone with a conodont fauna composed almost exclusively of Gnathodus bulbosus Thompson and Taphrognathus various Branson and Mehl, which may be regarded therefore as a subzone. (a) The Siphonodella spp. fauna is based on rather sparse conodonts recovered from the 23 m-thick oolite low in the Glenbawn sequence, to be named the Brushy Hill Limestone Member by Roberts and Oversby (in press), and best exposed in a large quarry immediately south-east of the dam wall. The few available siphono- dellids (e.g. PI. 1 19, fig. 10) are inadequate for specific allocation but some resemble late Kinderhookian forms. Also present are Patrognathus andersoni Klapper (PI. 1 19, figs. 25-27), Polygnathus communis communis Branson and Mehl, Pseudopolygnathus multistriatus Mehl and Thomas, Spathognathodus costatus costatus (E. R. Branson), Prioniodina prelaevipostica Rhodes et al., Neoprioniodus sp., Ozarkodina sp., and Elictognathusl sp. (b) The Gnathodus punctatus Zone is based on conodonts from the lowest portion of the Carellan section and from exposures alongside the Aberdeen- Rouchel Brook road 04 km east of Rockhill homestead, where limy beds alternate irregularly with elastics through about 27 m of strata. The index species (PI. 1 19, fig. 14) has also been found recently in the Gloucester district where it is reportedly accompanied in the lower part of its range by siphonodellids (pers. comm. D. T. Crane). Other elements of the fauna are Pseudopolygnathus multistriatus , Bactrognathus liamatus Branson and Mehl, Polygnathus communis carina, an unnamed subspecies of P. communis dis- tinguished by faintly crenulate margins, and, at one locality, many specimens of Gnathodus sp. C (PI. 119, fig. 28). (c) The Gnathodus semiglaber fauna occurs in the lower, but not basal, part of the Carellan section, in the lower part of the ‘crinoidal-coral limestone sequence’ at Glenbawn dam, and at the top of the section near Rockhill, Rouchel Brook. T. B. H. JENKINS: CONODONT BIOSTRATIGRAPHY 913 I Q QZ o U_ tO UJ o UJ oz oz o £ z £ o S 0 SIN/dOOn/dVD t isnsosina snaoHivNo I SNVIdVA SnHlVNDOVLIdVL NOliVWaOd ddVISOVId 3NO1S0NVS ivayav yaanaaiQNia NOIlVWdOd ■p j SnHiVNOOSiVd snivi sriHivNoonoa SnVHOHDNV snmvMOonvDs snivNNid smnoNvidi snHitNOAiOdoons^ SfUVNIOdVWOOON V SDHivjoAlOdOOndSd XOOtffl 13H 3DN3CI03S Z I GO Z UJ o o o oz c o X U z» O c* DOOM MV3N niHXDoa INOSddONV SHHlVNOOdlVd I dds VlldOONOHd/SUHW ods snaoHivNo snivNNid smnoNvm snmvNOAiodoamsd smvisoj snivisoD snaOHivNoomvds shividisumw s'nHivNOAiOdoandSd i V’ds snOOHIVNOf sinowwoo s/Nnwwco diavioiwi S SflOOHlVNO SniVlDNHd naoHivNO m NMV8N31© IV 33N3H03S TS1 1VMQ3 1VQIONIM3 SOHlVNOAlOd d38W3W 3N01S3WH IniH AHsnaal ds VNI0OH3NO1V13W uj y >5 uj ~r eo >. tO snauoins snivisoD snmsco snopHivNooHivds snmsco snaoHivNDOHivds i snmaisuinw sriHxvNDAiodoandSd S/NnwVVOD S/NnWWODl SflHXVNOAlOd vd5 snaoHivNo d38V10/W3S SnaOHIVNO smvaNnd snaoHivNo 3 DN30O3S NV113SVD h- O Z III O 00 V“/ ULJ to W O o £ o u O N 8 | ? &■ 2: u ■ST; d ■*- -c o §i 8.1 O ^3 "S 8 <2 O O D si c o a.S _a; a 0) o a ; -C to 5° == ^ 8| .«! O) -C_Q a oo -2 £ ~ai o -Q c E 3| O 0> text -fig. 2. Proposed conodont zones for New South Wales and ranges of some stratigraphically significant conodont elements. 914 PALAEONTOLOGY, VOLUME 17 Accompanying the index gnathodid (PI. 1 19, fig. 1 3) are the elements Ps. multistriatus, Sp. costatus costatus, Sp. costatus sulciferus (Branson and Mehl), Sp. plumulus cf. shirley ae Rhodes et al., Metalonchodina sp., P. communis communis and the marginally crenulate subspecies of communis mentioned above. The species name semiglaber is here used in the sense of recent North American authors, such as Thompson and Fellows (1970), which departs significantly from Bischoff’s (1957) original concept. Likewise the naming of spathognathodid and pseudopolygnathid elements follows recent practice, especially that of Rhodes et al (1969), although here again, as pointed out by Matthews and Naylor (1973, p. 364), there exists a case for systematic revision. Perhaps such revision should be integrated with a change from the prevailing element-based nomenclature to an apparatus-based nomenclature. (d) The Gnathodus sp. A Zone occurs in the upper part of the Glenbawn ‘crinoidal- coral limestone sequence’ and in a short road section about 1-6 km east of Belah EXPLANATION OF PLATE 119 All figures are x 40 unretouched SEM images. Specimens are registered in the palaeontological collection catalogues of the Department of Geology and Geophysics, University of Sydney. Figs. 1, 2, 4. Gnathodus bulbosus Thompson. 1, 2 oral and oblique oral views of SUP 22100. 4, oral view of SUP 22102; from limestones high in the Flagstaff Formation? in the Brownmore fault block, about 8 km north-west of Dungog, N.S.W. Fig. 3. Gnathodus texanus pseudo semiglaber Thompson and Fellows. Oral view of SUP 22101 ; from upper part of Baywulla Formation at Baywulla Crossing, Yarroll Basin, Queensland. Fig. 5. Taphrognathus varians Branson and Mehl. Oral view of SUP 22103; horizon and locality as for figs. 1, 2, 4. Figs. 6-9. Scaliognathus anchoralis Branson and Mehl. 6, oral view of SUP 22104. 7-9, oral, aboral, and side views of SUP 22105; from unnamed formation about 7 km west of Dungog on new road to Gresford, N.S.W. Fig. 10. Siphonodella sp. Oral view of SUP 22106; from 13-7 to 14-6 m below top of oolite (i.e. Brushy Hill Limestone Member) at Glenbawn Dam, Scone district, N.S.W. Fig. 11. Doliognathus latus Branson and Mehl. Oral view of SUP 22107; from top of supposed Wootton Beds, Walcha road, about 5 km west of Gloucester, N.S.W. Fig. 12. Pseudopoly gnathus triangulus pinnatus Voges. Oral view of SUP 22108 ; horizon and locality as for fig. IF Fig. 13. Gnathodus semiglaber Bischoff. Oral view of SUP 22109 from 26-5 to 26-8 m above the exposed base of the Carellan sequence, Keepit district, N.S.W. Fig. 14. Gnathodus punctatus (Cooper). Oral view of SUP 22110; from 4 0 to 4-3 m above the exposed base of the Carellan sequence, Keepit district, N.S.W. Fig. 15. Dollymae hassi Voges. Oral view of SUP 22111 ; from unknown horizon within ‘crinoidal-coral limestone sequence’ at Glenbawn Dam, Scone district, N.S.W. Figs. 16-18. Polygnathus sp. A. Oral, aboral, and side views of SUP 221 12; locality as for figs. 6-9. Figs. 19-21. Gnathodus sp. B. Aboral, side, and oral views of SUP 22113; locality as for figs. 6-9. Fig. 22. Gnathodus sp. A. Oral view of SUP 22114; from high horizon in the ‘crinoidal-coral limestone sequence’ at Glenbawn Dam, Scone district, N.S.W. Figs. 23, 24. Pseudopoly gnathus cf. nodomarginatus (E. R. Branson). Aboral and oral views of SUP 22115; from Raglan Limestone at Raglan homestead, 8 km west of Booral, N.S.W. Figs. 25-27. Patrognathus under son i Klapper. Oral, oblique, and side views of SUP 22116; from 11 -6 to 1 T9 m below top of oolite (i.e. Brushy Hill Limestone Member) at Glenbawn Dam, Scone district, N.S.W. Fig. 28. Gnathodus sp. C. Oral view of SUP 22117; from 24 0 to 24-3 m below top of limestone near Rockhill homestead, Rouchel, N.S.W. PLATE 119 JENKINS, Conodonts 916 PALAEONTOLOGY, VOLUME 17 homestead on the Scone-Gundy road. The index gnathodid (PI. 119, fig. 22) is distinguished by a broad nodose outer platform and an arcuate inner platform bear- ing transverse or slightly radiating ridgelets; it may be related to Gn. cf. bilineatus of Thompson 1967, p. 37, but is not identical with the species there illustrated (pi. 3, figs. 8, 10, 12, 17). Together with the gnathodid occur abundant Spathognathodus costatus sulciferus, Sp. costatus costatus , and Ps. multistriatus , which become dominant elements upwards as the gnathodids become rare or absent. In the upper part of the zone appears Ps. triangulus pinnatus Voges, Dollymae sp. (PI. 1 19, fig. 15), and a new subspecies of P. communis distinguished by a row of small nodes flanking the anterior medial carina on one or both sides and linking with transverse carinae of the anterior platform. These distinctive forms could provide grounds for recognizing a subzone or separate zone if they are found to occur sufficiently abundantly. ( e ) The Scaliognathus anchoralis Zone is known from five localities between Gresford and the coast at Taree, e.g. in the Taree Limestone at the railway cutting near the town; and in oolitic limestones crossing the Gresford-Bungog road 1-2 km east of its junction with the Gresford-Salisbury road. Field relations are such that at none of these localities can the zone’s vertical continuity from an underlying conodont horizon be established and at only one locality (text-fig. 1, right-hand column), the second cited above, can the zone, there falling in the Bingleburra Formation, be placed in sequence with higher conodont-bearing formations. However, the position of the distinctive anchoralis- Zone conodonts in the general sequence of mid-Binantian conodont faunas is sufficiently well established elsewhere to warrant the zonal sequence here given. Elements indicative of the zone include Scaliognathus anchoralis (PI. 119, figs. 6-9), Doliognathus latus (PI. 119, fig. 11), Polygnathus sp. A (PI. 119, figs. 16-18), and Gnathodus reversus Thompson, Ford and Sweet, while Ps. triangulus pinnatus (PI. 119, fig. 12) ranges up into the anchoralis Zone from the underlying zone, and Ps. cf. nodomarginatus and Gnathodus sp. B (PI. 119, figs. 19-21) continue upwards above the range of S. anchoralis. These overlapping ranges thus support the partly inferred zonal position of the anchoralis Zone. (/) The Pseudopolygnathus cf. nodomarginatus Zone is based on rather restricted faunas recovered from limestones within the Ararat Sandstone of the Lewinsbrook Syncline and about 8 km north-west of Gresford, and from the Raglan Limestone, outcropping near the homestead of that name 5 miles west of Booral. The com- monest elements are intermediate in form between Polygnathus and Pseudo- polygnathus, with elongate basal cavities and ribbed platforms which are usually asymmetrical in the manner of the latter genus. Such forms are widely reported under several names from mid-Dinantian strata and are here provisionally referred to Ps. cf. nodomarginatus (PI. 119, figs. 23, 24). Near the base of the Ararat Sandstone Gnathodus sp. B occurs, and towards the top Cavusgnathus sp. appears. (g) The Patrognatliusl cf. capricornis Zone, based on the faunas from four thin limestones within the Flagstaff Formation? in the Brownmore fault block and from the Verulam Limestone near Gloucester, is characterized by the index fossil, accompanied in the lower part of its range by Cavusgnathus sp., and in the middle two limestones of Brownmore by Gnathodus bulbosus (PI. 119, figs. 1, 2, 4), Taphrognathus various (PI. 119, fig. 5), and Mestognathus sp. 1 follow Klapper’s (1971, p. 8) suggestion in referring Bruce’s species capricornis, with some reservations, to Patrognathus. T. B H. JENKINS: CONODONT BIOSTRATIGRAPHY 917 CORRELATION Correlation with other Australian sequences Conodonts from the Rockhampton Group of the northern Yarrol Basin (Druce 1970) suggest possible approximate Queensland equivalents for several N.S.W. formations: the Gudman Oolite (Qld) for the Glenbawn oolite (N.S.W. ); the Gargoogie Oolite (Qld) for the Flagstaff Formation (N.S.W.); and the unnamed lime- stone of Druce’s fig. 2 for the various formations carrying the anchoralis fauna in N.S.W. However, the published descriptions of the Yarrol Basin faunas are rather meagre and in no case can these possible correlations be regarded as established with either precision or certainty. Correlation with the Bonaparte Gulf Basin is more firmly based, the conodonts having been monographed by Druce (1969). However, two factors operate against detailed biostratigraphic parallelism: firstly the absence from the northern faunas of the rich gnathodid component found in the N.S.W. faunas, and secondly, the termination of the calcareous sequences in such a manner as to provide no Visean sequence in the Bonaparte Gulf Basin equivalent to the upper zones of N.S.W., and no known well-developed N.S.W. sequence equivalent to the lowest Carboniferous zones in the northern basin. The fauna of the Glenbawn oolite (Brushy Hill Lime- stone Member of Roberts and Oversby (in press)), with its rare siphonodellids, correlates with some part of the siphonodellid range from 91 to 274 m above the base of the Burt Range Formation; additional material is necessary to achieve closer correlation of the Glenbawn oolite. The Clydagnathus nodosus Assemblage Zone of the northern basin cannot be recognized in the N.S.W. material but these eastern faunas do contain elements of the succeeding three zones recognized by Druce (1969). Included in the Gn. punctatus Zone in N.S.W. is Spathognathodus anteposicornis Scott; Sp. costatus costatus and Sp. costatus sulciferus become common in the upper part of the Gn. sp. A Zone. Ps. cf. nodomarginatus occurs in some of the succeeding anchoralis Zone faunas in N.S.W. and ranges upwards through the Ps. cf. nodomarginatus Zone. It seems likely that some part of this N.S.W. range includes the Ps. nodomarginatus Assemblage Zone of Druce and this raises the possibility that the S. anchoralis Zone may be represented in the Bonaparte Gulf Basin by sandstones above or within the upper part of the Septimus Limestone. The rich conodont fauna of the Utting Calcarenite of the Bonaparte Gulf Basin cannot at present be matched in N.S.W. However, its Taphrognathus sp. element (Druce 1969, p. 139) has been named by Druce (1970, p. 102) as T. capricornis and is close to the form here termed Pa. ? cf. capricornis , characteristic of the uppermost zone recognized in N.S.W. Correlations with sequences outside Australia The N.S.W. conodont sequence shows greater parallelism with certain sequences described from the northern continents than with those discussed above from northern areas of Australia. The Missouri conodont sequence for the Kinderhookian and Osagean described by Thompson (1967) and Thompson and Fellows (1970) is especially valuable in seeking to establish correlations with North America. On the 918 PALAEONTOLOGY, VOLUME 17 other hand, some recent British results (Rhodes et al. 1969) for the lower and middle Avonian, based primarily on a succession of spathognathodid, polygnathid, and clydagnathid elements, are generally difficult to apply to the N.S.W. sequence in which the stratigraphically useful elements are mainly gnathodids and pseudo- polygnathids. However, Butler’s (1973) account of conodont sequences from the eastern Mendips, reveals considerable parallelism with the faunas here reported. Three of the species names attached to the proposed N.S.W. conodont zones have previously been used, either alone or in conjunction, as name bearers for zones or subzones in Europe and North America: Gnathodus punctatus, by Hass 1959, p. 367; by Thompson and Fellows 1970, p. 571; Gnathodus semiglaber, by Collinson et al. 1962, p. 22; Thompson 1967, p. 17; Thompson and Fellows 1970, p. 58; and Scaliognathus anchoralis, by Bischoff 1957, p. 12; Voges 1959, p. 270; 1960, p. 216. The last of these zones is very widely distributed, the S. anchoralis fauna having been recognized by one or more of its distinctive elements from Czechoslovakia (Zikmundova 1967) to Ireland (Hill 1971) and North Africa (Remack-Petitot 1960) and from the Mississippi Valley (Collinson et al. 1971) to New Mexico (Burton 1964) as well as from Queensland (Druce 1970) and N.S.W. (Branagan et al. 1970). It should be noted that Groessens (1971) and Groessens et al. (in press) have shown that the anchoralis fauna characterizes the upper part of the topmost division (i.e. upper Tn3c) of the Tournaisian stratotype, correcting the previous information in Conil et al. (1969) and, consequentially, lowering the cully horizon in terms of the Belgian equivalents. The Scaliognathus anchoralis Zone can thus, by direct correlation with the stratotype, be taken as indicating a topmost Tournaisian segment of the time scale. The sequence of gnathodids forming the main basis for the Gn. punctatus , Gn. semiglaber , and Gn. sp. A zones in N.S.W. shows close parallelism with the Missouri sequence (Thompson and Fellows 1970, table 1). Replacement of punctatus by semi- glaber and of semiglaber by Gn. cf. bilineatus or Gn. sp. A seems to be equally abrupt in Missouri and N.S.W. These vertical distributions are so close as to warrant equating the boundary between the Gn. punctatus and Gn. semiglaber zones in N.S.W. with the boundary between the Siphonodella cooperi hassi-Gnathodus punctatus Zone and the Gnathodus semiglaber- Polygnathus communis carinus Zone of Missouri. Close parallelism between N.S.W. and Missouri gnathodid sequences contrasts with significant differences in the vertical distributions of other elements. In parti- cular the N.S.W. sequence shows no sudden incoming of Ps. multistriatus at the top of the semiglaber Zone such as occurs at about this level in Missouri and the Mississippi Valley, for in N.S.W. such pseudopolygnathids continue through the gnathodid zones, with the chief change occurring at or near the top of the Gn. sp. A Zone, where elongate forms with attenuate platforms become dominant. Polygnathus communis communis becomes extinct in N.S.W. at about this level, although descendant forms continue into the anchoralis Zone. A common element of the latter zone in N.S.W. is Doliognathus latus, name giver for a Missouri subzone which correlates with the lower part of the anchoralis Zone in N.S.W. if the ranges indicated in Thompson 1967, p. 18 are accepted. T. B. H. JENKINS: CONODONT BIOSTRATIGRAPHY 919 text-fig. 3. Comparison and correlation of conodont zones for the Tournaisian and Lower Visean stages in Belgium, U.S.A., and New South Wales. text-fig 3. Comparison and correlation of conodont zones for the Tournaisian and Lower Visean stages in Belgium, U.S.A., and New South Wales. 