o X if) if) m E t i > 2 ^'Q^_oc^ j, 2 ^ > xiOiiist^ 2 "^ > 2 if) Z W * Z V z yan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws saidvaan librar ~ CO 2 ^ — Vo o X»\ pc>^ — o JTION NOIinillSNI~'NVINOSHillNS SaidVdaiT LIBRARIES SMITHS0NIAN“^INSTITUTI0N NOlinili Z (- X, 2 r- 2 — rn '' \^' ^ m ^ rn CO ' r: CO £ co — dan LIBRARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHillMS S3ldVdail LIBRAR z , CO ^ z , CO 2 CO 2 ■. ’ - /f.rr*v 5 c/> z x. 1 1 I TT Z 1 LIBRAR CO JTION NOliniliSNI NVINOSHillMS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOlinili r- V 2 r- 2 r~ 2 O ^ X) CO X CO CO aan libraries Smithsonian institution NoiiniiiSNi nvinoshiiims saidvdan librar :;■ ^ _ CO 2 ^ ^ 5 ' < \^ 2 VjstitiTX < 2 < --cxrr^ 2 . < ,Cy' / 2 O CO X 3, 'v.jvAVi^' 2 ^ - 2 X^oMi^ > if) -Z CO ■• ‘ Z CO — z JTION NOliniliSNI NVINOSHillMS S3ldVaan LIBRARIES SMITHSONIAN INSTITUTION NOlinili' CO — •, CO — CO X CO X UJ ^rLTrirTX „ X LlI ✓^TITiX ^ i . I X -.-XnTX LlJ /r^sTiriX THSONIAN^INSTITUTlON^^NOliDiliSNI NVINOSHilWS^^S 3 iavyan_LlBRARIES SMITHSONIAN _ I NSTI7 co __ c/) 2 2 m — >^ 00 ?=; Vp;T NOSHilWS^S3 I ava a I1~*LI B RAR I ES^SMITHS0NIAN“‘lNSTlTUTI0N^N01inillSNI^NVIN0SHillAiS^S3 I a> fi/ :z ’" 5 CD PQ > ™ t^Ti THSONIAN~INSTITUTION NOlinillSNI ~ NVINOSHilWS S3iavaail L I B R A R I E S SMITHSONIAN INSTIl — f/5 _ T NOSHimS^^SH I ava a n^LI B RAR l ES^^SMITHSONIAN institution N01iniliSNI_NVIN0SHimS^S3 I a tn — in — CO -p' \ . . ^ ITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIIMS S3iavaai1 LIBRARIES SMITHSONIAN^ INSTI z r- 2 ^ NOSHiii*JS saiavaa CO DO 7i > -Xi CO ^ BRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHlil^S S 3 I a m — fnP'^ z: o CO X ^ V// >«— CO THSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS S3iavaa — (O — CO > . ^ I^LIBRARIES SMITHSONIAN — CO (X J§ ^ -• VC’ Z2 CQ 7^ xo] Q x^ pc^ ~ x^njii^^iX o ^ O xgN pc^ _ ^ O N0SHimS^S3 I ava a n*^LI B RAR l ES^SM1THS0N1AN“‘|NSTITUTI0N NOlinillSNI NVINOSHimS S3 I a z r- z: [[ z ' " ;;; aT3£5oX 9. ra DO X) > lp m ^5 ' m m CO ~ ^ 4/^ ITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS S3iavaai3 L I B R A R I E S SMITHSONIAN I NSTI ■y , in Z CO Z z » 5 '• ^ ^ ^ < ><^vct/7x < ». S ^sSi- . S ^ ^ — /u'2«T5!>K\ — X CO o z ^ > ' S '^osviiV >■ 2 ' 5 CO ■*'“ Z CO Z CO ^ 2 CO NOSHiiiNS S3iavaan libraries Smithsonian institution NoiiniiiSNi_NviNOSHims s3 tf\ — — tr\ — * > X CO CO VOLUME 15 ’ PART 3 Palaeontology SEPTEMBER 1972 PUBLISHED BY THE PALAEONTOLOGICAL ASSOCIATION LONDON Price £5 THE PALAEONTOLOGICAL ASSOCIATION The Association was founded in 1957 to further the study of palaeontology. It holds meetings and demonstrations, and publishes the quarterly journal Palaeontology and Special Papers in Palaeontology. Membership is open to individuals, institutions, hbraries, etc,, on payment of the appropriate subscription: Institute membership £10-00 (U.S, $28.00) Ordinary membership ..... £5-00 (U.S. $14.00) Student membership ..... £3-00 (U.S. $9.00) There is no admission fee. Institute membership is only available by direct appli- cation, not through agents. Student members are persons receiving full-time instruc- tion at educational institutions recognized by the Council; on first applying for membership, they should obtain an application form from the Membership Treasurer. All subscriptions are due each January, and should be sent to the Membership Treasurer, Dr. A. J. Lloyd, Department of Geology, University College, Gower Street, London, WCIE 6BT, England. COUNCIL 1972-3 President: Professor M. R. House, The University, Kingston upon Hull, Yorkshire Vice-Presidents: Dr. Gwyn Thomas, Department of Geology, Imperial College, London, S.W. 7 Mr. N. F. Hughes, Sedgwick Museum, Cambridge Treasurer: Dr. J. M. Hancock, Department of Geology, King’s College, London, W.C. 2 Membership Treasurer: Dr. A. J. Lloyd, Department of Geology, University CoUege, Gower Street, London, WCIE 6BT Secretary: Dr. W. D. I. Rolfe, Hunterian Museiun, The University, Glasgow, W. 2 Editors Dr. Isles Strachan, Department of Geology, The University, Birmingham, B15 2TT [Dr. R. Goldring, Department of Geology, The University, Reading, R66 2AB, Berks. Dr. J. D. Hudson, Department of Geology, The University, Leicester Dr. D. J. Gobbett, Sedgwick Museum, Cambridge Dr. L. R. M. Cocks, Department of Palaeontology, British Museum (Natural History), London, S.W. 7 Other members of Council Dr. M. G. Bassett, Cardiff Dr. E. N. K. Clarkson, Edinburgh Dr. R. H. Cummings, Abergele Prof. D. C. Dineley, Bristol Dr. Julia A. E. B. Hubbard, London Dr. J. K. Ingham, Glasgow Mr. M. Mitchell, Leeds Dr. Marjorie D. Muir, London Dr. B. Owens, Leeds Dr. W. H. C. Ramsbottom, Leeds Dr. P. Rawson, London Dr. Pamela L. Robinson, London Dr. A. D. Wright, Belfast 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-^-Pierre, Trinidad, West Indies Western U.S. A. : Professor J. Wyatt Durham, Department of Paleontology, University of California, Berkely 4, California Eastern U.S. A.: Professor J. W. Wells, Department of Geology, Cornell University, Ithaca, New York © The Palaeontological Association^ 1972 PHOSPHATIZED OSTRACODS WITH APPENDAGES FROM THE LOWER CRETACEOUS OF BRAZIL by RAYMOND HOLMES BATE Abstract. A new ostracod, Pattersoncypris micwpapiUosa gen. et sp. nov., is described from the Aptian to Albian Santana Formation, Serra do Araripe, Ceara, Brazil. Of the 253 specimens obtained, 138 are complete carapaces of which 103 contain appendages in an excellent state of preservation. Of the total number of speci- mens only 15 are pre-adult instars and 6 are males. A small number (24) of single valves possibly represent moulted carapaces, the remainder were living up to the time of burial and fossilization. The original calcium carbonate and chitin of the ostracod was replaced by apatite immediately after death. The ostracods were found entombed with a teleost fish, Cladocyclus garditeri Agassiz, the decaying body of which, on which the ostracods are considered to have been feeding, is considered to have been the source of the phosphate salts responsible for the mineralization of the ostracod anatomy. The Santana Formation of Serra do Araripe, Ceara, northern Brazil, is a gypsiferous marl succession dated as ranging from the Upper Aptian to the Lower Albian (Beurlen, 1970). Fish-bearing nodules are particularly common in the marl beds and are collected and sold commercially for their well-preserved fish remains. Several such nodules have been prepared recently by Dr. Colin Patterson of the British Museum [Natural Flistory], hereafter abbreviated (B.M.N.FI.); the enclosed fish skeletons being etched from the nodule with acetic acid. In all previous preparations the calcareous ostracods present dissolved in the acid and were lost. In one nodule, however, examination of the washed residue revealed a number of well-preserved ostracods, complete with their appendages. A careful examination of further residues from the same nodule resulted in the recovery of 253 specimens consisting of whole carapaces and single valves of both adult and juvenile instars. In addition, many hundreds of eggs were recovered (PI. 67, fig. 8) which may belong to this ostracod although they are small compared to the eggs of comparable living forms. In transmitted light the eggs appear completely transparent, with what was probably the yolk sac now apparently replaced by calcium carbonate. The ostracods all belong to the same species and are clearly resistant to the acetic acid. An X-ray powder analysis of the carapace and of the body and appendages, by Dr. R. Davies (B.M.N.H.), revealed that both are preserved as apatite. Some carapaces still retain a calcium carbonate infilling but none of this material appears in the carapace itself. Of the 138 complete carapaces, 103 retain appendages, many in an excellent state of preservation, and all represent individuals living up to the time of burial and mineraliza- tion. The remainder, together with the 24 single valves, probably represent moulted instars. Although an almost complete instar range has been obtained for Pattersoncypris micropapillosa only 15 pre-adult instars have been found; the remainder being adults falling into the final instar cluster as illustrated in text-fig. 1. Of the adult instars, only 6 have been identified as being males. Although this discovery represents the best-preserved fossil ostracod material ever found, it is not the first record of fossil appendages. Brongniart (1876) was the first to [Palaeontology, Vol. 15, Part 3, 1972, pp. 379A-393, pis. 66-71.] C 9016 C C 380A PALAEONTOLOGY, VOLUME 15 record these in his Palaeocypris edwardsii from the Carboniferous of Saint-Etienne, whilst Sylvester-Bradley (1941) records the presence of appendages in Pleistocene speci- mens of Cypris piibera O. F. Muller. Subsequently, Gocht and Goerlich (1957) obtained chitinous appendages from some Jurassic and Cretaceous ostracods by dissolving com- plete carapaces in dilute hydrochloric acid. Gramann (1962) similarly observed chitinous antennae and antennules in the Liassic Bairdia molesta Apostolescu. More recently, Schmidt and Sellmann (1966) and Eagar (1970) have described mummified ostracods from the Pleistocene of Alaska and New Zealand respectively. Sohn (personal communi- cation), however, is of the opinion that Eagar’s find is a fresh-water mite and not an ostracod. None of these previous records attains the excellent state of preservation of Pattersoncypris micropapUIosa. The only previous records of ostracods from the Santana Formation appear in Santos and Valen^a (1968), where Candonopsis sp. and Schideridea sp. were recognized, and more recently in Krommelbein and Weber (1972), see p. 388. The occurrence Patterson- cypris micropapiUosa in the Santana Formation has been briefly recorded elsewhere (Bate, 1971). The photographic illustrations for this paper were taken on a Cambridge Stereoscan scanning electron microscope whilst the text-figs, were drawn usinga Leitz camera lucida. All the specimens of Pattersoncypris micropapiUosa are registered in the Collections of the Depart- ment of Palaeontology, British Museum [Natural History]. SYSTEMATIC DESCRIPTION Subclass OSTRACODA Latreille 1806 Order podocopida Muller 1 894 Suborder podocopina Sars 1866 Family cyprididae Baird 1845 Subfamily cypridinae Baird 1845 Genus pattersoncypris nov. Type species. Pattersoncypris micropapiUosa sp. nov. Derivation of name: in honour of Dr. Colin Patterson. Diagnosis: Cypridinae having oval carapace with acute antero-dorsal hump. Description: Carapace ovoid in lateral view, convex in dorsal view. Greatest length of carapace passes through medial line. Shell surface without coarse ornamentation. Right valve with acutely projecting anterior cardinal angle and concave antero-dorsal slope. Hinge adont: groove situated in right valve, terminally expanded and accepting dorsal margin of left valve. Muscle scars as for family. First thoracic appendage adapted for feeding, with distal hook in both dimorphs. Antennules with 7 podomeres; antennae with 3 podomeres: first podomere with 6 swimming setae on inner face. Second and third thoracic appendages elongate, with 4 and 3 podomeres respectively. Furcae long. Paired hemipenes with broad, triangular proximal shield. Remarks. Pattersoncypris is a typical member of the Cyprididae, having the muscle scars, carapace detail, and appendages of that family. Although without the marginal spines R. H. BATE: PHOSPHATIZED OSTRACODS 381 of the Recent Cypris piibera O. F. Muller, there is a close similarity between the two genera with respect to the strong dorsal hump in the region of the anterior cardinal angle. The dorsal hump in Pattersoncypris is, however, a development of the anterior part of the hinge separate from the cardinal angle, and this, together with the more generally ovoid carapace outline serves to distinguish it from the genus Cypris. The only fossil genus which approaches Pattersoncypris in carapace outline is Brasa- cypris Krommelbein (1965a), described from the (?) Upper Jurassic to lower Cretaceous Bahia Series of Brazil. This genus, represented by the single species B. ovum Krommel- bein, differs from Pattersoncypris in the angular development of the posterior cardinal angle thereby giving it a more quadrate outline in lateral view. The right valve is also noticeably smaller than the left whilst in Pattersoncypris the carapace is virtually equivalve. The following are currently placed in Pattersoncypris: Pattersoncypris micropapiUosa — type species, Santana Formation, Serra do Araripe, Brazil. Plourcqia angiilata salitrensis Krommelbein and Weber, 1792, Santana Formation, Pernambuco, Brazil = Pattersoncypris angulata salitrensis (K. and W.). Hourcqia angulata sinuata Krommelbein and Weber, 1 972, Riachuelo beds, Alagoas, Brazil = Patterson- cypris angulata sinuata (K. and W.). Pattersoncypris micropapiUosa sp. nov. (Pis. 66-71, text-figs. 1-12) Derivation of name. With reference to the small papillae which cover the shell. Geological horizon and locality. Aptian/Albian, Santana Formation, Serra do Araripe, Ceara, northern Brazil and from Continental beds from Liberia, N.W. Africa. Material from Liberia is not, however, included in the description of the species. Description. Carapace oval in outline, virtually equivalve, with distinct antero-dorsal hump produced by the dorsal extension of the anterior hinge element, especially of the right valve (PI. 66, fig. 1). Antero-dorsal slope of right valve concave; convex in the left valve. Anterior and posterior margins rounded. Instars III; V; VTII; and IX have been recognized and the length/height measurements plotted (text-fig. 1). A single clustering of individuals in the adult (IX) instar indicates that although both males and females are present the sexes are indistinguishable on earapace outline. The variable gape of the carapaee in ventral view makes an accurate width measurement difficult. Eggs, which may through their association be regarded as belonging to this species despite their small size, have been plotted on text-fig. 1. The shell surface is covered by very small papillae which are espeeially apparent on the ventral surface (PI. 68, fig. 4). Normal pore canals have sensory bristles projecting, although much shortened by breaking off at the tips, and are particularly well developed and more closely spaced in the anterior part of the earapace (PI. 68, fig. 5). The hinge is adont with the terminally expanded groove situated in the right valve (text-fig. 2). The left valve hinge is simply the dorsal edge of the valve (text-fig. 9). Internally the selvage is set well baek from the anterior free margin (text-fig. 2), LENGTH 382 PALAEONTOLOGY, VOLUME 15 TEXT-FIG. 1. Size distribution of instars in Pattersoncypris micropapillosa sp. nov. EXPLANATION OF PLATE 66 Figs. 1-6. Pattersoncypris micropapillosa sp. nov. All figs. X70. 1. External view, right side, female carapace, paratype lo. 4692. 2. External view, right side showing extended furcal rami, female carapace, paratype lo. 4693. 3. External view, left side, female carapace, paratype lo. 4704. 4. External view, left side, juvenile carapace, paratype lo. 4706. 5. External view, right side, juvenile carapace, paratype lo. 4702. 6. Dorsal view, female carapace, paratype lo. 4704. EXPLANATION OF PLATE 67 Figs. 1-5. Pattersoncypris micropapillosa sp. nov. 1. Appendage and body muscles, some attached to inside of carapace, showing typical mammillated crystal growth of apatite. Paratype lo. 4712 X 230. 2. Egg with small surface fracture. X 337. 3. Ventral view of female carapace to show extension of selvage over body of the animal. Female paratype, lo. 4698 X 70. 4. Ventral view, female carapace to show outer and inner lamella and appendages, holotype lo. 4680 X 70. 5. Enlarged view of oral region of holotype lo. 4680, to show basal podomere and endopodite of mandible, upper lip, maxillae, and proximal end, with setae, of first thoracic appendages. X 280. Palaeontology, Vol. 15 PLATE 66 BATE, Phosphatized ostracods Palaeontology, Vol. 15 PLATE 67 BATE, Phosphatized ostracods o I I * td'V R. H. BATE: PHOSPHATIZED OSTRACODS 383 TEXT-FIG. 3. Left antennule, female paratype, lo. 4713. whilst ventrally it develops into a prominent flap which covers the thorax when the valves are closed (text-fig. 2, PI. 67, fig. 3). The inner margin and line of concrescence do not coincide and a distinct vestibule, broader anteriorly, is developed along the valve margin. The line of concrescence is so close to the valve margin that there is no fused duplicature for the development of radial pore canals, which are not developed. The muscle scars (PI. 68, fig. 1) are typical of the family. Antennules are long and possess seven elongate podomeres (text-fig. 3) as in the recent Cypridopsis vidua (Muller); but apart from those shown in PI. 69, fig. 1, they are not sufficiently well preserved to show the setae. The length of the antennules suggests that they were used in swimming and that normally long, swimming setae would be present. 384 PALAEONTOLOGY, VOLUME 15 TEXT-FIG. 4. a, outside face, left antenna, b, inside face, right antenna to show natatory setae. Female paratype, lo. 4707. The antennae are large and powerful (text-fig. 4n, b, PI. 69, figs. 3, 4) and were useful in both swimming and crawling over the substratum. Six long, swimming setae, one of which has a feathered distal end, are situated on the inner face of the first of the three endopodite podomeres, whilst additional long setae, approximately 5 in number, are present at the distal end of the second podomere. Further setae are also present at the distal end of the smaller third podomere. The protopodite of the antenna appears to consist of two podomeres which may have fused to form a single segment, but this is difficult to determine and may be a product of fossilization. The mandibles (text-fig. 5, PI. 67, fig. 5) bear well-developed terminal teeth situated EXPLANATION OF PLATE 68 Figs. 1-5. Pattersoncypris micropapillosa sp. nov. 1. Muscle scars, female carapace, paratype lo. 4693 X282. 2. Internal view, left valve x70 [specimen lost]. 3. Branchial plate of maxilla. Paratype lo. 4714 X 203. 4. Ventral view, female carapace to show papillose ventral surface and appendages. Paratype lo. 4709 X 238. 5. Bristle bases projecting through normal pore canal openings, anterior part of female carapace, paratype To. 4721 x 2,380. Palaeontology, Vol. 15 PLATE 68 BATE, Phosphatized ostracods n Si • A V L\ ■ '"^.v -■1? ..-y 4 R. H. BATE: PHOSPHATIZED OSTRACODS 385 at the distal end of the large basal podomere, the outer edge of which is serrated (text- fig. 5), the significance of the latter being uncertain. Long setae are present on both the endopodite and the exopodite, but the exact number has not been determined. The precise number of setae is generally not visible on other appendages but they are too brittle for dissection. The maxillae (text-fig. 6, PI. 67, fig. 5) consist of an outer palp and an outer, middle, and inner masticatory process. There appears to be a minimum of 5-6 setae on each process of the maxilla. The branchial plate has been observed only in a single specimen 012 m.m. outer TEXT-FIG. 5. Left mandible, female paratype, lo. 4680. 012 m.m. TEXT-FIG. 6. Right maxilla, female paratype, lo. 4720. (PI. 68, fig. 3) where it consists of a long, narrow arm and a paddle-shaped terminal blade bearing at least 10 short spines. The upper lip is posteriorly a broad and triangular-shaped structure having a serrated edge on both sides of the centre point (PI. 70, fig. 2). Anteriorly the upper lip becomes elongate and fuses with the forehead of the animal (PI. 67, fig. 5). The first thoracic appendage is, as in other Cyprididae, adapted as a second maxilla used in feeding rather than for locomotion (text-figs, la-c, PI. 70, fig. 1). Proximally there are approximately 13 setae used to propel food towards the mouth whilst distally the appendage terminates in a strong hook in both sexes. In the male, the first thoracic appendage is slightly longer than in the female. The second thoracic appendage (text-fig. ^b) has four elongate endopodite podomeres, each of which has a distal seta. The seta of the fourth podomere, being the terminal spine or claw, has a strongly serrated or saw-toothed edge (PI. 70, fig. 3). The third thoracic appendage (text-fig. 8n) has three endopodite podomeres. The protopodite and the first and second podomeres of the endopodite possess a distal seta, whilst the third podomere of the endopodite terminates in a claw composed of 2 setae. 386 PALAEONTOLOGY, VOLUME 15 TEXT-FIG. 7. a, left first thoracic appendage, male? paratype, lo. 4719. b, left first thoracic appendage, male paratype, lo. 4682. c, left first thoracic appendage, female paratype, lo. 4681. TEXT-FIG. 8. o, right third thoracic appendage, female paratype. To. 4681 . h, right second thoracic appendage, female paratype. To. 4681. R. H. BATE: PHOSPHATIZED OSTRACODS 387 The furca (text-fig. 10, PI. 71, fig. 6) are long and appear to have simple rami, the extreme tip of which has either been broken off or is obscured. So far, however, there is no evidence of bifurcation. The testes (PI. 71, fig. 4) are coiled and situated postero-ventrally. The hemipenes (PI. 71, figs. 1-5) are large, having a broad, triangular proximal shield, and in nearly every case project below the carapace due to the relaxation of the retaining musculature on the death of the animal. 9 0 • 5 mrn. H TEXT-FIG. 9. Anterior hinge, left valve paratype, lo. 4684. TEXT-FIG. 10. Furcal rami, female paratype, lo. 4715. No eyes have been observed in this material although possibly present in the species. An eye has been included in the reconstruction of the ostracod in text-fig. 12. Dimensions (in mm) Holotype, $ lo. 4680 Paratvpes, $ lo. 4681 (J lo. 4682 $ lo. 4692 (7 lo. 4696 $ lo. 4700 juv. R.V. lo. 4702 juv. To. 4705 juv. lo. 4706 R.V. lo. 4707 3 lo. 4710 length height width 0-93 0-67 0-61 0-98 0-62 0-54 0-95 0-65 0-58 105 0-72 0-62 0-98 0-64 0-58 1-03 0-68 0-55 0-65 0-43 0-49 0-33 0-27 0-64 0-43 0-36 0-70 (broken) 0-54 1 03 0-68 0-61 388 PALAEONTOLOGY, VOLUME 15 TAXONOMIC DISCUSSION In their account of the ostracod fauna of Pernambuco and Alagoas, Brazil, Krommelbein and Weber (1972) describe a new species of freshwater ostracod, Hourcqia angiilata, of which H. angiilato salitrensis comes from the Santana Formation of Pernambuco and H. angiilata siniiata comes from the Riachuela beds, Alagoas. Hourcqia Krommelbein \965b, was first described as a monotypic genus (type species, Hourcqia africana) from the Lower Cretaceous of the Congo, West Africa. It is clearly distinguished from Pattersoucypris by the more quadrate carapace outline, absence of the diagnostic umbonate dorsal outline of the latter, possession of reversed valvular overlap — the right valve being the larger, and the development of marked sexual dimorphism of the carapace — the males clearly identified by their more elongate carapace in lateral view. All these characters are considered to set Hourcqia well apart from Pattersoucypris. The assignment of Hourcqia aiigulata to Hourcqia by Krommelbein and Weber necessitated this comparison of Hourcqia and Pattersoucypris because Hourcqia angiilata is considered here to belong more naturally to Pattersoucypris. Hourcqia angiilata sinuata is a small subspecies of angiilata coming from a different basin of deposition to the larger salitrensis, and is probably slightly younger in age (Krommelbein, personal communication). Hourcqia angiilata salitrensis, although of somewhat similar carapace outline to Pattersoucypris micropapillosa, differs significantly in size (it is larger) and in details of carapace outline as compared in text-fig. 1 \a-f. The assignation of P. micropapillosa to the Cyprididae on the fundamental structure of the carapace is confirmed by the morphology of the appendages. The additional information of the soft parts places the genus in the Cypridinae rather than in the Cypridopsinae which possess short, much-reduced furcal rami. The phylogenetic impor- tance of the appendage detail observed in P. micropapillosa lies in the fact that it shows the Cypridinae to be an extremely conservative group which has remained largely unchanged over the last 100 million years, since Lower Cretaceous times. FUNCTIONAL MORPHOLOGY OF APPENDAGES Antenmtles (text-fig. 3). The paired antennules possess 7 rather square podomeres which almost certainly bore long, swimming setae in life. Although the other appendages of this ostracod are extremely well preserved, the antennules have lost their setae through breaking oflT, probably because of their rather exposed position, projecting from the antero-dorsal part of the carapace. The length of this slender appendage suggests con- siderable flexibility in movement and strongly supports the contention that the podo- meres bore long, swimming setae and that this animal was a powerful swimmer. Anletmae (text-fig. Aa, b). The paired antennae are powerful appendages bearing 6 long, swimming setae on the inner surface, distal end, of the first podomere with additional EXPLANATION OF PLATE 69 Figs. 1-4. Pattersoucypris micropapillosa sp. nov. 1 . Paired antennules (lower part of photograph) with swimming setae present on right antennule (inside face). Paired antennae (upper part of photograph) with setae visible. Photograph taken in dorsal view, female carapace, paratype lo. 4718 x555. 2. External view, right side, juvenile carapace, paratype lo. 4705 X70. 3. Antennules, antennae (with 6 swimming setae), mandible (with basal podomere, endo-, and expodites), maxilla (with arm of branchial plate), and first thoracic appendage. Paratype lo. 4707 X 222. 4. Enlarged view of endopodite podomeres 1 to 3 of antenna. Note feathered end (lower right of photograph) of swim- ming seta. Paratype lo. 4707 x 555. Palaeontology, Vol. 15 PLATE 69 BATE, Phosphatized ostracods R. H. BATE: PHOSPHATIZED OSTRACODS 389 TEXT-FIG. 11. a, b, e, Pattersojicypris micropapillosa sp. nov. a, right view, female carapace, holotype. lo. 4680. h, e, right and dorsal views, female carapace, paratype, lo. 4692. c, d, /, Pattersoucypris angidata salitrensis (Krommelbein and Weber). Right and dorsal views, complete carapaces. 390 PALAEONTOLOGY, VOLUME 15 TEXT-FIG. 12. Reconstruction of adult female with left valve removed. The severed ends of the central muscles are shown together with the appendages and a hypothetical eye at the base of the antennule. long setae situated at the distal ends of the second and third podomeres. The podomeres are more powerfully developed than those of the antennules and the antennae would have been used not only as swimming appendages, but also as walking limbs, to pull the animal forward over the substratum. In doing this, the antennae would be projected forward and, by pressing down and pulling back the animal would, with a forward push from the 2nd thoracic appendages, be projected forward in a slightly jerky movement. This type of walking may be observed in laboratory cultures of Cypridopsis vidua (Muller) and has been clearly described and illustrated by Kesling (1951, p. 85). Mandibles (text-fig. 5). The mandibles bear the only ‘teeth’ used in feeding and are powerful appendages indicating that the animal was a scavenger, feeding on decaying plant and animal debris and possibly rasping algal growth oft' plants and stones of the EXPLANATION OF PLATE 70 Figs. 1-3. Pcittersoncypris luicropapillosa sp. nov. L Ventral view of female carapace, to show appendages. Note distal hook of lirst thoracic appendage and elongate second and third thoracic appendages. Paratype, lo. 4708 X 230. 2. Enlarged view of oral region to show serrated edge of upper lip, maxillae, and first thoracic appendages. Lower lip or hypostome situated in lower centre of photograph between paired maxillae. Paratype lo. 4708 X 570. 3. Enlarged view second thoracic appendage to show serrated edge of terminal spine or claw. Paratype lo. 4708 X 570. Palaeontology, Vol. 15 PLATE 70 BATE, Phosphatized ostracods R. H. BATE: PHOSPHATIZED OSTRACODS 391 substratum. This type of feeding has been observed in laboratory cultures of both Cypri- dopsis vidua and Heterocypris incongnieus (Ramdohr). Upper lip. The broad, rather triangular upper lip has a strongly serrated edge used to assist the movement of food into the mouth. Maxillae (text-fig. 6). The maxillae are modified appendages used to push food forward to the mouth, where it is either pushed inside or held whilst being torn by the more powerful mandibles. In addition to feeding, the maxillae have the exopodite developed into a flattened branchial plate, rapid forward and backward movement of which assists in producing a flow of water through the carapace. In this way the branchial plate assists in respiration and possibly in flushing out smaller particles of unwanted debris. In P. micropapillosa the branchial plate is long and paddle-shaped. 1st thoracic appendage (text-fig. la-c). As in other members of the family Cyprididae, the first thoracic appendages are adapted for feeding rather than being used as walking legs. The proximal end of the appendage is adapted, as are the maxillae, for pushing food towards the mouth, the ‘masticatory’ setae curving inwards to the centre of the ventral surface. The distal end of the appendage (endopodite) terminates in a strong, inwardly curved hook in both male and female dimorphs. In living species of the Cypridinae the first thoracic appendage is dimorphic, terminating distally in a strong hook only in the male as a copulatory adaptation. The male has been observed in many species to adopt a postero-dorsal position relative to the female during mating, the distal hook serving to prevent him sliding off the carapace of the female. In P. micropapillosa, however, the distal hook is present in both dimorphs and, if used in copulation suggests the mating position was not postero-dorsal. Instead a venter to venter posture is implied. The increased number of papillae on the ventral surface could be corroborative evidence for the ventral mating position, a slightly roughened surface being much easier to hold on to than a smooth one. Such a position has not yet been recorded within the Cypri- dinae, although Elofson (1941, p. 359) does record this method of mating in a number of marine cytheracean species. There is a general paucity of information on the mating habits of ostracods (McGregor and Kesling 1969) and it should not be assumed that a venter to venter position is never adopted in the Cypridinae at the present time. 2nd thoracic appendage (text-fig. 86). These paired appendages are true walking limbs and are used to assist the antennae in forward movements. The appendages are bent under the body of the animal, the distal end facing anteriorly. Pressure downwards and backwards would propel the animal forwards when used in association with the forward pull of the antennae. The second thoracic appendage is an elongate, uniramous appen- dage consisting of 4 podomeres and terminating in a long spine or claw, as such it is ideally suited to locomotion. 3rd thoracic appendage (text fig. 8n). Attached immediately behind the 2nd thoracic appendages the 3rd thoracic limbs are also uniramous, almost certainly very flexible, and used in cleaning out the inside of the carapace. The 3rd podomere terminates in a double claw used to grasp unwanted particles from inside the carapace and to eject them. It is possible that the terminal claw also assisted in locomotion by gripping weeds or the substratum. 392 PALAEONTOLOGY, VOLUME 15 Furcae (text-fig. 10). The paired furcae are long, without any observable terminal bifurcation. In fact the furca appears to be a very simple elongate structure and in length differs from the rather short furcae which characterize the Cypridopsinae. Although the appendage detail of T. micropapillosa is very close to that of Cypridopsis vidua the imme- diately noticeable differences between the two are the size of the furcae and the shape of the branchial plate. Sensory bristles. The presence of sensory bristles projecting through pores in the cara- pace is a common feature of ostracods, particularly in the anterior region. The absence of true radial pore canals and associated bristles has been compensated for in P. micro- papillosa by the development of additional normal pore canal bristles in the anterior region of the carapace. ENVIRONMENT AND PRESERVATION The excellent state of preservation of these specimens of P. micropapillosa is the result of chance mineralization under unique conditions, the phosphate material causing the mineralization being almost certainly derived from the body of the dead teleost fish, Cladocyclus gardneri Agassiz, upon which the ostracods had been feeding. Other fish- bearing nodules are crowded with the same ostracod, yet these are merely calcareous carapaces without chitinous appendages. It seems probable that the ostracods were suddenly buried, and rapidly asphyxiated, dying with their valves slightly agape. Phos- phate salts derived from the decaying fish permeated the immediate environment of the dead ostracods and a rapid replacement of the calcium carbonate and chi tin of the animals by apatite took place. Freshwater environments are often characterized by the dominance of only one or two ostracod species and in general the number of species present is limited. The numbers of individuals varies depending on food supply and the nature of the water, whether freely flowing or static. In marine environments, however, the number of species is greatly increased. In the Santana Formation only a single ostracod species has so far been observed in the calcareous nodules. This is evidence of a freshwater environment, although it might be caused by other factors also. However, without more comprehen- sive sampling it is impossible to assume that only the single species Pattersoncypris micropapillosa is present in the Santana Formation. From the proportions of juvenile instars to adults, the population of P. micropapillosa does not appear to be normal and some selection between adults and juveniles might be possible. In this context it is suggested that juvenile instars most probably lived amongst aquatic vegetation and only a very small number descended upon the dead fish; the vast majority of the ostracods EXPLANATION OF PLATE 71 Figs. 1-6. Pattersoncypris micropapillosa sp. nov. I. Ventral view, male carapace to show position of appendages and male copulatory organ. Paratype lo. 4710 x70. 2. Enlarged view male copulatory organ to show broad proximal shield. Paratype lo. 4710 x225. 3. Ventral view, male carapace, showing copulatory organ. Paratype lo. 4696 X 70. 4. Fragmentary male right valve containing testes, copulatory organ, mandible, antenna, and antennule. Paratype lo. 4711 X 110. 5. Internal view male right valve with inner lamella folded back showing attached muscles. Paratype lo. 4695 x70. 6. Ventral view of paired furcal rami, female carapace, paratype lo. 4713 x225. Palaeontology, Vol. 15 PLATE 71 BATE, Phosphatized ostracods R. H. BATE: PHOSPHATIZED OSTRACODS 393 present being either eggs or adults. Alternatively, the relationship may result from seasonal fluctuation of instars leaving only a primarily adult population with a subordi- nate number of juveniles and a large number of eggs. A close examination of a much larger number of nodules would be needed to produce more evidence on these relation- ships, as it cannot be suggested that every nodule represents a group of ostracods feeding on a dead fish. Indeed, it is most probable that the present nodule was the only instance so far found where this was so, and represents an important factor in the preservation of the ostracods by mineral replacement. A second nodule containing specimens of P. nucropapillosa, similarly mineralized, has been found since this paper was written. The specimens, however, lacked appendages. Acknowledgements. 1 am indebted to my colleague. Dr. Colin Patterson, for drawing my attention to this important ostracod material. 1 should also like to record my thanks to Dr. R. Davies of the Mineralogy Department, for kindly undertaking the X-ray powder analysis of the specimens. Dr. K. Krommelbein (Kiel) assisted greatly in sending specimens of Honrccjia angulata salitrensis, thereby permitting a much closer comparison of these related ostracods. REFERENCES BATE, R. H. 1971. Phosphatized ostracods from the Cretaceous of Brazil. Nature, London, 230, 397-398. BEURLEN, K. 1970. Geologie von Brasilien, viii + 444 pp. Berlin. BRONGNiART, M. c. 1876. Note sur un nouveau genre d’entomostrace fossile provenant du terrain carbonifere des environs de Saint-Etienne. Ann. sd. Geol. Paris, 7, 1-6, pi. 6. EAGAR, s. H. 1970. A Pleistocene mummified ostracod from the Wairarapa District. N.Z. Jl. mar. fresliw. Res. 3, 607-609. ELOFSON, o. 1941. Zur Kenntnis der marinen Ostracoden Schwedens mit besonderer Beriicksichtigung des Skageraks. Uppsala Univ. Zool. Bigrag. 19, 215-534. GOCHT, H., andcoERLiCH, F. 1957. Reste des Chitin-Skelettes in fossilen Ostracoden-Gehausen. Geol. Jb. 13, 205-214, pi. 14. GRAMANN, F. 1962. Extremitatenfuiide an liassischen Bairdien (Ostracoda). Paldont. Z. 36, 28-32. RESTING, R. V. 1951. The morphology of ostracod molt stages. Illinois Biol. Monogr. 21, 1-324, pis. 1-96. KROMMELBEIN, K. 1965fl. Ncue, fur Vergleiche mit West-Afrika wichtige, Ostracoden-Arten der brasilianischen Bahia-Series (Ober-Jura?/Unter-Kreide in Wealden-Fazies). Senck. letli. 46a, 177- 213, pis. 11-15. 1965fi. Ostracoden aus der nicht-marinen Unter-Kreide (Westafrikanischer Wealden) des Congo- Kiistenbeckens. Meyniana, 15, 59-74, pis. 1-4. and WEBER, R. 1972 [in press]. MCGREGOR, D. L., and KESLiNG, R. V. 1969. Copulatory adaptations in ostracods. Part II. Adaptations in living ostracods. Contrib. Mas. Paleontol. Univ. Michigan, 22, 221-239. SANTOS, R. DA s., and VALENyA, J. G. 1968. A Formagao Santana e sua Paleoictifauna. Anais Acad. bras. Cienc., Rio de Janeiro, 40, 339-360. SCHMIDT, R. A. M., and SELLMANN, p. V. 1966. Mummified ostracods in Alaska. Science, 153, 167. SYLVESTER-BRADLEY, p. c. 1941. The Shell Structure of the Ostracoda and its application to their palaeontological investigation. Ann. Mag. nat. Hist. 8, 1-33. R. H. BATE Department of Palaeontology British Museum (Natural Elistory) Cromwell Road London, S.W. 7 Final typescript received 4 November 1971 A TEXANITES-SPINAPTYCHUS ASSOCIATION FROM THE UPPER CRETACEOUS OF ZULULAND by w. j, KENNEDY and h. c. klinger Abstract. Two specimens of the ammonite genus Texanites from the Santonian sediments of False Bay, Zululand, have aptychi, identified as Spinaptychus spimsus (Cox), in their body chambers. This association appears to be a valid one, and is compatible with previous records of Spinaptychus from Texanites-hcdix'ing strata in England, North Africa, the Middle East, and North America, although no previous associations have been described. The aptychus form-genus Spinaptychus Traiith (type species Aptychus spinosus Cox 1926, 577, pi. 24, figs. 1-3) is one of the most distinctive and bizarre ammonite operculae known. Representatives of this genus are recorded from the Upper Cretaceous of Great Britain (Cox 1926), North Africa (Collignon 1966), the Middle East (Picard 1926, Trauth 1930), and North America (Fischer and Fay 1953), but in no case has the ammonite genus or genera with which Spinaptychus is associated been proven, although texanitids have been suggested by Cox (Arkell et al. 1957, p. 440). Our independent discovery of aptychi of this type associated with the texanitid genus Texanites sensu stricto in the Santonian of Zululand, South Africa, is therefore of interest. These occurrences are described below, and previous records and inferred associations are discussed. It is not our intention to discuss here the functional aspects and interpretation of aptychi. Some uncalcified forms, in particular anaptychi, are now known to be parts of the ammonite jaw apparatus (Fehmann 1967, 1970, 1971, Closs 1967). Calcareous forms including Spinaptychus still seem best regarded as operculae, the classic view (Trauth 1927 onwards, Arkell et al. 1957). Occurrence. Both specimens described here come from Santonian silts exposed on the eastern shores of False Bay, WNW. of Hluhluwe, northern Zululand. BMNH C76773 comes from foreshore exposures with abundant texanitids at the southern end of Die Rooiwalle, 51 km north of Lister’s Point. S.A. Z2200 comes from foreshore exposures of a comparable horizon 2-2 km north of Picnic Point. DESCRIPTION OF THE SPECIMENS The British Museum Material, BMNH C76773. This specimen (Pis. 72-73) consists of almost two-thirds of a whorl of body chamber of a texanitid ammonite, 240 mm long, with a maximum whorl height of 90 mm. On one side the ornament is well preserved; the other is abraded as is much of the venter. The specimen is an internal mould; only traces of iridescent, nacreous test are pre- served (PI. 72, fig. 1 ; PI. 73, fig. \a). The mould itself is of buff- weathering grey-blue calcareous siltstone, with scattered shell fragments. The coiling of the ammonite was very evolute, with a compressed whorl section, the greatest breadth being at the lower lateral tubercle. The intercostal section is oval, with [Palaeontology, Vol. 15, Part 3, 1972, pp. 394-399, pis. 72-73.] KENNEDY AND KLINGER: SPINAPTYCHUS 395 a distinct rounded keel; the costal whorl-section is polygonal. The umbilical wall is steep, the umbilical shoulder sharply rounded, the flanks gently rounded and converging towards the venter. There are eighteen ribs on the two-thirds whorl remaining (PI. 72, fig. 1). They arise as faint broad undulations on the umbilical wall, but on the shoulder, strengthen into distinct, umbilical bullae. From these bullae extend strong, straight, rounded ribs, which are narrower than the interspaces. They bear a weak, radially elongate lower lateral tubercle, a somewhat stronger, rounded upper lateral tubercle, and an even stronger lower ventrolateral tubercle. The ribs broaden at the lower ventro-lateral tubercle, and connect to well developed, clavate upper ventro-lateral tubercles which are separated by a smooth band from the strong, rounded siphonal keel. Most of the ribs are simple, but towards the aperture, one of the umbilical bullae gives rise to a pair of ribs. The Spinaptychus lies in the ventral part of the body chamber close to the inferred position of the aperture (PI. 73, figs. \a-b), resting upon the lower inner surface of the wall of the body chamber of the shell as it was entombed (PI. 72, fig. 3). The harmonic margin (symphysis) lies parallel to, but about 10 mm from, the ventral keel of the body chamber (PL 72, fig. 3; PI. 73, fig. la), while the outer margin faces towards the aperture. The two valves are slightly separated, but in a closed position (PI. 72, fig. 3; PI. 73, figs, la, 2), with the harmonic margins displaced laterally by over 10 mm, and the lateral margins almost touching at one point (PI. 72, fig. 3). As found, the outer surface of the left valve was almost completely exposed (PI. 73, figs, la-b), with the distinctive ornament visible, although somewhat abraded (PI. 73, fig. lb, lower part of figure). Mechanical preparation exposed the inner surface of the right valve, and produced material suitable for thin section study, leaving an internal mould attached to the same block as the left valve (PI. 72, figs. 2-3; PI. 73, fig. 2). The total length of the valves is estimated as from 82 to 85 mm, and the greatest thickness noted is 0-8 mm. The ornament of the inner surface of the valves is typical for Spinaptychus', discrete, separate, spiny, sometimes perforate protuberances from 0-70 to 1 -00 mm in diameter, with a maximum preserved height of 1-50 mm and a perforation diameter of 0-25 mm (PI. 73, figs. la-b). The protuberances are rather irregularly disposed, but with a sug- gestion of concentric arrangement. The outer surface of the valves is ornamented by concentric growth striae of varying strengths, parallel or sub-parallel to the lateral margin. Major striae are spaced at intervals of 2 to 3 mm; between are much finer striae, approximately 5 per mm. This ornament is crossed by fine, but nevertheless distinct striae [as noted by Cox (1926), but doubted by Trauth (1928, p. 131)], running normal to the growth lines (PI. 72, fig. 2, lower right- hand part of figure). Macroscopic examination indicated that the specimen consisted of coarsely crystalline calcite mosaic with prismatic blocks normal to the surfaces of the aptychus. Thin sections show traces of internal structure; fine growth layers of varying magnitude inclined at an acute angle to the surface of the aptychus. These are prominent on the inner part of the specimen, but only traces remain in the outer parts. The structure thus corresponds to that given by Fischer and Fay (1953) rather than the tripartite division seen by Cox (1926). D d C9016 396 PALAEONTOLOGY, VOLUME 15 Discussion. The ornament of the body chamber indicates that the ammonite belongs to the genus Texanites sensu stricto, and to the group of Texanites soutoui (Baily) (Spath 1921, 1922), as is clear from a comparison with the type specimen, BMNH C47261. The aptychus is, without a doubt, a Spinaptychus, the described species of which are separated on minor details of ornament, as follows: 1. Spinaptychus spinosus Cox (1926, p. 577, pi. 24, figs. 1-3; Trauth 1927, pp. 193, 200, 220, 232, 244; 1928, p. 131, pi. 3, figs. 17-18; 1930, p. 339; Collignon 1966, p. 51, pi. 21, figs. 2-4). This species has a maximum length of 65 mm (Cox 1926) to 70 mm (Collignon 1966), and is characterized by an essentially random distribution of protuberances. 2. Spinaptychus picardi Trauth (1930, p. 340, pi. 5, fig. 19; Picard 1929, pp. 434, 435, 436, 455, pi. 9, fig. 1). This species is up to 55 mm long, and is characterized by a distinctive concentric arrangement of protuberances. 3. Spinaptychus perlatus Trauth (ex Fraas MS.) (1930, p. 341, pi. 5, fig. 20) is a somewhat smaller species — 45 mm long, with concentric ornament on the central area and irregular ornament in marginal regions. 4. Spinaptychus sternbergi Fischer and Fay (1953, pp. 77-92, pis. 1-2, fig. 1) is a gigantic species up to 170 mm long, characterized by a strongly aligned concentric ornament of rather distant protuberances, tubercles large and perforate in the central area of the aptychus, but smaller and non-perforate towards the margins. These differences are all rather trivial. Examination of the type series of Spinaptychus spinosus (BMNH 46576, 46770, 46772, 47719, 48077-48078, 62172, 70391, 70545, C3088-3092, 48741, 73994, C27299) shows that there is great variation in this ‘species’, perhaps as great as that seen in all the forms noted above. Small protuberances at the margins of specimens are imperforate; with growth, perforations appear. In some specimens (i.e. BMNH 46772) there is a striking concentric ornament, in others the ornament is distant, but on central parts of the test (i.e. Cox 1926, pi. 24, fig. 1), deposition of calcium carbonate gives a crowded and very irregular appearance to the surface. There is even a magnificent specimen (BMNH C32521, Uintacrinus Zone (Santonian), Margate, Kent, ex Rowe Collection) which is comparable to Spinaptychus sternbergi. We would thus refer our specimen with confidence to S. spinosus. EXPLANATION OF PLATE 72 Figs. 1-3. Texanites-Spiliaptyclius association. 1. Lateral view of the body chamber fragment of Texanites soutoni (Baily), BMNH C76773. Reduced xO-67. 2. The Spinaptychus in position, prior to development, viewed from behind. The positions of the left valve (LV) and right valve (RV) is indicated, x 1 . 3. Lateral view of the internal mould of the right valve, after development. Note the concentric and transverse striae. X 1. All specimens are coated with ammonium chloride. EXPLANATION OF PLATE 73 Figs. 1-2. Spinaptychus. la-l/>. The prior to development. Fig. lu shows the specimen in position; the inner face of the left valve (LV) is exposed, traces of the right valve (RV) are visible. Reduced xO-67. Fig. \b is a detail, showing the surface ornament X I -5. BMNH C76773. 2. The specimen after development; view showing the symphysis, the left valve with test, and the internal mould of the right valve. X I. All specimens are coated with ammonium chloride. Palaeontology, Vol. 15 PLATE 72 KENNEDY and KLINGER, Spinaptychus Palaeontology, Vol. 15 PLATE 73 KENNEDY and KLINGER, Spinaptychus \ KENNEDY AND KLINGER: SPINAPTYCHUS 397 It remains to confirm that the Spinaptychus really belongs to the Texanites soutoni it was buried with. Those parts of the body chamber in front of the aptychus are crowded with shell debris, while areas behind contain only a few fragments mostly near to the aperture; though one texanitid fragment 50 mm across lies behind the aptychus. The adapical portion of the body chamber contains a peculiar opaque black object lying some 8 mm from the ventral margin. Up to 5 mm across, and 18-5 mm long, tapering adaperturally, this structure bears some resemblance to the ink-sacs recently described by Lehmann (1967) and others in Jurassic and Cretaceous ammonites. How- ever, it might equally well be a burrow fill! The shell material in the body chamber suggests that either the aptychus was present all the time, and stopped most shell fragments getting into the body chamber, or that it was one of the earliest objects washed in. The Pretoria Specimen. S.A. Geological Survey, Z2200. This specimen was collected by the late E. C. N. Van Hoepen, and it was only during curation that traces of an aptychus were discovered. This second aptychus, like the British Museum specimen, lay in the body chamber, but at the adapical end, 70 mm from the final septum. The harmonic margin faced ventrally, with the inner margin facing the aperture. Both ammonite and aptychus are poorly preserved, although the one is definitely a Texanites and the other a Spinaptychus. The Pretoria specimen is to be further described by Klinger; it gives added weight to the validity of the Texanites-Spinaptychus association. One occurrence could well be chance; two seem less likely to be so. DISCUSSION The Santonian sequence yielding our specimens contains a range of texanitids, with both Texanites itself and Menabites species; other genera are Baculites, Scaphites, diplomoceratids, Hauericeras, Pseudoschloenbachia, and pachydiscids. All save the pachydiscids can be ruled out immediately as bearing Spinaptychus on the basis of morphology, while the known pachydiscid association with Pseudostriaptychus rules out this last contender. Only Texanites and Menabites remain. We therefore believe that Spinaptychus is the aptychus of Texanites', our specimens are of Santonian age. How does this conclusion match with previous records? The English Chalk. Spinoseaptychi were first recorded by Blackmore(1896, pp. 531-532), who believed that what is clearly Spinaptychus spinosus belonged to Parapuzosia lepto- phylla (Sharpe). Subsequently, Cox (1926) described and recorded the species from the Upper Chalk (Santonian, Micraster coranguinwn Zone) of Kent, Surrey, and Wiltshire. The only ammonites known at this level are Parapuzosia and Texanites (Wright and Wright 1951) and Cox (1926, p. 580) successfully rules out Parapuzosia on morpho- logical grounds. North Africa. Collignon (1966) recorded Spinaptychus spinosus from horizons above beds with Gauthiericeras margae (Schliiter), which he nevertheless regarded as Coniacian. 398 PALAEONTOLOGY, VOLUME 15 No ammonites occur with the Spinaptychus to date them precisely, so that the record is somewhat unsatisfactory. Texaniles is well known in North Africa. Middle East. Te.xanites is well known in this region. Picard (1929) recorded Spinaptychus from relict Senonian deposits on the Jordan plain west of Jericho. The age of the specimen was said to be Campanian, but there is no good evidence for this. Trauth (1930) recorded a further specimen from Alma, Syria, but could date it no closer than ‘Senonian’. North America. Fischer and Fay (1953) recorded many specimens of Spinaptychus sternbergi from the Upper Smokey Hill Member of the Niobrara Formation, in Kansas, and the species has been subsequently recorded by other workers. According to Miller (1968) these occurrences are of Upper Santonian age. Texanites is well known in the Santonian rocks of this area. CONCLUSIONS 1. The two specimens of Texanites and Spinaptychus from the Santonian of Zululand appear to be a real association. 2. Spinaptychus is known definitely from the Santonian of Britain, South Africa, and North America. Records at lower and higher levels in the Upper Cretaceous are dubious. 3. The known geographical distributions of Spinaptychus is within that known for Texanites (Collignon 1948) during the Santonian. Acknowledgements. We are indebted to Dr. H. W. Ball, Dr. E. G. Kauffman, Dr. M. K. Howarth, Mr. D. Phillips, Mr. P. J. Rossouw, Dr. W. S. McKerrow, and Dr. B. W. Sellwood for their help and assistance in many ways during the preparation of this paper. Klinger’s contribution published under copyright authority 4591 (19.10.1971) of the Government Printer of the Republic of South Africa. REFERENCES ARKELL, w. J., et at. 1957. Cephalopoda Ammonoidea. In Moore, R. C., ed.. Treatise on invertebrate palaeontology. Part L4. Kansas. BLACKMORE, H. p. 1896. Some notes on the aptychi from the Upper Chalk. Geol. Mag., n.s. 4 (3), 529- 533, pi. 16. CLOSS, D. 1967. Goniatiten mit Radula und Kieferapparat in der Itarare Formation von Uruguay. Paldont. Z. 41, 19-37, pis. 1-3. COLLIGNON, M. 1948. Ammonites neocretaces du Menabe, Madagascar I, Les Texanitidae. Ann. geol. Serv. min. Madagascar, 14, 101 pp. pis. 15-32. 1966. Les Cephalopodes cretaces du bassin cotier de Tarfaya. Notes mem. Serv. mines geol. Maroc. 175, 7-149, 35 pis. cox, L. R. 1926. Aptvclins spinosiis sp. n. from the Upper Chalk. Ann. Mag. nat. Hist. 9(17), 573-580, pi. 24. FISCHER, A. G., and FAY, R. o. 1953. A spiny aptychus from the Cretaceous of Kansas. Ball. State geol. Siirv., Kansas, 102, 77-92, pis. 1-2. LEHMANN, u. 1967. Ammoiiiten mit Kieferapparat und Radula aus Lias-Geschieben. Paldont. Z. 41, 38-45, pi. 4. 1970. New aspects in ammonite biology. Proc. North American Palaeontological Convention, Part 1, 1251- 1971. Jaws, radula and crop of Arnioceras (Ammonoidea). Palaeontology, 14, 338-341. MILLER, H. 1968. Invertebrate fauna and environment of deposition of the Niobrara Formation (Cretaceous) of Kansas. Fort Hayes Studies, 8, 90 pp., 9 pis. KENNEDY AND KLINGER: SPINAPTYCHUS 399 PICARD, L. 1929. On Upper Cretaceous (chiefly Maestrichtian) Ammonoidea from Palestine. Ami. Mag. nat. Hist. 10 (3), 433-456, pis. 9-10. SPATH, L. F. 1921. On Cretaceous Cephalopoda from Zululand. Ann. S. Afr. Mas. 12, 217-231, pis. 19-26. 1922. On the Senonian Ammonite fauna of Pondoland. Trans. R. Soc. S. Afr. 10, 113-147, pis. 5-9. TRAUTH, F. 1927. Aptychenstudien. I, Uber die Aptychen in allgemeinen. Annin naturli. Mas. Wien, 41, 171-259. 1928. Aptychenstudien. 11, Die Aptychen der Oberkreide. Ibid. 42, 121-193, pis. 2-4. 1930. Aptychenstudien. Ill-V. ibid. 44, 329-411, pis. 3-5. WRIGHT, c. w., and wright, e. v. 1951. A survey of the fossil Cephalopoda of the Chalk of Great Britain. Palaeontogr. Soc. (Monogr.), 1-40. W. J. KENNEDY Department of Geology and Mineralogy Parks Road Oxford, 0X1 3PR H. C. KLINGER South African Geological Survey Private Bag Xii2 Pretoria Final typescript received 25 September 1971 South Africa THE AFFINITIES OF IDIOHAMITES ELLIPTICOIDES SPATH (CRETACEOUS AMMONOIDEA) by W. J. KENNEDY Abstract. Idiohamites ellipticoides Spath is an aberrant species of the genus Idiohamites which shows the recoiling trend so widespread in Cretaceous heteromorph ammonites. Its ornament is identical with some members of the largely southern hemisphere family Labeceratidae, but study of suture lines show this to be merely a case of homeomorphy. The species is described and illustrated on the basis of large new collections. The species Idiohamites ellipticoides was described by L. F. Spath in 1939 on the basis of six fragments from the Upper Albian Mortoniceras inflatiim Zone, Hysteroceras varicoswn Sub-zone, Gault Clay of Kent. Since that time the species has received no notice from other workers and remains poorly known. In 1964 I collected a series of specimens from the varicoswn Sub-zone clays exposed at the Associated Portland Cement Manufacturer’s clay pit at Paddlesworth, Kent, and have subsequently seen many other specimens from this locality. The new material proved to be strikingly similar to some members of the heteromorph family Labeceratidae, best known from the southern hemisphere. Because of a lack of comparative material I was unable to confirm this suspicion, but large collections of Albian labeceratids made in South Africa during 1970 and new material from Australia now show that Idiohamites ellipticoides is, in fact, a labeceratid homeomorph. The species is therefore redescribed and discussed below, on the basis of Spath’s types and the new material. SYSTEMATIC DESCRIPTION Idiohamites ellipticoides Spath Plate 74, figs. 1-10, PI. 75, figs. 1-6, 9, 11, ?7a-c, ?10a-c 1939 Idiohamites ellipticoides Spath, p. 594, text-iig. 213, pi. 65, fig. 9. Holotype. BMNH C39746 from the Gault Clay of Folkestone, Kent. Other material. Several specimens from Folkestone and Merstham, Kent, mentioned by Spath (1939, p. 594), scores of specimens from the Hysteroceras varicosum Sub-zone Gault Clay, Associated Portland CenTent Co. clay pit, Paddlesworth, near Snodland, Kent (National Grid Reference 51/690620) in the collections of M. K. Durkin and J. D. Hollis. A series of specimens is deposited in the British Museum (BMNH C76840-C76857). Description. The coiling is rather tight for Idiohamites, the early whorls forming a loose planispiral. The body ehamber appears to have been straightened, but not, so far as is known, recurved into a erozier. The whorl section is compressed (whorl breadth; whorl height varies from 0-54 to 0-44) with rounded venter and dorsum, and battened sides. [Palaeontology, Vol. 15, Part 3, 1972, pp. 400-404, pis. 74—75.] W. J. KENNEDY: IDIOHAMITES 401 Ornament consists of primary ribs which arise at the base of the flanks, and branch at a point above mid-flank. At the same point, additional, shorter intercalated ribs may be inserted. The ribs are continuous across the venter; the dorsum is smooth. The strength, nature, degree of flexion, point and number of branches of the ribs is highly variable. I have tried to illustrate variation in Plate 74; figures \a-c to 5a-d show variation in ornament at comparable sizes. Figures 6a~d to lOa-d show changes in ornament through ontogeny. These should be compared with the holotype (figured on Plate 75 as figures 2a-b) and a series of larger specimens figured on the same Plate (figures 3a-c, 4a-d, 5a-d). Normally there are from five to six primary ribs in a distance equal to the whorl height. When branching is obvious, it is normally into two (i.e. Plate 74, fig. 8), but sometimes into three (i.e. Plate 74, fig. 6). Addition of intercalated ribs in these forms (i.e. Plate 74, fig. 6) is irregular, and results in their being 2 to 2-5 times as many secondary + intercalated ribs as primaries. It is of course difficult to draw a line between intercalated short ribs and branched primaries; the two grade into each other in a single specimen and between specimens (Plate 74, figs. 1, 2, 6). A further variation in ornament is seen in some of the more compressed forms, where fine striae appear between ribs; these seem to be little more than strengthened growth lines. The species thus, to a degree, follows Buckman’s first law of covariance. Tracing changes with ontogeny (Plate 74, figs. 6a~d to \0a-d) again reveals much variation. At the smallest diameters studied (whorl height = T5 mm; Plate 75, fig. 2a-b), the shell is almost smooth. With growth, ribs appear, as variable in their develop- ment as on larger specimens. The largest specimens available (Plate 74, fig. 6a-d‘, Plate 75, figs. 4a-d, 5a-d) show an increase in the number of secondary and intercalated ribs, and frequent incipient triplication. The holotype of Idiohamites ellipticoides is quite without the ventral tubercles so typical of Idiohamites, as are most other specimens. One individual which does have such tubercles is shown as Plate 74, fig. 9a-c; its ornament is atypical in some other respects, and it may belong to another species. The suture-line of /. ellipticoides is rather simple (perhaps in part a reflection of small- size) and little subdivided (text-figure 1, figs. 1, 2a-d, 4). The saddles are small, and the lobes are distinctly bifid. In these respects there is agreement with other Idiohamites at a comparable size (i.e. Spath 1939, text-figures 206, 209, 210). Discussion. Had this material been available to Spath in 1939 he would undoubtedly have divided it into a series of varieties, or even species. The presence of intermediate morphologies between the more extreme variants, as shown in the Plates, suggests a single population to me, especially in view of the widespread recognition of comparable variation in other Cretaceous ammonite groups (Reeside and Cobban 1960, Kennedy and Hancock 1970, Wiedmann 1969, for example), and the known highly variable nature of heteromorph species (Wiedmann and Dieni 1968, with discussion). The style of ribbing, whorl section, and coiling of Idiohamites ellipticoides dis- tinguishes it from all other Gault Clay Idiohamites and Hamites species, and the dubious ‘'Hamitoides' rusticus Spath. It thus stands unique amongst described Gault heteromorphs. 402 PALAEONTOLOGY, VOLUME 15 CONCLUSIONS Because of its peculiar ornament Idioliamites ellipticoides presents problems of affinity and particularly so as it has been referred to the Laberatidae (Owen 1970). There are in fact three groups to which the species shows resemblance, the Laberceratidae, and the genera Algerites and Idioliamites. The Laberceratidae are a family known exclusively from the southern hemisphere, save for Hamitoides, whose true affinities are in fact dubious (Spath 1939, Wright 1957). Three genera are definitely placed in the family : Labeceras, Myloceras, and Ellipsoceras. They all show an initial open spiral, parts of which may in fact touch, and some have a hooked body chamber, often lappeted. The ornament is highly distinctive, with striking branched ribs. In this respect they find an exact match in /. ellipticoides', L. {Labeceras) is the most similar genus, as can be seen from species figured by Whitehouse (1926), Spath (1925), and Collignon (1932, 1950, 1963). Plate 75, fig. ?>a-b shows a typical South African Labeceras for comparison. When the suture-lines of labeceratids are compared with those of I. ellipticoides, striking differences are apparent. The suture of /. ellipticoides has bifid lobes and narrow saddles; all labeceratids have trifid lobes and broad saddles, as can be seen from Text- figure 1. These differences are apparent even in early stages, as is clear from a series of juvenile labeceratids in the British Museum (BMNH C7678 1-76782, C76797). Algerites (Pervinquiere 1910) is a Cenomanian form, known only from Algeria. Its coiling is reminiscent of /. ellipticoides in some respects, but later whorls come to touch completely. In Algerites the suture-line is closely comparable to that of /. ellipticoides, and indeed other Idioliamites. Ornament links it to Idioliamites — close-spaced tubercu- late ribs, which thus differ markedly from I. ellipticoides. The coiling and suture-line of I. ellipticoides can be closely matched with Idioliamites species; in particular there is a close resemblance to I. incertus Spath. The ornament of Idioliamites is usually quite distinctive, however, with simple tuberculate ribs. Conclusions about the affinity of Idioliamites ellipticoides thus rest on three criteria: ornament, coiling, and suture-line. Branched ribbing occurs widely in Gault heteromorphs; Hamites shows branched ribs not uncommonly on curved parts of the shell, and some specimens of "Hamitoides’' EXPLANATION OF PLATE 74 Figs, la-c to lOtw/. Idioliamites ellipticoides Spath from the Hysteroceras varicosimi Sub-zone Gault Clay, Paddlesworth, Kent. BMNH C76848 to C76857. Specimens are figured natural size and x 2. EXPLANATION OF PLATE 75 Figs. \a-d,Aa-d\.o6a-c,9a-d, \ \a-d, Ila-c, ’’lOa-c. Idioliamites ellipticoidesS^tiX.hAvomihQ Hysteio- ceras varicosimi Sub-zone Gault Clay, Paddlesworth, Kent. BMNH C76840 to C76847. Fig. 2a-b. The holotype of /. ellipticoides, BMNH C39746, from the H. varicosimi Sub-zone Gault Clay, Folkestone, Kent. Fig. 3a-c. Idioliamites ellipticoides, body chamber fragment from the Upper Gault of Merstham, Surrey. BMNH C39942, figured by Spath 1939, pi. 65, fig. 9a-b. Fig. 8rt-/x L. (Labeceras) plasticiim Spath, specimen from the lower part of the Upper Albian, Munyuwana Creek, north-east of Hluwluwle, Zululand, South Africa. BMNH C76858. All specimens are figured natural size, and X 2. Palaeontology, Vol. 15 PLATE 74 10a 10b KENNEDY, Jdiohamites PLATE 75 Palaeontology, Vol. 15 KENNEDY, Idiohamites W. J. KENNEDY; IDIOHAMITES 403 TEXT-FIG. 1. Suture-lines of Idiohamites ellipticoides and various labeceratids compared. 1, 2, 4. /. ellipticoides. 1, the holotype, BMNH C39746, X 15. 2, BMNH C76845, x 12. 4, BMNH C76853, x 12. 3. Labeceras papiilatiini Whitehouse, X 3. 5. Mylocercis aff. cornucopia Spath, X 5. 6. Labeceras sp., X 5. 7. Labeceras corupressum Whitehouse, X 4. 8. Labeceras plasticuin S’pCiXh, x4. 9. Mylocercis serotiuum rugosa Spath, X 5. 10. M. serotiuum plana Spath, X 5. (3, 7 from Whitehouse 1926; 5, 6, 8, 9, 10 from Spath 1925; 1 from Spath 1939.) 404 PALAEONTOLOGY, VOLUME 15 represent nothing more than such fragments (Spath 1939, pi. 66, figs. 3-5). Some Idiohamites species show a precisely similar change in rib style when curved (Spath 1939, p. 593, pi. 65, fig. 6). Furthermore, we know that there is a repeated trend towards re-coiling in heteromorphs (Wiedmann 1969). Dismissing these two features, one is left with the sutural development, and this places /. ellipticoides close to true Idiohamites. The species represents a member of the group which is tending towards close coiling also seen in I. incertus, and at an extreme in Algerites. Branched ribbing is stabilized in the species, which is thus a homeomorph of the contemporary labeceratids, with which it has no close affinity. One might reasonably argue that the peculiar ornament of Idiohamites ellipticoides, plus lack of tubercles, merits its separation into a separate genus or subgenus of Aniso- ceratidae. To do this on the basis of a single species, whose precise phylogenetic position remains unknown, would be unwise at this time. Acknowledgements. My best thanks are due to Dr. J. Wiedmann (Tubingen) for his invaluable comments on the material described. Dr. M. K. Howarth, Mr. D. Phillips, Dr. H. G. Owen, Mr. J. Hollis, Mr. M. Durkin, Mr. C. W. Wright, Dr. J. M. Hancock, and Dr. W. S. McKerrow all aided me in various ways; their help is gratefully acknowledged. REFERENCES COLLIGNON, M. 1932. Les ammonites pyriteuses de FAlbien superieur du Mt. Raynaud a Madagascar. Ann. geol. Serv. min. Madagascar, 2, 5-36, pis. 1-4. 