920 PALAEONTOLOGY, VOLUME 17 The Ps. cf. nodomarginatus Zone of N.S.W. is not readily matched in North American sequences but the name giver seems to be close to Polygnathus mehli Thompson, a minor component of the fauna in the Missouri Zone following the latus subzone, and which is recently reported to appear in abundance in a slightly modified form to distinguish a new subzone coinciding with the Cedar Fork Member of the Burlington Limestone in Iowa and the Mississippi Valley (Collinson et al. 1971, p. 379). In south-west Britain the Avonian is reported to have similar forms under the names Ps. nodomarginatus and Polygnathus lacinatus Huddle, the latter being used as name giver, alone or in combination, for three successive zones in the scheme of Rhodes et al. (1969). Patrognathusl cf. capricornis seems to be known only from Australia but associated species have a wider distribution, notably Taphrognathus varians and Gnathodus bulbosus. In the Mississippi Valley sequence the former species enters sporadically in the uppermost subzone of the Bactrognathus-Taplirognathus Zone and persists up to the base of the Apatognathus scalenus-Cavusgnathus Zone (Collinson et al. 1971, p. 279). Gn. bulbosus is unknown from the Mississippi Valley sections but in Missouri it is the short ranged index for the Gn. bulbosus Zone of Thompson and Fellows (1970). There T. varians enters at the top of the bulbosus range. The North American ranges of the two species thus fall in the middle part of the Valmeyeran, in zones which Collinson et al. (1971, table 1), accepting Voges’s (1960) tentative equating of the anchoralis Zone with the lower Visean, correlate with the middle Visean. As pointed out above, the anchoralis Zone has been found by Groessens’s recent detailed work (1971, and in press) to lie within the Tn3c division, i.e. uppermost Tournaisian of the stratotype. The resulting recalibration of the conodont sequence with Belgian strato- types requires the ranges of Gn. bulbosus and T. varians to be stated as lower Visean, although neither species is presently known from Belgium. Recent work by Butler (1973) records Gn. bulbosus from the Mendips, near Bristol, England, where its range appears to fall within that of S. anchoralis. These distributions are therefore consistently in favour of a mid-Dinantian correla- tion for the Gn. bulbosus-T. varians fauna, a lower Visean level being indicated for these species by the position of anchoralis in the Belgian stratotype. Comparison of intercontinental correlations based on ammonoids and conodonts Previous extra-Australian correlations of the N.S.W. Dinantian have given weight to ammonoid occurrences, two of which can now be considered in context with the conodont faunas here reported. From the Tulcumba Sandstone and its Rangari Limestone Member in the Belvue and Werrie Synclines, Campbell and Engel (1963) described a large fauna including the ammonoids Muensteroeeras cf. oweni (Hall), Protocanites lyoni (Meek and Worthen), Pr. australis Delepine, Prionoceras ( Imitoceras ) werriense Campbell and Engel, and Prionoceras ( Imitoceras ) sp. One of several localities for the first three species listed above was given as ‘one mile north-west of Croydon Homestead’. The name Croydon was an error for which the present writer is responsible, having supplied it to Campbell and Engel in locality details accompanying the ammonoid specimens in question; the name of the homestead is Carellan, the ammonoid source being in what is here called the Carellan sequence. The ammonoid horizon at Carellan T B. H. JENKINS: CONODONT BIOSTRATIGRAPHY 921 lies in the lower half of the Gn. sp. A Zone. This zone, on comparison of the conodont data with that for the Belgian stratotypes, correlates with either Tn3a or, more prob- ably, Tn3b, and in terms of the Mississippi Valley sequence corresponds to a pre- pinnatus part of the Burlington Limestone. This correlation of the Carellan ammonoid bed modifies the previous conclusion of Campbell and Engel (1963, p. 63) who found that both their ammonoids and brachiopods indicate a correlation with the Kinder- hookian of North America, and the ammonoids suggest a correlation with cul or culla of the European succession’ with ‘a cul age more probable’. The present conodont evidence points to the substitution of early Valmeyeran for Kinderhookian as the North American stage correlative of the Carellan ammonoid horizon, and strongly supports culla in preference to Cul in the European standard, since Schmidts’s (1925) culla indices (Pericyclus princeps and Muensteroceras complanatus ) derived from an isolated Belgian section now correlated with the lower part of the Calcaire de Calonne (Tn3c) (Delepine 1940, p. 7) and/or the uppermost Calcaire de Vaulx (Tn3b2) (Mortelmans 1969, p. 39). The non-sequential nature of the German ammonoid standard has been made clear in recent years (Paproth 1969, pp. 286-289; Matthews 1970a, pp. 115-117). A consequence is that the discrepancy between the proposed conodont-based correla- tion of the Carellan ammonoid bed with Tn3a or Tn3b and the original comparison with culla (= Tn3c) made by Campbell and Engel is more apparent than real for between cul and culla there exists a substantial interval for which there are no presently available formal ammonoid zones, and in which the Carellan ammonoids may well belong. Another large hiatus in the German ammonoid 'standard’ seemingly covers the whole of the Lower and Middle Visean and the lowest part of the Upper Visean (Weyer 1972, pp. 175-184; Groessens 1971). This latter gap is relevant to consideration of the ammonoid fauna from Trevallyn, near Gresford, N.S.W., first reported by Roberts (1961), who recorded the presence of Prolecanites sp. Subse- quently, Roberts (1965a) figured and described his finds as did Brown et al. (1964) who discovered and named Beyrichoceras trevallynense from the same bed, then assigned to the Bingleburra Formation but now considered by Roberts (pers. comm.) to belong to the higher Bonnington Siltstone. The ammonoids were taken to indicate a cuIIIa-cuIIS age in terms of the German standard, later restated as being 'as young as the European Visean zone cullla’ (Campbell and Roberts 1969, p. 263). This correlation conflicts with the conodont evidence, for, regardless of whether the Trevallyn ammonoids belong to the Bingleburra Formation or the Bonnington Siltstone, their source underlies the Flagstaff Sandstone? with its Gn. bulbosus- T. varians fauna for which an early Visean age is indicated (see above), in opposition to the late Visean age deduced for the older ammonoids. To consider briefly the ammonoid evidence, the two internal moulds identified and figured as Prolecanites sp. by Roberts (1965a, pp. 74-75, pi. 12, figs. 7-8, text-fig. 6) yielded only an incomplete suture, lacking the internal segment and the ventral lobe, so that generic allocation is not completely satisfactory. Again, the sutures of Beyrichoceras trevallynensis (Brown et al. 1964, fig. 6) have a character seemingly intermediate between Muensteroceras and Beyrichoceras. Recent Russian work on Lower Visean ammonoids (e.g. Kusina 1971) has resulted in the recognition of several new genera, some with Beyrichoceras-hke sutures (e.g. Winchelloceras Ruzhencev N 922 PALAEONTOLOGY, VOLUME 17 1965, with the type species Goniatites allei Winchell = Beyrichoceras allei of Miller and Garner 1955) and others seemingly transitional with muensteroceratids (e.g. Terek tytes Librovitch 1957). In reviewing the significance of Dinantian conodont correlations and ammonoid distributions Matthews (19706, p. 1162, fig. 1) has inferred, without detailed documentation, that Beyrichoceras may have appeared in pre-anchoralis time. These developments imply that the correlation of the Trevallyn ammonoids with cuIIS or cullla may no longer be as firmly based as when first proposed. The discrepancy between ammonoid- and conodont-based correlations could be resolved by re-identification of the critical species or by revision of their time-ranges, or by showing that the conodonts had been reworked into late Visean strata. To state the case for the conodonts one observes that the critical species ( Gn . bulbosus and T. various ) are highly distinctive forms, unlikely to be confused with any other known species, and have time ranges established in well-documented overseas sections, as outlined above. The N.S.W. specimens show no detected signs of abrasion such as would be expected from reworking, and their occurrence in thin limestones within thick clastic sequences lacking known evidence of disconformity argues against reworking, as does the absence of admixture with forms indicating other horizons. Further, the conodonts in question occupy a distinct part of a well-characterized conodont sequence in N.S.W., and their position in that sequence is consistent with their position in extra-Australian sequences, allowing for minor differences of local time range. Again, Gn. bulbosus is claimed (Thompson and Fellows 1971, p. 89) to be a component of a phylogenetic progression of conodont elements (Gn. bulbosus- Gn. texanus pseudosemiglaber-Gn. texanus texanus) which, if true, enhances its significance as a time indicator, despite the absence from N.S.W. of the later members of the series. It is relevant here to note the occurrence of the second member, Gn. texanus pseudosemiglaber (PI. 119, fig. 3) in the Baywulla Formation of the Yarroll Basin, Queensland, a formation correlated on its Marginirugus barringtonensis fauna with an horizon in N.S.W. well above that with Gn. bulbosus. It is relevant also to note that Butler (1973, text-figs. 2, 9) records the same order of appearance for these three gnathodids in the Mendip area of south-west England as was first reported from Missouri. Finally, there is also the evidence of the brachiopods from Trevallyn, as stated by Roberts (19656, p. 115) and Campbell and Roberts (1969, p. 263); they favour an earlier age than that indicated by the associated ammonoids, specific brachiopod comparisons being possible with forms from the Keokuk and Burlington limestones of the Mississippi Valley. These Osagean limestones are respectively within and mainly below the combined ranges of Gn. bulbosus and T. various in North America. Brachio- pod and conodont correlations are thus in substantial agreement. Among the necessary corollaries of the proposed conodont correlation are great rapidity of sediment accumulation during mid-Dinantian times in the area north of Dungog and Gresford and the relative attenuation of the higher section which is available below the presently accepted base of the Permian for the later Visean and Silesian (i.e. Upper Carboniferous). This raises the possibility of there being one or more undetected hiatuses in the higher Carboniferous section in the Dungog- Paterson area of N.S.W. T. B. H. JENKINS: CONODONT BIOSTRATIGRAPHY 923 Acknowledgements. The writer thanks Dr. John Roberts and Mr. Leon McDonald of the University of New South Wales, who supplied most of the Brownmore samples and provided locality details; Dr. D. A. Bassett of the National Museum of Wales for facilities during 1973; and Mr. D. T. Crane, who assisted with all the later phases of the work. Financial support was received from the Nuffield Foundation, University of Sydney Research grants, and A.R.G.C. REFERENCES bischoff, G. 1957. Die Conodonten-Stratigraphie des rhenoherzynischen Unterkarbon. Abh. hess. Landesamt Bodenforsch. 19, 1-64, pis. 1-6. BRANAGAN, D. F., JENKINS, T. B. H., BRYAN, J. H., GLASSON, K. R., MARSHALL, B., PICKETT, J. W. and VERNON, r. h. 1970. The Carboniferous sequence at Glenbawn N.S.W. Search , 1, 127-129. brown, d. a., Campbell, K. s. w. and Roberts, j. 1964. A Visean cephalopod fauna from New South Wales. Palaeontology, 7, 682-694, pis. 102-103. burton, r. c. 1964. A preliminary range chart of Lake Valley formation (Osage) conodonts in the southern Sacramento Mountains, New Mexico. New Mexico geol. Soc. 15th Field Conference, 73-75. butler, m. 1973. Lower Carboniferous conodont faunas from the eastern Mendips, England. Palaeontology, 16, 477-517, pis. 56-59. Campbell, k. s. w. and engel, b. a. 1963. The faunas of the Tournaisian Tulcumbah Sandstone and its members in the Werrie and Belvue Synclines, New South Wales. J. geol. Soc. Aust. 10, 55-122, pis. 1-9. - mckelvey, j., Roberts, J. and nashar, b. 1969. Carboniferous System. In packham, g. h. (ed.). The geology of New South Wales. J. geol. Soc. Aust. 16, 245-265. and Roberts, J. 1969. Carboniferous System: the faunal sequence and overseas correlation. In packham, G. h. (ed.). The geology of New South Wales. J. geol. Soc. Aust. 16, 261-264. collin son, c., rexroad, c. b. and Thompson, t. l. 1971. Conodont zonation of the North American Mississippian. In sweet, c. w. and Bergstrom, s. m. (eds.). Symposium on conodont biostratigraphy. Mem. geol. Soc. Am. 127, 353-394. collinson, c. w., scott, a. j. and rexroad, c. b. 1962. Six charts showing biostratigraphic zones and correlations based on conodonts from the Devonian and Mississippian rocks of the LJpper Mississippi Valley. Circ. III. St. geol. Surv. 328, 1 32. conil, r., Austin, r. l., lys, M. and Rhodes, f. h. t. 1969. La limite des etages tournaisien et viseen ou stratotype de l’assise de Dinant. Bull. Soc. beige Geol. Paleont. Hydro!. 77, 39-69, pis. 1-2. delepine, g. 1940. Les goniatites du Dmantien de la Belgique. Mem. Mus. R. Hist. nat. Beige, 91, 1-71, pis. 1-5. druce, e. c. 1969. Devonian and Carboniferous conodonts from the Bonaparte Gulf Basin, Northern Australia and their use in international correlation. Bull. Bur. Miner. Resourc. Geol. Geophys. Aust. 98, 1-242, pis. 1-43. — 1970. Lower Carboniferous conodonts from the northern Yarrol Basin, Qld. Ibid. 108, 91-112, pis. 15-18. glenister, b. f. 1960. Carboniferous conodonts and ammonoids from northwestern Australia. C.R. 4 Congr. Intern. Strat. Geol. Carbonif. 1, 213-217. groessens, e. 1971. Les conodontes du Tournaisien superieur de la Belgique. Note preliminaire. Service geol. de Belgique Prof. Paper, 1971, 4, 1-29. — conil, r. and lees, a. (in press). Problemes relatifs a la limite du Tournaisien et du Viseen en Belgique. Bull. Soc. beige Geol. Paleont. Hydro!. hass, w. H. 1959. Conodonts from the Chappel Limestone of Texas. Prof. Pap. U.S. geol. Surv. 294-J, 365-399, pis. 46-50. hill, p. 1971. Carboniferous conodonts from southern Ireland. Geol. Mag. 108, 69-71. jenkins, t. b. h. 1969. Devonian of the Keepit Inlier. In packham, g. h. (ed.). The geology of New South Wales. J. geol. Soc. Aust. 16, 242-243. klapper, g. 1971. Patrognathus and Siphonodella (Conodonta) from the Kinderhookian (Lower Mississippian) of western Kansas and southwestern Nebraska. Bull. St. geol. Surv. Univ. Kansas, 202, 3-14, pis. 1, 2. kusina, l. f. 1971. O nekotorykh novykh i maloizvestnykh rannevizeyskikh (saurskikh) ammonoideyakh. Paleont. Zhur. 1971, 37-48, pi. 4. 924 PALAEONTOLOGY, VOLUME 17 mcwhae, j. r. h., playford, p. E., lindner, a. w. and glenister, b. f. 1958. The stratigraphy of Western Australia. J. geol. Soc. Aust. 4, 1-161. Matthews, s. c. 1970a. A new cephalopod fauna from the Lower Carboniferous of east Cornwall. Palaeonto- logy, 13, 112-131, pis. 25-28. 19706. Comments on palaeontological standards for the Dinantian. C.R. 6e Congr. Intern. Strat. Geol. Carbonif. 3, 1 159-1 163. — and naylor, d. 1973. Lower Carboniferous conodont faunas from south-west Ireland. Palaeontology, 16, 335-380, pis. 35-38. mortelmans, g. 1969. L’Etage Tournaisien dans sa localite type. C.R. 6e Congr. Intern. Strat. Geol. Carbonif. 1, 19-43. palmieri, v. 1967. Upper Silurian-Lower Devonian conodonts from the Wondai ‘Series’ (Murgon District), Queensland. Aust. J. Sci. 30, 67. — 1969. Upper Carboniferous conodonts from limestones near Murgon, South-east Queensland. Pubis, geol. Surv. Qd , 341, 1-13, pis. 1-7. paproth, e. 1969. Die parallelisierung von Kohlenkalk und Kulm. C.R. &’ Congr. Intern. Strat. Geol. Carbonif. 1, 279-291, pis. 1-2. remack-petitot, m. l. 1960. Contribution a l’etude des Conodontes du Sahara (bassins de Fort-Polignac, d’Adrar Reggane et du J. Bechar). Comparaison avec les Pyrenees et la Montagne Noire. Bull. Soc. geol. Fr. 7, 240-262. Rhodes, F. h. t., Austin, r. l. and druce, e. c. 1969. British Avonian (Carboniferous) conodont faunas, and their value in local and intercontinental correlation. Bull. Br. Mus. nat. Hist. Geol. suppl. 5, 1-313, pis. 1-31. Roberts, J. 1961. The geology of the Gresford district, N.S.W. J. Proc. roy. Soc. N.S.W. 95, 77-91. — 1965a. A Lower Carboniferous fauna from Trevallyn, New South Wales. Palaeontology , 8, 54-81. 19656. Lower Carboniferous zones and correlation based on faunas from the Gresford-Dungog district. New South Wales. J. geol. Soc. Aust. 12, 105-122. — and oversby, b. s. (in press). The Lower Carboniferous geology of the Rouchel district. New South Wales. Bull. Bur. Miner. Resour. Geol. Geophys. Aust. schmidt, h. 1925. Die carbonischen Goniatiten Deutschlands. Jb. preuss. Geol. Landesanst. 45 (1924), 489-609, pis. 19-26. Thompson, t. l. 1967. Conodont zonation of Lower Osagean rocks (Lower Mississippian) of southwestern Missouri. Rep. Invest. Mo. geol. Surv. 39, 1-88, pis. 1-6. — and fellows, l. d. 1970. Stratigraphy and conodont biostratigraphy of Kinderhookian and Osagean rocks of southwestern Missouri and adjacent areas. Ibid. 45, 1-263, pis. 1-8. voges, a. 1959. Conodonten aus dem Untercarbon I und II (Gattendorfia- und Pericyclus-Stufe) des Sauerlandes. Paldont. Z. 33, 266-314, pis. 33-35. — 1960. Die Bedeutung der Conodonten fur die Stratigraphie des Unterkarbons I und II (Gattendorha- und Pericyclus-Stufe) im Sauerland. Fortschr. Geol. Rheinld Westf. 3, 197-228. weyer, d. 1972. Trilobiten und Ammonoideen aus der Entogonites nasutus-Zone (Unterkarbon) des Biichenberg-Sattels (Elbingeroder Komplex, Elarz), Teil I. Geologie . 21, 166-184. zikmundova, j. 1967. Conodont fauna of the Scaliognathus anchoralis Branson and Mehl Zone in the Ponikev Shales of Nisky Jesenik area. Vest, ustred. Ust. geol. 42, 449-451. pis. 1-4. [In Czech.] T. B. H. JENKINS Department of Geology and Geophysics University of Sydney Revised manuscript received 14 February 1974 New South Wales, Australia Note added in proof. Taxonomic revisions of some of the species cited in this paper are in press. Dr. R. L. Austin (pers. comm. 23 July 1974) is proposing (in press, Bull. geol. Surv. Gt Br.) a new genus to receive forms similar to those herein referred to Patrognathus ? cf. capricornis. In another paper, Zeigler, Sandberg and Austin (in press, Geologica et Palaeontologica ) revise the double-rowed spathognathodid elements of the late Famennian and early Tournaisian, resurrecting Bispathodus Muller, 1962 to receive, inter alia , Spathognathodus costatus costatus of this paper as Bispathodus aculeatus aculeatus (Branson and Mehl 1934), Sp. plumulus cf. shirleyae as B. aculeatus plumulus, Sp. anteposicornis as B. aculeatus anteposicornis and Sp. costatus sulciferus as B. spinulico status (E. R. Branson). NON-VASCULAR LAND PLANTS FROM THE DEVONIAN OF GHANA by W. G. CHALONER, M. K. MENSAH and M. D. CRANE Abstract. Two species of fossil plant ( Spongiophyton nanum Krausel, S. lenticulare (Barbosa) Krausel) preserved as compression fossils with cuticles, are described from the ?Middle Devonian of Ghana. S. nanum , first described from the Devonian of Parana, Brazil, is a plant of dorsiventral thalloid organization, branching dichotomously, and with a series of large pores on the presumed upper surface. It has a cuticle far thicker than that of most vascular land plants. On its inner surface the outlines of cells of the underlying tissue may be seen. Nothing else is known of the inner tissue of the plant. Material of 5. lenticulare is similar, but is only seen as small fragments. Spongiophyton combines an external morphology resembling that of some algae and liverworts, with a thick cuticle unknown in those groups. In this and its dorsiventral organization, it appears to show adaptation to a terrestrial environment. It may be compared with Foerstia (? = Protosalvinia) and with Parka , but shows significant differences from these and other genera of Devonian thalloid plants. This paper is an account of two species of Spongiophyton , a thalloid plant, repre- sented by compression fossils with ‘cuticle’, from the Takoradi Sandstone of Ghana. The material comes mainly from a horizon within the Sekondi Series lower than that which yielded lycopods believed to be Lower Carboniferous (Mensah and Chaloner 1971). The plants described here are preserved as compression fossils in a matrix of shaly sandstone and are believed to be of Middle Devonian age. This age is based on the identity of the plants themselves with specimens from Brazil dated as Middle Devonian on palynological evidence. The plants are in the form of fragments of branched tubes, representing the outer membrane of an originally more or less cylindrical structure. Two species are recognized, Spongiophyton nanum Krausel and S. lenticulare (Barb.) Krausel, which were described from the Middle Devonian of Parana, Brazil, by Krausel (1960). Examination of these Ghanaian fossils and parti- cularly the use of thin sections and scanning electron microscopy reveals new features not previously reported in the Brazilian material. These two species are redescribed on the basis of our observations of the Ghanaian specimens. Their bearing on the age of the Takoradi Sandstone is then considered, and the affinity of Spongiophyton and some comparable genera reviewed. PREPARATION AND EXAMINATION OF SPECIMENS The plants described here may be observed directly on the bedding surfaces and in the matrix of sandstone but all the material illustrated was extracted either by maceration, or by removing fragments with needles. Most of the specimens were obtained from bulk maceration, by treating hand specimens of sandstone with concentrated nitric acid for up to 24 hours, followed by decanting and repeated washing in water. The swelling effect of the nitric acid acting on the plant material effectively disaggregated the matrix, so that after washing, the plant fragments were sieved from the sand and clay fraction. The larger pieces were then examined under a binocular microscope with top illumination. Individual specimens were given further oxidative treatment with Schulze’s solution (a saturated solution of potassium chlorate in nitric acid), for from 5 to 20 minutes, until the cuticular material became translucent. Subsequent treatment with ammonia, commonly used for coalified plant cuticles, was found to be unnecessary, and merely caused darkening of the [Palaeontology, Vol. 17, Part 4, 1974, pp. 925-947, pis. 120-124.] 926 PALAEONTOLOGY, VOLUME 17 plant material. The macerated fragments were then mounted directly in glycerine jelly, or were dehydrated in alcohol and mounted in Canada balsam for observation by transmitted light. Specimens used for SEM were mounted on stubs using either double-sided Sellotape or ‘Durofix’ cement, or occasionally by using silver dag as an adhesive. They were then coated with carbon followed by gold-palladium, using an SEM planetary specimen holder in an Edwards High Vacuum Coating Unit. Photographs were taken on both a Cambridge ‘Stereoscan’ Mark II and an S600 scanning microscope. In order to eliminate the possibility of artifacts of cuticle structure arising from the maceration procedure, some SEM observation was carried out on fragments of the plant removed from the matrix with a mounted needle, without further chemical treatment. The structure of the plant was further studied by embedding and sectioning pieces of the cuticle. Portions of the plant were transferred to acetone after Schulze’s solution, and were then embedded in Araldite resin in capsules. The resin-embedded material was then sectioned on a Reichert microtome and the sections mounted in Canada balsam. DESCRIPTIONS Spongiophyton nanum Krausel (1960) Plate 120, figs. 1-4, 7-9; Plate 121, figs. 1, 3-7; Plate 122, fig. 1 ; Plate 123, figs. 1-3; Plate 124, figs. 1-2; text-fig. 1 General features. This description, based on observation by light microscopy, SEM, and thin section of the new Ghanaian material, amplifies that given by Krausel (1960). A formal emendation of his diagnosis is given below (p. 945). The specimens here attributed to S. nanum consist of flattened tubes of cuticle, with rounded apices, sometimes branching once or twice. The flattened tubes range from 2 to 5 mm across, the majority of specimens being at the upper end of this range. The longest fragment found is 25 mm long. A few specimens show one or two suc- cessive more or less equal dichotomies, apparently in the same plane. The rounded apices, where seen, are of thinner cuticle than the remainder perhaps representing regions of growth; they are frequently damaged, or a branch is simply truncated, with the apex missing. No apparently intact basal end of any of the fragments was seen; all have a truncated (broken) base. In a few specimens, a dome-like, or basally contracted globular lobe appears laterally on one branch of the thallus, representing a relatively short branch at right angles to those normally lying in the plane of compression (PI. 120, figs. 8, 9). Each of the flattened tubes shows a differentiation in the morphology of its two surfaces; one, the ’poral surface’, has the thickest cuticle, and shows a series of circular to oval EXPLANATION OF PLATE 120 Spongiophyton from Komenda, Ghana. Figs. 1, 2. Spongiophyton nanum, apical portion of thallus extracted by maceration, x 5. 1, poral surface. Note broken edge of dichotomy at left. 2, aporal surface of same. Figs. 3, 4. 5. nanum. Poral and aporal surfaces of thallus showing dichotomy, x 5. Figs. 5, 6. S. lenticulare. Short length of intact cuticular tube (long axis upright, torn edges above and below), x 3. Pores are present on both surfaces, but are more abundant in fig. 5. Fig. 7. Piece of S. nanum thallus flattened so that poral face appears at left side, aporal at right, x 5. Figs. 8, 9. S. nanum , portion of thallus with short ?upright lateral branch, with its apex collapsed showing cellular reticulum on inner cuticle surface. (Fig. 8, x 20; fig. 9, x 50.) (Scanning electron micrograph.) PLATE 120 CHALONER et al., Spongiophyton 928 PALAEONTOLOGY, VOLUME 17 text-fig. 1. Diagrammatic reconstruction of the Spongiophyton nanum cuticle, at different magnifications, summarizing details revealed by light microscopy and scanning electron microscopy. a, external detail, showing two successive unequal dichotomies and one additional small lobe on the poral surface. The density of distribution of pores varies along the length of the thallus. b, reconstructed segment of thallus showing pores on thicker, supposedly upper, face and slits and folding of thinner aporal face. c, section of pore showing bevelled edges, and cellular reticulum on inner face of cuticle. Cut edges of cellular reticulum are brought together, where the tube is folded at left. d, section of cuticle shows borings in various orientations, fusiform and more or less spherical cavities, and ridges of cellular reticulum on inner face. pores typically 200-300 ^ m in diameter scattered more or less irregularly over the surface. The other, the ‘aporal surface’, has a thinner cuticle, and commonly shows some longitudinal folding and tearing; this surface generally lacks pores, although a few may be present along the edges (PI. 120, figs. 1-4). The ‘poral’ cuticle ranges from 24 ^m to 84 in thickness, while the ‘aporal’ cuticle ranges from 24 to 36 /xm. These measurements are based on ten specimens showing suitable more or less transverse fractures of the two surfaces. The ratio of poral to aporal cuticle thickness ranges typically between 2:1 and 3:1, although occasionally specimens are seen which show little difference in thickness between the two surfaces. Where the aporal surface has longitudinal tears (PI. 120, figs. 2, 4) these show match- ing edges, and appear to represent a simple physical splitting in the thinnest part of the cuticular tube. This may have occurred simply as a result of the mechanism of the collapse of the tube, with the thinner aporal surface coming under tension as the thicker poral surface became splayed out; or it may be as a result of greater vulner- ability of the thinner cuticle to the maceration process, although this seems less likely. In most cases the flattening of the cuticular tube on fossilization was such that the CHALONER ET AL.\ S PO NG IO P H YTO N 929 poral and aporal surfaces coincided with the upper and lower faces of the flattened tube. This suggests that there was therefore some preferred orientation related to the poral/aporal differentiation. The simplest explanation would be that in life the tube was somewhat flattened in the poral/aporal plane. However, a few specimens show part of each flattened tubular cuticle surface with pores, and part without (PI. 120, fig. 7). Seventy-two fragments of S. nanum were picked from the Komenda material, show- ing clearly the flattened tube structure with both edges intact ; of these, 56 showed the two surfaces to coincide with the poral/aporal faces, while 13 showed the poral/ aporal boundary lying within the flattened face (cf. PI. 120, fig. 7). This suggests some degree of preferred orientation, but clearly not such as would be shown by a flattened thallus as of, say, Fucus or a thalloid liverwort. The preferred plane of orientation in the sediment might have been simply the result of dichotomous branching of a more or less cylindrical thallus in the plane of the poral/aporal differentiation, as seen in the specimen of Plate 1 20, figs. 3-4. Of the total 72 specimens examined in this respect, only three showed pores more or less uniformly on both surfaces. These are considered further below. The pores are typically about five diameters apart, but there is a good deal of varia- tion. Along the length of a tube there are zones of relatively high concentration (with pores up to about one diameter apart) and others where they are more sparse. The resulting zonation of pore density (text-fig. 1) appears to show no consistent relation- ship with dichotomy of the thallus. Each pore is bevelled or countersunk as seen from the outer surface (PI. 121, fig. 5; text-fig. lc); from the inner surface it lies in a slight depression (i.e. the cuticle thins gradually towards the pore). While the pores are more or less randomly scattered, there are never large pore-free areas on the pore- bearing surface, nor are two pores ever closer than an average pore radius. Cuticle structure. The cuticle forming the tubes superficially resembles fossil cuticle of vascular land plants (e.g. conifers) prepared by maceration. It is brown to black in colour, with a good deal of variation presumably due to vagaries of preservation history. The better-preserved fragments are relatively tough and even flexible to some extent. The outer surface is more or less smooth, although it may be marked with an irregular series of vermiform borings attributed to post-mortem microbial attack (see below). We use the term ‘cuticle’ throughout this article in the broad sense of an outer resistant covering of what was evidently a cellular organism. Although its super- ficial appearance and behaviour under maceration are similar to fossil vascular plant cuticles we do not wish to press the implications that the covering of Spongiophyton is cuticle in this rather precise sense. We prefer this term, used broadly and variously in different biological contexts, to any alternative (e.g. meristoderm, used of the Foerstia ‘cuticle’) which has more specific implications of algal affinity. The inner cuticle surface shows a cellular reticulum formed by cuticular ridges comparable to those left on the inner face of typical higher plant cuticles after macera- tion. The cellular reticulum shows rather elongated cells, typically 40 long and 20-30 (tun broad, generally aligned more or less parallel to the long axis of the tube, except in the vicinity of pores where a more or less radiating orientation may be 930 PALAEONTOLOGY, VOLUME 17 followed. These cell outlines show clearly by transmitted light. Scanning microscopy (PI. 121, fig. 3) and thin sections (PI. 124, figs. 1, 2) show that the ridges representing cell outlines are on the inner surface of the cuticle. The cellular outlines are less clear on the cuticle of the thin, aporal surface. They are there also of more markedly elongated shape and linear alignment. The extent to which the cuticular ridges appear pressed together at the edge of the flattened tube (PI. 124, fig. 2) suggests that the flattening was not an original feature of the plant. This implies that the thallus might have been more or less elliptical in cross-section, but was not a flat ribbon-like struc- ture in life. The cuticle shows no sign of having been compressed during compaction of the sediment ; it appears equally thick where it lay horizontally, as at the folded edge. We can see no evident explanation for this, but note that Carboniferous mega- spores seen in coal thin-sections, even when completely flattened, similarly show no reduction of exine thickness in the vertical dimension compared with that seen at the folded edge. The cuticle was examined in thin section, following resin embedding, and broken faces were examined by SEM. The texture and colour of the cuticle is not uniform, being generally darker towards the outer surface. Within the cuticle there are a num- ber of cavities giving it a locally spongy or foam-like texture (PI. 121, fig. 7; PI. 124, fig. 1), and this seems to become more pronounced in macerated material. These lacunae are largely enclosed within the cuticle although they may abut on the inner or outer surface. There are in addition occasional lens-shaped cavities in the cuticle perhaps associated with separation of layers of lamellated cuticular material. Indepen- dent of these, and more irregular in distribution, are a series of what are here referred to as ‘borings’, which abut mainly on the outer surface, apparently representing the site occupied by some micro-organism (PL 121, figs. 4, 6, and PI. 124, fig. 1). They vary from 2 to 5 /xm in diameter, and occasionally appear to divide. They appear to be concentrated particularly towards the outer surface of the cuticle. They show various orientations within the cuticle, and sometimes appear at the surface as grooves or gouges. It is possible that they represent the activity of some parasitic organism during the life of the Spongiophyton, but in view of their extensive occurrence they seem more likely to be a post-mortem (saprophytic) invasion of the plant. The result- ing grooves in the surface may take the form of a ‘negative reticulum’ pattern, simulating that of the underlying cell outlines (and perhaps influenced by them?). This is seen in the outer cuticle surfaces shown in Plate 121, figs. 1 and 4. EXPLANATION OF PLATE 121 Scanning electron micrographs of Spongiophyton from Komenda, Ghana. Fig. 1. S. nanum, apex of thallus. Note irregular, cell-like pattern of borings, x 60. Fig. 2. S. lenticulare. Inner surface of cuticle, x 300. Fig. 3. S. nanum. Inner surface of cuticle, showing pores, x 100. Fig. 4. S. nanum. Outer surface of cuticle, showing borings, x 100. Fig. 5. S. nanum. Outer surface of cuticle of another specimen, showing bevelled edges of pores, x 150. Fig. 6. S. nanum. Outer surface of cuticle showing borings simulating cellular pattern, x 75. Fig. 7. S. nanum. Oblique view of fractured edge of cuticle, x 400; outer surface is at top left. (Compare text-fig. 1 and PL 5, fig. 1.) PLATE 121 CHALONER et al., Spongiophyton cuticle 932 PALAEONTOLOGY, VOLUME 17 These borings are very similar to those described in Silurian arthropod cuticles as ‘vermiform tubules’ by Rolfe (1962) and as ‘thallophyte borings’ in the cuticle of Cretaceous decapods by Taylor (1971). We follow Taylor in believing that chitrids (a group of aquatic phycomycetes) are the most likely causal organisms, both in his structures and our own. Rolfe’s description of ‘haphazardly arranged, anastomosing, meandering and convoluted tubules’ aptly describes the borings that we have observed in the Spongiophyton cuticles. Taylor’s borings are unbranched, but branching is simulated by their intersection; their diameters range from 0-3 to 15 ^m, which straddles the range of our observations. The only reservation that might be made concerns Taylor’s suggestion that the chitinous content of the substrate may have been important in the metabolism of the invading chitrid. If, as we believe, our Spongiophyton cuticle was comparable to that of higher plants, and so largely of lipid composition, then the invading fungus cannot have been chitino- philic. It might also be observed that when fungal hyphae penetrate leaf cuticle they usually do so by a single direct passage of a hypha through the cuticle, in contrast to the meandering borings seen here. This adds to our uncertainty about the role of the cuticle as a metabolic substrate for the micro-organism. However, the size and character of the borings favours fungi, rather than any other group, as the most likely causal organisms. Unmacerated specimens of Spongiophyton nanum removed from the rock matrix with a needle and embedded and sectioned, show only occasional patches of black structureless substance between the two cuticle layers. Presumably this represents a residue of the original internal tissue but its featureless appearance did not encourage us to attempt any further investigation. Incorporation and fossilization. The conditions of deposition of the basal Takoradi Sandstone are regarded by Crow (1952) as representing a transition from estuarine to fluviatile conditions during a phase of emergence. There are no invertebrate fossils associated with the Spongiophyton. While the majority of specimens, as indicated, are flattened in the poral/aporal plane, they are not all oriented the same way in the sedi- ment. Specimens from Komenda showed some tendency for the poral surface to lie uppermost in the sediment; of 30 fragments of S. nanum showing both faces, on five orientated rock hand-specimens, 22 (rather over 70%) had the poral surface facing upwards. From the coarse-grained, clastic nature of the sediment, and hence its presumably relative rapid rate of accumulation, it is assumed that these plant fossils were transported down the drainage system into the site of deposition, rather than representing marine organisms washed into a non-marine environment. This is also consistent with their fragmentary nature. The preferred orientation in the sediment is not very great and might be the result of some internal tissue differentiation associated with the poral surface (e.g. air cavities with trapped air or gas underlying the pores— cf. the liverwort Marchantia). Spongiophyton lenticular e Krausel (1960) Plate 120, figs. 5-6; Plate 121, fig- 2; Plate 122, figs. 2-3 This species has been fully described by Krausel (1960), and the Ghanaian material adds relatively little to our knowledge. As in his case, our specimens are much more fragmentary than those of S. nanum. The largest of the Ghanaian specimens is that in Plate 120, figs. 5-6, which shows part of a cuticular tube 12 mm in diameter, and irregularly broken at both ends. The fusiform pores are clearly seen, and show much the same range in size and density of distribution that he reports. The pores in the Ghanaian S. lenticulare have not the clearly defined circular to oval outline typical of S. nanum but rather fusiform slits, at the margin of which the edges of thin cuticle may be seen to curl back (PI. 121, fig. 2). The cuticle itself is basically similar to that of S. nanum , but the cell outlines, showing as a reticulum on the inner face of the cuticle CHALONER ET AL .: S PO NG IO P H YTO N 933 are larger and narrower (PI. 121, fig. 2) but also generally more obscure. They give a more markedly striate or longitudinally grained appearance to S. lenticulare (PI. 122, fig. 2). Krausel has described vividly how prolonged maceration causes tears to appear and expand in the mem- brane, so forming ‘pores’ where they did not exist previously, except perhaps in the form of incipient lines of weakness. Although we have not repeated Krausel’s progressive maceration sequence, it is evident that his basic observations are applicable to our material; while fusiform pores are present in unmacerated cuticle, the formation and expansion of further slits takes place with prolonged oxidative maceration in Schulze’s solution. We accept Krausel’s interpretation that S. lenticulare is a distinct species of the genus Spongiophyton. However, the fragmentary nature of the material leaves uncertain the external form of the whole plant. We have seen no pieces indicating a clear dichotomy, so that we have direct evidence only that the thallus was covered with a (tubular) cuticular layer, showing pores and an internal cellular reticulum. We have not sufficient pieces of intact tube to establish a clear differentiation of poral and aporal surfaces; indeed, some pieces (e.g. PI. 120, figs. 5, 6) show the pores distributed irregularly on both faces, while others suggest some differentiation of pore concentration between the two surfaces, as in S. nanum. We do not regard the supposed ‘internal’ structure of irregular wrinkles and depressions described and illustrated by Krausel as representing original features of the cuticle; this is discussed further below. LOCALITY DATA The plants forming the basis of this study come principally from the base of the Takoradi Sandstone of the Sekondi Series close to its contact with the underlying Elmina Sandstone at Komenda, on the coast of Ghana about 28 miles east of Sekondi. The exposure (our locality 1) is in the beach below Komenda College (longitude 1 ° 29' 30" West, latitude 5° 2' 40" North) at the bottom of the cliff directly opposite the Principal’s bungalow. Here, the basal part of the Takoradi Sandstone consists of 30 m of thin- to medium-bedded sandstone with shaly partings. The material was collected from the uppermost 10 m of this member which becomes exposed only at low tide. More fragmentary material, typically fawn or yellow in colour as though ‘naturally macerated’ before being incorporated in the matrix, occurs (our locality 2) in grey fine-grained argillaceous sandstone from Essipon, 500 m east of the Coconut Beach Hotel, from the upper part of the Takoradi Shales on top of the shaly member which yielded the Essipon invertebrate fauna reviewed by Crow (1952, p. 32). This appears to be at a horizon above that yielding the plant macrofossils of ArchaeosigiUaria and Lepidodendropsis (Mensah and Chaloner 1971). In addition, a single specimen of Spongiophyton nanum has been identified on a single unlogged borehole core (our locality 3) from the Accra Shales near the UTC warehouse and wholesale stores on the High Street in Accra (over 150 km to the east of localities 1 and 2). Fragments of Spongiophyton sp. have also been seen in the Takoradi Sandstone (our locality 4) just above the Elmina Sandstone on the southern side of the hill on which Fort Convaadsborg (Fort St. Jago) is situated at Elmina (1° 21' West, 5° 5' North). All the figured material, and that on which the descriptions are based, comes from locality 1. We believe that this (and probably that from locality 4) represents primary fossilized material, whereas the more fragmentary paler specimens from localities 2 and 3 could well be reworked. This supposition is based partly on the state of pre- servation, and the obvious facility with which this material would survive reworking, but also partly on our belief that the earlier (locality 1 ) occurrence is M iddle Devonian, while the horizon represented by locality 2 is probably early Carboniferous in age. All the figured and cited material of Ghanaian specimens has been deposited in the Geology Department, University of Ghana. Duplicate material is in the Department of Palaeontology, British Museum, Natural History. 