1950. Recherches sur les faunes albiennes de Madagascar. Ibid. 12, 7-85, 14 pis. 1963. Atlas des fossiles caracteristiques de Madagascar (Ammonites) X, Albien, xvA183 pp., pis. 241-317. Service geologique, Tananarive. KENNEDY, w. J., and HANCOCK, J. M. 1970. Amiiionites of the genus Acanthoceras from the Cenomanian of Rouen, France. Palaeontology, 13, 462-490, pis. 88-97. OWEN, H. G. 1970. The stratigraphy of the Gault in the Thames Estuary and its bearing on the Mesozoic tectonic history of the area. Proc. Geol. Ass. 82, 187-207. PERViNQUiERE, L. 1910. Sur quelques ammonites du cretace algerien. Meni. Soc. geol.fr. 42, 86 pp., 7 pis. REESiDE, J. B., and COBBAN, w. A. 1960. Studies of the Mowry Shale (Cretaceous) and contemporary formations in the United States and Canada. U.S. Geol. Siirv. Prof. Paper 355, 126 pp., 58 pis. SPATH, L. F. 1925. On Upper Albian Ammonoidea from Portuguese East Africa. With an appendix on Upper Cretaceous Ammonites from Maputoland. Ann. Trans. Mas. Pretoria, 11, 179-200, pis. 28-37. 1939. A monograph of the Ammonoidea of the Gault. Part 13, pp. 541-608, pis. 59-64. Palaeont. Soc. (Monogr.). WHiTEHOUSE, F. w. 1926. The Cretaceous Ammonoidea of eastern Australia. Mem. Queensland Mas. 8, 195-242, pis. 34-41. WIEDMANN, J. 1969. The heteromorphs and ammonoid extinction. Biol. Rev. 44, 563-602. and DiENi, I. 1968. Die Kreide Sardiniens und ihre Cephalopoden. Pal. Ital. 34, 1-171, 18 pis. WRIGHT, c. w. 1957. In moore, r. c. (editor). Treatise on Invertebrate Palaeontology, Part L, 4, L80-490, Kansas. W. ,1. KENNEDY Department of Geology and Mineralogy Parks Road Oxford, 0X1 3PR Final typescript received 18 November 1971 THE SYSTEMATIC POSITION OF THE JURASSIC BRACHIOPOD CADOMELLA by C. H. C. BRUNTON aild D. L. MACKINNON Abstract. Shell structural and morphological studies of Cadomella and various species of the Spiriferida and Chonetacea show that Cadomella belongs to the Koninckinacea within the Spiriferida. The term supra-apical is considered inappropriate for the pedicle foramen of the Koninckinacea. The relationship between pseudo- punctae and tubercles is discussed. Working independently (MacKinnon on the Spiriferida and Brunton on the Choneti- dina) we reached the same conclusions as to the systematic position of the Koninckinacea and it seemed sensible, therefore, to combine our information into one paper. Our results show that Cadomella should be retained within the Koninckinacea and that this superfamily belongs to the Spiriferida. Unless otherwise stated the classification here employed follows the brachiopod ‘Treatise’ (Muir-Wood 1965). During the past ten years the Lower Jurassic genus Cadomella, the sole member of the Cadomellacea (Muir-Wood 1955, p. 90) has received attention because of its unusual combination of morphology and shell structure; features which have resulted in its being classified within either the Strophomenida or Spiriferida. The species assigned to Cadomella, C. moorei (Davidson), the type species, C. quenstedti Ran and C. davidsoni (Eudes-Deslongchamps) were originally described as strophomenaceans. Muir-Wood (1955) wrote of strophomenide characteristics, including a supra-apical foramen, in Cadomella, and in 1962, in her monograph entitled ‘Chonetoidea’, she included the Cadomellacea in her new classification of the Chonetidina. She wrote (1962, p. 30) that the ‘Cadomellacea are dealt with elsewhere’, unfortunately without actually having done so. It seems, however, that she may have thought that genera other than Cadomella alone were involved, for she stated that the geological range was ‘ ? Trias to Upper Lias’ in the familial diagnosis. In the brachiopod Treatise Muir-Wood (1965) followed her 1962 classification, retaining the Cadomellacea in the Chonetidina, but only Cadomella was included and the range was limited to Lower Jurassic. The diagnosis includes mention of a functional supra-apical pedicle foramen, lack of hinge spines and a lamellar, fibrous, and internally pseudopunctate shell structure. Boucot, Johnson, and Staton (1965), in the Treatise, placed the Triassic and Jurassic Koninckinacea within the athyrididinid Spiriferida. Previously constituent genera had commonly been assigned to the Strophomenida. Cowen and Rudwick (1966) described the discovery of a spiral brachidium in Cadomella davidsoni from the Upper Lias of Normandy. They recognized the similarity between Cadomella and the Koninckinacea, but placed this superfamily within the Chonetacea. This assignment of the Koninckina- cea to the Strophomenida was made in the belief that foramina were supra-apical and on the basis of morphological features such as profile, strophic hinge, and convex pseudodeltidium and notothyrium. In support of the implication that in this group a spiral brachidium must have evolved independently from that of the Spiriferida Cowen [Palaeontology, Vol. 15, Part 3, 1972, pp. 405^11, pis. 76-78.] 406 PALAEONTOLOGY, VOLUME 15 and Rudwick (1966) quote Williams (1953) who, at that time, suggested that the spire- bearing Thecospira was a Triassic davidsoniacean. Williams (1968), in his survey of articulate brachiopod shell structure, noted that the shell of Cadomella and Kouinckina was identical, but not strophomenide. He described a thin primary layer followed by secondary fibres, forming an internal mosaic com- parable to the standard shell structure seen in spiriferides. In addition a tertiary prismatic layer, again as in the Spiriferida, was recognized. He showed too that the shell of Thecospira consisted of two standard calcareous layers and reassigned both it ‘and Cadomella and its associates to the Spiriferida’ (1968, p. 48). Recently Jaanusson (1971, p. 44), in his study of brachiopod articulation, has concluded that Cadomella is ‘a late member of the order Spiriferida’. Our observations confirm the opinion of Cowen and Rudwick (1966) and Williams (1968) that Cadomella is a member of the Koninckinacea, a relationship hinted at by Muir-Wood in 1962. In 1968 Williams was able only to use replicas for transmission electron-microscopy. Since then the use of the scanning electron microscope has greatly facilitated shell structural studies, in particular in allowing complete or fragmentary specimens to be searched for unusual or poorly known structures at a very wide range of magnifications. Furthermore, the process of replication may diminish or destroy remnants of primary shell on fossil species, a danger which is eliminated by direct study of the material under the scanning electron microscope. SHELL STRUCTURE The shell structures in all three species of Cadomella are the same, but differences occur in the absolute sizes and presumed duration of secretion of these structures, probably related to the much greater size of C. davidsorti. Details of the shell structure of C. moorei and C. quenstedti appear to be identical and here examples are figured only for the former species. The shell of C. davidsoni is considerably thicker than that of EXPLANATION OF PLATE 76 Specimens with B, BB, or ZB registration numbers are housed in the British Museum (Natural History), London. Eigs. 1-4. Scanning electron micrographs of the shell of Cadomella davidsoni (Eudes-Deslongchamps) from the Lias near Caen, Calvados, Erance. 1. Etched transverse section through primary (top left) and secondary shell layers, x 1050, BB58575, from The Queen's University, Belfast (QUB) collec- tions. 2. Eracture surface within the secondary layer, near the postero-median margin, showing the typical parallel stacking of fibres. Posterior to the top, viewed towards the exterior. X 450, BB7242, from May, 8 km S. of Caen. 3. Transverse etched section through secondary and tertiary shell layers. X 500, BB58575, from QUB collections. 4. Median longitudinal fracture surface, showing the junction of the secondary and tertiary layers. A temporary reversion to the development of fibres can be seen within the tertiary layer at the bottom left, x 525, BB7242. Figs. 5-6. Scanning electron micrographs of the shell of Cadomella (Davidson) from the Upper Lias of Curey, about 20 km SSW. of Caen, Calvados, France. B14642. 5. Median longitudinal fracture surface through a ventral valve showing primary shell covering secondary layer fibres. Anterior is to the right and the dififerent orientations of fibres can be seen. X 550. 6. Fractured surface within the secondary layer showing the typical shape and stacking of standard fibres, x 1400. In all figures, except figure 2, the valve exteriors are to the top. Palaeontology, Vol. 15 PLATE 76 BRUNTON and MACKINNON, Cadomella • fl-' ■ ■■ BRUNTON AND MACKINNON: CADOM ELLA 407 C. moorei, to a large extent being composed of the tertiary shell layer. In C. moorei the primary layer is thin, apparently only a few ^^m thick (PI. 76, fig. 5) while in C. davidsoiii this layer reaches 20 /xm thick (PI. 76, fig. 1). Secondary fibres have a more or less parallel standard stacking and cross-sectional shape (PI. 76, fig. 6). In C. moorei they measure 15-20 ^tm wide and about 4/xm thick (PI. 76, figs. 5,6) but measure up to40^m wide and 8 fj.m thick (PI. 76, figs. 1-3), and appear more triangular in cross-section in C. davidsoiii. During the ontogeny of C. moorei the first fibres grew tangentially to the valve margins, but within a few rows of the primary/secondary layer boundary they became reorientated to lie more or less radially (PI. 76, fig. 5). A tertiary layer was secreted in which columnar prisms of calcite grew nearly perpendicularly inwards from the secondary fibres in both C. moorei and C. davidsoni (PI. 76, figs. 3, 4). Within the much thicker tertiary layer of C. davidsoni sporadic reversions to the temporary secretion of fibres (PI. 76, fig. 4) have been recognized and probably resulted from mantle regres- sion close to the secondary /tertiary layer boundary. The fact that Cadome/la is reported to be pseudopunctate (Muir-Wood 1965, p. H438) has been used, in the past, as evidence of its strophomenide affinities. Williams (1968, p. 48) has recognized that the shell layers are inwardly deflected around rods of calcite which give rise to a tuberculate pattern on the internal surface of both valves. The rods, which measure up to 60 /xm in diameter, exhibit a fine porous or granular texture (PI. 77, fig. 1 ) so that they can even be recognized in sections through parts of the more massive- looking tertiary layer as well as the secondary layer. In most respects they resemble taleolae, the rod-like bodies which permeate the shells of many Strophomenida (e.g. PI. 78, fig. 9), but equally justifiable is a comparison with the tubercles of Megeriia, a Recent terebratulide, the Jurassic thecideacean Moorel/ina, or the Triassic spire- bearing Thecospira, all of which exhibit a shell fabric bearing a likeness much closer to that of Spiriferida than any Strophomenida. Unlike MoorelUna (Baker 1970, p. 87) but as in Megeriia and Thecospira, the tubercle cores are not continuous with material of the primary layer but appear to arise from clusters of fairly small secondary layer fibres which must have lain, initially, only a short distance from the primary/secondary layer boundary. Only those tubercles of Cadomella situated close to the valve margins appear to have been functional, for those rods located further from the shell edge (and hence secreted at an earlier stage in growth) appear to be overlapped by later secondary layer fibres. This situation is reminiscent of that in Megeriia where tubercles forming con- spicuous outgrowths at the valve margins (PI. 77, fig. 2) are, on occupying a position increasingly distant from the expanding shell edge, reduced to flattened scars (PI. 77, fig. 3) by the processes of resorption and overgrowth of new shell material. Around such scars fibres are still deflected until finally they are submerged below a stream of secondary layer fibres whose ragged outlines are the only indication of any disruption of the under- lying shell succession (PI. 77, fig. 4). At present it is uncertain what advantages were to be gained from having tuberculate shell margins, although an examination of the outer epithelial cells overlying tubercles in living Megeriia should give the answer. Provisionally it is noteworthy that well- developed tubercles in Megeriia are built up of a central complex of irregular calcite shreds which give way peripherally to more conventional secondary layer fibres (PI. 77, fig. 5). Around the base of each tubercle well-developed fibres with smooth terminal faces are deflected on both sides; but near its flattened head the regular outlines break 408 PALAEONTOLOGY, VOLUME 15 down in a manner similar to that affecting fibres located on the periphery of muscle scars (PL 77, fig. 6) so that it is possible that the overlying outer epithelial cells were internally supported by dense concentrations of tonofibrils and served as attachment areas for some specialized component(s) of the mantle. On the other hand, as tubercles are only upstanding around the periphery of both valves, and are very rapidly resorbed and overlapped posteriorly, it is possible that these unusual outgrowths may have func- tioned as nothing more than a skeletal support for a fleshy protective grille. DISCUSSION AND CONCLUSIONS From studies of Recent brachiopods it is known that secretion of the shell is controlled by the outer mantle epithelium and, therefore, intimately bound up with the biochemistry of the animal. Thus it seems logical to assume a high level of systematic importance for shell structure. Work in progress (D. I. MacKinnon) shows that the shell structure of Thecospira and several koninckinacean genera is essentially spiriferide in nature, i.e. standard primary and fibrous secondary layers, with or without the addition of a tertiary prismatic layer internally. Observations recorded above show that Cadomella has an EXPLANATION OF PLATE 77 Fig. 1 . Scanning electron micrograph of the shell of Cadomella clavidsoiii (Eudes-Deslongchamps) from the Lias of Curey, near Caen, France. BB58575, from QUB collections. Etched section through a tubercle submerged within the secondary layer of a brachial valve, X 525. Figs. 2-6. Scanning electron micrographs of the shell of Megerlia tnmcata (Linnaeus); Recent. Medi- terranean Sea. ZB3318, from QUB collections. 2. View of a tubercle located close to the valve margin showing the disposition of secondary fibres around its base and wall, and the breakdown of the mosaic when traced towards the porous core (see fig. 6). X 425. 3. View of a tubercle submerged within secondary fibres and appearing as a flattened scar, x 525. 4. View of the secondary layer internal mosaic modified where the fibres are overlapping a submerged tubercle. X 525. 5. Etched section through part of a tubercle (top right) showing the inward deflection of secondary layer fibres. Valve interior to the left, x 1015. ZB3319. 6. View of the tip surface of a tubercle showing the central region of irregular calcite shreds. X 825. EXPLANATION OF PLATE 78 Figs. 1-7 illustrate the umbonal regions of a variety of Spiriferida showing apical pedicle foramina. Fig. 1. Dorsal view of Amphiclina amoena Bittner (Koninckinidae) from the Triassic St. Cassian Beds near Cortina, Italy. x4-4. BB58567. Fig. 2. Dorsal view of Cadomella davidsoni (Eudes-Deslongchamps) (Cadomellacea) from the Upper Lias of May, Calvados, France, x 7. BB58568. Figs. 3, 4. Dorsal views of two specimens of Koninckella triassina Bittner (Koninckinacea) from the Triassic St. Cassian Beds near Corvara, Italy. X 8. BB58565-58566. Fig. 5. Dorsal view of Homoeospira evax (Retziacea) from the Upper Silurian of Indiana, U-S. A. x4-2. B5355, Davidson Collection. Fig. 6. Dorsal view of the umbonal region of Athyris vittata Hall ( Athyridacea) from the Mid Devonian of Ontario, Canada. x6-3. B7796, Davidson Collection. Fig. 7. Dorsal view of the umbonal region of Tetractinella trigoiiella (Schlotheim) (Athyridacea) from the Trias of Italy. x5-5. B7801, Davidson Collection. Figs. 8, 9. Scanning electron micrographs of the shell of a dorsal valve of a Permian chonetacean, Dyoros sp., from Russia, B2149. 8. Deeply exfoliated external view showing laminar sheets com- posed of calcite blades. Anterior to the bottom. X 2400. 9. Exfoliated internal surface, close to the antero-lateral margin (to bottom) showing shell lamellae deflected around the taleola of a pseudo- punctum, X1150. Palaeontology, Vol. 15 PLATE 77 BRUNTON and MACKINNON, Cadomella Palaeontology, Vol. 15 PLATE 78 w?Q'; BRUNTON and MACKINNON, Cadomella ■ . t-' r ■. ■i:- ^■. ■■ BRUNTON AND MACKINNON: CADOMELLA 409 identical shell structure, which supports the view of Williams (1968) that the shell structure of this genus, like other Koninckinacea, is typical of the Spiriferida. In Cadomella the position of the pedicle foramen and the nature of restrictions of the delthyrial and notothyrial cavities have been cited as evidence in support of its having interarea a pedicle foramen symphytium TEXT-FIG. 1 . Postero-dorsal views of the umbonal regions of AmphicUna anioena Bittner (a), KoninckeUa triassina Bittner (b), and Cadomella davidsoni (Eudes-Deslongchamps) (c), showing the similar nature of the pedicle openings. In AmphicUna the hinge line, and consequently the interareas, are only narrowly developed and the posterior width of the shell accommodates wide cardinal areas. In both KoninckeUa and Cadomella the hinge lines and interareas are wide. See Plate 78, figs. 1-4. strophomenide affinities. According to Cowen and Rudwick (1966, p. 404) the pedicle foramina of KoninckeUa Uassina Bouchard, K. triassina Bittner, and AmphicUna suessi Laube, as well as Cadomella, ‘are definitely supra-apical’. The term ‘supra-apical’ was first used by Arber (1940, p. 162) to describe a foramen when it is separated from the ventral interarea by a solid rim of the brephic pedicle valve; presumably Cowen and Rudwick employed the term in the same sense. However, the foramina of a number of Koninckinacea examined by us (including specimens of 410 PALAEONTOLOGY, VOLUME 15 Koninckella triassina Bittner (PI. 78, figs. 3, 4 and text-fig. \b), Amphiclina amoena Bittner (PI. 78, fig. 1 and text-fig. In), as well as Cadomella davidsoni (Eiides-Deslong- champs) (PI. 78, fig. 2 and text-fig. le), do not fall into this category. The foramina of Koninckella and Amphiclina are situated apically like those of many Retziidines (PI. 78, fig. 5) and Athyrididines (PI. 78, figs. 6, 7) and no wedge of shell material has been found to isolate the pedicle opening completely from a ventral interarea as occurs in many Strophomenida. In some mature Cadomella davidsoni the pedicle foramen may appear to be supra-apical, but in younger specimens the foramen is in contact with the anterior- most apex of the ventral interarea (PI. 78, fig. 2). In C./uoore/ the pedicle foramen is open to the delthyrium (see Treatise, fig. 285, Ic). Wear and tear around the pedicle opening is recognizable in many Recent and fossil Brachiopoda, and the slight anterior migration of the pedicle opening in the genera under discussion as well as an over-all increase in size may be accounted for by a combination of abrasive and/or resorptive processes accompanying normal shell growth. A significant feature is the fine external ornamentation, seen only on unabraded areas of a few specimens of C. davidsoni, in which the finely granular surface consists of closely packed small elongate pustules arranged radially. Between thirty to forty such pustules occur within 1 mm width of the shell surface. This type of surface ornamentation is unknown in the Chonetacea but common in the Spiriferida. As the last undoubted chonetacean is known from Permian strata it is impossible to compare contemporaneous shell of true chonetaceans with any koninckinacean. Work by Brunton (1972) on the shell of chonetaceans and their ancestors shows that the shell structure of Permian representatives is essentially laminar (PI. 78, figs. 8, 9). Each lamina is composed of blades which do not run parallel to those of the previously deposited lamina. Neither a finely ‘granular’ primary layer nor a prismatic tertiary layer is found in any chonetacean. During the Silurian, Devonian, and much of the Carboni- ferous the chonetacean shell structure became progressively more laminar, having had its origins in the orthodox shell structure of Ordovician plectambonitaceans. It seems highly unlikely that a reversion to standard primary/secondary shell secretion, plus the addition of the tertiary layer, could have been achieved between the late Permian and mid Triassic, when the Koninckinacea appear. Besides the lack of hollow posteriorly directed marginal spines and presence of a spiral brachidium there are morphological features in the Koninckinacea quite unlike those of chonetaceans, the most important of which are discussed above. Bearing in mind the great morphological variation in the Spiriferida and the umbonal features of such stocks as the Athyridacea and Retziacea, which range into the Triassic with Koninckinacea-like shell, the gross morphology of the Koninckinacea seems closer to that of the Spiriferida than to any strophonienide. Add to this the spiral brachidium and shell structural evidence and their spiriferide affinities seem incontestable. Acknowledgements. We gratefully acknowledge our thanks to Professor Alwyn Williams for construc- tive comments on the typescript, and to NERC for providing a research studentship to D. I. Mac- Kinnon. REFERENCES ARBER, M. A. 1940. The relation of the valves to the pedicle in the strophomenid brachiopods. GeoL Mag. 77, 161-174. BRUNTON AND MACKINNON: CADOMELLA 411 BAKER, p. 1970. The growth and shell microstructure of the thecideacean brachiopod Moorellina granulosa (Moore) from the Middle Jurassic of England. Palaeontology, 13, 76-99, pis. 18-21. BOUCOT, A. J., JOHNSON, J. G., and STATON R. D. 1965. In Williams, A. et al., Treatise on invertebrate paleontology, pt. H, Brachiopoda, H632-667, Kansas. BRUNTON, c. H. c. (1972). The shell structure of chonetacean brachiopods and their ancestors. Bull. Br. Mus. nat. Hist. (GeoL), 21, 1-26, pis. 1-9 cowEN, R., and rudwick, m. j. s. 1966. A spiral brachidium in the Jurassic chonetid brachiopod Cadomella. Geol. Mag. 103, 403-406. JAANUSSON, V. 1971. Evolution of the brachiopod hinge. Smithsonian Contribs. to Paleobiol. 3, 33-46. MUiR-wooD, H. M. 1955. A history of the phylum Brachiopoda. Brit. Mus. Nat. Hist., London. 124 pp. 1962. On the morphology and classification of the brachiopod suborder Chonetoidea. Brit. Mus. Nat. Hist., London. Pp. 132, 16 pis. 1965. In Williams, A. et al.. Treatise on invertebrate paleontology, pt. H, Brachiopoda, Kansas, H41 2-439. RAU, K. 1905. Die Brachiopoden des mitteren Lias Schwabens mit Ausschluss der Spiriferinen. Geol. Pal. Abh. 6, 263-355, pis. 32-34. WILLIAMS, A. 1953. The classification of the strophomenoid brachiopods. J. Wash. Acad. Sci. 43, 1-13. 1968. Evolution of the shell structure of the articulate brachiopods. Special Papers in Palaeonto- logy, 2, 1-55, pis. 1-24. C. H. C. BRUNTON Department of Palaeontology British Museum (Natural History) London, S.W.7 D. I. MACKINNON Department of Geology University of Canterbury Pinal Typescript received 7 November 1971 Christchurch, New Zealand C9016 Ee OBSERVATIONS ON SOME LOWER PALAEOZOIC TREMANOTIFORM BELLEROPHONTACEA (GASTROPODA) FROM NORTH AMERICA by JOHN S. PEEL Abstract. Muscle scars are described in Salpingostoma biielli (Whitfield 1878) and Tremanotus alpheiis Hall 1865 with reference to finely preserved specimens of Bellerophon spp. The taxon Salpingostomatidae Koken 1925 is abandoned with transferal of Salpingostoma Roemer 1876 to the subfamily Bucaniinae of the Bellerophontidae. A new subfamily of the Sinuitidae, the Tremanotinae, is proposed to accommodate Tremanotus Hall 1865 and Boiotremiis Horny 1962. Brief discussion is given of described species of the three genera from the Silurian of North America. It is over a hundred years since F. B. Meek (1866) discussed the affinity of the family Bellerophontidae M’Coy 1851 and gave ‘. . . as nearly a positive demonstration of . . . the affinities of the Bellerophon-gxow'p, as we can probably ever expect in such a case’. In so doing, he ended much of the controversy that had developed regarding the systematic position of this extinct molluscan group since von Hupsch first noted their existence in 1781. By reference to Tremanotus chicagoeusis (McChesney 1859), Meek reiterated de Koninck’s (1842-1844) interpretation of the bellerophontaceans as proso- branchiate gastropods. The presence of tremata was held to indicate that Tremanotus ‘. . . bears exactly the same relations to Bueania, that PoJytremaria does to Pleurotomaria, and Rimula to Emarginula'. Since that time there have been considerable advances in our knowledge concerning the bellerophontiform molluscs. A notable contribution was the demonstration by Wenz (1940) of the tryblidiid affinity of the coiled Cyrtonel/a mitella (Hall 1862) which, of necessity, cast doubt upon the status of the remaining members of the group. Wenz (1940) presumed that all bellerophontaceans possessed similar musculature indicating a lack of torsion and placed the superfamily alongside the Tryblidiacea in the subclass Amphigastropoda. However, the description by Knight (1947) of muscle scars, sugges- tive of torsion having taken place, in Bellerophon gibsoni White 1882 and Sinuites caneellatus corrugatus (Hall 1847) clearly indicated that the Amphigastropoda of Wenz (1940) was not a homogeneous unit. Furthermore, the Bellerophontacea of then current usage could itself no longer be maintained as a single entity. The distinction between the coiled Cyclomya (Monoplacophora) and the Bellero- phontacea (Gastropoda) is now well recognized, principally through the works of Horny (1962, 1963r/, 19636, 1965a, 19656, 1966). Recent summaries are given by Yochelson (1967), Rollins (1969), and Starobogatov (1970). Muscle scars are known in an increasing number of bellerophontiform molluscs, though the description by Rollins and Batten (1968) of the sinus-bearing monoplacophoran Sinuitopsis acutilira (Hall 1861) serves to emphasize some of the problems accompanying systematic determination of the many remaining taxa. The intention of this paper is to describe structures of the shell interior in the Ordo- [Palacontology, Vol. 15, Part 3, 1972, pp. 412-422, pi. 79.] PEEL: LOWER PALAEOZOIC TREM ANOTIFORM BELLEROPHONTACEA 413 vician Sa/pingostoma buelli (Whitfield 1878) and a specimen of Tremcmotiis alpheus Hall 1865 from the Niagaran (Silurian) dolomites of Illinois. Owing to the coarseness of preservation, interpretation of these structures is attempted with reference to well- preserved specimens of Bellerophon spp. of Carboniferous age. It is of historical interest that the conclusions reached by Meek (1866) after examination of one member of this group of dorsally perforate molluscs receive further support, the described internal structures being interpreted as muscle scars of gastropod type. On the basis of general morphology the suprageneric classification of Tremanolus, SaJpingostoma, and Boio- tremus Horny 1962 is revised and a brief discussion given of the described species of those genera from the Silurian of North America. Repositories. The following Institutions are denoted by abbreviations in the text : USNM, U.S. National Museum, Washington D.C. ; AMNH, American Museum of Natural History, New York; GPM, Greene Paleontology Museum, Univ. of Wisconsin at Milwaukee; YPM, Yale Peabody Museum, New Haven, Conn.; OUM, Oxford University Museum. MUSCLE SCARS IN BELLEROPHON The most readily observed indication of the position of supposed rectractor muscles of the type described by Knight (1947) is a fine spiral ridge on each umbilical shoulder of natural internal moulds of the shell. Under exceptional conditions of preservation a series of fine crescentic striae can be seen on the mould at the adapertural extremity of each ridge. The crescents are convex adaperturally and coalesce dorsally at the spiral ridge. Each spiral ridge was interpreted by Knight (1947) as forming the dorsal margin of a muscle attachment scar, with successive positions of the adapertural scar margin producing the crescentic striae. In some cases a groove on the mould may replace the spiral ridge while a number of minor structures are commonly developed in association with the general muscle attachment area. The latter include local swellings of the mould, corresponding to slight hollows on the shell interior, and changes in the degree of angularity of the umbilical shoulders. Two internal moulds assigned to Bellerophon show well-preserved muscle scars. The larger specimen (USNM 169475), apparently from the Pennsylvanian of the U.S. A. and hereafter referred to as Bellerophon specimen A, lacks the apertural margin, though the last preserved growth stage is almost certainly less than one tenth of a whorl back from the true aperture (PI. 79, fig. 1). Three whorls are present, the nuclear whorls having been lost. Although the shell infilling is rather coarse the surface of the mould is smooth and polished. A well-marked spiral ridge is observed on each umbilical shoulder with crescentic scars developed at the adapertural extremity of the ridges. The crescentic scars lie half a whorl back from the final preserved growth stage. Both spiral ridges terminate abruptly one half whorl back from their anterior extremities. In detail, the nature of the spiral ridges is seen to vary over the half whorl extension. Abaperturally, each spiral ridge has greater relief and is overturned towards the dorsum. As a conse- quence, a narrow well-defined channel exists between the acute overturned crest of the spiral ridge and the dorso-lateral surface of the mould. In terms of the shell interior, this channel equates with a low-angled projection of shell towards the umbilical shoulder. Proceeding adaperturally, the relief of each ridge decreases, the crest becomes less acute and the channel is lost. However, the general form of a gently convex slope on the umbilical side of the ridge and a steep surface on the dorsal side, is retained. A series of grooves occupies the crest of each spiral ridge, the various grooves being continuous with the several crescentic scars. 414 PALAEONTOLOGY, VOLUME 15 A broad, elongate swelling is seen on the dorsal side of each spiral ridge. This structure, representing a hollow in the inner side of the shell, is subparallel to the ridges but commences at a slightly earlier growth stage. Proceeding towards the aperture, the swelling and spiral ridge diverge. It is not possible to trace the swelling beyond the midpoint of the spiral ridge on either shoulder. The second internal mould, Bellerophon specimen B, is from the Carboniferous Lime- stone of Armagh, Northern Ireland (OUM E2214). The apertural margin is not pre- served and the left side is obscured by coarsely recrystallized calcite [left and right are used in the sense of Knight (1947). In lateral view with the aperture facing to the left, the side visible is the left side]. The muscle scar on the right side is visible (PI. 79, fig. 2) and is of particular interest in that the abapertural margin of the muscle attachment area can be observed. In addition, the scar shows notable morphologic differences from the form seen in Bellerophon specimen A. TEXT-FIG. 1. Lateral view of the right muscle scar of Bellerophon specimen B (OUM E2214). The arrow indicates the direction of growth. For explanation see text ( X 6). The spiral ridge (text-fig. 1 ; r) is symmetrical with a convex upper surface in contrast to the ad-dorsal ly overturned acute ridges of the previously described specimen. The abapertural termination is sudden and a faint scar passing around the extremity indicates the margin of the muscle attachment area. The umbilical margin of the scar is visible for about one-third of the length of the spiral ridge and encloses a spirally striated muscle attachment area. Dorsally, the scar margin is continued as the acute dorsal edge of a flat topped swelling (p) which parallels the abapertural third of the spiral ridge. The swelling has a concave slope passing from this dorsal edge to the normal surface level of the dorsum and fades away suddenly when the dorsal edge converges with, and meets, the spiral ridge. Two obscure swellings (si; 5.,) diverge from the spiral ridge at this point at approximate angles of 30° and 45° and pass obliquely across the dorsum, almost to the median plane. Approach of the spiral ridge towards the umbilical shoulder causes an increase in the angularity of the mould shoulder which reaches a maximum at the point of divergence of the oblique transdorsal swellings. Proceeding adaperturally, the spiral ridge increases in relief but loses the initial strong delimitation from the dorsal surface by the acquisition of more gently sloping sides. The upper surface is marked with a prominent median groove (g) from which adumbilically concave striae diverge in the direction of the umbilical shoulder. After a total length of half a whorl, the spiral ridge gradually disappears and obscure crescentic striae (c) are developed with the median groove forming their dorsal margin. The muscle scars in Bellerophon specimen A are essentially identical to those described by Knight (1947, pi. 42, figs. 2, 3r/, h) in B. gibsoni White 1882. The scar in Bellerophon PEEL: LOWER PALAEOZOIC TREMANOTIFORM BELLEROPHONTACE A 415 specimen B retains the over-all form but differs with regard to expression of the detail. In the absence of variation studies of bellerophontacean muscle scars it would be unwise to attempt any assessment of the importance of the differences. The spiral ridges were interpreted by Knight (1947) as the dorsal margins of muscle attachment scars, the crescentic grooves marking successive ontogenetic positions of the adapertural margin of the scars. Although not observing the umbilical and abapertural margins of the muscle attachment areas, Knight (1947, p. 266) considered the length of the spiral ridges to approximate to the length of attachment of the retractor muscles. While the abrupt abapertural termination of the spiral ridges in many observed speci- mens of Bellerophon tends to support this view, the evidence available from Bellerophon specimen B, above, adds direct confirmation. However, this would not appear to be the case in all bellerophontaceans and, as a general rule, the spiral ridges are best regarded as the locus of the muscle attachment scars unless detail of the scar margins permits additional qualification. Both of the described specimens of Bellerophon have swellings variously associated with the spiral ridges, though the equivalence of these structures is uncertain. The prominent fiat-topped swelling seen in specimen B is closely related to the ridge itself, and lies within the muscle attaehment area. The swellings on specimen A lack this specific association with the spiral ridges and the embedded nature of the dorsal margin of the retractor muscles, indicated by the overturned spiral ridges, suggests that the swellings were outside the muscle attachment areas. It is more probable that the oblique trans- dorsal swellings which diverge from the spiral ridge in specimen B are comparable to the structures seen in the North American example. While the location of the various swellings on the moulds demonstrates some connection with the musculature, the nature and purpose of the structures is quite unknown. MUSCLE SCARS IN S ALPINGO STO MA BUELLI (WHITFIELD 1878) A full description of this species is given by Ulrich and Scofield (1897, pp. 900-901). The two specimens discussed here are internal moulds from the Middle Ordovician (Black River) Plattville Fm. at Beloit, Wisconsin. The more complete (USNM 15641) shows the bell-shaped final growth stage and explanate aperture characteristic of the genus (PI. 79, figs. 3, 6, 8) but in the other (USNM 169472) the explanate stage is not preserved. The median dorsal slit is represented by a raised spiral ridge on each specimen. Circumbilical spiral ridges have been observed in both specimens but crescentic scars of the type noted above have not been seen (PI. 79, figs. 3-5). In view of the relative coarseness of the matrix, it is unlikely that such delicate structures would be preserved. Specimen USNM 15641 shows the right and left spiral ridges but the former is rather obscure. Eaeh ridge is loeated on the umbilical wall about halfway between the whorl periphery and the suture with the previous whorl. The adapertural extremity of each ridge lies vertically above the plane of the aperture, high on the side of the bell-shaped final growth stage. A low callosity is developed on the floor of the whorl at this point. It is difficult to ascertain the length of the spiral ridges but it would appear that the left ridge (text-fig. 2b, r) extends over at least half a whorl. In USNM 169472 only the right spiral ridge is seen, the left umbilical wall being 416 PALAEONTOLOGY, VOLUME 15 damaged. The ridge is essentially identical with those visible in the more complete specimen but is only observable over about one quarter of a whorl. The spiral ridges in Salpingostoma hiielli differ from those in Bellerophon specimens A and B with regard to position, relief, and relationship to the slit. In 5’. buelli the ridges are located on the umbilical walls, well within the umbilicus, while in specimens A and B of Bellerophon the ridges are situated on the umbilical shoulders, essentially at the maximum width of the whorl. The positional differences can be partly accounted for by change in whorl profile. The latter specimens have whorls which are reniform in cross- section with relatively high impression of the earlier whorl and a deep, narrow umbilicus. TEXT-FIG. 2a. Lateral view (x 1) of Tremanotits alpheiis Hall 1865 (GPM 21123) showing the position of the left swelling. The cross indicates the point of maximum relief of the structure. Numbers around the periphery mark the location of the tremata though an earlier probably incomplete trema noted in text-fig. 3 is omitted. The origin of the radial scale employed in text-fig. 3 is shown. B, lateral view (xl) of Salpingostoma bnelli (Whitfield 1878) (USNM 15641) illustrating the left spiral ridge (/ ). The extent of the open dorsal slit which is partly infilled with matrix is denoted by the dashed line. The fine dotted line indicates maximum whorl width. S. buelli is evolute with minor impression of the previous whorl and a lenticular whorl profile. The relatively acute nature of the junction between the dorso-lateral and umbili- cal surfaces in S. buelli possibly affords a less suitable site for muscle attachment than the corresponding more convex site in Bellerophon Specimens A and B. It is difficult to assess the effects of acquisition of the bell-shaped late growth stage EXPLANATION OF PLATE 79 Fig. 1. Bellerophon specimen A (USNM 169475), Pennsylvanian, U.S.A. {non loc.)', lateral view of internal mould showing right muscle scar, the arrow indicates the obscure anterior margin crescentic scar, X 1 . Fig. 2. Bellerophon specimen B (OUM E2214), Armagh, N. Ireland; lateral view of internal mould showing right muscle scar, X 1 . Figs. 3-6, 8. Salpingostoma bnelli (Whitfield 1878), Beloit, Wisconsin, U.S.A. 3, 6, 8 (USNM 15641); 3, oblique umbilical view showing circumbilical ridge, X 1 ; 6, dorsal view, X 0-75 ; 8, posterior view, xO-75; 4, 5 (USNM 169472); oblique views illustrating the circumbilical spiral ridge, X 1. Figs. 7, 9-11. Tremanotus alpheiis Hall 1865 (GPM 21123), Hawthorne, Illinois, U.S.A. 7, oblique lateral view showing ridge on ad-dorsal margin of swelling, xO-75; 9, posterior view showing swell- ings, xO-75; 10, lateral view, xO-75; 1 1, dorsal view illustrating the dorsal tremata, xO-75. Palaeontology, Vol. 15 PLATE 79 PEEL, Lower Palaeozoic Bellerophontacea PEEL: LOWER PALAEOZOIC TREM ANOTIFORM BELLEROPHONTACE A 417 upon the musculatory requirements of the individual. It is probable that the more central position of the retractor muscles in S. buelli enabled better control of the body mass than muscles situated peripherally. The relationship between the presently observed musculature and that prior to attainment of the bell-shaped shell is unknown. The spiral ridges in S. biielli are much broader than those in Bellerophon specimens A and B and lack the strong delimitation from the normal steinkern surface typical of their development in the Carboniferous examples. Although the adapertural extremities of the ridges are comparable distances back from the aperture, there is variation with respect to position relative to the dorsal emargina- tion. In Bellerophon specimen A the slit is very short, the deepest portion being half a whorl forward from the ends of the spiral ridges. In S. buelli the deepest part of the long slit lies one quarter of a whorl back from the apertural termination of the ridges. The importance of this dissimilarity is unknown since studies into the nature, particularly depth, of the bellerophontacean or pleurotomariacean slit have not been made. Anatomi- cal ditferences are presumably involved but, at present, there is no way of resolving these. Indeed, neither is it possible to make analogies concerning the slits for there is no reason to presume that the respective anal openings bore similar relationships within those structures. MUSCLE SCARS IN TREMANOTUS ALPHEUS HALL 1865 An internal mould of Tremanotus alpheiis Hall 1865 from the Niagaran dolomites at Hawthorne, Illinois (GPM 21123) shows paired swellings on the umbilical shoulders (PI. 79, figs. 7, 9-11). The specimen differs from the holotype and materials described by Clarke and Ruedemann (1903) and Knight (1941) in its lack of a well-developed flared aperture. However, measurement of gradients of growth (text-fig. 3;a,b) indicates the characteristic rapid expansion of the later portions of the whorl and, since the development of the feature is cyclic, it is probable that the specimen represents a growth stage intermediate between successive flared stages. The form of the tremata (text-fig. 3; d) is in close agreement with the holotype and other specimens in the large sample available at Milwaukee do have flared apertures preserved. The structures take the form of a discrete swelling on each umbilical shoulder pro- ducing a slight increase in the total whorl width (text-fig. 3; c). The increase commences suddenly at the beginning of the final whorl (text-fig. 2a) and the internal mould has returned to normal width by the appearance of the first trema, a fifth of a whorl later. An increment of 2 mm is added to the width by the swellings. Similar swellings are apparently present at approximately one quarter of a whorl earlier. The left side is slightly damaged, but on the right side a crude spiral marking forms the dorsal limit of the swelling (PI. 79, fig. 7). The specimen has no structures comparable to the fine spiral ridges and crescentic scars which characterize the muscle attachment areas in the described specimens of Bellerophon but the nature of the matrix precludes preservation of these. However, equivalence can be suggested between the undamaged right swelling in the specimen of T. alpheus and the similar swelling located on the abapertural portion of the right spiral ridge in Bellerophon specimen B. The strong dorsal margin to the swelling in T. alpheus favours comparison with the latter-mentioned struc- ture rather than with the more subdued paired swellings of Bellerophon specimen A. Wing (unpublished Ph.D. thesis, 1923, The Silurian Gastropoda of northeastern 418 PALAEONTOLOGY, VOLUME 15 Illinois, University of Chicago) in his description of Tremcmotus iimbonus nom. mid., commented on intermittent expansion of the final whorl causing development of distinct shoulders. The description and the illustration do not permit precise comparison with the swellings observed here, but Wing’s comment that three expansions appear on the last half whorl must surely cast doubt on any interpretation as muscle scars. As noted in the final section of this paper, the holotype is too poorly preserved to warrant description. TEXT-FIG. 3. Growth gradients in Tremanotiis alpheiis Hall 1865 (GPM 21123). a. Radial vector from axis of coiling to periphery of final whorl showing increased rate of expansion in latest growth stage. b. Radial vector from axis of coiling to suture between final and penultimate whorls. The discrepancy in time of commencement of the increased rates of expansion in a and b is an apparent effect produced by the radial method of measurement. Since the lines of growth are opisthocline, a point on the whorl periphery lies on a different radius than an ontogenetically equivalent point on the suture. c. Increase in whorl width during the final whorl illustrating the paired swellings. d. Location of the dorsal tremata. The tremata are represented in diagrammatic form at their true width (not scale width). Incomplete rectangles denote damaged tremata. The vertical scale is logarithmic. The horizontal scale is an angular measure of growth stage from an arbitrarily selected point on the final whorl (text-fig. 2a). SUPRAGENERIC CLASSIEICATION OF TREMANOTUS, BOIOTREMUS, AND SALPINGOSTOMA Horny (1962) elevated the subfamily Salpingostomatinae Koken 1925 to familial status, placing his new genus Boiotremus therein. In a later discussion (Horny 1963) he considered the family containing Salpingosioma Roemer 1 876, Tremanotiis Hall 1 865, and Boiotremus to lie close to the Sinuitidae but to have certain connections with the Bellerophontidae. In discussing Tremanotiis Horny (1963u, p. 97) noted that the genus developed tremata only in the later growth stages and that early growth stages had no such openings or sign of them ever having been present. The tremata were considered to be related to the PEEL: LOWER PALAEOZOIC TREM ANOTIEORM BELLEROPHONT ACE A 419 periodic development in late ontogenetic stages of the flared aperture and did not occur prior to the acquisition of this feature. Boiotremus was erected by Horny (1962) to include those species previously assigned to Trewauoius which developed tremata and an associated flaring aperture throughout ontogeny. The type species of Salpiugostoma is the poorly known S. megalostoma (Eichwald 1 840) from the Ordovician of Esthonia. Following the generally accepted interpretation of the genus by Ulrich and Scofield (1897), Salpingosloina is characterized by the possession of a slit in all stages of growth but the very latest. The slit is deep and narrow and generates a true selenizone. An open slit is not present in the latest portions of the bell-shaped expansion typical of the genus, or in the explanate part of the aperture. In terms of the characters of the emargination, Tremanotus prior to the development of the flared aperture and its associated tremata has afflnity with the Sinuitidae in its possession of a V-shaped sinus. As Ulrich and Scofield (1897) have observed, prior to attainment of the flared aperture, Salpiugostoma agrees in all aspects with Bucania Hall 1847. The two forms are comparable only in the development of a flared aperture which produces a closing of the emargination. In Salpiugostoma the apertural portion of the true slit is closed while in Tremanotus the deepest part of the sinus is left as an open trema at the commencement of apertural expansion. Boiotremus is closely related to Tremanotus, differing in its production of tremata and flared aperture throughout ontogeny. Accepting the differences in the nature of the emargination it is not possible to uphold the family Salpingostomatidae and relocation of the three genera previously placed there is necessary. A new subfamily — Tremanotinae — of the family Sinuitidae is proposed to contain Tremanotus and Boiotremus. Saipingostoma is transferred to the subfamily Bucaniinae of the family Bellerophontidae. Class GASTROPODA Subclass PROSOBRANCHIA Miliie-Edwards 1848 Order archaeogastropoda Thiele 1925 Suborder bellerophontina Ulrich and Scofield 1 897 Superfamily bellerophontacea M’Coy 1851 Family sinuitidae Dali in Zittel-Eastman 1913 Subfamily tremanotinae nov. Diagnosis. Essentially sinuitid bellerophontaceans developing dorsal tremata in associa- tion with a widely expanded aperture. Discussion. This taxon is proposed to accommodate Tremanotus Hall 1865 and Boiotremus Horny 1962. The latter genus represents the endpoint of a sequence more clearly seen in Tremanotus whereby dorsal tremata are developed in relation to rapid expansion of the shell aperture. In Tremanotus the periodic expansions are commenced only in later ontogenetic stages, while in Boiotremus they occur throughout ontogeny. It is presumed that the successive flared apertures are resorbed with continuation in growth. SILURIAN SPECIES OF TREMANOTUS, BOIOTREMUS, AND SALPINGOSTOMA FROM NORTH AMERICA Bassler (1915) recorded the following described species of Tremanotus, all of Silurian age: T. alpheiis Hall 1865 T. crassolare (McChesney 1861) T. angustata (Hall 1852) T. pervoluta (McChesney 1861) T. chicagoensis (McChesney 1859) T.t trigonostoma Hall and Whitfield 1875 420 PALAEONTOLOGY, VOLUME 15 to which may be added; T. longitiidinalis Lindstrom 1884 T. Northrop 1939. of Northrop 1939 T. alpheus is well known from the works of Clarke and Ruedemann (1903) and Knight (1941) and need not be further discussed. The holotype of T. augiistata (AMNH 2235) is a rather poorly preserved internal mould from the Guelph Formation of Ontario. Although the general form is moderately well displayed, there is no indication of the nature of the dorsal emargination. This is possibly the reason for some of the confusion that has arisen concerning the relationship of this form to T. alpheus (see Whiteaves 1895, pp. 70-71). Comparison of the two holotypes shows that T. angustata is more laterally compressed, with a much wider umbilicus. A specimen in the U.S. National Museum (USNM 67088), referred with confidence to T. angustata, has the dorsal tremata clearly preserved. Much of the expanded portion of the shell is missing but there are at least ten open tremata borne on a raised median ridge on the mould. External shell characters remain unknown but the tremata justify transferal to Boio- tremus. The genus is typified by abundant, short, and closely spaced tremata, while in Tremauotus the dorsal openings are numerically fewer, elongate, and more widely spaced. The type of T. chicagoeusis is lost, but a plastercast in the U.S. National Museum (USNM 67594), purporting to be of that specimen, agrees exactly with McChesney’s (1867, pi. 8, fig. 5) illustrafion. The very slowly expanding whorl profile prior fo the bell-shaped final growth stage seems to be characteristic. Details of the dorsal emargination are not visible, but internal moulds from Huntington, Indiana (USNM 67128) show elongate, widely spaced tremata of Tremanotus-Xypt, carried on a dorsal ridge. A fragmental external mould has some suggestion of the distinctive tremanotinid ornament of prominent spiral cords and posteriorly directed growth lines. McChesney’s Bucaiiia crassolare and Bucania pervoluta are imperfectly known. The original descrip- tions (McChesney 1861) are unaccompanied by illustrations and provide no definite criteria for delimitation from co-existent species. In the absence of type specimens, it is not possible to make any reliable determination. Wing (1923 MS.) referred forms with a broad whorl profile to T. crassolare but failed to recognize Bucania pervoluta. T. Itrigonostoma is similarly in need of reinvestigation. One of the major problems in determination of the Niagaran species of Tremauotus is a lack of knowledge concerning variation. This would appear to be considerable in terms of whorl tumidity and apertural shape. As a consequence, determination of such poorly defined forms as the three previously mentioned species must at best be delayed and might prove to be undesirable. Horny (1963) considered T. longitudinalis Lindstrom 1884 to be a typical Boiotremus. Although the specimen assigned to that species by Northrop (1939) has not been examined, an example from the Gascons Formation of the Port Daniel area, Quebec (USNM 169474) would appear comparable to Lindstrom’s species. The dorsum is partly damaged but shows at least eleven tremata of Boiotremus aspect. Northrop (1939) mentioned the presence of tremata in T. miuutus, but his descriptions and illustra- tions do not demonstrate the form of the openings sufficiently to permit assignment to either Tremauotus or Boiotremus. In an unpublished thesis Wing (see above) described a new species of Tremauotus from the Niagaran of Illinois under the name T. umbouus (above). The name was included in the published abstract of that thesis (Wing 1925) without description, illustration, or reference to the earlier manuscript descrip- tion and is consequently a uomeii nudum. The specimen, in the collections of the University of Chicago at the Field Museum, Chicago, is in a thoroughly fragmented state and is not considered worthy of description. Gross form suggests T. chicagoeusis. Four recorded species of Salpiugostoma from the Silurian of North America have been noted: S. horeale Whiteaves 1904 S. iuoruatum Northrop 1939 S. dilatus (Sowerby 1839) of S. orieutalis Twenhofel 1928 Northrop 1939 S. boreale is a diminutive form unusual in its subcircular whorl profile. Type material in the collec- tions of the Geological Survey of Canada, Ottawa indicates a selenizone of Salpingostoma-type. In the specimen examined, the expanded apertural portion has no slit while earlier growth stages show a narrow selenizone with lunulae. Damage has removed the portion of the dorsum where the open PEEL: LOWER PALAEOZOIC TREM ANOTIFORM BELLEROPHONTACEA 421 emargination might be expected to occur. The depth of the slit must be considerably less than that seen in S. biielli. The two species recorded by Northrop (1939) from Gaspe are each known only from a single specimen. S. clilaiatiim (Sowerby 1839) of Northrop 1939 cannot be placed in Sowerby’s species if it is a true Salpingostonia. In his description of the holotype Reed ( 1921, p. 82) commented on the distinct traces of foramina (tremata) which are indicative of Tremanoliis or Boiotreimis rather than Salpingo- sloma. The specimen, in the Yale Peabody Museum, is a fragment of the aperture rather too poor for comparison. The holotype of S. inoniatiim (YPM 13314) is of particular interest because of Northrop’s (1939, p. 207) claim of ‘two curious bodies, probably opercula . . .’ which occupy the aperture. Examination of the specimen reveals that the bodies are respectively a trilobite cephalon and a bivalve of Paracyclas- type. The holotype is badly abraded, comparison at generic and specific levels being hardly possible. However, the aperture can be examined and a median dorsal groove is present for a short distance just within the edge of the smooth lip. The structure was apparently interpreted as a selenizone by Northrop (1939). Specimens of Salpingostoiua, Tremanotus, and Boiotremus often show a raised dorsal band on the mould, the tremata being situated upon it in the last two named forms. Such a ridge would correspond to a groove on the inside of the shell of the type seen here. There is some suggestion of tremata on the dorsum but the evidence is inconclusive. The type of S. orientalLs from Anticosti Is. has not been examined but referal to Salpingosloma would seem justified, on the basis of Twenhofel’s (1928) description and illustrations. Acknowledgements. This study forms part of a research project on Silurian gastropods under the supervision of Prof. P. C. Sylvester-Bradley (University of Leicester). The support of a NATO Science Studentship is gratefully acknowledged. I should like to thank Porter M. Kier, Smithsonian Institution, and Norman F. Sohl, U.S. Geological Survey, for facilities extended during the tenure of a Visiting Research Associateship (1969-1970) at the U.S. National Museum of Natural History, Washington, D.C. Ellis L. Yochelson, U.S. Geological Survey, gave generously of his advice and hospitality and assisted in the borrowing of specimens. H. P. Powell, Oxford University Museum, and Katherine F. Nelson, University of Wisconsin at Milwaukee, kindly loaned specimens in their care. Mrs. N. Far- quharson and Mrs. Mary Cutting helped in the preparation of the manuscript, which also benefited from the advice of J. D. Hudson, University of Leicester. REFERENCES CLARKE, J. M., and RUEDEMANN, R. 1903. Guelpli fauna in the State of New York. Mem. N. Y. St. Mas. not. Hist. 5, 1-195, 21 pis. HORNY, R. 1962. New genera of Bohemian Lower Paleozoic Bellerophontina. Vest.Ustfed. list.geol. 37, 473-476. 1963fl. Lower Paleozoic Bellerophontina (Gastropoda) of Bohemia. Sb. geol. Ved. Pcdeontologie, 2, 57-164, pis. 1-44. 1963Z?. On the systematic position of Cyrtonelloids (Mollusca). Cos. ndrod. Miis. 132 (2), 90-94, 2 pis. (In Czech, English summary.) 1965fl. On the systematical position of Cyrtolites Conrad, 1838 (Mollusca). Ibid. 134 (I), 8-10. (In Czech, English summary.) 19657). Cyrtolites Conrad, 1838 and its position among the Monoplacophora (Mollusca). Sb. ndr. Mils. Praze, 21 (2), 57-70, pis. 1-2. 1966. Yoclielsonellis, new name for Yochelsonia Horny, 1962. Cos. ndrod. Miis., prlrod., 135 (2) 69. (In Czech, English summary.) HUPSCH, J. w. K. A. F., BARON VON. 1781. Natiirgescliic/ite des Niederdeiitscheslandes nnd anderer Gegenden, nebst haiifigen neiien Entdeckiingen nnd Beobachtungen vershiedener seltenen, merkwiirdigen nnd wenigbekannten Naturwerke. Erster Theil. Nurenberg. KNIGHT, J. B. 1941. Paleozoic gastropod genotypes. Spec. Pap. geol. Soc. Am. 32, 1-510, 96 pis. 1947. Bellerophont muscle scars. Jour. Paleont. 21, 264-267, pi. 42. KONINCK, L. G. DE. 1 842-1 844. Description des animan.x fossiles, qiii se trouvent dans le terrain carbonifere de Belgique. Liege. 422 PALAEONTOLOGY, VOLUME 15 MCCHESNEY, A. M. 1861. Descriptions of new species of fossils from the Paleozoic rocks of the Western States: (Extract two). Chicago Acad. Sci. Trans. 77-96. 1867. Descriptions of fossils from the Paleozoic rocks of the Western States, with illustrations. Ibid. 1, 1-57, pis. 1-9. MEEK, F. B. 1866. Note on the affinities of the Bellerophontidae. Chicago Acad. Sci. Proc. 1, 9-11. NORTHROP, s. A. 1939. Palcontology and stratigraphy of the Silurian rocks of the Port Daniel-Black Cape Region, Gaspe. Spec. Pap. Geol. Soc. Am. 21, 1-302, pis. 1-28. REED, F. R. c. 1920-1921. British Ordovician and Silurian Bellerophontacea. Soc. [Monogr.]: 1-92, pis. 1-13. ROLLINS, H. B. 1969. The Taxonomic position of Cyrtonella mitella (Hall) (Mollusca, Monoplacophora). Jour. Paleout. 43, 136-140. and BATTEN, R. L. 1968. A sinus-bearing monoplacophoran and its role in the classification of primitive molluscs. Palaeontology, 11, 132-140, pi. 28. STAROBOGATOV, YA. I. 1970. Systematics of early Paleozoic Monoplacophora. Paleont. Zh. 1970 (3), 293-302 (translation). TWENHOFEL, w. H. 1928. Geology of Anticosti Island. Mem. geol. Siirv. Can. 154, 1-351, 60 pis. ULRICH, E. o., and scofield, w. h. 1897. The Lower Silurian Gastropoda of Minnesota, in Geology of Minnesota. Final Kept. Minnesota Geol. Survey, 3 (2), 813-1081, pis. 61-82. WENZ, w. 1940. Ursprung und friihe Stammesgeschichte der Gastropoden. Arch. Molluskenk. 72, 1-10. WHiTEAVES, j. f. 1 895. Revision of the fauna of the Guelph Formation of Ontario, with descriptions of new species. Palaeozoic fossils, 3 (2), 45-109, pis. 9-14. WING, M. E. 1925. The Silurian Gastropoda of northeastern Illinois. Abst. of thesis, Chicago Univ. Sci. Series, 1, 311-317. YOCHELSON, E. L. 1967. Quo vadis, Bellerophon? in teichert, c., and yochelson, e. l. (eds.). Essays in Paleontology and Stratigraphy, Kansas Univ. Press, 141-161. J. S. PEEL Department of Geology University of Leicester Leicester LEI 7RH Typescript received 23 June 1971 MARCOUIA, GEN. NOV., A PROBLEMATICAL PLANT LROM THE LATE TRIASSIC OE THE SOUTHWESTERN U.S.A. by SIDNEY R. ASH Abstract. Marcouia, gen. nov. has large palmately compound leaves in which the pinnae are deeply divided into linear, lateral segments. The segments (pinnules?) contain broad midribs and parallel veins that often divide and anastomose one or more times before reaching the margins. Leaves from the Chinle Formation that were formerly referred to as Ctenis are here assigned to Marcouia. Marcouia is not classified and is known only from the lower part of the Upper Triassic Chinle Formation in eastern Arizona and western New Mexico. The purpose of this report is to redescribe some unusual pinnate leaves that occur in the Upper Triassic Chinle Formation in the southwestern United States. Previously the leaves were known only from Arizona, but recently I collected many additional speci- mens in western New Mexico. The new specimens give us so much information about this unusual leaf that a redescription is thought necessary. When these leaves were first described they were called Ctenis neuropteroides Daugherty (1941, p. 80). Their assignment to the genus Ctenis Lindley and Hutton (1834, p. 63, pi. 103), however, is untenable (Ash 1966, p. 146). In Ctenis the ultimate segments (pinnae) do not contain a midrib and the veins more or less parallel the pinnae margins (see Florin 1933, p. 81, and Harris 1964, p. 102 for recent treatments of the genus). On the other hand, the ultimate segments (pinnules) of the fossils from the southwestern United States contain broad midribs that extend to within a short distance of the pinnule apices and secondary (lateral) veins lie at a high angle and may parallel the margins for only a very short distance just before they disappear close to the margins (see text-fig. Ic). Since these fossils do not compare very closely with Ctenis or any other plant known to me, they are referred to a new genus which I will call Marcouia. The new material considered in this report has been deposited in the U.S. National Museum (USNM) in Washington, D.C. and the Museum of Northern Arizona (MNA) in Flagstaff. The specimens described earlier by Daugherty are in the University of California Museum of Paleontology (UCMP) at Berkeley. SYSTEMATIC DESCRIPTION Genus marcouia gen. nov. Type species. Marcouia neuropteroides (Daugherty) Ash, comb. nov. Diagnosis. Known portion of plant consisting of a palmately compound leaf composed of several segments (pinnae). Pinnae linear lanceolate, lamina divided into lateral seg- ments (pinnules). Pinnules arising at a high angle from sides of main rachis of pinna, oval to linear, margins wavy to lobed, apex obtusely pointed, upper margin strongly contracted, lower margin strongly decurrent on the rachis, a narrow flange of tissue [Palaeontology, Vol. 15, Part 3, 1972, pp. 423-429, pi. 80.] 424 PALAEONTOLOGY, VOLUME 15 running decurrently from one pinnule to the next below. Pinnule midrib broad, well- defined, disappearing a short distance below apex by dissolving into veins. Lateral veins numerous, arising mainly from pinnule midrib at a high angle, several arising from main rachis of pinna and entering decurrent portion of pinnule lamina, typically dividing and anastomosing one or more times with adjacent veins, usually united at margins. Derivation of name. The name commemorates Jules Marcou, the French-Swiss geologist who accom- panied the Whipple expedition through the southwestern United States in 1853-1854. He observed the petrified wood-bearing rocks (which also contain the fossils described here) in what is now Petrified Forest National Park, Arizona, and correlated them with the Keuper stage of the Upper Triassic in Furope, a correlation that is still generally accepted (Ash 1970, pp. D5-D6). Comparisons. Some of the features shown by Marcouia gen. nov. can be matched in certain other fossils that have linear pinnules. For example, in Gleuopteris Sellards (1900), Pachypteris Brongniart (1829) em. Harris (1964), and Protoblechmim Lesquereux (1880) em. Halle (1927) the pinnules have decurrent lower margins and contracted upper margins at the rachis as in Marcouia gen. nov. In the three older genera, however, the lateral veins are simple to forked and do not anastomose as they do in the present genus. The venation in the pinnules of Marcouia gen. nov. compares fairly closely with that in the pinnae of Scoresbya Harris 1932 (also see Krausel and Schaarschmidt 1968). The lateral margins of the pinnae of Scoresbya are only strongly dentate and the pinnae are not divided into distinct pinnules as in Marcouia. Recently Bock (1969, p. 231) referred to Strangerites (sometimes called Stangerites) Bornemann (1856), the fossils which Daugherty (1941) assigned to Ctenis. Although somewhat similar, those fossils and the new ones described here are distinguished from Strangerites by having anastomosing venation while the older genus has free dichoto- mous venation. Bock (1969, p. 231) also considered Pseudodanaeopsis Fontaine (1883) to be a synonym of Strangerites but Marcouia is distinguished from Pseudodanaeopsis by having strongly contracted upper margins of the pinnules at the pinna rachis while the pinnules of the older genus are attached by the whole base. Marcouia neuropteroides (Daugherty) Ash, comb. nov. Plate 80; text-fig. 1 1941 Ctenis neuropteroides Daugherty, pp. 80-81, pi. 13, fig. 3, pi. 14, fig. 2 (non 3). Holotype. UCMP 1571. Paratypes UCMP 1572, MNA P4. 102, USNM 172271, 172273. Distribution. This species has been collected from the Monitor Butte Member of the Chinle Formation at U.S. Geological Survey (USGS) paleobotany localities 10059 and 10060 in the Fort Wingate area and from the lower part of the Petrified Forest Member of the Chinle Formation at USGS paleobotany locality 10062 in Petrified Forest National Park, Arizona. Detailed data on these localities has been presented elsewhere (Ash 1970, p. D25). Emended diagnosis. Known portion of plant consisting of a palmately compound leaf, composed of several pinnae united at their bases. Pinnae linear lanceolate as a whole, large, estimated to have been 30 cm or more in length, 15 cm or more in width, lamina divided into lateral segments (here termed pinnules). Pinnules oval to linear more or less opposite, arising at a high angle (typically 60°-75°) from lateral margins of pinna rachis, linear, typically T5-2-0 cm wide, 3-5-8-5 cm long (range noted 0-6-4 cm wide, 1-10 cm S. R. ASH: PROBLEMATICAL TRIASSIC PLANT 425 TEXT-FIG. 1. Marcouia ueuropteroides (Daugherty) Ash, comb. nov. a, Reconstruction of the upper part of a pinna, approximately x J. b. Seed-like structure on the lamina of the pinnule shown on the left in PI. 80, fig. 6. MNA P4. 102, X 2. c, Apical region of a pinnule showing the venation. Note that the vein-meshes occur mainly near the margins and that the veins commonly do not divide or anasto- mose near the midrib. Drawn from a photograph of USNM 172281(7, X 5. D, Venation near the margin (on the right) of a pinnule. Note the irregularly shaped vein meshes and that the free veins frequently follow the margin for a short distance, transfer preparation USNM 172281fi, X 10. e. Epidermal cells on the lamina between veins, transfer preparation USNM 172280, x 100. f. Epidermal cells near a vein (at the right), transfer preparation USNM 172279, x 100. Specimens in b and f from USGS fossil plant locality 10062, lower part of the Petrified Eorest Member, Chinle Eormation, Petrified Forest National Park, Arizona. Those in c-e from USGS fossil plant locality 10061, Monitor Butte Member, Chinle Formation, Fort Wingate area, New Mexico. 