934 PALAEONTOLOGY, VOLUME 17 DISCUSSION Comparison of Ghanaian and Brazilian material. Our study of Krausel’s material of Brazilian Spongiophyton housed in the Forschungsinstitut Senckenberg at Frankfurt am Main was confined to light microscopy. We attempted to remount some specimens for observation by SEM, but were unable to free them sufficiently from the mounting medium, which appeared to be denatured glycerine jelly. Microscopic observation confirmed the identity of our two species with Krausel’s, with the exception of one of his specimens recorded as S. lenticular e, which was evidently S. nanum inadvertently mislabelled. Agreement extended to the presence of the ‘borings’ clearly recognizable in the Brazilian S. nanum. Krausel put much emphasis on an aspect of his cuticles which he regarded as representing a characteristic feature of Spongiophyton, namely a series of irregular anastomosing ridges or corrugations of the cuticle which he interpreted as remains of original structure. This he referred to as a spongy structure (Schwammstruktur) com- posed of irregular longitudinal ribs (Langsbalken). Having carefully examined Krausel’s specimens, we believe that these irregularities are an effect of preservation peculiar to his material, perhaps associated with imprinting of the angular matrix particles on the cuticle. This ‘spongy structure’ is evident only in some parts of his specimens and may appear (PI. 123, fig. 3) in immediate juxtaposition to clearer areas of the cuticle showing a cellular reticulum agreeing closely with that seen in our Ghanaian material (PI. 123, fig. 2). We therefore regard this supposedly original ‘Schwammstruktur’ of Spongiophyton as a product of a particular environment of preservation. We have seen no evidence of preservation of any internal tissue of either Spongiophyton species in Krausel’s material or our own. But the reticulum of elongated cell outlines on our cuticles of S. nanum strongly suggests the presence of an internal cellular tissue probably of parenchymatous organization as seen in typical higher plants, rather than the more or less rounded cells seen in the ‘cuticle’ of Foerstia (figured here on PI. 124, fig. 3) or the meristoderm of a brown alga (PI. 124, fig. 4). We believe that the cellular fragments of early Devonian age figured by Mortimer and Chaloner (1972, PI. Ill, figs. 1-3) from South Wales and boreholes in south-east England, referred to ‘ Spongiophyton sp.’ and ‘cf. Spongiophyton' should probably not be referred to this genus. They might equally well be com- pared to Foerstia (' 1 Protosalvinia ) or Orestovia (see table 2). In the present state of our knowledge, such fragments are perhaps better left generically unassigned. THE AGE OF SPONGIOPHYTON FROM GHANA AND BRAZIL Krausel (1960) originally described five species of Spongiophyton from Brazil ( S . lenticulare, S. nanum, S. minutissimum, S. articulatum, and S. hirsutum). The last of these he later (Krausel and Venkatachala 1966) removed from the genus, and assigned EXPLANATION OF PLATE 122 Fig. 1. Cuticle of S. nanum from Komenda, photographed by transmitted light, showing cellular reticulum, two pores, and borings (at left-hand margin), x 160. Fig. 2. Cuticle of S. lenticulare from Komenda showing pore and cellular reticulum, x 64. Fig. 3. Cuticle of S. lenticulare from the Ponta Grossa Formation, Parana, Brazil (Krausel Collection), x 64. (Forschungsinstitut Senckenberg, slide no. FO 258/1.) PLATE 122 CHALONER et al., Spongiophyton cuticle 936 PALAEONTOLOGY, VOLUME 17 to a new genus Aculeophyton based on a Siberian Devonian species. Of Krausel’s four remaining species of Spongiophyton we have found only two in Ghana— S. lenticulare and S. nanum. Krausel believed his material from Brazil to be of early Devonian age, but subsequent work by Lange and Petri (1967) taking account of palynology and marine faunas favours a Middle or Upper Devonian assignment. Lange and Petri (1967, p. 25) state that "all the six localities from which Krausel, 1954, described different species of Spongiophyton belong to the Middle Devonian Sao Domingo Member’ (of the Ponta Grossa Formation, Parana Group). The Sao Domingo Member actually transgresses two palynologically based zones (‘bio- stratigraphic intervals’ of those authors, p. 26). These are the D4 and D5 Zones of Daemon et al. (1967) which are on well-documented palynological evidence equated approximately with the Givetian (Middle Devonian) and Frasnian (Late Devonian) respectively. However, this Frasnian palynological zone is recognized only in sub- surface rock in Parana. The best approximation for the age of the Brazilian Spongio- phyton is therefore, on available palynological evidence, Givetian (late Middle Devonian). This suggests that the Takoradi Beds (Takoradi Shales plus Takoradi Sandstone) probably span an interval from approximately Givetian to early Carboni- ferous age. This and other stratigraphic implications of the occurrence of Spongio- phyton in the Sekondi Series are considered further by one of us in a paper now in press (Mensah 1973). THE NATURE OF SPONGIOPHYTON Krausel (1960) and Krausel and Venkatachala (1966) have discussed the possible affinity of Spongiophyton. The Brazilian species of that genus, and in particular S. lenticulare, were at first thought to be cuticle remains of a lycopod. They were indeed initially assigned to a genus of Devonian lycopod, as Haplostigma lentieularis Barbosa (1949). Krausel (1960) subsequently showed that this assignment was un- acceptable. On Barbosa’s interpretation the holes (pores) in the cuticular tube were interpreted as the sites of (abscissed) leaves of a lycopod, of which the cuticle repre- sented the stem surface (compare, for example, PI. 122, fig. 1 with the lycopod stem cuticle of Arehaeosigi/laria essiponensis ; Mensah and Chaloner 1971, PI. 65, fig. 1). When we first encountered fragments of Spongiophyton nanum , we tried to interpret them as lycopod cuticles just as Barbosa had done, before realizing their identity with Krausel’s genus. Two principal features of Spongiophyton ‘cuticles’ weigh against their representing stem surfaces of lycopods or other vascular plants. Firstly the dorsiventral nature of EXPLANATION OF PLATE 123 Figs. 1-3. Cuticle of Spongiophyton nanum photographed by transmitted light. 1, from poral surface of a specimen from Komenda, x 250. 2, 3, from poral surface of a specimen from Parana, Brazil (Forschungs Institut Senckenberg, slide no. TO 265), x 160. 2, shows the cellular reticulum, abundant trace of borings, and the irregularity of cuticular surface which grades into the condition seen in the following figure. 3, part of the same cuticle, showing the irregularity (due to preservation?) referred to by Krausel as ‘spongy structure’. PLATE 123 CHALONER et al., Spongiophyton cuticle 938 PALAEONTOLOGY, VOLUME 17 the tubes both in the distribution of the pores and in cuticle thickness. This lack of pores (putative sites of leaf attachments) on one surface is unknown in any lycopod, living or fossil. Secondly, and related to this, the pore arrangement is apparently random on the poral surface, and does not show the regular whorled or spiral con- figuration characteristic of all known lycopods. These two features taken together seem to rule out their attribution to any known vascular plant. Once we dismiss a relationship of these cuticular tubes to a particular group of plants, their affinity must be reconsidered ; even the possibility of animal versus plant origin must be entertained. Spongiophyton nanum could be compared superficially with at least three groups of colonial animal; the graptolites (in the broadest sense), the sponges, and the bryozoa. Several specimens and SEMs were examined by Dr. Adrian Rushton and Dr. Dennis White (Institute of Geological Sciences) who reject the possibility of graptolite affinity. Miss P. L. Cook (Department of Zoology, British Museum, Natural History) and Dr. P. Sandberg (Department of Geology, University of Illinois) examined specimens as possible bryozoa. They set aside this possibility emphasizing particularly that no known Palaeozoic bryozoa have an organic outer membrane of the considerable thickness of Spongiophyton nanum. They also suggest that the lack of any demarcation between the pores (if these were to be construed as the zoecia) is entirely unlike any bryozoan. Dr. R. P. S. Jefferies (British Museum, Natural History) also examined material, and expressed the opinion that its structure could not be interpreted in terms of any animal group known to him. We should like to thank these colleagues for their help, and permission to record their opinions here. An attempt to postulate the systematic position of a fossil known only from its cuticular covering, must be both tentative and hazardous. In the case of Spongio- phyton nanum the principal features on which such speculation can be based are as follows: (1) It has a thalloid body, known to dichotomize at least twice successively, with rounded apices to the lobes. (2) The whole organism has a very thick resistant cuticle, far thicker than most land plant cuticles, with an elongate-cellular reticulum on the inner surface. (3) The cuticle is penetrated by pores mainly on one surface only; this, and differentiation in cuticle thickness, demonstrates a dorsiventral organization. EXPLANATION OF PLATE 124 Figs. 1,2. Vertical section of cuticle of Spongiophyton nanum transverse to thallus long axis, from Komenda, photographed by transmitted light. 1, the inner face of the cuticle is below, showing the ridges corre- sponding with underlying cells of the plant, in life. Note borings, and the various small lacunae (cavities) within the cuticle, x 250. 2, the same specimen, showing the folding of the cuticle (right) at one edge of the flattened cuticular tube, where the ridges on the inner cuticle surface are brought into juxtaposition at the fold, x 85. Fig. 3. ‘Cuticle’ (outer surface of meristoderm?) of Foerstia ( IProtosalvinia ) ohioensis from the Upper Devonian of Ohio, U.S.A., photographed by transmitted light, x 250. Fig. 4. Paradermal section through the meristoderm of Fucus vesiculosus, photographed by transmitted light, x 640. These two figures show the characteristically rounded appearance of the cellular reticulum of Foerstia and Fucus which contrasts with the elongated cells of Spongiophyton (cf. PI. 123, fig. 1). PLATE 124 -J r CHALONER et al., Spongiophyton, Foerstia , and Fucus cuticles 940 PALAEONTOLOGY, VOLUME 17 (4) The fossilized (coalified) substance of the Spongiophyton cuticle shows neither the laminar units nor the fibrous elements demonstrated for some invertebrate cuticles, e.g. the Eurypterid cuticles of Dalingwater (1973). The conclusion that Spongiophyton is a plant is based on consideration of the four items above taken collectively, plus the rejection of the organism from each of the three plausible animal phyla cited earlier. It has to be conceded that there is no basis for totally rejecting animal affinity. Accepting a plant affinity on this tentative basis the character of the original organism can be considered. A flattened, thalloid dichotomizing habit occurs in three groups of living plants— algae (principally red, Rhodophyceae, and brown, Phaeophyceae) ; in liverworts (Hepaticae), and in lichens. Within these several groups, two situations may be con- trasted. Large thalloid algae, adapted to live in the sea as anchored, submerged organisms, normally of more or less upright habit, show no differentiation of the two surfaces of the thallus. These contrast with the thalloid liverworts and lichens (both basically terrestrial organisms, although a few of the former are aquatic) which typically grow with one face applied to a rock or soil surface and show in varying degree both internal and external dorsiventral differentiation. The thalloid liverwort Marchantia, for example, shows gas-exchange pores in its upper surface only, with rhizoids and scales on the lower surface. It is worth noting that those few thalloid brown algae (e.g. Fucus vesiculosus , Pelvetia canaliculata ) which have developed free- living growth forms adapted to a mud-flats environment in salt marsh (see Baker 1912) show no external differentiation of an upper and lower surface. This is pre- sumably because as free-living (non-attached) forms they can be turned over freely with each rising tide, although they remain more or less safely ‘trapped’ within the marsh-pan environment. There are marine algae with a more or less thalloid habit which grow encrusting or appressed to rocky substrates, and which show some associated dorsiventrality of internal organization (e.g. Ralfsia , Zanardinia , and the Aglaozonia (sporophyte) stage of Cutleria , all in the Phaeophyceae; and Peysonnelia , Hildenbrandia, and Rhizophyllis among Rhodophyceae; see Fritsch 1952). The dorsiventral character of S. nanum taken on its own should therefore not perhaps be construed unequivocally as an adaptation to a terrestrial habit. But its enormously thick cuticle certainly suggests an environment in which it was necessary for the plant to drastically restrict water loss. The so-called cuticle of the Fucales is said by Fritsch to be of a mucilaginous composition, while that of the red algae is said to be ‘probably of a pectic nature'. The labile character of both these groups of substances makes the preservation of a cuticle of such composition extremely unlikely in the sedimentary environment in which Spongiophyton occurs. Among living vascular plants it may fairly be said that there is a broad correlation between cuticle thickness and aridity of the environment. However, nearly all conifers— regardless of habit— have relatively thick cuticles, as do many flowering plants of saline habitats. But the cuticle of S. nanum , over 80 gm in thickness, exceeds by a factor of four times what is rated as a ‘thick cuticle’ in fossil gymnosperm leaves (cf. 6-8 jit m in Pagiophyllum , Kendall 1948; 10-20 g.m in Pachypteris, Harris 1964). This cuticle thickness, alone, is perhaps one of the most remarkable features of Spongiophyton. CHALONER ET AL.: S PO NG I O P H YTO N 941 It must be conceded that a thick waxy covering could confer flotation on the Spongiophyton plant, much as the waxy cell walls of Botryococcus may be associated with its floating habit. One could then visualize a growth habit such as that of the water fern Azolla, or the liverwort Riccia natans, free-floating on a freshwater surface. However, we feel that a more plausible habitat for the plant, suggested by its thick cuticular covering and dorsiventral organization, is a terrestrial soil surface. This would also account for the few specimens which, while showing dorsiventral organiza- tion and branching in the assumed horizontal plane, occasionally have an upwardly directed lobe of quite limited growth (PI. 120, figs. 8, 9 and text-fig. 1). It is possible that such lobes (and indeed the rounded apices, as in PI. 120, figs. 1 and 2) represent reproductive regions comparable to the receptacles of living brown algae such as Fucus. The density of pores in such regions appears to be equivalent to that seen else- where, and does not particularly encourage this hypothesis. The specimen of Plate 120, figs. 8 and 9 shows such a lobe, of which the apex (probably with an originally thinner cuticle) has collapsed, revealing the cellular reticulum on its inner face. The few specimens of S. nanum with pores on all faces of a short cuticular tube may also repre- sent such upright lobes of the generally dorsiventral thallus. Further speculations about the life form of Spongiophyton devolve on the role of the pores. Despite Krausel’s suggestion of a doubtful— and ambiguous— group of dark bodies in one pore in Spongiophyton, we have seen no evidence in our material of spores being formed beneath the pores, as are regularly seen in Foerstia. None the less, it is possible that the pores represent such sites of release of reproductive bodies —either spores (presumably without any exine, as none are preserved) or motile propagules (cf. the osteole of the conceptacle of Fucus). Other possibilities are pores for secretion of mucilage (cf. various brown algae), or apertures leading to internal, aeration tissue (cf. stomata of vascular plants, or the barrel-shaped pores of Marchantia). There are doubtless many other possibilities; we merely suggest here that Spongiophyton (or, at least, the species S. nanum) was a dorsiventral, terrestrial plant, dichotomizing in the plane of the surface on which it grew, with the pores on the upper surface, from which occasional upright lobes developed. The systematic assignment of Spongiophyton will be further considered after a note on the composition of the cuticle, and a review of other comparable plant fossils. Composition of Spongiophyton cuticle. In an attempt to get further evidence on the nature of the Spongiophyton organism, an elemental analysis for carbon, hydrogen, nitrogen, and oxygen of fragments of S. nanum was carried out. For comparison, similar analyses of two other types of fossil plant material were made. These were of Foerstia ohioensis White, from the Upper Devonian Ohio Shale of Ohio, U.S.A. (material kindly supplied by Professor J. M. Schopf), and leaf material of Ptilophyllum pecten (a gymnosperm) from the middle Jurassic of Cloughton, Yorkshire. These were both prepared in the same way. The relevance of the Foerstia is that it represents a plant perhaps closer in its structure (and affinity?) to Spongiophyton than any other available fossil material. The Ptilophyllum, in con- trast, is perhaps adequately representative of coalified tissue of a land plant leaf with a rather thick cuticle. While the Spongiophyton and Foerstia specimens consisted of more or less coal-free ‘cuticle’ material which had to some extent been naturally 942 PALAEONTOLOGY, VOLUME 17 macerated during fossilization, the Ptilophyllum leaves certainly contained some coalified leaf mesophyll tissue. The specimens used for the analysis were extracted from the matrix with cold 40% hydrofluoric acid, washed repeatedly with distilled water, and then dried at 100 °C. The analyses were carried out on duplicate samples, made (in the case of Spongiophyton) by splitting individual 'tubes' of the plant into more or less equivalent halves (‘matched pairs’). It was hoped in this way to eliminate minor differences in original composition and effects of the microenvironment at the site of incorporation in the sediment. The result- ing samples, composed of several individual fragments, weighed about 1 mg. The analysis was carried out in the Chemistry Department of University College, London, through the kindness of Dr. A. J. Layton, using a routine analysis procedure. This basically involves heating a weighed sample of fossil material in a furnace at 900 °C, flushing initially with helium, followed by a stream of oxygen. Metallic copper removes excess oxygen from the resulting gas mixture, and the carbon, hydrogen, and nitrogen are assayed as C02, H20, and N, by means of a detector measuring the thermal conductivity of the gas. The results are given in table 1. In addition, comparable data are given for a bitu- minous coal and for two graptolites. These results are obviously only a very general guide to the original composition of the organic material at the time of fossilization. The coalification process, of course, alters the elemental ratios ; and the three samples have had rather different histories of incorporation and subsequent diagenesis. Even so, some guarded conclusions may be drawn. table 1 Elemental composition of Spongiophyton nanum and other coalified fossil plant material. For the first three plants, the figures given are the means of duplicate analyses. Carbon (C), hydrogen (H), and nitrogen (N) are determined directly, and the oxygen is obtained by difference. The figures for an average bituminous coal are from Swain (1970); the nitrogen: carbon ratios for the two graptolites are means of 3 and 2 determinations respectively, and are from Wiman 1902. C H N O H/C % N/C % Spongiophyton nanum 78-4 8-4 2-7 10 5 10-7 3-4 Foerstia ohioensis 63-5 5-5 1-7 29-3 8-6 2-7 Ptilophyllum pecten 67-2 6-2 1-6 250 9-2 2-4 Bituminous coal 79 5-4 1-6 14 6-9 2-0 Desmograptus 5 0 Gothograptus 6-7 Krausel has placed emphasis on the significance of the nitrogen fraction in such analyses as implying the presence of chitin in the original organic material. In his analysis of Foerstia ohioensis (Krausel 1941) he obtained 2-5% nitrogen, for which he invokes an original 35% of chitin in that plant. This he used as a basis for suggesting fungal affinity for Foerstia , and hence for the recognition of a new group of plants, the ‘Algomycetes’ based on that genus. The danger of ascribing organic nitrogen in fossil material directly and solely to chitin as its source is well illustrated by recent work on graptolites. Many early workers regarded the carbonaceous substance of graptolites as chitinous (e.g. Kraft 1926; see also Wiman 1902, whose analyses are quoted in table 1). But it has now been shown that a major nitrogenous constituent of fossil graptolite organic matter (perhaps all) is scleroprotein, which on hydrolysis yields a range of amino-acids (see Florkin’s 1969 discussion of the earlier work of Foucart). Krausel’s supposition that Foerstia contained chitin is perhaps no better founded than in the case of the graptolites. Our interpretation of the data in table 1 may be summarized as follows: (1) The N/C ratio in Spongiophyton is 25% greater than in Foerstia. This is at CHALONER ET AL.