426 PALAEONTOLOGY, VOLUME 15 long). Margins wavy to lobed, sometimes toothed, apex obtusely pointed, upper margin strongly contracted at the rachis, lower margin strongly decurrent, narrow flange of lamina running decurrently from each pinnule along the lateral margins of pinna rachis, joining lamina of next pinna below. Near pinna apex, division of lamina into separate pinnules incomplete and a series of short rounded segments decreasing in size toward tip is usually present. Pinnule midrib well defined, 1-3 mm broad, arising at an angle of 30°-40° to the pinna rachis, typically bending outward near the base to an angle of about 60°-75°, then following a more or less straight course disappearing 1-2 cm below apex by dissolving into lateral veins. Lateral veins numerous, slender, about 0-1 mm wide, arising mainly from pinna midrib at a high angle (70°-80°), 1-5 veins arising from pinna rachis and entering decurrent portion of lamina, often dividing near base, typically anastomosing one or more times with adjacent veins in marginal area forming elongated, irregular meshes. Lateral veins usually united at margins, rarely free, free vein endings occasionally following margin for a short distance. Tracheids of lateral veins showing annular, helical, and scalariform thickenings. Epidermal cells rectangular to square, rarely polygonal, 35-80 fxm wide, 46-120 i^m long, rectangular cells usually adjacent to veins with long axis oriented parallel to veins, elsewhere cells more nearly square and irregularly oriented, anticlinal cell walls fairly straight, about 2 jum thick. Stomata oval, scattered sparse, guard cell pair about 45-60 /xm in diameter. Discussion. A complete or even nearly complete leaf of M. neuropteroides is not known. Most examples consist of just fragments of the pinna rachis with a few attached pinnules. Only one shows the base of the leaf (see PI. 80, fig. 7) and it is poorly preserved. The fossil consists of the remains of four or five pinnae which are clearly joined at their EXPLANATION OF PLATE 80 Figs. 1-9. Marcouia neuropteroides (Daugherty) Ash, comb. nov. All X 1 . 1 , A nearly complete pinnule that bears a large oval structure below the midrib which may be the remains of a seed. This structure, however, is almost twice as large as those in figure 6 and also is below, not above, the midrib. Note the wavy to lobed margins in this specimen. USNM 172271. 2, Main rachis of a pinna bearing several fragmentary pinnules showing contracted upper margins and decurrent lower margins at the rachis, USNM 172272. 3, 4, Apical region of two pinnae showing characteristic lobingof the lamina. Note that in figure 3 the uppermost pinnule is broadly fused with the apical lamina. In contrast the next lower pinnule is nearly free except for a narrow band of tissue that runs decurrently along the rachis to the apical lamina, USNM 1 72273 and 1 72274. 5, Apex of a sterile pinnule in which the venation is particularly clear, USNM 172275. 6, Two pinnules bearing the remains of oval seed-like structure on the lamina. Compare with tf. IB. MNA P4. 102. 7, Two nearly complete and unusually small pinnae. Note that the rachises and midveins are nearly the same size as those in the larger examples shown on the plate but the lamina is much smaller. Here again the decurrent lower margins and contracted upper margins of the pinnules are clearly visible, USNM 172276A. 8, Por- tion of the main rachis of a pinnae bearing the remains of several pinnules, one of which is compound. The contracted upper margin at the rachis is evident in the uppermost pinnule on the left, USNM 172277. 9, The base of a palmately compound leaf composed of five pinnae, USNM 172278. Specimens in figures 1-5, 7, 8 from USGS fossil plant locality 10060 in the Monitor Butte Member of the Chinle Formation near Fort Wingate, New Mexico. The specimen in figure 9 is from USGS fossil plant locality 10059 in the Monitor Butte Member of the Chinle Formation in the Fort Wingate, New Mexico area. The specimen in figure 6 is from USGS fossil plant locality 10062 in the lower part of the Petrified Forest Member of the Chinle Formation in Petrified Forest National Park, Arizona. Palaeontology, Vol. 15 PLATE 80 ASH, Problematical Triassic plant S. R. ASH: PROBLEMATICAL TRIASSIC PLANT 427 bases, demonstrating that the leaf is palmately compound. It shows a remarkable resemblance to the base of Scoresbya dentata Harris (1932, pi. 2, fig. 9). The small pinna to the left in PI. 80, fig. 7 is probably nearly complete. It is about 2 cm wide and originally may have been 6 cm or more in length. The rachis bears 5 pairs of unusually small lateral pinnules and an apical lamina. A larger (7 cmx 13 cm) but very poorly preserved and less complete pinna (USNM 172282) shows 4 pairs of average-size pinnules together with the remains of an apical lamina. About 40 fairly complete pinnules of M. neiiropteroides are now known. They vary somewhat in outline and show a wide range in size. Pinnules from the Petrified Forest are often 3-5 to 4-0 cm wide and 6 to 8 cm long, although some are as much as 5 cm wide and 1 1 cm long. Usually they are twice as long as wide and are somewhat oval in outline (see Daugherty 1941, pi. 14, fig. 2). In comparison, the pinnules from Fort Wingate typically are 2 to 2-5 cm wide and 7 to 9 cm long. Exceptionally small pinnules (5 mmx 18 mm) are preserved in one example (see PI. 80, fig. 7). Most of the pinnules from the Fort Wingate area are three times (or more) as long as wide and are distinctly linear in outline (see PI. 80, fig. 1). Small, round to oval structures which may be the remains of seeds occur on the laminae of three pinnules of M. newopteroides. Two of the pinnules bearing such struc- tures are attached opposite each other on a short length of the main rachis of a pinna (PI. 80, fig. 6). The structures show as small (about 2 mm X 3 mm) oval gaps surrounded by narrow bands of carbonaceous material. Both structures occur above the pinnule midribs. One is located about half-way between the base and apex of the pinnules while the other is somewhat closer to the base than to the apex of the other pinnule. Both more or less parallel adjacent veins (see text-fig. 1b and PI. 80, fig. 6). A third pinnule bears a similar but longer (2 mm X 4 mm) structure. It also seems to be surrounded in places by a carbonaceous band within which there is an oval gap. However, in contrast to the other oval structures this one occurs below, not above, the midrib and lies practically transverse to the surrounding veins with one end nearly touching the midrib (PL 80, fig. 1). Although there is always the possibility that these structures are merely aceidents of preservation or are pathologic, their regularity and definite outline suggests not. In addition, the gaps on the lamina clearly suggest the sedimentary filling of hollow struc- tures and the surrounding bands of carbonaceous material could be the compressed walls or shells of seeds. Nevertheless, since so few of these structures have been seen and none clearly shows definite seed features, such as a micropyle, stony layer, etc., they are regarded as only problematical structures which could be seeds. Although the veins are clear in most specimens, the substance of the laminae has almost completely disappeared and it is impossible to make a typical cuticle preparation from them. A few acetate film transfers, however, do show anticlinal walls of the epidermal cells rather faintly. The stomata are not preserved on the transfers, but there are oval spaces scattered among the epidermal cells. It is thought that they mark the position of the guard cells which have totally disappeared. Vascular tissue is preserved in many of the veins on the transfers and the thickenings on the side walls of the tracheids are clearly visible in some. Comparisons. M. newopteroides can be distinguished from Strangerites obliqims Emmons F f C901C 428 PALAEONTOLOGY, VOLUME 15 (1857, pp. 121-122, fig. 89) from the Newark series of the eastern United States by having anastomosing venation whereas the venation in the older species is free dichoto- mous. The pinnules of M. neuropleroides can be differentiated from Pseiidodanaeopsis nervosa Fontaine (1883, pp. 61-63, pi. 31, figs. 1-2) by having wavy to toothed lateral margins and contracted upper margins at the rachis while the pinnules of P. nervosa have smooth lateral margins and the upper margin is not contracted at the rachis. Further- more, the venation in M. neuropleroides is anastomosing, not free dichotomous, as in P. nervosa. The fossils described in this paper can be easily distinguished from the specimen described from the Newark series by Bock (1969, pp. 236-239, fig. 392) as St anger ites obliqua. According to Bock the lateral veins in his fossil frequently divide very near the midrib and anastomose fairly regularly in the vicinity of the margins forming a ‘W’ or ‘X’ pattern. Also, the lateral margins are straight to curving and are slightly constricted at the rachis. In M. neuropleroides the lateral veins rarely divide near the midrib and anastomose very irregularly near the lateral margins which are wavy to lobed. Also, the upper margin is strongly contracted at the rachis and the lower is decurrent. M. neuropleroides can be distinguished from the fossil (USNM 8208) called Siranger- iles planus Emmons (1857, p. 122) and the fossil (USNM 30958) referred to Pseudo- danaeopteris reticulata Fontaine (1883, pp. 59-60, pi. 30) by having short pinnules with wavy to lobed or toothed margins and strongly contracted upper margins at the rachises. In the two older fossils the pinnules are comparatively long with smooth margins and have slightly constricted upper margins at the rachises. The specimens described here can also be differentiated from the fossils Bock (1969, figs. 385-387) called Stangerites planus by having wavy to lobed or toothed margins, strongly contracted upper margins, and strongly decurrent margins at the rachis. The pinnules of Bock’s fossils have smooth margins, slightly contracted upper margins, and slightly decurrent margins at the rachis. Additionally, the veins in M. neuropleroides do not anastomose as regularly as they do in the fossils Bock calls Stangerites planus. Classification. Daugherty (1941, p. 80) considered this plant to be a bennettite. However, the anticlinal walls of its epidermal cells are not sinuous as they usually are in the bennettitales. Marcouia is not classified at this time because we have no clear evidence of its true affinities. Acknowledgements. I am grateful to Dr. Chester A. Arnold, University of Michigan, for critically reviewing this report and to Dr. Francis M. Hueber of the U.S. National Museum who loaned me the fossils from the Newark Series that were referred to Pseiidodanaeopsis by Fontaine. Research on the project was supported by National Science Foundation Grant GA-25620. REFERENCES ASH, s. R. 1966. The Upper Triassic Cliinle flora of the southwestern United States. Ph.D. thesis, Univ. Reading, 212 pp., 11 pis. 1970. Ferns from the Chinle Formation (Upper Triassic) in the Fort Wingate area, New Mexico. Prof Pap. U.S. Geol. Siirv. 613-D, 1-52, 5 pis. BOCK, w. 1969. The American Triassic flora and global distribution. GeoL Center Res. Ser. 3/4, 406 pp. BORNEMANN, J. B. 1 856. Uber organische Reste der Lettenkohlengriippe Thiiringens. Leipzig, 85 pp., 1 2 pis. S. R. ASH: PROBLEMATICAL TRIASSIC PLANT 429 BRONGNIART, A. 1829 (1828-1838). Histoire des vegetaux fossiles, on recherches botaniques et geologiqiies siir les vegetaux renfevmes dans les diverses couches du globe. Paris. G. Dufour and Ed. D’Ocagne, 2 vols. DAUGHERTY, L. H. 1941. The Upper Triassic flora of Arizona. Carnegie Inst. Wash. Publ. 526, 108 pp., 34 pis. EMMONS, E. 1857. American Geology, pt. 6 (3). Albany, N.Y. 155 pp. ELORiN, R. 1933. Studien liber die Cycadales des Mesozoikums nebst Erorterungen liber die spaltdff- nungsapparate der Bennettitales. K. svenska Vetensk-Akad. Handl. 12 (5), 134 pp., 16 pis. FONTAINE, w. M. 1883. Contributions to the knowledge of the older Mesozoic flora of Virginia. Mon. U.S. Geol. Survey, 6, 144 pp., 54 pis. 1900. Status of the Mesozoic floras of the United States. First paper: The older Mesozoic. U.S. Geol. Survey 20th Annual Kept. pt. 2, 213-748, pis. 21-179. HALLE, T. G. 1927. Paleozoic plants from central Shansi. Paleontologica Sinica, ser. A, 2, pt. 1,316 pp., 64 pis. HARRIS, T. M. 1932. The fossil flora of Scoresby Sound, east Greenland. 2. Medd. Gronland, 85 (3), 1 12 pp., 9 pis. 1964. The Yorkshire Jurassic Flora. II. Caytoniales, Cycadales, and Pteridosperms. Brit. Mus. (Nat. Hist.). 212 pp., 7 pis. KRAUSEL, R., and SCHAARSCHMIDT, F. 1968. Scoresbyo Harris (Dipteridaceae ‘i’) aus dem unteren Jura von Sassendorf. Palaeontographica, B 123, 124-131, 9 figs., 1 pi. LESQUEREUX, L. 1880. Description of the coal flora of the Carboniferous formation in Pennsylvania and throughout the United States. Penn. 2nd Geol. Survey, Kept. Progress, 2 vols., atlas, 1879. LiNDLEY, j., and HUTTON, w. 1831-1837. The fossil flora of Great Britain. 3 vols. Eondon. SELLARDS, E. H. 1900. A new genus of ferns from the Permian of Kansas. Kansas Univ. Quart. 9, 179-189, pis. 37-42. SIDNEY R. ASH Weber State College Typescript received 17 May 1971 Ogden, Utah, U.S. A. THE UPPER DEVONIAN SALTERN COVE GONIATITE BED IS AN INTRAFORMATIONAL SLUMP hy PETER VAN STRAATEN and MAURICE E. TUCKER Abstract. The Saltern Cove Goniatite Bed, which contains upper Frasnian conodonts, ostracods, and gonia- tites, is shown also to contain blocks of Famennian limestone and to lie within a sequence of Famennian sediments. From the sedimentology, it is considered that Famennian limestone clasts were reworked into Frasnian muds which were then transported en bloc into Famennian sediments. A Schwellen area (rise) nearby is suggested by the nature of the derived limestone blocks. Saltern Cove (Grid Ref. SX 895585) 5 km south of Torquay, south Devon (text- fig. 1) is a classic locality for Upper Devonian goniatites. It was first described by Lee (1877) and subsequently by Ussher (1903), Anniss (1927), Lloyd (1933), Donovan (1942), and House (1963). The goniatites are now ascribed to the holzapfeli Zone of the Frasnian (IS), Upper Devonian. Recent work by the authors (briefly reported in Tucker and van Straaten 1970n) has shown that the stratigraphical context of the goniatites is not quite so simple as would first appear. Previously, all the Upper Devonian in Saltern Cove and Waterside Cove has been considered to be of Frasnian age (apart from the last 20 to 30 m at the north end of Waterside Cove, which were attributed to the lower Famennian). Results presented here show that all the succession from the north end of Saltern Cove is Famennian in age, though with several beds of older intraformational derived sediment. These include the Goniatite Bed. STRATIGRAPHY AND STRUCTURE The Devonian in South Devon is tectonically complicated and in the Torquay area Richter (1969) has recognized four phases of structural deformation. In the Saltern Cove area the bedding is near vertical and youngs northwards. There are several small faults and a strong cleavage is present, which is almost horizontal or dips south-east at a few degrees. In the central part of Saltern Cove a 3 m thick bed of massive limestone (dated on coral evidence as middle Frasnian by Scrutton 1965) is succeeded by about 20 m of alternating limestone and red shale bands. The limestones are composed of corals, crinoids, and other carbonate debris, and represent the talus from a carbonate-producing area nearby. The northern part of the bay is occupied by some 25 m of purple and red mudstones, with rare crinoidal limestones, which are of lower Famennian age (Chei/o- ceras Strife) at the top. These are terminated by a fault running ESE.-WNW. which forms the northern side of the cove (text-fig. 1). The succession (with dated ranges) from this fault to the north end of Waterside Cove is (youngest beds, most northern, at the top): [Palaeontology, Vol. 15, Part 3, 1972, pp. 430-438.] VAN STRAATEN AND TUCKER: SLUMPED GONIATITE BED 431 TEXT-FIG. 1. Geological sketch-map of Saltern Cove and Waterside Cove, South Devon (after mapping by the authors), showing location of conodont samples, 1, 2, 3, and 24. 432 PALAEONTOLOGY, VOLUME 15 Lower Devonian Fault — ■ — 40 m 1 m 7 m 1 m 5 m 20 m Purple mudstones with tuff and shell bands (cen- tral part of Waterside Cove). Limestone conglomerate (described by Holwill 1966). Red shales with nodules and tuff bands (southern part of Waterside Cove). Red shale with numerous derived limestone and tuff fragments. The Saltern Cove Goniatite Bed. Red silty shales with many limestone clasts (most southern part of Waterside Cove and promontory between Water- side and Saltern Cove). Red and purple shales with many tuff bands (coastal platform between Waterside and Saltern Cove). U^-lUoc clasts: IIj8-IIIa, I, Givetian Il/3-lIIa ll/S-IIIa shales: 18 clasts: IljS-IIIa Il^-IIIa Fault forming northern side of Saltern Cove AGE DETERMINATIONS IN THE VICINITY OE THE GONIATITE BED Beds above and below the Goniatite Bed Samples from tuffs and calcareous nodules collected above and below the Goniatite Bed all yield conodonts of the quadrantinodosa Zone (Il^-IIIa, upper CheilocerasAovjQV Platyclynienia, Famennian). Location of the samples is shown in text-figs. 1 and 2, and conodont determinations in text-fig. 3. Limestone clasts from the conglomerate (Holwill 1966) give a range of ages; blocks with corals are Givetian-lower Frasnian (Holwill 1966), whilst fine-grained clasts give Frasnian and lower Famennian ages on conodonts (samples 18, 19, and 20). Immediately below the Goniatite Bed, and cropping out on the south face of the promontory, is a thin but continuous calcareous band (sample 6) which is particularly rich in conodonts (often broken), thin-shelled bivalves, crinoid fragments, and ostracods. The ostracods, with their stratigraphic range after Blumenstengel (1965) are Acratia spp., Arnphissites bispinosus(upper C/ieiloceras-P!atycIyinenia,}\^-lU), and Ceratacratia cerata (Famennian, Ila-VI). Purple mudstones at the north end of Saltern Cove (sample 1, text-fig. 1) yielded Entoinozoe (Nehdentoniis) nehdensis Matern 1929 indicative of the lower Famennian (Cheiloceras Strife). The Goniatite Bed {a) Shales. The goniatite fauna, recently revised by House (1963), belongs to the holzapfeli Zone (18) of the upper Frasnian. The fauna is dominated by the genus Archoceras. The goniatites are usually a centimetre or less in diameter and are haema- tized. Most have been obtained from the shales in the cliff section just north of the promontory (text-fig. 2). Goniatites also occur on the south face of the promontory VAN STRAATEN AND TUCKER: SLUMPED GONIATITE BED 433 (M. R. House, pers. comm., 1969). In addition, the authors have found goniatites at the same horizon in the vicinity of the large tuff block figured by Holwill (1966), 10 m west of the promontory. Other fossils present in the shales include Buchiola spp., ortho- cones, trilobites, brachiopods, crinoids (Anniss 1927), and haematized ostracods identi- fied as Eutomoprimitia spp. (upper Frasnian) and Entomozoe (Nehdeiitomis) tenera Rabien 1954 (Frasnian-lower Famennian, I-Il). Conodonts, occurring in small ‘reduc- tion’ centres, were obtained by breaking open shales. Generally the conodonts are rather TEXT-FIG. 2. Sketch of the promontory between Saltern Cove and Waterside Cove, viewed obliquely, showing location of conodont samples 4 to 23. more common than the ammonoids. Of some 50 collected, half were identifiable, and indicated an upper Frasnian age: Palmatolepis gigas Miller and Youngquist 1947, Pahnatolepis subrecta Miller and Youngquist 1947, Palmatolepis triangularis Sannemann 1955, and Polygnathus spp. The goniatites, ostracods, and conodonts all indicate a similar age. No Famennian conodonts were found. {b) Calcareous clasts ^vitliin the Goniatite Bed. Within the Goniatite Bed are a number of angular and rounded limestone blocks (sample numbers 7 to 16, text-fig. 3). Most are about 20 cm x 10 cm x 10 cm, but some reach up to 2 m x 50 cm x 30 cm. The block of ‘calcareous tufl'’ figured by Holwill (1966) measures 3 mx 1 mx50 cm (sample 8). All yielded conodonts (by formic acid digestion) of the c/uadrantinodosa Zone ( upper lower Ilia) or rhomboidea-quadrantinodosa Zones (Il^-lower Ilia). Other fossils found in the limestones were crinoid fragments, orthocones, ostracods, gastropods, and one goniatite (species indeterminate). The Goniatite Bed (red shales with limestone clasts) and the limestone conglomerate can also be found at the north-west corner of Saltern Cove. Conodonts from limestone blocks here (samples 2 and 3) again show the presence of the lower Famennian. 434 PALAEONTOLOGY, VOLUME 15 STRATIGRAPHICAL INTERPRETATION The data presented above show the apparently anomalous situation in the Goniatite Bed where an argillaceous sediment older than the adjacent shales contains limestone blocks which are two whole goniatite zones younger than the enclosing sediment. There are three possible explanations: 1. Mis-coirelation of goniatites and conodonls. The ostracods and conodonts in the shales are the same age as the goniatites. This shows that there is no local mis-correlation Ancyrodelta rotundiloba BRYANT 1921 Icriodus spp. Polmatolepis distorla BRANSON a MEHL 1934 Pal glabra lepta ZIEGLER 8 HUDDLE 1969 Pal. glabra pectmata ZIEGLER i960 Pol. glabra pnma ZIEGLER a HUDDLE 1969 Pol. gracilis gracilis BRANSON a MEHL 1934 Pal. minuto minuta BRANSON 8 MEHL 1934 Pol perlobolo schmdewolfi MULLER 1956 Pol quodrantinodosa inflexa MULLER 1956 Pol quodronUnodoso hflexoidea ZiEGLER 1962 Pal quodrantinodosa margmifera ZIEGLER 1960 Pal spp. Polygnathus diversa HELMS 1959 Pol. glabra glabra ULRICH a BASSLER 1926 Pol. nodocostofa s\ BRANSON 8 MEHL 1934 Pol nodocosloia cf. pennaluloidea HOLMES 1928 Pol. nodoundaio HELMS I96i Pol Iriphyllola ZIEGLER 1960 Pol spp. Polylophodonta confluens ULRICH 8 BASSLER Spothognothodus amplus BRANSON 8 MEHL Spalhognathodus spp. TEXT-FIG. 3. Conodont determinations from Saltern Cove and Waterside Cove, South Devon. 0 = cf. determination. The specimens are deposited in the archives of the Department of Geology, University of Reading, where they are numbered S22901-22924. The conodont zones represented by the samples (after Ziegler 1962, 1965, and Klapper 1966) are as follows: Sample 20 asymmetrica Zone (la). Samples 1, 5, 7, 11, 12, 15, 16 rhomboidea-quadrantiiioclosa Zones (ll)3-lower Ilia). Samples 3, 4, 6, 8, 10, 17, 18, 19, 21, 24 quadrantinodosa Zone (upper Il/S-lower Ilia). Samples 13, 14, 22 lower quadrantinodosa Zone (upper IIj8). Sample 2 qnadrantinodosa-mxAA\Q velifera Zones (upper 11(8-111^). Sample 9 upper triangidaris-uppQt: velifera Zones (IS-III^). of the goniatite and conodont chronologies and that the German system (Ziegler 1962) is applicable to the area. Indeed, at Chudleigh, 30 km NNW. of Saltern Cove, the conodont succession from the Clieiloceras Strife is essentially in agreement with the German chronology and with the goniatite succession (Tucker and van Straaten 1970/)). 2. A complication of the tectonics. From the bedding/cleavage relationship around the promontory (bedding vertical, cleavage horizontal) a fold cannot possibly be constructed so that the Goniatite Bed occurs in the centre of a tight anticline, unless such a fold was developed during an earlier deformational phase. However, the beds above and below the Goniatite Bed show normal grading (northwards), indicating that early folding did not take place. From text-fig. 4 it is clear that faulting and/or shearing has not occurred around the Goniatite Bed. The tectonic deformation is only such that the limestone VAN STRAATEN AND TUCKER: SLUMPED GONIATITE BED 435 blocks have been rotated by the cleavage, and once continuous ash bands are sheared. A complication of the tectonics then is most unlikely to have caused these Frasnian shales to occur in a Famennian succession. 3. Large-scale transportation of sediment. A sedimentary explanation means that the youngest elements in the Goniatite Bed, the limestone clasts, were reworked into older argillaceous sediments and that the whole was then transported en bloc into a muddy TEXT-FIG. 4. The Saltern Cove Goniatite Bed. Location of samples indicated. succession indistinguishable in age from the clasts. The presence of goniatites, conodonts, and ostracods all of the upper Frasnian, and the absence of Famennian goniatites and conodonts within the shales of the Goniatite Bed, indicate transportation en bloc of the Frasnian muds (with their Famennian limestone clasts) rather than a simple reworking of the Frasnian fauna into Famennian muds. Evidence that this was the case can be seen from an examination of the sedimentology of the succession. LAMINATION OF SEDIMENTS. A petrographic examination of the shales of the Goniatite Bed shows that they are quite distinct from the shales above and below in one important respect, that good lamination is absent. In spite of the strong cleavage, thin continuous laminae of medium to coarse silt, 0-5 mm thick, occur in the shales above and below. In the goniatite shales, however, the same amount of silt is present but is either irregularly distributed or present in discontinuous lenses. Numerous graded tuff bands, from a few millimetres to 40 cm thick, occur at intervals of 20-30 cm above and below the Goniatite t 436 PALAEONTOLOGY, VOLUME 15 Bed, and can be traced for up to 70 m across the wave-cut platform. A small channel structure is also exposed here. However, in the shales of the Goniatite Bed macroscopic laminations of this type are completely absent. It is suggested that transportation of the sediment destroyed its primary lamination. There is no evidence of any erosional struc- ture between the Goniatite Bed and the underlying laminated shales. LIMESTONE CLASTS. The blocks of Famennian limestone and tuff are clearly derived. There is some concentration of clasts towards the base of the Goniatite Bed, but many occur randomly scattered throughout the shale. The limestones are micritic with an appreciable amount of silt. Some are almost completely dolomitized. Prominent pressure solution planes (stylolites) occur in some blocks and there are also cavity-fill structures. A few clasts contain a network of cavities filled first by fibrous calcite, and then silty shale material. Many of the limestone blocks are reminiscent of the condensed pelagic facies (Schwellen facies, Cephalopodenkalk) which at Chudleigh, and commonly in Germany, follows Givetian/lower Frasnian shallow-water massive limestones (Tucker unpub. thesis, Reading 1971). The events which led to the formation of the Goniatite Bed are as follows: 1. Deposition of Frasnian muds (containing goniatites, conodonts, and ostracods) perhaps on the slopes of a Schwelle (submarine rise). 2. Deposition of Famennian limestones on a Schwelle. 3. Reworking of clasts of the Famennian limestone into the Frasnian muds. 4. En bloc transportation of the Frasnian sediments with clasts of Famennian lime- stone into Famennian muds. Another band with many tuff and limestone fragments occurs immediately above the Goniatite Bed, and conodonts (sample 17) suggest that the clasts are of the same zonal age as the shales. The limestone conglomerate (7 m above the Goniatite Bed) described by Holwill (1966) contains limestone fragments of various types and ages. The bed is graded and Holwill suggested a proximal turbidity current origin. COMPARISONS Three types of sediment transportation have occurred at Saltern Cove, (a) Of more frequent occurrence, movement of lithified shallow-water sediments (limestones) down- slope into a deeper water environment, where the transported material is generally older than the host sediments. An example of this is the limestone conglomerate of Holwill (1966), mainly composed of Givetian and Frasnian clasts, but with some micritic lime- stones of Famennian age (samples 18 and 19). (b) Transportation of lithified shallow- water sediments downslope where the reworked clasts are of the same zonal age as the deeper water sediments. An example of this is the bed of reworked tuff and limestone fragments occurring immediately above the Goniatite Bed (sample 17). Other bands containing calcareous clasts occur higher up the succession in Waterside Cove and are probably also of this type. Slumping of this second type, with no stratigraphical break, is typical of most intraformational conglomerates and is known in a similar situation in the Upper Devonian of the Harz Mountains, Germany (Stoppel and Zscheked 1963). (c) Eu bloc transportation of argillaceous sediments which are older than the host sediments, e.g. shales of the Goniatite Bed. Slumping of the third type involving transportation of muds VAN STRAATEN AND TUCKER: SLUMPED GONIATITE BED 437 is less common and may be difficult to prove: it is detectable only where either the stratigraphical break is large enough, or sufficiently fine dating is possible. In the absence of the latter, only the lack of lamination might give a clue to the detailed history. SIGNIFICANCE OF SLUMPING AT SALTERN COVE The two pelagic facies of the Upper Devonian, the condensed limestones (Schwellen or rise facies) and ostracod shales (Becken or basinal facies) are developed in south-west England, though the rise facies is only well known at Chudleigh (House 1963, Tucker and Straaten 19706). However, it may be present at Petit Tor Combe (Grid Ref. SX 926665, 4 km north-east of Torquay) where patches of nodular limestone containing middle Frasnian goniatites occur (Ussher 1890; Lloyd 1933; and House 1963). The basinal facies is well developed in the Torquay district. The blocks of Famennian lime- stone at Saltern Cove indicate that pelagic carbonate was deposited in the vicinity on a topographic high. In the Upper Devonian of Germany condensed pelagic limestones occur above submerged ‘reefs’, on volcanic ridges and basement rises (geanticlines) within the Rhenish Geosyncline. The Torquay and Brixham shallow-water limestones, or the massive limestone immediately to the west at Goodrington may have been positive areas during the Famennian where pelagic carbonate accumulated. A slope facies can be recognized in the German Devonian in the transitional sediments between Schwellen and Becken. The slope facies consists mostly of shales with bands of calcareous nodules, but is characterized by the presence of slumped or reworked horizons. The Saltern Cove Goniatite Bed shows that a similar situation of submarine rise and slope existed during the Upper Devonian of this area. Holwill (1966) considered volcanism as the immediate cause for the limestone con- glomerate. However, for the Goniatite Bed, instability on a slope and movement of the underlying sediment is as likely an explanation, although the movement may have been initiated by volcanic activity. CONCLUSIONS 1. The Saltern Cove Goniatite Bed is a transported deposit of Frasnian shales con- taining Famennian limestone blocks, developed within a sequence of early Famennian sediments. The sedimentology of the finer-grained sediments provides a clue to the stratigraphical interpretation. The type of slumping is unusual since en bloc movement of shales has occurred, which are older than the contained limestone clasts and adjacent sediments. 2. The derived Famennian pelagic limestones indicate the former existence of a Schwellen area. 3. Where tested, the German goniatite, conodont, and ostracod zonation is applicable, and confirmed in South Devon. Acknowledgements. We are very grateful to Dr. Helga Gross (Gottingen) for determining the ostracods, and to Dr. R. Goldring and Professor M. R. House for criticizing the manuscript. 438 PALAEONTOLOGY, VOLUME 15 REFERENCES ANNiss, L. G. 1927. The Geology of the Saltern Cove Area, Torbay. Q. Jl geol. Soc. Land. 83, 492-500. BLUMENSTENGEL, H. 1965. Zur Taxionomie und Biostratigraphie verkieselter Ostracoden aus dem Thiiringer, Oberdevon. Freiberger Forschwigshefte, C183, 127 pp. DONOVAN, D. T. 1942. Specles of Archoceras from Saltern Cove, Devon. Proc. Bristol Nat. Soc. 9, 375-380. HOLWiLL, F. j. w. 1966. Conglomerates, tuffs and concretionary beds in the Upper Devonian of Waterside Cove, near Goodrington Sands, Torbay. Proc. Usslier Soc. 1, 238-241. HOUSE, M. R. 1963. Devonian ammonoid successions and facies in Devon and Cornwall. Q. Jl geol. Soc. Loud. 119, 1-27. KLAPPER, G. 1966. Upper Devonian and Lower Mississippian conodont zones in Montana, Wyoming, and South Dakota. Univ. Kansas Paleont. Contr. 3, 1-43. LLOYD, w. 1933. The geology of the country around Torquay. (2nd edn.) Mem. Geol. Surv. U.K. RICHTER, D. 1969. Structures and Metamorphism of the Devonian Rocks south of Torquay, S.E. Devon (England). Geol. Mitt. Aachen, 9, 109-178. SCRUTTON, c. T. 1965. The ages of some coral faunas in the Torquay area. Proc. Usslier Soc. 1, 186-188. STOPPEL, D. and zscheked, j. g. 1963. Friihdiagenetische Sedifluktionen im Mittel und Oberdevon des West Harzes. Berlin Naturhist. Ges. 107, 5-18. TUCKER, M. E. and STRAATEN, p. VAN. 1970(7. Conodoiits from the Upper Devonian of the Saltern Cove- Elberry Cove area (Abstract). Proc. Usslier Soc. 2, 159. 19706. Conodonts and Facies on the Chudleigh Schwelle. Ibid. 2, 160-170. ussher, w. a. e. 1890. The Devonian rocks of South Devon. Q. Jl geol. Soc. Loud. 46, 487-517. 1903. The Geology of the country around Torquay. Mem. Geol. Sitrv. U.K. ZIEGLER, w. 1962. Taxionomie und Phylogenie oberdevonischer Conodonten und ihre stratigraphische Bedeutung. Abli. liess. Landesamt. Bodenforscli. Wiesbaden, 38, 166 pp. 1965. Eine Verfeinerung der Conodontengliederung an der Grenze Mittel- Oberdevon. Fortsclir. Geol. Reiiikl. Westf. 9, 647-676. PETER VAN STRAATEN Geologisch-Palaeontologisches Institut Gottingen W. Germany MAURICE E. TUCKER Department of Geology Fourah Bay College University of Sierra Leone Typescript received 10 June 1971 Freetown, Sierra Leone PTYCHODUS PREDATION UPON A CRETACEOUS INOCERAMUS by E. G. KAUFFMAN Abstract. The type specimen of Inoceramus tenuis Mantell displays probable tooth impressions of Ptychodus on the ventral margin of the left valve. The predation is interpreted, and the general problem of vertebrate predation on Inoceramidae, which were the dominant Cretaceous bivalves, is discussed. Bivalves of the families Inoceramidae and Ostreidae dominate the epifaunal benthos of many non-Tethyan Cretaceous assemblages and are an important element of nearly all of them (e.g. Kaulfman 1967, 1969). As such they should have been a primary food source for a variety of contemporary shell-feeding vertebrates; skates, rays, pycnodonts, shell-crushing sharks like Ptychodus, and reptiles like the mosasaurian Globidens. Most Cretaceous ostreids were cemented epizoans with thickened, hard calcitic shells. They primarily attached to hard substrates which were difficult feeding areas for preda- tory vertebrates adapted to sifting softer sediments in search of bivalves, echinoids, and similar food. Thus, the Inoceramidae, epifaunal echinoderms, and various infaunal organisms were probably their principal prey. This is consistent with the inferred range of inoceramid adaptations and habitats, which were primarily related to unconsolidated and semiconsolidated substrates, with or without weak byssal attachment (Kauffman 1965, 1968, 1969). Surprisingly, no definite cases of marine vertebrate predation on Inoceramidae have been previously documented. The impression that such predation was an important factor in the ecology and fossil preservation of inoceramids is purely circumstantial — selective marginal shell fragmentation and highly broken shell beds in quiet water deposits, marginal fragmentation of bivalved (living?) shells, common association of Ptychodus teeth with highly fragmented inoceramid assemblages, and common healed shell injuries which resemble the impressions of blunt teeth. This will be subsequently discussed. The discovery of two parallel rows of blunt tooth impressions along the ventral margin of the type of Inoceramus tenuis, with spacing and size characteristics closely matching those of known jaw assemblages of the contemporaneous shell crushing shark, Ptychodus, is the first good evidence for vertebrate predation on Inoceramus and is described below. PTYCHODUS PREDATION IN INOCERAMUS TENUIS Inoceramus tenuis Mantell occurs sparsely through chalks and marls of the Lower Cenomanian Hypoturrilites carcitanensis Zone and extends upward into the Middle Cenomanian Turrilites costatus Zone of Kennedy (1969). The type is from the Chalk Marl at Hamsey, Sussex (Woods 1911, p. 271, figs. 31, 32: PI. 81, figs. \a, b herein). These are all levels where fragmentation of Inoceramus shells (primarily within the Inoceramus ? crippsi lineage) is common. The holotype (PI. 81, fig. \a, b) is completely [Palaeontology, Vol. 15, Part 3, T972, pp. 439-444, pi. 81.] 440 PALAEONTOLOGY, VOLUME 15 uncrushed except for a series of moderately deep, subrounded, subequally spaced indentations — circular zones of crushing which do not perforate the shell — extending in two parallel rows across the mid-ventral and postero-ventral margins of the left valve (PI. 81, fig. \a, left lateral view; fig. \b, diagonal left lateral view). The innermost row of three indentations is the best-developed. It is situated 10 to 20 mm in from the commissure, with tooth impressions 15-17 mm in diameter, and their centres 16 and 17 mm apart (measurements taken left to right along axis of row, PI. 81, fig. \b). The central impression is very slightly ofifset toward the ventral margin of the shell from the axis of the row (defined by terminal impressions). At least two very shallow, somewhat more irregular impressions lie in a second row marginal to the first, one each below and slightly oflset to the right of the middle and right-hand impressions of the main row (as viewed in PI. 81, fig. \b). These have diameters of 13 and 15 mm respectively; the distance between them is 18 mm. Significantly, the margin of the left valve is broken off just below the zone of tooth impressions, including the area where additional impressions of the second (marginal) row would be expected to occur (left side of row, fig. 16), and nowhere else on the shell. The shape, size, and subregular arrangement of the impressions indicate a single bite by a vertebrate predator with rounded, elevated tooth crests which were closely spaced but not in contact. The teeth occurred in parallel rows (trending left to right in PI. 81, fig. 16) but adjacent teeth were moderately offset from one row to the next; some within each row were slightly out of line. The impressions were made by teeth near or at the lateral edge of a jaw apparatus. Evidence for this is the lack of impressions nearer the centre of the shell, orientation of the rows of teeth (anterior-posterior trending) impressions parallel to the ventral margin of the Inoceramus shell, and curvature of the main row slightly concave toward the centre of the shell, ruling out the possibility that this row of teeth was situated at the anterior (convex outward) edge of the predator’s jaw. This set of characteristics rules out normal skates and rays, which have smaller and generally flatter interlocking pavement teeth. It also eliminates reptiles like Globidens from contention, the teeth of which are larger, more widely spaced, and would have done much greater damage to the shell. Cretaceous pycnodont fishes had small mouths filled with very small, low, rounded, pebble-like teeth probably adapted for coral and algae ‘grazing’ on hard substrates (Romer 1955, p. 99). These do not compare with the observed tooth marks. Most Teleost fish have pointed teeth which are more close-set, and would have punctured the shell on contact, leaving small, close-set perforations. Described Cretaceous Ptychodus, especially forms like P. deciiirens Agassiz (Wood- ward 1911, pi. 51, fig. 4), possess all of the criteria defined above for the predator. Their teeth are knob-like, blunt, and moderately elevated on stout bases, with crests closely spaced but not in contact. The teeth are arranged in closely spaced parallel rows (anterior to posterior) with adjacent teeth in alternate rows slightly offset. Along lateral marginal EXPLANATION OF PLATE 81 (pars', SCC cllSO p. 446) Fig. \a, h. Lateral and oblique ventrolateral views, respectively, X 1, left valve of Inoceraimis tenuis Mantell; holotype, BMNH 5890, showing probable Ptychodus tooth impressions in two parallel rows (axes left to right as oriented) and broken edge of ventral margin caused by single bite. Mantell Collection, from Cenomanian Chalk Marl of Hamsey, Sussex. Palaeontology, Vol. 15 PLATE 81 KAUFFMAN, Predation on Inoceramus HANCOCK, KENNEDY, and KLAUMANN, Ammonites from Rhenish Cretaceous E. G. KAUFFMAN; PREDATION ON INOCERAMUS 441 and siibmarginal rows, individual teeth within each row may also be slightly offset from the axis of the row. Woodward (1911, pi. 51) shows typical examples. Significantly, the spacing between tooth crests and rows and the diameters of individual tooth crests in jaw assemblages of Ptychodus decwrens Agassiz illustrated by Woodward from the English Chalk (191 1, pi. 51 , fig. 4) closely match the size and spacing of tooth impressions in Inoceramus tenuis. I conclude from these data that the tooth marks along the ventral margin of /. tenuis are probably those from the lateral margin of a Ptychodus conspecific with or closely related to P. decwrens. The two species coexisted: Inoceramus tenuis has a Lower to Middle Cenomanian range; Ptychodus decwrens occurs sparsely in the Lower Ceno- manian (Dibley 1912; Kennedy 1969) and more commonly in Middle Cenomanian to Turonian zones of the English Cretaceous sequence (Woodward 1911; Dibley 1912; Kennedy 1969). An interesting problem connected with these observations is the absence of com- parable tooth impressions on the smaller right valve of the type of /. tenuis. It is nearly complete, with only small pieces of the margin broken, and uncrushed (Woods 1911, fig. 32). From this, the orientation of the shell and predator at the time of attack can be determined, and is compatible with the implied life habits of both. Weak pedalbyssal musculature and a thin, sinuous dorsoanterior byssal slit in Ino- ceramus tenuis indicate that it lived weakly attached to the substrate by a narrow row of byssal threads, with the dorsoanterior margin close to the bottom. Inoceramus tenuis is moderately inequivalve, with the left valve larger, considerably more inflated, and with the umbo projecting well beyond that of the right valve (Woods 191 1, fig. 31). This further infers that in life the shell lay primarily on the dorsal and anterior flanks of the left valve, with the anterior flank and umbone held close to the substrate by byssal threads or even partially buried, but with the commissural plane of the shell inclined upwards and tilted so that the inhalent and exhalent areas of the mantle margin were elevated well above the sediment-water interface. In this position, with the valves gaping moderately (as now preserved; Woods 1911, fig. 31), the ventral margin of the left valve projected outward, beyond the more elevated margins of the gaping right valve by more than 20 mm (measured in the plane of the substrate surface; as reconstructed from the type specimen). The inferred orientation can be reproduced by orienting Woods’s anterior view of I. tenuis (1911, fig. 31) so that the ventral surface of the umbo on the left valve lies tangential to a horizontal plane representing the substrate surface. A predatory fish, like Ptychodus, feeding in characteristic fashion close and parallel to the bottom (or even slightly immersed in soft substrate) would first encounter the projecting left valve margin of the gaping shell if approaching the vulnerable feeding (ventral) edge. Distribution of tooth marks on the shell infer such an approach, but from a ventrolateral direction so that only the side of the fish jaw apparatus closed on the projecting left valve margin. Impression of tooth marks on only the ventral edge of the left valve indicates that the shell remained partially gaping during the attack. Failure of the animal to close the valves at or before predator contact, bringing the right valve within range of the teeth, may indicate that the bivalve was sick or recently dead from other causes. Most epifaunal bivalves have excellent sensory perception at the mantle margins and close the valves well in advance of near approach or contact by a large fish (or divers). Further, the Ptychodus made only a single bite, and then abandoned the 442 PALAEONTOLOGY, VOLUME 15 luoceramus. This may further indicate a dead or dying organism. There is no evidence for subsequent healing of shell fractures caused by the bite. Burial was prior to complete organic decay, however, as the ligament remained intact long enough to keep the valves together until they were covered, and the crushed marginal shell areas did not fragment and disseminate. Broken shell pieces were probably held in place by bands of mantle muscles associated with, and inside of, the pallial line; the mantle tissue was presumably still in place at the time of the attack. INDIRECT EVIDENCE FOR VERTEBRATE PREDATION IN INOCERAMIDAE Although the preceding example is the first good evidence for vertebrate predation in Inoceramidae, a number of other observations strongly suggest that this was a wide- spread and ecologically important phenomenon. This largely indirect evidence, discussed below, has led a number of workers to propose limited to extensive vertebrate predation on these and other bivalves during the Cretaceous. Frey (1972, in press) and D. E. Hattin (personal communication) have suggested that fragmentation of luoceramus cuvieri Sowerby in the Fairport Member of the Carlile Shale (Turonian; chalky limestone and shaly chalk facies), and of I? deformis Meek, Volviceramiis grandis (Conrad), and Platy- ceramus plalinus (Logan) in chalks, limestones, and chalky to calcareous clay shales of the Niobrara Group (Coniacian-Earliest Campanian), Western Interior United States, may well have been caused by vertebrate predation, especially by the commonly asso- ciated Piychodus. These are all moderately deep- and quiet-water facies (Kauffman 1967, 1969). Speden (1971, pp. 56-60) has documented numerous occurrences of luoceramus prism and shell fragment aggregations in the Albian-Cenomanian ? Clarence Series of New Zealand which he interprets as disgorged material and fecal pellets from vertebrate predators. These occur in particular zones, usually between beds which bear whole to non-selectively fragmented or concentrated luoceramus material. Of special interest is the scarcity of fragments from the thickened hinge areas of inoceramids in these regurgi- tated and fecal concentrations, possibly inferring that the predator selectively bit off the thin-walled part of the shell containing most of the animal. Additional observations which suggest widespread vertebrate predation on Ino- ceramidae are as follows: 1. In Cretaceous strata which physically suggestveryquiet-waterdepositional environ- ments by their fine grain size, undisturbed thin bedding, and great lateral persistence of individual laminae, Inoceramidae up to several feet in diameter are often selectively crushed and fragmented around the ventral and ventrolateral margins. This is an area where the shell is thinnest and which is preferentially attacked by a variety of predators on living epifaunal and semi-infaunal bivalves. In some cases entire luoceramus ‘beds’ are highly fragmented into angular pieces of various size, forming thin local calcarenites and coquinas in evenly and thinly bedded carbonates and clay shales which lack evidence of significant current scour or wave action. These fragmented and crushed shells are particularly common in the Fairport, Niobrara, and Austin chalks and shaly chalks, in the Blue Hill and Pierre dark gray shales of the United States Cretaceous, and in parts of the English, Irish, French, Danish, and German chalk sequences. Where shells show only marginal fracturing it commonly extends to areas within the E. G. KAUFFMAN; PREDATION ON INOCERAMUS 443 pallial line which are underlain by a strong, generally inflexible nacreous layer; this suggests that forces greater than gentle bottom currents were involved in breakage. Sedimentary compaction might account for some of this crushing, but fine-grained carbonates in particular have a relatively low compaction factor, and it is in these rocks that zones of crushed or broken shells are most widespread. Further, compaction should effect fracturing of the entire shell, at least in broad, thin-shelled species of low valve convexity; yet many shells have only the ventral and ventrolateral margins crushed. Some other factor must be called upon to account for selective crushing of inoceramid shells in these environments and predation by blunt-toothed, bottom-feeding vertebrates is a primary alternative explanation. As pointed out by Speden (1971) and Frey (1972, in press), and corroborated by much field observation by myself and colleagues, this is a common phenomenon in Cretaceous rocks throughout the world. 2. In Cretaceous sequences of the Western Interior United States the principal verte- brate fossils found associated with Inoceramidae, other than normal teleost scales and bones, are teeth of the shell-crushing shark Ptychodus. Significantly, these teeth are most abundant in carbonate units containing zones of Inoceramus shell coquina and margin- ally crushed shells, suggesting a direct predator-prey relationship between the two. 3. Many large Inoceramus shells which display selective crushing along the ventral and ventrolateral margins are bivalved individuals in life position (lying on the left valve), and have the crushed fragments still in or close to normal position. Assuming that selective marginal crushing is primary (at the surface) rather than diagenetic (see discussion under (1)), these observations infer that the organism was alive at the time of crushing, and that the mantle was still in place along the inner edge of the shell, allowing the mantle muscles to hold the shell fragments in near normal relationship after crushing, and that any bottom currents were very weak. Sedimentation in many of these areas was apparently slow, so that crushed shells remained exposed for some time after attack. The ligament of Inoceramidae is weak and lacks strong calcified supports; it would have been quickly destroyed through organic decay after death. Yet both valves and marginal crushed shell fragments remained in normal or near normal position after death; thus low current energy levels are inferred, insufficient to cause much of the observed crushing. 4. In carbonate and dark shale units where fragmentation of inoceramid shell beds and marginal crushing of individual shells is most common, and also where Ptychodus teeth are most abundant, inoceramids are by far the dominant macrofossils preserved, and commonly the only ones. They were probably the principal components of these Cretaceous ‘palaeocomm unities’ (Kauffman 1967, 1969). As such they must have been the primary food of various known contemporaneous vertebrates which were specifically adapted for bottom feeding on large molluscs and echinoderms. 5. A great number of Cretaceous Inoceramidae display healed shell injuries which seem to reflect crushing of the ventral and ventrolateral margins by some blunt object(s). Many of these crushed areas are about the size of the Ptychodus impressions described here in Inoceramus tenuis, but rarely are there more than two and none have been observed to occur so clearly in rows as in this specimen. Nevertheless, at least some could have been made by the blunt teeth of shell-feeding vertebrates. A search of existing collections for these types of injuries will doubtless turn up additional support for vertebrate predation; the author is presently conducting such a search. Gg C 9016 444 PALAEONTOLOGY, VOLUME 15 In summary, a single specimen of Inoceramus shows definite evidence of vertebrate predation, but considerable indirect evidence — marginal crushing of shells in quiet- water environments, numerous shell injuries made by blunt objects, the role of Ino- ceramidae as dominant benthic invertebrates in ‘paleocommunities’ also characterized by the remains of shell-feeding vertebrates, especially Ptychodiis teeth, and the common occurrence of inoceramid shell debris in presumed vertebrate fecal pellets and disgorged masses — all suggest that vertebrate predation was a common and important ecological factor among Cretaceous Inoceramidae. REFERENCES DiBLEY, G. E. 1912. Additional notes on the Chalk of the Medway Valley, Gravesend, West Kent, North-east Surrey, and Grays (Essex); with paleontological notes and appendices. Proc. Geol. ^55. 29, 68-96, pis. 7-9. FREY, R. In press. Stratigraphy, paleoecology, and depositional environment of the Fort Hays Lime- stone Member, Niobrara Chalk (Upper Cretaceous), west-central Kansas. Bull. Kaus. Univ. geol. Surv. HATTiN, D. E. 1962. Stratigraphy of the Carlile shale (Upper Cretaceous) in Kansas. Ibid. 156, 155 pp., 27 pis. KAUFFMAN, E. G. 1965. Taxoiiomic, ecologic, and evolutionary significance of interior shell morphology in the Inoceramidae (Mesozoic Bivalvia). Progr. a. Mtgs Geol. Soc. Am., Kansas City, 85. 1967. Coloradoan macroinvertebrate assemblages, central Western Interior, United States. In KAUFFMAN, E. G., and KENT, H. c. (cds.), Palcocnvironments of the Cretaceous Seaway — a symposium; Spec. Pub. Colorado School of Mines, 67-143, 12 figs. 1968. The Upper Cretaceous Inoceramus of Puerto Rico. Proc. 4th Carib. Geol. Conf, Trinidad, 1965, 203-218, pis. 1-2, figs. 1-6. 1969. Cretaceous marine cycles of the Western Interior. Mountain Geol. 6 (4), 227-245, figs. 1-4. KENNEDY, w. j. 1969. The correlation of the Lower Chalk of southeast England. Proc. Geol. Ass. 80, 459-560, 22 pis., 16 figs., 10 tabs. ROMER, A. s. 1955. Vertebrate Paleontology (6th edn.). Univ. Chicago Press, 687 pp., 4 tab., 377 text-figs. SPEDEN, I. G. 1971. Notes on New Zealand Fossil Mollusca — 2. Predation on New Zealand Cretaceous species of Inoceramus (Bivalvia). N.Z. Jour. Geol. Geophys. 14, pp. 56-60, figs. 1-4. WOODS, H. 1911. Monograph ofthe Cretaceous Lamellibranchia of England. Prt/eo/;/og7-. Soc. [Monogr.'] 2, pt. 7, 1-271. WOODWARD, A. s. 1911. The fossil fishes of the English Chalk. Paleontogr. Soc. [Monogr.] 2, pt. 7, 225-264, pis. 47-54. ERLE G. KAUFFMAN Department of Paleobiology E-307 Natural History Building U.S. National Museum Washington, D.C. 20560, U.S.A. Typescript received 23 June 1971 AMMONITES FROM THE TRANSGRESSIVE CRETACEOUS ON THE RHENISH MASSIF, GERMANY by J. M. HANCOCK, W. J. KENNEDY, and H. KLAUMANN Abstract. Ammonites from the transgressive Cretaceous at Miilheim-Broich near Essen in Germany are from the Zone of Hypotun ilites carcitanensis, at the base of the Cenomanian. In western Germany, on the southern flank of the Munster basin, the Upper Cretaceous is transgressive southwards on the Sauerland region of the Rhenish massif (for a general account, including an extensive summary in English, see Thiermann and Arnold 1964). Kahrs (1927), in a study of the palaeogeography of the Upper Cretaceous of Westphalia at its southern limits, described in detail exposures at Kassenberg in Mtilheim-Broich, some 13 km west -south-west of Essen. In Rauen quarry, cut in the east slope of the hill, there were pockets of Cretaceous resting in hollows about half a metre deep in the upper surface of Upper Carboniferous sandstone; Kahrs called these sediments in pockets ‘klippenfacies’ to distinguish them from the flat-lying more extensive Cretaceous where it rests on Upper Carboniferous shales on the other side of the hill in Becker quarry, some 500 m to the west (Kahrs 1927, figs. 1 and 2). The lower klippenfacies (‘Rotkalk mit Brauneisen-schwarte’) is a conglomeratic sandy limestone with a very rich Ceno- manian fauna. Kahrs listed two species of porifera, ten species of scleractinid corals, nine species of echinoids, 16 species of brachiopods, 65 species of bivalves, 77 species of gastropods, two species of ammonites, four species of serpulids, and one crab; many of these fossils were collected by J. Bohm. The ammonites listed by Kahrs were Schloenbachia varians Sow. and Hyphoplites laurenti Joh. Bohm. Until recently nearly every Schloenbachia has been called S. varians. H. laurenti, according to Wright and Wright (1949), is a synonym of Hyphoplites curvatus (Mantell); this strongly suggests a Lower Cenomanian age. LENTICLE OF KLIPPENFACIES IN THE RAUEN QUARRY For over 12 years one of us (H. K.) has kept observations on the Rauen quarry, east of Holzstrasse and just west of the railway beside the Ruhr in Miilheim-Broich. In the north-eastern quarter of the quarry the basal conglomerate has been found to occur as an infilling of a shallow trough, rather than in pockets. The trough was aligned roughly north-south, with a maximum width of 100 m and a length of possibly 150-180 m, cut into Carboniferous shales. Kahrs’s horizontal section (his fig. 1) may have been drawn through the southern margin. Even in the centre the lenticle of sediment does not exceed 1-5 m in thickness, and is the same facies as Kahrs’s older klippenfacies which can still be found in pockets elsewhere in the quarry: a pink to rusty-brown micritic limestone with pebbles and boulders of Carboniferous sandstone, granules of limonite, and grains of quartz. Limonitic crusts have formed in the top 0-25 metre on the south-east flank [Palaeontology, Vol. 15, Part 3, 1972, pp. 445-449, pi. 81.] 446 PALAEONTOLOGY, VOLUME 15 W E TEXT-FIG. 1. Section at the Rauen quarry, Mulheim-Broich. The small patches of klippenfacies west of the lenticle are the pockets in which Kahrs recorded Plenus Zone sedim.ents above the Lower Cenomanian Rotkalk mit Brauneisen-schwarte and below the Labiatus Mergel. EXPLANATION OF PLATE 81 (pars) (See opposite p. 440) All figures are of natural size except 2c; all specimens are coated with ammonium chloride. Figs. 2a-c. Hyphoplites aff. campichei Spath transitional to falcaliis aurora Wright and Wright. H. Klaumann coll. 14. Fig. 2c is a side view X 2 to show the feather structure preserved on the body chamber and the last few chambers of the phragmocone. Feather structure has previously been described in Cretaceous placenticeratidsl Meek 1876; Hyatt 1903, p. 222, pi. 47, figs. 3-5; Haas 1961) and desmoceratids (Wright in Arkell et at. 1957, p. L92), Jurassic oppeliids (Waagen 1869, pi. 18, fig. 5; Petitclerc 1918) and haploceratids (Arkell in Arkell et al. 1957, p. L92; Holder 1955). This specimen compares best with the example figured by Haas, but all examples are on the outer whorls of ammonites with markedly compressed whorl sections. Feather structure has previously been regarded as a structure of the inner nacreous shell layer, but on this Hyphoplites it is preserved on an internal mould. Figs. 2>a-b. Schloeiibachia variaiis )uv. ?var. subvariaiis etuct. H. Klaumann coll. 11. Conch preserved, probably as a calcite replacement. Figs. 4a-c. Schloeiibachia varians var. siibplana (Mantell). H. Klaumann coll. 10. Figs. 5a-c. Schloenbachia varians (J. Sowerby) forma typica. H. Klaumann coll. 3. Note the complete absence of umbilical tuberculation. Figs. ba-c. Schloenbachia varians var. subplana (Mantell) H. Klaumann coll. 7. A nearly complete adult with more than half a whorl of body chamber much of which has conch preserved. Figs. la-c. Schloenbachia varians var. subtiiberculata (Sharpe) H. Klaumann coll. 15. Appreciably more compressed than the lectotype. Figs. 8n-c. Sciponoceras roto Cieslinski. Side, ventral, and dorsal views of H. Klaumann coll. 13. A wholly septate internal mould. The circular cross-section and constrictions spaced at one in a length equivalent to about three diameters, distinguish this species from the later S. baculoide (Mantell) which has a markedly compressed whorl section and more closely spaced constrictions (Kennedy 1971), as does S. glaessneri Wright. Figs. 9a-b. Anisoceras aff. picteti Spath. H. Klaumann coll. 12. Side and ventral views. The body cham- ber begins at the first long rib. Whorl breadth 86 per cent of whorl height. Kennedy (1971) discussed this form. There arc around four ribs in a length equal to the whorl height. This internal mould shows a fiattening on the top of each shoulder tubercle, indicating that there was a septum across the base of each spine on the conch. This animal was injured in the last formed part of the phrag- moconc : a pit on each flank may be bite-marks, and there is also a tubercle missing on one shoulder. HANCOCK ET AL.: AMMONITES FROM RHENISH CRETACEOUS 447 of the lenticle. The lithology of the matrix is remarkably similar to the Shenley Lime- stone of the English Lower Albian, and in many places has a similar-looking brachiopod fauna and the bivalves Septifer lineatus and Chlamys robinaldina. Like the Shenley Limestone also is the fact that the fauna varies from place to place. The lenticle contains a fauna similar to that recorded by Kahrs from the pocket occurrences of Rotkalk mit Brauneisen-schwarte klippenfacies, with the addition in places of a rich bryozoan fauna. Moreover, the lenticle has yielded many more cephalo- pods. We have: Eutrephoceras sublaevigatum (d’Orbigny) (2): Anisoceras aff. picteti Spath (1): Sciponoceras roto Cies'lihski (1): Hyphoplites aff. campichei Spath transitional to H.falcatiis aurora Wright and Wright (1): Scliloenbacliia varians (J. Sowerby) forma typica (1): Schloenbachia varians aff. var. subtiiberciilata (Sharpe) (2): Schloenbacbia varians iuv. ?var. subvarians auct. (1): Schloenbachia varians transitional between var. subvarians auct. and subplana (Mantell) (2): Schloenbachia varians var. subplana (Man- tell) (5). Most of the specimens are preserved as internal moulds, but some of the ammonites retain some of the conch, probably as a calcite replacement of the original aragonite. The Hyphoplites, Sciponoceras, and Anisoceras were all collected in the range 70-108 cm above the base of the conglomerate, whilst most of the Schloenbachia came from 45-60 cm above the base. By comparison with the English succession the whole faunule belongs to the Zone of Hypoturrilites carcitanensis, although this ammonite itself has not yet been found at Miilheim-Broich. This Zone is the lowest of three that have been recognized in the Lower Cenomanian in England (Kennedy 1969, 1971 ; Kennedy and Elancock 1971 ), and equates with the lower part of the Zone of Hypoturrilites schneegansi in Algeria-Tunisia (Dubourdieu 1956). The assignment of the faunule to the Carcitanensis Zone is sup- ported by the occurrence of Inoceranius anglicus elongatus'l Pergament. REGIONAL SETTING Since the basal klippenfacies at Miilheim-Broich belongs to the Carcitanensis Zone, some of the Rhenish massif in this area must have been submerged very early in the Cenomanian. Some 50 km to the north-east towards Munster, on the hidden extension of the Rhenish massif, the Carboniferous was submerged even earlier, some time in the Albian (Arnold 1964). The subsequent sedimentological history on the massif is compli- cated, and is now difficult to work out because most of the region is built over, but it is certain that there is a variety of breaks in the succession. In the Rauen quarry itself the lenticle of klippenfacies conglomerate is overlain by 0T5 to 3 m of chalky marl (Labiatuston or Labiatus Mergel) from which we have obtained inoceramids that Dr. E. G. Kauffman has kindly identified as Mytiloides labiatus labiatus (Schlotheim) and M. labiatus transitional to M. latus (Mantell), which together indicate low Turonian. But on the western side of the quarry, in the pockets, a glauconitic marl with pebbles and phosphatic nodules (younger klippenfacies) inter- venes between the Rotkalk mit Brauneisen-schwarte and the Labiatuston, and was assigned by Kahrs to the Plenus Zone. In the Becker quarry, now built over, there was no klippenfacies but beneath the glaueonitic marl there was ‘Glauconitische Schwammergel’. This is a dark green rock of glauconite in a clay matrix which has often been called the Essen Greensand. But it 448 PALAEONTOLOGY, VOLUME 15 is by no means certain that ‘Essen Greensand’ has been used only for this formation. Bartling and Breddin (1931), for example, seem to have used the name for any basal greensand. On the west side of Rauen quarry, and probably at other localities, one can get Turonian greensand (Bochumer Griinsand) resting directly on the Carboniferous. Yet a further complication is that some authors (e.g. Schliiter) have distinguished another formation, the Tourtia, above the Essen Greensand, whilst others have treated these two names as synonyms. In view of all this confusion we feet it is not possible to give an exact date for this oft-quoted formation. We would only mention that most of the 14 ammonites from Essen Greensand and Essen Tourtia figured by Schliiter (1871-1876) are Lower Cenomanian, and all of them could be. On the other hand, the whole of the Cenomanian in the Essen area is shown as ‘greensand’ on the facies map by Arnold (1964). The two statements are not necessarily contradictory because (i) the ammonites may not have occurred through the whole thickness of the Essen Greensand, and/or (ii) much of the Middle and Upper Cenomanian may be missing in most of the region, as is now known to be the case in the Rauen quarry. Acknowledgements. This paper was made possible by the kindness of Professor Dr. E. Voigt who introduced the German author to his English colleagues, advised on the stratigraphy, and criticized a first draft of the paper. Dr. Erie G. Kauffman generously identified and commented on inoceramids. REFERENCES ARKELL, w. j., et al. 1957. Cephalopoda Ammonoidea. Treatise on Invertebrate Paleontology, Part L, Mollusca 4 Kansas University Press. ARNOLD, H. 1964. Fazies und Machtigkeit von Kreidestufen im Munsterland. Fortschr. Geol. Rheinid. iL Westf. 7, 599-610. BARTLING, R., and BREDDIN, H. 1931. Erldiiteriingen zur Geologischen Karte von Preiissen, 295, Blatt Midheim (Ruhr). Berlin. DUBOURDiEU, G. 1956. Etude geologique de la region de I’Ouenza (Confins Algero-Tunisiens). Bull. Carte geol. Alger. 10, 1-659, 22 pis., map. HAAS, o. 1961. A Placenticeras with feather structure. J. Paleont. 35, 230. HOLDER, H. 1955. Die Gattung Taramelliceras im siid-westdeutschen Unter- und Mittelmalm. Palae- ontographica Abt. A, 106, 37-153, pis. 16-19. HYATT, A. 1903. Pseudoceratites of the Cretaceous, stanton, t. w. (ed.). Monogr. U.S. geol. Surv. 44, 351 pp., 47 pis. KAHRS, E. 1927. Zur Palaogeographie der Oberkreide in Rheinland-Westfalen. Neues Jb. Miner. Geol. Paldont. Beil Bd 58B, (festband Pompeckj), 626-687, pis. 42-44. KENNEDY, w. J. 1969. The Correlation of the Lower Chalk of South-East England. Proc. Geol. 80, 459-560, pis. 15-22. 1971. Cenomanian ammonites from southern England. Special Papers in Palaeontology, 8, 64 pis. and HANCOCK, j. m. 1971. ManteUiceras saxbii, and the horizon of the Martimpreyi Zone in the Cenomanian of England. Palaeontology, 14, 437-454, pis. 79-82. MEEK, F. B. 1876. A report on the invertebrate Cretaceous and Tertiary fossils of the upper Missouri Country. In meek, f. b., and hayden, e. v. U.S. Geol. Geog. Survey Terr., Mon. 9, lxiv+629 pp., 44 pis. PETiTCLERC, p. 1918. Ornamentation peu connue chez certaines ammonites jurassiques. Bull. Soc. geol. Fr. (ser. 4) 18, 233-234. SCHLUTER, c. 1871-1876. Die Cephalopoden der oberen deutschen Kreide. Palaeontographica, 21, 1-24, pis. 1-8; 22, 25-120, pis. 9-35; 24, 121-264, pis. 36-55. THiERMANN, VON A. and ARNOLD, H. 1964. Die Kreide im Munsterland und in Nordwestfalen. Fortschr. Geol. Rheinid. u. Westf. 7, 691-724. WAAGEN, w. 1869. Die Formenreihe des Ammonites subradiatiis. Geogn.-Paldont. Bcitr. 2, 181-256, pis. 16-20. HANCOCK ET AL.: AMMONITES FROM RHENISH CRETACEOUS 449 WRIGHT, c. w. and wright, e. v. 1949. The Cretaceous ammonite genera Discohoplites and Hyphoplites Spath. Q. Jl. geol. Soc. Land. 104, 477-497, pis. 28-32. J. M. HANCOCK Department of Geology King’s College Strand London wc2r 2ls W. J. KENNEDY Department of Geology and Mineralogy Parks Road Oxford 0x1 3pr H. KLAUMANN 433 Miilheim-Ruhr Quellenstrasse 93 Typescript received 23 June 1971 West Germany THE DEVELOPMENT OF THE LOOP IN THE JURASSIC BRACHIOPOD ZEILLERIA LECKENBYI by P. G. BAKER Abstract. An ontogenetic series of Z. leckenbyi has been obtained from a locality in the mid-Cotsvvolds. Serial sectioning has enabled determination of the microstructure and development regime of the loop. The work reveals that a loop of adult character is formed by the time the brachial valve is about 4 0 mm long, confirming Elliott’s suspicion that in zeilleriids the early stages of loop development were passed through very quickly. During early development phases the loop is connected to a septal pillar rising from the floor of the brachial valve. The general pattern of loop development appears to combine terebratellid and dallinid characters. It is found that the descending elements play only a subsidiary role during loop development. That they become relatively massive fairly early appears to be due to the fact that they are required to support the ascending complex after resorption of the connection with the median septum has occurred. The growth pattern of the median septum indicates that it may be regarded as a secondary character and therefore makes only a very limited contribution to loop development in the Zeilleriidae. Attention is focused on the gross inadequacy of our knowledge of the actual growth of juvenile loops during the recognized stages passed through during ontogeny. Analysis of the development regime illustrates some of the dangers of recording growth stages which are essentially momentary phenomena in what must necessarily be a cumulative process. The current work indicates that a cryptacanthiinin of Glossothyropsis type may be ancestral to the Zeilleriidae. It is concluded that the possession of spinose ascending and descending elements is a more important ancestral character than the absence of a median septum and that the microstructure of developing loops will provide the key to the solution of the complex phylogeny of the Terebratulida. The loop ontogeny of Terebratulides is not well known and even the more recent inter- pretations (Dagis 1958, 1959; Babanova 1965) rely heavily on the work of Elliott ( 1948, 1953). Undoubtedly the main reason for the lack of information is the scarcity of juveniles of representative genera. It is surprising, therefore, that despite our general lack of data, high taxonomic significance is attributed to the loop ontogeny of terebratulides. It appears that microstructural analysis of loop elements will enable some of the gaps in our knowledge to be bridged and that bulk sampling may reveal the presence of very small juveniles previously overlooked. The material used in these investigations was recovered during the search for micro- morphic brachiopod faunas in the Oolite Marl. This deposit is a weakly-coherent inter- bedded marl and biomicrite of Upper Aalenian {murchisonae zone) age occurring in the Inferior Oolite of the mid-Cotswolds around Cheltenham. The two best remaining exposures are at Cleeve Cloud (SO 984261) and Westington Hill Quarry (SP 142368) from which locality the current material was obtained. The stratigraphy of the Oolite Marl and the horizon from which collections have been made are outlined in Baker (1969, p. 388). A third, excellent exposure in the old cutting at Notgrove Railway Station (SP 094213) is now unfortunately no longer available as the area has been taken over for site development. In addition to a rich organo-detrital residue and micromorphic brachiopods (Baker 1969) the Oolite Marl yields juvenile rhynchonellides and terebratulides and adults assigned to the species Glohirhyuchia subobsoleta (Dav.), EpUhyris submaxiUata (Morris), Plectothyris fimbria (Sow.), ‘‘Terebratiiki whitakeri Walker MS, and ZeiUeria leckenbyi (Davidson ex Walker MS). [Palaeontology, Vol. 15, Part 3, 1972, pp. 450-472, pis. 82-85.] P. G. BAKER; ZEILLERIID LOOP 451 A collection of 250 juvenile terebratulides, ranging in size from 0-6 to 14-0 mm in length, was analysed on the basis of 27 morphological characteristics. The data obtained from this analysis is too voluminous to be included in the present work and is to be published later. Briefly it may be stated that the larger juveniles may be readily correlated with their adult counterparts and on the basis of character evaluation may also be correlated with progressively smaller juveniles. Fortunately, Zeilleria leckenbyi is the only long-looped species present so the possibility of error is greatly reduced. Thirty specimens ranging from 0-8 to 25-6 mm in length were selected, which, on the basis of the results obtained from morphological analysis, were anticipated to be juveniles and adults of Z. leckenbyi (PI. 82, figs. 1-12). These specimens were serially sectioned and the results obtained are summarized in Table 1. The external morpho- logical characters of all the specimens anticipated to be Z. leckenbyi were supported, with only two exceptions, by zeilleriid internal characters. The obvious developmental ‘progression’ leaves little room for doubt that the remaining 28 shells represent the ontogenetic stages of a single species. Fundamental to all thinking regarding the interpretation of secondary shell fabric must be the realization that the smaller the unit considered, the closer it must approach Rudwick’s (1959) ‘momentary’ conception of growth and the larger the unit, the closer it approaches his ‘cumulative’ conception. This means that the various parts of the developing loop are the cumulative product of a series of momentary units. As West- broek (1967, p. 29) has pointed out, increase in size of internal structures is the result of deposition of shell at their distal ends but thickening is usually the result of shell deposition proximally, i.e. in posterior zones of the shell where structures arise apically. It follows, therefore, that in the apical region of a shell, early growth stages will only be preserved (where no resorption of material has taken place) as cores buried in the shell material of subsequently enlarged structures. Added to this is the problem that growth of the valve must necessarily bring about a change in the orientation of an early structure relative to the commissural plane. As the early growth phases of the pedicle valve are eliminated by resorption in Z. leckenbyi, attempts to trace continuity of structures must be based on the study of brachial valves. A consideration of critical importance (Westbroek 1969), apparently largely ignored by many workers, is that owing to the change in orientation of structures during growth, the orientation of the plane of section must be adjusted accordingly if ‘buried’ structures are to be clearly identifiable in sections of larger shells. For instance, transverse sections through units of the very early cardinalia will be encountered in near horizontal sections through adult shells. In view of the need for accurate correlation between specimens in different size ranges, trios of specimens were selected which were as near identical as possible, morphologically and in size. Of these, one was retained as a reference, one was sectioned, normal, trans- verse, and the third specimen was sectioned growth oriented according to the normal transversely sectioned specimen of the preceding size range. In the text, therefore, hori- zontal sections refer to the orientation of the plane of section relative to the loop rather than to the orientation of the shell, which is usually low oblique. Orientation cannot be exact because the loop is obviously not visible, but critical sections may be correlated to a greater extent using this method. There can be no hard and fast correlation between size of animal and stage of loop development, for at a given size, loop development will 452 PALAEONTOLOGY, VOLUME 15 TABLE 1. Tabulation of the structures present in 28 specimens of Z. leckenbyi out of a sample of 30 shells anticipated to belong to that species. Of the remaining two shells, one, 1 -0 mm long showed no dental plates, the other, 9 0 mm long and having a damaged beak, proved to be a short looped form. Specimen Dimensions m m . Orientation Internal characters present Developmnt. phase represented Dental plates Crural plates Septalium Median septum Asc. elements Desc. elts. Loop of adult form Septal pillar Postr. arching spurs Asc. lamellae Anterior spurs Connection with median septum Lacunae Asc. branches with sour remnants pesc. branches Asc. and desc. 1 elements united Spinose L. w. Th. 37628/23 0-8 0-7 0-28 H.S. + Pre paramagad- iniform 37629/29 0-9 0-8 0-3 H.S. + 37630/19 1-2 10 0-35 H.S. + 37629/9 1-2 1-1 0-4 H.S. + + 37629/4 1-3 1-3 0-4 H.S. + + + 37629/10 1-3 1-1 0-4 H. S. + + 37629 1-3 1-1 0-4 H.S. + + + 37589/7 1-5 1-5 0-5 H.S. + + + + + 37630/28 1-7 1-6 0-6 H.S. + + + + + + + 37589/3 2-1 2-1 0-7 H.S. + + + + + + + + + Paramagad- iniform 37556 2-5 2-4 1-1 T.S. + + + + + + + + + 37570/1 30 30 1-2 H. S. + + + + + + + + + + Syncampag- iform 37580 3-9 3-4 1-4 T. S. + + + + + + + + + + 7 Frenuliform 37581 50 50 1-7 T. S. + + + + + + + + + Terebratal - iiform 37617 5-4 5-2 1-8 H. S. + + + + + + 7 + 37582 5-9 5-4 2-4 H.S. + + + + + + + + 37583 6-5 60 2-5 T. S. + + + + + + + + 37622 7-2 70 30 H.S. + + -f + + + + + 37663 10-5 10-4 4-1 H.S. + + + + 7 + Dalliniform 37664 12-2 12-5 5-5 H. S. + + + + + + 37665 14-5 14-5 7-3 T. S. + + + + + + 37667 17-2 18-5 7-5 T.S. + + + + + + 37660 17-5 16-7 8-2 H.S. + + + + + + 37657 190 180 8-0 T.S. + + + + + + 37658 190 190 90 H.S. + + + + + + 37659 21-0 200 90 H.S. + + + + + + Zeilleriiform 37661 23-5 20-0 110 H.S. + + + + + + 37666 25-6 220 130 T.S. + + + + + + P. G. BAKER: ZEILLERIID LOOP 453 be either precocious or retarded according to the momentary point which has been adopted as the mean. The closest approach to such a correlation must, therefore, lie in the relation of development phases to approximate size categories (Table 1). Owing to the very small size of the early juveniles and the delicate nature of their loop elements, it was virtually impossible to recognize loop elements in transverse section. This became particularly apparent where the inhlling matrix was not of uniform texture. Most of the work was therefore based on horizontal sections. This orientation offered the greatest chance of success as the component fibres of the elements would be more or less length-sectioned and, therefore, more easily visible in the matrix. Traces of trans- verse sections through specimens are included (text-fig. 1a-c) to conform with the accepted method of illustrating serially sectioned brachiopod loops. However, at this early stage of development the various elements may be only 3-4 fibres thick and do not respond well to photography. For recording loop development photographically, hori- zontal sections showing length-sectioned fibres yield far better results. Accordingly the evidence provided by horizontal sections has played an important part in the reconstruc- tion of early loops. Those sections which are regarded as critical are also figured (PI. 84, figs. 1-12). Ackiiowleclgeineuts. The author is indebted to Dr. J. D. Hudson, Department of Geology, The Univer- sity of Leicester, for discussion during the preparation of this paper. Thanks are due to Mr. G. McTurk for preparation of the stereoscan negatives and to Professor Sylvester-Bradley for use of the research facilities of the Department of Geology, The University of Leicester. Registration of Material. The material figured in this paper, together with original and duplicate peels, is to be housed in the Museum collection of the Department of Geology, University of Leicester under the catalogue numbers quoted. TECHNIQUES AND PREPARATION OF MATERIAL A comprehensive account of the preparation of Oolite Marl material is given in Baker (1969) and the material studied in the present paper was obtained by the same method. The smaller specimens up to 4-0 mm in length were investigated using the techniques developed for the study of the micromorphic thecidellinid Moorellina granulosa ( Moore) (Baker 1970). The larger specimens were studied using the techniques developed by Hendry et ai (1963). It has not been possible to locate a separated juvenile brachial valve in whieh the loop is complete although the bifid appearance of the ascending lamellae is quite common (PI. 82, figs. 16-18). NOMENCLATURAL PROBLEMS This and other studies show that the differences, discussed later, between dallinid and terebratellid loop ontogenies are often very subtle. Although all the changes are variants of two common plans they are complicated by precocious or retarded development and by changes in the relative proportions of the same structures in successive growth stages of a single genus also as a result of changes in the relative proportions of different parts of the loop in different genera. From an evolutionary point of view it is obvious that loop development within the suborder Terebratellidina forms a fairly intimate complex and it would, therefore, be wrong to introduce an entirely new nomenclature for zeilleriid loop 454 PALAEONTOLOGY, VOLUME 15 TEXT-FIG. 1a-c. Series of serial transverse sections through specimens of Z. leckenbyi, drawn from microprojectcd cellulose acetate peels, showing features of the species during different phases of development. A. Frenuliform specimen (37580). n. Terebrataliiform specimen (37583). c. Adult speci- men (37666). Numbers refer to the distance in mm of the sections from the beak. Lettering as Figs. 3, 4. Scale represents 2 mm. P. G. BAKER: ZEILLERIID LOOP 455 ontogeny. New terms must be introduced with due regard for their probable affinity with comparable structures in other forms. Great care has been taken to avoid the use of the term median septum in description of ontogenetic stages as it can be shown that the median septum as observed in the adult Z. leckenbyi plays no part in the development of the ascending elements of the loop. Whether the structure plays a valid role in the development of the loops of other terebratellidines is not yet clear. Muir-Wood (1934) noted that in the Zeilleriidae the septalium is formed by two plates, the septalial plates, which converge and fuse together to form a septum. She states ( 1 934, p. 529) that the septum appears to be distinct from the true median septum in many species. In very young specimens of Z. leckenbyi there is no septalium and the plates extending from the crural bases to the floor of the valve (PI. 82, figs. 14, 15) should, according to the Treatise definition (Williams and Rowell 1965) be regarded as crural plates. Muir-Wood also noted that in Digonella the dorsal end of the median septum appears to be inserted in the wall of the brachial valve. Study of oriented valves of Z. leckenbyi indicates that both the above structures are represented by the sessile bundle of secondary fibres which gives rise to the septal pillar of this species, in which case the median septum as observed in adult shells is a bi-component subsequent valve element. It post-dates the septal pillar, as the remnant of this structure (PI. 85, figs. 1, 2) is enveloped as the septum is extended anteriorly. In Z. leckenbyi, therefore, the median septum sensu lato, makes only a limited, if any, contribution to the development of the ascending elements. Elliott (1953, p. 263) refers to the often pillar-like upgrowth from the valve floor which constitutes the precursor of the ascending elements in the Terebratellidina. The use of Elliott’s term, septal pillar, being most apt, is adopted to describe the almost cylindrical structure arising from the valve floor and leaning anteriorly in a characteristic terebratellid (Elliott 1953, p. 267) manner. Following the appearance of the pillar the ascending elements develop rapidly and as a number of components appear almost simultaneously the term ascending complex is introduced (text-fig. 2c, d). It is felt that this term is required to describe structures which are the precursors of the ascending elements proper (ascending branches and transverse bar) and also those which regulate the early location and anterior extension of the descending branches. The component parts of the axial complex in order of appearance are as follows: Posteriorly arching spurs. Posteriorly directed outgrowths from the sides of the septal pillar near its distal end. The descending elements unite with them and grow along their ventral edge to fuse with the material of the ascending lamellae which lies between the anterior spurs (text-figs. 2a, b, 3a; PI. 83, figs. 2, 3). Ascending septum. Small, vertical, axially aligned plate, developed on top of the septal pillar and replacing the dallinid hood which is sometimes preserved as a rudiment on the posterior edge (text-fig. 2c, d). Ascending lamellae. A pair of diverging plates which arise from the anterior edge of the ascending septum and are subsequently extended along its ventral edge. These lamellae are the precursors of the ascending branches. As they increase in size they give rise to 456 PALAEONTOLOGY, VOLUME 15 P. G. BAKER; ZEILLERIID LOOP 457 horns of material posteriorly, which are deflected towards the mid-line until they unite to form the transverse bar (text-fig. 3b-d; PI. 83, fig. 6). Anterior spurs. A pair of prismatic calcite spurs developed from the base of each ascend- ing lamella. They regulate the position of the descending branches relative to the ascending branches and become ensheathed in secondary fibres as the descending branches are extended anteriorly (text-figs. 2e, f, 4, 5; PI. 83, figs. 1, 8). From the host of modifications described by Elliott (1953, 1960) and Muir-Wood et al. (1965) it becomes clear that terms such as campagiform and frenuliform must not be applied too rigidly. Also the use of the term stage is deplored. It is thought that the term stage is likely to lead to inflexibility as it implies a growth attainment of a momentary nature (Rudwick 1959). Accordingly, the author proposes to use the term phase as this suggests the more real, cumulative growth pattern of the loop. While the ascending complex is developing it shows a close resemblance to the modi- fied magadiniform type seen in Australiarcula artesiana Elliott and Bouchardia rosea (Mawe) (Elliott 1960). As the microstructure of the magadiniform loop is unknown, the term paramagadiniform will be used to describe the loop of Z. leckenbyi during this growth phase. DifiTerences of opinion exist (Thomson 1927; Elliott 1947, 1953) with regard to the correct definition of the campagiform loop, but all specify the presence of a hood. There- fore, by definition Z. leckenbyi does not possess a campagiform stage because the hood, even if present, is never more than a rudiment. Elowever, the growth phase succeeding the paramagadiniform culminates in a structure resembling the campagiform loop but arrived at by a different development sequence. Therefore, the phase of development succeeding the paramagadiniform phase will be termed syncampagiforin to describe the loop of campagiform appearance which arises without the involvement of a hood. ZEILLERIID DEVELOPMENT PHASES The developing loop of Z. leckenbyi passes through seven recognizable phases of growth, which, on the basis of experience may be anticipated to coincide with certain approximate size ranges. TEXT-FIG. 2a-h. Reconstructions of early juvenile brachial valves of Zeilleria leckenbyi (Davidson). Sequence of interior and lateral views to show the morphology of the developing loop and its promi- nence relative to the plane of the commissure. The figures are based essentially on data obtained from stereoscanned cellulose acetate peels, but the evidence was reinforced by polished sections and separated valves. A, B. Preparamagadiniform phase (37589/7). a. Brachial view, crural plates sloping down to unite with valve floor, b. Lateral view showing the relatively low pillar with an anterior inclination, c, D. Paramagadiniform phase (37589/3). c. Brachial view showing the form of the ascending complex. Median septum still not properly developed but the anteriors of the crural plates are separating from the valve floor and beginning to form a septalium. d. Lateral view showing the increased development of the descending elements and the ascending complex much higher relative to the commissural plane. E, F. Syncampagiform phase (37570/1). Lateral and brachial views showing the well-developed ascend- ing lamellae, anterior spurs, and transverse bar. Descending branches now united with the sides of the ascending lamellae. Septalium formed and septal pillar enlarged. G, h. Erenuliform phase (37580). Lateral and brachial views showing the location of the lacunae, a.c. ascending complex, a.sm. ascending septum, h.r. hood rudiment, i.s.r. inner socket ridge, 1. lacuna, s. dental socket. Numbers refer to the length of the shell in mm from which the peels were obtained. 458 PALAEONTOLOGY, VOLUME 15 TEXT-FIG. 3. Series of Ihreeqiiarters profile reconstructions (not to scale) of the loop of Z. leckenbyi showing the various elements in the complete development sequence, a. Preparamagadiniform. B, c. Early and late magadiniform. d. syncampagiform. e. Frenuliform. f. Terebrataliiform. g. Dal- liniform. ii. Zeilleriiform. Figs, c, e, g show only the distal portions of the descending branches, a.br. ascending branch, a.I. ascending lamella, a.s. anterior spur, a.sm. ascending septum, d.br. descending branch, d.el. descending element, h.r. hood rudiment, I. lacuna, m.s. median septum, p. septal pillar, p.a.s. posteriorly arching spur, r.a.l. resorbed ascending lamella, t.b. transverse bar. P. G. BAKER: ZEILLERIID LOOP 459 Phase 1. Pre-paramagadinifonn. It seems probable that the floor of the brachial valve is at first featureless. The precursor of the ascending elements of the loop appears very early as a bundle of fibres, longitudinally arranged, lying almost parallel with the floor of the valve (PI. 83, fig. 1 1). Growth increments are added to the ventral surface of this bundle, as might logically be expected from the position it occupies on the valve floor. An abrupt change in this growth pattern occurs when the animal is about T3 mm long. The bundle of fibres becomes re-oriented to rise as a septal pillar projecting from the valve floor (text-fig. 2a, b; PI. 82, figs. 13-15). This change in fibre orientation is accom- panied by a change in growth pattern as the new growth increments are now added anteriorly, i.e. on the originally dorsal surface of the fibre bundle (PI. 83, figs. 9, 10). The explanation for this change is obvious, for the new development regime will carry the developing loop anteriorly. Soon after the formation of the septal pillar, two ribbons of calcite, arched posteriorly, arise laterally at a point close to its distal end (PI. 83, figs. 2, 3, 5; text-fig. 2a, b, 3a). The descending elements at this time are very short. Before the shell reaches a length of T8 mm a further development phase begins. Phase 2. Paramagadiniform. Shell length T8-2-5 mm. The distal end of the septal pillar becomes laterally flattened and rises vertically as an ascending septum (text-figs. 2c, d, 3b). The descending elements become more strongly developed and there is a propor- tionate increase in the size of the posteriorly arching spurs, so that the two soon unite. A swelling appears about half way up the posterior edge of the septum. On occasional specimens this may be seen to develop into a low ring-like collar at its distal end. The structure actually plays no part in the formation of the loop but may be interpreted as a rudimentary hood of the type found in Aiistraliarcida artesiana Elliott. At the same time two divergent lamellae, the ascending lamellae, develop at the anterior edge of the ascending septum (PI. 82, fig. 18, PI. 83, fig. 6; text-fig. 3b, c). During this phase of development the crural plates become raised above the floor of the valve. In the mid-line they become turned down to form a shallow septalium (PI. 82, fig. 17) and extend as an overlay of secondary fibres above and on either side of the earlier deposited fibre bundle which represents the sessile portion of the septal pillar (PI. 83, fig. 11). The divergent lamellae are of the utmost importance for they represent the precursors of the whole of the subsequently developed loop. Within eaeh lamella arise the proximal ends of a pair of spurs which herald what is fundamentally the most important phase of development of the entire loop. Phase 3. Syncampagiform. Shell length 2-5-3-5 mm. No direct evidence is available but it is believed that the ascending lamellae extend up the anterior edge of the ascending septum and along its ventral edge, at first diverging and subsequently converging posteriorly to form the rudiment of the transverse bar of the adult loop (text-fig. 3b-d). This mode of origin is favoured because the campagiform hood is never more than a rudiment and the transverse connection when first formed is always thread-like. Following the initiation of the spurs (occasionally three on each lamella) their develop- ment proceeds rapidly and is accompanied by thickening of the distal ends of the descending branches and broadening of the ascending lamellae so that the loop becomes relatively massive (text-figs. 2e, f, 3d). Investigation shows that the anterior spurs are composed of prismatic calcite (PI. 83, fig. 4; text-figs. 4, 5). As they are extended H h C9016 460 PALAEONTOLOGY, VOLUME 15 TEXT-FIG 4. Serial block reconstructions to show the microstructure of the loop of Z. leckeiihyi as it appears in transverse section during the syncampagiform, a, and terebrataliiform, b, phases of develop- ment. For the purposes of clarity only the distal portions of the descending branches are included, a.l. ascending lamella, a.s. anterior spur, d.br. descending brancli, e.s. enveloped spur, I.r. lamella remnant, m.s. median septum, p. septal pillar, p.p. pillar precursor, r.a.I. resorbed ascending lamella, r.s. resorbed spur, t.b. transverse bar. Scale represents 0-2 mm. P. G. BAKER: ZEILLERIID LOOP 461 TEXT-FIG. 5. Serial block reconstructions to show the microstructure of the loop of Z. leckenbyi as it appears in horizontal section during the syncampagiform, a, and terebrataliiform, b, phases of develop- ment. Lettering as in Fig. 4, except p.r. pillar remnant. Scale represents 0-2 mm. 462 PALAEONTOLOGY, VOLUME 15 anteriorly, they are responsible for controlling and facilitating the location of the descending branches (PI. 83, figs. 1, 8). The spurs act as girders for the anterior extension of the descending branches, becoming at first partially, then completely enveloped by secondary fibres as the descending elements are extended (text-figs. 4, 5). The culmina- tion of this development phase is the development of a structure which morphologically bears a close resemblance to the dallinid campagiform stage. At the moment, however, there is no data on the dallinid type to show whether the two forms are microstructural equivalents. EXPLANATION OF PLATE 82 Figs. 1-18. Stereoscan photomicrographs, except 10-12, of various specimens of Z. leckenbyi (David- son) from the Oolite Marl, Westington Hill Quarry near Chipping Campden, showing the general morphology during ontogeny. Specimens coated with evaporated aluminium before photography. 1-3. Brachial, anterior, and lateral views of a preparamagadiniform juvenile (37629). Note the rounded appearance of the brachial valve and the beginning of apical resorption of the delthyrium during this phase of development. X 35. 4-6. Lateral, brachial, and anterior views of a syncampagi- form juvenile (37530). Apical resorption is now advanced and a low anterior sulcus is present. X 15. 7-9. Anterior, brachial, and lateral views of a terebrataliiform juvenile (37581). Brachial valve now becoming more elliptical, x 8. 10-12. Brachial, lateral, and anterior views of an adult (37668) for comparison purposes. X 1. 13. Interior of a brachial valve (37671) during the preparamagadiniform phase of development, showing the septal pillar rudiment in the floor of the valve prior to the forma- tion of the median septum, x 18. 14. Three-quarters profile view of specimen (37671) showing the septal pillar rudiment and crural plate (arrowed) descending to the valve floor. X 50. 15. Interior of a brachial valve (37672) showing the septal pillar rising from the valve floor. The crural plates are more well developed but still united with the valve floor. X 20. 1 6. Interior of a brachial valve (37669) show- ing the median septum and septal pillar with divergent ascending lamellae. X 18. 17. Three-quarters profile view of specimen (37669) showing the crural plates now raised above the floor of the valve and forming a shallow septalium. X 45. 18. High incidence profile view of specimen (37584) showing the ascending lamellae. The loop of this specimen is clearly damaged and was probably syncampagiform or frenuliform. Xl2. EXPLANATION OF PLATE 83 Figs. 1-11. Stereoscan photomicrographs of specimens of Z. leckenbyi. Material for all figures, except figs. 2, 3, obtained from cellulose acetate peels coated with evaporated aluminium before photo- graphy. 1. Montage of a horizontal section through a syncampagiform ascending lamella (37570/1, 12) showing the anterior spur, a.s., ensheathed in secondary fibres posteriorly and ‘locating’ the fibres of the descending element (arrowed). X 500. 2. Photomicrograph, reflected light, of a horizontal section through a polished specimen (37589/3) showing the posteriorly arching spurs arising from the septal pillar, x 40. 3. Retouched copy of fig. 2. 4. Transverse section through an anterior spur of a frenuliform specimen (37580/102) showing the prismatic calcite core. X 525. 5. Micrograph of a peel obtained from the etched surface of specimen (37589/3), fig. 2 above, enlarged to show detail of the junction of the posteriorly arching spurs with the septal pillar. X 1 80. 6. Horizontal sec- tion through a paramagadiniform specimen (37589/1 1) showing the ascending septum and divergent ascending lamellae arising from its anterior edge. X 70. 7, 8. Transverse sections through the anterior spurs of the left, fig. 7, and right, fig. 8, ascending lamellae of specimen (37580) showing the enveloped prismatic cores and their controlling influence on the location of the descending branches, fig. 8, centre, x 200. The actual separation between the left and right lamellae at this magnification would be approximately 65 mm. 9. Horizontal section through a syncampagiform specimen (37570/1, 7) showing the septal pillar in near transverse section with the change in the orientation of the secondary fibres anteriorly. X 100. 10. Enlarged view of the anterior portion of the septal pillar shown in fig. 9, to show detail of the fibre mosaic. x275. 11. Montage of a transverse section through a frenuliform specimen (37580/48) showing the septal pillar precursor fibre bundle (arrowed) overlain by secondary fibres as a result of the anterior extension of the median septum. X 300. Palaeontology, Vol. 15 PLATE 82 BAKER, Zeilleriid loop Palaeontology, Vol. 15 PLATE 83 10 11 BAKER, Zeilleriid loop ) P. G. BAKER: ZEILLERIID LOOP 463 Subsequently deposited material continues to accumulate posteriorly in the mid-line of the valve to form a recognizable median septum which is extended anteriorly so that the septal pillar itself becomes partially enveloped (PI. 83, fig. 1 1 ; text-figs. 4, 5). Although this has a strengthening effect on the pillar the primary objective appears to be the provision of a base for the attachment of adductor muscles. Development through the three phases described may be regarded as being aimed at the attainment of a basic skeletal structure which is capable of being translated into a loop able to support a plectolophe. Metamorphosis of this existing framework is neces- sary to elaborate and alter the relative proportions of the various loop elements produced by the initial development regime. Shell resorption, until now not a prerequisite of loop development, quite suddenly assumes a critical role in current and all subsequent development. Phase 4. Fremiliform. Shell length 3-5-4-0 mm. Although this phase of development probably does not reflect any change in the form of the lophophore it does illustrate that important physiological changes are taking place. Whilst the anterior of the loop is developing in the manner described above, resorptive activity begins at the posterior of the ascending complex. Two lacunae appear close to the point of divergence of the ascending lamellae (text-figs. 2g, h, 3e). Phase 5. Terebrataliiform. Shell length 4-0-7-2 mm. As the lacunae increase in size the distal portion of the pillar is also resorbed so that the loop becomes freed from its con- nection with the septal pillar (text-figs. 3f, 6a, b). None of the material sectioned showed the retention of a connection between the descending branches and the septal pillar so that a form corresponding to the true terebrataliiform stage of the dallinids has not been seen in Z. leckenbyi. However, if the mechanics of resorption are considered it seems probable that the delicate strips of material posterior to the lacunae would be lost before the descending branch connections and it is therefore logical to assume a short terebra- taliiform phase. It is the intimate relationship between the processes of accretion and resorption which during this phase sculpture the approximate configuration of the adult loop. The anterior spurs have by now apparently fulfilled their purpose as they cease to develop although their remnants are still visible (PI. 85, fig. 3; text-fig. 3f) and these continue to regulate the development of the descending branches. Apparently, the posterior end of the lower spur of each lamella is resorbed more slowly as forms passing through this phase of development normally show a small, posteriorly pointing projec- tion, posterior to the point of union of the lamella with the descending elements. Resorption, after it has begun, apparently proceeds rapidly as the transition from syncampagiform to late terebrataliiform is accomplished quickly and as far as can be ascertained, during the time that the shell is between 3-5 and 7-2 mm in length. The median septum continues to extend anteriorly after resorption of the connection of the septal pillar with the ascending elements has occurred. The pillar remnant, enveloped by subsequently deposited fibres, can be clearly seen in horizontal sections through young shells, some distance from the anterior termination of the median septum (PI. 85, figs. 1, 2). The median septum, therefore, as seen in zeilleriids must be regarded as a median septum sensu lato. 464 PALAEONTOLOGY, VOLUME 15 P. G. BAKER: ZEILLERIID LOOP 465 Phase 6. Dallinifonn. Shell length 7-2-19-0 mm. The dalliniform phase of development is essentially concerned with smoothing out the irregularities of the crudely adult loop developed during the preceding development phase. The loop thus formed (text-figs. 6c, D, 3g), although quite symmetrical is still relatively heavy and the anterior of the ascending branches still has a plate-like aspect. Phase 7. ZeiUeriiform. Shell length > 19-0 mm. The culmination of the accretion- resorption regime produces an adult loop in which all the elements are comparatively slender and ribbon-like (Ager 1956) with the descending branches often densely spinose (text-fig. 7a, b). The growth and resorption zones of the adult loop show the same basic pattern as that illustrated by Williams (1968, p. 25) in Magellania flavescens (Lamarck). The main difference lies in the fact that the spinose zones of the descending branches of the adult loop of Z. leckenbyi consist of a double ribbon of shell material (PI. 85, figs. 4, 6). The spines develop in the same manner as the anterior spurs and show a similar fibre enveloped prismatic core (PI. 85, fig. 5). DEVELOPMENT OF SPINES Occasional spines appear on the descending elements even during the syncampagi- form phase but they only become numerous during the late dalliniform phase. The development of spines is considered by Elliott to be indicative of some power of secretion of calcite by cirri. In Z. leckenbyi the spines can be shown to be a fundamental part of the loop. At first it was thought that the spines might be ‘unused’ anteriors of spurs which had become isolated by resorption and owed their orientation to rotation of the loop axis during growth. Later it was realized that the spurs always lie relatively close to the mid-line of the valve and could in no way migrate to the observed position of the spines on the descending elements. It appears, therefore, that the spines and spurs follow the same pattern of development. The implication of this will be discussed later. FUNCTION OF THE LOOP From a consideration of Rudwick's work (1962) on filter-feeding mechanisms in brachiopods. It seems certain that during the pre-paramagadiniform growth phase the brachial apparatus supported a schizolophe and that during the paramagadiniform phase TEXT-FIG. 6a-d. Reconstructions of juvenile brachial valves of Z. leckenbyi. The figures are based essentially on data obtained from stereoscanned cellulose acetate peels but the evidence was reinforced by polished sections and separated valves, a, b. Brachial and lateral views of a late terebrataliiform shell (37582). All connections with the septal pillar have been resorbed. The descending branches are relatively massive and the ascending branches and transverse bar are differentiated. The ascending lamellae are now only represented by unresorbed lamella remnants, l.r. The loop rises very high above the commissural plane, almost touching the floor of the pedicle valve. The septalium is fully developed and the median septum is formed by the rapid extension of the septalial plates to envelop and extend beyond the remains of the septal pillar, c, d. Brachial and lateral views of a typically dalliniform early adult loop (37664). The descending branches, ascending branches, and transverse bar are of relatively massive proportion. The anterior spurs have by now been eliminated by resorption. The anterior extension of the median septum is relatively slower and the lateral profile of the loop is becoming flatter. Numbers indicate the length in mm of the shell from which the peels were obtained. 466 PALAEONTOLOGY, VOLUME 15 it supported a zygolophe. As the ascending lamellae diverged and the transverse bar appeared, the zygolophe would develop into a plectolophe. At the beginning of the syncampagiform phase material is added rapidly to the descending elements so that the descending branches become relatively massive. It seems that the posterior projections from the ascending complex provide an initial framework for the accelerated development of the descending elements, i.e. they act as ‘formers’ to enable the descending elements to progress rapidly but the directive control over the morphology of the anterior of the loop is exercised by the ascending elements. Elliott (1953) suggests that selection pressure is directed towards the development of a more efficient lophophore and the fact that most terebratulide loops are believed to have supported a plectolophe indicates that this form of organization was in some way advantageous to the animal. Elliott further maintains that any mutation favouring the earlier attainment of an adult pattern of lophophore would be selected and correlates this with the advantages conferred by earlier attainment of an adult loop. As in Z. leckenbyi the anterior spurs and spines follow the same development pattern it seems probable that they appeared as undifferentiated structures on the descending and ascending elements of ancestral forms. At some point in time, possibly under the influence of strong selection pressure certain spines on the ascending elements could be ‘fortuitously’ utilized for the extension of the descending branches and, therefore, for the increase in size of the loop. The importance of a mutation which enabled the animal to utilize the anterior spines in a locating role, as a means for rapidly extending the anterior portion of the loop, will be readily appreciated. EXPLANATION OF PLATE 84 Figs. 1-12. Stereoscan photomicrographs of cellulose acetate peels of selected horizontal serial sections through a juvenile Z. leckenbyi (37570/1 ) showing the various elements of the syncampagiform loop. Peel interval 20 [xm. Material of all figures coated with evaporated aluminium before photography. 1. Ascending lamellae and connection with the dorsal edge of the descending branches. Peel 10, X 40. 2. Enlarged view of the ascending complex shown in fig. 1. X 120. 3, 4. Ascending lamellae with the anterior spurs beginning to develop. Peel 1 1, original and retouched copy, X 1 10. 5, 6. Ascending lamellae with the lower anterior spurs and descending branches. Peel 12, original and retouched copy, x70. 7, 8. Descending branches (left one with spine) beginning to extend along the spurs. Peel 14, original and retouched copy, X 50. 9, 10. Ascending lamellae separated, with the upper anterior spurs almost reaching the anterior margin of the shell. Peel 25, original and retouched copy, X 36. 11,12. Ascending lamellae united posteriorly to form the transverse bar. Peel 32, original and retouched copy, x 20. EXPLANATION OF PLATE 85 Figs. 1-6. Stereoscan photomicrographs of Z. leckenbyi. Material for all figures obtained from cellulose acetate peels coated with evaporated aluminium before photography. 1 . Horizontal section through the median septum of an early terebrataliiform specimen (37617/7). x45. 2. Montage of an enlarged portion of the proximal end of the septum shown in fig. 1, to show the pillar remnant enveloped by the development of the septum. X 260. 3. Montage of the point of union between the ascending and descending elements, upper right, of a late terebrataliiform specimen (37582/25) showing the remnant of the anterior spur, lower centre. Horizontal section. X 275. 4. Enlarged portion of fig. 6 showing the detail of the prismatic inner, left, and fibrous outer, right, double ribbon of the descending branches of the loop. X 265. 5. Montage of a horizontal section through a portion of a descending branch of an adult loop (37661/13) showing the prismatic core of a spine sectioned obliquely. X 260. 6. Montage of a horizontal section through a portion of the distal end of a descending branch of an adult loop (37661/11) showing the prismatic inner ribbon repeatedly deflecting the outer fibrous ribbon to form cored spines. Arrow indicates anterior of loop. X 105. Palaeontology, Vol. 15 PLATE 84 1 8 10 11 12 BAKER, Zeilleriid loop Palaeontology, Vol. 15 PLATE 85 BAKER, Zeilleriid loop P. G. BAKER: ZEILLERIID LOOP 467 TEXT-FIG. 7a, b. Brachial and lateral reconstructions of the adult (zeilleriiform) loop of Z. leckenbyi (37661). Loop long, with spinose descending branches. It may be distinguished from the dalliniform type by its more slender descending branches and ribbon-like ascending branches and transverse bar. c.p.l. cardinal process lobe. Number refers to the length in mm of the shell from which peels were obtained. 468 PALAEONTOLOGY, VOLUME 15 Elliott (1948) regards the peak of brachial evolution, expressed by the Terebratellidina, to be intraregressional and coinciding with the radiation of the relatively more mobile Bivalvia. The arguments advanced appear logical and the zeilleriids, which developed adult-pattern (in the terebratellid sense) loops very early, should have remained domi- nant. In fact they were the first to die out during the Lower Cretaceous, although other terebratellidine forms, e.g. Kingena, Trigonelliua, and Ziltelina, having juvenile loops in the zeilleriid sense but presumed to be adults, appear later and have loops capable of supporting a small plectolophe. The Zeilleriacea are largely replaced by the Terebratel- lidae and the later representatives of the Dallinidae and Terebratulidina. It is felt that it is incorrect to correlate density of population with elaboration of loop as there is no evidence to suggest that competition was any less severe in terebratulidine colonies where forms with short loops appear to be no less successful. Indeed, the terebratulacean Terebratulina, with its very strong, fused loop with heavy anterior spiculation, is probably one of the most successful of all Mesozoic-Recent brachiopods. The advantage, there- fore, appears to lie in the possession of a plectolophe, and it appears that selection pressure was geared to the early attainment of this organ however it may be arrived at. A view of this sort is necessary to explain the continuance of short- and long-looped forms and also the various apparently neotenous forms which co-exist with them. With regard to loop form it seems probable that long loops have evolved several times. Williams and Wright (1961) regard the enteletacean Tropidoleptus as a homeomorph of the juvenile terebratellacean loop. However, the development pattern of zeilleriid, dallinid, and terebratellid loops displays such community of character that it seems likely that they are very closely related. All patterns of loop can be accounted for by differing degree and time of onset of calcification during development of the same type of lophophore. The unifying theme is apparently the early attainment of an ascending complex from an outgrowth from the floor of the brachial valve. INTER-RELATIONSHIPS OF MESOZOIC LONG LOOPED TEREBRATULIDA Phylogenetic aspects of brachiopod microstructure have been studied previously by Baker (1970) and Williams (1968, 1970). Elliott (1953) considers the median septum fundamental to the Terebratellacea, for in modern short-looped forms the septum is absent and there are various extinct long-looped forms, not terebratelloids, where, so far as is known, the septum played little part in the development of the loop. Therefore it seems that the demonstration that the median septum sensu lato may be relieved of its role as a loop former in zeilleriids, may be important in a phylogenetic sense. It has been demonstrated that the early phases in the development of the zeilleriid loop have terebratellid counterparts whilst the later phases have a dallinid aspect. At present there is no information as to whether the same similarities exist at the microstructure level. However, it seems probable, as divergence of the anterior of the lower edges of the campagiform hood to produce long spurs is well known in dallinids (Elliott 1953). In addition Elliott noted that in growth stages of higher genera and when first attained in adult terebrataliiform genera, the terebrataliiform stage is somewhat angular, often with short spines at the anterior points of recurvature of the descending and ascending branches. It appears, therefore, that when the ontogeny of the loops of other spinose P. G. BAKER: ZEILLERIID LOOP 469 forms is studied more critically, the role that the anterior spurs play in the extension of the descending elements will prove to be a character not unique to Z. leckenbyi. Asymmetry of loop development is described by Fischer and Oehlert (1892) and in several juveniles of Z. leckenbyi, the right ascending lamella was found to be more strongly developed. Investigation by Babanova (1965) of Aulacothyris pah (Buch) and A. karabugasensis Moisseiv confirmed the presence of a connection between the loop and the median septum early in ontogeny. Babanova proposed that the genus Aulacothyris Douville be removed from the Zeilleriidae and together with several similar forms should be placed in a new tribe Aulacothyrini assigned to the Dallinidae. The basis for this argument is that the Zeilleriidae are characterized by the absence of a connection between the loop and the median septum even in the earliest stages of ontogeny. This view, propounded by Stehli (1956h) and Dagis (1958, 1959), if carried to its conclusion, will in the light of the present discovery necessitate the removal of typical zeilleriids from the Zeilleriidae. Z. leckenbyi may be shown to pass through a connected phase quickly and very early in ontogeny, i.e. size range 2-00-4-00 mm length. Similar phases of development are seen in Aulacothyris karabugasensis and A. pah but over a much larger size range, i.e. 4-00-9-5 mm length. In A. karabugasensis the campagiform stage (Babanova 1965, fig. 3n) has a short broad loop, not deeply dissected into two lobes anteriorly. The general morphology and absence of spurs indicates that the early development regime of the aulacothyrid loop differs from that of Z. leckenbyi. However, the late terebrataliiform loop of Z. leckenbyi, particularly the morphology of the ascending elements closely resembles the terebrataliiform stage of A. karabugasensis. The specimens studied by Stehli and Dagis are also significantly larger than the early growth phases recorded in Z. leckenbyi. Stehli ( 19566) studied specimens of an unnamed terebratellacean from Peru which were 7-5 mm in length. This was subsequently (Elliott 1959) named Eodallina peruviana, although its systematic position remains uncertain (Muir-Wood et al. 1965, p. H844). Dagis (1958) described the development of Zeilleria agechbokensis Moiss. through a size range 5-2-8-7 mm length. The loop of this species is very similar to the syncampagiform loop of Z. leckenbyi, though not actually united with the median septum in the material studied. In 1959 Dagis published a further paper on the development of Aulacothyropsis reflexa (Bittner) and Pseudorugiteh pulchelh (Bittner). The actual dimensions of the specimens are not listed, but the distance from the umbo to the end of the loop is 5-6 mm and 10-00 mm respectively. As Z. leckenbyi appears to be closely related to ornithellids it is probable that a similar ontogeny will be present in that genus also. Obviously the early juveniles of Z. leckenbyi represent some of the smallest zeilleriid material as yet available for study. In view of the evidence, therefore, it seems that possession of a connection with the median septum during the early phases of development is not valid justification for the removal of genera from the Zeilleriidae. It is hoped that early juveniles of other genera will eventually become available, together with data on the development regimes of other known ontogenies. It is felt that a basic pattern of loop development in dallinids and related forms exists. Correlation of the various development regimes will enable the systematic position of the Zeilleriidae to be fixed more accurately. With regard to the systematic position of the Zeilleriidae relative to that of the Terebratellacea it is necessary to review the criterion for separation of the Dallinidae 470 PALAEONTOLOGY, VOLUME 15 and Terebratellidae. This criterion at present lies in the dallinid hood and the terebra- tellid ring and the mode of development of the descending branches of the loop. The difference between the hood and the ring may not be as fundamental as it appears. Elliott (1960) has described an Upper Cretaceous terebratellid, Australiarcula artesiana sp. nov. which possesses a primitive magadiniform loop with well-developed descending branches and also a rudimentary hood. The same type of structure has been noted in some paramagadiniform specimens of Z. leckeribyi, the entire ascending complex during this phase of development bearing a striking resemblance to the loop of Bouchardia rosea (Mawe). There is also a resemblance between the early ascending lamellae of Z. leckeubyi and the two divergent plates which constitute the early development of the loop of Kraussina. However, it is unlikely that these forms are derived neotenously from the zeilleriid type as all three genera are without the dental plates which are invariably present in Z. leckeubyi, even in the earliest juveniles it has been possible to isolate (0-8-0-9 mm long). Elliott also noted (1953) the spiny nature of many mature and immature dallinid loops and the almost complete absence of spines in the terebratellids. In view of the latter point it appears that the bulk of the development pattern of the loop of Z. leckeubyi, despite its early terebratellid aspect, is much closer to that of the Dallinidae. It is hoped that the present paper will stimulate research into the continuous growth phases of the Juvenile loops of other forms, for critical investigation of the fabric of the developing loop must provide the solution to many of the systematic problems posed by the varied assemblage of genera at present comprising the Jurassic Terebratellidina. The evidence now available shows that some terebratellid forms exist which possess a rudimentary hood characteristic of the dallinids. Also, that zeilleriids pass through a development phase comparable with these terebratellid types. The inference being, there- fore, that by varying the degree and/or order of expression of the elements of the ascending complex, zeilleriid stock is potentially ancestral to both the Dallinidae and the Terebratellidae. Suppression of the posteriorly arching spurs and accelerated develop- ment of the hood would initiate a dallinid line, whereas elimination of the hood would consolidate a terebratellid line. AFFINITIES WITH PALAEOZOIC TEREBRATULIDES The concept of investigating developing loops at the microstructure level is quite novel and work in the hrst instance, therefore, must be largely descriptive. However, there is some merit in attempting to analyse the results from an evolutionary point of view. It is clear that if the septal pillar is disregarded the anterior of the ascending complex of Z. leckeubyi during the paramagadiniform phase of development has a distinct affinity with the divided septum of certain Lower Carboniferous centronelliform types, e.g. Gaciua (Stehli 1956n, p. 197). It is generally agreed by most workers that selection pressure favoured rapid attainment of a larger lophophore and presumably a larger loop. The most obvious advantage of a terebratellid loop is that it allows loop development to proceed from two points, i.e. the median septum and the dorsal cardinalia. It has been demonstrated that in Z. leckeubyi the ascending elements can arise from a septal pillar without the involvement of a median septum sensu lato. All that is needed, therefore, to derive the zeilleriid ascending complex from Carboniferous centronelliform types, is a P. G. BAKER: ZEILLERIID LOOP 471 slight change of organization to allow the ascending complex to be supported by a septal pillar. However, as previously mentioned, the presence of spines on the ascending elements appears to be a very important ancestral character for these appear to be the mechanism by which rapid anterior advancement of the loop was attained. Dental plates, being the first internal structures to appear, are also presumed to be important. It seems, therefore, that in evolution the zeilleriid loop progressed via a modified centronelliform type through a cryptonelliform type in which the ascending and descend- ing elements were marginally spinose, e.g. Glossothyropsis type. The characters, dental plates, cardinal plate (supported by median septum or unsup- ported) primitively centronelliform and spinose cryptonelliform loop are to be found in cryptacanthiinin centronellidines, allowing a Gacina-Glossothyropsis type trend. Spinose cryptonelliform loops, dental plates, and a supported or unsupported cardinal plate are, however, also found in cryptonellacean genera such as Heterelasma. Unfortunately information on the early ontogeny of cryptonelliform loops is not at present available and passage through a modified centronelliform phase of development cannot be demonstrated. The zeilleriid median septum sensu lato, being a subsequent development, need not be regarded as a critical factor when considering ancestry and appears to be antedated by the cardinal plate. The one basic problem remaining, therefore, whether the zeilleriid loop be derived from centronellidine or cryptonellid stock, is the appearance and resorp- tion of the septal pillar. A great deal more work is necessary but having brought the importance of the median septum into perspective, initial studies indicate that the origin of the Zeilleriidae may be found in cryptacanthiinin stock of Glossothyropsis type and that zeilleriid and cryptonellid loops may have arisen independently. The cryptonellacean loop has a degree of symmetry which suggests that it may itself be the end product of an evolutionary trend. Probably the most important single contribution made by this paper is the demonstration that in attempting to trace the origin of the Zeilleriidae the presence of spinose ascending and descending elements is likely to be of more significance than the absence of a median septum. REFERENCES ACER, D. V. 1956. Some new Liassic terebratuloids. Proc. Geol. Assoc., Land. 67, 1-14. BABANOVA, L. I. 1965. New data on Jurassic brachiopods. Intermit. Geol. Rev. 7 (8), 1450-1455. BAKER, p. G. 1969. The ontogeny of the thecideacean brachiopod Moorellina granidosa (Moore) from the Middle Jurassic of England. Palaeontology, 12, 388-399. 1970. The growth and shell microstructure of the thecideacean brachiopod Moorellina granidosa (Moore) from the Middle Jurassic of England. Ibid. 13, 76-99. DAGis, A. s. 1958. Loop development in some Triassic Terebratulida. Lett. TSR Mokslii, Akad. Geol. ir Geog. Inst. Mokslinai Dranesiniai, SSR, Trudy, ser. B, 3 (15), 175-182. (In Russian.) 1959. New genera of Triassic Terebratulida. Ibid. 9, 23-41. (In Russian.) ELLIOTT, G. F. 1947. The development of a British Aptian brachiopod. Proc. Geol. Assoc., Lond. 58, 144-159. • 1948. The evolutionary significance of brachial development in terebratelloid brachiopods. Ann. Mag. nat. Hist. (12), 1 , 297. 1953. Brachial development and evolution in terebratelloid brachiopods. Biol. Rev. 28, 261-279. 1959. Six new genera of Mesozoic brachiopoda. Geol. Mag. 96, 146-148. 1960. A new Mesozoic terebratellid brachiopod. Proc. Geol. Assoc., Lond. 71, 25-30. 472 PALAEONTOLOGY, VOLUME 15 FISCHER, p., and oehlert, d. p. 1892. Mission scientifique du Cap Horn (1882-1883). Brachiopodes. Bull. Soc. Hist. nat. Autun, 5, 254. HENDRY, R. D., ROWELL, A. J., and STANLEY, j. w. 1963. A rapid parallel grinding machine for serial sectioning of fossils. Palaeontology, 6, 145-147. MUIR-WOOD, H. M. 1934. On the internal structure of some Mesozoic Brachiopoda. Phil. Trans. Roy. Soc. Loncl. 223, B, 511-567. ELLIOTT, G. E., and HATAi, K. 1965. Mesozoic and Cenozoic Terebratellidina. In moore, r. c. (ed.). Treatise on invertebrate paleontology, part H, Brachiopoda, H8 16-857. Kansas Univ. Press. RUDWicK, M. J. s. 1959. Growth and form of brachiopod shells. Geol. Mag. 96, 1-24. 1962. Filter-feeding mechanisms in some brachiopods from New Zealand. Linnaean Soc. Lond. Zool. Jour. 44, 592-615. STEHLi, F. G. 1956rt. Evolution of the loop and lophophore in terebratuloid brachiopods. Evolution, 10, 187-200. \956b. A late Triassic terebratellacean from Peru. Washington Acad. Sci. Jour. 46, 101-103. THOMSON, J. A. 1927. Brachiopod morphology and genera (Recent and Tertiary). N.Z. Bd. Sci. Art. 7, 338 pp. WESTBROEK, p. 1967. Morphological observations with systematic implications on some Palaeozoic Rhynchonellida from Europe, with special emphasis on the Uncinulidae. Leidse Geologische Mededelingen, 41, 1-82. 1969. The interpretation of growth and form in serial sections through brachiopods, exemplified by the trigonorhynchiid septalium. Palaeontology, 12, 321-332. WILLIAMS, A. 1968. Evolution of the shell structure of articulate brachiopods. Spec. Pap. Palaeont. 2, 55 pp. 1970. Origin of laminar-shelled articulate brachiopods. Lethaia, 3, 329-342. and ROWELL, a. j. 1965. Morphology. In moore, r. c. (ed.). Treatise on invertebrate paleontology, part H, Brachiopoda, H57-155. Kansas Univ. Press. and WRIGHT, A. D. 1961. The origin of the loop in articulate brachiopods. Palaeontology, 4, 149-176. P. G. BAKER Department of Biological Sciences Derby and District College of Technology Kedleston Road Derby, DE3 IGB Typescript received 16 June 1971 THE BRACHIOPOD ACANTHOCRANIA IN THE ORDOVICIAN OF WALES by A. D. WRIGHT Abstract. Acanthocrania, a rare craniacean brachiopod in the Ordovician of Europe, is described for the first time from Wales, where it occurs in the form of A. papillifera (Roemer). The inarticulate brachiopod Acanthocrania is distinguished from other attached crania- ceans by having its brachial valve ‘ornamented by fine papillae or spines’ (J. S. Williams 1943, p. 71). The type species, Crania spicuJata Rowley 1908, is a Carboniferous form from North America, where the genus is also recorded by Cooper (1956, p. 283) as being widely distributed and fairly abundant in the Ordovician. In contrast, Acanthocrania is rather uncommon in the European Ordovician. The first specimens attributed to the genus in the British Isles were from the Ashgillian Portrane Limestone of Eastern Ireland (Wright 1963, p. 250), from which a small number of silicified brachial valves were classified as Acanthocrania cracentis sp. nov., and Acantho- crania sp. Although the valves of the latter are incomplete, their much coarser ornament in particular enables them to be readily separated from A. cracentis. A subsequent record from the British Isles is that of Temple (1968, p. 18) from Keisley, Westmorland, in beds which must be very close to the Ordovician-Silurian junction; although Temple favours an early Llandoverian age for these beds, a late Ashgillian age cannot be discounted. A specific name was not proposed for the contained Acantho- crania, and indeed the presence of ribs in addition to spines on the exterior of these shells casts some doubt on the inclusion of the species in Acanthocrania. For, as Temple says (p. 19), the presence of both ribs and spines makes it possible to refer the form to either Acanthocrania or Philhedra Koken 1889, and it is the unusually fine nature of the ribs which cause him to include the form in Acanthocrania. The present writer has always taken the presence of any radial ribs as being a characteristic of Philhedra and not Acanthocrania', this would seem also to be the interpretation of Cooper (1956). However, there is some ambiguity in Williams’s original definition, for although he states categorically that ‘radial costae do not occur’, he also says that ‘fine radial striae may be present’ (op. cit., p. 71). Nevertheless in earlier discussion (p. 70) he comments that the ornamentation of his ‘third type’ {Acanthocrania) ‘consists of fine papillae or spines and concentric growth lines’, without any mention of either radial striae or ribs. Thus, although there is some uncertainty, I would interpret Williams’s intention to be that forms included in Acanthocrania should not have obvious radial ornament; and accord- ingly I would regard the Keisley specimens as being better assigned to Philhedra than Acanthocrania. Elsewhere in the European Ordovician several of the forms treated by von Heune (1899) have recently been transferred to Acanthocrania (Wright 1963, p. 249). These comprise three species: Philhedra pustidosa (Kutorga 1846), Philhedra hemipnstidosa Heune 1899 and Craniella (?) papillifera (Roemer 1861). This last is a particularly [Palaeontology, Vol. 15, Part 3, 1972, pp. 473^75, pi. 86.] 474 PALAEONTOLOGY, VOLUME 15 distinctive species of Acanthocrania, and may be readily distinguished from all other described Ordovician forms by the development of spines of two distinct sizes (Heune 1899, p. 318); a similar ornamentation is, however, apparent on the illustrations of a form recently described from the Carboniferous of Montana as A. spinosa by Rodriguez and Gutschick (1967, pi. 41, figs. 10-13). Heune’s uncertain generic placing of Roemer’s species is a reflection of the total lack of evidence at the time concerning the nature of the shell interior, for the species is rare with only two exteriors and an impression being available to Heune including the original single specimen of Roemer. The latter came from an erratic block in Silesia, and was considered by Heune to be of the same age as his own specimens from the Lyckholm Beds (Fib) of Estonia (Vormsi Stage), i.e. Ashgillian. In sorting through material collected from the Ashgill Series of the Llanfyllin District, Montgomeryshire, by a past student of this department. Miss S. E. Devonald, a specimen of a craniacean was found which on close examination turns out to belong to Acantho- crania. The specimen comes from the old quarry behind Belan Farm, situated about 2 miles NW. of Llanfyllin (SJ 116209), which is listed by the Geological Survey in the Oswestry Memoir as locality 22 (Wedd et al. 1929, p. 62). As may be seen from the figures on Plate 86, the specimen possesses spines of two distinct series as in A. papillifera (Roemer), and comparison with the description and illustrations of that species given by Heune (1899, 317-318, pi. 5, figs. 10, 11) leaves little doubt that the two are conspecific; minor variations in form between the specimens are not regarded as having taxonomic significance in these attached craniaceans. The dis- position of the adductor scars and other features of the brachial valve interior is con- sistent with that of other species of Acanthocrania, and confirms the earlier placing of Crania papillifera Roemer in this genus (Wright 1963, p. 249). The pedicle valve is not known for either this or any other species of Acanthocrania, and it would seem that this valve was essentially composed of organic material. Recent examination of the sectioned margin of an attached shell has, however, revealed the presence of a lens of calcite in the position where the marginal rim of the pedicle valve would be expected to occur (Williams and Wright 1970, p. 39). SYSTEMATIC DESCRIPTION Family craniidae Menke 1828 Genus acanthocrania Williams 1943 Acanthocrania papillifera (Roemer 1861) Plate 86, figs. 1-9 1861 Crania papillifera Roemer, pp. 48-49, pi. 5, fig. 14. 1899 Craniella‘1 papillifera (Roemer); Heune, pp. 317-318, pi. 5, figs. 10, 11. EXPLANATION OF PLATE 86 Figs. 1-9. Acanthocrania papillifera (Roemer); Ashgill Series, Belan Quarry, Llanfyllin, Montgomery- shire. 1-4. Latex cast of brachial valve external mould. BB 34089 b. 1, view of antero-lateral surface to show detail of ornamentation, x5; 2-4, lateral, dorsal, and postero-dorsal views of cast, X 3. 5-8. Dorsal, lateral, postero-dorsal, antero-dorsal views of brachial valve internal mould. BB 34089 a. X 3. 9. View of anterior surface of brachial valve external mould showing detail of ornamentation. BB 34089 b. X 5. Palaeontology, Vol. 15 PLATE 86 WRIGHT, Acanthocrania Kiny;.