\ S P O N G I O P H YTO N 943 variance with Krausel’s (1960) results for Spongiophyton, which showed a very low (unstated) nitrogen content, which he interpreted as equivalent to a chitin content of less than 1%. (2) The N/C ratio in Spongiophyton is also appreciably higher than in the Ptilo- phyllum leaf material. (3) The nitrogen content of all the plant specimens (expressed as a nitrogen to carbon ratio) is considerably less (2:3-4) than that shown by the graptolite analyses of Wiman (5:6-7). Other Palaeozoic fossil plants comparable to Spongiophyton. We now know of a num- ber of Devonian plants which, while lacking evidence that they were tracheophytes, show one or more attributes of land plants (e.g. a thick cuticle, spores with an exine). Four of these— Prototaxites, Parka , Foerstia (? Protosalvinia) and Nematothallus are thoroughly discussed by Lang (1945). Foerstia (of the Upper Devonian of the United States) and its relationship to the Brazilian Protosalvinia , has been exten- sively reviewed by Krausel (1941), Arnold (1954), Schopf and Schwietering (1970), and Phillips et al. (1972). Krausel and Yenkatachala (1966) have also reviewed Spongiophyton , Aculeophyton , and Orestovia. Excluding the relatively massive tree-sized Prototaxites and the probably related thalloid Nematothallus, the remaining genera show one or more significant features of comparison with Spongiophyton. The main features of these five Devonian genera are listed in table 2. Of the genera there listed, three are based on plants known only as a tubular cuticular covering of a more or less cylindrical plant body : Spongiophyton ; the similar genus Aculeophyton (Lower Devonian of Western Siberia and possibly from Brazil); and the enigmatic Russian genus Orestovia. These three genera, of which the internal structure and reproductive organs are unknown, are grouped by Krausel and Venkatachala in a family, the Spongiophytaceae. But the status of the family remains obscure; Krausel and Venkatachala leave it unassigned, but make guarded comparison with the Rhodophyceae. The state of uncertainty of our know- ledge of these plants is reflected in the fact that one of them, Orestovia , is treated by Ananiev and Senkevitch (1963) as a psilopsid (a vascular plant). Views on Foerstia (? Protosalvinia) are equally diverse; Krausel (1941) regarded Foerstia as a member of a new class of Thallophytes, the ‘Algomycetes’, while Schopf and Schwietering believe that the Protosalvinia/ Foerstia complex ‘probably should be assigned to the Fucales" (Phaeophyta, brown algae). We remain doubtful of this assignment, since the thick cuticle-like covering and resistant spore tetrads have no close counterpart in any living member of that order. The essential features of the early Devonian plant Parka decipiens were demon- strated by Don and Hickling (1917) and it is a great tribute to the thoroughness of their work that no substantially new information on the plant has been forthcoming since that time. Parka may be compared to a limited extent with both Foerstia and Spongiophyton. It resembles both these genera in having been a somewhat flattened thalloid plant with a cuticular covering. Like Foerstia , Parka had large spore- containing cavities in the thallus, which gave the whole plant a rather net-like appearance. But although the spores of Parka have a resistant covering, as in Foerstia , they were not formed in tetrads as in that genus. There are several other poorly known £ TABLE 2 Genus SPONGIOPHYTON Krauscl ACULEOPHYTON Krauscl and Vcnkalachala Age Geographical External Cuticle location morphology Reproductive Systematic bodies position Important references Middle Devonian Parana, Brazil (Krauscl) Ghana (this work) Flattened, cylindrical Thick, with pores (on None seen dorsivcntral dicho- one face only?) inner tomizing thallus reticulum of elongated cells 'Thallophylc'— Spongiophytaceae (Krauscl and Vcnkalachala 1966) Krauscl I960 Krauscl and Vcnkalachala 1966 Lower Devonian Kuznclzk Basin (and ?Middlc ?Parana, Brazil Devonian) Unknown, perhaps Thin, with conical to None seen similar to spine-like papillae, Spongiopliyton and cellular reticulum 'Thallophylc'— Krauscl and Spongiophytaceae Venkatachala 1966 (Krauscl and Venkatachala 1966) Lower Devonian Kuznctzk Basin Yunnan, China Ergolskaya, amend. Krausel and Venkatachala ORESTOVIA Zalessky ex ?Cylindrical thallus with pscudo- monopodial branching Thin, with cellular reticulum, pores of two sizes ('large and small') 'Thallophylc' — Spongiophytaceae (Krauscl and Venkatachala 1966) or Psilopsida (Ananiev and Senkevitch 1963) Krauscl and Venkatachala 1966 Ananiev and Senkevitch 1963 Snigirevskaya 1971 foerstia White (? Proiosalvinia Dawson) Upper Devonian Mid-western U.S.A. Brazil Flattened, disc-like or upright club- shaped or dichoto- mizing pincer-shaped thallus Thick, with inner reticulum of isobilatcral cells Tetrads of large (200 m/i) triradiatc spores in depressions in thallus 'Algomycclcs', Foerstiales (Krausel) or Phaeophyta, Fucales (Schopf and Schwietcring 1970) Krauscl 1941 Schopf and Schwietering 1970 Phillips ei at. 1972 parka Lower Devonian England Fleming Upper Silurian Wales Scotland Flattened, rounded to Very thin, with irregularly lobed cellular reticulum dorsivcntral thallus Masses of small (30 ji) spores in cavities in thallus, not in Anomalous plants of ?algal affinity Don and Hickling 1917 Lang 1945 944 PALAEONTOLOGY, VOLUME 17 H ^ 1 VO vO — 1 2 S- 03 OS ^ X _ — 1 cd 3 -3 X >v : *2 g -2 cd 3 S “> £ * I 'c > > G k- :03 (U 2 a ^ < ■g o3 ® 00 5 C/3 C < 00 i -a -r = cd | a % .£■ O X X 1 _G o r— 1 O C/3 -G ■ 00 Oh 2 c 6 .2 i <-> I g CL •o -2 G cd cd 43 >% O L_ oo r/t rv g :* •* O I- c CL^ '/■v ^ 5 ^ g >V C G *g o 13 cd g* o . '_ vu o3 X! O 2P 3 2 X g :cd ^ cd O ^ G X CL S/ H w w> I g , cd ti X 43 cd £ a^g^.-2 § •J o’S g S' u ° I1 1 J2 ti = |§:Sc£ X CL^2 « V. (_ OO w > O T3 ^ cd v> 2 c t> ,—• £.2 « e ^ ? >^X 2? 43 O G CL OO '£ :§ 7? g > c/3 <; tL * o Ph co Cl tj X O 05 X) 00 cd O g CL o *s • - X O 5 -S 3 ^ r, tu S s o . ^ w u 3 Id .2 .2 8 §. S3 ;£ x j3 ^ a) O a .2 x ’rS ££ 3 ^ ^ ^ u- -S o Cd bo | ^ *3 ‘ C •" ° 73 3 " S 5 00 X X -2 S 3 -O c o o ,5 Td q, -G g a o o ■ 5 X G G ">v- ° 2 U ^ S x G c*_ O o - 43 - g ■H E O *■13 •a -a s 4^ ■*-> J3 X A O ^ “ -2 -3 ? a 3 .S3 s S u •a y S' S.-S -3 3 =1 .CS c3 o'SH5 cd cd OQ ^ *2 U X cd O 73 Pi o ^ X s a s £ G ^ ■>■ ^ G « cd Q ? x Z cd 2 "2 a >> S S3 3 ^ CL (U cd B 3 ^ £ « ^ -o a _r c x g cd o < >v > 7^ • w cd h; oS N W O G X :cd c > ^ 1 a < O o H £ & I^q § Fleming Upper Silurian Wales irregularly lobed cellular reticulum (30 ft) spores in of ?algal affinity 1917 Scotland dorsi ventral thallus cavities in Lang 1945 thallus. not in tetrads CHALONER ET AL.\ S PO N G 10 P H YTO N 945 fossil plants, to which comparison with Spongiophyton may be guardedly extended. They include the thalloid fossil Thallomia Heard, and Eohostimella Schopf. Thallomia is an enigmatic genus in which modern plant material (leaf tissue) may have been erroneously linked with an undoubted fossil organism (see Heard and Jones 1931; Lang 1937, p. 247; and Krausel and Venkatachala 1966, p. 222). Although its thalloid form might invite comparison with Spongiophyton, its putative internal structure and cuticle need reinvestigation. Eohostimella Schopf et al. , 1966 (from the Silurian of Maine, U.S.A.), consists of upright tubes of coalified tissue (?cuticle or cortex) of what were apparently plants, but we know nothing of their microscopic structure or internal organization. We believe that systematic assignment of any of the genera of table 2 must be tentative, but this does not diminish their biological interest. Their greatest significance is that although they are not apparently tracheophytes, and cannot be closely matched in any extant plant group, they share adaptations to terrestrial life (a cuticle, resistant spores) which are now seen only in tracheophytes and the sporophytes of bryophytes. In these attributes they may be thought of as showing parallel development to the early vascular plants. EMENDED DIAGNOSES OF TAXA Genus spongiophyton Krausel 1954 Type species. S. lenticulare (Barbosa) Krausel 1954, p. 206. (See also Krausel 1960.) Emended diagnosis. Tubular thallus with cuticular covering, branching dichotomously or subdichotomously with rounded apices; base unknown. Cuticle with internal cellular reticulum and circular to fusiform pores, largely confined to one surface of the thallus. Spongiophyton nanum Krausel 1960, p. 32 Emended diagnosis. Thallus cylindrical, of originally circular or elliptical cross- section, dichotomizing several times, with rounded apices. Branches typically 2-5 mm wide by up to 25 mm long (incomplete). Cuticle penetrated by numerous circular to elliptical pores 200-300 pm in largest diameter, with edges bevelled on the outer face. Pores principally confined to one face, this poral surface having the thicker cuticle (typically 60 ^m); cuticle of aporal face, typically 30 thick. Inner face of cuticle with ridges forming a cellular reticulum, cells typically 40 long by 20-30 pm broad, oriented with the longer axis parallel to the length of the thallus. Spongiophyton lenticulare (Barbosa) Krausel 1954 Emended diagnosis. Thallus cylindrical, its cuticular covering forming a tube up to 12 mm diameter. Lenticular pores of various sizes present, up to 0-8 mm in longest dimension, elongated parallel to long axis of thallus; pores of greater concentration on one surface of cuticular tube. Inner surface of cuticle with cellular reticulum, cells typically 50 ^m wide by 100 ^m long, with pronounced arrangement in longitudinal series. 946 PALAEONTOLOGY, VOLUME 17 Acknowledgements. We wish to express our thanks to the Royal Society whose generous grant to one of us (W. G. C.) made it possible to purchase the scanning electron microscope S600 used in this study. We also wish to thank Professor T. M. Harris, F.R.S., Dr. B. E. Plunkett, and Professor A. F. J. Smit for helpful discussion, and Dr. F. Schaarschmidt for granting access to the Brazilian material of Spongiophyton in the Forschungsinstitut Senckenberg, Frankfurt am Main. REFERENCES ananiev, a. r. and senkevitch, m. a. 1963. Psilopsida; Systematic part. In vachrameev, v. a., Radchenko, g. p. and takhtajan, a. i. (eds.). Osnovii Palaeontologii, 15, 325-343. Moscow. [In Russian.] Arnold, c. a. 1954. Fossil sporocarps of the genus Protosalvinia Dawson, with special reference to P.furcata (Dawson) comb. nov. Svensk Botanisk Tidskr. 48, 292-300. baker, s. m. 1912. On the brown seaweeds of the salt marsh. J. Linn. Soc. Bot., Loud. 40, 275-291. barbosa, o. 1949. Vegetais fosseis do Devoniano do Brasil e da Bolivia. Mineratydo e Metalurgia , 14, 81-84. crow, a. t. 1952. The rocks of the Sekondi Series of the Gold Coast. Gold Coast Geol. Surv. Bull. 18, 1-68. daemon, r. f., quadros, l. p. and da silva, l. c. 1967. Devonian Palynology and Biostratigraphy of the Parana Basin, 99-132. In bigarella, j. j. (ed.). Problems in Brazilian Devonian Geology (Bot. Paranaense Geosciencias, 21/22). 151 pp. dalingwater, j. 1973. The cuticle of a eurypterid. Lethaia , 6, 179-186. don, a. w. R. and hickling, g. 1917. On Parka decipiens. Q. Jl geol. Soc. Lond. 71, 648-666. florkin, m. 1969. Fossil shell ‘conchiolin’ and other preserved biopolymers, pp. 498-520. In eglington, g. and murphy, M. T. J. (eds.). Organic Geochemistry , 828 pp. Springer, Berlin/New York. fritsch, f. e. 1952. The structure and reproduction of the algae, II. Cambridge University Press. 939 pp. Harris, T. m. 1964. The Yorkshire Jurassic Flora. II. Caytoniales, Cycadales and Pteridosperms. British Museum (Nat. Hist.) London. 191 pp. heard, a. and jones, J. f. 1931 . A new plant (Thallomia) showing structure from the Downtonian rocks of Llandovery, Carmarthenshire. Q. Jl Geol. Soc. Lond. 87, 551-562. kendall, M. w. 1948. On six species of Pagiophyllum from the Jurassic of Yorkshire and Southern England. Ann. Mag. Nat. Hist. 12, 73-108. kraft, p. 1926. Ontogenetische Entwicklung und Biologie von Diplograptus und Monograptus. Palaont. Zeit. 7, 207-249. krausel, r. 1941. Die Sporokarpon Dawsons, eine neue Thallophytenklasse des Devons. Palaeontographica , B. 86, 113-133. 1954. Spongiophyton nov. gen. (Thallophyta) e Haplostigma Seward (Pteridophyta) no Devoniano Inferior do Parana. In Paleontologia do Parana: vol. Comemorativo do 1° Centenario do Estado do Parana, Publicado pela Comissao de Comemoracoes do Centenario do Parana, 195-210. Curitiba. — 1960. Spongiophyton nov. gen. (Thallophyta) e Haplostigma Seward (Pteridophyta) no Devoniano inferior do Parana. Monogr. Dep. nac. Prod. min.. Div. Geol. Miner. 15, 1-41. and venkatachala, b. s. 1966. Devonische Spongiophytaceen aus Ost- und West-Asien. Senckenberg. leth. 47, 215-251. lang, w. H. 1937. On plant remains from the Downtonian of England and Wales. Phil. Trans. R. Soc. B. 227, 245-291. — 1945. Pachytheca and some anomalous early plants. J. Linn. Soc., Bot. 53, 535-552. lange, f. w. and petri, s. 1967. The Devonian of the Parana Basin, pp. 5-55. In bigarella, j. j. (ed.). Problems in Brazilian Devonian Geology ( Bol . Paranaense Geosciencias, 21/22). 151 pp. mensah, m. k. 1973. On the question of the age of the Sekondi Series, Upper Devonian or Lower Carboni- ferous Rocks of Ghana. Ghana J. Sci. 14 (1). (In press.) — and chaloner, w. G. 1971. Lower Carboniferous lycopods from Ghana. Palaeontology , 14, 237-269. mortimer, m. g. and chaloner, w. G. 1972. The palynology of the concealed Devonian rocks of southern England. Bull. Geol. Surv. Gt Brit. 39, 1-56. Phillips, T. L., niklas, K. j. and ANDREWS, H. N. 1972. Morphology and vertical distribution of Proto- salvinia ( Foerstia ) from the New Albany Shale (Upper Devonian). Rev. Palaeobot. Palynol. 14, 171-196. rolfe, w. d. I. 1962. The cuticle of some ceratiocaridid Crustacea from Scotland. Palaeontology, 5, 30-51. CHALONER ET AL. S P O N G 10 P H YTO N 947 schopf, j. m., mencher, e., boucot, a. j. and Andrews, h. n. 1966. Erect plants in the early Silurian of Maine. Prof. Pap. U.S. Geol. Surv. D69-D75. - and schwietering, J. f. 1970. The Foerstia zone of the Ohio and Chattanooga Shales. U.S. Geol. Surv. Bull. 1294-H, 1-15. SNiGiREVSKAYA, n. s. 1971. Application of the scanning electron microscope in botany. Botanischeskii Zhurnal, 56 (4), 549-558. [In Russian.] swain, F. M. 1970. Non-marine organic geochemistry. Cambridge University Press. 445 pp. taylor, b. j. 1971. Thallophyte borings in phosphatic fossils from the Lower Cretaceous of south-east Alexander Island, Antarctica. Palaeontology , 14, 294-302. wiman, c. 1902. Uber die Borkholmer Schicht im Mittelbaltischen Silurgebiet. Bull. Geol. Inst. Univ. Uppsala , 5, 149-222. W. G. CHALONER Department of Botany Birkbeck College University of London Malet Street London, WC1E7HX M. K. MENSAH Geology Department University of Ghana Legon, Ghana M. D. CRANE City of Portsmouth Museum Alexandra Road Portsmouth, POl 2LJ Typescript received 16 November 1973 ECOLOGICAL SUCCESSION IN INTRAFORMATIONAL HARDGROUND FORMATION by R. goldring and j. kazmierczak Abstract. A review of discontinuity hardgrounds shows that an ecological succession can be recognized accompany- ing the gradual increase in lithification. The burrowing, boring, and encrusting biota is divided into five groups: soft to firm substrate burrowers, animals that penetrate firm or cemented substrates, borers restricted to cemented sub- strates, non-restricted encrusters on firm to cemented substrates, and encrusters restricted to cemented substrates. The type of ecological succession present depends on the lithology, and four types of hardground are recognized reflecting differences in lithification potential: calcarenite, calcirudite, calcilutite with very low clay content, and calcilutite with about 2% clay. The principal factor influencing marine organisms in colonizing the substrate is the degree of consolidation, which depends on grain size and shape distribution, and mineralogy, together with external factors such as temperature, turbulence, and salinity. Much is known about the ecology of rocky shorelines and also, though to a lesser extent, of submarine rocky surfaces, submarine canyon walls, and the exten- sive submarine lithified pavements that have been described from the Persian Gulf and elsewhere (Shinn 1969; Taft et at. 1968). The latter type have received particular attention from geologists because they appear to be the modern analogues of fossil intraformational hardgrounds (see Bathurst 1971 for summary). Such hardgrounds represent stratigraphical discontinuities in calcareous sediment where the substrate became lithified before a permanent cover was established. Their recognition is chiefly by the boring and encrusting fauna. In North America the term hardground has only recently been applied to such discontinuities (Halleck 1 973). Other geologists have used hardground synonymously with hard substrate, regardless of its strati- graphical and sedimentological context, e.g. Krantz (1972). This is quite valid bio- logically and it may be useful to distinguish intraformational hardgrounds from other types. Intraformational hardgrounds must have formed from unlithified sediment. Since benthonic organisms have a limited range of tolerance to degree of consolidation an ecological succession of organisms able to cope with the different stages of surface consolidation is to be expected. That such successions in fossil hardgrounds are found confirms that these discontinuity horizons did pass through various stages of con- solidation. By comparison with the tolerance ranges of modern taxa, we can infer the degree of consolidation attained. Kazmierczak and Pszczolkowski (1968, 1969) and, independently, Bromley (1968) recognized that a succession of biocoenoses had occurred when hardgrounds could be shown to have passed through earlier softer stages. Fiirsich (1971) and Palmer and Fiirsich (1974) have described successions from Middle Jurassic hardgrounds. The many interacting factors which influence substrate colonization, in addition [Palaeontology, Vol. 17, Part 4, 1974, pp. 949-962, pis. 125-126.] 