* A. D. WRIGHT: AC ANTHOC RAN I A 475 Material. A single internal and external mould from the grey green mudstones of the Ashgill Series, Belan Quarry, Llanfyllin, Montgomeryshire. British Museum No. BB 34089. Approximate dimensions: length 13 mm, width 12 mm, thickness 7 mm. Description. Brachial valve deep, of irregularly subcircular outline. Anterior profile strongly convex, lateral profile steep and concave posterior to umbo, convex anteriorly; umbo situated at about one-quarter of valve length, below greatest thickness of valve located at about two-fifths of valve length. Valve ornamented by concentric growth lines and irregularly developed growth stages, and spinose ornament consisting of fine spines arranged along growth lines with density of about 4-6 per mm; and much larger hollow spines commonly scattered at 1 mm distance apart, but distance variable, ranging from about 0-5 to 2-0 mm. Exact relation of coarse spines to growth lines, and hollow nature of smaller spines, not established from this material. Punctation present in the form of fine pitting on umbonal region of external cast where spines are less well developed, and in places on internal mould. Brachial valve interior with short limbus partially preserved; limbus with median indentation on posterior margin. Adductor muscle scars well developed with elongatedly suboval anterior pair larger than transversely suboval posterior pair. Muscle field extending to almost half valve length; maximum width across posterior pair 5-2 mm, across anterior pair 7-0 mm. Posterior scars separated by shell thickening anterior to median invagination; anterior scars separated by deep groove along which faint median ridge develops and continues in front of adductor muscle scars where faint, radially disposed markings of mantle canals are visible. REFERENCES COOPER, G. A. 1956. Chazyan and related Brachiopods. Smithson. Misc. Coll. 127, 1-1245, pis. 1-269. HEUNE, F. VON. 1899. Die silurischen Craniaden der Ostseelander mit Ausschluss Gotlands. Verb. Russ.-Kais. Min. Ges. St. Petersb. (2) 36, 181-359, pis. 1-6. RODRIGUEZ, J., and GUTSCHiCK, R. c. 1967. Brachiopods from the Sappington Formation (Devonian- Mississippian) of Western Montana. J. Paleont. 41, 364-384, pis. 41-44. ROEMER, c. F. 1861. Die fossHe Fauna der silurischen Diliivialgeschiebe von Sadewitz bei Dels in Nieder- Schlesien. Fine paldontologische Monographic. Breslau. TEMPLE, J. T. 1968. The Lower Llandovery (Silurian) Brachiopods from Keisley, Westmorland. Palaeontogr. Soc. [Monogr.], 58 pp., 10 pis. London. WEDD, c. B., SMITH, B., KING, w. B. R., and WRAY, D. A. 1929. The Country around Oswestry. Mem. geol. Sun’. Eng. and Wales. Sheet 137. 234 pp. WILLIAMS, A., and wright, a. d. 1970. Shell structure of the Craniacea and other calcareous inarticulate Brachiopoda. Palaeont. Assn. Spec. Pap. 7, 1-51. WILLIAMS, J. s. 1943. Stratigraphy and Fauna of the Louisiana Limestone of Missouri. Prof. Pap. U.S. Geol. Surv. 203, 1-133, pis. 6-9. Washington. WRIGHT, A. D. 1963. The Fauna of the Portrane Limestone. 1. The Inarticulate Brachiopods. Bull. Brit. Mus. (Nat. Hist.) Geol. 8, 221-254, pis. 1-4. A. D. WRIGHT Department of Geology The Queen’s University Typescript received 27 July 1971 Belfast BT7 INN 1 1 C9016 CRYSTAL DEVELOPMENT IN D1 SCOASTERACEAE AND BRAARUDOSPHAERACEAE (PLANKTONIC ALGAE) by MAURICE BLACK Abstract. In both families the skeleton consists of little rosettes of calcite crystals grouped round a central axis. In the Discoasteraceae the crystals are arranged each with its optic axis parallel with the principal ( = central) axis; the individual crystals have a bilateral symmetry with virtual suppression of the trigonal symmetry charac- teristic of inorganically grown crystals. The Braarudosphaeraceae differ in having their crystals arranged with a cleavage plane at right angles to the principal axis, to which the optic axes are oblique; the calcite is internally laminated parallel to this cleavage, and all three cleavages respond differently to etching. It is suggested that these peculiarities in morphology and chemical behaviour result from biochemical controls exerted by the organism during crystal growth, producing complex internal structures on a very fine scale, yet not fine enough to interfere with the crystal-lattice. I HAVE commented elsewhere upon the remarkable control which coccolithophorid cells are able to exert over the crystallization of calcite, enabling them to shape the crystals that they secrete with great precision (Black 1963). Two other families of planktonic algae, the Discoasteraceae and Braarudosphaeraceae, share this ability to regulate the growth of crystal faces, and although the resulting patterns are different, the crystallo- graphic control is equally striking. At present we do not know how this is achieved. The physiological processes leading up to the precipitation of calcium carbonate in the coccolithophorid cell have been investigated in some detail (Paasche 1968, Watabe and Wilbur 1966), and the same general principles are presumably applicable to the discoasters and braarudosphaerids. The morphology of the resulting crystals, on the other hand, is clearly dependent upon controls of a much more specific kind; it is perhaps connected with biochemical pecu- liarities which are constant in any given species, but vary from one species to another and from one family to another. Looked at in this way, structural details of the crystal- line skeleton may be expected to reflect biochemical characters, at present unknown, but likely to have an important bearing upon the taxonomy of these organisms. TERMINOLOGY AND CHOICE OF INDICES Miller’s three-index notation for the trigonal system has considerable advantages over a four-index method for representing the crystal faces in the arms of a discoaster, and for this reason has been adopted here. When the rhombohedron face inclined towards the principal axis is taken as (010), the indices display the bilateral symmetry of a dis- coaster arm with a clarity that would be impossible if a four-index notation were used. The word face is used in its crystallographical sense, and the obverse and reverse sides of the asterolith are referred to as surfaces. Since we do not know which of the two surfaces was external in relation to the cell, we cannot logically speak of them as upper and lower, or distal and proximal. The two surfaces are nevertheless morphologically [Palaeontology, Vol. 15, Part 3, 1972, pp. 476-489, pis. 87-96.] M. BLACK: CRYSTAL DEVELOPMENT IN PLANKTONIC ALGAE 477 unlike, and we can make a crystallographical distinction between the F surface, with a ring of rhombohedron faces round the principal axis, and the E surface with a little star of radiating polar edges. DISCO ASTER ACEAE Discoasters are small fossils, commonly between 10 ftm and 20 ftm in diameter, con- sisting of calcite crystals grouped radially round a central axis so as to form patterns resembling rosettes or snowflakes (PI. 89, fig. 4); these structures are commonly referred to as asteroliths. In most species the crystals lie in a single plane, which may be called the principal plane of the asterolith, at right angles to the principal axis, which behaves as an axis of symmetry for the whole body (these are the Hauptebene and Hauptachse of Stradner, in Stradner and Papp 1961, pp. 48, 49). Optica! orientation. One fundamental property of all discoasters is that their constituent crystals are invariably arranged so that the optic axis of each is parallel with the principal axis. In microscope slides prepared in the usual way, with the asteroliths lying flat against the object-glass, they remain dark between crossed nicols, and are thus easily distinguishable from coccoliths and most other nannofossils. Because of this, it has sometimes been assumed that each asterolith consists of a single crystal. This does not, however, necessarily follow from the optical behaviour, sinee a group of crystals with diflferently oriented lattices would give the same effect provided that the c-axes were all parallel. Examination under the electron-microscope of specimens with recognizable crystal-faces shows that as a general rule each arm of a discoaster is an independent crystal, and it is only in rather special circumstances that several rays are crystallo- graphically united to form a single crystal. Crystallographic observations under the electron-microscope. When discoasters are pre- pared by standard methods for examination under the microscope, their discoidal shape usually causes them to lie flat upon the supporting film or slide, with the principal plane at right angles to the line of vision. From their optical behaviour in polarized light we know that the trigonal axis of each crystal is then aligned parallel with the line of vision. Hence we can treat electron-micrographs of discoasters as orthographic projections on the principal plane, which coincides with the basal pinacoid {111} of the constituent crystals. This provides a convenient method of working out the development of faces on these crystals, and their relation to the general morphology of the complete discoaster. There is an additional advantage in that the plane of projection (1 1 1} is the same as that used in a conventional stereographic projection, and observations made on electron- micrographs can be quite simply related to the positions of poles and the traces of zones plotted on a standard stereogram. On an orthographic projection, and hence on a suitably oriented electron-micrograph, all edges between faces in the same zone are represented by parallel lines, and the common direction of these lines is normal to the diameter of the zone-circle in the stereogram. By measuring the direction of a crystal-edge on an electron-micrograph we ean thus find the projected direction of the zone-axis to which the adjacent faces are related. In the more simply constructed discoasters this is often all that is needed to 478 PALAEONTOLOGY, VOLUME 15 confirm the identity of a face recognized by inspection, for experience shows that the faces most commonly developed all have simple indices. Occasionally the identity of a face can be settled by looking at the symmetry of etch- figures or overgrowths. For example, the basal pinacoid can often be recognized by the presence of overgrowths in the form of regular trigonal pyramids. The orientation of these pyramids fixes the cleavage directions and hence the orientation of the crystal- lattice; it also serves to distinguish (111) from (TTT), and thus provides a means of identifying the upper and lower surfaces of an asterolith. Simple six-armed discoasters. These principles can now be applied to some of the very simply constructed discoasters that are found in Tertiary pelagic sediments. Two species of this kind, Discoaster adamanteus Bramlette and Wilcoxon, and D. obtusus Gartner, can be obtained in an excellent state of preservation from Oligocene and Miocene deep- sea oozes. The shape of the asteroliths is determined entirely by well-developed crystal faces, the actual combination of faces varying slightly in different samples. Specimens of D. obtusus resembling the holotype consist of six primitive rhombohedra, and have the simplest structure possible in a six-armed discoaster (text-fig. 1 ; PI. 87, figs. 3, 4). In the holotype of D. adamanteus the rhombohedra are modified by the addition of faces which give the asterolith a star-shaped instead of a hexagonal outline (text-fig. 2; PI. 88, figs. 2-4). As seen under a light-microscope, the two holotypes are quite distinct in outward shape, but they are accompanied by a range of intermediate forms in which the extra faces are developed to various degrees. This leads to difficulty in drawing a line between the two species, and raises an interesting taxonomic problem. The simplest form of D. obtusus is hexagonal in outline, and consists of six simple rhombohedral crystals each of which has a polar edge running radially along the middle of the arm, flanked by a pair of steeply sloping rhombohedron faces (text-fig. \a\ PI. 87, fig. 4). The third rhombohedron face is directed towards the centre, and is much smaller than the other two. Measurements on suitably tilted specimens confirm the identification of the three faces on this surface of the arm as belonging to the primitive rhombohedron r = {100}. As this unequal development of individual faces belonging to the same form is found repeatedly in the arms of discoasters, it is desirable to adopt a consistent scheme for labelling such faces. For asteroliths constructed on the same plan as D. obtusus, with EXPLANATION OF PLATE 87 Figs. 1-2. Discoaster adamanteus Bramlette and Wilcoxon, Middle Miocene, Trinidad (H851). 1, oblique view of F surface. Face (010) very small, at the centre; (100) and (001) make the flanks of the arms; (101) is not developed on the F surface of this specimen. 24455, x7500. 2. Hemi- discoaster condition, orthographic view of the F surface. A simple combination of r faces without e. Arms 1, 3, and 5 are united to form a single crystal, with suppression of (010). In the alternate arms (010) is represented by very small triangular faces. Figs. 3-4. Discoaster obtusus Gartner. 3, Flemidiscoaster condition, oblique view of F surface. Alternate arms are united to form a single crystal with its trigonal axis at the centre of the asterolith. (100) and (001) are the only faces developed on this surface of the arms. Lower Miocene, Jamaica (H719). 24407, X 12 500. 4, normal (Eudiscoaster) condition, orthographic view of F surface. Face (010) is developed as a ring of very small faces surrounding the centre. Compare text-fig. \a. Lower Miocene, Pacific Ocean (H852). 24420, x7500. Palaeontology, Vol. 15 PLATE 87 BLACK, Planktonic algae M. BLACK: CRYSTAL DEVELOPMENT IN PLANKTONIC ALGAE 479 1 1 TEXT-FIG. 1. Discoaster obtiisiis Gartner in orthographic projection on (1 1 1). The asterolith consists of six arms, each of which is part of a primitive rhombohedron. a-c. F surface (face (010) to centre), d-f. E surface (edge to centre), a, d. The complete asterolith; compare PI. 87, fig. 4 and PI. 88, fig. 1. b, e. A single arm. c,f A single arm in relation to the fully developed rhombohedron. a crystallographic plane of symmetry running along the length of the arm (text-figs. 1, 2), the r face directed towards the centre is selected as (010), and the parallel face which cuts off the tip of the arm on the reverse surface of the asterolith is then (OTO). The two opposite surfaces of the asterolith can thus be distinguished from each other: one has a ring of crystal faces surrounding the centre, and the other has a set of edges radiating away from it (PI. 88, figs. 1, 2). They may conveniently be referred to as the F surface (face-to-centre, with positive indices for r) and the E surface (edge-to-centre, with nega- tive indices for r). Specimens of D. obtusus almost invariably place themselves on 480 PALAEONTOLOGY, VOLUME 15 4 4 ToT TEXT-FIG. 2. Discoaster adauianteiis Bramlette and Wilcoxon in orthographic projection on (111). The arms are parts of crystals with a special habit combining the primitive rhombohedron r = {100} and the obtuse rhombohedron e = {101} with unequal development of the faces of both forms. a-c. F surface, r/. The complete asterolith; compare with PI. 88, fig. 2. Zi. A single arm. c. A single arm in relation to the upper surface of the crystal, in which ( 101) is developed without (110) and (Oil). d-f. E surface, d. The complete asterolith; compare with PI. 88, fig. 4, which shows a slightly different development of faces, e. A single arm. /. A single arm in relation to the lower surface of the crystal, which unlike the upper surface includes (TTO) and (OlT). a microscope slide with F uppermost because this side is steeply conical, and the fossils naturally come to rest on the flatter E surface. Each of the six arms shows a similar development of faces successively rotated through an angle of 60°, and since the lattice must also be rotated through the same angle there can be no crystalline continuity between adjacent arms. In the orthographic projection M. BLACK: CRYSTAL DEVELOPMENT IN PLANKTONIC ALGAE 481 the solid angles that form the distant ends of the rays give an internal angle of 120°, and six-fold repetition produces a regular hexagon. Electron-micrographs of specimens lying flat on the supporting film show a very close approximation to this ideal shape (PI. 87, fig. 4). In examples of D. adamant eus agreeing closely with the type (Bramlette and Wilcoxon 1967, pi. 7, fig. 6) the shape of the arms is modified by additional faces belonging to the obtuse rhombohedron e = {110}. The six faces of this form are hardly ever all developed, and the omission of individual faces is always effected in such a way as to emphasize the bilateral symmetry of each arm (PI. 88, fig. 3; PI. 89, fig. 2). The crystallographic plane of symmetry running along the length of the arm thus remains operative, whereas the other two planes of symmetry are, in effect, suppressed in the external morphology of the crystal. For example, a pair of e faces, (TTO) and (OTT), is conspicuously developed on the E surface, whereas the third face (TOT) is almost invariably suppressed (PI. 88, fig. 4). These two faces intersect the (100) and (001) faces of the F surface, producing a notch in the outline of the asterolith at the distant end of each interradial suture. In many specimens the obtuse rhombohedron is represented by the (TTO) and (OTT) faces only, and there are no faces belonging to this form on the opposite F surface. Occasionally a specimen presenting \\ve; F surface is seen with each of the polar edges of the primitive rhombohedron r truncated by a narrow, parallel-sided e face (PI. 88, fig. 2); more commonly, however, only (101) is developed and (110) and (01 1) are suppressed. A further modification is sometimes introduced by the addition of a basal plane, either (111) or (TTT) (PI. 88, fig. 1). The identity of these faces can often be confirmed by the presence of little_outgrowths in the shape of equilateral pyramids, which on ( 1 1 1) have a face and on (TTT) an edge directed towards the centre of the asterolith. Some of the patterns resulting from these various combinations are shown in Plates 87-89. Judged as discoasters, they all have a primitive appearance. Nevertheless, the organism has clearly exerted a firm control over the growth of its crystals, and in par- ticular has taken considerable liberties with the trigonal symmetry of calcite to achieve some of these results. Each of the six crystals has been prevented from growing larger than a certain pre-determined size, and the pair of faces (100) and (001) has been forced to develop quite differently from (010). Although the interfacial angles necessarily pre- serve a trigonal symmetry, the selective development of faces gives an external form suggestive of a much lower degree of symmetry. The triad axis, all three diad axes, and two of the planes of symmetry are suppressed, leaving the third plane as the only opera- tive element of symmetry in the crystal. This imparts a pseudomonoclinic appearance to the outward shape of each arm, mimicking a crystal of the monoclinic clinohedral class. The discoaster gives the impression of obeying the letter of crystallographic law and disregarding its spirit. This becomes more noticeable with increasing complexity of structure; indeed the elaborate shapes achieved by the more advanced discoasters are a measure of their success in overcoming the morphological limitations which trigonal symmetry imposes upon inorganically grown crystals (PI. 89, fig. 4; PI. 93, fig. 1). Hemidiscoaster structure. We have already seen that there can be no crystalline con- tinuity between adjacent arms of a six-rayed discoaster, since the crystal lattice is rotated through 60°, giving a different orientation in space. Rotation through 120°, on the other hand, brings the lattices into a parallel position, so that crystalline continuity between 482 PALAEONTOLOGY, VOLUME 15 alternate rays is theoretically possible. Fusion of crystals in this way does in fact take place, producing asteroliths in which three of the six rays are united to form a single crystal, separated from the alternate rays by well-marked interradial sutures (PI. 87, figs. 2, 3 ; PI. 89, fig. 2). This structure can easily be seen under a good optical microscope, and was used by Tan Sin Hok (1927, p. 120) to separate the genus Hemidiscoaster, with fusion of alternate rays, from Eiidiscoaster in which all the rays retain their independence. Bramlette and Riedel (1954, p. 394) found it impossible to maintain the distinction between these two genera because several otherwise homogeneous species were found to include both variants. In fact, the special relationship between the crystals in a six-rayed asterolith would lead us to expect such a state of affairs; the possibility of crystalline continuity is always present, but it may well be a matter of chance whether it is actually achieved or not. Pauciramous asteroliths. Some six-armed species of Discoaster are accompanied by a few similarly constructed asteroliths which differ in having a smaller number of arms. Some of these have been given separate specific names, but there are good reasons for regarding them as aberrant representatives of the six-armed forms which they otherwise resemble. Just before its extinction at the end of the Pliocene Period, D. brouweri produced an abundance of such pauciramous forms (Pis. 90-92). These clearly are not all of the same kind. In one type, asteroliths with 3, 4, or 5 arms preserve a vestige of hexagonal sym- metry in that the angles between the arms are always either 60° or 120°, and under the electron-microscope the stump of an undeveloped arm can often be seen bisecting the wider angle. Pauciramous asteroliths of this kind result simply from the failure of some EXPLANATION OF PLATE 88 Fig. 1. Discoaster obtusus, orthographic view of E surface. (TOO) and (OOT) make an arrowhead pair of faces pointing towards the centre; (OTO) is large, sloping outwards. The small equilateral triangles on some of the arms are (TTl). In the upper part of the figure, the median edge of the F surface is faintly visible through the transparent replica. Compare text-fig. Id. Lower Miocene, Jamaica (H719). 24413, X9000. Figs. 2-4. D. adamanteus. Lower Miocene, Pacific Ocean (H852). 2, orthographic view of F surface. The ring of (010) faces makes a small hexagonal crater at the centre; (101) is narrow and not developed on all the arms. Compare text-fig. 2a. 22291, X 9000. 3, orthographic view of F surface \vith (101) in addition to (010), (100), and (001). 22274, x4500. 4, orthographic view of E surface. (TOO) and (OOT) are developed as a pair of very small faces pointing to the centre; (TTO) and (OTT) large, (TOT) not developed; (OTO) is at the outer ends of the arms, not always clearly distinguishable. Compare with text-fig. Id, which differs in having the additional face (TOT). 24442, x4500. EXPLANATION OF PLATE 89 Fig. 1. Discoaster cf. adamanteus. E surface. Near the centre are (TOO) and (OOT) with the intervening edge truncated by (TOT). The lozenge-shaped face is (TTT). (TTO) and (Oil) are relatively large. (OTO) cuts off the tip of the arm. Lower Oligocene, Pacific Ocean (H861). 24051, x4500. Figs. 2-3. D. adamanteus. Lower Miocene, Pacific Ocean (H852). Hemidiscoaster condition, F surface. Arms 2, 4, and 6 fused to form a single crystal, arms 1 , 3, and 5 independent. Each arm is a combina- tion of (100), (001), and (101). The asterolith has three radial planes of symmetry. 24435, x6750. 3, D. adamanteus, abnormal condition, F surface. Arm 2 is fused crystallographically with arm 6, and arm 3 with 5; 1 and 4 are still independent. Each arm is a combination of (010), (100), (001), and (101). The asterolith has only two planes of symmetry. 24444, x4500. Eig. 4. Discoaster chaUengeri Bramlette and Riedel, with bilaterally symmetrical arms giving radial symmetry to the asterolith. Middle Pliocene, Indian Ocean (H845). 24326, x4500. Palaeontology, Vol. 15 PLATE 88 BLACK, Planktonic algae Palaeontology, Vol. 15 PLATE 89 BLACK, Planktonic algae «■ ^1 ,"'f '1". ■ _ ■1:? I ‘••4 s kk '■ f S- : i M. BLACK: CRYSTAL DEVELOPMENT IN PLANKTONIC ALGAE 483 of the arms to grow to a normal length, and no fresh crystallographic problems are involved (PI. 90, fig. 3-Pl. 91, fig. 3). In a second type, there is no vestige of a hexagonal plan, and the asteroliths have a new 3-, 4-, or 5-fold symmetry with angles of 120°, 90°, or 72° between the arms (PI. 91, fig. 4-Pl. 92, fig. 3). In asteroliths with four or five arms the general plan of construction is similar to that of six-armed individuals. Each arm is an independent crystal, separated from its neighbours by interradial sutures and presenting a similar face or edge towards the centre. The lattice having been rotated through 90° or 72° in adjacent arms, the crystals are brought into incongruent positions and have sharply defined contacts. With three-armed asteroliths, the situation is different. By virtue of the trigonal sym- metry of calcite, a rotation of 120° brings the lattices of all three arms into the same orientation. The arms then no longer behave as twinned individuals but act as a single crystal with a homogeneous lattice throughout (PI. 92, figs. 2, 3). For this reason inter- radial sutures are never found in three-armed discoasters or in Marthasterites, which has exactly the same crystallographic structure, and differs only in the sculpturing at the tips of its arms. Departures from radial symmetry. In the great majority of discoasters, the angle between adjacent arms is constant for any particular number of arms, so that the symmetry is truly radial around the principal axis. Occasionally individuals of D. brouweri are found to depart from this rule, having the six arms at unequal angles such as 30°, 60°, and 90°. Such aberrant forms have been observed in other six-armed species, and Gartner (1969, p. 598) has recently described a five-armed derivative of D. brouweri as a new species, D. asymmetrieus, in which unequal spacing is the rule rather than an exception (PI. 91, fig. 4). When such forms are examined in detail, it is often observed that the interradial sutures are regularly disposed at angles of 60° or 72°, and that the irregular spacing is caused by a departure from bilateral symmetry in the arms themselves. Specimens of D. asymmetrieus from Pliocene oozes in the Pacific Ocean have inter- radial sutures close to 60°, 72°, and 84°, which are the sums of possible combinations of 30° and 42°, namely 30+30, 30+42, and 42+42. The arms themselves are asym- metrical, the length of the arm making angles of 30° and 42° with the interradial sutures. If all the five arms were either right-handed or left-handed, a regular spacing of 72° would result, but if arms of both kinds are combined in a single asterolith the observed angles are inevitably produced. The six-armed specimen shown in PI. 92, fig. 4 similarly has interradial sutures making angles of 15° and 45° with the length of the arm, and the introduction of two adjacent left-handed arms gives one angle of 15+15 = 30° and another of 45+45 = 90° instead of the normal 60°. In asteroliths with this kind of arm, the ideal symmetry about the principal axis is of a rotary nature, and obviously cannot be strictly radial even when the arms are all spaced at equal angles. Further development of this type of construction probably leads to such forms as D. lodoeusis with falcate arms and a much more pronounced rotary symmetry (PI. 93, fig. 1). In these more advanced asteroliths recognizable crystal faces are almost invariably replaced by curved surfaces, and the material at present available is not suitable for crystallographic study. 484 PALAEONTOLOGY, VOLUME 15 BRAARUDOSPHAERACEAE This family also builds skeletal plates consisting of crystals grouped radially round a principal axis, but it differs from the Discoasteraceae in the orientation of the crystals. In the genera under consideration, each skeletal plate consists of five units which behave optically as single crystals; for this reason the plates are usually referred to as pentaliths. The pentalith-bearing species are customarily grouped into three genera, Braarudo- sphaera, Micrantholithus, and Pemma, but the distinctions between them are not at all sharp. It is not always easy to draw a definite line between Pemma and Micrantholithus, and some species that have been assigned to Braarudosphaera might equally well have been placed in one of the other genera. In all the species so far examined, each crystal lies with one cleavage in the principal plane of the pentalith, and hence at right angles to the principal axis (PI. 93, figs. 3, 4). For consistency of description, this cleavage is selected as (010). When the crystals are in the form of simply rhombohedra, as in B. africana and M. concimms (PI. 94, fig. 1; PI. 96), they are arranged with their acute angles meeting centrally at the principal axis, and their (100) and (001) cleavages intersect at an angle of 78° on the (010) face. Hence the cleavages are not exactly parallel with the interradial sutures, which are spaced at 72°, and some adjustment is necessary to the shape of the crystals in order to make them fit compactly into a whorl of five sectors. In these two species the adjustment is made symmetrically, and this is probably true also of most species of Micrantholithus and Pemma (PI. 94, figs. 3, 4; PI. 95, figs. 1, 4). It is not true, however, for all species of Braarudosphaera, and in B. bigelowi there is a considerable latitude in the way that the sectors are pared down to prevent the sum of their radial angles exceeding 360°. Optica! orientation. Optically, the slow directions are arranged radially round the princi- pal axis and the optic axes lie, not in the principal plane, but inclined to it at an angle of about 45°. From this it follows that there can be no planes of symmetry in the archi- tecture of a pentalith, whatever the external form may suggest, the only symmetry possible being of a rotary kind around the principal axis. Pentaliths are more akin to the coccoliths in this respect than to the discoasters. EXPLANATION OF PLATE 90 Figs. 1-4. Discoaster brouweri Tan Sin Hok. 1, oblique view of convex surface. Middle Pliocene, Indian Ocean (H845). 24320, X7500. 2, orthographic view of convex surface. Lower Pliocene, Pacific Ocean (H850). 22203, X 10000. 3, convex surface, showing unequal development of arms. Lower Pliocene, Pacific Ocean (H850). 22204, X 5000. 4, convex surface. Five-armed modification of six-armed form by failure of the sixth arm to develop fully. Upper Pliocene, Pacific Ocean (H849). 24353, X5000. EXPLANATION OF PLATE 91 Figs. 1-3. D. brouweri. 1, concave surface. Five-armed modification of six-armed form by failure of the sixth arm to develop. Middle Pliocene, Indian Ocean (H845). 24278, X 9000. 2, convex surface. Four-armed modification of six-armed form by failure of two opposite arms. Hemidiscoaster condi- tion with fusion of alternate arms across the centre. Middle Pliocene, Indian Ocean (H845). 24299, X 6000. 3, convex surface. Three-armed modification of six-armed form, by failure of alternate arms. Upper Pliocene, Pacific Ocean (H849). 24371, x6000. Fig. 4. Discoaster asymnietricus Gartner, convex surface. This is a five-armed species with unequal spacing of the arms. Pliocene, Pacific Ocean (H728). 16984, x9000. Palaeontology, Vol. 15 PLATE 90 BLACK, Planktonic algae Palaeontology, Vol. 15 PLATE 91 BLACK, Planktonic algae M. BLACK: CRYSTAL DEVELOPMENT IN PLANKTONIC ALGAE 485 In Pemma and Micrantholithus the slow direction nearly, but often not exactly, bisects the radial angle of each sector, which consequently shows approximately symmetrical extinction under the polarizing microscope. In Braanidosphaera bigelowi and probably in most species of this genus the extinction directions are less regular; they are not always spaced at equal angles round the principal axis, nor do they show any constant relationship to the radial edges of the sectors. Their only constant feature is that the slow-vibration direction always lies within the radial angle of each sector. Quite clearly the pentagonal symmetry suggested by the external shape of Braanidosphaera does not apply to the internal structure of the pentaliths. In the other two genera, the optical properties suggest a closer approximation to five-fold rotary symmetry; this is confirmed by the cleavage directions, which are often visible under the electron-microscope (PI. 95, figs. 1-3). Details of fine structure. The pentaliths of B. bigelowi have a peculiar layered structure, giving an appearance rather like the cleavage in a sheet of mica (PI. 93, fig. 4). Each of the five segments is constructed of about 20 very thin laminae parallel with (010). Although the layering is thus parallel with one of the cleavage directions of the crystal, it is not in itself a cleavage in the mineralogical sense, since each layer is a sheet of finite thickness with sharply defined upper and lower boundaries. The true mineral cleavage is impossible to see in well-preserved specimens. When corrosion has cut obliquely across the laminae, each layer is seen to be crossed by two sets of cleavage planes which do not pass from one lamina to the next. Apparently the calcite layers are separated from each other by some substance which prevents cleavage- fractures from passing beyond it, on the principle of laminated safety-glass. In a pentalith 2-0 ftm thick the calcite laminae are about OT pm in thickness, and the separating layers appear to be much thinner. Perhaps the most remarkable feature of this arrangement is that the organism has picked out one of the three crystallographically identical cleavage directions and has treated it differently from the other two, which never show any sign of developing a comparable interlamination. The structure within the pentalith is thus more complex than would be expected from observations in polarized light. A similar laminar structure has been observed in other species of Braanidosphaera and also in Micrantholithus and Penvna. In B. discula (PI. 93, fig. 2) the layered structure is much like that of B. bigelowi, and there is a similar inconstancy of extinction directions. In this and in most other species of the genus, the distal margins of the segments show little or no relation to crystallo- graphic directions. The earliest known species, B. africana (PI. 94, fig. 1), is exceptional in having star-shaped pentaliths whose segments have an almost unmodified rhombo- hedral form. Its outward shape is thus much like that of Micrantholithus concinniis, but without the localized thickening that is characteristic of this and other species of Micrantholithus. The pentaliths of Penvna differ from those of other Braarudosphaeraceae by the presence of a small pit or perforation near the centre of each segment (PI. 95, fig. 2). In polarized light they react much like the pentaliths of Braanidosphaera, each segment behaving as a single crystal with the slow vibration direction arranged radially. The spacing is more regular than in Braanidosphaera', in P. rotiindiini, the type species, each 486 PALAEONTOLOGY, VOLUME 15 segment is very close to 72°, and is bisected by the slow-extinction direction, usually within the accuracy of measurement. Measurements of intersecting cleavages on the faces of the segments, if made well away from the radial sutures, always give angles close to 78° and 102°, with the acute angle towards the centre; the third cleavage (010) must therefore lie parallel to the principal plane of the fossil. The segments are laminated parallel to (010) as in Braarudo- sphaera, and this structure gives rise to striking effects in etched specimens (PI. 95, fig. 3). Pentaliths in the Bracklesham Beds, for example, are often deeply corroded with destruction of the original outline and enlargement of the pits to form large rhomb- shaped cavities with stepped surfaces. So far as is known, the intimate structure of Micrantholithiis resembles that of Pemma in all respects. The difference between these genera lies in the outward shape of the segments and the etched forms resulting from natural corrosion. In the type species, M.flos, each segment has a deep triangular notch cutting into it from the circumference, leaving the segment V-shaped (PI. 95, fig. 4). This shape is characteristic of numerous species of Micrantholithus in much the same way that a pit in the centre of each segment is characteristic of most species of Pemma. Response to corrosion is curiously different in the three genera under review, and it is difficult to see why this should be so. In Braanidosphaera corrosion is apt to start at the interradial sutures, producing narrowed segments quite unlike those of the original fossil (PI. 94, fig. 2). In Pemma it usually attacks the pits, which become enlarged by etching of the component laminae to produce peculiar stepped surfaces (PI. 95, fig. 3). EXPLANATION OF PLATE 92 Figs. 1-4. D. brouweri, convex surface. 1, 2, 4. Middle Pliocene, Indian Ocean (H845). 1, four- armed form showing interradial sutures. 24294, x6750. 2, three-armed form with the arms fused into a single crystal, and no interradial sutures. 24286, X 9000. 3, corroded specimen with etch- pits. Upper Pliocene, Pacific Ocean (H849). 24364, x4500. 4, irregular spacing of arms discussed on page 483. 24297, x6750. EXPLANATION OF PLATE 93 Fig. 1. Discoaster lodoensis Bramlette and Riedel. Asymmetrical arms giving rotary symmetry to the asterolith. Lower Eocene, Donzacq, France (H695). 15771, X 3750. Fig. 2. Braanidosphaera discida Bramlette and Riedel, distal surface of slightly corroded specimen showing lamination parallel to the principal plane, (010). Lower Eocene, Donzacq, France (H695). 15782, X6000. Figs. 3-4. Braanidosphaera bigelowi (Gran and Braarud) Deflandre. 3, distal surface, which is parallel to (010). Upper Eocene, Mississippi (H696). 15750, x 9000. 4, B. bigelowi, proximal surface. Three sectors of a pentalith showing lamination parallel to (010). Middle Eocene, Bracklesham Bay, Sussex (H807). 19895, x9000. Fig. 5. Braanidosphaera hoschiiltzi Reinhardt., distal surface. Three sections of a naturally corroded pentalith with the (010) lamination picked out by differential etching. Lower Cretaceous, North Sea (H1055). 29388, X 13 000. EXPLANATION OF PLATE 94 Figs. 1-2. Braanidosphaera africana Stradner. 1, distal surface of relatively undamaged specimen. Lower Cretaceous (Aptian), Alford, Lincolnshire (H599). 25334, x4500. 2, corroded specimen. Lower Cretaceous (Albian), Mildenhall, Suffolk (H934). 26491, x9000. Figs. 3-4. Pemma papillatiim Martini. Middle Eocene, Bracklesham Bay (H807). 3, non-punctate surface. 19920, X6750. 4, punctate surface. Corroded specimen showing unilateral etching. 19893, X 6750. Palaeontology, Vol. 15 PLATE 92 BLACK, Planktonic algae ^4. Palaeontology, Vol. 15 PLATE 93 BLACK, Planktonic algae Palaeontology, Vol. 15 PLATE 94 BLACK, Planktonic algae M. BLACK: CRYSTAL DEVELOPMENT IN PLANKTONIC ALGAE 487 The effects in MicrcmihoUthus are in some respects the most remarkable of all. Species with V-shaped notches are attacked from the notch, leaving resistant bars along the interradial sutures, often more pronounced at one arm of the V than at the other. This selective one-sided etching is most clearly seen in M. conciiimis, which has simple unnotched rhombohedral segments (PI. 96). The original surface in the principal plane is by definition (010), and the laminated structure within each segment is parallel with this. The (100) and (001) cleavage traces on the (010) surface are nearly parallel with the interradial sutures. Corrosion works back from the outer margin, not sym- metrically, but by cutting a series of steps controlled by the intersection of the (010) and (001 ) cleavages, thinning the segment at one of its outer margins and leaving a thick rib along the suture which is nearly parallel with the (001) cleavage trace. The (100) cleavage plays no part at all in this effect. We thus have the extraordinary situation that the three cleavages, crystallographically indistinguishable, respond differently and quite indepen- dently to solution during the corrosion of the pentalith. This independent behaviour of the cleavages presumably results from small-scale structural peculiarities that were built into the mineral substance of the pentaliths during the lifetime of the organism. The unique behaviour of the (010) cleavage can be readily explained by the presence of interlaminated sheets of isotropic material, which are easily seen at magnifications of a few thousand times. If, as seems likely, the different behaviour of the (100) and (001) cleavages is similarly due to associated structural features, these must be on such an extremely fine scale that magnifications up to 50 000 are not adequate to resolve them. DISCUSSION AND CONCLUSIONS From this brief review it is evident that although the Discoasteraceae and Braarudo- sphaeraceae both construct their skeletal plates by grouping a few crystals more or less symmetrically round a central axis, they have adopted fundamentally different plans for arranging these crystals and controlling their growth. The Discoasteraceae orient all their crystals with the trigonal axes parallel to the principal axis of the asterolith; elonga- tion is at right angles to the trigonal axis, and the development of each crystal is related to a single plane of symmetry, the other two planes being suppressed. The Braarudo- sphaeraceae, on the other hand, arrange their crystals so that one cleavage always lies in the principal plane of the pentalith, and the optic axes are oblique to both the principal plane and the principal axis. Moreover, the calcite of each crystal is interlaminated with sheets of some other substance parallel to the (010) cleavage. The other two cleavages play a different and quite subordinate role in the structure of the pentalith. There is thus a clear morphological difference between the two families which ought to be given taxonomic recognition in any classification based upon purely morphological characters. Deflandre (1950, p. 1 158; in Piveteau 1952, p. 109) divided the Coccolitho- phoridae into two groups, the Heliolithae, a structure spheroUthique, and the Ortho- lithae, a structure crystalline. These two groups have come to be treated as orders, the coccoliths proper being placed in the Heliolithae, and the discoasters and braarudo- sphaeres united together in the Ortholithae. The fossils of both groups are of course all crystalline, the structure spheroUthique being merely a special way of arranging the crystals in which a rotary symmetry is usually involved. We have seen that rotary symmetry is also characteristic of the Braarudosphaeraceae which, in this respect, have 488 PALAEONTOLOGY, VOLUME 15 more in common with the coccoliths than with the discoasters, whose symmetry is prevalently radial. The distinction between Heliolithae and Ortholithae, useful though it was at the time of its original introduction, now needs reconsideration in view of our more detailed knowledge of fine structure. The structural and crystallographical differences between the Discoasteraceae and Braarudosphaeraceae clearly result from biological controls which the living cells were able to exert over the crystallization of calcite. Research into the nature of these controls by biochemical methods is ruled out as far as the Discoasteraceae are concerned, for this family appears to be totally extinct. There is, however, one species of the Braarudo- sphaeraceae still living, and an investigation of carbonate secretion in cultures of B. bigelowi is needed to discover the biochemical conditions within the cell that enable the organism to differentiate between crystallographic directions that are indistinguishable according to inorganic laws. The purely palaeontological evidence reviewed above cannot carry us very far along this line of inquiry, but it does nevertheless give a glimpse of the structural devices by which the properties of calcite can be modified without interfering with the crystal lattice. Acknowledgements. It is a pleasure to thank Professor P. Allen for reading the manuscript, and for his many helpful suggestions. Carbon replicas for the illustrations were prepared and photographed by Mr. D. Stubbings, using an A.E.I. EM6 electron-microscope provided by D.S.I.R. (now N.E.R.C.). Prints are by Mrs. Rolfe and Mr. M. H. Waring. I am most grateful to the Council of Trinity College, Cambridge for a generous contribution towards the cost of publication of this paper. REEERENCES BLACK, M. 1963. The fine structure of the mineral parts of the Coccolithophoridae. Proc. Linn. Soc. London, 174, 41-46. BRAMLETTE, M. N., and MARTINI, E. 1964. The great change in calcareous nannoplankton fossils between the Maestrichtian and Danian. Micropaleontology, 10, 291-322. and RIEDEL, w. R. 1954. Stratigraphic value of discoasters and some other microfossils related to recent coccolithophores. J. Paleont. 28, 385-403. and SULLIVAN, f. r. 1961. Coccolithophorids and related nannoplankton of the early Tertiary in California. Micropaleontology, 7, 129-188. and WILCOXON, j. a. 1967. Middle Tertiary calcareous nannoplankton of the Cipero section, Trinidad, W.l. Tulane Stud. Geol. 5, 93-131. EXPLANATION OF PLATE 95 Figs. 1-3. Pemma rotundum Klumpp. Middle Eocene, Bracklesham Bay (H807). 1, non-punctate surface with rosette of rhombohedra indicating crystal-orientation. The visible face of each crystal is (OTO). 19867, x4500. 2, punctate surface, slightly corroded and showing pits of normal size. Relics of the original (010) surface are preserved near the centre. 19908, x4500. 3, punctate surface, deeply corroded with great enlargement of the pits and displaying lamination parallel with (010). The (100) and (001) cleavage directions are picked out near the periphery. 19894, x6750. Fig. 4. Micrantholithus pinguis Bramlette and Sullivan. Upper Paleocene, Lodo Canyon, California (H1053). 28674, x9000. EXPLANATION OF PLATE 96 Fig. 1. Micrantholithus concinnus Bramlette and Sullivan, corroded specimen showing internal lamina- tion parallel with (010) and the effects of unilateral corrosion controlled by the intersection between this lamination and the (001) cleavage. Upper Paleocene, Lodo Canyon, California (HI053). 28663, X 1 1 000. Palaeontology, Vol. 15 PLATE 95 BLACK, Planktonic algae Palaeontology, Vol. 15 PLATE 96 BLACK, Planktonic algae M. BLACK: CRYSTAL DEVELOPMENT IN PLANKTONIC ALGAE 489 DEFLANDRE, G. 1934. Lcs Discoasterides, microfossiles calcaires incertae sedis. Bull. Soc.frang. micr. 3, 59-67. 1947. Braaruclospliaera nov. gen., type d’une faniille nouvelle de coccolithophorides actuels a elements composites. C. R. Acad. sc. Paris, 225, 439-441. 1950. Observations sur les coccolithophorides, a propos d’un nouveau type de braarudo- sphaeride, Micrantholitlius, a elements clastiques. Ibid. 231, 1156-1158. 1952. Classe de coccolithophorides, in J. piveteau. Trade de paleontologie, 1, 107-115. Paris: Masson. GARTNER, s. 1967. Calcarcous nannofossils from Neogene of Trinidad, Jamaica and Gulf of Mexico. Univ. Kan. Paleont. Contrib., Paper 29, 1-7. 1969. Correlation of Neogene planktonic foraminifer and calcareous nannofossil zones. Trans. Gulf-Cst. Ass. geol. Socs. 19, 585-599. MANiviT, H. 1965. Nannofossiles calcaires de I’Albo-Aptien. Rev. nn'cropaleont. 8, 189-201. MARTINI, E. 1965. Mid-Tertiary calcareous nannoplankton from Pacific deep-sea cores. Colston Papers, 17, 393-411. and BRAMLETTE, M. N. 1963. Calcareous nannoplankton from the experimental Mohole drilling. J. Paleont. 37, 845-856. NOEL, D. 1960. Revision du genre Discoaster Tan Sin Hok. Bull. Soc. Hist. nat. Afr. Nord, 51, 201-299. PAASCHE, E. 1968. Biology and physiology of coccolithophorids. Annual Review of Microbiologv, 22, 71-86. ROTH, p. H. 1970. Oligocene calcareous nannoplankton biostratigraphy. Eel. Geol. Helv. 63, 799-881. STRADNER, H. 1963. Ncw Contributions to Mesozoic stratigraphy by means of nannofossils. Proc. Sixth World Petrol. Congr., Sec. 1, 167-183. and PAPP, A. 1961. Tertiiire Discoasteriden aus Osterreich und deren stratigraphischen Bedeutung. Jahrb. Geol. Bundesanst. Wien, Sonderbd. 7, 1-159. TAN SIN HOK. 1927. Over de sammenstellung en het onstaan van krijt en mergelgesteenten van de Molukken. Jaarb. Mijnw. Nederl.-Indie, 55, 1-165. WATABE, N., and WILBUR, K. M. 1966. Effects of temperature on growth, calcification, and coccolith form in Coccolithus huxleyi (Coccolithineae). Limnol. Oceanogr. 11, 567-575. M. BLACK Department of Geology Sedgwick Museum Downing Street Cambridge CB2 3EQ Typescript received 13 October 1971 A NEW SPECIES OF PROTOCETUS (CETACEA) FROM THE MIDDLE EOCENE OF KUTCH, WESTERN INDIA by A. SAHNi ami v. p. mishra Abstract. The recent discovery of a cetacean from the Middle Eocene (Lutetian) beds of south-western Kutch is the earliest record of mammals from India. The Archaeoceti are represented in India by a new species of Protocetus, Protocetus sloani sp. nov., based on a partial skull and two mandibular fragments. The age of the bone-bearing horizon is Lutetian on the basis of an associated foraminiferal assemblage. Studies of Tertiary vertebrates in India have mainly centred on the Neogene Siwalik beds of the Himalayan foothills, but recent finds of fossil cetaceans and other vertebrates in the Babia Stage (Lutetian) of the Berwali Series of Kutch have opened up new areas of palaeontological interest, having a significant bearing on the Palaeogene palaeobio- geography of the subcontinent. The vertebrates were collected in situ in a dry stream section, about 3 km south-east of Baranda (lat. 23° 34' 20" N., long. 68° 43' 10" E.) in south-western Kutch. The beds are highly fossiliferous shallow-water marine deposits with abundant gastropods and pelecypods. On the basis of a rich foraminiferal assemblage, a Lutetian age has been determined for the sequence which is given below: Oligocene (Lattorfian) Glauconitic limestone Unconformity Cream limestone White marl Calcareous sandstone Grey marl Middle Local unconformity Eocene (Lutetian) Gritty sandstone Gypsiferous clay Chocolate limestone Green clay Variegated clay Ferruginous shale U nconformity — ^ — Lower Eocene (Ypresian) Yellow marl Bones are obtained from three main horizons. The oldest producing horizon is the chocolate limestone which contains well-preserved bones and vertebrae. The main horizon, however, appears to be the overlying Gypsiferous clay from which the poorly preserved cranial and mandibular fragments were collected. The third bone bed is the slightly younger Grey marl overlying a local unconformity. Associated with the cetacean material are some turtle plastron fragments, and an isolated tooth, probably of a fish. The Tertiaries are particularly well developed in south-western Kutch and have been studied in considerable detail by Biswas (1965) and Biswas and Deshpande (1970). The (Palaeontology, Vol. 15, Part 3, 1972, pp. 490-495, pl. 97.] SAHNI AND MISHRA: EOCENE CETACEAN FROM INDIA 491 sequence ranges from Upper Cretaceous to Recent. The Deccan basalts are overlain by lateritic conglomerates and tuffaceous shales with dicotyledonous leaf impressions and other plant remains of Palaeocene age (Madh Series). Unconformably overlying these beds are shales, marls, and limestones of Eocene age. The Eocene in Kutch has been divided on the basis of foraminifera into a lower and upper division, corresponding to the Laki and Kirthar Series of the type-section of Sind- Baluchistan, respectively. The foraminiferal assemblage from the Middle Eocene rocks south-east of Baranda, has been described by Tandon (unpub. thesis, Lucknow, 1966) and includes Nummulites acutus, N. stamineits, N. beaumonti, Discocyclina dispansa, D. javana, D. sowerbyi, Dictyocorwides cooki, Halkyardia minima, and A/veolina elliptica. Nuttall (1926), Tewari (1957), and Biswas and Deshpande (1970) have clearly demon- strated that the Middle Eocene beds of south-western Kutch are of Lutetian age. The Eocene beds are overlain successively by a thin band of Oligocene arenaceous marls, a thick sequence of Mio-Pliocene sandstones and carbonates, and a thin bed of Miliolitic Limestone of Pleistocene age. Recent and subrecent alluvium and wind-blown sand blanket parts of the Tertiary sequence. There is no previous record of a Palaeogene mammal from India. Of the countries adjacent to India, Palaeogene mammal localities are so far known only from West Pakistan and Central Burma, two widely separated regions. In the north-western region of West Pakistan a mammalian fauna has been described from the Chharat Series of Middle Eocene age, and includes taeniodonts, creodonts, condylarths, brontotheres, anthracotheres, and other Artiodactyla (Pilgrim 1940, Dehm and Oettingen Spielberg 1958). The Creodonta are represented by at least four genera, of which two belong to the family Mesonychidae. The presence of mesonychids is significant in the context of Van Valen’s (1966, p. 93) suggestion that this group is ancestral to the Archaeoceti. In Central Burma, Primates, Artiodactyla, and Perissodactyla have been reported from the Pondaung Sandstones of Upper Eocene (Auversian) age (Pilgrim and Cotter 1916, Pilgrim 1925, 1928). Other records of Protocetidae from the Eocene are: a partial scapula of Anglocetus from the London Clay (Lower Eocene) of England (Tarlo 1964), Protocetiis from the Lower Lutetian of Egypt and Texas, and Pappocetus from the Lutetian of Southern Nigeria, and Eocetus from the Upper Eocene of Egypt. The other two families of Eocene whales are the shorter-skulled Dorudontidae and the gigantic Basilosauridae. SYSTEMATIC DESCRIPTION Order cetacea Brisson 1762 Suborder archaeoceti Flower 1883 Family protocetidae Stromer 1908 Genus protocetus Fraas 1904 Protocetus sloani sp. nov. Plate 97, figs. 1-7; text-fig. 1a, b, c Etymology. The species is named after Dr, Robert E. Sloan, Department of Geology, University of Minnesota, U.S.A. Diagnosis. Skull with a low sagittal crest, and a small lambdoidal crest; parietal narrow and elongated; temporal fossa large. Teeth extending far behind the anterior part of the orbits. Skull similar to that K k C 9016 492 PALAEONTOLOGY, VOLUME 15 of Protocetus atavus (Fraas 1904) but shorter in length with a larger foramen magnum and tympanic bullae. Lower dentition differing from that of Pappocetus lugardi (Andrews 1920); canine small, no diastema between C and Pj; and the symphysis extends posterior to Pj. Posterior part of lower jaw massive. P4 longer than M^, outer surface of jaw strongly curved, inner surface concave particularly near the ventral side. Holotype. Anterior mandibular fragment, Specimen Lucknow University Vertebrate Palaeontology (L.U.V.P.) No. 11002 (PI. 97, figs. 4, 5; text-fig. 1a, b). Museum, Geology Department, Lucknow University, Lucknow. Locality. Stream section approximately 3 km south-east of Baranda (23° 34' 20" N., 68° 43' 10" E.) in the Gypsiferous clays of the Berwali Series, south-western Kutch, India (Toposheet No. 41 A/10). Age. Middle Eocene (Lutetian). TEXT-FIG. 1. Protocetus sloani sp. nov. A, posterior cross-sectional profile (Specimen 11002) xl; B, anterior cross-sectional profile (Specimen 1 1002) X 1 ; c, posterior cross-sectional profile (Specimen 11003) X 1. DESCRIPTION The material of Protocetus sloaui sp. nov. consists of a partial skull and two edentulous jaw fragments. The skull (L.U.V.P. 1 1001), which was found some distance away from the lower jaws, is in a highly gypsified and friable condition and lacks the portion anterior to the frontals (PI. 97, figs. 1, 2, 3). The skull closely resembles that of Protocetus atavus (Fraas 1904) in dimensions and morphological features and is regarded as congeneric with it. Both the condyles are present in the specimen and the foramen magnum is large. As in P. atavus, the supraoccipital forms an almost vertical wall supporting the posterior part of the brain case. Although some of the postero-dorsal bones of the cranium are broken, it is apparent that neither the lambdoidal nor the sagittal crests were well developed, a condition EXPLANATION OF PLATE 97 Figs. 1-7, Protocetus shaui sp. nov. 1-3 xf; 4-7 X approx, f; 1-3, Right lateral, dorsal, and posterior views of L.U.V.P. 11001. 4-5, Left lateral, and occlusal views of L.U.V.P. 11002 (Holotype). 6-7, Left lateral, and occlusal views of L.U.V.P. 11003. Palaeontology, Vol. 15 PLATE 97 SAHNI and MISHRA, Eocene cetacean from India SAHNI AND MISHRA: EOCENE CETACEAN FROM INDIA 493 characteristic of the Protocetidae but differing from other Archaeoceti, such as Dorudontidae and Basilosauridae, where both the crests become very prominent. The highly altered and gypsified nature of the skull has obliterated many details, but on the right side, the contact of the parietal with the squamosal and frontal can be made out. These bones occupy the same position as in other members of the Protocetidae. The brain case is narrow and elongated, and the temporal fossa is large. The post-orbital process of the frontal is broken. The bulla is well developed, and there appears to be a wide area between the petrosal and the basioccipital bone, a condition characteristic of mesonychids (Van Valen 1966, p. 92). The basioccipitals are covered by matrix. The palatine is narrow and elongated, and its contact with the maxilla is not clear. The maxilla bears a three-rooted molar series, spaced together and extending posteriorly behind the anterior rim of the orbits. The crowns of the teeth have been eroded away, only the roots and alveoli are visible. The lower dentition consists of two mandibular fragments, which were found close together, but not in definite association with each other. They probably represent two different individuals of the same species. Of these L.U.V.P. 1 1002containsthesymphysisandalveolifor C, Pi, Pj.and Pg, while L.U.V.P. 1 1003 is more robust, with alveoli for P4 and Mi and the anterior alveolus for Mg. Though the length and depth of these jaws are almost identical to those of Pappocetus liigardi (Andrews 1920), the dimensions of the teeth are consistently smaller. L.U.V.P. 11002 is an anterior mandibular fragment with right and left rami (PI. 97, figs. 4, 5; text-fig. 1a, b). The symphysis is strong and is marked by a deep groove which becomes more pro- nounced posteriorly up to the posterior alveolus for Pg whence the right and left rami begin to diverge. The outer sides of the rami on the dorsal surface are convex at the symphysial region, a condition that was noted in Pappocetus htgardi by Andrews (1920). In general, the jaw is laterally compressed, and the cross-sectional profiles are given in text-fig. 1a and b. Near the ventral side of each ramus, there is a deep groove which becomes shallower posteriorly, fading out completely at the point of divergence of the two rami. As the jaw is edentulous, only tentative views can be expressed concerning the teeth that occupied the alveoli in the jaws. The assessment has been made on the basis of the following assumptions. Firstly, that C, Pj are single-rooted and the teeth posterior to Pi are two-rooted; and, secondly, that a condition similar to that in other protocetids, where the point of divergence of the two rami is either at the posterior end of Pg or the anterior root of Pg, prevails. The alveolus for the left canine is incomplete and there is no diastema between it and Pj. Judging from the shape of the alveolus for Pj, the first premolar was directed forward and slightly upwards. Pg is separated from Pi by a diastema of 16 mm. Pg appears to be two-rooted and the posterior alveolus is larger. Pg is separated from Pg by a diastema of 8 mm and, as in Pg, the anterior alevolus is much smaller than the posterior. Measurements (in cm) of L.U.V.P. 11002 Eength of right mandibular fragment 9-2 Estimated length of C alveolus 0-65 Length of left mandibular fragment 7-5 Length of Pj alveolus 1-0 Depth of right mandibular fragment at Pi 4-5 Length of Pg alveoli 2-65 Depth of right mandibular fragment at Pg 5-2 Estimated length of Pg alveoli 2T The only other mandibular fragment in the collection is a robust left jaw (L.U.V.P. 11003) with an asymmetrical cross-sectional profile (PI. 97, figs. 6, 7; text-fig. Ic). The outer side is strongly convex near the dorsal surface of the jaw, while the inner is only slightly convex and becomes concave near the ventral border. The anterior tooth is probably P4. The anterior root of P4 which is exposed, is marked by longitudinal ridges. The alveoli of the Mg are unequal, the anterior alveolus being larger. The anterior alveolus for Mg is larger. Measurements (in cm) of L.U.V.P. 11003 Length 7-3 Length of P4 3-7 Depth at P4 6 0 Width of P4 1-5 Depth at Mi 6-3 Length of Mj 2-5 Width of anterior alveolus of Ml 1-5 494 PALAEONTOLOGY, VOLUME 15 REMARKS. Protocetus is based on a complete skull from Egypt, while Pappocetus is based on two mandibular fragments from Nigeria. The possibility that these genera are not distinct was suggested by Van Valen (1966, p. 92). Judging from the present material, however, it appears that the genera Protocetus and Pappocetus are independent but related. The skull from south-western Kutch closely resembles that of Protocetus, while the lower jaw fragments are clearly different to those of Pappocetus. As the association of the mandibular fragments and the skull from Kutch is uncertain, the distinction between the two genera still remains tentative. From the chocolate-coloured limestones underlying the Gypsiferous clays, some vertebrae, turtle carapace fragments, and an isolated tooth were collected. The vertebrae, which are four in number, closely resemble the vertebrae described by Andrews (1920) and others, and are cetacean. The turtle carapace fragments have a characteristic surface texture of closely spaced, rounded tubercules, similar to the texture of some species of Aspideretes. The isolated tooth is laterally compressed and acutely pointed, about 9 mm in length and probably belongs to a teleost. The grey coloured marls overlying the local unconformity have yielded only broken rib pieces and unidentified bones. CONCLUSIONS The significance of the discovery of a fossil cetacean from the Middle Eocene beds of south-western Kutch is five-fold. First, evidence in hand seems to suggest that the genera Protocetus and Pappocetus are distinct. Secondly, the new species represents one of the older records of the group, from beds that have with certainty been dated as Lutetian. Thirdly, its occurrence in western India extends the geographical distribution of Middle Eocene Archaeoceti to southern Asia. At present primitive archaeocetes have been described from the Lower Eocene of England (Tarlo 1964) as well as the Lutetian of Nigeria, Egypt, and Texas, U.S.A. Upper Eocene finds are confined to North America, North Africa, and Europe. Fourthly, the occurrence of protocetids in widely separated areas indicates that the forms could freely migrate across interconnecting seaways. One such route could have been a seaway extending continuously from Africa to its eastern limit in western India. Fifthly, in view of the fact that the new species Protocetus sJoani is the oldest mammal (Middle Eocene) to be reported from India, its bearing on the Palaeogene palaeobiogeography of southern Asia becomes significant. REFERENCES ANDREWS, c. w. 1920. A description of new species of Zeuglodont and of Leathery Turtle from the Eocene of Southern Nigeria. Proc. zool. Soc. Loud. 22, 309-319, pis. 1, 2. BISWAS, s. K. 1965. A new classification of the Tertiary rocks of Kutch, Western India. Bull. geol. min. metall. Soc. India, 35, 1-6. -and DESHPANDE, s. V. 1970. Geological and Tectonic maps of Kutch. Bull. O.N.G.C. {India), 7 (2), 115-120. DEHM, R., and OETTiNGEN SPIELBERG, T. zu. 1958. Paliiontologische und geologische Untersuchungen im Tertiar von Pakistan; 2, Die mitteleocanen Siiugetiere von Ganda Kas bei Basal in Nordwest Pakistan. Bayer. Akad. Wiss., mat.-nat. Kl., Abh. 91, 5-54. FRAAS, E. 1904. Neue Zeuglodonten aus den unteren Mitteleociin vom Mokattam bei Kairo. Geol. palaeont. Abh. 10 (3), 199-220, pis. 10-12. SAHNI AND MISHRA: EOCENE CETACEAN FROM INDIA 495 NUTTALL, w. L. F. 1926. The zonal distribution and description of the larger Foraminifera of the middle and lower Kirthar (Middle Eocene) of parts of Western India. Rec. geol. Siirv. India, 59, 115-164, pis. 1-8. PILGRIM, G. E. 1925. The Perissodactyla of the Eocene of Burma. Palaeont. indica, n.s. 8 (3), 1-28. 1928. The Artiodactyla of the Eocene of Burma. Ibid. 13, 1-39. 1940. Middle Eocene mammals from north-west India. Proc. zool. Soc. Land. IlOA, 127-152. and COTTER, G. de p. 1916. Eocene mammals from Burma. Rec. geol. Surv. India, 47, 42-77. TARLO, L. B. H. 1964. A primitive whale from the London Clay of the Isle of Sheppey. Proc. Geol. Ass. 74 [for 1963], 319-323. TEWARi, B. s. 1957. Geology and the stratigraphy of the area between Waghpadar and Cheropadi, Kutch, Western India. J. palaeont. Soc. India, D. N. Wadia Jubilee, 2, 136-148. VAN VALEN, L. 1966. Deltathcridia, a new order of mammals. Bull. Am. Mas. nat. Hist. 132, 1-126. 1968. Monophyly or Diphyly in the origin of whales. Evolution, 22, 37-41. A. SAHNI V. P. MISHRA Department of Geology University of Lucknow Lucknow 7, India Typescript received 30 June 1971 FOSSIL WOOD OF PLATANUS FROM THE BRITISH EOCENE by DONALD W. BRETT Abstract. The fossil wood studied comes from the Landenian, Ypresian, and Pleistocene (Red Crag, presumed derived from Ypresian) of south-east England. Two specimens anatomically indistinguishable from the trunk wood of living species of Platanus are described as Platanus sp. : other specimens are described as Plataninium decipiens sp. nov. ; these have certain features seen in wood of branch bases and roots of Platanus, but differ slightly from ordinary trunk wood of the living genus. Twelve additional specimens are assigned to the new species. An emended diagnosis of the organ-genus Plataninium Unger is given. The material described was collected at coastal localities in Suffolk, Essex, and Kent, and derives from three geological horizons, Landenian, Ypresian, and Pleistocene. The two principal specimens were given to me for investigation in 1954 by H. E. P. Spencer {Platanus sp.) and G. E. Elliott {Plataninium). Fossil wood in the Landenian beds in south-east England is silicified, whilst similarly well-preserved wood in the London Clay is calcified. Pyritized and carbonized twigs and wood fragments are abundant in the London Clay, along with the well-known pyritized fruits and seeds. Soft humified wood and lignite are probably present in most of the British lower Tertiary horizons, but all this other material is usually poorly preserved. The fossil wood in the Red Crag (Pleistocene) has been generally regarded as derived from the Tertiary beds that underlie the Crags in Suffolk. This wood is calcified, like that of the London Clay, but it is orange-rust coloured and contains very little organic matter. Although polished, the Red Crag wood does not show signs of excessive abra- sion: many pieces have bark attached, and there is a specimen in the Ipswich Museum with a thin vine coiled round it. These pieces of wood have presumably been re-deposited close to their original source and so it is most likely that they are from the local London Clay. These records of Platanus extend the list of occurrences of non-tropical types of trees in the Eocene of south-east England. DESCRIPTION OF THE FOSSILS FAMILY Platanaceae Platanus sp. Plate 98, figs. 1-4; Plate 99, figs. 1, 2 V44298 (including slides) in Department of Palaeontology, British Museum (Nat. Hist.), London; cut from a larger piece in the Museum, Ipswich, Suffolk. From the London Clay at Harwich, Essex. Stem about 10 cm across; medulla 1-5 cm wide, xylem cylinder with distinct growth rings. Vessels 1 50 per mm^, tangential diameter 20-72 pm (average 50 /rm), mostly crowded in multiples and irregular clusters, also solitary especially in late wood where vessels are less numerous; intervascular pitting opposite, perforations scalariform with 8-16 bars. Fibres 15-20//m across, with bordered pits; several rows of flattened fibres at end of late wood. Rays about 3 per mm, 1-22 cells wide (average 7) in transverse section; in tangential section uniseriate or part-uniseriate rays very rare, most rays broad, [Palaeontology, Vol. 15, Part 3, 1972, pp. 496-500, pis. 98-99.] D. W. BRETT: EOCENE PLATANUS WOOD 497 fusiform, 2-3 mm high, occasionally to 6 mm high and dissected diagonally by fibres. Ray cells mostly procumbent, narrow, with broader or square cells at margins. Crystals common in procumbent cells. Wood parenchyma diffuse and in short wavy strings of cells. Secondary phloem with clusters of large stone cells, phloem ray cells packed with crystals. V8036-8 (not illustrated). From the Red Crag at Woodbridge, Suffolk. Specimen consists of the inner 6 rings of wood with part of medulla. Vessels distorted by compression, about 120 per mm^. Rays 2-4 per mm, up to 15 cells wide; many rays 3 mm high, dissected rays 7 mm and more. FAMILY ?Platanaceae Organ-genus Plataninhim Unger emend, herein Emended diagnosis Fossil secondary wood, or stems or roots with some secondary wood. Rays of all widths commonly averaging 5-15 cells, uniseriate rays rare; about 3 per mm; height to above 1 mm, highest rays several mm sometimes diagonally dissected by fibres. Ray cells mostly procumbent, usually square or upright in marginal tiers; tangential walls often oblique in transverse view, especially at growth ring boundary; crystals common in procumbent cells. Diffuse or graded porous; vessels mostly solitary, or in small clusters and tangential multiples especially in early wood; 20-150 per mm'^; tangential diameter rarely above 100 ^m, narrower vessels in late wood when graded porous. Perforation plates mostly scalari- form, 1-30 bars, or simple in larger vessels in early wood. Fibres with bordered pits. Parenchyma scanty where vessels are crowded, otherwise abundant in wavy tangential strings of cells often extending from ray to ray. Plataninium decipieiis sp. nov. Plate 99, figs. 3-6 Diagnosis. Secondary xylem; growth rings mostly obscure. Vessels mostly solitary, about 30 per mm^, tangential diameter 35-105 p-m (average 80 p-m), intervascular pitting opposite to scalariform, perfora- tion plates scalariform with 13-25 bars (average 18); rays about 3 per mm, 2-18 cells wide (average 6) in transverse section; in tangential section rays mostly 1-6 mm high; ray cells mostly procumbent except at ray margins, richly pitted, commonly crystalliferous. Wood parenchyma diffuse and in strings of cells, abundant. Holotype. V45684 in Department of Palaeontology, British Museum (Nat. Hist.), London. Horizon. London Clay (Ypresian); Isle of Sheppey, Kent. Additional specimens. V45685: from the beach at Herne Bay, Kent, this small piece of silicified wood with some bark is presumed to have come from the Thanet or Woolwich Beds (Landenian) in the cliffs at this locality. Vessels mostly solitary, except for overlapping ends, occasionally in groups of 2-4 forming a short tangential row; 42 per mm^ excluding the wider rays; tangential diameter 54-120 pm (average 88 pm). Scalariform perforation plates with 17-32 bars (average 25). Rays 1-3 per mm in transverse section, 4-25 cells wide; in tangential section the wider rays reaching height of 5 mm, acutely tapered at margins except where two rays are vertically in line. Bulk of ray cells procumbent, marginal cells more or less square. The following specimens in the British Museum (Natural History) agree fairly closely in structure with the type of the new species and are also assigned to it: London Clay; Sheppey, Kent, 32662, V8354-5; Herne Bay, Kent, V21733; Harwich, Essex, V10253-4. Red Crag: Woodbridge, Suffolk, 52696, V7815, V8009, V8011, V8018, V8031, V8054-5. DISCUSSION The woods described above do not differ in any major anatomical detail from the wood of living species of Platamis. Whilst there can be little doubt about the identity of 498 PALAEONTOLOGY, VOLUME 15 the specimen described above as "Plat anus sp.’ I do not wish to ignore the differences which exist between this specimen and the others I have described in this paper. Thus I have placed these other specimens in a new form-species, Plataninhim decipiens. In doing so I have recognized the possibility that these fossils belonged to a natural genus other than Platanus. Nevertheless, the structure of these fossils seems to fall within the known range of variation of wood from the extant genus. It is unfortunate that mature wood (as distinct from small branches on herbarium sheets) is not available of some species of Platanus likely to be of great interest to the palaeobotanist — namely those from Mexico, Guatemala, and Laos. Available timber specimens and published descrip- tions of the major species {P. occidentalis, P. orientalis, P. racemosa, P. nrightii, P. lindeniana, P. mexicana) show that the wood varies little from one species to another, and a 2-year-old branch of P. kerrii gives no indication that its trunk would be dis- tinguishable from any of the other species. The development of late-wood is very variable between one growth increment and another in timber specimens, suggesting that this is a feature readily influenced by the environmental or climatic conditions during the growing period, but the early-wood almost always commences with rather larger, crowded pores. This is not so, however, in the root. I have examined a root of about 6 cm diameter from a large Platanus Xacerif alia. Compared to normal trunk wood, the root has fewer and more rounded pores with no obvious diminution in size through the growth increment; growth ring boundaries are obscure; and the broad rays are closer together. The bases of branches may also be different, with less crowded pores and a lot of parenchyma. This can be seen in Plate 98, fig. 3, which shows the wood in a small branch base (‘knot’) of the fossil Platanus', in this case the parenchyma and many other cell cavities are filled with dark material, and the distribution of the wood parenchyma can be easily seen. At least these give some indication of possible variation in wood of Platanus. In some respects these variants show similarities to Plataninium decipiens sp. nov. and to another species recently described as Plataninium californicum (Page 1968). One specimen of P. californicum is a small branch so there is no question of regarding this type of wood as solely root material, although this is clearly a possibility in the case of the holotype of P. decipiens. EXPLANATION OF PLATE 98 Figs. 1-4. Platanus sp. 1, Transverse section showing full width of a growth ring, early wood at bottom. X 40. 2, Transverse section at growth ring boundary showing the crowded early wood vessels at top and characteristic widening of the broad ray at the boundary. Many oblique tangential walls of the ray cells are also visible. X 220. 3, Transverse section of wood in branch base embedded in the wood of the specimen described. The abundant wood parenchyma is revealed clearly by its dark contents. X 70. 4, Tangential section showing ends of broad rays. X 70. EXPLANATION OF PLATE 99 Figs. 1, 2. Platanus sp. 1, Radial section showing the procumbent cells which constitute the bulk of the ray tissue. 2, Radial section showing heterogeneous ray cells such as occur along the ray margins, and elsewhere in the ray at the growth ring boundary and in narrower rays. X 100. Figs. 3-6. Plataninium decipiens. 3, Transverse section; parenchyma revealed by dark contents. X 70. 4 Transverse section at growth ring boundary showing the widening of the ray and oblique tangential walls. X 140. 5, Tangential section showing the broad rays. x70. 6, Radial section showing the heterogeneous composition of a narrow ray or ray margin; numerous crystal pseudomorphs are visible as small light patches in the cells. x70. Palaeontology, Vol. 15 PLATE 98 BRETT, Eocene Platanus wood Palaeontology, Vol. 15 PLATE 99 BRETT, Eocene Plantanus wood D. W. BRETT: EOCENE PLATANUS WOOD 499 Several fossils agreeing very closely with wood of the common Platanus spp. have been described from the Tertiary of Europe and North America (Edwards 1931, Mathieson 1932, Slijper 1932, Hofmann 1952, Selmeier 1957, Prakash and Barghoorn 1961). Most of them are so similar to one another, so far as one can judge from the descriptions, that it is hard to justify the conventional use of a different specific name for each. The older descriptions (v. Edwards 1931) omitted details that a wood anatomist would wish to know about today, but the material was usually well preserved and the authors quite convinced about the identity of their material with the living genus. A consequence of this is that the fossil genus Plataniniiim (along with many others) was never given a proper diagnosis, and came to be accepted as simply the name for fossil E/utou/s' wood. (Whether or not the newer names ending in -xylon are accepted and used still depends on one’s regard for the priority of Unger’s generic names. This is probably a trivial matter com- pared with the fact that so few of these ‘genera’, whatever suffix they bear, are provided with an adequate diagnosis.) Mathieson incorrectly used the name Plataniies for wood from Greenland; there is no evidence for connection with the leaves so named, although it seems more than likely that they were from the same natural species of tree. A diagnosis could be constructed on this basis (i.e. fossil wood of Platanus L.) using such information as we at present have about the wood of extant species, but there is little advantage in that, and the description of any part of the living species is simply an extension of the diagnosis of the natural genus. If there exists a group of modern genera, related or otherwise, which cannot be separated on the grounds of wood anatomy alone, it should be possible to construct a diagnosis for a form-genus to include the whole range of wood structure in the natural genera. Kriiusel (1949) has discussed this fully in relation to the wood of conifers. I attempted to do this for the fossil wood genus Querciniwn (syn. Qiiercoxylon) which is best used only for certain types of wood found in the Eagaceae, but not for all (Brett 1960). An emended diagnosis for Plataninium has been given by Page (1968). It is in general terms, not in the usual style of a diagnosis, and includes the statement that Plataninium is ‘a form-genus for fossil woods resembling certain members of the Eagaceae (Fagus), Platanaceae (Platanus), Eupteleaceae (Euptelea), and Icacinaceae (Citronella, Otto- schultzia), whose familial relationships cannot be determined with certainty’. ADDENDUM Since the above was written some new fossil Platanus and Platanus-WkQ woods have been described from Oligocene deposits in Northern Bohemia (Prakash, Bfezinova, and Buzek 1971). The new Platanus wood has been named Platanoxylon hohemicwn sp. nov. and it is virtually indistinguishable from that which I have described above and from the wood of living species. They have also described a Platanus-Wko. wood, Plataninium europeanum sp. nov., which appears to be very similar to the specimens I have described in the present paper as P. clecipiens sp. nov. and there is therefore a stronger possibility that these belong to a distinct natural genus apart from Platanus. Also of interest is the occurrence with Platanus in the Northern Bohemian Oligocene of Cercidiphyllum (Cercidiphylloxylon kadanense sp. nov.) since fossil wood of this genus also occurs in the London Clay (Brett 1956). 500 PALAEONTOLOGY, VOLUME 15 REFERENCES BRETT, D. w. 1956. Fossil wood of Cercidiphyllum Sieb. & Zucc. from the London Clay. Ann. Mag. nat. Hist. (12) 9, 657-665. 1960. Fossil oak wood from the British Eocene. Palaeontology, 3, 86-92. EDWARDS, w. N. 1931. FossiUuni catalogiis, 2 : Plantae, pars 17, Dicotyledones (Ligna). Berlin. HOFMANN, E. 1952. Pflanzeiireste aus dem Phosphoritvorkommen von Prambachkirchen in Ober- osterreich. II. Palaeontographica, 92, B, 122-183. KRAUSEL, R. 1949. Die fossilen Coniferen-Holzer (ausgeschlossen Araucarioxylon). II. Kritische Untersuchungen zur Diagnostik lebender und fossiler Koniferenholzer. Palaeontographica, 89, B, 83-203. MATHiESON, F. j. 1932. Notes on some fossil plants from East Greenland. Medd. om Gronland, 85 (4), 16-18. PAGE, V. 1968. Angiosperm wood from the Upper Cretaceous of Central California. II. Am. J. Bot. 55, 168-172. PRAKASH, u., and barghoorn, e. s. 1961. Miocene fossil woods from the Columbia Basalts of Central Washington. J. Arn. Arbor. 42, 165-195. brezinova, d., and buzek, c. 1971. Fossil woods from the Doupovske hory and Ceske stfedohofl Mountains in Northern Bohemia. Palaeontographica, B, 133, 103-128. SELMEiER, A. 1957. Die Kieselholzer des bayerischen Miozans. Sonderdr. 23. Jahresbericht Naturwiss. Ver. Land. SLUPER, E. J. 1932. liber pliozaner Holzer aus dem Ton von Reuver (Limburg, Holland). Rec. trav. bot. neerl., 29, 18-35. D. W. BRETT Botany Department Bedford College Regent’s Park Typescript received 4 August 1971 London, N.W.l CRETACICRUSTA GEN. NOV., A POSSIBLE ALGA FROM THE ENGLISH CRETACEOUS by GRAHAM F. ELLIOTT Abstract. An encrusting organism which shows zonal cell-differentiation and patterns of cells reminiscent of both Melobesieae and Solenoporaceae (Algae), and cell-sizes more in keeping with certain invertebrate organisms, is described as Cretacicnista duhiosa gen. et sp. nov., and its nature discussed. The organism described below comes from the well-known Lower Cretaceous Sponge- Gravel of Faringdon, Berkshire, southern England. This is a current-bedded accumula- tion of calcareous marine organisms — mostly sponges, brachiopods, and bryozoa with subordinate echinoids and mollusca — in a matrix of iron-stained quartz-sand and pebbles. It is of Upper Aptian age (niitfie/densis zone): a summary account is given by Casey (1961), and the fossil now described came from the Lower or Yellow Gravels of the Little Coxwell Pit, Faringdon. The specimen is a calcareous encrustation on sponge material collected by the late H. D. Thomas. It was recognized by him as possibly algal. PROBLEMATICA (? ALGAE ?RH0D0PHYTA) CRETACICRUSTA gen. nov. Diagnosis. Encrusting calcareous organism showing cell-diflerentiation into basal and subsequent layers reminiscent of the Melobesieae (Algae), with some cell-structures like those of the Solenoporaceae (Algae), and with cell-sizes larger than normal for these algae. Type-species. C. ditbiosa sp. nov.; Lower Cretaceous (Aptian) of England. Cretacicnista diibiosa sp. nov. Plates 100-101 Diagnosis. Characters those of the genus, of which it is the only known species. Description. This organism encrusts a fragment of one of the calcisponges which, whole or broken, make up much of the bulk of the yellow sponge- gravel. The encrustation is 25 mm in diameter and about 3 mm thick, with outer surface smooth to the naked eye and finely perforate under magnification. In thin-section calcisponge tissue forms the core of the section. Resting on the outer surface of this core is a conspicuous basal single layer of cells. In vertical section these are irregularly four-sided with diameters of up to 0-180 mm. Although somewhat irregular in shape and variable in size they form a markedly differentiated layer, with conspicuous dark cell- walls of about 0-010 mm thickness and with clear calcite cell- fillings. This layer underlies almost the whole of the subsequent encrustation, though at one spot it appears to have been destroyed by replacement. [Palaeontology, Vol. 15, Part 3, 1972, pp. 501-503, pis. 100-101.] 502 PALAEONTOLOGY, VOLUME 15 Where the plane of section cuts through this layer at varying angles, due to its close application to an irregular or twisted surface, the cells are seen to be elongate, up to a length of 0-630 mm in one case, though only half this dimension is shown by several other examples. From the limited evidence available it is not clear what is a normal cell-length. The succeeding layer of cells is, for some reason, mostly destroyed and replaced by calcite, and so appears as a conspicuous clear seam. But at one spot it is seen to consist of squarish cells, each matching those of the basal layer in diameter and a little higher (up to 0-225 mm) in vertical section. This layer is succeeded by the main portion of the organism. This consists of adjacent, parallel, slightly irregular, and densely septate tubes, set vertically to the exterior, and each apparently following on a cell of the subjacent layer. Preservation is not as good as for the basal layer (the sponge-gravel matrix is of quartz grains with much irregular calcite cementation and iron-staining). Intercalation and bifurcation are not clearly dis- tinguishable, the tubes increasing only slightly in diameter outwards, perhaps most so for accommodation of growth from where the basal layer is curved. On a better-preserved portion near the periphery the inner tube diameters are 0-090-0-108 mm with wall- thickness very irregular but with a mean of about 0-045 mm. Mid-wall to mid-wall the diameter is about 0-180 mm. The crowded septa are gently curved in section (concave when viewed with the outer surface of the encrustation uppermost), 0-010 mm thick, and occur at intervals of 0-027-0-036 mm. Holotype. The specimen figured in Pis. 100-101, from the Yellow Sponge-Gravel (Lower Cretaceous, Upper Aptian, juitfieldensis zone)', Little Coxwell Pit, Faringdon, Berkshire. Brit. Mus. (Nat. Hist.) Dept. Palaeont., reg. no. V27998. Discussion of affinities. The clear differentiation of this organism into two layers, the basal and adjacent cell-layers together on the one hand, and the main septate-tubular portion on the other, is reminiscent of the hypothallus and perithallus in many members of the Melobesieae or calcareous encrusting red algae. Here the hypothallus of squarish- sectioned, elongate cells, is the initial spreading layer, adapting itself to irregularities in EXPLANATION OF PLATE 100 Figs. 1-4. Cretacicnista ditbiosa gen. et sp. nov., holotype; thin-section. From the Yellow Sponge- Gravel (Lower Cretaceous, Aptian, mitfieUiensis zone); Little Coxwell Pit, Faringdon, Berkshire, England. Brit. Mus. (Nat. Hist), Dept. Palaeont., reg. no. V27998. 1. Calcisponge tissue on right, succeeded by basal structure of Cretacicnista (two cell-layers), then by naain tubular portion of the organism to the left; X 35. 2. Enlarged view of part of fig. 1 , showing clear dark-walled cells of the basal layer in transverse section, and the second cell-layer (to the left) which is re-crystallized; X 1 10. 3. The same view and structures as in fig. 1, from a different part of the thin-section, to show the elongate-irregular form of the basal cells on a twisted sponge-surface; x35. 4. Enlarged view of part of fig. 3; x 1 10. EXPLANATION OF PLATE 101 Figs. 1-2. Cretacicnista diihiosa gen. et sp. nov., holotype; details as PI. 100. 1. Calcisponge tissue bottom left, then dark-walled basal cells of Cretacicnista, succeeded by a portion of the second cell-layer showing individual cell-outlines corresponding to the individual basal cells. At the top right the main septate-tubular part of the organism commences, the tubes corresponding to the underlying cells; x 1 10. 2. Details of well-preserved peripheral portion of tubes, showing close-set concave septa; X 140. Palaeontology, Vol. 15 PLATE 100 ELLIOTT, Cretacicrusta Palaeontology, Vol. 15 PLATE 101 ELLIOTT, Cretacicrusta G. F. ELLIOTT: CRETACICRUSTA 503 the substratum. The succeeding thicker perithallus, of vegetative and reproductive cells, is of different cell-structure. In Cretacicrusta, however, the main layer ( ?perithallus equivalent) is in its septate-tubular structure much more like the Solenoporaceae, whose hypothallus is rudimentary and not well differentiated. That the two layers are parts of the same organism and not two separate successive encrusting growths seems probable from the close correspondence and apparent continuity of the tubes with the subjacent cells. No reproductive structures have been recognized, which is a typical negative character for the Solenoporaceae. A basal layer which is perhaps comparable is figured by Opik and Thomson (1933) for Solenopora cf. nigra Brown, but the cells are much smaller. In dimensions, however, the cells and tubes of Cretacicrusta are much larger than those of most algae. They are about double those of a solenoporoid, and melobesioids are smaller still. Comparison with invertebrate encrusting organisms yields no close parallel. In a British example of a Mesozoic Chaetetid, Blastochaetetes bathoniciis J.-C. Fischer, recently recognized by me from the Great Oolite (M. Jurassic) of Gloucestershire, the tubes are twice the diameter of those in Cretacicrusta, and the septa are mostly straight in section and very irregularly set. Cretacicrusta does not show typical hydrozoan struc- ture (the stromatoporoid Actinostromaria faringdonensis Thomas occurs very rarely at Little Coxwell), nor typical bryozoan structure. In view of these comparisons, Cretacicrusta can only be considered a problematic organism, possibly algal. It is anomalous in combining melobesioid and solenoporoid characters with unusually large cells. I have elsewhere (Elliott 1965) reviewed the relation- ship and possible evolutionary connection between the Solenoporaceae and Melobesieae, the latter having replaced the former during geological time. In the Solenoporaceae occasional rare developments, possibly to be interpreted as reproductive structures like those of the Melobesieae, were unsuccessful in evolution. If Cretacicrusta is a similar solenoporoid achievement of a true hypothallus it seems to have been equally unsuccessful. REFERENCES CASEY, R. 1961. The stratigraphical palaeontology of the Lower Greensand. Palaeoutotogv, 3, 487-621, pis. 77-84. ELLIOTT, G. F. 1965. Tertiary solenoporacean algae and the reproductive structures of the Soleno- poraceae. Palaeontology, 1, 695-702, pis. 104-108. OPIK, A., and THOMSON, p. w. 1933. fiber Konzeptakeln von Solenopora. Tartn Ulik. Geol.-Inst. Toiin. 36, 1-8. G. E. ELLIOTT Dept, of Palaeontology British Museum (Natural History) London, S.W. 7 Final typescript received 4 December 1971 THE BAJOCIAN AMMONITE DORSETENSIA IN SKYE, SCOTLAND by N. MORTON Abstract. The genus Dorseteusia is revised to include Ammonites pingiiis Roemer and related species. Dimorph- ism is possible but separate specific names for macroconchs and microconchs are retained for the present. In the basal part of the Humphriesianum Zone and Subzone D. haimoverana (Hiltermann) (?M), D. hebridica sp. nov. (? M), and D. pingiiis (Roemer) (? m) occur, while stratigraphically higher in the Humphriesianum Subzone are D. liostraca Buckman (? M) and D. romani (Oppel) (? m). In western Scotland Bajocian ammonites occur mainly in the Bearreraig Sandstone of Trotternish, Isle of Skye. Ammonites from the Humphriesianum Zone have been found in the Upper Sandstones (Morton 1965), and the fauna is dominated by species of Stephanoceras and Dorseteusia. The former are discussed by Morton (1971), the latter genus is the subject of this paper. STRATIGRAPHY The genus Dorseteusia, as revised herein, occurs in three localities in Trotternish. From south to north these are: Torvaig (NG 502444) : from the dark sandstone bed at the base of the Upper Sandstones — Dorseteusia piuguis (Roemer), D. hauuoveraua (Hiltermann), and D. hebridica sp. nov. Bearreraig (Rudha Sughar NG 518537) — Dorseteusia rouiaui (Oppel), D. liostraca Buck- man, from loose blocks of the lower part of the Upper Sandstones. Rigg (NG 521566): from lower part of the Upper Sandstones on the shore south of the waterfall — Dorseteusia rouiaui (Oppel), D. liostraca Buckman; from c. 12 m higher in the lower part of the Upper Sandstones, shore north of waterfall — D. liostraca Buckman. The associated ammonite faunas are listed in Morton (1971). Placing of the faunas in stratigraphical sequence must depend on correlation of more or less isolated sections, but those at Bearreraig and Rigg are at approximately the same position in the lower part of the Upper Sandstones. The ammonite bed at Torvaig is taken as the basal bed of the Upper Sandstones, so that the species of Dorseteusia from Torvaig are probably slightly older than those from Bearreraig and Rigg. Similar relationships between piuguis and rouiaui were found in Germany by Wester- mann (1954, pp. 22-23) and Huf (1968, pp. 14-16), and are partly the basis for subzonal division of the Humphriesianum Zone. The 'piuguis Schichten’ (Westermann) or ‘Zone’ (Huf) is equivalent to the Frechi (+Ur>'ibilicum) Subzone, while the 'rouiaui Schichten or Zone’ is equivalent to the Humphriesianum Subzone (s.s.) (Westermann 1967, table 10). A similar subzonal sequence probably exists in Skye, but absence of direct evidence of stratigraphical sequence precludes categorical conhrmation. Recognition of an upper Blagdeni Subzone is, however, already achieved (Morton 1971). [Palaeontology, Vol. 15, Part 3, 1972, pp. 504-518, pis. 102-105.] N. MORTON: THE AMMONITE DORSETENSIA IN SKYE 505 Dimensions. The dimensions given for the specimens are as follows : D. Diameter of specimen. Wh. Whorl height. Wb. Whorl breadth. Ud. Diameter of umbilicus. bih. Whorl breadth divided by whorl height. The letters A and P in front of the diameter indicate that the measurements were taken at the aperture and the end of the phragmocone respectively. In the graphs the latter is indicated by a circle. Abbreviations. Numbers prefixed by HMS are specimens from the Hunterian Museum, University of Glasgow, by Call. J. from the collection of Dr. J. H. Callomon, and by GSE from the Geological Survey, Edinburgh. SYSTEMATIC PALAEONTOLOGY Genus dorsetensia Buckman 1892 Type species. Ammonites edouardiamis d’Orbigny 1 844, original designation by Buckman ( 1 892, p. 302). Discussion. The genus Dorsetensia was created by Buckman for a group of sonniniids from the Humphriesianum Zone. Subsequent experience suggests that the type species edouardiana is less common than the widespread species rornani (including complanata) and liostraca (and closely related species subtecta and tecta) which occur in the lower part of the Humphriesianum Zone, for example in England (Buckman 1892), France (Haug 1893, Brasil 1895, Gillet 1937, Maubeuge 1949, 1951, Roche 1943), Germany (Dorn 1935, Huf 1968), Switzerland (Maubeuge 1955, 1967), Spain (Fallot and Blanchet 1923), Poland (Kopik 1967), U.S.S.R. (Stankewicz 1964). There are many similarities between Dorsetensia and Witchellia, such that some authors (e.g. Haug 1893, Gillet 1937, Krimholz 1958) have regarded Dorsetensia as a junior synonym of Witchellia, while others (e.g. Oechsle 1958) have included both in Sonninia. Other authors, notably Dorn (1935) and Maubeuge (1951), recognized Dorset- ensia and Witchellia as separate genera but confused the species, including the type species, belonging to each. Distinction of Dorsetensia and Witchellia can be justified on morphological and stratigraphical grounds. In Witchellia the radial line is more flexed on the whorl sides and the venter remains distinctly separate and tabulate even in involute smooth forms, while in Dorsetensia such forms have an acutely fastigate venter. The two genera are also stratigraphically separate — Witchellia from Laeviuscula Sub- zone, Sowerbyi Zone, Dorsetensia from Humphriesianum Zone. As previously defined the genus Dorsetensia includes species ranging from evolute ribbed forms with tabulate venter (the type species edouardiana) to involute smooth forms with acute venter (e.g. liostraca). In the lower part of the Humphriesianum Zone there occur also more or less evolute species with tabulate to bisulcate venters and ribbing which branches near the umbilical edge and is very slightly flexed on the whorl sides (e.g. pinguis, hannoverana). These species are variously classed as Sonninia (e.g. Gillet 1937, Hiltermann 1939, Westermann 1954, Oechsle 1958), Witchellia (e.g. Dorn 1935, Maubeuge 1967), Dorsetensia (Jager 1952), Poecilomor pints (Huf 1968), or IPelekodites (Westermann 1967). Their morphological features, particularly style of ornament, and stratigraphical occurrence suggest that they should be included in the genus Dorsetensia (see also discussion of pinguis below). Huf (1968, p. 54) commented on the occurrence of intermediates between hannoverana and Dorsetensia deltafalcata (Quenstedt). 506 PALAEONTOLOGY, VOLUME 15 Dorseteusia Uostraca Buckman Plate 102; Plate 103, figs. 1-2; Plate 104, figs. 1-2 1892 Dorseteusia Uostraca S. Buckman, pp. 301-311, pi. 53, figs. 11-16, pi. 55, figs. 3-5, pi. 56, fig. 1. 1935 Dorseteusia Uostraca S. Buckman, Dorn, pp. 101-102, p). 9, fig. 5, pi. 22, fig. 3, pi. 27, fig. 1, text-fig. pi. 8, figs. 5-8. 1967 Dorseteusia Uostraca Buckman, Kopik, pp. 25-27, pi. 6, fig. 4, pi. 7, figs. 1-4. 1968 Dorseteusia Uostraca Uostraca S. Buckman, Huf, pp. 97-103, pis. 30-40. Material. Ten specimens, some fragmentary; HMS26380/1-2, HMS26381/1-6, HMS26382, GSE7234. Dimeusious: D. Wh. Wb. Ud. HMS26380/1 — 32-6 16-8 HMS26381/1 ? A c. 196 c. 84-8 (43) — 41-5 (21) 168 79-4 (47) 310 (19) 31 -2 (19) P 140 66-8 (48) 28-4 (20) 27-4 (20) HMS26381/2 142 72-5 (51) — 28-2 (20) c. 117 c. 54-8 (47) — 25-3 (22) c. 102 r. 47-5 (46) — 22-4 (22) HMS26381/3 A c. 192 c. 82-0 (43) 27-2 (14) 45-0 (23) 172 76-9 (45) 27-6 (16) 37-0 (22) 154 720 26-7 (17) 29 0 (19) HMS26381/5 c. 144 70-2 (49) 27-2 (19) 28-2 (20) HMS26382 A c. 158 c. 80 0 (51) c. 27-0 (17) 24-4 (15) 128 66-1 (52) 23-3 (18) 21-0 (17) P 960 50-0 (52) 20-8 (22) 14-5 (15) 81-6 42-0 (52) 16-2 (20) 13-7 (16) GSE7234 39-2 18-9 (48) 9-3 (24) 8-2 (21) Description. Involute, compressed, oxyconic; whorl section high and narrow, with rela- tively sharp umbilical angle and umbilical wall becoming vertical; venter acute with prominent keel, hollow on most of body chamber but septicarinate on phragmocone; apart from faint ribbing on the innermost whorls, ornament consists of only growth lines and faint strigation on the body chamber; the radial line is straight or only very slightly flexed on the whorl sides, but is strongly projected ventrally; the body chamber is just over half a whorl in length— 185° in HMS26382, 190° in HMS26381/3, and 210° in HMS26381/1 ; the suture line is comparatively uncomplicated with the lateral lobe open and approximately as broad as long, and three well-developed umbilical lobes. Discussion. The Skye specimens are typical of the Dorseteusia Uostraca group. Buckman ( 1 892) named three species — Uostraca, tecta, and subtecta distinguished mainly by width of umbilicus — but commented on the occurrence of intermediates. It is likely that all three, together with lotharingica Maubeuge and thilense Maubeuge are variants of one species. The relative umbilical diameter of the specimens from Skye is comparable with EXPLANATION OF PLATE 102 Fig. 1. Dorseteusia Uostraca S. Buckman; HMS26381/3, Humphriesianum Zone, lower part of Upper Sandstones, shore just south of Rigg Waterfall, Trotternish, Skye. X 1. Palaeontology, Vol. 15 PLATE 102 MORTON, The ammonite Dorsetensia in Skye 3 -i' '6 r I I i •jAjjrj ■ v> N. MORTON: THE AMMONITE DORSETENSIA IN SKYE 507 that of D. liostraca. Buckman (1892, p. 310) regarded form ^ as type, of which pi. 53, figs. 11-15 and pi. 55, figs. 3-4 are examples. Of these pi. 55, fig. 3 is the only large complete specimen and is refigured by Huf (1968) as lectotype. The whorl height and umbilical diameter of this specimen (SMJ6260) are plotted in text-fig. 1 with data for the Skye specimens. TEXT-FIG. 1. Whorl height and umbilical diameter plotted against diameter for Dorsetensia liostraca S. Buckman from Skye, also for the lectotype (L) from Dorset. D. liostraca is characteristic of the lower part of the Humphriesianum Zone (i.e. the Humphriesianum Subzone) and is recorded by Buckman (1892, p. 310) from England, by Buck, Hahn, and Schadel (1966, pi. 4) and Huf (1968, pp. 102-103) from Germany, by Kopik (1967, p. 11) from Poland, and by Pavia and Sturani (1968, p. 312) from SE. France. However, Dorn (1935, p. 120) shows the species as also ranging down into the Sauzei Subzone. Locality. Humphriesianum Zone, lower part of Upper Sandstones; shore just south and just north of Rigg waterfall. Specimens were also observed at Bearreraig. l1 C9016 508 PALAEONTOLOGY, VOLUME 15 Dorsetensia romani (Oppel) Plate 103, figs. 3-8; Plate 104, figs. 3-6 1862 Ammonites Romani, Oppel, p. 145, pi. 46, figs. 2a-b. 1892 Dorsetensia complanata, S. Buckman, pp. 306-307, pi. 53, figs. 1-10, pi. 54, figs. 1-2. 1893 Witchellia complanata (Buckm.), Haug, pp. 312-315, pi. 10, figs. 4-5. 1935 Dorsetensia complanata Buckman, Dorn, p. 98, pi. 9, fig. 4, pi. 10, fig. 5, text-fig. pi. 8, fig. 1. 1935 Dorsetensia romani OppQ\,T>oxn,px>. 100-101, pi. 9, fig. 5, pi. 1 1, fig. 4, pi. 13, fig. 2, text-fig. pi. 8, fig. 4. 1937 Witchellia complanata Buckman, Gillet, pp. 66-67, text-figs. 47-48. 1951 Dorsetensia complanata Buckman, Maubeuge, pp. 40-45, pi. 1, fig. 2a-c, pi. 14, fig. 3. 1967 Dorsetensia complanata Buckm., Kopik, pp. 22-24, pi. 4, fig. 4, pi. 5, figs. 1-2, pi. 6, fig. 1. 1967 Dorsetensia aff". complanata Buckm., Kopik, pp. 24-25, pi. 6, figs. 2-3. 1968 Dorsetensia rornani romani (Oppel), Huf, pp. 86-93, pi. 13, fig. 6, pis. 14-27, pi. 28, figs. 1-2. Material. Thirteen specimens and three fragments: Call. J. 461, Call. J. 462, Call. J. 463, Call. J. 464, Call. J. 465, HMS26383/1-3, HMS26384/1-2, HMS26385, HMS26386/1-2, HMS26387, HMS26388/ Dimensions: D Wh Wb Ud Call. J. 461 61-0 25-3 (42) 14-2 (23) 18-2 (30) 5M 22 0 (43) 12-8 (25) 16-4 (32) Call. J. 462 P 78-5 34-2 (44) 16-1 (21) 21-1 (27) Call. J. 463 c. 71-0 c. 30 0 (42) 14-4 (20) 18-1 (26) P 490 21-5 (44) 9-7 (20) c. 12-0 (25) Call. J. 464 64-0 28-1 (44) c. 13-4 (21) 16-9 (26) P £•. 510 c. 22-5 (44) — 13-1 (26) 43-3 20 0 (46) — 12-3 (28) Call. J. 465 72-3 35-1 (49) 12-1 (17) 15-9 (22) 54-0 26-3 (49) 11-5 (21) c. 12-0 (22) HMS26383/1 32-6 14-6 (45) — 8-6 (26) HMS26383/2 c. 53-0 f. 22-0 (42) — c. 15-5 (29) HMS26383/3 39-6 17-8 (45) 7-8 (20) c. 10-0 (25) HMS26384/1 c. 78-0 c. 310 (40) c. 12-5 (16) c. 24-0 (31) 640 27-0 (42) 12-1 (19) 19-8 (30) HMS26384/2 50-7 25-7 (51) — • 9-3 (19) 42-4 21-4 (51) — 8-4 (20) HMS26385 — 10-1 7-1 — HMS26386/1 48-3 21-1 (44) 8-7 (18) 13-8 (29) 41 0 17-0 (42) 7-8 (19) 13-0 (32) P 36-4 14-2 (39) 7-4 (20) 10-5 (29) HMS26386/2 47-2 16-6 (35) — 16-5 (35) 40-4 15-5 (38) — 15-0 (37) P 32-3 13-7 (42) — • c. 11-5 (36) HMS26387 — 25-3 9-5 — HMS26388/1 — 36-2 15-5 — HMS26388/2 c. 71-5 c. 30-0 (42) — c. 22-3 (31) c. 56 0 c. 22-3 (40) 9-7 (17) 14-9 (27) P 521 21-5 (41) — 13-9 (27) All figures natural size. EXPLANATION OF PLATE 103 Figs. 1, 2. Dorsetensia liostraca, S. Buckman. 1, HMS26381/3, ventral view of Plate 102 fig. 1; 2, HMS2638I/I, ventral view. Figs. 3-8. Dorsetensia romani (Oppel). Loose blocks from lower part of Upper Sandstones, Rudha Sughar, Bearrcraig, Trotternish, Skye. 3, 4, Call. J. 461 ; 5, 6, HMS26383/3; 7, 8, Call. J. 462. Palaeontology, Vol. 15 PLATE 103 MORTON, The ammonite Dorsetensia in Skye N. MORTON: THE AMMONITE DORSETENSIA IN SKYE 509 Description. Moderately involute, compressed, becoming suboxyconic in some; whorl section narrow, acute, with venter scarcely differentiated from whorl sides; umbilical angle sharp, with umbilical wall steep or vertical; venter acutely fastigate with promi- nent hollow keel; ornament of distant broad ribs, mostly simple, and developed mainly on the lower part of the whorl sides, fading towards the venter and also on the body chamber towards the aperture; the radial line is very slightly flexed on the whorl sides r40 -30 E E 40 - - 20 j 0 1 30--10 e20--0 10 04- 20 30 40 50 60 70 Diometer (mm) 80 90 WO 110 TEXT-FIG. 2. Whorl height and umbilical diameter plotted against diameter for Dorsetensia romani (Oppel) from Skye, and for the holotype of D. romani (Oppel) (H) and the lectotype of D. complanata Buckman (L). Data from Huf (1968), for umbilical diameter of D. romani (Oppel) from Gerzen, Germany, are also shown. but is strongly projected ventrally; most of the specimens are incomplete, but the length of body chamber is 190° in Call. J. 463, 200° in HMS26388/2, 210° in Call. J. 465, and 220° in HMS26386/2; the suture line is moderately complex, but with the lateral lobe broad-stemmed. Discussion. The whorl shape, involution, and ornamentation of the specimens is typical of Dorsetensia romani (Oppel), abundantly figured by Huf (1968). Two specific names have been widely used in the past (see synonymy and Huf 1968, pp. 86-88): romani Oppel 1862 and complanata Buckman 1892. The holotype of romani (figd. Huf 1968, pi. 13, fig. 6a-e) and the lectotype of complanata (Buckman 1892, pi. 53, figs. 3-5, and Huf 1968, pi. 15, fig. \a-d) are very similar, differing only in that complanata is slightly more evolute (Ud. = 32-38% D. cf. 30% in romani) and has thinner whorls (Wb. = 510 PALAEONTOLOGY, VOLUME 15 17% D. cf. 19-21 % in romani). However, both fall within the ranges of variation of the Skye specimens and of specimens from Gerzen, Germany, described by Huf (1968) (text-fig. 2), so that complamta Buckman must be regarded as a junior synonym of romani (Oppel). This confirms Huf’s conclusion. Comparing the Skye specimens with those from Gerzen, there is considerable overlap in the range of variation, but the Skye specimens tend to be more involute (text-fig. 2), grouping below the line Ud. = 30% D. rather than over this line as in those from Gerzen. The occurrence of D. romani is closely paralleled by that of D. liostraca, both in Skye and in other areas (see synonymy lists). It is possible that D. romani may be the micro- conch of D. liostraca, although no specimens with lappets are recorded. Dorsetensia romani (inch complanata) is widely recorded from the lower part of the Humphriesianum Zone, and is used as subzonal index by some authors (see synonymy list and Pavia and Sturani 1968, p. 312, Westermann 1967, pp. 133-134). Localities. Humphriesianum Zone; lower part of Upper Sandstones. (u) Call. J. 461-465, HMS26383/1-3 from loose blocks of the Upper Sandstones, Rudha Sughar, Bearreraig. ib) HMS26384/1-2, HMS26385, HMS26386/1-2, HMS26387, HMS26388/1-2 from lower part of Upper Sandstones, shore south of Rigg waterfall. Dorsetensia pinguis (Roemer) Plate 105, figs. 1-12, 19-20 1836 Ammonites pinguis, Roemer, p. 186, pi. 12, fig. 3. 1939 Sonninia pinguis (Roemer), Hiltermann, pp. 164-167. 1968 Sonninia {PoecHomorphus) pinguis pinguis (Roemer), Huf, pp. 54-61, pi. 4, figs. 7-12, pi. 5, figs. 1-8. Material. Fourteen specimens and one fragment; HMS15347/1, 2, HMS15349/1, 3, HMS26389/1-1 1, Dimensions: D Wh Wb Ud bib HMS15347/1 27-5 11-5 (42) 8-8 (32) 8-8 (32) 0-77 191 7-4 (39) 7-3 (38) 6-3 (33) 0-99 HMS15347/2 c. 21-2 9 0 (43) 8-4 (40) 7-8 (37) 0-93 13-5 5 0 (37) c. 5 0 (37) 4-7 (35) 100 HMS15349/1 A c. 30-0 12-4 (41) 12-4 (41) c. 9-6 (32) 0-77 24-8 10-6 (43) c. 8-9 (36) 7-2 (29) 0-84 HMS26389/1 A 39-6 14-9 (38) 11-3 (29) 13-5 (34) 0-76 33-2 12-6 (38) 101 (30) 12-2 (37) 0-80 30-7 11-2 (37) 8-7 (28) 11-3 (37) 0-78 P 24-5 9-4 (38) 8-3 (34) 8-4 (34) 0-88 HMS26389/2 A 28-8 10-8 (38) 9-5 (33) 10-4 (36) 0-88 22-1 8-7 (39) 7-8 (35) 7-3 (33) 0-90 P c. 190 8 0 (42) 7-0 (37) 4-2 (33) 0-87 All figures natural size. EXPLANATION OF PLATE 104 Figs. 1, 2. Dorsetensia liostraca S. Buckman; HMS26382, Humphriesianum Zone, lower part of Upper Sandstones, shore just north of Rigg waterfall, Trotternish, Skye. Figs. 3-6. Dorsetensia romani (Oppel). 3, 4, HMS26387, Humphriesianum Zone, lower part of Upper Sandstones, shore just south of Rigg waterfall, Trotternish, Skye. 5, 6, Call. J. 464, Humphriesianum Zone, loose block from lower part of Upper Sandstones, Rudha Sughar, Bearreraig, Trotternish, Skye. Palaeontology, Vol. 15 PLATE 104 MORTON, The ammonite Dorsetensia in Skye 511 N. MORTON: THE AMMONITE DORSETENSIA IN SKYE Dimensions : D Wh Wb Ud bjh HMS26389/3 ? A 28-6 11-6 (41) 9-6 (34) 8-8 (31) 0-83 21-2 9 0 (42) 7-9 (37) 6-5 (31) 0-88 P c. 18-5 8-0 (43) 6-9 (37) 61 (33) 0-86 HMS26389/4 A 31-5 13-2 (42) 10 0 (32) 10 0 (32) 0-76 24-1 9-6 (40) 81 (34) 8 0 (33) 0-84 P c. 19-5 8-6 (44) 7 0 (36) 6 0 (31) 0-81 HMS26389/5 20-8 8-7 (42) 7-5 (36) 6-2 (30) 0-86 15-3 6-2 (41) 61 (40) 4-6 (30) 0-99 HMS26389/6 ? A 26-2 12-6 (48) 8-6 (33) 6-6 (25) 0-68 19-7 9-4 (48) 7-2 (37) 5-5 (28) 0-77 P 13-8 5-9 (43) 6 0 (43) 4-2 (30) 101 HMS26389/7 28-1 f. 13 0 (46) c. 9 0 (32) c. 7 0 (25) 0-69 21-5 9-1 (42) 71 (33) 6-3 (29) 0-78 ?P 14-9 6-6 (44) 6 0 (40) 4 0 (27) 0-91 12-2 5-3 (43) 5 0 (41) 3-4 (28) 0-94 HMS26389/8 A 35-3 16 0 (45) r. n o (31) 9-2 (26) 0-69 28-2 12-8 (45) 101 (36) 8-1 (29) 0-79 P 211 9-5 (45) 8-2 (39) 5-9 (28) 0-86 180 7-5 (42) 7-3 (41) 5-1 (28) 0-97 HMS26389/9 A 22-8 9 0 (40) 7-3 (32) 7-4 (33) 0-81 170 7-3 (43) 6-4 (38) 5-3 (31) 0-88 P c. 13-5 5-5 (41) 5-4 (40) 4 0 (30) 0-98 HMS26389/11 A c. 21-5 10 0 (47) 81 (38) 6-2 (29) 0-81 16-1 7-1 (44) 6-2 (39) 5-0 (31) 0-87 10-3 5 0 (49) 5-0 (49) 3-1 (30) 100 Description. Moderately evolute, planulate; whorl section subquadrangular, with venter fairly well differentiated from whorl sides; umbilical angle rounded, with umbilical wall becoming steep but rarely vertical; venter broad, tabulate with prominent keel and becoming bisulcate in some; ornament of broad distant ribs, mostly bifurcating at the umbilical angle and fading on the top of the whorl sides before reaching the venter; the strength of the ribbing varies considerably; the radial line is slightly flexed on the whorl sides, but is strongly projected ventrally; the peristome is at least partially preserved on five specimens, and the lengths of the body chambers vary: 205° in HMS26389/2, 215° in HMS26389/8, 225° in HMS26389/9, 235° in HMS26389/4, and 245° in HMS26389/1; none shows lappets, and crowding of sutures is evident only in HMS26389/2; the suture line is moderately simple, with the lateral lobe fairly broad-stemmed. Discussion. There is considerable variation in relative dimensions (text-fig. 3) and orna- mentation. In style of ornament, whorl section, and venter, and also in the range of variation, the Skye specimens are very similar to those figured by Huf (1968) as Sonninia (Poecilomorphus) pinguis pinguis (Roemer). This species is herein transferred from Poeci- lomorphus mainly because of differences in the style of the oxmmQnV. Poecilomorphus has strongly falcate ribbing which is typically stronger on the outer part of the whorl sides, while pinguis has almost straight ribbing which fades on the ventral shoulder. It is placed in the genus Dorsetensia because in the almost straight ribbing and tabulate keeled venter it is similar to the type species D. edouardiana (d’Orbigny), although these features are more strongly developed in pinguis. Comparisons with D. deltafcdcata (Quenstedt) were made by Huf (1968, pp. 53-54) and in his detailed synonymy list (pp. 54-55) pinguis is twice recorded as a Dorsetensia. 512 PALAEONTOLOGY, VOLUME 15 TEXT-FIG. 3. Umbilical diameter, whorl height, and whorl breadth; height ratio plotted against diameter for Dorseteusia pinguis (Roemer) from Skye, and for the holotype from Germany (H). N. MORTON: THE AMMONITE DORSLTENSIA IN SKYE 513 The Skye specimens of D. pinguis are all of relatively small size, even when complete. Most have the body chamber preserved, but in the absence of lappets or crowding of sutures (except in HMS26389/2) evidence of maturity is lacking and the specimens may be juveniles. The specimens figured by Huf (1968, pis. 4-5) are of similar small size, but many of these (including the holotype) lack body chamber. ^Poecilomorphus' pinguis is used as index species for a Pinguis Subzone of the Hum- phriesianum Zone by Buck, Hahn, and Schadel (1966, pi. 4), and Huf (1968, pp. 12-16). Westermann (1967, p. 123) suggested a new name (Frechi Subzone) for this subzone. Pavia and Sturani (1968, p. 311) record the species from a lower horizon — Sauzei Subzone. Locality. Humphriesianum Zone; basal bed of Upper Sandstones; foot of cliff overlooking shore east of Torvaig, near Portree. Dorsetensia haimoverana (Hiltermann) Plate 105, figs. 15-18, 23-24 1939 Sonninia pinguis hannoverana, Hiltermann, p. 167, pi. 11, figs. 8-9. 1968 Sonninia (Poeciloinorplnts) pinguis hannoverana (Hiltermann), Huf, pp. 64-69, pi. 6, figs. 5-12, pi. 7, figs. 1-3, pi. 10, fig. 1. Material. Four specimens and three fragments: HMS15348/1-3, HMS26390/1-4. Dimensions: D Wh Wb Ucl blh HMS15348/1 P 41-4 16-4(40) 15-8 (38) 14-8 (36) 0-96 30-8 11-2 (36) 12-0 (39) 12-2 (40) 1-07 HMS26390/1 39-0 17-4 (45) 14-1 (36) 11-9 (31) 0-81 311 13-3 (43) 12-0 (39) 10-4 (33) 0-90 HMS26390/2 39-7 16-9 (43) 12-4 (31) 12-1 (31) 0-73 30-6 12-2 (40) 101 (33) 9-8 (32) 0-83 HMS26390/3 33-2 13-4 (40) 10-0 (30) 11-3 (34) 0-75 25-3 10-2 (40) 8-3 (33) 8-2 (32) 0-81 Description. Moderately evolute; whorl section subquadrangular with venter well differentiated from whorl sides; umbilical angle rounded, sometimes sharply so, with umbilical wall becoming steep and vertical in some; venter broad, tabulate with promi- nent keel and sometimes bisulcate; ornament of strong broad distant ribs, mostly bifurcating from the umbilical angle and fading on the ventral shoulder; the radial line is slightly ffexed laterally and strongly projected ventrally; the specimens are incomplete, lacking body chamber. Discussion. The specimens grouped in this species differ from the other specimens from the same locality described under D. pinguis (Roemer) in having thicker whorls and stronger ribbing. They are also larger, especially when allowance is made for lack of body chamber. There is very good agreement in these features with the species described by Hiltermann (1939) and Huf (1968) as Sonninia (Poeciloniorphus) pinguis hannoverana (see text-fig. 4). Huf’s measured specimens (p. 67) are smaller than the Skye specimens, but those figured on pi. 6, figs. 11 and 12 are comparable. This species is transferred from Poecilomorphus to Dorsetensia for the same reasons as D. pinguis discussed above. According to Huf (1968, p. 69) D. hannoverana comes from the ’’pingitis-Zont', i.e. the lowermost part of the Humphriesianum Zone. 514 PALAEONTOLOGY, VOLUME 15 Text-fig. 4. Umbilical diameter, whorl height, and whorl breadth : height ratio plotted against diameter for Dorsetensia liannoveraiia (Hiltermann) from Skye, and for the lectotype from Germany (L). Diameter of umbilicus(mm) N. MORTON: THE AMMONITE DORSETENSIA IN SKYE 515 Q4t i 0 6- “O o 4> 0 7- O 08- 09- -30 e o 20--10 10- -0 0-- 10 T" 20 R JJ-— - — Holotype of D . romanl » ■' X D.hebridico ® End of phragmocone “I 1 1 1 1 1 30 40 50 60 70 80 Diameter (m m) TEXT-FIG. 5. Umbilical diameter, whorl height, and whorl breadth : height ratio plotted against diameter for Dorsetensia hebridica sp. nov. from Skye, and for the lectotype of D. deltafalcata (Quenstedt) {D) and the holotype of D. romani (Oppel) (R). 516 PALAEONTOLOGY, VOLUME 15 Locality. Humphriesianum Zone; basal bed of Upper Sandstones; foot of cliff overlooking shore east of Torvaig, near Portree. Dorsetensia hebridica sp. nov. Plate 105, figs. 13-14, 21-22, 25-26 Material. Holotype HMS26391 and four syntypes: HMS26392, HMS26393/1-3. Dimensions: D. Wh Wb Ud bjh HMS26391 ? A 49-8 22-8 (46) 14-7 (30) 13-3 (27) 0-64 41-6 19-9 (48) 12-8 (31) 11-2 (27) 0-64 P c. 29 0 12-8 (44) 10-9 (38) 8-9 (31) 0-85 HMS26393/1 c. 74 0 c. 33-0 (45) 21-0 (28) 18-5 (25) 0-64 53-2 24-9 (47) 17-0 (32) 14 0 (26) 0-68 ?P 43-7 20-1 (46) 15-6 (36) 12-8 (29) 0-78 HMS26393/2 P 68-9 29-9 (43) 18-3 (27) 18-2 (27) 0-61 56-8 22-8 (40) 16-5 (29) 16-4 (29) 0-72 48-8 21 -1 (43) 13-9 (29) 141 (29) 0-66 37-5 16-4 (44) 121 (32) 10-6 (28) 0-74 HMS26393/3 ? A c. 51 0 24-5 (48) 11-2 (22)* 11-9 (23) 0-46* 38-5 17-9 (47) 9-7 (25)* 10 0 (26) 0-54* ?P c. 30 0 13-8 (46) 10-0 (33) 8-4 (26) 0-72 * affected by partial crushing Description. Moderately involute; whorl section high subquadrangular, with venter fairly well differentiated from whole sides; umbilical angle becoming sharp, with umbili- cal wall becoming vertical; venter fairly broad, tabulate or subtabulate, sometimes bisulcate, with prominent keel; ornament of strong distant ribs on the inner whorls but fading on the outer whorl so that the larger specimens are almost smooth; on the body chamber the ribbing is more strongly developed on the middle of the whorl sides; the radial line is very slightly flexed on the whorl sides but is strongly projected ventrally; none of the specimens shows the peristome clearly, but HMS26391 and HMS26393/3 appear to be complete, with 230° and 215° of body chamber respectively; the suture line is moderately complex, with large broad-stemmed lateral lobe. Discussion. The specimens grouped here are larger and slightly more involute than the other species from Torvaig — D. pinguis and D. hannoverana. They differ also in that all show decline of ribbing on the body chamber, while only one or two of D. pinguis show this. Decline of ornamentation is also evident in D. deltafalcata (Quenstedt) and D. roman i (Oppel). D. hebridica is more involute than D. deltafalcata, with relative umbilical . ^ . EXPLANATION OF PLATE f05 All figures natural size. All specimens from Humphriesianum Zone, basal bed of Upper Sandstones, Torvaig, Trotternish, Skye. All HMS collection. Figs. 1-10, 17-20. Dorsetensia pinguis (Roemer); 1, 2, 26389/4; 3, 4, 26389/3; 5, 6, 26389/2; 7, 8, 15347/1 ; 9, 10, 26389/8; 17, 18, 26389/1; 19, 20, 15349/1. Figs. 11, 12, 15, 16, 23, 24. Dorsetensia hannoverana (Hiltermann); 11, 12, 26390/3; 15, 16, 15348/1; 23, 24, 26390/2. Figs. 13, 14, 21, 22, 25, 26. Dorsetensia hebridica sp. nov.; 13, 14, 26393/1; 21, 22, 26393/3; 25, 26, holotype, 26391. Palaeontology, Vol. 15 PLATE 105 MORTON, The ammonite Dorsetensia in Skye N. MORTON: THE AMMONITE DORSETENSIA IN SKYE 517 diameter 25-30% D compared with generally 35-40%, and correspondingly higher relative whorl height (40-48 % cf. 35-40%). The relative whorl breadth (and b:h ratio) is, however, comparable. The relative umbilical diameter and whorl height of D. hebridica are generally comparable with D. rotnani, but the relative whorl breadth (and b:h ratio), and breadth of venter, are much greater in hebridica than in romani. The appropriate dimensions of D. hebridica are summarized graphically in text-fig. 5, with data for the lectotype of D. deltafalcala and the holotype of D. romani (from Huf 1968, pp. 82, 90) shown for comparison. It is not certain whether the specimens are adults or not. Certainly they are larger than the specimens of D. pinguis from the same bed and show decline of the ornamentation on the body chamber which is common in adult macroconchs. Locality. Humphriesianum Zone; basal bed of the Upper Sandstones; foot of cliff overlooking shore east of Torvaig, near Portree. DISCUSSION The specimens of Dorsetensia from Skye may be divided into two distinct groups in terms of size. At Rigg and Bearreraig specimens of liostraca are much larger than those of romani, while at Torvaig complete specimens of hannoverana and hebridica would be larger than those of pinguis. This is also evident from the literature on other areas, for example, Buckman 0892), Dorn (1935), Kopik (1967), and Huf (1968). Some of the small specimens figured show lappets (Buckman 1892, pi. 52, figs. 15-17, and Dorn 1935, pi. 3, fig. 2 and pi. 9, fig. 5) and presumably Dorsetensia is dimorphic. However, micro- conch specimens with lappets are exceptional — none has been found in Skye and two specimens figured by Buckman (pi. 53, figs. 1-2, SMJ6250, and pi. 53, figs. 3-4, SMJ6251) are adults showing crowding of the last two or three sutures but lacking lappets — so that it seems possible that lappets were not developed on all Dorsetensia microconchs. The macroconch: microconch pairings suggested for Skye Dorsetensia are D. liostraca (M) : D. romani (m) in beds stratigraphically higher than those with D. hannoverana (M) and D. hebridica (M):Z). pinguis (m). In the latter case the microconchs of hannoverana and hebridica are presumably less variable and have been grouped in pinguis. A conser- vative taxonomic approach is used herein, because revision of species of Dorsetensia into bisexual taxa should be based on a larger and wider sample than that from Skye alone. Acknowledgements. The work was commenced as part of a Ph.D. thesis in the University of Glasgow supervised by Professor T. N. George. Further fieldwork was carried out with the assistance of a grant from the Central Research Fund of the University of London. For access to collections or loan of specimens I am grateful to Dr. W. D. I. Rolfe (Hunterian Museum, Glasgow), Dr. J. H. Callomon (University College, London), Mr. R. B. Wilson (Geological Survey, Edinburgh), Dr. C. L. Forbes and Dr. R. B. Rickards (Sedgwick Museum, Cambridge), Dr. F. Westphal and Dr. J. Wendt (Tubingen, Germany). Dr. J. D. Hudson (Leicester) and Dr. G. E. G. Westermann (Hamilton, Canada) presented specimens from Torvaig. I thank Dr. M. K. Howarth, Dr. J. H. Callomon, and Dr. H. S. Torrens for their comments on the manuscript. The photographs are by Mr. M. Hobbs and Mr. E. L. Cory. REEERENCES BRASIL, L. 1895. Cephalopodes nouveaux ou peu connus des etages jurassiques de Normandie. Bull. Soc. geol. Normandie, 16, 27-46, pis. 1-4. 518 PALAEONTOLOGY, VOLUME 15 BUCK, E., HAHN, w., and SCHADEL, K. 1966. Zur Stratigraphic des Bajociums und Bathoniums (Dogger (S-e) der schwabischen Alb. Jh. gcol. Landesamt Baden-Wurttemberg, 9, 23-46. BUCKMAN, s. s. 1892 (1887-1907). A monograph of the ammonites of the Inferior Oolite Series. Palaeontogr. Soc. [Mouogr.]. DORN, p. 1935. Die Hammatoceraten, Sonninien, Ludwigien, Dorsetensien und Witchellien des siiddeutschen insbesondere frankischen Doggers. Palaeontographica, 82A, 1-124, pis. 1-29. FALLOT, p., and BLANCHET, F. 1923. Observations sur la faune des terrains jurassiques de la region de Cardo et de Tortosa (Province de Tarragone). Treb. Inst, catal. Hist. nat. 1921-1922, fasc. II, 73-264, pis. 1-13. GILLET, s. 1937. Les ammonites du Bajocien d’Alsace et de Lorraine. Mem. Serv. Carte geol. Als.-Lorr. 5, 1-130, pis. 1-5. HAUG, E. 1893. Etude sur les ammonites des etages moyens du systeme jurassique. Bid/. Soc. geol.fr., ser. 3, 20, 277-333, pis. 8-10. HiLTERMANN, H. 1939. Stratigraphic und Palaeontologie der Sonninienschichten von Osnabriick und Bielefeld. I Teil. Stratigraphic und Ammonitenfauna. Palaeontographica, 90A, 109-209, pis. 9-13. HUF, w. 1968. t)ber Sonninien und Dorsetensien aus dem Bajocium von Nordwestdeutschland. Beih. geol. Jb. 64, 1-126, pis. 1-51. jAGER,w. 1952. Der geologischeWer degang des Salzstoches Wietze-Hambiihren. Be/A. geo/. 6, 1-104. KOPiK, J. 1967. Bajocian ammonites from the Koscielisko Beds in the vicinity of Przystajri (Cracow- Wieluri Jura). Biul. Inst. Geol. 209, 5-50, pis. 1-12. KRIMHOLZ, G. JA. 1958. In Luppov, N. p. (cd.), 1958, Osnovi Paleontologii. Molluska-Cephalopoda 2, Ammonitina. Akad. nauk. SSSR, Moscow [in Russian]. MAUBEUGE, p. L. 1949. Notcs palcontologiques sur quelques ammonites jurassiques rares ou nouvelles de la region frontiere franco-luxembourgeoise et de la Lorraine centrale. Arch. Inst, grand-ducal Lux., Sect. Sci. nat., phys. math., n. ser., 18, 149-78, pis. 1-17. 1951. Les ammonites du bajocien de la region frontiere franco-beige. Mem. Inst. r. sci. nat. Belg., ser. 2, 42, 1-104, pis. 1-16. 1955. Les ammonites aaleniennes, bajociennes et bathoniennes du Jura suisse septentrionale. Mem. suisses paleont. 71, 1^8, pis. 1-11. 1967. Catalogue des ammonites du jurassique inferieur et moyen (Hettangien a Bathonien) du Musee cantonal de Bale-Campagne. Deuxieme partie. Tdtigkeitsbericht der Naturforschenden Gesell- schaft Baselland, 25, 43-130. MORTON, N. 1965. The Bearreraig Sandstone Series (Middle Jurassic) of Skye and Raasay. Scott. J. Geol. 1, 189-216. 1971. Some Bajocian ammonites from western Scotland. Palaeontology, 14, 266-93, pis. 40-51. OECHSLE, E. 1958. Stratigraphic und Ammonitenfauna der Sonninien-Schichten des Filsgebiets. Palaeontographica, lllA, 47-129, pis. 10-20. OPPEL, A. 1862 (1862-1863.) Palaeontologische Mittheilungen. III. liber jurassische Cephalopoden. Mitt. Mus. K. bayer. Staates, 3, 127-266, pis. 40-74. PAVIA, G., and STURANi, c. 1968. Etude biostratigraphique du bajocien des terrains subalpines aux environs de Digne (Basses-Alpes) (note preliminaire). Boll. Soc. geol. Ital. 87, 306-16. ROCHE, p. 1943. Sur les couches dites a Ammonites Blagdeni du Mont d’Or lyonnais. Trav. Lab. geol. Univ. Lyon, mhn. 30, 1-36, pis. 1-2. ROEMER, F. A. 1836 (1836-1839.) Die Versteinerungen des norddeutschen Oolithen-Gebirges. 277 pp., 21 pis. Hannover (Hahn). STANKEViCH, E. T. 1964. Aminonity yiirskikh peschano-glinistykh otlozhenii severo-zapadnogo Kavkaza. Akad. nauk SSSR, Moskva-Leningrad. WESTERMANN, G. E. G. 1954. Monographic der Otoitidae (Ammonoidea). Beih. geol. Jb., 15, 1-364, pis. 1-33. ■ 1967. Lexique stratigraphique international, vol. 1, Europe, fasc. 5/2, Allemagne, Jurassique moyen (Alpes exclues). Paris. N. MORTON Department of Geology, Birkbeck College University of London 7-15 Gresse Street, London WIP IPA Final typescript received 19 January 1972 PALAEONTOLOGY The journal Palaeontology is devoted to the publication of papers on all aspects of palaeon- tology. 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 at least are published each year and are sent free to all members of the Association. Members who join for 1972 will receive Volume 15, parts 1 to 4. All back numbers are still in print and may be ordered from B. H. Blackwell, Broad Street, Oxford, England, at £5 per part (post free). A complete set. Volumes 1-14, consists of 55 parts and costs £275. SPECIAL PAPERS IN PALAEONTOLOGY This is a series of substantial separate works published by the Association. The subscription rate is £8 (U.S. $22.00) for Institute Members and £4(U.S. $11.00) for Ordinary and Student Members. Subscriptions and orders by members of the Association should be placed through the Membership Treasurer, Dr. A. J. Lloyd, Department of Geology, University College, Gower Street, London, WCIE 6BT, England. The following Special Papers are available. Members may obtain them at reduced rates through the Membership Treasurer. Non-members may obtain them from B. H. Blackwell, Broad Street, Oxford, England, at the prices indicated. 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. $14.00). 3. (for 1968): Upper Maestrichtian Radiolaria of California, by Helen p. foreman. 82 pp., 8 plates. Price £3 (U.S. $9.00). 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. $9.00). 5. (for 1969): Chitinozoa from the Ordovician Viola and Femvale Limestones of the Arbuckle Mountains, Oklahoma, by w. A. m. jenkins. 44 pp., 10 text-figs., 9 plates. Price £2 (U.S. $5.50). 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. $9.00). 7. (for 1970) Shell Structure of the Craniacea and other Calcareous Inarticulate Brachio- poda, by a. williams and a. d. wright. 51 pp., 17 text-figs., 15 plates. Price £1.50 (U.S. $4.50). 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 Palaeoenvironment, by m.waldman. 130 pp., 37 text-figs., 18 plates. Price £5 (U.S. $14.00). 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. $14.00). 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, 10, pp. 707-12). PALAEONTOLOGY VOLUME 15 • PART 3 CONTENTS Phosphatized ostracods with appendages from the Lower Cretaceous of Brazil. By RAYMOND HOLMES BATE 379 A Texanites-Spinaptychus association from the Upper Cretaceous of Zululand. By W. J. KENNEDY and H. C. KLINGER 394 The afl&nities of Idiohamites ellipticoides Spath (Cretaceous Ammonoidea). By W. J. KENNEDY 400 The systematic position of the Jurassic brachiopod Cadomella. By c. h. c. BRUNTON and D. L. MACKINNON 405 Observations on some Lower Palaeozoic tremanotiform Bellerophontacea (Gastropoda) from North America. By John s. peel 412 Marcouia gen. nov., a problematical plant from the late Triassic of the south- western U.S.A. By SIDNEY R. ASH 423 The Upper Devonian Saltern Cove goniatite bed is an intraformational slump. By PETER VAN STRAATEN and MAURICE E. TUCKER 430 Ptychodus predation upon a Cretaceous Inoceramus. By e. g. bcauffman 439 Ammonites from the transgressive Cretaceous on the Rhenish Massif, Germany. By J. M. HANCOCK, W. J. KENNEDY, and H. KLAUMANN 445 The development of the loop in the Jurassic brachiopod Zeilleria leckenbyi. By P. G. BAKER 450 The brachiopod Acanthocrania in the Ordovician of Wales. By A. D. wright 473 Crystal development in Discoasteraceae and Braarudosphaeraceae (planktonic algae). By Maurice black 476 A new species of Protocetus (Cetacea) from the Middle Eocene of Kutch, western India. By a. sahni and v. p. mishra 490 Fossil wood of Platanus from the British Eocene. By donald w. brett 496 Cretacicrusta gen. nov., a possible alga from the English Cretaceous. By GRAHAM F. ELLIOTT 501 The Bajocian ammonite Dorsetensia in Skye, Scotland. By n. morton 504 PRINTED IN GREAT BRITAIN AT THE UNIVERSITY PRESS, OXFORD BY VIVIAN RIDLER, PRINTER TO THE UNIVERSITY VOLUME 15 • PART 4 Palaeontology DECEMBER 1972 PUBLISHED BY THE PALAEONTOLOGICAL ASSOCIATION LONDON Price £5 THE PALAEONTOLOGICAL ASSOCIATION The Association was founded in 1957 to further the study of palaeontology. It holds meetings and demonstrations, and publishes the quarterly journal Palaeontology and Special Papers in Palaeontology. Membership is open to individuals, institutions, libraries, etc., on payment of the appropriate annual subscription: Institute membership ..... £10. Os. (U.S. $26.00) Ordinary membership . . . £5. Oj. (U.S. $13.00) Student membership . . . £3. O5. (U.S. $8.00) There is no admission fee. Institute membership is only available by direct appli- cation, not through agents. Student members are persons receiving full-time instruc- tion at educational institutions recognized by the Council; on first applying for membership, they should obtain an application form from the Membership Treasurer. All subscriptions are due each January, and should be sent to the Member- ship Treasurer, Dr. A. J. Lloyd, Department of Geology, University College, Gower Street, London, WCIE 6BT, England. COUNCIL 1972-3 President: Professor M. R. House, The University, Kingston upon Hull, Yorkshire Vice-Presidents: Dr. Gwyn Thomas, Department of Geology, Imperial College, London, S.W. 7 Mr. N. F. Hughes, Sedgwick Museum, Cambridge Treasurer: Dr. J. M. Hancock, Department of Geology, King’s College, London, W.C.2 Membership Treasurer: Dr. A. J. Lloyd, Department of Geology, University College, Gower Street, London, WCIE 6BT Secretary: Dr. W. D. I. Rolfe, Hunterian Museum, The University, Glasgow, W. 2 Editors Dr. Isles Strachan, Department of Geology, The University, Birmingham, B15 2TT Dr. R. Goldring, Department of Geology, The University, Reading, RG6 2AB, Berks. Dr. J. D. Hudson, Department of Geology, The University, Leicester Dr. D. J. Gobbett, Sedgwick Museum, Cambridge Dr. L. R. M. Cocks, Department of Palaeontology, British Museum (Natural History), London, S.W.7 Other members of Council Dr. M. G. Bassett, Cardiff Dr. E. N. K. Clarkson, Edinburgh Dr. R. H. Cummings, Abergele Prof. D. C. Dineley, Bristol Dr. Julia A. E. B. Hubbard, London Dr. J. K. Ingham, Glasgow Mr. M. Mitchell, Leeds Dr. Marjorie D. Muir, London Dr. B. Owens, Leeds Dr. W. H. C. Ramsbottom, Leeds Dr. P. Rawson, London Dr. Pamela L. Robinson, London Dr. A. D. Wright, Belfast 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 JVest Indies and Central America: Mr. John B. Saunders, Geological Laboratory, Texaco Trinidad, Inc., Pointe-^-Pierre, Trinidad, West Indies Western t/.S./l.:ProfessorJ. 