950 PALAEONTOLOGY, VOLUME 17 to the degree of consolidation, means that community successions are likely to be complex. Any discontinuity surface which is being colonized and is undergoing change in degree of consolidation must also be undergoing a sere. Further, inter- ruption of the ecological succession is possible, at any stage of consolidation, by deposition of a permanent sedimentary cover. Recognition of an interrupted sere is difficult. Frequently, lithification appears to have taken place over a period of dis- continuous sedimentation and erosion. Kazmierczak and Pszczolkowski (1968, 1969) recognized that discontinuity surfaces in the Polish Trias and Jurassic must have reached different stages of consolidation before being permanently smothered. SEDIMENTOLOGICAL ASPECTS Little work has yet been done on several important sedimentological aspects of hard- grounds, but it is not the purpose of this paper to investigate the processes of sub- marine consolidation. Whilst corrosion and bioerosion were undoubtedly active on the substrate, the smoothness of many truncated surfaces, especially those on calcilutites, suggests that degradation by corrasion was often dominant. Likely corrasion agents were the coarse sediment seen infilling burrows and borings (PI. 125, fig. 1) and evidently derived from temporary mobile covers. Occasionally such covers were themselves lithified (PI. 126, fig. 3; text-fig. 1) as in the re-bored borings of Rose (1970), suggesting that consolidation and lithification proceeded very quickly. In the example figured (PI. 125, fig. 1) the coarse crypt fill lithified more quickly than the micritic host sediment. Similar examples have been observed in Canadian Ordovician hardgrounds (M. E. Brookfield pers. comm.). Hardgrounds in pelagic facies (Fabricus 1968; Wendt 1970; Jenkyns 1971 ; Tucker 1973) are often more irregular, probably reflecting the greater role played by corrosion in their formation. BIOLOGICAL ASPECTS There is relatively little information on the way infaunal and sedentary taxa are affected by different substrates. Ekman (1947), Trueman et al. (1966), and Evans (1968) have discussed bivalve penetration into different types of soft and hard EXPLANATION OF PLATE 125 Fig. 1. Successive generations of borings in micritic, slightly clayey limestone in Lower Kimmeridgian, top of unit 13 in Kazmierczak and Pszczolkowski (1968), Bolmin village, south-western Holy Cross Mountains, Poland, x3. Inset, 1, 2, 3 first generation of borings, deformed by compaction and in 2, possibly with injection of soft sediment; 4, boring of second generation less deformed; 5, boring of third generation also slightly deformed at point indicated. When boring 5 was formed the infilling of 4 was probably fully cemented (shells truncated) although the surrounding sediment was only firm. Fig. 2. Successive generations of burrows and borings into low clay calcilutite. Pskov Formation, Upper Devonian, River Velikaya at Vybuty village (Porogi Vybutskiye), Pskov district, U.S.S.R., x 1. Figs. 3, 4. Impressions of atrypid valves in low clay calcilutite. Same horizon and locality as fig. 2. 3, bedding surface with impressions of two valves, x 0-85. 4, section normal to bedding cutting atrypid valve and showing early generations of burrows and later Trypanites borings, x 3. Fig. 5. Trypanites borings cutting oolitic calcirudite of Snetogorsk Formation, Upper Devonian, at River Velikaya section near Snetogorsk Monastery, Pskov district, U.S.S.R., x2. Fig. 6. Irboskites encrusting shelly low clay calcarenite. Pskov Formation, Upper Devonian, at Pskov quarry (east side of River Velikaya), Pskov district, U.S.S.R., x 1. PLATE 125 GOLDRING and KAZMIERCZAK, hardground ecology 952 PALAEONTOLOGY, VOLUME 17 text -fig. 1 . Different types of hard- ground in the Upper Devonian, Pskov Formation, River Velikaya section at Vybuty village, Pskov district, U.S.S.R. At A a hard- ground state was not attained. Hardground was attained at B and C. I, fine calcarenite, slightly mottled. 2, marly, high clay calcilutite with burrows of Balanoglossites- type somewhat deformed by compaction. Top of unit truncated and burrows infil- trated by limonite. 3, slightly marly calcilutite (type 3b) with burrows (undeformed) of Balanoglossites- type. Top of unit truncated and densely bored by Trypanites (PI. 126, figs. 3, 4). 4, brachiopod- crinoidal calcirudite (type 2) with truncated top bored by Trypanites. 5, very fine calcarenite with inclu- sions of coarser material. 6, brachio- pod coquina, calcirudite, without hardground. 7, bioturbated marls. 0 cm sediment and Wilson (1952) has reviewed aspects of larval settlement. Rhoads (1970) distinguished between bioturbation structures made in thixotropic sediment from those made in plastic sediment and recognized both types in fossil sediments. In thixotropic sediment the structures have an indistinct outline whereas in plastic sediment the outline is sharp and well-defined. Changes in burrow morphology, reflecting the change from thixotropic to plastic state can be seen in calcilutites from the Devonian of the Pskov area, U.S.S.R. On polished surfaces and peels the succes- sive burrow generations have increasingly sharp outlines and increasingly circular cross-sections (PI. 125, fig. 2; PI. 126, fig. 1). EXPLANATION OF PLATE 126 Fig. 1. Successive generation of burrows and borings into low clay calcilutite. Same locality and horizon as Plate 125, fig. 2, x 1-8. Fig. 2. Burrows, probably of crustacean origin and later Trypanites borings (white spots) in pure calcilutite. Lower Kimmeridgian (top of unit 11 in Kazmierczak and Pszczolkowski 1968, text-fig. 2), locality as for Plate 125, fig. 1, x 1-5. Figs. 3, 4. Two hardgrounds, shown diagrammatically in text-fig. 1 ; levels B, C truncating lithologies 3 and 4 respectively. Locality and horizon given in explanation to text-fig. 3, x 5. 4, part of hardground encrusted by nebecularid-like foraminifera, x 20. Fig. 5. Exogyra sp. encrusting surface with abraded boring of Gastrochena sp. Banded oolitic calcilutite with chert. Lower Kimmeridgian (top of unit 8 in Kazmierczak and Pszczolkowski 1968, text-fig. 2), Skorkow village, south-western Floly Cross Mountains, Poland, x 1. Fig. 6. Trypanites borings into banded, fine calcarenite (oolite), from upper surface and from within earlier formed burrows (lower right). Same locality and horizon as fig. 5, x 0-8. (= Kazmierczak and Pszczol- kowski 1968, pi. 4, fig. 3.) PLATE 126 GOLDRING and KAZMIERCZAK, hardground ecology 954 PALAEONTOLOGY, VOLUME 17 Schafer (1962, 1972), Trueman (1968), and others have described how animals move into and through non-lithified sediment. Undulatory movement (Schlangel- bewegung) and ‘swimming’ is probably confined to thixotropic and liquid sediment whilst several other modes of burrowing occur in thixotropic and plastic sediment, e.g. peristaltic movement and the movement of burrowing scaphopods, gastropods, and bivalves. (Animals employing certain types of movement, of course, change the physical state of the sediment during penetration.) The polychaete Polydora ciliata is capable of penetrating a wide range of sediments from mud to limestone, boring into the latter chemically. The trace fossil Trypanites (generally restricted to a straight tunnel, but more widely interpreted by Bromley 1972) seems to have been made in a similarly wide range of sediment. Many bivalves that penetrate the substrate mechanically, using the armed shell, are relatively tolerant : burrowing into firm substrates and thereby displacing particles, and boring into cemented substrates by cutting the fabric. Boring and burrowing have been used more or less synonymously by authors but it is useful to define them more narrowly (following Shinn 1969; Bromley 1970, 1974 in press; Perkins 1971). Organisms that bore or drill have been reviewed by Yonge (1963) and Bromley (1970). Boring may be difficult to prove in fine-grained sediment but deformed crypts indicate a somewhat plastic substrate. Borings may enter from open galleries, older burrows (PI. 126, fig. 6) and crevice roofs as well as from the upper sedimentary surface. In Plate 125, fig. 1 (at 5) the substrate was partly burrowed and partly bored. Similar situations have been observed where a bivalve, having bored through a cemented crust, continues penetration into firm sediment. The bored margin to the crypt is sharp whilst the burrowed margin is ragged. A more restricted group of boring organisms only penetrates sediment where the grains are cemented together. This includes boring sponges, algae, fungi, bivalves which bore by chemical means, boring bryozoans, and the barnacle Lithotrya. Warme (1970) has pointed out that rocks with only 5% carbonate may be chemically bored. Some encrusting organisms can also cope with a considerable range of substrate consolidation and cementation. In part this depends not only on the degree of con- solidation and cementation but also on the distribution and size of clasts sufficiently attractive for larval settlement. The roughness and erodability of a substrate and the presence of an organic film are also important factors. Today, algae, encrusting serpu- lids and encrusting bryozoa attach to weed, shells, or to smooth and cemented surfaces. From geological observations (fig. 3) the tabulate coral Aulopora and encrusting foraminifera such as Tolypammina and Bdelloidina also required a cemented surface. In contrast, animals less specialized in attachment, such as oysters, may only require a small area of hard substrate such as a shell fragment to attach to on an otherwise firm but uncemented substrate. The Devonian brachiopod Irboskites (PI. 125, fig. 6) apparently attached in a manner similar to oysters (PI. 126, fig. 5), though with most of its ventral valve attached. Crinoids, and other pelmatozoans known from some hardgrounds were probably able to attach themselves to firm as well as to hard surfaces. In attaching to hard surfaces ‘roots’ or discs of attachment were usually covered by stereomal secretion (Ehrenberg 1929). Organism encrustations and bivalve crypts are not, in themselves, conclusive evidence for full lithification. GOLDRING AND KAZMIERCZAK: HARDGROUND ECOLOGY 955 Classifications of organisms in relation to hardgrounds From the above, five groups of living and fossil organisms may be distinguished: Group 1. Burrowers in loose to firm substrates. (a) Animals moving by swimming and undulatory movement into very soft or thixotropic sediment and making impermanent burrows include many polychaetes, oligochaetes, echiurids, enteropneusts, certain holothurians, and certain arthropods together with nuculid bivalves. Fossil forms include the branching burrow Balano- glossites (Ord. -Trias). (b) Firm substrate burrowers include many burrowing bivalves (e.g. My a, Ensis ), echinoderms, arthropods, burrowing coelenterates, and burrowing polychaetes producing permanent burrows. Fossil forms include the trace fossils Thalassinoides, unlined Ophiomorpha, Arenicolites , Dip/ocraterion, and Coropliioides. In incohesive sediment Ophiomorpha is lined by pellets. Group 2. Animals that penetrate firm or cemented substrates (burrowers or borers) include those bivalves using mechanical means (e.g. Pho/as), and the polychaete Polydora. The trace fossil Trypanites seems to have been similarly unrestricted. Group 3. Borers restricted to hard substrates. Bivalves which bore by chemical means (e.g. Hiatella ) together with boring sponges, algae, fungi, and boring phoronids. Fossil forms include the boring Entobia attributed to clionid sponges. Group 4. Encrusters on firm or cemented substrates (non-restricted encrusters) include some oysters, byssally attached bivalves, and crinoids. Fossil forms include the productid brachiopod Irboskites, Exogyra , Apiocrinus, some edrioasteroids, and other primitive echinoderms. Group 5. Encrusters restricted to hard substrates. Serpulid polychaetes (e.g. Spirorbis), encrusting calcareous algae, bryozoans, cirripedes, foraminifera, thecideidinid and craniid brachiopods. Fossil forms include encrusting tabulate corals, encrusting foraminifera such as Tolypammina , Bdelloidina, encrusting serpulids, and bryozoans. STRATONOMICAL CRITERIA There are several stratonomical criteria for determining the degree of consolidation achieved. 1. Burrow-in-burrow structure (PI. 125, fig. 2; PI. 126, fig. 1), where successive generations of burrow show progressively sharper margins, less distortion, and increasingly circular cross-sections, indicating that the sediment was undergoing an increase in degree of consolidation. 2. Deformed crypts (PI. 125, fig. 1) indicate that the organism penetrated firm but not lithified sediment and deformation occurred with subsequent compaction of the sediment. 3. Borings truncating evenly across shells, ooids, oncoids, and older crypt fills (PI. 126, fig. 3, also Purser 1969, figs. 4, 12) indicate that the matrix was as hard as the shells and other clasts. 4. Discontinuity surfaces evenly truncating clasts, shells, and matrix, likewise indicate full lithification of the surface (PI. 125, fig. 5). 956 PALAEONTOLOGY, VOLUME 17 5. The hardness of the substrate at the time of penetration may be estimated from the form of the shell and borings of pholads (Evans 1968, 1970). 6. The absence from a sedimentary unit of burrows penetrating down from the overlying unit may indicate that an increase in consolidation had occurred before the overlying unit was deposited, if it can be shown that non-penetration was unlikely to have been because of other factors (e.g. depth of penetration required); text-fig. 1, horizon A. 7. Shells and other objects of known hardness introduced above the discontinuity surface in the smothering layer and pressed into the surface (PL 125, figs. 3, 4) show that the discontinuity surface was sufficiently plastic to take an impression. 8. Toolmarks scratched on the discontinuity surface provide information, especially if the tool can be identified. A. Pszczolkowski (pers. comm, to J. K.) has observed, on the top of a Lower Kimmeridgian discontinuity surface penetrated by bivalves, prod or impact marks made probably by the small calcareous algae Marinella. TYPES OF INTRAFORMATIONAL HARDGROUND Direct measurement of the physical state of a fossil intraformational hardground at the time of penetration or encrustation is not possible. Resort has to be made to biological and sedimentological criteria. However, intraformational hardgrounds are known only from calcareous sediments whose diagenetic history is to a large degree dependent on grain size and sedimentation rate (Shinn 1969) and the proportion of clay minerals. Some indication of the rate of lithification may be shown by the frequency and stratigraphical spacing of hardgrounds. Sugden and McKerrow (1962) recognized that a carbonate to clay ratio of four to one was critical in separating lime- stone and marly limestone from marl. Subsequently, the work of Bausch (1968) and Zankl (1969) has indicated that where clay exceeds 2% in calcilutites early recrystalliza- tion is inhibited and compaction will occur on subsequent loading. The proportion of clay must have influenced the maximum degree of consolidation attained and consequently the ecological succession. In the literature authors have, unfortunately, only infrequently identified the lithology associated with hardgrounds and whether or not earlier bioturbation preceded the actual hardground colonization. This limits our ability to interpret previously described hardgrounds. Four types of ecological succession can be recognized (text-fig. 2) on palaeontological evidence and, whilst we have not, at this stage, made extensive determinations of the clay content of various hardgrounds we have attempted a correlation with lithological type. Type 1 . The most common type of fossil hardground is formed of calcarenite (oosparite, biosparite) with low clay content (e.g. Voigt 1959, 1970; Purser 1969; Halleck 1973; Palmer and Fiirsich 1974), similar to that of modern hardgrounds in the Persian Gulf (Shinn 1969). Prior to lithification the sandy loose sediment was burrowed and on lithification the discontinuity crust was bored and encrusted. Type 2 (PI. 125, fig. 5). Less commonly hardgrounds are formed in calcirudites. Burrowing in a coarse substrate leaves little evidence, but where truncation, con- solidation, and cementation have led to a lithified surface, this surface may become encrusted and bored. text -fig. 2. Schematic diagram to illustrate relationships between hardground formation in different lithologies (sedimentation rate, degree of consolidation, and cementation) and ecological succession. 958 PALAEONTOLOGY, VOLUME 17 Type 3. It is clear that hardgrounds of types 1 and 2 cemented relatively quickly compared with calcilutitic hardgrounds. Where the proportion of clay minerals was sufficient to prevent early diagenetic recrystallization an actual hardground state could not be attained and the discontinuity surface must have remained no firmer than plastic, akin to discontinuity and bedding surfaces in clastic sediments. Where, how- ever, clay was largely absent calcilutites were able to attain the hardground state. Type 3a (PI. 125, figs. 1, 2; PI. 126, figs. 1, 2). Discontinuity hardgrounds in low clay calcilutites show a similar ecological succession to those in calcarenites. The principal difference is that more extensive burrowing took place during the thixo- tropic and plastic stages. Several examples of such hardgrounds have been figured in the literature. 1. Upper Devonian, U.S.S.R. (Hecker 1960, pi. 4, fig. 11; also figured in Hecker 1965, pi. 7, fig. 2), with Aulopora, Irboskites, and Trypanites. Although Hecker does not describe the lithology, from our observa- tions of the Pskov Formation with Professor Hecker we consider it to be a calcilutite of this type. 2. Upper Devonian, U.S.A. (Koch and Strimple 1968) with Aulopora , Spirorbis, edrioasteroids and cystids, Trypanites. Although Koch and Strimple considered that lithification had probably taken place subaerially, submarine lithification is more likely. A specimen collected and donated by Dr. C. R. C. Paul (University of Liverpool) has an insoluble residue of about 3%. The mottling in Koch and Strimple’s fig. 2 suggests that an ecological succession may be determinable. 3. Upper Jurassic, Poland (Kazmierczak and Pszczolkowski 1968, fig. 2, tops of horizons 11 and 13 and pi. 3, fig. 4) with serpulids on walls of Rhizocorallium burrows near to their apertures. Type 3b (PI. 125, figs. 3, 4; text-fig. 1, level B). Where the percentage of clay in calcilutites was intermediate consolidation was somewhat hindered so that the climax fauna included only members of groups 2 and 4 (e.g. Trypanites, Irboskites, and crinoids). In the example figured (insoluble residue 5%) the critical evidence for the final hardness of the discontinuity surface prior to a permanent cover being estab- lished is that Atrypa valves, introduced with the covering sediment, were impressed into the discontinuity surface. Earlier stages of burrowing can be recognized as in type 3a. Deformed crypts may also demonstrate that the substrate was still incom- pletely lithified. Some examples are listed below. 1. Triassic, Poland (Kazmierczak and Pszczolkowski 1969, pi. 4, fig. 1) with Trypanites. 