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 © The Palaeontological Association, 1972 THE WATER-VASCULAR SYSTEM IN LIVING AND FOSSIL ECHINODERMS by DAVID NICHOLS The Fourteenth Annual Address, delivered 3 March 1971 Abstract. All Recent echinoderms possess extensile tube-feet, and probably all extinct groups had them too. In the most primitive of the living groups, the Crinozoa, the tube-feet are extended by muscular contractions of the water-vascular canals that link together all tube-feet in the animal; most probably, all fossil crinozoans share this method. In the Asterozoa, as well as the agency of the canals, various accessory structures, such as bulbs and ampullae, also become involved. The Homalozoa appear to have most resembled the asterozoans in the operation of their tube-feet. In the Echinozoa, either the accessory ampullae assume the dominant role in tube-foot protractions or, as in aspidochirote holothuroids, protraction is brought about solely by retrac- tion of adjacent tube-feet; some extinct groups could have adopted this method too. The evolution of the water-vascular system is discussed. When a series of hydraulic structures protrudes into the surrounding sea -water from an otherwise almost totally enclosed animal theca, they may be expected to assume considerable importance in the life of the animal possessing them. Such a situation occurs in the echinoderms, where tube-feet emerge from a rigid test to provide highly mobile, thin-walled, sometimes sucker-ended projections which may play a part in feeding, respiration, burrowing, reproduction, locomotion, and even excretion. They are very important effector organs to members of the phylum, and occur, so far as is known, in all living members; in fact, their presence is diagnostic for the phylum. They are connected together within the body by a series of canals constituting the water- vascular system, which is anatomically a part of the secondary body cavity or coelom. All parts of the system are fluid-filled, the fluid being virtually sea-water with a dense population of coelomocytes suspended in it, and sometimes migrating through its walls to other body cavities, and also out to the surrounding sea-water. The function of the system is to generate, distribute, and control the hydrostatic pressure necessary for the operation of the tube-feet, but secondarily it may also serve other functions which make use of the fact that it carries a fluid of the body into close proximity with the surrounding sea-water; for instance, it may assist in gaseous exchange to the inner parts of the body and in removal of waste. A major problem with any hydraulic system is the recognition of changes in the head of water in which the animal is living (Nichols 1966), so that appropriate adaptations to the changing hydrostatic situation can be brought about. A part of the water-vascular system, the hydropore and its associated structures, is devoted to this aspect, and is a place where the water- vascular system opens to the surrounding sea-water. The system is not wholly unique in the animal kingdom; though nothing precisely comparable occurs elsewhere, it is likely that the ancestry of the system lies in the lophophoral canal system in that group of animals generally referred to as the minor coelomate phyla (Nichols 1967Z>; Smith, Carthy, Chapman, Clark, and Nichols 1971), the canal system which creates and distributes fluid pressures associated with [Palaeontology, Vol. 15, Part 4, 1972, pp. 519-538.] C 9202 M m 520 PALAEONTOLOGY, VOLUME 15 the protraction of the lophophoral tentacles. In most of these phyla the lumina of the lophophore tentacles are connected by a circum-oral coelomic canal, even in some cases with a hydropore-like opening to the exterior. Among comparable systems of this sort are, first the brachial canal system in brachiopods, which is a tubular part of the coelom concerned with maintaining turgor in the lophophoral filaments, and secondly, the mesocoelic cavity of pterobranch hemichordates, which maintains turgor in the tentacles and even has an opening to the exterior, though this is not to say that the opening has a similar function to the echinoderm hydropore. The purpose of this paper is to review our knowledge of the mode of action of the water-vascular system in present-day echinoderms and see how far the structure and function of the system in fossil groups can be inferred from this. THE STRUCTURE AND OPERATION OF THE ECHINODERM WATER-VASCULAR SYSTEM There are three main parts to the system (text-fig. 1) : i. The ring canal, encircling the oesophagus, which may have accessory bulges and pouches off it. ii. The stone canal, which originates at one point on the ring canal and leads to an opening to the exterior, usually called in modern forms the madreporite, because of its resemblance to madreporarian coral. Typically sieve-like in modern forms, it was more usually a single pore in early echinoderms, and hence is referred to as the hydropore, and this term will be used throughout to avoid confusion. The ring canal and stone canal, with the hydropore, may be thought of as the eentral parts of the system. iii. The peripheral system, composed of radial canals and the tube-feet and their accessory structures. The radial canals, usually five or multiples of five, arise from the ring canal and pass centrifugally along each arm or other radial structure. They may be internal or external, and the tube-feet may either lead straight off them, or there may be lateral canals leading to the origin of the tube-feet. The radial and lateral canals may be variously modified with muscular or elastic tissues and with valves to assist in protraction of the tube-feet. The mechanics of the system The central parts of the system maintain a supply of coelomic fluid to each radial canal, as required. Some of the accessory structures which branch from the ring canal, such as the muscular, bag-like polian vesicles, which are present in some members of all groups except crinoids, appear to hold water-vascular fluid in reserve and under slight pressure until it is required by the animal; other accessory structures, such as Tiede- mann’s bodies, are cytopoietic, and manufacture some of the coelomocytes which float in the coelomic fluid and act in transport and excretion (Bargmann and Behrens 1964). The fluid pressure needed to operate the tube-feet is most often generated mainly within the peripheral canals and to do this the canal walls are either muscular or elastic. In advanced echinoderms, however, various accessory structures, such as bulbs and ampullae, may assist the contractile canals in forcing out the tube-feet. On mechanical grounds, such a system requires a ‘safety-valve’ for the release of excess fluid when, say, all tube-feet contract together to cause an unusual increase in pressure within the test. NICHOLS: WATER-VASCULAR SYSTEM 521 Conversely, when a tube-foot or canal is ruptured, as happens quite frequently in life, the system needs to replenish lost coelomic fluid with sea-water from its surrounding. One might expect that the hydropore acts in both these capacities, since it is the only opening to the exterior, and many writers have assumed this to be the case. But a series of experiments by Fechter (1965) and Buchanan (1969) has shown that this is not so: when an unusually high fluid pressure is generated within the animal, for whatever TEXT-FIG. 1. Diagram of a generalized echinoderm water-vascular system. The ring- canal, with its accessory vesicles and bodies, surrounds the first part of the gut. The radial canals lie on the oral side of the arms or brachioles, and the tube-feet, drawn here directed upwards, as in crinozoans and some echinozoans, form avenues to the food grooves. reason, the peristome and periproct first bulge outwards, and then fluid may escape from the anus; but no fluid has been observed to escape from the hydropore, as evi- denced by a manometer sealed over it (text-fig. 2). Similarly, excision of tube-feet and even arm severance does not lead to an incursion of fluid into the hydropore. So neither replenishment of fluid nor the release of excess pressure appear to take place via the hydropore. One must therefore ask: what is the function of the hydropore? The obvious experi- ment of blocking it up and observing the result was first performed by Dakin (1923) who blocked the hydropore of echinoids with shellac and returned them to his shallow aquarium; their behaviour or capabilities did not appear to have been altered. Fechter (1965), however, plunged similarly-treated echinoids to the bottom of a deep aquarium, about 2 m below the water-surface (text-fig. 2, Ci). In this case, the echinoid was unable to protract its tube-feet normally. In the converse experiment, in which a head of water was sealed over the hydropore (text-fig. 2, Cii) so that a positive pressure was created within the water-vascular system relative to the surrounding sea-water, the tube-feet 522 PALAEONTOLOGY, VOLUME 15 all extended to their maximum length and could not be retracted. A head of water of only 20 cm in the tube, Fechter reports, is sufficient to bring about tube-foot extension so that they protrude ‘like taut tubes, incapable of executing lateral [pointing] move- ments’ (my translation). A. Manometer on hydropore — no fluid movement B. Hydropore sealed in shallow water — tube-foot activity normal C. Varying pressure on water -vascular system i. hydropore sealed i i. head of water on in deep water hydropore TEXT-FIG. 2. Fechter’s (1965) experiments on the function of the hydropore in echinoids. A, With a manometer sealed over the hydropore openings and tube-feet in normal activity (left side), no move- ment of the mercury bubble in the capillary, indicating no fluid movement across the hydropore; similarly, if all tube-feet are induced to retract together (right side), there is still no fluid movement across the hydropore. b. If the hydropore is sealed and the animal returned to shallow water, there is no noticeable effect on tube-foot activity, c, i. If the hydropore is sealed and the echinoid is plunged into deep water, so that there is a positive ambient pressure on the water-vascular system, then the tube-feet are unable to protract normally, ii. In the converse situation, in which a head of water is sealed on to the hydropore, the tube-feet extend maximally, unable to retract. I have been unable to repeat Fechter’s results using a tube of water sealed over the hydropore, even with the level of water in the tube as much as 2 m higher than that of the surrounding water: the tube-feet of my experimental animals {Psammechimis) did not protrude like taut tubes. A possible explanation is that care was taken to ensure that the water tube surmounted only the hydropore, and did not overlap any of the gonopores. NICHOLS: WATER- VASCULAR SYSTEM 523 If pressure is applied to the gonopore, then indeed the tube-feet do extend markedly, pressure presumably being transmitted from the gonad sac to the perivisceral coelom and thence to the water-vascular system. One must be cautious in interpreting such experiments; one must remember that the animal will almost certainly never in its life experience a pressure situation like that in the ‘water-tube’ experiment. Nonetheless, experiments like these seem to indicate that however it works, and whatever protective devices, such as valves and sphincters, it may subsequently be shown to possess, the hydropore is concerned at least in part with pressure equalization, necessary when the head of water above the animal changes, as, for instance, in tidal rise and fall. For it to be effective, no fluid need pass across the hydropore: it suffices to be merely an opening, so that the dynamic fluid system operating the tube-feet is confluent with the external milieu. THE CRINOZOA Tube-foot operation in a living crinoid On many grounds, including the structure and operation of the water-vascular system, the Class Crinoidea is the most primitive living class of the echinoderms. An under- standing of the mode of operation of the system in this class is therefore vital to a true understanding of the system in all other living classes. The modern crinoid on which most of the work on the system has been done is the British comatulid Antedon, an unstalked form from shallow water. The following account is based on this genus (Nichols 1960), though the system in other crinoids probably works in much the same way. The ‘central’ part of the system consists solely of a ring vessel: there is no stone canal or hydropore as such. The central disc is not rigid, and it may well be that the adjust- ment to pressure change differs from that of rigid echinoderms. This may be why the hydropore as such is absent. Instead, the whole tegmen (the upper surface of the animals’ disc) is perforated by tiny pores, and the whole of the ring canal is similarly perforated, like an irrigation pipe. Perhaps such a confluence of water-vascular fluid via perivisceral coelom to the outside water suffices the needs for pressure regulation. A comparable simpliflcation of the hydropore may also be seen in some flexible-bodied holothuroids. The peripheral part of the system consists of radial canals running along the upper (oral) side of each arm, of which there are ten in Antedon, with branches to each pinnule. The canals are flattened, not cylindrical. From both brachial and pinnular canals there arise the tube-feet, in groups of three. Each arm group has tube-feet of similar size, but in each pinnule group there is one long, one medium, and one short tube-foot: the long ones are held out laterally, and nearly touch their equivalents from the adjacent pinnule, the medium ones project dorso-laterally and the small ones project nearly vertically upwards on either side of the food groove, like a fence bordering a shallow ditch (text- fig. 3). Between each group of tube-feet the brachial or pinnular canal is constricted transversely, so that it forms a linear series of interconnecting compartments. Muscle fibres stretch across the actual lumen of the canal. Additionally, in the brachial canal there is a longitudinal gulley in the floor of the canal. Protraction is brought about mainly if not entirely by generation of pressure produced by contraction of the muscles within the brachial and pinnular canals: there are no 524 PALAEONTOLOGY, VOLUME 15 TEXT-FIG. 3. Diagram to show the structure of the water-vascular system in a modern crinozoan, based on the British comatulid Antedon. A radial water-vascular canal, traversed by muscles, runs along the oral side of each of the ten arms and their alternating pinnules, giving off groups of tube- feet on either side of the food groove; the canals are constricted at intervals, prominently so in the pinnules. The arm canal, shown in section at lower left, has a ventral by-pass running along its length, to ensure a through-channel to more distal parts of the system; no such channel exists in the pinnule canal, shown in section at lower right. accessory structures. Considering first the situation in each arm, the sequence of events as the canal muscles contract is probably this; first, the ventral gulley is closed off so that water-vascular fluid can still by-pass the region undergoing activity; then the trans- verse constrictions on either side of the tube-feet to be protracted are closed off; then finally, the volume of the compartment so formed is reduced, to drive fluid into the lumina of the tube-feet, extending them. In the pinnules, the situation is much the same, except that there is not the same problem of maintaining a by-pass for fluid to distal regions, so the ventral gulley is missing. In both arms and pinnules the change in shape of the canal is brought about at the expense of cavities lying above and below, that is, the sub-neural and sub-tentacular NICHOLS: WATER-VASCULAR SYSTEM 525 canals. The walls separating the canals are thin yet tough, to permit flexibility without rupture. The walls of the tube-feet are muscular, to bring about both bending and retraction. By their movement the tube-feet can distribute a sticky mucous food net, collect it up again and pass it into the food groove, whence it is transported to the mouth. To bring about retraction, the longitudinal muscles of the tube-foot contract and the canal muscles relax, so that fluid from the tube-foot lumen is transferred back into the canal. The water-vascular system in fossil crinozoans In modern crinozoans, the tube-foot system is very extensive, and one presumes that the tube-feet subserve almost all the respiratory needs of the animal in addition to being the main feeding organs. Certainly no other special respiratory organs occur. This is not so in many extinct crinozoan groups, where more often than not there are additional respiratory devices, such as the dichopore and fistulipore systems of cystoids and some paracrinoids, the hydrospires of blastoids, the epispires of eocrinoids and the cataspires of parablastoids. Perhaps it is safe to infer from the presence of these other respiratory surfaces that the tube-foot systems of these extinct groups were not so exten- sive as that of today’s crinoids. On the fossil evidence the system of brachioles and pinnules in some of these groups does not appear to have been very extensive, parti- cularly in those forms which we regard as the more primitive members of the groups, in which the brachioles arise from facets very close to the mouth, often only two or three to an animal. But then in later forms of the same groups there is sometimes a trend to increase the number of brachioles, and they come to arise from a larger and larger part of the theca. It does appear that the blastoids may have had a crinoid-like water-vascular system. Well-preserved specimens sometimes show a ring-like groove in the wall of the peristome which probably held a ring canal in life (Breimer and Macurda 1972) and this may have opened, in the crinoid manner, by a number of hydropores. With the exception of the blastoids and most of the crinoids, all other crinozoan groups appear to have had a hydropore. Sometimes it becomes double, and in some cystoids, such as Jaekelocystis (Kesling 1967), it even has a sieve plate across it, as in the more advanced echinoderm groups. This suggests that in all these cases the central parts of the water-vascular system, that is, the ring vessel and associated organs, were internal, and in a few cases it is possible to see the pores through the theca by which the radial canals left the inside to travel up the outside of each arm or brachiole. So beyond making somewhat vague assertions about the extent of the water-vascular system in fossil crinozoans, we cannot go far in interpreting details of its structure, be- cause, as in modern crinoids, most of the peripheral parts of the system probably lay away from the ossicles and left no fossilizable impression. But the crinozoan system, as interpreted from living forms, sets the scene for an understanding of the evolution of the system in the other, more advanced, echinoderm groups. THE ASTEROZOA From the viewpoint of the state of the water-vascular system, the next evolutionary grade after the Crinozoa is seen in the Asterozoa, at least so far as echinoderm groups 526 PALAEONTOLOGY, VOLUME 15 with living members are concerned. When this Class first appears in the Ordovician, with forms like Chiniamster, the system is remarkably crinoid-like, though there are important differences (Fell 1963o, b). Of course, there is in the Asterozoa a radical change in the posture of the body, in that the mouth now faces downwards. Some crinoids, such as Edriocrinus, may well have foreshadowed this reversal of their normal feeding posture by turning mouth-downwards, to feed from the sea-bottom, rather than from suspended matter in the water. But in asterozoans the ‘mouth downwards’ posture is the rule, and has been accompanied by a corresponding change in the uses to which the tube-feet are put, so they must be constructed so as to adhere to, or burrow into, the sea-bottom while still usually also playing a part in feeding. In asterozoans, the main structures for operating the tube-feet are still the canals, but there is a new departure, in that the canals are aided by special extra organs, such as bulbs and ampullae, one to each tube-foot, so that the strength and extent of pro- traction of the tube-feet can be increased. Somewhat similar to crinoids in the general design of the water- vascular system, and the least sophisticated of the asterozoans are the ophiuroids. Though exhibiting a wide variety of habitat, more often than not the ophiuroid arm is used to capture food, either from the water or from the film on the sea-bottom. The tube-feet, single rather than grouped, very often create, spread, then gather up a mucous food net, and, like crinoids, they are highly active when the animal is feeding. What follows is a generalized picture, based on several different British ophiuroids (Buchanan and Woodley 1963; Woodley 1967). While the radial canal running up each arm contributes to the build-up of hydrostatic pressure for tube-foot operation, as in crinoids, it does so by virtue of the elasticity of its walls, not by inherent musculature. A further difference from crinoids is that the canal leading out to each tube-foot is expanded into a muscular bulb (text-fig. 4), and it is this that apparently provides the principal muscular power for tube-foot extension, though one must not disregard the possible contribution made by other tube-feet when contracting (see below, p. 532). Just as the crinoid canal has muscular constrictions to confine the increase of hydrostatic pressure in the canal to where it is required, so the ophiuroid lateral canal has a muscle-operated valve on the proximal side of the bulb, so that any increase in water pressure within the bulb can be directed into the lumen of the tube-foot rather than dissipated down the canal. The tube-feet are, of course, themselves muscular for retraction. But the lumen of the extended tube-foot holds more fluid than does that of the head-bulb, so the excess fluid must escape down the lateral canal to the neighbouring part of the radial canal, which expands against the elasticity of its walls to accommodate this fluid. Additional elastic-walled reservoirs arise from the dorsal side of the radial canals in some ophiuroids, to help store additional water- vascular fluid. For fluid to escape from the tube-foot and its head-bulb, the tube-foot valve must be endowed with its own musculature so that it can open against a pressure on the distal side. The sequence of events in ophiuroid tube-foot protraction (text-fig. 4) is probably this: 1 . The radial canal partly contracts by elasticity, while the head-bulb relaxes, allowing the tube-foot to extend slightly. 2. The valve shuts and the head-bulb contracts, further forcing out the tube-foot. NICHOLS: WATER-VASCULAR SYSTEM 527 TEXT-FIG. 4. Diagram to show the probable sequence of events taking place in tube-foot pro- traction in a generalized ophiuroid, after Woodley (1967). a, Part of the arm, showing a short series of ambulacral ossicles (‘vertebrae’) and, superimposed in section, the radial canal and a pair of tube-feet, retracted, b. The first part of the protraction cycle, in which the muscles of the head-bulb relax, while the radial canal partly contracts by elasticity; the tube-foot partly extends, c. The head-bulb muscles contract, forcing fluid from the head- bulb into the tube-foot lumen, d. The muscles of the head-bulb relax, allowing further elastic contraction of the radial canal, so that the tube-foot extends to its maximum. 3. The valve opens and the head-bulb relaxes and the radial canal fully contracts by elasticity, to produce maximum tube-foot protraction. So apart from the possible involvement of neighbouring tube-feet in the protractive process, the principal muscular component is the head-bulb, and this is anchored proximally in a concavity in the underside of the arm plates. A similar concavity is found 528 PALAEONTOLOGY, VOLUME 15 in some of the earliest asterozoans to occur in the fossil record, such as Chinianaster, which is regarded by some (e.g. Fell \963a, b; Ubaghs 1967; Nichols 1969) as lying close to the point where the ophiuroids and asteroids diverged. In the subsequent lines from this early fossil, while the ophiuroids appear to have retained the chinianasterid con- dition of the water-vascular system, the asteroids adopted a different method of creating pore to interior ambulacral ossicle Chinianasterid TEXT-FIG. 5. A possible phyletic diagram of the major stages in the evolution of the peripheral water- vascular system of the Asterozoa. In each case a short section of the ambulacrum is shown in ventral view. The somasteroid Chinianaster is seen as lying close to the common stem from which both modern ophiuroids (top left) and modern asteroids (top right) have arisen. protractive pressures; they evolved accessor ampullae within the body cavity of the arm, the presence of which is shown in the fossils by pores between the arm plates carrying canals to these structures in life. The probable course of evolution from Chinianaster through such forms as Villebrnnaster to the modern asteroids can be traced with some confidence (text-fig. 5), and shows the gradual reduction in the size of the ambulacral plates from the ophiuroid-like ‘vertebrae’ to the delicate bridal arch of ambulacral plates of modern asteroids (Fell 1963<7). NICHOLS: WATER-VASCULAR SYSTEM 529 The acquisition of ampullae did not, however, relieve the radial canals from parti- cipating in tube-foot protraction : the canal is highly muscular in modern starfishes and can distend to receive fluid from retracting feet, and contract to distribute it (Nichols 1969; Blackman 1971). The valve guarding the tube-foot/ampulla system is, as in ophiuroids, muscle-operated, and can both shut and open against pressures on either side of it. Like the head-bulb of the ophiuroid tube-foot, the cavity of the ampulla is of TEXT-FIG. 6. Diagram of tube-foot operation in modern asteroids. When the tube-foot is retracted (1) , the ampulla and adjacent part of the radial canal are expanded. The first part of protraction (2) usually involves the contraction of only the appropriate section of the radial canal, and this con- tinues (3) until the tube-foot is extended to between half and three-quarters of its maximum length. The final stage for maximum contraction is brought about by contraction of the ampulla (4). On retraction, the ampulla usually fills first (5), then the canal (6), and fluid may also pass to other parts of the system, which can be detected by leaks in the nearby parts (7). very much smaller volume than the lumen of the fully extended tube-foot, so excess fluid must be able to escape down the radial canal and perhaps assist in the protraction of other tube-feet nearby, or at least be accommodated by the radial canal for future use. So a tube-foot and its ampulla are not necessarily antagonistic in activity, as was previously thought. One can check this directly by cutting a window in the dorsal inte- gument of a starfish using a miniature circular saw, or other fast cutting tool that renders minimum damage to the basi-epithelial nerve plexus. Very often, this can be done without introducing excessive stimulation to the tube-feet, so one can watch them under near-normal activity, and can ‘match up’ a tube-foot and its ampulla during the stepping cycle (text-fig. 6). Usually a tube-foot will begin to extend unaccompanied by a contraction of its ampulla — in fact, its ampulla may also expand during this phase. The ampulla, if it takes part at all, contracts to provide about the last half of tube-foot extension. When the foot retracts, its ampulla usually takes up the first portion of the 530 PALAEONTOLOGY, VOLUME 15 fluid to be displaced, and so it expands. Then the valve presumably opens, against pressure, and releases further fluid back into the rest of the system. THE HOMALOZOANS This, the only subphylum of the echinoderms (on Ubagh’s (1967) classification) with no living representative, has a water-vascular system that most resembles that of the TEXT-FIG. 7. Diagrams of the ambulacra in the Homalozoa. a, A homostelean, such as Trochocystites, in which there are food grooves in the plates of the margin, b, A homoiostelean, such as Deudrocystoides, in which the food groove is borne on a single projecting arm. c, A stylophoran, such as Cothurnocystis, in which the food groove is also on a single arm; an enlarged view of part of this arm shows an ophiuroid-like arrangement of ambulacral ossicles. asterozoans, apparently, so it is appropriate to consider it here. Formerly called ‘car- poids’ or ‘heterosteles’, its members have irregular symmetry and flexible theca. Gas exchange, apart from that taking place across the tube-feet, was probably anal, as in some of today’s holothuroids, the flexible theca aiding this. Three classes are generally recognized, each of which has members with a typical crinozoan-like food groove (text- fig. 7), with cover-plates, which, distally at least, could most probably be opened to allow protraction of tube-feet. The arrangement of food grooves in the three classes is as follows: 1 . Homostelea — grooves round the front of the theca leading to the mouth. NICHOLS: WATER-VASCULAR SYSTEM 531 2. Homoiostelea — single feeding arm, with groove down one side, on the side oppo- site the stem. 3. Stylophora — single feeding arm, with groove down one side, from one side of the boot-shaped theca. It is in the Stylophora that we know the ambulacral structure best (text-fig. 7, c). The medial groove, which probably contained a radial nerve and radial water-vascular canal, has side-branches from it leading to lateral depressions, in very much the same arrangement as in early asterozoans (compare text-figs. 5 and 7, c), and one can assume that here too the cupules marked the site of tube-foot origin, probably with head-bulbs operating in much the same way as those in today’s ophiuroids (see p. 526 above). Each cupule is joined to its ipsilateral neighbour by a short groove, which may have held lateral ambulacral nerves, as in modern asteroids (Nichols 1967<7). THE ECHINOZOANS Echinoids and holothuroids Here, as in asterozoans, tube-foot operation is usually assisted by accessory ampullae, though it appears that some echinozoans have delegated the task of protraction almost solely to the ampullae; here, it seems that the radial and lateral canals take only a minor part in pressure generation. Modern echinoids, for instance, usually have very flat ampullae (text-flg. 8), set close together, and there is hardly any distance of radial canal between the origins of successive branches to the tube-feet which might be concerned with the development of pressure, though the canal as a whole can expand to receive excess water-vascular fluid, say, when all tube-feet in a segment of one ambulacrum retract at the same time. In other modern echinozoans, such as some holothuroids, the specialization in the protractive process has been in another direction, namely, to make no use of contractile canals and ampullae, but rely solely on the retraetion of neighbouring tube-feet. One interesting aspect of the evolution of the echinoid water-vascular system which emerges from a study of the fossil record (see, for instance, Kier 1965) is that early eehinoids had the radial canal enclosed within the calcite of the ambulacral plates, whereas with the passing of time the canal was freed from this enclosure and came to lie against the inside surface of the plates, that is, within the body cavity. In view of what we know about the participation of the canal in primitive echinoderms, one wonders whether this ‘emancipation’ of the canal might not have allowed its greater participation in the protraetive process, unrestricted by enclosure within a calcite tube. The question arises, however: why did the canal become enclosed in the first place, if this group arose, as seems likely, from a primitive group with external canals ? Possibly with the increased importance of the ampullae in the protractive process the protection of the canal within the plates assumed a selective advantage, and this advantage was maintained when the eanal ‘broke through’ to the interior of the theca, added to which it could again parti- cipate in protraction. In a normal, active echinoderm, usually only about half its tube-feet are protracted at any one time (Fechter 1965; Blackman 1971) the other half being retracted. This period of retraction may be necessary to permit the mucous glands of the tube-foot sucker to recharge. What is interesting, from the viewpoint of tube-foot mechanics. 532 PALAEONTOLOGY, VOLUME 15 however, is that usually as one tube-foot withdraws, another nearby extends, and, as mentioned above, it seems probable that the contraction of the one may be assisting in the protraction of the other. Some holothuroids demonstrate this principle well: while some, such as the British dendrochirote Pawsonia (= Cucumaria), have accessory TEXT-FIG. 8. Diagram of the peripheral water-vascular system in the Echinoidea. The ampullae take the major part in protraction, and since the tube-foot and its ampulla are an important agency in gas exchange, circulation of water-vascular fluid is assisted by having the ambulacral pore subdivided, the circulating currents passing through each in opposite directions. A valve can isolate a tube-foot and its ampulla from the rest of the system. ampullae, others, such as the British aspidochirote Holotlmria, have no visible internal structures which might play the major part in protraction, such as ampullae or muscular or elastic canals. But it is a fact that whenever a tube-foot of Holotlmria is extended, a neighbouring one is being withdrawn (text-fig. 9). A prerequisite for this system to work is, of course, that there must be a muscle-operated valve at the origin of each tube- foot, and such a structure is indeed present. NICHOLS: WATER-VASCULAR SYSTEM 533 The implication of this principle, used to such good effect by Holothwia and clearly of importance to other groups too, is that it is not necessary for an echinoderm to show evidence of contractile structures in the water-vascular system to possess, or have possessed, highly extensile tube-feet. This principle is particularly pertinent to one group of echinozoans, the cyclocystoids, which are entirely extinct. TEXT-FIG. 9. Diagram of tube-foot operation in an aspidochirote holothuroid, based on Holothiiria. The ‘central’ parts of the system are shown, and the origins of the five radial canals; the ventral radial canal is drawn with two pairs of tube-feet. Tube-foot extension is brought about by retraction of neighbouring tube-feet, the process being controlled by valves at the head of each tube-foot and by the tube-foot retractor muscles. The hydropore is internal, because the body is flexible, and therefore transmits changes in external hydro- static pressure to the coelomic fluid. The cyclocystoids This is one of the most fascinating and controversial of the extinct echinozoan groups. Its members are disc-like fossils (text-fig. 10) with an outer ring of hefty ossicles each bearing two (rarely one or three) large pores which are confluent by canals with the interior of the theca. Outside the ring is a skirt of small plates. There is an upper (aboral) and lower (oral) integument, with loose-fitting plates, within the ring. What is generally regarded as the upper surface has pores between the plates which most likely bore papula-like structures, that is, external blisters of the integument which have a 534 PALAEONTOLOGY: VOLUME 15 lumen continuous with the perivisceral body coelom, but unconnected with the water- vascular system. On the inner face of the upper integument is a series of grooves radiating from the centre of the theca and passing beneath the pores between the plates. These are thought by some workers all to have contained elements of the water-vascular system, and therefore the soft structures protruding from the pores would have been tube-feet. But it seems more likely that the grooves merely directed the coelomic ciliary currents to the region of each papula, as an aid to respiratory efficiency, though some may have held radial canals leading towards the periphery. As for the large pores in the respiratory papulae TEXT-FIG. 10. One interpretation of Cyclocystoides (Camb.). In this attempt, the plates within the submarginal ring are seen as the aboral surface, bearing respiratory papulae from pores between the plates. The large pores in the sub- marginal plates are interpreted as bearing downwardly directed tube-feet in life, the retraction of one aiding the extension of its neighbour. submarginal ring of plates, it seems highly probable that they gave rise to large tube- feet for loeomotion, and the skirt of small plates would then have acted to protect both these tube-feet, and the downward-direeted mouth in the centre of the undersurface. Other interpretations of this fossil have been given (see, for instance, Sieverts-Doreck 1951, Kesling 1967; Durham, in Nichols 1969; Henderson and Shergold 1971). Some think that the large tube-feet could not have been locomotory, because they would have required large ampullae. But suppose they operated in the manner of modern holothuroids, the retraction of one foot providing the main hydraulic pressure for the protraction of its neighbour: the tube-feet are, after all, often in pairs, and this is parti- cularly well shown in those specimens which are well enough preserved for an accurate count to be made. Sueh a mechanism could explain the functional link between the two tube-feet arising from one submarginal plate. But it must be emphasized that this is only another possible interpretation of this fossil : until it is better known, it eannot be said definitely which orientation is correct. Ophiocislioids, camploslromatoids, and lepidocystoids Some interpretations of the little-known ophiocistioids (see, for instance, Ubaghs 1966; Fell and Pawson 1966) have it that the fiexible, plated structures arising from the NICHOLS: WATER-VASCULAR SYSTEM 535 under-surface of the theca were heavy, plated, permanently retracted tube-feet. By com- parison with other echinoderms this seems very strange, since nowhere else in the phylum are tube-feet so cumbersome or irretractible. Unfortunately, the morphology of these large, plated structures is imperfectly known, so one cannot be sure what they are; the fact that they were almost certainly hollow does indeed suggest that they are tube-feet, but the possibility of their being brachioles must not be ruled out. One could perhaps point to the camptostromatoids and lepidocystoids as echinoderms with similar plated structures arising, as in ophiocistioids, in five columns radiating from the mouth, and in these groups the structures are normally regarded as brachioles. Certainly, in Lepido- cystis, in which the structure of these projections is better known, there does appear to be a food groove down one side (Durham 1967a), and maybe further finds in the other two groups will clarify the structure of these projections in them. The edriosteroids and helicoplacoids The most interesting fossil echinozoans are those that were on the scene apparently long before the first echinoids and holothuroids appeared. These are the stalked or flattened edrioasteroids (Regnell 1966) and the spirally plated helicoplacoids (Durham and Caster 1963). Both these groups first occur in the lowest Cambrian rocks (Durham 1967c), and in both groups some members at least show evidence of pores between the ambulacral plates, indicating that a water-vascular system was present very early in the history of the group. In the edrioasteroids it is possible to say with fair certainty that the peripheral water-vascular system was external, because in some there is a pattern of external grooves in each ambulacrum which most probably cradled the radial water- vascular canal. There are also pores between the ambulacral plates through which canals probably passed to ampullae within the body. But not all edrioasteroids show either the external radial grooves or the interplate pores. Perhaps these edrioasteroids lacked a water-vascular system altogether, and relied on external ciliary currents for feeding; or perhaps the external parts of the water- vascular system were, like modern crinoids, contained in the soft tissues lying well away from the ambulacral ossicles, and not requiring the use of accessory ampullae. Among the helicoplacoids, Waucobella (Durham 1967Z?) is a genus that shows ambulacral structures particularly well. The column of ambulacral pores is single, and there is no evidence of external or internal grooves which might have held a water- vascular canal, so it is not possible yet to say whether the system was internal or external. The helicoplacoids probably adopted an upright posture in a depression in the slit, and could untwist to extend for feeding, and twist up again to retreat. In twisting and untwisting they could open and close the plates covering the ambulacra and also could open up a spiral series of channels in the interambulacral areas of the animal’s outer surface. Though tube-feet were almost certainly present, as evidenced by the pores in the ambulacra, it seems possible, as Durham (1961b) has said, that they were for respiration only, since in most helicoplacoids the ambulacral system is ridiculously short to cater for the feeding requirements of the animal. Perhaps food collected all over the animal’s exterior, was passed to the mouth at the upper pole, along the interambulacral channels. N n C 9202 536 PALAEONTOLOGY: VOLUME 15 SUMMARY AND CONCLUSIONS Only in the living classes of echinoderms have we direct evidence for the structure and operation of the water- vascular system. So, using the living forms as a basis, the evolu- tionary trends in the elaboration of the water-vascular system may be summarized as follows (see also text-fig. 11): 1. The most primitive situation is seen in the crinozoans, in which the radial canal itself is the sole pressure generator. TEXT-FIG. 11. Summary diagram of the various kinds of peripheral water-vascular system seen in the three extant sub-phyla of echinoderms; the suggested evolutionary lines are based on the con- figuration of this system only. 2. In ophiuroids the lateral canal to each tube-foot plays a part. 3. In asteroids the radial and lateral canals are the main pressure generators, but assisted by ampullae to increase the effective extent of tube-foot protraction. 4. In echinoids and some holothuroids, independently, the greater part of pressure generation is performed by ampullae. 5. In other holothuroids it is other tube-feet which create the necessary pressure. NICHOLS: WATER-VASCULAR SYSTEM 537 On this framework can be hung an interpretation of the system in some extinct forms : 6. Most extinct crinozoans, such as blastoids, cystoids, eocrinoids, paracrinoids, parablastoids and edrioblastoids, probably had structures very similar to those of living crinoids, with radial canals supplying the main protractive force. 7. The homalozoans may well have adopted an ophiuroid-like system, using head- bulbs. 8. The helicoplacoids possibly had a single column of tube-feet worked on the echinoid principle. 9. The edrioasteroids probably used canals and ampullae in an asteroid-like manner. 10. The cyclocystoids may have used the retraction of one tube-foot to protract another, as in some modern holothuroids. The history of higher invertebrate animals is largely a story of elaboration of the coelom. The original function of the secondary body cavity, which is the function re- tained for it by the great majority of invertebrates, is to transfer pressure changes from one place to another, chiefly in connexion with the extension of soft structures. The echinoderms have retained this function too, but not for the whole coelom: the main part of the coelom is enclosed within a more or less rigid skeletal capsule, and is not involved in pressure-changes, except passively ; but the water-vascular system represents a part of it which does involve the primitive function, in that it creates and controls the pressure changes which operate a vitally important set of soft parts, the tube-feet. Acknowledgements. The work summarized here has been done over several years, and it is a pleasure to record my indebtedness to my colleagues and students, past and present, for many discussions. I mention particularly Dr. R. A. A. Blackman, Dr. A. C. Campbell, Dr. D. Heddle, Dr. R. C. Higgins, Miss C. Swann, and Dr. J. D. Woodley; responsibility for the opinions expressed, and any errors, however, rests with the author. Part of the work was assisted by grants from the Science Research Council, to whom grateful acknowledgement is made. REFERENCES BARGMANN, w. and BEHRENS, B. 1964. tJber die Tiedemannschen Organe des Seesterns (Asterjas nibens L.). Z. Zellforscli. mikrosk. Anat. 63, 120-133. BLACKMAN, R. A. A. 1971. Functional morphology and hydrodynamics of the asteroid peripheral water- vascular system. Thesis for Ph.D., University of Newcastle upon Tyne (unpublished). BREiMER, A. and MACURDA, D. B. 1972. The phylogeny of the fissiculate blastoids. Kon. ned. Akad. Wetensch. 26, 375 pp. (in the press). BUCHANAN, J. B. 1969. Feeding and the control of volume within the tests of regular sea-urchins. J. Zool. Lond. 159, 51-64. and WOODLEY, j. d. 1963. Extension and retraction of the tube-feet of ophiuroids. Nature, Lond. 197, 616-617. DAKIN, w. 1923. Function of the water-vascular system of echinoderms. Proc. Lpool biol. Soc. 37, 70-73. DURHAM, J. w. 1964. The Helicoplacoidea and some possible implications. Yale sci. Mag. 39 (2), 24-28. 1967fl. Camptostromatoids and Lepidocystoids, in moore, r. c. (ed.). Treatise on Invertebrate Paleontology, Part S, Echinodermata 1 (2), S631-637. University of Kansas. 19676. Notes on the Helicoplacoidea and early echinoderms. /. Paleont. 41, 97-102. 1967c. The incompleteness of our knowledge of the fossil record. J. Paleont. 41, 559-565. and CASTER, k. e. 1963. Helicoplacoidea : a new Class of echinoderms. Science, N. Y. 140, 820-822. 538 PALAEONTOLOGY, VOLUME 15 FECHTER, H. 1965. t)ber die Funktion der Madreporenplatte der Echinoidea. Z. vergl. Physiol. 51, 227-257. FELL, H. B. 1963fl. The phylogeny of sea stars. Phil. Trans. R. Soc. Ser. B, 246, 381-435. 19636. The evolution of the echinoderms. Smithson, aim. Rep. 1962, 457-490. ■ and PAWsoN, d. l. 1 966. General biology of echinoderms, in boolootian, r. a. (ed.). Physiology of Echinodermata, 1-48. Wiley Interscience, New York. HENDERSON, R. A. and SHERGOLD, J. H. 1971. Cyclocystoides from early Middle Cambrian rocks of Northwestern Queensland, Australia. Palaeontology, 14, 704-710. KESLiNG, R. V. 1967. Cystoids, in moore, r. c. (ed.). Treatise on Invertebrate Paleontology, Part S, Echinodermata 1 (1), S85-267. University of Kansas. KiER, p. M. 1965. Evolutionary trends in Paleozoic echinoids. /. Paleont. 39, 436-465. NICHOLS, D. 1960. The histology and activities of the tube-feet of Antedon bifida. Q. J. microsc. Sci. 101, 105-117. 1966. Functional morphology of the water-vascular system, in boolootian, r. a. (ed.). Physiology of Echinodermata, 219-244. Wiley Interscience, New York. \961a. In discussion after Jefferies, r. p. s. Some fossil chordates with echinoderm affinities. Symp. zool. Soc. Loud. 20, 204-205. 1961 b. The origin of echinoderms. Symp. zool. Soc. Lond. 20, 209-229. 1969. Echinoderms, 4th edn. 192 pp. Hutchinson University Library, London. REGNELL, G. 1 966. Edrioasteroids, in moore, r. c. (ed.). Treatise on Invertebrate Paleontology, Part U, Echinodermata 3 (1), U136-173. University of Kansas. sieverts-doreck, h. 1 95 1 . Uber Cyclocystoides Salter and Billings und eine Neue Art aus dem belgischen und rheinischen Devon. Senckenbergiana, 32, 9-30. SMITH, J. E., CARTHY, J. D., CHAPMAN, G., CLARK, R. B., and NICHOLS, D. 1971. The invertebrate panorama, x+406 pp. Weidenfeld and Nicholson, London. UBAGHS, G. 1966. Ophiocistioids, in moore, r. c. (ed.). Treatise on Invertebrate Paleontology, Part U, Echinodermata 3 (1), U174-188. University of Kansas. 1967. General characters of Echinodermata, in moore, r. c. (ed.). Treatise on Invertebrate Paleon- tology, Part S, Echinodermata 1 (1), S3-60. University of Kansas. WOODLEY, J. D. 1967. Problems in the ophiuroid water- vascular system. Symp. zool. Soc. Lond. 20, 75-104. DAVID NICHOLS Department of Biological Sciences University of Exeter Exeter, EX4 4PS Typescript received 24 January 1972 HIATUS CONCRETIONS AND HARDGROUND HORIZONS IN THE CRETACEOUS OF ZULULAND (SOUTH AFRICA) by W. J. KENNEDY and H. c. klinger Abstract. Horizons of bored concretions, ‘hardgrounds’ or ‘hiatus concretions’ of European authors, occur at several levels in the Aptian to Coniacian marine sediments of Zululand (South Africa). They show signs of a complex burial and excavation history, are bored by lithophagous bivalves and encrusted by a variety of epizoans, while vacated borings have been secondarily occupied by nestling bivalves and other organisms. Recognition of these horizons provides insight into substrate conditions, bathymetry and diagenetic history, whilst explaining the incomplete nature of the Cretaceous succession in the area. In Zululand, Cretaceous rocks outcrop over a broad belt extending for 250 km south of the Mozambique border (text-fig. 1). Actual exposures are poor (less than 1% of the area), due to an extensive mantle of Tertiary and Recent sediments. The Cretaceous succession is over a kilometre in thickness, and marine horizons from Barremian to Maastrichtian have been demonstrated on macro- and microfaunal evidence at the surface, whilst a continuous marine sequence up into the Palaeocene is inferred from borehole evidence (Davey 1969, Pienaar 1969). Recent field-work in this area (Kennedy and Klinger 1971, in preparation) has shown that the succession is far from complete, and that previous doubts as to the presence of parts of the Lower Albian, Upper Cenomanian and the Turonian are confirmed. At these levels (and at several others), horizons of winnowed, bored and encrusted con- cretions, sometimes rolled and incorporated into later concretions, can be recognized. These horizons resemble the hardgrounds of the European Mesozoic described by Hallam (1969) and others, and the ‘Hiatus Konkretionen’ of Voigt (1968). The period of formation of these horizons in one case spanned more than a stage, whilst others developed within the duration of a single ammonite subzone. The Cretaceous history of Zululand is outlined below, occurrences are described, and the sedimentary, diagenetic, and palaeoecological significance of the bored concretion horizons is discussed. STRATIGRAPHIC SUCCESSION In northern Zululand, Jurassic Lebombo Volcanics are overlain by pre-Upper Aptian to pre-Barremian coarse clastic Cretaceous sediments. These consist of interbedded conglomerates and sandstones, with log beds, and are presumably of fluviatile origin. They are followed by similar sediments with conglomerates becoming less important, trigoniid shell-pebble beds appearing and Teredo-bored logs abundant; these pass up into a variable series of Barremian to Aptian, or Upper Aptian, silts and shell beds, also with logs. The lowest horizons of bored concretions noted are in the Aptian. The Albian /Aptian boundary is a non-sequence, marked by a bed of bored concretions. Above, the Albian is an expanded sequence with faunas extending up to the Stoliczkaia [Palaeontology, Vol. 15, Part 4, 1972, pp. 539-549, pis. 106-108.] 540 PALAEONTOLOGY, VOLUME 15 dispar Zone. Lithologies are usually silts, with shell beds. A number of bored concretion horizons have been recognized in the Lower Albian. The Cenomanian again consists of silts with shell beds and concretions; at the Mozambique border only the Lower Cenomanian is preserved, but at the Skoenberg, along the Mzinene River north-east of Hluhluwe, the succeeding Middle Cenomanian and part of the Upper Cenomanian are represented. Turonian rocks are absent at the surface in Zululand. The Cenomanian/Coniacian boundary is exposed along the Mzinene River, and is a slight angular unconformity. A horizon of bored concretions with a Cenomanian fauna lies beneath a thin Ptero- trigouia conglomerate with Coniacian Proplacenticeras. South of the Mzinene, at Riverview on the north bank of the Mfolozi near Mtuba- tuba, this unconformity increases, and a slightly higher horizon in the Coniacian rests on Lower Cretaceous fluviatile conglomerates. At Umkwelane Hill, south of the river, the Coniacian rests first on Stormberg Basalts, and then on granitic basement. Above this unconformity the Coniacian to Lower Maastrichtian consists of a thick sequence of silts, fine sandstones, shell-beds and concretionary layers, known chiefly from the area around Lake St. Lucia. We have seen no examples of bored concretions in this part of the succession. LOCALITIES Localities at which we have seen bored concretions are shown in text-figure 1 . Precise latitudes and longitudes are given below. (1) Upper Aptian Mfongosi Spruit, 8 km north-north-east of Otobotini, exposes horizons from Lebombo Volcanics through conglomerates and sands and up into marine Aptian and Albian sedi- ments (Haughton 1936: 286). Degraded cliffs at locality A (Latitude 27° 22' 04" south. Longitude 32° 09' 03" east) and locality B (Latitude 27° 2T 43" south. Longitude 32° 09' 03" east) on the north and south sides of the spruit (intermittent stream) approximately 700 m downstream from the old drift show Upper Aptian silts, shell beds and concretion layers. One concretionary shell bed contains bored concretions. The sequence can be broadly interpreted as a series of small-scale sedimentary cycles: burrowed silts with a partly in situ molluscan fauna alternating with winnowed shell and wood beds. In the silts, infaunal bivalves are prominent; thick-shelled trigoniids, Veniella, Gervillella and oysters dominate shell beds. Silts clearly represent quieter water deposition, whilst shell beds record high energy episodes or non-sedimentation. (2) The Albianj Aptian boundary. Mlambongwenya Spruit lies 20 km north-east of the Mfongosi. The boundary is exposed on the north bank, west of the drift and south of Mlamhongwenya Store at locality C (Latitude 27° 10' 59" south. Longitude 32° IT 8" east). The junction is also seen along the Mfongosi at locality D (Latitude 27° 2T 38" south. Longitude 32° 09' 57" east). At both localities there is a non-sequence at the junction. Silts below yield Diadochoceras and Tropaeum; above is a fauna rich in Douvilleiceras. The actual contact is a line of bored concretions, overlain by a drifted shell bed. Along the Mzinene River, a rather similar sequence is exposed at several localities, although the actual junction is not seen. (3) Lower Albian. Bored concretions occur at several levels in the thick sequence of silts, concretions and shell beds with Douvilleiceras at several localities in northern Zululand. There are good exposures along the Mfongosi and Mlambongwenya, at localities C and D. To the north, we have seen loose bored concretions derived from similar horizons on the west and north banks of Qotho Pan, west of Ndumu (locality E, Latitude 26° 56' 22" south. Longitude 32° 12' 48" east and locality F, Latitude 26° 55' 59" south. Longitude 32° 18' 04" east). KENNEDY AND KLINGER: HIATUS CONCRETIONS AND HARDGROUNDS 541 TEXT-FIG. 1. Location of the area studied. (4) Middle! Upper Albion. Loose bored concretions were noted in fields on the north side of Msunduzi Pan, east of Ndumu at locality G (Latitude 26° 56' 08" south, Longitude 32° 13' 57" east). They are associated with loose Middle ( ?) and Upper Albian fossils. (5) The Cenomanian jConiacian junction. Locality H, a degraded river clilf on the north bank of the Mzinene, 100 m due north of the farm Belvedere and north-east of Hluhluwe (Latitude 27° 52' 45" south. Longitude 32° 20' 44" east) provides the best section of this contact. The Cenomanian consists of yellow silts with concretions and shell-beds. The highest course of concretions seen is truncated, bored by lithophagids and overlain by a Pterotrigonia shell conglomerate with rare Coniacian Proplacenticeras. THE CONCRETIONS Lithologically, the concretions are nearly all hard, brown-weathering blue-hearted shelly calcite-cemented siltstones and fine sandstones, only rarely containing lenticles of fine conglomerate. Most are less than 30 cm long. Some are cut by calcite veins. Concretion formation seems to have taken place at an early pre-compaction stage in diagenesis, as many included fossils and burrows are uncrushed and undistorted. 542 PALAEONTOLOGY, VOLUME 15 Borings Most concretions are bored on one or several sides (PI. 106, figs. 3-5; PI. 108, figs. 1-2, 5). The borings themselves are from 3 to 25 mm in diameter, and when complete, may be up to 50 mm long, with a constricted aperture. The boring organisms in every case are mytilid bivalves best referred to the Tertiary- Recent genera Lithophaga and Botula, and several species are represented, as might be expected in material ranging from Aptian to Coniacian age (PI. 106, figs, la-c; PI. 108, figs. la-b). Lithophaga is a chemical borer, dissolving substrates by secretion of mucus containing a calcium complexing compound (Jaccarini, Bannister and Micallef 1968; Bromley 1970). It is thus normally restricted to calcareous substrates, as is the case with the present occurrences and other fossil records (i.e. Radwanski 1964, 1965, 1968, 1970; Hecker, Ossipovaand Belskaya 1963; Roniewicz 1970; Holder and Hollmann 1969; Purser 1969; Warme and Marshall 1969; Bromley 1970, all with bibliographies). Yonge (1955) has, however, shown that some mytilids may bore mechanically. In some cases, traces of a calcareous lining to the boring is preserved (PI. 106, fig. 3, C). There can be no doubt that these are indeed borings, for grains and shell-fragments in concretions are truncated against the sides of the bore, as in Jurassic examples described by Purser (1969). Borings are usually most densely developed (up to 1400 m^) on the upper surfaces of concretions as they are found (PI. 106, fig. 3; PI. 108, fig. 2), and in some cases actual bioerosion due to intensive attack can be recognized. Some concretions also have borings on their edges and marginal regions of undersurfaces, showing that they stood proud of the sea-floor; yet others are bored all over, and have quite definitely been rolled and overturned. Orientation of borings varies from normal to the surface to highly inclined. Many borings are incomplete; the constricted aperture may be missing, or only the rounded basal termination may remain, whilst in many cases, the bivalve has been washed out. This points to quite extensive abrasion of concretion surfaces after boring (PI. 108, fig. 2, A). Bore fillings. The fillings of borings also give evidence of quite complex post-boring history. Where bored concretions have been enveloped in a second, later course of con- cretions, some have been left as voids which developed a subsequent fill of sparry calcite. EXPLANATION OF PLATE 106 Fig. la-c. Botula' sp. from bored concretions at the Albian/Aptian junction, locality D, Mfongosi Spruit, northern Zululand, X 1-5, BMNH LL 27575. Fig. 2. Proliserpula sp. encrusting a concretion from the Lower Albian at locality C, Mlambongwenya Spruit, northern Zululand, BMNH A 102802. Fig. 3. Upper surface of bored concretion from the same horizon and locality as fig. 1. The surface shows numerous part-eroded Lithophaga crypts; some, as at C, retain traces of calcareous lining. A small caryophyllid coral is at B, whilst an arcid at A nestles in a vacated lithophagid crypt. There are also encrusting serpulids and oysters, BMNH LL 27574. Figs. 4, 5. Vertical sections of borings in concretions from the same horizon and locality as fig. 2. Both show sections of surface oxidation zones. In 4, A and 5, A zones follow the outline of borings. In 5, B both boring and fill are cut, BMNH LL 27580, 27582. Bar scales are 5 mm. Palaeontology, Vol. 15 PLATE 106 KENNEDY and KLINGER, Hardgrounds j ( KENNEDY AND KLINGER: HIATUS CONCRETIONS AND HARDGROUNDS 543 In other cases, the bore is sediment-filled whilst the lithophagid has a coarsely crystal- line calcite fill. The nature of the sediment infilling of borings within the same concretion may also vary. At locality H, the Cenomanian/Coniacian contact, some borings are filled with silt (PI. 108, fig. 2, B) whilst others are filled by the overlying conglomerate (PI. 108, fig. 2, C). Cross-cutting relations point to two phases of filling (and perhaps burial and re-exhumation?), silt pre-dating conglomerate fill. Secondary inhabitants. Bore infilling has not been entirely passive. In some cases, borings are crusted and lined by serpulids, oysters, and in one case a bryozoan. In some Lower Albian occurrences, eroded bores have been occupied by small non-boring bivalves (PI. 107, fig. 3, A; PI. 108, fig. 3). These show taxodont hinges, bear radial and concentric ornament, may have flared concentric ribs, and appear to be a species of Barbatia (Acar). These later inhabitants were presumably byssate nestlers/crevice dwellers like their recent counterparts (Kauffman 1969; Stanley 1970). Traces of yet a third type of secondary inhabitant of borings are seen in some Upper Aptian concretions. Borings are stuffed with ovoid faecal pellets 1 mm diameter and 2 mm long. These might reasonably be interpreted as bivalve pseudofaeces, but for the fact that some lie in cylindrical vermiform burrows within the sediment filling the borings ; they seem more likely to be traces of polychaetes which lived in vacated bores. Secondary inhabitation of borings is, of course, well-known in Recent environments; Evans (1967) lists no less than 30 species which nestle in vacated Penitella penita bores, half of them potentially fossilisable, and there is a wide literature on the subject. Fossil nestlers have also been reported by several authors: Addicot (1963) and Radwanski (1970), pi. 2, fig. a, for instance, whilst a nestling arcid is figured by Masuda (1968, pi. 39, fig. 6) from Miocene Pholadidea borings in andesites. Marginal weathering. Borings show some interesting relationships to the marginal oxida- tion zones of concretions. They cut the weathered zone and thus post-date it (PI. 108, fig. 2) in some cases, whilst in others, the zone traces the outline of the bore, post-dating excavation, but pre-dating filling (PI. 106, figs. 4, 5, A). Yet others cut both bore and filling (PI. 106, fig. 5, B), again stressing the complex burial history of concretions. Discussion. Apart from lithophagids, there is a singular lack of other sorts of boring organisms (i.e. PL 108, fig. 2). In this respect, the South African examples match certain European Middle Jurassic occurrences, but differ greatly from the diverse boring associa- tions which Bromley (1967, 1968, 1970), Voigt (1959, 1968), Radwanski (1970), and others have documented. Lithophagids tend to occur in high densities (Radwanski 1970), and this, or narrowly defined environmental conditions, may have discouraged other borers. Epizoans Three groups occur commonly as cemented epizoans ; bivalves, serpulid polychaetes and corals. To a degree, the presence of hard substrates is also reflected in the faunas of supra-adjacent shell beds, and these are discussed as ‘others’. Bivalves. Two genera are represented : an Exogyra and an Ostrea. Both occur in pro- fusion, plastering the surfaces of concretions, often to the virtual exclusion of other groups (text-fig. 2). 544 PALAEONTOLOGY, VOLUME 15 Exogyra occur at every growth stage from spat to 20-30 mm individuals: only lower, attached valves are preserved (PI, 107, fig. 3). Ostrea. Again vary from spat to individuals several centimetres long, occasionally with both valves (PI. 108, fig. 2). Shape may be strongly influenced by substrate morphology: (i.e. xenomorphic, as defined by Stenzel, Krause and Twining 1957). Some are bored by bryozoans. Serpulids. Two types of serpulid occur. A large form with a tube up to 7 mm in diameter an early planispiral coil, and a later irregular portion, is referred to the genus ProU- serpula (PI. 106, fig. 2; PI. 107, fig. 1). A smaller form, with a 1 to T5 mm tube, coiled irregularly, meandering across concretions and forming felted coverings is referred to SpiroserpuJa (PI, 107, fig. 1). Corals. A small caryophyllid hexacoral occurs sparingly (PI. 106, fig. 3; PI. 108, fig. 4). Distribution. Epizoans are common on concretions at the Albian/Aptian junction, and in the Lower Albian above. They also occur on Aptian concretions, but none have been noted at the Cenomanian/Coniacian junction. Associations are usually single-species dominated. Oysters plaster upper surfaces, and are well-developed on sides of con- cretions. They occur only sparingly on undersurfaces. Serpulids do occur on upper surfaces, but are commonest on sides and overhung edges. They are usually the only epizoans on the undersides of many concretions, and also liked living in borings. The few corals occur on upper surfaces and sides. These features suggest that settlement was influenced by two main factors; presence of adults, and surface orientation. This is in keeping with what we know of larval EXPLANATION OF PLATE 107 Epizoans on surfaces of concretions from the Lower Albian at locality C, Mlambongwenya Spruit, northern Zululand. Fig. 1. ProHserpula and Spiroserpitla, BMNH A 102801. Fig. 2. Ostrea and serpulids, BMNH LL 27581. Fig. 3. Exogyra, BMNH LL 27577. Bar scales are 5 mm. EXPLANATION OF PLATE 108 Figs, la-b, 2. Vertical sections of bored concretions at the Cenomanian/Coniacian junction, from locality H, on the north bank of the Mzinene River north-east of Hluwluwle, Zululand. Figs, la-b show internal and external moulds of Eithoplmga', and one specimen retaining shell. Fig. 2 shows the Cenomanian siltstone concretion below, and Coniacian shell conglomerate above. Note the lack of all but Eithophaga' borings; an eroded boring at A, a silt-filled boring at B, cut by a later, conglomerate-filled boring at C. A section of an in-situ Eithopliaga' is indicated by D, BMNH LL 27583-5. Fig. 3. Arcid bivalve nestling in eroded Eithophaga' crypt. Lower Albian, locality C, Mlambongwenya Spruit, northern Zululand, BMNH LL 27579. Fig. 4. Encrusting caryophyllid coral. Horizon and locality as for fig. 3, BMNH LL 27579. Fig. 5. Calcite-lined Eitbophaga' boring, from the Upper Aptian of locality A, Mfongosi Spruit, northern Zululand, BMNH LL 27573. Bar scales are 5 mm. Palaeontology, Vol. 15 PLATE 107 KENNEDY and KLINGER, Hardgrounds Palaeontology, Vol. 15 PLATE 108 KENNEDY and KLINGER, Hardgrounds KENNEDY AND KLINGER: HIATUS CONCRETIONS AND HARDGROUNDS 545 TEXT-FIG. 2. A, lower surface; B, upper surface; C, side of Ostrea encrusted concretion from the Lower Albian at locality C, Mlambongwenya Spruit, northern Zululand. Note occurrence of oysters on all surfaces. 8MNH LL 27581, X 1 approx. settlement in the living organisms (Johnson 1964) and matches some other fossil occur- rences (e.g. Hallam 1969, fig. 4); although markedly different from others (e.g. Cope 1968). Other epizoans. Shell beds above bored concretions at several localities contain bi- valves which might have utilized them for attachment, in particular large Gervillella. These also occur in other shell beds, where they presumably utilized dead shells for 546 PALAEONTOLOGY, VOLUME 15 attachment. At Mlambongwenya Spruit, however, brachiopods occur in the matrix of bored concretions in the Lower Albian. Since brachiopods are otherwise rare in the Zululand Cretaceous, their presence in all probability reflects the unusual substrate conditions. CONCLUSIONS Sedimentary history. Uncrushed fossils and burrows show concretion formation to have been an early diagenetic event, probably occurring only a little distance below the sedi- ment— water interface. After formation, erosion has removed unconsolidated sediment and exposed concretions on the sea floor to form a hard substrate — a discontinuous hardground; they have then been encrusted and subjected to the boring activities of lithdomous bivalves. In some cases, concretions have been winnowed completely free of sediment, so that epizoans extend over the sides and base, or have been flipped over, exposing the underside to attack by borers. Following this, the history of a concretion may follow one of several directions — simple re-burial, incorporation into a later course of concretions, or re-exhumation and renewed boring, as recorded in a series of differing sedimentary fills to successive generations of borings. Permutations of these and other processes produce a whole series of types of bored and encrusted concretion, as we have tried to summarize in text-fig. 3. Palaeoecology. From our descriptions and figures the following conclusions emerge : 1 . 'Lithophaga" preferred boring in the upper surfaces and sides of concretions. 2. Oysters prefer the tops and sides of concretions but also grow beneath them. 3. Serpulids prefer sides and lower surfaces. 4. Secondary inhabitation of borings is widespread, with recognizable traces of arcid bivalves, serpulids, bryozoans, and perhaps polychaete worms. 5. Encrustation usually post-dates boring. 6. Hard substrates provided attachment sites for byssate bivalves and brachiopods now found in the overlying sediments. Bathymetry. The general facies association of the marine Zululand Cretaceous suggests shallow water, with depths probably never greater than a hundred metres. Hardgrounds and bored horizons represent higher energy and probably shallower water episodes than the rest of the succession. Turner and Boss (1962) show that Lithophaga occurs most abundantly intertidally or in depths of a few metres. It is a common down to 10 m and more, and occurs only occasionally at greater depths, with dead shells recorded from water depths of up to 250 m. The bored horizons described here thus indicate depths of perhaps only a few metres, and probably no more than 10-30 m. Chronology. From the biostratigraphic data available at present, it is clear that the con- cretion formation, excavation and re-burial history of the bored concretion horizons took place in some cases within only part of a single ammonite subzone, a period of the order of a few hundred thousand years (Upper Aptian and Lower Albian occurrences). On the other hand, at least one occurrence (the Cenomanian/Coniacian contact) represents the duration of more than a stage: several million years. TEXT-FIG. 3. Burial-exhumation history of bored concretions. Case 1 : normal concretions. Case 2 : exhumed and bored, with or without transport. Case 3 : re-exhumation, with complex boring history. Case 4: incorporation into a second concretion layer. Case 5 : exhumation, renewed attack and formation of ‘hiatus concretions’. 548 PALAEONTOLOGY, VOLUME 15 Stratigraphic significance. Bored concretion horizons are a few tens of centimetres in thickness in a sequence of over a kilometre of rather uniform sedimentary facies. Location and recognition of these horizons provides an explanation of the long suspected absence of Lower Albian, uppermost Cenomanian and Turonian sediments in this part of South Africa. They also indicate (by virtue of their wide extent) tectonic and erosional events on a regional scale. Acknowledgements. Field-work in South Africa for Kennedy was made possibly by a grant from the Trustees of the late Sir Henry Strakosch, and through the assistance of the staff of Union Corpora- tion (Johannesburg). The assistanee of Professor L. C. King and his colleagues at Durban is gratefully aeknowledged. We are indebted to Dr. H. W. Ball (British Museum, Nat. Hist.) and Mr. P. J. Rossouw (South African Survey) for their help and encouragement, and also thank N. J. Morris, J. D. Taylor, S. Ware, B. W. Sellwood, A. Hallam, E. G. Kauffman, W. S. McKerrow and A. Kendall. This paper is published under copyright authority 4591 of 19-10-1971 of the Government printer of the Republic of South Africa. REFERENCES ADDicoT, w. o. 1963. An unusual occurrence of Tresiis nuttali (Conrad 1837) (Mollusca: Pelecypoda). Veliger, 5, 143-145. BROMLEY, R. G. 1967. Some observations on burrows of thalassinidean Crustacea in ehalk hardgrounds. Quart. J. geol. Soc. Land. 123, 157-182. • 1968. Burrows and borings in hardgrounds. Meddl. dansk. geol. Foren. 