2. Upper Jurassic, Poland (Kazmierczak and Pszczolkowski 1968, pi. 4, fig. 3). 3. Upper Cretaceous, Europe (Voigt, 1959, p. 134, types 2 and 3). Certain of the hardgrounds described by Bromley (1968) probably fall into this group, e.g. those from the Chalk Rock showing vermiform and sponge borings and occasional bivalve borings but not polyzoan or polychaete encrustations. Type 3c. In clayey calcilutites (text-fig. 1, level A; text-fig. 2) only two stages of substrate colonization can be recognized and the fauna shows no evidence that hardening and lithification proceeded beyond that required for firm substrate burrow- ing; no hardground state was attained. The form of the early mottling suggests that the sediment was thixotropic at that stage and the sediment was virtually completely destratified. Increasing firmness of the substrate (probably reflecting the change to a firm stage of consolidation) can be seen by the later, more distinct burrow system. Thalassinoides is a common burrow system below many hardgrounds and Kazmierczak and Pszczolkowski (1969) also recognized the importance of enteropneusts. On no occasion, in their examples, have encrusters or structures attributable to boring organisms been observed, indicating that insufficient time for the state of lithification GOLDRING AND KAZMIERCZAK: HARDGROUND ECOLOGY 959 required to be attained was not a reason for their absence. It may be significant that bioturbation from overlying units did not penetrate below the discontinuity surface (above— stratonomical criterion 6). Similar surfaces are common in shallow-water clastic facies (e.g. Goldring 1964; Farrow 1966). 1. Middle Triassic, Poland (Kazmierczak and Pszczolkowski 1969), where most of the discontinuity surfaces are of this type. 2. Upper Jurassic, Poland (Kazmierczak and Pszczolkowski 1968, text -fig 3). HIATUS CONCRETIONS, CONCRETION HORIZONS, AND CREVICES Concretions and concretion horizons exhumed on the sea-floor differ from the hard- grounds discussed in that the surface was generally already lithified before being exposed to organic colonization. Even in those described by Voigt (1968), Hallam (1969), and Kennedy and Klinger (1972) the repeated borings record subsequent burial and re-exhumation rather than diagenetic change whilst the concretions were actually forming the substrate. Hardgrounds lithified subaerially at an earlier stage should also be distinguished from intraformational hardgrounds formed below low-water mark since they too will not show biotal succession relative to gradual lithification. This criterion may be added to those given by Rose (1970) for distinguishing between submarine and subaerial discontinuity surfaces. The formation of crevices on the sea-floor emphasizes how thin the hardground crust may be. Purser (1969) and Palmer and Fiirsich (1974) have described crevices below Jurassic hardgrounds and others have been described from pelagic facies (Tucker 1973). The latter were never colonized although they may contain a hydraulic- ally introduced fauna. Palmer and Fiirsich describe the ecological distinction between the biota colonizing the upper surface and the crevice roof, analogous to the dis- tribution in modern crevices. EVOLUTION OF THE HARDGROUND BIOTA The number of hardgrounds described in the literature is still small and narrowly distributed through the stratigraphic column. An attempt has been made to show the geological range of organisms having colonized hardgrounds (text-fig. 3). Informa- tion on encrusting organisms is not easy to assemble and although borers have been reviewed by Boekschoten (1965) and Bromley (1970) it is not always clear whether the borings are into the actual hardground surface or into shells which may or may not be associated with a hardground. The trace fossil assemblage of hardgrounds, like that of any trace fossil facies, essentially reflects certain behavioural patterns and shows little change with time, although the organisms responsible for particular traces certainly changed. Hecker (1935) has suggested that the productid brachiopod Irboskites is paralleled in younger environments by encrusting barnacles. A major feature of Mesozoic and younger shallow-water hardgrounds is encrustation by oysters which also act as a site for other encrusters and borers. Pelagic hardgrounds encrusted by 960 PALAEONTOLOGY, VOLUME 17 text-fig. 3. Distribution of disconformity hard- ground biota. (Borers and encrusters presently known to be associated only with mineralized skeletons and shells are excluded.) the foraminifera Tolypammina and BdeUoidina show no change from the Devonian to the Mesozoic. There is a notable absence of any described intraformational hardgrounds from the Cambrian, and only sparse records from the Carboniferous and Permian. As Hecker (1970) has mentioned, this is an enigma. The most likely explanation, out- side the Carboniferous and Permian of the Russian platform where hardgrounds have been searched for, is that they have not yet been fully recognized in these systems. Hardgrounds have not yet been recognized in the Precambrian. This is almost certainly because of the difficulty of recognizing such surfaces in the absence of biological evidence. The Cretaceous-Tertiary unconformity. It is pertinent to note that the unconformity surface between the Cretaceous (Chalk) and Tertiary in south-east England and Germany was sufficiently soft for arthropods to leave distinctive scratch marks on the burrow walls (see Kennedy 1967). For the Chalk to have remained as soft as this at the unconformity, the surface must have been cut below low-water mark, since it does not seem possible for Chalk to have remained sufficiently soft intertidally. GOLDRING AND KAZMIERCZAK: HARDGROUND ECOLOGY 961 Acknowledgements. The evidence on which this contribution is based comes to a large extent from the Upper Devonian of the Main Devonian Field in the Pskov area of Russia. The palaeoecology of this area has been documented by Hecker (1935, 1960, 1970) and we were privileged in having been shown several sections by Professor R. F. Hecker. One of us (J. K.) has investigated Mesozoic hardgrounds in Poland (Kazmierczak and Pszczolkowski 1968, 1969) and Dr. A. Kendall (Department of Mineral Resources, Regina, Canada) demonstrated Middle Jurassic hardgrounds in England to R. G. We have also benefited from discussion with Messrs. F. Fiirsich and T. Palmer (Oxford), Dr. R. G. Bromley (Copenhagen), and Dr. G. Warner (Reading). REFERENCES bathurst, R. 1971. Carbonate sediments and their diagenesis. 620 pp. Elsevier, Amsterdam. bausch, w. m. 1968. Clay content and calcite crystal size of limestones. Sedimentology, 10, 71-75. boekschoten, G. j. 1966. Shell borings of sessile epibiontic organisms as palaeoecological guides. Palaeogeogr. Palaeoclimatol. Palaeoecol. 2, 333-379. bromley, R. G. 1968. Burrows and borings in hardgrounds. Medder. Dansk Geol. Foren. 18, 247-250. 1970. Borings as trace fossils and Entobia cretacea Portlock, as an example. In crimes, t. p. and harper, j. c. (eds.). Trace Fossils. Geol. Jour. Spec. Issue, 3, 49-90. — 1972. On some ichnotaxa in hard substrates, with a redefinition of Trypanites , Magdefrau. Palaeont. Zeit. 46, 93-98. (in press). Trace fossils at omission surfaces. In FREY, r. w. (ed.). The study of trace fossils. ehrenberg, K. 1929. Pelmatozoan root-forms (Fixation). Bull. Amer. Mus. Nat. Hist. 59, I 76. ekman, s. 1947. Uber die Festigkeit der marinen Sedimente als Factor der Tier Verbreitung. Zool. Bidr. Uppsala, 25, 1-20. evans, j. w. 1968. The effect of rock hardness and other factors on the shape of the burrow of the rock- boring clam Penitella penita. Palaeogeogr. Palaeoclimatol. Palaeoecol. 4, 271-278. — 1970. Palaeontological implications of a biological study of rock-boring clams (Family Pholadidae). In crimes, T. p. and harper, j. c. (eds.). Trace Fossils. Geol. Jour. Spec. Issue, 3, 127-140. fabricius, f. h. 1968. Calcareous sea bottoms of the Rhaetian and Lower Jurassic Sea from the west part of the northern Calcareous Alps. In muller, g. and friedman, g. m. (eds.). Carbonate Sedimentology in Central Europe, 240-249, Springer, Berlin. farrow, g. 1966. Bathymetric zonation of Jurassic trace fossils from the coast of Yorkshire, England. Palaeogeogr. Palaeoclimatol. Palaeoecol. 2, 103-151. fursich, F. 1971. Hartgrunde und Kondensation im Dogger von Calvados. Neues. Jb. Geol. Palaont. Abh. 138, 313-342. goldring, r. 1964. Trace-fossils and the sedimentary surface in shallow-water marine sediments. In van straaten, l. m. j. u. (ed.). Developments in Sedimentology, 1. Deltaic and shallow marine deposits , 136-143, Elsevier, Amsterdam. hallam, a. 1969. A pyritised limestone hardground in the Lower Jurassic of Dorset (England). Sedi- mentology, 12, 231-240. halleck, m. s. 1973. Crinoids, hardgrounds and community succession: the Silurian Laurel- Waldron contact in southern Indiana. Lethaia, 6, 239-252. hecker, r. f. 1935. Phenomena of overgrowth and attachment in Upper Devonian fauna and flora of Main Devonian Field. Trudy Paleo. Inst. Acad. Sci. U.S.S.R. 4, 149-280. (In Russian with German summary.) — 1960. Fossil fauna of smooth rocky marine bottom. Trudy Geol. Inst. Acad. Sci. Estonian S.S.R. 5, 199-227. (In Russian with German summary.) 1965. Introduction to Paleoecology. 166 pp. American Elsevier, New York. 1970. Palaeoichnological research in the Palaeontological Institute of the Academy of Sciences of the U.S.S.R. In crimes, t. p. and harper, j. c. (eds.). Trace Fossils. Geol. Jour. Spec. Issue, 3, 215-226. jenkyns, h. c. 1971. The genesis of condensed sequences in the Tethyan Jurassic. Lethaia, 4, 327-352. kazmierczak, j. and pszczolkowski, a. 1968. Sedimentary discontinuities in the Lower Kimmeridgian of the Holy Cross Mts. Acta Geol. Pol. 18, 587-612. (In Polish with English summary.) 1969. Burrows of Enteropneusta in Muschelkalk (Middle Triassic) of the Holy Cross Mountains, Poland. Acta Palaeont. Pol. 14, 299-324. Q 962 PALAEONTOLOGY, VOLUME 17 Kennedy, w. j. 1967. Burrows and surface traces from the Lower Chalk of southern England. Bull. Brit. Mus. Nat. Hist. ( Geol. ), 15, 125-167. and klinger, h. c. 1972. Hiatus concretions and hardground horizons in the Cretaceous of Zululand (South Africa). Palaeontology, 15, 539-549. koch, d. l. and strimple, h. l. 1968. A new Upper Devonian cystoid attached to a discontinuity surface. Iowa Geol. Surv. Rept. Invest. 5, 1-49. krantz, R. 1972. Die Sponge-Gravels von Faringdon (England). Neues Jb. Geol. Palaeont. Abh. 140, 207-231. palmer, T. J. and fursich, F. T. 1974. The ecology of a Middle Jurassic hardground and crevice fauna. Palaeontology, 17, 507-524. perkins, b. f. 197 1 . Traces of rock-boring organisms in the Comanche Cretaceous of Texas. In perkins, b. f. Trace Fossils, Louisiana State University Miscellaneous Publication, 71-1, 137-148. purser, b. h. 1969. Syn-sedimentary marine lithification of Middle Jurassic limestones in the Paris Basin. Sedimentology , 12, 205-230. rhoads, d. c. 1970. Mass properties, stability and ecology of marine muds related to burrowing activity. In crimes, t. p. and harper, j. c. (eds.). Trace Fossils, Geol. Jour. Spec. Issue, 3, 391-406. rose, p. r. 1970. Stratigraphic interpretation of submarine versus subaerial discontinuity surfaces: an example from the Cretaceous of Texas. Bull. Geol. Soc. Amer. 81, 2787-2798. schafer, w. 1962. 1972. Aktuo-Paldontologie nach Studien in der Nordsee. 666 pp. Kramer Verlag, Frankfurt am Main. (English translation 1972: Ecology and Palaeoecology of marine environments. 568 pp. Oliver and Boyd, Edinburgh.) shinn, E. a. 1969. Submarine lithification of Holocene carbonate sediments in the Persian Gulf. Sedi- mentology, 12, 109-144. sugden, w. and mckerrow, w. s. 1962. The composition of marls and limestones in the Great Oolite Series of Oxfordshire. Geol. Mag. 99, 363-368. TAFT, W. H., ARRINGTON, F., HA1MORITZ, A., MACDONALD, C. and WOOLHEATER, C. 1968. Lithification of modern carbonate sediments at Yellow Bank, Bahamas. Bull. Marine Sci. Gulf Carribean. 18, 762-828. trueman, e. r. 1968. The burrowing activities of bivalves. Symp. zoo. Soc. Loud. 22, 167-186. brand, a. r. and davis, p. 1966. The effect of substrate and shell shape on the burrowing of some common bivalves. Proc. malac. Soc. London, 37, 97-109. tucker, m. e. 1973. Sedimentology and diagenesis of Devonian pelagic limestones (Cephalopodenkalk) and associated sediments of the Rhenohercynian geosyncline, West Germany. Neues Jb. Geol. Paldont. Abh. 142, 320-350. voigt, e. 1959. Die okologische Bedeutung der Hartgriinde (Hardgrounds) in der oberen Kreide. Palaeont. Zeit. 33, 129-147. - 1968. Uber Hiatus-Konkretionen (dargestellt am Beispielen aus dem Lias). Geol. Rundschau, 58, 281-296. 1970. Foraminiferen und (?) Phoronoidea als Kommensalen auf den Hartgriinden der Maastrichter Tuffkreide. Palaeont. Zeit. 44, 86-92. warme, j. e. 1970. Traces and significance of marine rock borers. In crimes, t. p. and harper, j. c. (eds.). Trace Fossils. Geol. Jour. Spec. Issue, 3, 515-526. wendt, i. 1970. Stratigraphische Kondensation in triadischen und jurassischen Cephalopodenkalkender Tethys. Neues Jb. Geol. Palaeont. Mh. 1970, 433-448. wilson, d. p. 1952. The influence of the nature of the substratum on the metamorphosis of the larvae of marine animals, especially the larvae of Ophelia bicornis , Savigny. Ann. Inst, oceanog. Monaco, 27, 49-156. yonge, c. m. 1963. Rock-boring organisms. In sognnaes, r. f. (ed.). Mechanisms of Hard Tissue Destruction, Pubis. Amer. Assn. Adv. Sci. 75, 1-24. zankl, h. 1969. Structural and textural evidence of early lithification in fine-grained carbonate rocks. Sedimentology, 12, 241-256. r. goldring j. kazmierczak Department of Geology University of Reading Reading, RG6 2AB Polska Akademia Nauk Zaklad Paleozoologii Al. Zwirki i Wigury 93 02-089 Warsaw, Poland Revised typescript received 3 April 1974 THE PALAEONTOLOGICAL ASSOCIATION Annual Report of the Council for 1973 Membership. On31 December 1973 there were 1366 members (808 Ordinary, 155Student, and 403 Institu- tional), a net increase of 45 members during the year. Subscriptions. There were 216 subscribers to Special Papers through the Association. Through Blackwell’s agency, there were 324 subscriptions to Palaeontology and 120 to Special Papers. Finance. During 1973 the Association published Volume 16 of Palaeontology at a cost of £13,348 and Special Paper 12 at a cost of £4,174. This printing bill of £17,522 is only £457 less than 1972, although in that year we published two Special Papers', printing charges continue to rise every year. Total expenditure was £18,645, which is the lowest figure since 1970 because of the relatively modest printing programme and savings due to the adoption of filmset printing in the last year, combined with an excess provision of £857 for 1972. There will be no such artificial modification to the account for 1974: for the first time in the history of the Association we have paid for, or know the exact cost of, all the publications scheduled for the past year. Total income in 1973 was £29,456 which is by far the largest the Association has ever enjoyed. Some of this is from donations towards the cost of individual publications and the Association is indebted to the Universities of Ohio, Reading, Natal, Sydney and the Cite Universitaire of Quebec, the Carnegie Trust, and D. Naylor for their help in this way, and to the Abu Dhabi Petroleum Company, the Quatar Petroleum Company, and Shell Oil Company for their contributions for Special Paper 13. But the bulk of our excess income of £10,81 1 was from the good sales of Special Papers , including Special Paper 12 our joint publica- tion with the Systematics Association on ‘Organisms and Continents through Time’. We are much indebted to the Systematics Association for their loan to us of £1,200 to help get this printed. The Association’s reserves now stand at £23,281. This is a satisfactory figure since it covers our aim to have one year’s printing costs in hand. Publications. Four parts of Volume 16 of Palaeontology were published during 1973; they contained 46 papers and consisted of 849 pages and 109 plates. Special Paper 12, the Cambridge Symposium volume ‘Organisms and Continents through Time' was produced early in 1973. Sales of this Special Paper proved most encouraging, and many new subscribers to the series have been obtained. The next Special Paper will be Number 13 for 1974— it will be a commemorative volume in honour of Professor O. M. B. Bulman, comprising papers dealing with current graptolite research. Meetings. Five meetings were held during 1973. The Association is indebted to Professor J. Zussman (Manchester University) and the President, Professor M. R. House (Hull University), for granting facilities for meetings, to the leaders of the two field excursions, and to the local secretaries for their efficient services in organizing the meetings. a. The Sixteenth Annual General Meeting was held in the Meeting Room of The Zoological Society of London on 7 March 1973. Professor P. C. Sylvester-Bradley of Leicester University delivered the Sixteenth Annual Address on ‘Oysters and Jurassic shorelines’. b. A Field Demonstration Meeting was organized by the Carboniferous Group on ‘Lower Carboniferous of the Grange area, north Lancashire’ and led by M. Mitchell, W. H. C. Ramsbottom, W. C. C. Rose (Institute of Geological Sciences, Leeds), and R. F. Grayson (Manchester University), on 28-29 April 1973. It was well attended by about 70 members. c. An extra Field Demonstration Meeting on ‘The Magnesian Limestone in Durham— a study in facies analysis’ was held on 29-30 September 1973 and led by Dr. D. B. Smith (I.G.S., Leeds). Dr. J. Senior (Durham University) acted as local secretary. d. A Symposium on ‘Communities through time’ was held as the Association’s contribution to the second co-ordinated meeting of geological societies of the British Isles, at Manchester University on 20 September 1973. The meeting was organized by Dr. J. E. Pollard. 964 THE PALAEONTOLOGICAL ASSOCIATION e. An Open Discussion Meeting was held at Hull University on 16-19 December 1973. Despite national travel difficulties, over 70 members attended to hear 22 papers and to see 15 demonstrations. Two field excursions were organized to local important sections, led by Drs. J. W. Neale, P. F. Rawson and Professor M. R. House and Dr. M. A. Whyte. The local secretary was Mr. M. Ashton. Council. The following were elected members of Council for 1973-1974 at the Annual General Meeting on 7 March 1973 : President: Professor M. R. House; Vice-Presidents: Mr. N. F. Hughes, Dr. Isles Strachan; Treasurer: Dr. J. M. Hancock (Deputy Treasurer during Treasurer’s absence abroad: Dr. L. R. M. Cocks); Membership Treasurer: Dr. E. P. F. Rose; Secretary: Dr. W. D. I. Rolfe; Assistant Secretary: Dr. C. T. Scrutton ; Editors : Dr. R. Goldring, Dr. J. D. Hudson, Dr. D. J. Gobbett, and Dr. L. R. M. Cocks. Dr. A. J. Lloyd was co-opted to Council to assist the new Membership Treasurer take over his office. Other members: Dr. M. Bassett, Dr. D. D. Bayliss, Professor D. L. Dineley, Dr. Julia Hubbard (Circular Reporter), Dr. C. P. Hughes*, Dr. J. K. Ingham, Mr. M. Mitchell, Dr. M. Muir, Dr. J. W. Murray*, Dr. B. Owens, Dr. P. F. Rawson, Professor D. Skevington. *Co-opted to Publications Subcommittee. Circulars. Many favourable comments have been received on the information service provided by the enlarged Circulars, and this is an encouragement to those involved— particularly the Circular Reporter, Dr. Julia Hubbard. Three Circulars (72-74) were distributed to members, and to over 100 Institutional Members on demand. Council activities. The stimulus committees for 1972-1973 and 1973-1974 have again produced valuable suggestions for improving the Association’s activities. As a result a Conservation Working Group has been set up: Dr. P. Rawson (Convener), Dr. E. Robinson, Professor D. Dineley, and Dr. P. Toghill. It will inquire into other organizations’ views on sites of palaeontological interest, and will explore the publication of a list of sites where collecting would occasion little harm. A questionnaire was circulated to members to elicit the level of support that exists for small group meetings of members. Only 15 responses were received to the c. 1000 questionnaires sent out: a disappointing return, suggesting no great enthusiasm for the organization of such meetings through the Association. Small grants-in-aid from the Association's funds have been suggested to support new important revisionary work. The idea of a correspondence column in Palaeontology was also resurrected by the current stimulus committee. Other suggestions under consideration by Council were that in addition to the new master classes in palaeontology, teach-ins on sites of unusual interest should be held, as should ‘workshops’ on data presenta- tion and more meetings on controversial topics in addition to meetings that would attempt to define guide lines for future development of the science. The President presented proposals, which were implemented by Council, for the introduction of fixed terms of office for officers. This should ensure a future orderly progression of officers through Council, and enable long-term detailed planning of its composition. Through Dr. Porter Kier’s good offices, a copy of his spectacular time-lapse colour film of echinoid behaviour was obtained jointly by The British Universities Film Council and the Association. Dr. Goldring has made two other films available to members; they are on Limulus behaviour. Professor Alwyn Williams has been appointed ‘Treatise’ Adviser for a three-year term. Dr. M. G. Bassett has taken over the co-ordination of members’ orders for ‘Treatise’ volumes. This task was performed for many years by Dr. C. Downie, to whom grateful thanks are extended. The Association has become a Corporate Member of the International Palaeontological Association (formerly I.P.U.). BALANCE SHEET AND ACCOUNTS FOR THE YEAR ENDING 31 DECEMBER 1973 o r- oo r- o o o o in — O o in on in o o on on r- o r- o m o on o n Q .2 c od Dh > •c CD G X) .£ G GO g 5 o cu ’1) ^ ’> ^ o a, CP y X JO PJ G . 'G U 3 £ G G -*-* ■§ < w G -/5 o hJ .O o » 3 CD 5 0) <1 m on ON — < on o m on o on no m ON .2 § . . X> *a 3 t & X) r? C 03 n j- .q u CJ X C3 rto o O (TJ 25 ft g oa nj d GO £ Oh d GO GO C ;§ i5 o X c/5 (2s E< pf o o o z < X M - N Ol iri On M M on r~~ — m W ft! P H 5 z w ft X W ON o v~) (N ON Tl O ^ O O >D O o o NOON'OOO'CONO-OO'O Ci (N Y) Cl ^ -7 p oo ^i^i^^QNO^OON^t^NO c+j E' OO NO NO OO OO OO cn o 0\DDr^ir, 'tODNCrciNbob OO fO NO cd cd cd fti C— ft. o H cd cd ft ft O > 3 o ft El X 3 ft. £ C/5 O Cv, CJ *3 <2^ §•§ O c^ | £ O * 3 . C 3 s ft o .3 H 2 o 3 -S2 O fc(j2i/il!.E 13 •§ a 8 x O Eh HjH w O ft 2 oj 2 > X Jr, 'x f-. S | ec & CD O c3 u OhS S aj 3 O O O '~ H H ON O'! c+i OO i— < ' ON r5* ON ON On . New microfossils from the Silurian (Llandovery, Stage 6) of the Oslo region, Norway, 707. Leangella segmentum, 106. Lehua princeps, 848, 113. Lepidocyclina, 499; {Eulepidina) andrewsiana , 499, 74; (Eulepidina) ephippioides, 500, 74; sp., 502, 74; (Nephrolepidina) spp., 503. Lepidodendroid stoma, 525. Lepidodendron veltheimii, 527, 78, 81. Lepidophloios acerosus, 530, 78, 82; fuliginosum, 83; scottii , 83. Leptostrophia filosa, 106. Levinton, J. S. Trophic group and evolution in bivalve molluscs, 579. Lindstrom, M. The conodont apparatus as a food- gathering mechanism, 729. Lingula sp., 106. Liothyrella neozelanica, 21-25; shell structure, 181. Lithophaga sp., 75, 76. Llandovery : microfossils, 707. Lophophy llidium, 441; extumidum, 65; hadrum , 65, 67; proliferum , 60, 62, 63; simulans, 63; sp. nov. A, 60, 62, 64, 66, 68-70; sp. nov. B, 67; sp. nov. C, 61, 62; sp. nov. D, 61, 62, 64, 67, 68; sp., 60. Lophothyris etheridgii, 27. Lord, A. Ostracods from the Domerian and Toarcian of England, 599. Ludlow: marine communities, 779. M Mackmnon, D. I. and Williams, A. Shell structure of terebratulid brachiopods, 179. McLean, D. M. Two new Paleocene dinoflagellates from Virginia and Maryland, 65. McLean, R. A. Chonophyllinid corals from the Silurian of New South Wales, 655. Maglio, V. J. A new proboscidean from the late Miocene of Kenya, 699. Malongullia, 231 ; oepiki, 232, 30-32. Malonophyllum kansasense, 64. Marginopora vertebralis, 488, 74. Maryland: Paleocene dinoflagellates, 65. Mastodonsaurus lavisi, 265, 282. Matthews, S. C. M. and Thomas, J. M. Lower Car- boniferous conodont faunas from North-East Devon- shire, 371. Megalosaurids: Bajocian of Dorset, 325. Megalosaurus, 326; hesperis, 326, 42, 43; nether com- bensis, 333, 44. Meiourogonyaulax bu/loidea, 533, 92; psoros , 634, 92; stoveri, 633, 93; sp., 635, 92. Mensah, M. K. See Chaloner, W. G. Mesenteripora sp., 77. Mesopholidostrophia sp., 106. Mestognathus beckmanni, 50. 970 INDEX Microfossils: Llandovery, 707. Mictocystis , 662; endophylloides , 664, 95. Midlands: Permian Pelycosaurs, 541. Miocene: Foraminifera, 475; proboscidean, 699; spider- crab, 869. Miogypsina (Miogypsinoides) bantamensis, 496, 73; complanata, 494, 73; dehaarti , 497, 73; neodispansa, 497, 71. Miospores: Devonian assemblages, 322. Moody, R. T. J. See Walker, C. A. Moorallina sp., 77. Muderongia staurota, 644, 91. M urania , 341 ; lefeldi , 341, 45, 46. N Neseuretus , 75; ( Neseuretus ) cf. N. complanatus, 83, 9; tristani, 76, 9. New Brunswick: Devonian heterosporous plant, 387. New South Wales: Carboniferous conodonts, 909; Ordovician trilobites, 203; Silurian corals, 655. New Zealand: Miocene spider-crab, 869. North America: Ordovician graptolites, 1. Norway: Silurian microfossils, 707. Nubeculinella sp., 76. Nucleospira pisum, 106. O Oligocene: Foraminifera. 475. Oligosphaeridium nannum, 636. Onycopyge, 409, 41 1 ; liversidgei, 413, 58. Operculina, 49 1 . Ophiacodon sp., 548, 84. Opipeuter , 112; inconnivus, 112, 13, 14. Ordovician : graptolites from eastern North America, 1 ; Scharyia, 685; trilobites from Cornwall, 71 ; trilobites from New South Wales, 203; trilobites from South Wales, 841. Orthograptus amplexicultis, 29, 2. Ostracods : English Jurassic, 599 ; Quaternary of Argen- tina, 669. Owens, R. M. The affinities of the trilobite genus Scharyia , with a description of two new species, 685. Oxytoma costatum, 75. P Pachydyptes, 293 ; simpsoni, 294, 37-39. Palaeocene: dinoflagella tes, 65. Palaeoecology : conodont, 740; Eocene penguin, 304. Palaeogeography : Wenlock, 770. Palaeozoic: acritarchs, 41 ; tetracorals, 441. Palmer, C. P. A new genus of Jurassic bivalve mollusc ancestral to Globocardium, 165. Palmer, T. J. and Fiirsich, F. T. The ecology of a Middle Jurassic hardground and crevice fauna, 507. Pampacythere, 674; multiperforata , 675, 96; solum, 678, 97. Parkesolithus, 224; dictyotos, 221, 29-31. Paroedinia, 644; aceras, 645, 93; sp., 645, 93. Paton, R. L. Capitosauroid labyrinthodonts from the Trias of England, 253; Lower Permian pelycosaurs from the English Midlands, 541. Patrognathus andersoni, 119. Pelycosaurs: Permian of England, 541. Penguin: Eocene of Australia, 291. Penn, I. E. The production of stratigraphical range- diagrams by automatic methods, 553. Permian: English pelycosaurs, 541 ; tetracorals, 441. Perotr Hites microbaculatus, 41. Phacops rana, 349, 353; rana africanus , 353, 47, 48; rana crassituberculata, 47; rana milleri, 47; rana tindoufensis, 355, 47; schlotheimi, 48. Phlebopteris dunkeri , 87. Phylogeny: Capitosauroid labyrinthodonts, 283. Pickett, J. W. and Jell, J. S. The Australian tabulate coral genus Hattonia, 715. Plagioecia, 77. Plants: Devonian, 311, 387, 565, 925; Lepidodendroid stoma, 525; Podocarpus, 365; Weichselia leaves, 587; Williamsoniella, 125. Pleuropugnoides, 819; pleurodon, 819, 108. Plicatula fistulosa, 76; sp., 77. Podocarpus, 365; tzagajanicus, 368, 49. Pollen: compression of, 125; Williamsoniella lignieri, 125, 15. Polygnathus communis Carina , 50; communis communis, 50; inornatus, 50; sp. A, 119. Praelongithyris praelonga, 27. Price, D. Trilobites from the Sholeshook Limestone (Ashgill) of South Wales, 841. Priscogalea, 41, 47; cortinula, 50, 4; distincta, 50, 4; fimbria, 47, 3; simplex, 47, 3. Proboscidean, 699. Prolixosphaeridium, 636 ; parvispinum, 637. Protocardia, 176; vicaryi, 176, 20. Protochonetes ludloviensis, 106; minimus, 106. Protoellipsodinium clavulum, 637, 93. Prototaxites sp., 318, 40. Pseudoclirnacograptus modestus, 24; scharenbergi, 27, 2; scharenbergi stenostoma, 26. Pseudoglossothyris curvifrons, 27. Pseudopoly gnathus nodomarginatus , 119; triangulus pin- natus, 51, 119; triangulus triangulus, 50. Pseudo sphaerexochus boops, 850, 113; juvenis , 849, 113. Q Quaternary: ostracod, 669. R Range-diagrams, 553. Rasul, S. M. The Lower Palaeozoic acritarchs Prisco- galea and Cymatiogalea, 41. Regnellia, 712; camera, 712, 102. Remopleurides exallos, 29. Reptiles: Permian pelycosaurs, 541. Richardson. J. B. See Edwards, D. E. Riding, R. The Devonian genus Keega (Algae) re- interpreted as a stromatoporoid basal layer, 565. Riva, J. A revision of some Ordovician graptolites of eastern North America, 1 . INDEX 971 S Sadler, P. M. Trilobites from the Gorran Quartzites, Ordovician of South Cornwall, 71. Salopella, 315; allenii, 315, 40, 41. Salopina lunata, 106. Sandvikina , 707; brachiata, 707, 101, 102. Scaliognathus anchoralis, 50, 119. Scharyia, 685, 688; heothina , 693, 99; siceropotrix, 689, 98; sp. 1, 694, 99. Schuchertella , 815; sp., 107. Serpula (Cycloserpula) sp., 77; ( Dorsoserpula ) sp., 77; (Tetraserpula) sp., 77. Silurian: brachiopod distribution, 879; corals, 655; epizoa on brachiopods, 423; Hattonia, 715; micro- fossils, 707 ; Wenlock stratigraphy, 745. Siphonodella cooperi, 50; isosticha, 50; sp., 50, 51, 119. Skenidioides lewisii , 106. Slovakia: Cretaceous sclerosponge, 341. Sorites orbiculus , 487, 74. Spathognathodus stabilis , 51. Spliaerirhynchia wilsoni , 106. Sphaerocodium , 85. Sphaerocoryphe exserta, 237, 33. Sphenacodon britannicus, 542, 84. Spider-crab, 869. Spiniferites sp., 638, 91. Spiriferellina perplicata, 836, 111. Spiroclypeus, 491 ; margaritaceus, 492, 72. Spitsbergen : Ordovician trilobite. 111. Spongiophyton lenticulare, 932, 945, 120-122; nanum, 926, 945, 120-124. Spores: Chaleuria cirrosa , 390, 55-57; Devonian, 318. Stachyodes, 570; australe , 572, 85; sp., 86. Staffordshire: Carboniferous brachiopods, 811. Stegacanthia sp. indet., 818, 107. Stenopareia bowmanni, 842, 112. Stereostylus lenis , 62. Stoma: lepidodendroid, 525. Stomatopora dichotoma, 77. Streptis grayii, 106. Stromatoporoid : Devonian, 565. Systematophora schindewolfi, 640. T Tabulate: Hattonia, 715. Taphrognathus varians , 119. Tasmania: Cambrian trilobite, 95. Tavener-Smith, R. Early growth stages in rhabdomesoid bryozoans from the Lower Carboniferous of Hook Head, Ireland, 149. Tayamaia marianensis, 71, 74. Techniques: mounting material for SEM examination, 431; specimen location technique for SEM strew mounts, 435. Tetracorals: Upper Palaeozoic, 441. Tetraspis moeldenensis, 844, 112, 113. Texas: Permian tetracorals, 441. Thomas, B. A. The Lepidodendroid stoma, 525. Thomas, J. M. See Matthews, S. C. M. Timorphyllum, 441, 470; wanneri, 61-63, 66, 70. Toarcian: ostracods, 599. Toernquistia arguta, 223, 29. Tracheophyta, 313. Tremadocian : acritarchs, 41. Triarthrus sp., 214, 32. Trias: labyrinthodonts from England, 253. Trichodinium sp., 640, 92. Trilobites: Cambrian of Tasmania, 95; mode of life, 419; Onycopyge, 409; Ordovician of Cornwall, 71; Spitsbergen, Ireland, and Utah, 111; New South Wales, 203; Scharyia, 685; South Wales, 841. Trionyx, 901 ; silvestris, 901, 118, Trypanites , 125, 126; sp., 75. Turtle: Eocene, 901. U Utah: Ordovician trilobite. 111. V Verdier, J. P. See Davey, R. J. Vertebrates: Eocene turtle, 901; labyrinthodonts, 253; megalosaurids, 325; Miocene proboscidean, 699; pelycosaurs, 541 ; penguin, 291. Virginia: Paleocene dinoflagellates, 65. Visbyella trewerna, 106. W Waldman, M. Megalosaurids from the Bajocian (Middle Jurassic) of Dorset, 325. Wales: Ordovician trilobites, 841; Wenlock marine communities, 779; Wenlock stratigraphy, 745. Walker, C. A. and Moody, R. J. J. A new trionychid turtle from the Lower Eocene of Kent, 901. Wallodinium tuna, 645. Webby, B. D. Upper Ordovician trilobites from Central New South Wales, 203. Weichselia, 587, 87-89. Welsh Borderland: Devonian plants, 311; Wenlock communities, 779; Wenlock stratigraphy, 745. Wenlock: stratigraphy, 745. Whatley, R. C. and Cholich, T. D. C. A new Quaternary ostracod genus from Argentina, 669. Williams, A. See Mackinnon, D. I. Williamsoniella lignieri, 125, 15. Y Yassia, 665 ; enormis, 666, 95. Z Zoogeography: Ordovician trilobites, 208. Zosterophyllum sp., 313, 40. THE PALAEONTOLOGICAL ASSOCIATION The Association was founded in 1957 to further the study of palaeontology. It holds meetings and demonstrations as well as publishing Palaeontology and Special Papers in Palaeontology. Membership is open to individuals and to institutions on payment of the appropriate annual subscription : Institutional membership .... £10 00 (U.S. $26.00) Ordinary membership .... £5-00 (U.S. $13.00) Student membership .... £3-00 (U.S. $8.00) There is no admission fee. Institutional membership is only available by direct application, not through agents. Student members are persons receiving full-time instruction at educational institutions recognized by the Council. On first applying for membership, an application form should be obtained from the Membership Treasurer. Subscriptions cover one calendar year and are due each January; they should be sent to the Membership Treasurer, Dr. E. P. F. Rose, Department of Geology, Bedford College, Regent’s Park, London, NW1 4NS, England. PALAEONTOLOGY All members who join for 1974 will receive Volume 17, Parts 1-4. All back numbers are still in print and may be ordered from B. H. Blackwell, Broad Street, Oxford, OX1 3BQ, England, at £5 per part (post free). A complete set, Volumes 1-16, consists of 63 parts and costs £315. SPECIAL PAPERS IN PALAEONTOLOGY The subscription rate is £8 (U.S. $22.00) for Institutional Members and £4 (U.S. $11.00) for Ordinary and Student Members. Subscriptions should be placed through the Membership Treasurer, Dr. E. P. F. Rose, Department of Geology, Bedford College, Regent’s Park, London, NW1 4NS, England. Ordinary and Student members only may obtain individual Special Papers from Dr. Rose at reduced rates. Non-members may obtain them at the stated prices from B. H. Blackwell, Broad Street, Oxford, OX1 3BQ, England. COUNCIL 1974-1975 President : Professor C. H. Holland, Department of Geology, Trinity College, Dublin, 2, Ireland Vice-Presidents : Dr. R. Goldring, Department of Geology, The University, Reading, RG6 2AB Dr. W. D. I. Rolfe, The Hunterian Museum, The University. Glasgow, G12 8QQ Treasurer : Dr. J. M. Hancock, Department of Geology, King’s College, Strand, London, WC2R 2LS Membership Treasurer: Dr. E. P. F. Rose, Department of Geology, Bedford College, Regent’s Park, London, NW1 4NS Secretary: Dr. C. T. Scrutton, Department of Geology, The University, Newcastle upon Tyne, NE1 7RU Editors Dr. J. D. Hudson, Department of Geology, The University, Leicester, LEI 7RH Dr. L. R. M. Cocks, Department of Palaeontology, British Museum (Natural History), Cromwell Road, London, SW7 5BD Dr. C. P. Hughes, Department of Geology, Sedgwick Museum, Downing Street, Cambridge, CB2 3EQ Dr. J. W. Murray, Department of Geology, The University, Bristol, BS8 1TR Other Members of Council Dr. M. G. Bassett, Cardiff Dr. M. C. Boulter, London Professor D. L. Dineley, Bristol Dr. J. K. Ingham, Glasgow Dr. J. E. Pollard, Manchester Dr. A. W. A. Rushton, London Dr. P. Wallace, London Dr. D. D. Bayliss, Llandudno Dr. C. H. C. Brunton, London Dr. J. A. E. B. Hubbard, London Dr. C. R. C. Paul, Liverpool Dr. P. F. Rawson, London Professor D. Skevington, Galway Overseas Representatives Australia: Professor Dorothy Hill, Department of Geology, University of Queensland, Brisbane Canada: Dr. B. S. Norford, Institute of Sedimentary and Petroleum Geology, 3303-33rd Street NW, Calgary, Alberta India: Professor M. R. Sahni, 98 Mahatma Gandhi Marg, Lucknow (U.P.), India New Zealand: Dr. C. A. Fleming, New Zealand Geological Survey, P.O. Box 30368, Lower Hutt West Indies and Central America: Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad, Inc., Pointe-a-Pierre, Trinidad, West Indies Western U.S. A.: Professor J. Wyatt Durham, Department of Paleontology, University of California, Berkeley 4, California Eastern U.S. A. : Professor J. W. Wells, Department of Geology, Cornell University, Ithaca, New York Palaeontology VOLUME 17 • PART 4 CONTENTS The conodont apparatus as a food-gathering mechanism MAURITS LINDSTROM 729 Review of the stratigraphy of the Wenlock Series in the Welsh borderland and South Wales MICHAEL G. BASSETT 745 Wenlock and Ludlow marine communities in Wales and the Welsh borderland C. E. CALEF and N. J. HANCOCK 779 A Lower Carboniferous brachiopod fauna from the Manifold Valley, Staffordshire C. H. C. BRUNTON and C. CHAMPION 811 Trilobites from the Sholeshook Limestone (Ashgill) of South Wales DAVID PRICE 841 A new spider-crab from the Miocene of New Zealand R. J. F. JENKINS 869 A new trionychid turtle from the Lower Eocene of Kent C. A. WALKER and R. T. J. MOODY 901 Lower Carboniferous conodont biostratigraphy of New South Wales T. B. H. JENKINS 909 Non-vascular land plants from the Devonian of Ghana W. G. CHALONER, M. K. MENSAH and M. D. CRANE 925 Ecological succession in intraformational hardground formation R. GOLDRING and J. KAZMIERCZAK 949 R Palaeontological Association Report and Accounts for 1973 963 Index to Volume 17 967 Printed in Great Britain at the University Press, Oxford by Vivian Ridler, Printer to the University * SNI NVIN0SH1IIMS S3iaVaai3 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSH1II' 2 CO z . CO 2 * CO 2 fc § I » I 'W, | | _JP|s 1 1 X CO § , ^ /'■•' ' f- _ >■ s " 5 ^ >* i CO z CO Z CO '•■“ Z CO ES SMITHSONIAN INSTITUTION NOlinillSNI _ NVINOSH1IWS SSIHVaail LIBRARIES SMITHSON!/1 co zz co 2 \ 00 - z co o z CO z o SNI NVINOSH1IWS S3iavaeil LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSH1M r- „ ^ Z <“ 2 r- ~ ^ ' - 0 " 00 ^ co m co fz CO — -- — U/ IES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3IHVUan LIBRARIES SMITHSON!/ CO 2 .... CO Z CO 2 '« 2 (/) *■■ Z CO — -- — .sni NviNosHiiws ssiavaan libraries Smithsonian institution NoiiniusNi nvinoshiiv — ui — co ~ co — co CO q XX 2 XCOjuA^Z O ^ oC-Z _ •W*®' O x^misj^z q ES*2 SMITHSON I AN-* INSTITUTION NOlinillSN 1 ^NVINOSHIIIAIS^S 3 I dVHa M^LIBRAR I E S^ SMITHSONI/ 2 <2 z r- , 2 (- 2 O m xTToT/oX 2 /TTLILToX m an. CD “ Z'S'TvJJx *-* ^ i— 03 TO t— 00 /GjjffFp&S co Z m x^vos^iz :z ytAs>z m ' zz x^vas^z m _ co _ co _ co 2 .sni nvinoshhws ssiavaan libraries Smithsonian institution NoiiniusNi nvinoshiii — CO Z CO 2 V CO Z co § \v,, > 5 ■ ' vr > 2 • \X' > C/5 X CO ■*— CO ** " Z C/5 ES SMITHSONIAN INSTITUTION NOlinillSNI _ NVIN0SH1HMS S3IHVaaiT LIBRARIES SMITHSONI/ C/5 — C/5 \ C/5 SNI NVIN0SH1IIAJS SSIBVaail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIII/ avaan libraries Smithsonian institution NoiiniiiSNi nvinoshhiajs SBiavaan libr 2 , CO 2 v CO 2 CO 2 < 2 AV > '"-££1^ 2 CO 2 CO ■- 2 CO 2 CO , 'ITUTION NOlifUllSNI NVINOSHIIIAIS SBiaVaail LIBRARIES SMITHSONIAN INSTITUTION NOlif avaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiins saiavaan libr I- 2 I- 2 f- i'uiistjx m CO — CO X CO riTUTION NOlifUllSNI NVINOSHIIIAIS SBiaVaaiT LIBRARIES SMITHSONIAN INSTITUTION NOIll x CO o > 2 J, 2 y > *■ 2 co 2 co * 2 co avaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws saiavaan libr co X co — , . co lx) :c\ — ■*\ ct: < _ a. jpV m co q \"NdC/ q ITUTION NOlinillSNI ^NVINOSHIIIAIS ^S 3 lavaail^LIBRARI E S^ SMITHSONIAN-* INSTITUTION ^ NOlif 2 r~ , 2 r- 2 r- 2 O 2CIVriTo>v ~ ax O /JJSirSX /"Vi vSv O „ /£>a ^oX O > ;o m - m x^vosHix X m co — co X co BRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS SBIBVaail LI CO 2 « CO 2 CO 2 co ■— — * Y ^ — y > 5 s ^ 2, ' 2 X^VOS^A > 2 Z CO ■*' 2 CO 2 CO ITUTION NOlinillSNI NVIN0SH1IWS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOlif co — co — co U 5 #4-. | ^ ^ CO o 2 avaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws saiavaan libr C^ijii v*/ m co m x, m co X co £ — co V _ riTUTION ’/^^cv'.SNI NVINOSHIIIAIS SBiaVaaiT LIBRARIES SMITHSONIAN INSTITUTION N0I1 CO # k\ ^ Z CO 2 CO 2 v CO < 2 .< aVx 2 2 co X