18, 247-250. 1970. Borings as trace-fossils and Entobia cretacea Portlock as an example. Geol. J. Special Issue 3, 49-90. COPE, J. c. w. 1968. Epizoic oysters on Kimmeridgian ammonites. Palaeontology, 11, 16-18. DAVEY, R. J. 1969. Some dinoflagellate cysts from the Upper Cretaceous of northern Natal, South Africa. Palaeont. Afr. 12, 1-23, 4 pis. EVANS, J. w. 1967. Relationship between Penitella penita (Conrad 1837) and other organisms of the rocky shore. Veliger, 10, 148-151. HALLAM, A. 1969. A pyritiscd limestone hardground in the Lower Jurassic of Dorset (England). Sediment ology, 12, 231-240. HAUGHTON, s. H. 1936. Account of the geology of the Cretaceous beds of northern Zululand. Ann. S. Afr. Mas. 31, 283-295. HECKER, R. TH., ossiPOVA, A. I. and BELSKAYA, T. N. 1963. Fergana gulf of Palaeogene sea of Central Asia, its history, sediments, fauna and flora, their environment and evolution. Bull. Am. Ass. Petrol. Geol. 47, 617-631. HOLDER, H. and HOLLMANN, R. 1969. Bohrgange mariner organismen in jurassischen hart-und Fels- boden. N. Jb. Geol. Paldont. Abb. 133, 79-88. JACCARiNi, V., BANNISTER, w. H., and MicALLEF, N. 1968. The palHal glands and rock boring in Litho- phaga litopliaga (Lamellibranchia, Mytilidae). J. Zool. Loud. 154, 397-401. JOHNSON, R. G. 1964. The community approach to Palaeoecology, pp. 107-134. In imbrie, j. and NEWELL, N. D., Approaches to Palaeoecology. Wiley, New York. KAUFFMAN, E. G. 1969. In MOORE, R. c. (ed.). Treatise on Invertebrate Paleontology, Part N (1) Bivalvia, N129-N205. Univ. Kansas Press and Geol. Soc. Amer. KENNEDY, w. J. and KLINGER, H. c. 1971. A major intra-Cretaceous unconformity in eastern South Africa. /. geol. Soc. 127, 183-186. MASUDA, K. 1968. Sandpipes penetrating igneous rocks in the environs of Sendai, Japan. Trans. Proc. Palaeont. Soc. Japan, N.s. 72, 351-362. PIENAAR, R. N. 1969. Upper Cretaceous calcareous nannoplankton from Zululand, South Africa. Palaeont. Afr. 12, 75-128, 11 pis. PURSER, B. H. 1969. Syn-sedimentary marine lithification of Middle Jurassic limestones in the Paris Basin. Sedimentology, 12, 205-230. KENNEDY AND KLINGER: HIATUS CONCRETIONS AND HARDGROUNDS 549 RADWANSKi, A. 1964. Boring animals in Miocene littoral environments of southern Poland. Bull. Acad. pal. Sci. Ser. Sci geol. geogr. 12, 57-62, 6 pis. 1965. Additional remarks on Miocene littoral structures of southern Poland. Ibid. 13, 167-173, 4 pis. 1968. Tortonian cliff deposits at Zahorska Bystrica near Bratislava (Southern Slovakia). Ibid. 16, 97. 1970. Dependence of rock-borers and borrowers on the environmental conditions within the Toronian littoral zone of southern Poland. Geol. J. Special Issue 3, 371-390. RONiEWicz, p. 1970. Borings and burrows in the Eocene littoral deposits of the Tatra Mountains, Poland, Ibid. 439-446. STANLEY, s. M. 1970. Relation of shell form to life habits of the Bivalvia (Mollusca). Mem. Geol. Soc. Amer. 125. STENZEL, H. B., KRAUSE, E. K. and TWINING, J. T. 1957. Pelecypoda from the type locality of the Stone City beds (Middle Eocene) of Texas. Gniv. Texas Bur. Ecoii. Geol. Publ. 5704, 237 pp. 22 pi. TURNER, R. D. and BOSS, K. J. 1962. The genus Litliophaga in the Western Atlantic. Jolmsoniana, 1, 81-115. VOIGT, E. 1968. fiber Hiatus-Konkretionen (dargestellt an Biespielen aus dem Lias). Geol. Rundschau, 58, 281-296. WARME, J. E. and MARSHALL, N. F. 1969. Marine borers in calcareous terrigenous rocks of the Pacific Coast. Am. Zoologist, 9, 765-774. YONGE, c. M. 1955. Adaption to rock-boring in Botula and Litliophaga (Lamellibranchiata, Mytilidae) with a discussion of the evolution of the habit. Quart. J. micr. Sci. 96, 383-410. W. J. KENNEDY Department of Geology and Mineralogy Parks Road Oxford 0X1 3 PR H. C. KLINGER Geological Survey of South Africa Private Box 112 Pretoria Republic of South Africa Typescript received 10 August 1971 A LOWER CARBONIFEROUS CONODONT FAUNA FROM CHILLATON, SOUTHWEST DEVONSHIRE by S. C. MATTHEWS, P. M. SADLER and E. B. SELWOOD Abstract. Siliceous shale in the Lower Carboniferous at Chillaton, Devonshire, has abundant moulds of conodonts. The relatively rare genus DoUymae is represented by the species D. hassi, which has previously been found only in the upper part of the German anchoralis-Zonc. The presence of this form might seem to give a precise indication of the age of the Chillaton fauna. However, the three primary indices recommended for the anchomlis-Zone by Voges are missing. It is therefore necessary to take account of information from Texas and Belgium, where DoUymae species (although, so far, not D. hassi) are known to occur before the first appearance of Scaliognathus anchoralis. It emerges that the new fauna from Devonshire has much in common with what has been found in Texas, but rather less in common with what has been reported from Belgium. The Chillaton fauna is regarded, for the present, as being approximately of anchoralis-Zone age. The systematic section deals principally with DoUymae and GnathoUiis. Present information on the genus DoUymae is reviewed. Hass’ and Voges’ information is included, with a corrected rendering of assumptions on the orientation of this conodont. In GnathoUus, a wide variety of form is referred to G. pimctatus. It is suggested that particular variants of G. piaictatus might have been the sources from which particular species of DoUymae were derived. Lower Carboniferous conodont faunas attributable to the German anchoralis-Zono, (Bischoff 1957; Voges 1959, 1960) have a wide distribution in Europe, North Africa, and North America. The record runs from Austria (Fliigel and Ziegler 1957; Schulze 1968) and Czechoslovakia (Zikmundova 1967; Friakova 1968; Conil, Dvorak and Freyer 1971) to North Africa (Remack- Petitot 1960), Spain (Ziegler 1959; Higgins, Wagner-Gentis and Wagner 1964; Budinger 1965; van Adrichem Boogaert 1967; Marks and Wensink 1970), Portugal (van den Boogard 1963), France (Remack-Petitot 1960; Pelhate 1969), Belgium (Conil, Lys and Mauvier 1964; Conil, Austin, Lys and Rhodes 1969; Groessens 1971), England (Matthews 1961, 1969u, 19696; Morris 1970) and Ireland (Hill 1971). In North America there are reports of Scaliognathus anchoralis from Missouri and Oklahoma (Branson and Mehl 1941), Missouri (Thompson 1967), Missouri and Arkansas (Thompson and Fellows 1970), Texas (Hass 1959) and New Mexico (Burton 1964). Meischner (1971) has recently reviewed the succession of conodont faunas found in the Carboniferous of Germany. He has been able to confirm the essentials of Voges’ reading of the early Carboniferous sequence of forms and he supplies numerous observations on anchoralis-Zont faunas. Meischner has found that the genus DoUymae is confined to an upper part of the anchoralis-ZonQ, as Voges (1959, table 1) and Boger (1962) had already indicated. This would seem immediately to provide a basis for dating a conodont fauna recently discovered in the Lower Carboniferous of the south-western part of Devonshire. The new fauna has DoUymae in association with abundant gnatho- dids, and includes a few representatives of the genus Siphonodella (S. cremdata and S. obsoleta). These last, according to the German evidence (Voges 1959 and Meischner 1971 rather than Boger 1962), range upward into the higher parts of the anchoralis- Zone. [Palaeontology, Vol. 15, Part 4, 1972, pp. 550-568, pis. 109-111.] MATTHEWS ET AL.: LOWER CARBONIFEROUS CONODONTS 551 THE COMPOSITION OF THE CHILLATON CONODONT FAUNA The conodonts were found (by E.B.S.) as moulds on fine, hard siliceous shale in Marlow’s Quarry (SX 4349 8178), at Chillaton, which lies about 8 km north-west of Tavistock. Preservation and preparation are exactly as described in Matthews (1969a, 1969Z)). The forms identified (from latex pulls of the two surfaces produced by parting a single bedding-plane) are: Dollymae hassi Voges. Gnathodus delicatiis Branson and Mehl. Guathodiis pimctatiis (Cooper). Gnathodus semiglaber Bischoff. Polygnathus communis communis Branson and Melil. Polygnathus communis carina Hass. Pseudopolygnathus triangulus Voges subsp. indet. Siphonodella crenulata (Cooper). Siphonodella obsoieta Hass. Spathognathodiis cf. stabilis (Branson and Mehl). Chitinophosphatic brachiopods. Over 16,000 moulds (representing half that number of conodonts) are seen on approxi- mately 800 cm‘^ of rock surface. Over 80% of the total number of conodonts are bar types. They are omitted from the list above because they have no significance in an estimation of the age of the fauna. Their distribution on the rock-surface gives no direct suggestion of the presence of assemblages. A later communication will deal with the form-relationships and relative abundance of the bar-types available on this and other surfaces in the Chillaton siliceous shale. German evidence, as mentioned above, would suggest that a fauna of this composition belongs in an upper part of the aiichoralis-Zone. It is therefore surprising that Scaliog- nathus anchoralis, Hindeodella segaformis and Doliognathus latus, the three distinctive forms nominated by Voges as indices to the anchoralis-Zone, are not represented. Meischner (1971, p. 1176) offers one possible explanation of such a case. He remarks that in Germany a distinction can be made between basin-associated aiichoralis-ZonQ faunas (with S. anchorolis, siphonodellids and polygnathids related to P. inomatus) and Schwelle-associated faunas (gnathodids very much dominant and S. anchoralis rare or even absent, in which case the attribution to the anchoralfs-ZonQ may not be entirely straightforward). The Chillaton fauna, which has abundant gnathodids, few siphono- dellids and no Scaliognathus anchoralis nor Polygnathus inornatus, shows some resem- blance to those of Schwelle type, a suggestion which could draw support from work now in progress on the stratigraphy of the Tavistock-Launceston area. Work in progress in Germany should show whether Dollymae is consistently more abundant on Schwellen sites. If this proves to be so, the distinction Meischner makes could perhaps be restated as one in which the basinal faunas have a relatively high number of forms with restricted basal cavities (polygnathids, Scaliognathus) and the Schwellen faunas a greater propor- tion of forms with widely flared basal excavations (gnathodids, Dollymae). While these matters are under investigation, it would be well to consider any alter- native explanation of the absence of the three anchoralis-Zone, indices from this fauna. It is, for example, necessary to note the fact that Dollymae (although not, as yet, any example of D. hassi) is known to occur in Texas (Hass 1959) and in Belgium oo C 9202 552 PALAEONTOLOGY, VOLUME 15 (Groessens 1971) before the first emergence of S. anchoralis. Both of these cases deserve to be examined here for any evidence of comparability with the Chillaton occurrence. In the Chappel Limestone of Texas, Hass (1959; table 1, collections 9307, 15569, 15570, 15581, 9377) found Dollymae sagittula in his Gmthodus pimctatm Zone. Forms such as ScaJiognathus anchoralis and Doliognathus excavatus (which is close to Doliog- nathus latus — see Voges 1959, p. 275, and Thompson, 1967, p. 34), first appear in his Bactroguathus communis Zone above. Hass encountered some Devonian and even Ordovician forms in the Chappel Limestone. These obviously indicate reworking, and it has been suggested that the process of reworking might also have effected some rearrange- ment of the Mississippian conodonts present. Thompson and Fellows (1970, p. 60), who are of this opinion, mention the thinness of the Chappel Limestone, and suggest that deposition was slow and recycling of conodont material common. Hass’s sampling, according to Thompson and Fellows, tended to lump together faunas which they them- selves would claim to have separated by detailed sampling of sections elsewhere. Hass’s (1959) paper provides for a response to these criticisms. First, it can be observed (e.g. in table 1 of Hass 1959) that Hass’s samples were closely spaced, and that the collections he made from his samples show evidence of relatively coherent associations of forms: one notes, for example, the series of G. punctatus specimens, all from collection 9301, illustrated by Hass (1959, pi. 47, figs. 11-18). Or, taking the samples that produced Dollymae sagittula, one finds that its associates are present in fairly regular numerical proportions. Among these samples one finds cases (9037, 15569, and 15570) in which the number of pre-Chappel forms is quite small. Further, it would be reasonable to consider the fact that Lindstrom (1964, p. 97, reporting the work of A. J. Scott) was able to refer to meaningful gradations of form in conodonts recovered from a single sample collected in the (presumably) upper part of the Chappel Limestone. The view taken here is that although there is clear evidence that pre-Mississippian conodonts were reworked into the Chappel Limestone, there may not have been any considerable re-arrangement of the Mississippian conodonts themselves. Hass’s Gnathodus punctatus Zone faunas com- pare well with what is found at Chillaton. Abundant G. punctatus and G. delicatus are common to the two cases (there are, however, some minor differences between the two sets of G. punctatus — see below). Both have siphonodellids and P. communis carina. The resemblance is close, and yet it is Dollymae sagittula that appears in Texas, and D. liassi in Devonshire. The Belgian evidence which should be considered here comes from Groessens (1971) who has found yet another species of Dollymae, D. bouckaerti (a relatively simple form, which Groessens takes to be the same as Voges’ Dollymae sp. B) in the late Tournaisian (Tn 3c). Immediately above, the conodonts of Groessens’ Scaliognathus anchoralis- Hindeodella segaformis Assemblage Zone make their appearance. Groessens’ first report of his findings (Groessens 1971) seems to offer little for comparison with the Chillaton evidence. D. bouckaerti and D. liassi are quite different. Groessens makes no mention of G. punctatus. Siphonodellids occur at Chillaton, but Groessens would regard these as having met extinction in Tn 3a of the Belgian succession. The Belgian pseudo- polygnathids, which are different from those found at Chillaton, appear to have more in common with pseudopolygnathids found in the German anchoralis-ZonQ. The only distinctive form common to all these occurrences — Texas, Germany, Belgium, Devon- shire— is Polygnathus communis carina. MATTHEWS ET AL.: LOWER CARBONIFEROUS CONODONTS 553 There is one further record of DoUymae to be mentioned. Boyer, Krylatov, Le Fevre and Stoppel (1968, fig. 8) show their sample SK 260+CE 154 to include D. hassi and Protognathodus kockeli as well as numerous other forms. They refer this fauna to a high Gattendorfia-Stage horizon. Obviously, this particular record requires re-examination. Summarizing this discussion, one would say that until now DoUymae hassi has been encountered only in the late anchoralis-ZonQ of Germany (with the exception of the puzzling French case mentioned above). There are, however, records from Texas and from Belgium which show that other forms of DoUymae can occur before S. anchoraUs. The Chillaton fauna has much in common with the Texas case, but a great deal less in common with what has been reported from Belgium. The age of the Chillaton fauna may be taken, for the present, to be approximately in the range of the German anchoralis- ZonQ. Future inquiries may hope to show whether different modes of the genus DoUymae could have been generated at slightly different times. Any such inquiry may derive some assistance from the review of the present state of information on the genus which is included below, and which corrects a conspicuous error that exists in much of the descriptive material so far published. SYSTEMATIC NOTES Numbers prefixed BU refer to the collections in the Geology Museum, University of Bristol. Each five-figure number identifies one surface of a rock specimen. Suffixes to a five-figure number locate particular moulds present on that surface. It will be understood that two different numbers, each with its suffix, may refer to two aspects of a single conodont. The illustrations show latex (‘Revultex’) pulls dusted with ammonium chloride. Deeper parts of the moulds (e.g. the crest of the blade in the mould of an oral surface) of these small fossils will often test the pull technique to its limits — any local incompleteness of particular specimens as seen in the illustrations should be assumed to be due to this cause rather than taken as evidence of abrasion of the conodont. Complete counts of specimens are given only for DoUymae, Siphoiiodella and Spathognathodas. Total numbers of the other forms (gnathodids, polygnathids) will be supplied when the bar-type conodonts have been studied on this and other surfaces in the siliceous shale. Any count of individual Gnathodus ‘species’ will involve numerous decisions on the specific identity of the many ‘transitional’ forms available. The synonymy lists carry some of the signs proposed by R. Richter {Einfidmmg in die zoologische Nomenklatiir. Kramer Verlag, Frankfurt-a-M. (2nd edition), 1948). These signs are intended to indicate the different levels of confidence with which an author might insert items in his synonymy lists. They are widely used in German language publications. Genus dollymae F[ass 1959 1959 DoUymae gen. nov. Hass, p. 394. 1959 DoUyjnae Hass; Voges, p. 275. 1964 DoUymae Hass; Lindstrom, p. 168. Remarks. Finds of the genus DoUymae are recorded in papers by Hass (1959), Voges (1959), Boger (1962), Ziegler (1963), Krebs (1968), Boyer et al. (1968) and Groessens (1971). Groessens (1971) mentions that further discoveries have been made in Belgium and Ireland. Hass’s (1959) brief first description of the genus (restated in Hass 1962) is, apart from what appears in Lindstrom’s book of 1964, the sole systematic statement on the genus in the English language. Voges (1959) made much fuller reference to DoUymae, and 554 PALAEONTOLOGY, VOLUME 15 brought three new forms, Dollymae hassi, DoUymae sp. A and DoUymae sp. B, to join DoUymae sagittida, the single species Hass had proposed. Boger (1962) would have established DoUymae sp. B of Voges as the species Dollymae vogesi; but, as Ziegler (1963) and Conil and Paproth (1968) have already pointed out, his Dollymae vogesi must be regarded as a nomen nudum. Groessens (1971) has now proposed that Dollymae sp. B of Voges be absorbed in his own Dollymae boiickaerti. Scott, Ellison, Rexroad, and Ziegler (1962) have called attention to the fact that the conventions on the orientation of conodonts (especially the sense of the terms ‘anterior’ and ‘posterior’) employed by Hass differ from those used by the majority of conodont workers. Hass’s system of orientation appears in his descriptive references (1959, 1962) to Dollymae. It is perhaps not widely realized that Voges (1959) followed the Hass scheme of orientation when referring to German occurrences of this genus (lapsing into ‘normality’ on one single occasion — Voges 1959, p. 275) although taking the more con- ventional course in all the rest of his systematic descriptions. One finds, therefore, that all of the descriptive references to the genus Dollymae in the present literature, with the exception of Lindstrom’s(1964) brief note and Groessens’s (1971) relatively brief descrip- tion (in French) of D. bouckaerti, have a sense of the terms anterior and posterior that is the reverse of what is usually accepted in work on conodonts. It may be of some ser- vice to ofter here a summary of current information on the genus, with Voges’ (1959, pp. 275-277) observations rendered into English (see passages headed ‘Translation’) and with the terms anterior and posterior now taken as they are normally understood. Corrections of this kind, inserted by the translator (S.C.M.), are square bracketed in the translated sections. The genus Dollymae (Hass 1959, p. 394) has the form of an inverted cup, whose upper (i.e. the oral) surface bears a blade-carina and two subsidiary carinae. The free blade is situated anteriorly. It extends in carina form along the cup-surface and may project, spike-like, at the posterior margin. The blade-carina is slightly curved and is regarded as being convex toward the outer side. The outer portion of the cup is wider than the inner. The blade-carina and the two antero-laterally directed subsidiary carinae diverge from the posterior part of the oral surface to give a clear impression of sagittate (arrow-like) form. A radial carina may be developed within the angle between the outer subsidiary carina and the blade. The broadly excavated aboral surface shows its maximum vertical dimension at a point which lies near the posterior end and which corresponds with the point of convergence of the carinae on the oral surface. Four distinct forms of Dollymae have been recognized. They are: 1. DoUymae sagittula Hass (Hass 1959, p. 394; pi. 47, figs. 7, 10): Description. Sagittate, slightly asymmetrical unit. Near-straight blade is free anteriorly and bears denticles that are either erect or posteriorly directed. Blade-denticles fuse with one another along the oral surface of the cup to give a narrow carina. Relatively large terminal denticle of carina projects at posterior end. Two subsidiary carinae, the outer slightly curved, the inner almost straight, each bearing a single row of fused denticles. Crest line of denticles is highest near mid-length of each subsidiary carina and becomes lower near confluence with posterior part of blade-carina, whose distinctly large terminal denticle is of the same character as denticles on subsidiary carinae. Oral surface of cup smooth in areas away from carinae. Excavated aboral side bears grooves corresponding to courses of the oral surface’s carinae. 2. DoUymae hassi Voges 1959 (PI. 33, figs. 5-10 of Voges 1959, holotype, Vo 59/4, shown there in figures 5, 6); MATTHEWS ET AL.: LOWER CARBONIFEROUS CONODONTS 555 Translation. Diagnosis, a species of the genus DoUymae with parapet-like or nodose-ridged subsidiary carinae and with a radial carina (on the outer side). The upper surface of the blade bears two rows of nodes. In adult specimens the [posterior] margin of the cup has strengthening in the style of the carinae. Description of the liolotype. The arcuate cup makes a right angle with the slightly bent blade as it crosses it. The outer portion of the cup is the larger. The subsidiary carinae on the crest of the cup are low and discontinuously developed in the holotype, and arise from a point situated slightly [in front of] the [posterior] end of the ridge-like blade. The angle between blade and subsidiary carina is acute on the inner side and almost reaches a right angle on the outer side. It is divided by the radial carina, whose upper margin is nodose. The cup has a concave [anterior] margin on the inner side. The outer [anterior] margin is divided into two embayed parts by the projecting free termination of the radial carina. The lateral margins of the cup are restricted, the [posterior] margin broadly rounded. Near the free terminations of the subsidiary carinae the [posterior] margin is strengthened by a wavy parapet. Short ridges reach [forward] from this parapet to bring about an almost complete ornamentation of the oral surface of the cup. The nodes developed in two rows along the oral surface of the blade are arranged in pairs with linking low ridges. On the cup the nodes fuse, forming a ridge which reaches a short way beyond the point of origin of the two subsidiary carinae but which fails to reach the thickened [posterior] margin of the cup. The conodont is broadly excavated aborally. Greatest depth is found below the [posterior] termina- tion of the blade. The courses of the two subsidiary carinae, the radial carina and the blade are indicated by grooves which originate from the point of greatest depth. Toward the [anterior] end of the blade the sides of the excavation converge to produce narrower, trench-like form. Juvenile specimens. Here again the aligned nodes on the oral surface of the blade are paired and fuse in the [posterior] part of their course to give a ridge. The subsidiary carinae, the radial carina and (if present) the parapet-like thickening at the [posterior] margin of the cup are made up of simple transverse ridges or discontinuous series of nodes. The angle between subsidiary carina and blade is acute on the inner side and approximately right on the outer. The outline and the excavation of the aboral surface are essentially as given for the holotype. Relationships. The deep excavation of the aboral side, the grooves below the subsidiary carinae and blade and the crudely arcuate arrangement of the subsidiary carinae at the [posterior] end of the blade are characteristic of the genus DoUymae. In this species the spike at the [posterior] end of the blade is stunted. The arrow-like shape and the difference in ornament separate D. sagittula from D. liassi. The forms DoUymae sp. A and DoUymae sp. B are distinct chiefly by their lack of a radial carina. DoUymae was probably derived from Scaliognathns. 3. DoUymae sp. A (Plate 33, figs. 11-14 of Voges 1959). Translation. Description. The cup has an arcuate [posterior] margin and slightly concave to convex inner and outer [anterior] margins. The outer portion of the cup is the larger. The two subsidiary carinae lie in a curve which is sited close to the [posterior] margin of the cup. They originate from a point slightly [in front of] the [posterior] end of the ridged blade. The carinae are simple, ridge-like, or (in large specimens) have nodes and transverse ridges. The oral edge of the blade always bears a row of denticles. These are fused with one another almost to their free terminations, where they can be seen to be of circular or oval cross-section. Set lower, on either side of the oral edge of the free blade, there are rows of irregular nodes. In one specimen they appear in simple ridged form, and in another (smaller) they are not yet developed. On the cup, the teeth of the oral edge of the blade are fused to produce a ridge which goes beyond the point of origin of the subsidiary carinae and projects at the [posterior] margin in the form of a spike. The aboral surface of the conodont is excavated and the course of the blade and of the subsidiary carinae are indicated by grooves whose courses converge at the deepest point of the aboral surface, situated below a point near the [posterior] end of the ridged blade. The free blade may be grooved, or merely slit, along its length. 556 PALAEONTOLOGY, VOLUME 15 Relationships. To Doiiyinae hassi, see under that species. The arrow-like shape agrees with that of Doiiymae sagittuia, but the upper surface ornaments are not at all comparable. For the distinction from Doiiymae sp. B, see below, under ‘Description’. 4. Doiiymae sp. B. (pi. 33, figs. 15-17 of Voges 1959), now referred to D. bouckaerti by Groessens (1971, p. 14, pi. 1, figs. 6-8). Transiation. Description. This form shows strong resemblance to Doiiymae sp. A, but the free blade carries only one row of denticles and the subsidiary carinae have a middle position on the cup. These carinae, in large specimens, are made up of nodes and transverse ridges. They arise near the [posterior] end of the ridged blade but not necessarily both from the same point. The angle between subsidiary carina and blade is obtuse to right on the outer side and right to acute on the inner. The curved blade, which in its free part bears a row of fused teeth, goes over into ridged form on the cup and extends spike- like beyond the [posterior] margin of the cup. The small number of specimens available does not provide for a definitive statement on the form of the cup, but there does seem to be a tendency toward relative narrowness of the two lateral portions of the cup. A middle line drawn through these would run oblique to the blade. The aboral surface of the conodont is excavated. The deepest point is found where the transverse depressions running below the subsidiary carinae meet the groove coming from the aboral margin of the blade. The free blade is cut by a slit along part of its length only. Reiationships. To Doiiymae hassi see above. For the distinction from Doiiymae sagittuia, the state- ments made under Doiiymae sp. A would apply again. Groessens, as noted above, has referred Doiiymae sp. B to his new species D. bouckaerti. But his description of the new species supplies less detail than Voges offered for Doiiymae sp. B, so it is still worthwhile to refer to the Voges observations. One finds, for example, that Groessens (1971, p. 14) makes no mention of the siting of the subsidiary carinae (Voges specified a medial position on the lateral extensions of the cup), nor does he make the point that these carinae need not originate both from exactly the same point on the axis of the conodont. On the other hand, Groessens has noted distal bifurcation of the subsidiary carinae, and Voges made no mention of any such feature. The two descriptions clash in what they specify for the form of the blade: bent according to Voges, straight according to Groessens. Groessens’s illustration of the holotype of D. bouckaerti does indicate a resemblance to Voges’s Doiiymae sp. B, but his diagnosis and description are less precise than one would wish. Since Groessens and Voges have different views on the stratigraphic level (relative to the first appearance of Scaliognathus anchoralis) at which their representatives of Doiiymae emerge it is particularly important that the degree of resemblance of Doiiymae sp. B and Doiiymae bouckaerti, complete or otherwise, should be clearly documented. The literature carries occasional comments on relationships between Doiiymae and other forms of conodont. Hass (1959, p. 394) noted a superficial resemblance to Ancyro- della. Voges (1959, p. 276) briefly remarked that Doiiymae was probably derived from Scaliognathus. Lindstrom (1970) tentatively referred Doiiymae to his family Bactrog- nathidae, thus associating it with Bactrognathus, Doliognathus, Scaliognathus and Staurognathus. An interesting observation made by Groessens (1971) is that immature stages of his D. bouckaerti can be distinguished from his Spathognathus bultyncki only by the presence of nodes on the oral lateral surfaces of the cup. Groessens’s suggestion would link Doiiymae to forms whose basal excavation is relatively large and open, rather than to forms with more restricted basal features, such as Scaliognathus. MATTHEWS ET AL.: LOWER CARBONIFEROUS CONODONTS 557 The Chillaton conodonts suggest another possible relationship. Dollymae hassi and Gnathodus pimctalus may be compared in terms of the features of their aboral surfaces. The two surfaces have the same general scheme of topography, but the gnathodid has a relatively well developed posteriorward groove and Dollymae hassi is relatively well developed along laterally directed axes. One axis, directed antero-laterally, represents the course of the radial carina. The suggestion could be rendered in the terms Lindstrom (1964) used to describe the Prioniodus plan, referring to what are here called ‘axes’ as the I TEXT-FIG. 1 . Six major growth vectors identified in Gnathodus punctatus (drawn from BU 22088/21, cf. PI. 1 10, fig. 13) and in Dollymae hassi (drawn from BU 22090/3, cf. PI. 109, fig. 1). Anterior lateral process and inner lateral flare identified in G. punctatus as sug- gested by Lindstrom (1964). Note that the inner lateral flare appears on what is by normal convention the outer side of the conodont. Anterior lateral processes (vertical lines) and inner lateral flares (extra outer contour) indicated following the scheme of ornamenta- tion used by Lindstrom (1964, fig. 33). branchings of what Lindstrom called the inner lateral flare and the anterior lateral process. Text-figure 1 attempts to identify these features in the two forms. It should be said that there is more than their common adherence to the Prioniodus plan to suggest a relationship linking G. punctatus and D. hassi. There is some strati- graphic evidence that the two might be in some way associated (note the present case, in which G. punctatus is the most common gnathodid, and also Voges’ faunas 30 and 32 in which relative abundance of G. punctatus coincides with relative, if much less im- pressive, abundance of D. hassi), and there is a clear resemblance to be seen between details of ornament found here in G. punctatus (nodes, and less common short ridges) and details found in the first group of variants of D. hassi described below. Groessens (1971) has suggested that his Dollymae bouckaerti might have been derived from a spathognathodid. The suggestion here is that Dollymae hassi might have been 558 PALAEONTOLOGY, VOLUME 15 derived from a gnathodid. It should not be thought that these are two conflicting views on the origin of the ‘genus’ Dollyniae. Both might be valid. If both are valid, this, in turn, need not be taken to mean that the ‘genus’ Dollyniae is diphyletic. In the systematic section below, the discussion of G. pimctatus points to the possibility of yet another inde- pendent production of DoUymae. The source would appear to be again G. punctatus, but this time a G. punctatus in which the postero-laterally directed radial feature of oral surface ornament (arranged along the number 6 vector shown in text-figure 1) is relatively well developed. The associated DoUymae would be in this case the ‘species’ D. sagittula, whose aboral character (see Hass 1959, p. 47, fig. 7) seems to resemble that of the local variety (i.e. number 6 vector prominent) of G. pimctatus (see Hass 1959, pi. 47, fig. 18). If the resemblance is a genuine one, it would seem to follow that in D. sagittula the ‘inner lateral flare’ is relatively well developed, whereas in D. Iiassi growth seems to have favoured the ‘anterior lateral process’. These are proposals which should be checked when a greater abundance of DoUymae material is available. It will be necessary to consider both ‘left’ and ‘right’ forms (shown by Hass 1959, pi. 47 — note that his fig. 10 and fig. 7 refer to two different specimens— by Voges 1959, pi. 33 and here on Plate 109) and to discover whether accelerated development of an anterior lateral process, as opposed to an inner lateral flare, might even involve departures from total mirror-image symmetry in left and right forms. DoUymae Jiassi Voges 1959 Plate 109, figs. 1-4, 6-10, 12; text-fig. 1 v*1959 DoUymae Iiassi Voges, pp. 275-276, pi. 33, figs. 5-10 Material. BU 22088/1, 19, 22, 26, 27; BU 22089/1 ; BU 22090/3, 8, 30 (all figured). BU 22088/17, 20; BU 22090/5, 22, 23, 24, 46 (not figured). Remarks. It is convenient to refer to a number of variants seen here; 1. Some forms have discrete, rather punctate ornament found on the posterior part of the oral surface of the cup. The crestal features of the subsidiary carinae are more continuously developed. The terminal (i.e. most posteriorly situated) single node of the main carina is relatively large (e.g. PI. 109, fig. 4). These are the specimens which may be compared with Gnathodus punctatus as discussed above. 2. Certain other specimens, whose cup is more widely extended in the lateral sense, show ridge-like development in all of the crestal features of the oral surface, especially in the crest situated near the posterior margin. The ridge-like development may involve lateral mergings of the elements of a more punctate scheme of ornament. The angle EXPLANATION OF PLATE 109 Revultex pulls dusted with ammonium chloride. All X30. Figs. 1-4, 6-10, 12. DoUymae hassi Voges. 1, 4 (BU 22090/3, BU 22088/1) are aboral and oral views of one conodont. 8, 9 (BU 22088/27, BU 22090/30) are aboral and oral views of one conodont. 2, BU 22088/22. 3, BU 22090/8. 6, BU 22088/26. 7, BU 22089/1. 10, BU 22088/19. 12, BU 22088/17. Figs. 5, 13. Gnathodus punctatus (Cooper). 5 (BU 22090/25) inner lateral view. 12 (BU 22090/27) oral view. Fig. 11. Spathognathodus cf. stabilis (Branson and Mehl). BU 22090/10. Palaeontology, Vol. 15 PLATE 109 MATTHEWS, SADLER and SELWOOD, Carboniferous conodonts -¥ I ;:W»f ( MATTHEWS ET AL.: LOWER CARBONIFEROUS CONODONTS 559 between radial carina and outer subsidiary carina is here smaller than that between radial carina and main carina (PI. 109, fig. 9). 3. Certain forms show a closer approach to bilateral symmetry (but clearly do not achieve this). The outer subsidiary carina may follow a slightly curved (convex- posteriorward) course. Ornament is relatively delicate here and elements of the low crest situated near the posterior margin become discrete, ridge-like, and fade as they run towards the crests of the subsidiary carinae (PI. 109, fig. 3). 4. One small specimen shows near-continuous development of the posteriorly situated ridge and the carinal crests (PI. 109, fig. 12). All of these have radial carinae, and should therefore be referred to D. hassi rather than to any other described species of the genus. The same conclusion is suggested by their lack of any conspicuous, spike-like posterior projection, although group 1 forms do show a local slight bulging of the posterior margin, and in group 3 the fine ridge situated near the posterior margin migrates toward that margin (and may possibly over- ride it) in the neighbourhood of the length-axis of the conodont. The most robust fonn found here (PI. 109, fig. 2) is one that does not easily fall into any of the four informal groupings suggested above. Features of both group 1 (conspicuous single node) and of group 3 (posterior ridge migration toward posterior margin in neighbourhood of length- axis) can be seen; possibly a better understanding of interrelationships between these different groups (accounting for ontogenetic variation perhaps) might dispose of this apparent anomaly. None of the specimens in the Chillaton fauna shows the yet more robust ornament found in the holotype. Genus gnathodus Pander 1855 GualJiodus delicatus Branson and Mehl 1938 Plate 110, figs. 5, 7, 8, 9 *1938 Gnathodus delicatus Branson and Mehl; 145, pi. 34, figs. 25-27. 1963 Gnathodus delicatus Branson and Mehl; Ziegler, 327, pi. 2, figs. 5, 7, 9, 12, 14 (?figs. 8, 1 3 = G. punctatus). 1965 Gnathodus delicatus Branson and Mehl; Budinger, 56-57, pi. 2, figs. 9-13. 1968 Gnathodus delicatus Branson and Mehl; Canis, 74, fig. 7 only (fig. 8 = G. punctatiisl). vl969 Gnathodus delicatus Branson and Mehl; Rhodes, Austin and Druce, 97-98, pi. 30, figs. 6a-c only (non pi. 18, figs. 12a-d == G. punctatus). vl969 Gnathodus delicatus Branson and Mehl; Matthews (1969n), 267, pi. 46, fig. 4 (with synonymy). vl969 Gnathodus delicatus Branson and Mehl; Matthews (1969Z)), pi. 51, fig. 7. 1969 Gnathodus delicatus Branson and Mehl; Rexroad, 18-19, pi. 4, fig. 1. 1970 Gnathodus delicatus Branson and Mehl; Marks and Wensink, 261-262, pi. 3, figs. 8, 9, 1 1. 1970 Gnathodus delicatus Branson and Mehl; Thompson and Fellows, 85, pi. 1, figs. 14, 17, 18 only (non pi. 2, figs. 1-5 = G. punctatus). Figured specimens. BU 22088/27; BU 22090/28, 47, 49. Remarks. In G. delicatus the parapet found alongside the carina on the inner oral surface of the cup runs from a point near the posterior end and becomes slightly broader in a short anteriorly situated segment of its length. The broader outer oral surface bears a line of nodes beside a long posteriorly situated segment of the Carina’s course. The remainder of the outer oral surface may carry further nodes. These specifications are 560 PALAEONTOLOGY, VOLUME 15 met by the majority of the forms referred to G. delicatus here (see examples in PI. 110, figs. 5, 7), but one individual deserves special comment. The specimen (PI. 110, fig. 9) is relatively slim and carries its parapet and outer line of nodes high on either side of the Carina. It seems to bear some resemblance to G. ciineiformis. Ziegler (1963, pi. 2, figs. 5, 12) has figured specimens which are transitional between G. delicatus and G. cuueiformis; but the present individual shows a clear broadening of its inner parapet at a point situated near the anterior end, and this is taken to suggest an affinity with G. delicatus. Meischner (1971, fig. 2: "G. cf. cuueiformis") has sketched a somewhat similar case, and has suggested it to be related to G. punctatus and G. delicatus. The relationship between G. delicatus and G. punctatus is evident in the specimens figured on plate 2 of Ziegler (1963). Marks and Wensink (1970) have noted transitions from G. delicatus to G. cuueiformis and from G. delicatus to G. punctatus in their Spanish material. See below for further observations on the G. delicatus-G. punctatus transition. Matthews (1969a) suggested that Thompson’s (1967) G. sp. cf. G. bilineatus might be referred to G. deli- catus. The suggestion could be made again for G. sp. cf. G. bilineatus as figured by Thompson and Fellows (1970), and might apply also to the G. cf. G. bilineatus men- tioned in Thompson, Ford and Sweet (1971, 707). *1939 1959 71963 1965 1967 71967 1968 v 71 969 V.1969 v.1969 vl969 vl969 1970 Gnathodus punctatus (Cooper 1939) Plate 109, figs. 5, 13; Plate 110, figs. 1^, 11-15; text-fig. 1 Dryphenotus punctatus Cooper, 386, pi. 41, figs. 42, 43; pi. 42, figs. 10, 11. Gnathodus punctatus (Cooper); Hass, 395, pi. 47, 11-18 (7also G. delicatus, pi. 48, fig. 4 only). Gnathodus delicatus Branson and Mehl; Ziegler, pi. 2, figs. 8, 13 only. Gnathodus punctatus (Cooper); Budinger, 58-59 (with synonymy). Gnathodus punctatus (Cooper); Thompson, 40-41, pi. 5, figs. 12-15. Gnathodus n. sp. B Thompson; 43, pi. 4, figs. 1-4. Gnathodus punctatus (Cooper); Canis, 538, pi. 74, fig. 21. Gnathodus punctatus (Cooper); Rhodes, Austin and Druce, 105-106, pi. 18, figs, la-c, lOa-lld. Gnathodus delicatus Branson and Mehl; Rhodes, Austin and Druce, pi. 18, figs. 12a-b only (pi. 30, figs. 6a-c = G. delicatus). Gnathodus bilineatus (Roundy) transitional from G. punctatus (Cooper); Rhodes, Austin and Druce, pi. 30, fig. 18. Gnathodus punctatus (Cooper); Matthews (1969a), 267-268; pi. 46, fig. 2. Gnathodus punctatus (Cooper); Matthews (19696), pi. 51, fig. 12. Gnathodus cf. G. punctatus (Cooper); Marks and Wensink, 263, pi. 3, fig. 10. EXPLANATION OF PLATE 110 Revultex pulls dusted with ammonium chloride. All X 30. Figs. 1-4, 11-15. Gnathodus punctatus (Cooper). 1, 2(BU 22090/44, BU 22089/15) are oral and aboral views of one conodont. Similarly with 3, 4 (BU 22090/19 aboral and BU 22088/12 oral) and 13, 14 (BU 22088/21 aboral and BU 22090/1 3 oral). 11,BU 22090/21. 1 2, BU 22088/23. 15, BU 22088/6. Figs. 5, 7, 8, 9. Gnathodus delicatus Branson and Mehl. 5, BU 22088/27. 7, BU 22090/49. 8, BU 22090/28. 9, BU 22090/47. Fig. 6. Gnathodus sp. juv. BU 22088/2. Fig. 10. Gnathodus seiniglaber BischofT. BU 22090/20. Palaeontology, Vol. 15 PLATE 110 MATTHEWS, SADLER and SELWOOD, Carboniferous conodonts MATTHEWS ET AL.: LOWER CARBONIFEROUS CONODONTS 561 1970 Gnathodiis pwictatus (Cooper); Thompson and Fellows, 86-87, pi. 1, figs. 15, 16, 19; pi. 2, figs. 14-17. .1970 Gnathodiis delicatiis Branson and Mehl; Thompson and Fellows, pi. 2, figs. 1, 5 only (pi. 1, figs. 14, 17, 18 = C. delicatiis). Figured specimens. BU 22088/6, 12, 21; BU 22089/15; BU 22090/13, 19, 21, 25, Ti, 44. Remarks. A wide range of form is referred to G. pwictatus here. Particular variants are: 1. Forms transitional to G. delicatiis: note especially the specimen figured on PI. 109, fig. 13, which has all of the characters specified above for G. delicatiis (with each indi- vidual feature now more robustly developed) plus here a rudimentary, curved (convex toward the carina) inner parapet. 2. Relatively small forms which do not have the distinctively curved inner parapet of G. pwictatus but which do have more than one row of nodes on that side (PI. 1 10, fig. 11): Voges (1959, p. 284) referred specimens of this kind to G. pwictatus and Matthews (1969a, pi. 46, fig. 2) followed suit. Two further cases recorded in the literature might deserve the same interpretation. They are Gnathodiis n. sp. B of Thompson (1967) — ^but note that Thompson and Fellows (1970, pp. 90-91, pi. 3, figs. 11, 15) would now refer a rather wider range of form, including some relatively poorly ornamented individuals, to G. sp. B — ^and Gnathodiis cf. G. pwictatus of Marks and Wensink (1970). Marks and Wensink suggest that their forms resemble those of Rhodes, Austin and Druce and also one figured by Ziegler (1963, pi. 2, fig. 4). Neither suggestion seems particularly apt. A comparison with Ziegler (1963, pi. 2, figs. 8, 13) would be better. 3. Forms which have well developed ornament closely adjacent to the posterior part of the carina on either side (PI. 1 10, figs. 1, 2, 15). Ziegler’s (1963, pi. 2, fig. 4) specimen might be better compared with these. This ornament does not merge with the carina and this provides a means of distinguishing these forms from G. semiglaber even where the curved inner parapet is not conspicuously well developed. The specimen figured on PI. 110, fig. 15 is of this kind. So, too, perhaps is Burton’s (1964, table) G. bilineatus, which Thompson and Fellows (1970, p. 87) would refer to G. semiglaber. G. semiglaber of Canis (1968, pi. 74, fig. 5) is again vaguely of this character. 4. A single specimen which shows much stronger resemblance to G. semiglaber is illustrated on PI. 1 10, figs. 3, 4. The form of the cup, and its relatively poorly ornamented upper surface, would clearly suggest G. semiglaber-, but the growth of the outer side shows (PI. 1 10, fig. 3) a radial effect strongly developed towards the posterolateral angle, and the posterior part of the carina has nothing of the thickening commonly found in G. semiglaber. The specimen is not greatly different from one referred to G. semiglaber by Thompson and Fellows (1970, pi. 2, figs. 7-10). 5. Specimens unequivocally referable to G. pwictatus have on their inner oral side a short curved parapet which is convex toward the carina and on their outer oral side a broad noded surface. Only one of the present specimens (PI. 110, fig. 12) has a parapet set clearly apart from the carina in the manner seen in some of the specimens figured by Hass (1959: e.g. his pi. 47, figs. 14, 15, 17). In the present material one more commonly finds the condition shown in Hass’s (1959), pi. 47, fig. 16, where nodes other than those of the parapet are present on the inner oral surface. There is a further difference between G. punctatus as figured by Hass and G. pwictatus as found here: Hass’s specimens show a clear radial (along a line bisecting the angle between the carina and the anterior margin of the outer side of the cup) effect in the ornament of the outer oral surface of the 562 PALAEONTOLOGY, VOLUME 15 cup. No such effect is evident in the Chillaton specimens, although a similarly directed effect is plainly seen in the growth-lines of the aboral surfaces. It was suggested above that the Chappel Limestone form of G. punctatus, with this strong radial element, may be associated with the D. sagittuJa mode of DoUymae, and the Chillaton form, lacking that feature, linked instead with the D. hassi mode. It is interesting to observe that Voges (1959, p. 284) has noted the absence of the radial feature of outer oral surface ornamentation from his specimens of G. punctatus. He found G. punctatus to be especially common in his faunas 30 and 32, both of which have produced D. hassi. Gnathodus semiglabev Bischoff 1957 Plate 1 10, fig. 10 v*1957 vl959 1964 1965 1967 1967 1968 V non 1969 v?1969 1970 1970 Gnathodus bilineatus semiglaber Bischoff, 22, pi. 3, figs. 1-10, 12-14 (8-10, 12-14 are juveniles according to Bischoff). Gnathodus semiglaber (Bischoff); Voges, 284, pi. 33, tigs. 38, 39. Gnathodus semiglaber (Bischoff); Rexroad and Scott, 30, pi. 2, figs. 1, 2. Gnathodus semiglaber Bischoff; Budinger, 59-60, pi. 1, figs. 14-20; pi. 3, figs. 1, 4-6. (with synonymy). Gnathodus semiglaber (Bischoff); Thompson, 41, pi. 4, figs. 11-14. Gnathodus semiglaber (Bischoff); van Adrichem Boogaert, 179-180, pi. 2, tig. 20 only. Gnathodus semiglaber (Bischoff); Canis, 538, pi. 74, tig. 19 only (tig. 5 = G. punctatusl). Gnathodus semiglaber Bischoff; Rhodes, Austin, and Druce, 106-107, pi. 30, fig. 1 (= G. delicatiisl) Gnathodus antetexanus Rexroad and Scott; Rhodes, Austin, and Druce, 93-94, pi. 18, figs. 13a-d only. Gnathodus semiglaber Bischoff; Marks and Wensink, 264, pi. 3, figs. 19, 20. Gnathodus semiglaber Bishoff; Thompson and Fellows, 87, pi. 2, figs. 2-4, 7, 10. Material. BU 22090/20 (figured). Remarks. The discussion of G. punctatus (above) shows that the difference between G. punctatus and G. semiglaber is not dear cut (see, especially, group 3 and 4 there). Rhodes, Austin and Druce (1969, pi. 30, figs. 2, 8) have figured forms which they regard as transitional between G. punctatus and G. semiglaber. However, the specimen shown in their figure 8 appears to have more in common with the gnathodids attributed to G. typicus by Thompson (pi. 4, figs. 5, 7, 8, 10) and later transferred to G. antetexanus by Thompson and Fellows (1970). The single Chillaton specimen referred to G. semi- glaber shows some resemblance to G. punctatus in the character of its outer platform, but has a posterior carina of a kind more common in G. semiglaber. Its inner oral surface, although much narrower than is typical of G. semiglaber, carries a very brief parapet of a kind that suggests the character of G. semiglaber and not at all that of G. punctatus. This brief parapet shows some resemblance to the relatively widely developed anterior part of the parapet seen in specimens referred here to G. delicatus. Some recent authors (e.g. Rhodes, Austin, and Druce 1969; Thompson and Fellows 1970) have followed Rexroad and Scott (1964) in attributing Mehl and Thomas’s (1947) specimen of G. per plexus to G. semiglaber. A specimen of that character (see Mehl and Thomas 1947, pi. 1, fig. 4) would be identified as G. delicatus here (see also Rexroad 1969, synonymy list on p. 18). MATTHEWS ET AL.: LOWER CARBONIFEROUS CONODONTS 563 Genus polygnathus Hinde 1879 Polygnathus communis Branson and Mehl 1934 Remarks. The nominate subspecies is known to range from the Famennian {styriacus- Zone according to Ziegler 1962, 1971; or even earlier according to recent American information in Klapper et al. 1971; see also van Adrichem Boogaert 1967) into the Dinantian. It is abundant in the Chillaton fauna. So, too, is the subspecies P. communis Carina (treated below). A single small polygnathid (PI. 3, fig. 12) has the general form of P. communis but is distinct in having a line of nodes arranged parallel to the carina on either side of the platform. Hass (1959, pi. 49, fig. 11) has figured a specimen of P. communis which is relatively well provided with nodose ornament, although there the nodes run wider on either side to affect the form of the platform margins. Druce's (1969) P. communis dentatus may be of similar character, with the nodose effect confined to the anterior parts of the platform. Polygnathus communis carina Hass 1959 Plate 111, figs. 6, 7, 13 *1959 Polygnathus comtmmis var. carina Hass, 391, pi. 47, figs. 8, 9. vl959 Polygnathus communis carina Hass; Voges, 289, pi. 34, figs. 5, 6. .1963 Polygnathus communis Branson and Mehl; Ziegler, pi. 1, figs. 5, 6 only. 1964 Polygnathus communis carina Hass; Rexroad and Scott, 34, pi. 2, figs. 24, 25. .1965 Polygnathus communis Branson and Mehl; Budinger, pi. 1, figs. 12, 13 only. 1967 Polygnathus communis carina Hass; Thompson, 45, pi. 2, figs. 2, 10; pi. 4, figs. 6, 9. 1968 Polygnathus communis carina Hass; van Adrichem Boogaert, 184, pi. 2, figs. 43a, b. 1968 Polygnathus communis carina Hass; Canis, 544, pi. 72, figs. 18-20. 1968 Polygnathus communis var. carina Hass; Manzoni, 666-667, pi. 62, figs. 2, 3. 1969 Polygnathus communis carinus Hass; Druce, 95, pi. 18, figs. 12a-c. 1970 Polygnathus communis carinus Hass; Thompson and Fellows, 92-93, pi. 3, fig. 14. Figured specimens. BU 22088/11, 25; BU 22090/14, 18, 45. Nomenclatural note. Species-group names in the form of adjectives in the nominative singular are required to agree in gender with the generic name with which they are combined (ICZN Article 30). The sub-species name carina, however, is a noun in the nominative singular (ICZN Article 11 g i 2), and is therefore not subject to any such requirement. Remarks. The present material shows all transitions between P. communis communis and P. communis carina (compare remarks in Voges 1959, pp. 289-290). Voges en- countered P. communis carina in significant numbers in his faunas 30 and 32. In the Chillaton fauna the carinate ornament at the anterior end of the platform appears to consist, on the inner side, of fine transverse ridges (up to 3 in number) rather than a transverse arrangement of nodes. Hass (1959, pi. 47, fig. 8) has illustrated a comparable case. It can further be observed in the present material that the basal cavity is situated at the blade-platform junction in smaller specimens but lies enveloped in the platform growth of more mature specimens, at which stage it appears to be of relatively small size. Budinger (1965) has made similar observations for P. communis. Druce’s (1969, p. 94) suggestion, that in the P. communis group the basal cavity is at the blade- platform junction, is inexact. Cooper’s (1939) pre-Welden shale conodonts may include 564 PALAEONTOLOGY, VOLUME 15 P. communis carina (see, for example, Cooper’s 1939, pi. 39, figs. 1, 2, 9, 10, 23, 24, 33-36, all of which were referred to P. communis communis by Rexroad and Scott 1964). Genus pseudopolygnathus Branson and Mehl 1934 Pseudopolygnathus triangulus Voges subsp. indet. Plate 111, figs. 14-18 Figured specimens. BU 22088/9; BU 22089/8, 13; BU 22090/11, 32. Remarks. A number of the pseudopolygnathids encountered in conodont faunas of approximately this age have platforms broader anteriorly than those of Ps. inulti- striatus but less broad and less straight at their anterior margins than is common in Ps. triangulus pinnalus. Also, they lack the pinnate development of the inner antero- lateral margin characteristic of the latter form (Voges 1959). Pseudopolygnathids of this apparently intermediate kind are seen in Hass (1959: 'Pseudopolygnathus asymmetrica Cooper’) and in Thompson and Fellows (1970: 'Pseudopolygnathus triangulus pinnatus Voges’). Thompson and Fellows’s specimens, like the one figured by Thompson in 1967, pi. 4, figs. 17, 18 (note the more satisfactorily pinnate character here) are relatively small and have relatively large basal cavities. A larger specimen figured by Rexroad and Scott (1964, pi. 2, fig. 28: 'Pseudopolygnathus triangukd) shows a fair degree of resemblance to one form (PI. 1 1 1, fig. 15) encountered here. Ziegler (1963, p. 324, pi. 1, fig. 1) has figured a specimen — 'Pseudopolygnathus triangula subsp. indet. (wahrscheinlich pinnata)' — which has finer ribs than are seen in Rexroad and Scott’s or any of the present specimens, and which may be closer than either of these to Ps. triangulus pinnatus. Ziegler’s specimen and two of the Chillaton forms (PI. 3, figs. 14, 18) have each a relatively restricted basal cavity of the kind seen in Ps. triangulus pinnatus. Genus siphonodella Branson and Mehl 1944 Siphonodella erenulata (Cooper 1939) Plate 111, figs. 1, 11 *1939 Siphonognathiis erenulata Cooper, 409, pi. 41, figs. 1, 2. 1966 Siphonodella erenulata (Cooper); Klapper, 18, pi. 3, figs. 5-8 (with synonymy). EXPLANATION OF PLATE 111 Revultex pulls dusted with ammonium chloride. All x 30. Figs. 1, 11. Siphonodella erenulata (Cooper). 1, BU 22091/2. 11, BU 22088/13. Figs. 2, 3. Siphonodella cf. erenulata (Cooper). Oral (BU 22090/42) and aboral (BU 22089/3) views of one conodont. Figs. 4, 5. Siphonodella obsoleta Hass. Aboral (BU 22088/5) and oral (BU 22090/1) views of one conodont. Fig. 10. Siphonodella cf. obsoleta Hass. BU 22090/31. Figs. 6, 7, 13. Polygnathus communis carina Hass. 6, 7 oral (BU 22088/11) and aboral (BU 22090/14) views of one conodont. 13, BU 22090/45. Figs. 8, 9. Polygnathus communis subsp. Oral (BU 22088/25) and aboral (BU 22090/18) views of one conodont. Fig. 1 2. Polygnathus communis subsp. BU 22088/4. Note nodes on oral surface. Figs. 14-18. Pseudopolygnathus triangulus Voges subsp. indet. 14, 15 aboral (BU 22090/11) and oral (BU 22088/9) views of one conodont. Similarly with 17 (BU 22090/32, oral) and 18 (BU 22089/8, aboral). 16, BU 22089/13. Palaeontology, Vol. 15 PLATE 111 MATTHEWS, SADLER and SELWOOD, Carboniferous conodonts MATTHEWS ET AL.\ LOWER CARBONIFEROUS CONODONTS 565 1968 Siphonodella cremilata (Cooper); Canis, 548, pi. 72, fig. 21. 1969 Siphonodella crenidata (Cooper); Rexroad, 42, pi. 2, figs. 9, 10. 1970 Siphonodella crenidata (Cooper); Thompson and Fellows, 105, pi. 6, figs. 7, 10. Material. BU 22088/13; BU 22091/2 (figured). S. cf. crenidata-. BU 22089/3, BU 22090/42 (figured). BU 22090/29 (not figured). Remarks. A single small unornamented specimen is referred to S. cremilata principally because of the shape of the platform (cf Voges, 1959, pp. 307-309). Further observations on siphonodellids of this type are to be found in Matthews and Butler (in press). Siphonodella obsolete! Hass 1959 Plate 111, figs. 4, 5 *1959 Siphonodella obsoleta Hass, 392-393, pi. 47, figs. 1, 2. 1969 Siphonodella obsoleta Hass; Rhodes, Austin and Druce, 220-221, pi. 12, figs. 13a-c. vl969 Siphonodella obsoleta Hass; Matthews (1969u), 273-274, pi. 46, fig. 1. 1969 Siphonodella obsoleta Hass; Rexroad, 44, pi. 3, figs. 5-7. 71969 Siphonodella obsoleta Hass; Anderson, 924-925, ol. 108, figs. 3-5, pi. 109, figs. 12, 21, 23, 25. 1970 Siphonodella obsoleta Hass; Thompson and Fellows, 107-108, pi. 7, fig. 10; pi. 8, figs. 8, 9 (with synonymy). Material. BU 22088/5, BU 22090/1 (figured). A. cf. obsoleta: BU 22090/31 (figured), BU 22091/1 (not figured). Remarks. The above synonymy takes account of papers published too late to be con- sidered by Thompson and Fellows (1970). Certain of Anderson’s (1969) specimens of S. obsoleta deserve to be checked against Thompson and Fellows’ new S. cooperi hassi. The single specimen referred to S. obsoleta here is, again, not far removed from S. cooperi hassi, but is thought to belong to S. obsoleta because of the character of its outer rostral ridge, which continues, eventually as a line of nodes, to merge with the outer margin in the posterior half of the platform. The specimen will be seen to have the relatively extensive basal feature that occurs fairly frequently among siphonodellids. Genus spathognathodus Branson and Mehl 1941 Spathognathodus cf. stabilis (Branson and Mehl 1934) Plate 109, fig. 11 Material. BU 22090/10 (figured). Remarks. An especially robust single spathognathodid is compared with S. stabilis (Branson and Mehl) as interpreted by Klapper in 1966. Thompson and Fellows (1970) have recently suggested that in S. stabilis the basal cavity continues to the posterior end, whereas it is restricted to the middle third of the conodont in S. macer (Branson and Mehl). The present specimen might therefore better deserve to be compared with S. macer (see also Rexroad 1969, p. 48, and especially fig. 10 on his pi. 6). The decision would be simpler if Thompson and Fellows’ (1970, p. 114) remarks on the dentition of S. stabilis and S. macer were clearer. What may be more important than any of these questions of comparison is the resemblance that is seen here between the crestal features 566 PALAEONTOLOGY, VOLUME 15 and blade-denticle frequency of the spathognathodid (which has a restricted basal cavity) and those same characters in certain of the gnathodids (whose basal cavity is widely expanded and extends to the posterior end). A lateral view of a gnathodid is shown, for comparison, on PI, 109, fig. 5. Acknowledgements. Matthews and Sadler’s work on conodont faunas from south Devon and Cornwall is supported by an N.E.R.C. research grant, which is here gratefully acknowledged. Mr. Malcolm Butler has kindly assisted in discussions of the present paper. The photographic illustrations are the work of Mr. E. W. Seavill. REFERENCES ANDERSON, w. I. 1969. Lower Mississippian conodonts from northern Iowa. J. Paleont. 43, 916-928, pi. 107-109. BiscHOFF, G. 1957. Die Conodonten-Stratigraphie des rheno-herzynischen Unterkarbons mit Beriick- sichtigung der WocklumeriaSlvdQ and der Devon/Karbon-Grenze. Abh. hess. Landesamt. Boden- forsch. 19, 7-64, pi. 1-6. BOOGARD, M. VAN DEN 1963. Conodonts of Upper Devonian and Lower Carboniferous age from southern Portugal. Geologic Mijnb. 42, 248-259, pi. 1. BOOGAERT, H. A. VAN A. 1967. Devonian and Lower Carboniferous conodonts of the Cantabrian Moun- tains (Spain) and their stratigraphic application. Proefschrift, University of Leiden, 129-192, pis. 1-3. BOGER, H. 1962. Zur Stratigraphic des Unterkarbons im Velberter Sattel. Decheniana, 114, 133-170. BOYER, F., KRYLATOV, s., LE f' VRE, J., and STOPPEL, D. 1968. Lc Devonicn superieur et la limite Devono- Carbonifere en Montagne Noire (France). Lithostratigraphie-biostratigraphie (Conodontes). Bull. Cent. Rech. Pan, 2, 5-33. BRANSON, E. B. and MEHL, M. G. 1934. Conodonts from the Grassy Creek Shale of Missouri. Univ. Missouri Studies 8 (1933), 171-259, pis. 13-21. • 1938. Conodonts from the Lower Mississippian of Missouri. Ibid. 13, 128-148, pis. 33, 34. 1941. New and little known Carboniferous conodont genera. /. Paleont. 15, 97-106, pi. 19. BUDINGER, p. 1965. Conodonten aus dem Oberdevon and Karbon des Kantabrischen Gebirges (Nord- spanien). Inauguraldissertation, University of Tubingen. 1-103, pis. 1-5. BURTON, R. c. 1964. A preliminary range chart of Lake Valley (Osage) conodonts in the southern Sacramento Mountains, New Mexico. New Mexico Geological Society, 15th Field Conference, 73-75. CANis, w. F. 1968. Conodonts and biostratigraphy of the Lower Mississippian of Missouri. J. Paleont. 42, 525-555, pis. 72-74. CONIL, R., LYS, M. and MAUViER, A. 1964. Criteres micropaleontologiques essentiels des formations- types du Carbonifere (Dinantien) du bassin franco-beige. C.R. Cong, avanc. etud. stratigr. Carb. (Paris 1963), 1, 325-332. • — ^ — and PAPROTH, E. 1968. Mit Foraminiferen gegliederte Profile aus dem nordwestdeutschen Kohlenkalk und Kulm. Decheniana, 119, 51-94. — — AUSTIN, R. L., LYS, M., and RHODES, F. H. T. 1969. La limite des etages tournaisien et viseen au stratotype de I’assise de Dinant. Bidl. Soc. beige Geol. Paleont. Hydrol. 77, 39-69, pis. 1-2. DVORAK, J. and freyer, g. 1971. Lower Carboniferous from the cement-works quarry near Mokra (southern part of the Moravian Karst). Vest, listred. Ost. geol. 46, 9-18. COOPER, c. L. 1939. Conodonts from a Bushberg-Hannibal horizon in Oklahoma. J. Paleont. 13, 379- 422, pis. 39-47. DRUCE, E. c. 1969. Devonian and Carboniferous conodonts from the Bonaparte Gulf Basin, northern Australia. Bull. Bur. Miner. Resour. Geol. Geophys. Aust. 98, 3-157, pis. 1^3. FLUGEL, H. and ZIEGLER, w. 1957. Die Gliederung des Oberdevons und Unterkarbons am Steinberg westlich von Graz mit Conodonten. Mitt, naturw. Ver. Steierm. 87, 25-60, pis. 1-5. FRiAKOVA, o. 1968. Zprava o vyzkumu konodontove fauny z okoli Hranic na Morave. Zpr. geol. vyzk. 1967, 111-113. GROESSENS, E. 1971. Lcs conodontcs du Tournaisien superieur de la Belgique. Prof. Pap. Serv. geol. Belgique, 4 (1971), 1-29, pis. 1, 2. MATTHEWS ET AL.: LOWER CARBONIFEROUS CONODONTS 567 HASS, w. H. 1959. Conodonts from the Chappel Limestone of Texas. Prof. Pap. U.S. geol. Surv. 294-J, 365-399, pis. 46-50. ■ 1962. In MOORE, R. c. (ed.). Treatise on Invertebrate Paleontology, Part W (Miscellanea), W 3- W 69, Univ. Kansas Press and Geol. Soc. Amer. HIGGINS, A. c., WAGNER-GENTis, c. H. T. and WAGNER, R. H. 1964. Basal Carboniferous strata in parts of Northern Leon, N. W. Spain : stratigraphy, conodont and goniatite faunas. Bull. Soc. beige Geol. Paleont. Hydrol. 72, 205-248, pis. 1-5. HILL, p. 1971. Carboniferous conodonts from southern Ireland. Geol. Mag. 108, 69-71. HiNDE, G. J. 1879. On conodonts from the Chazy and Cincinnati Group of the Cambro-Silurian, and from the Hamilton and Genesee Shale division of the Devonian, in Canada and the United States. Quart. J. geol. Soc. Loud. 35, 351-369. KLAPPER, G. 1966. Upper Devonian and Lower Mississippian conodont zones in Montana, Wyoming and South Dakota. Paleont. Contr. Univ. Kansas, 3, 1-^3, pis. 1-6. ■ SANDBERG, C. A., COLLINSON, C., HUDDLE, J. W., ORR, W. R., RICHARD, L. V., SCHUMACHER, D., SEDDON, G., and uyeno, t. t. 1971. North American Devonian conodont biostratigraphy. Mem. geol. Soc. Am. 127, 285-316. KREBS, w. 1968. Die Lagerungsverhaltnisse des Erdbacher Kalkes (Unterkarbon II) bei Langenaubach- Breitscheid (Rheinisches Schiefergebirge). Geotekt. Forscli. 28, 72-103. LiNDSTROM, M. 1964. Conodonts. Elsevier, London. Amsterdam, New York. 196 pp. ■ 1970. A suprageneric taxonomy of the conodonts. Lethaia, 3, 427-455. MANZONi, M. 1968. II Devoniano superiore e il Carbonifero inferiore nelle serie pelagiche di Val Uqua (Tarvisio). G. Geol. (2) 34 (for 1966), 641-684, pis. 60-63. MARKS, D. and WENSiNK, H. 1970. Conodonts and the age of the ‘Griotte’ limestone formation in the Upper Aragon Valley (Huesca, Spain), I. Koninkl. Nederl. Akad. Wetens. Series B 73, 238-275, pis. 1-4. MATTHEWS, s. c. 1961. A Carboniferous conodont fauna from Callington, east Cornwall. Abstr. Proc. Conf. Geol. Geomorph. S.fV. England, R. Geol. Soc. Cornwall, Penzance 1963, 13-14. 1969fl. A Lower Carboniferous conodont fauna from east Cornwall. Palaeontology, 12, 262- 275, pis. 46-50. 19696. Two conodont faunas from the Lower Carboniferous of Chudleigh, south Devon. Ibid. 12, 276-280, pi. 51. and BUTLER, M. (in press). Siphonodellid conodonts in the Lower Carboniferous. Congres avanc. etiid. stratigr. carb. Krefeld 1971. MEiscHNER, K.-D. 1971. Conodonten-chronologie des deutschen Karbons. C.R. Cong, avanc. etud. stratigr. Carb. (Sheffield 1967), 3, 1169-1180. MORRIS, p. Cr. 1970. Carboniferous conodonts in the south-western Pennines. Geol. Mag. 106, 497-499. PANDER, c. H. 1 856. Monographic der fossilen Fische des silurischen Systems der russich-baltischen Gouvernements. St. Petersburg, Kaiser!. Akad. Wiss., 91 pp. PELHATE, A. 1969. Micropaleontologie des calcaires dinantiens du bassin de Laval. Bull. Soc. geol. mineral. Bretagne, Nouvelle Serie 1967, 27-76, pis. 1-6. REMACK-PETiTOT, M. L. 1960. Contribution a I’etude des conodontes du Sahara (bassins de Fort- Polignac, d’Adrar Reganne et du Jebel Bechar): comparaison avec les Pyrenees et la Montagne Noire. Bull. Soc. geol. Fr. 7 (2), 240-262. REXROAD, c. B. 1969. Conodonts from the Jacobs Chapel Bed (Mississippian) of the New Albany Shale in southern Indiana. Bull. Indiana Dep. Conserv. Geol. Surv. 41, 1-55, pis. 1-9. ■ and SCOTT, a. j. 1964. Conodont zones in the Rockford Limestone and the lower part of the New Albany Shale (Mississippian) in Indiana. Bull. Indiana Dep. Conserv. Geol. Surv. 30, 7-54, pis. 2, 3. 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.), 5, 4-313, pis. 1-31. SCHULZE, R. 1968. Die Conodonten aus dem Palaozoikum der mittleren Karawanken (Seeberggebiet). Neues Jb. Geol. Paldont. Abb. 130, 133-245, pis. 16-20. SCOTT, A. J., ELLISON, s. p., REXROAD, c. B., and ZIEGLER, w. 1962. Comments on the orientation of conodonts. J. Paleont. 36, 1394-1396. C 9202 pp 568 PALAEONTOLOGY, VOLUME 15 THOMPSON, T. L. 1967. Conodont zonation of Lower Osagean rocks (Lower Mississippian) of south- western Missouri. Missouri Geol. Siirv. Wat. Resour. Kept. Invest. 39, 88 pp., pis. 1-6. and FELLOWS, l. d. 1970. Stratigraphy and conodont biostratigraphy of Kinderhookian and Osagean rocks of south-western Missouri and adjacent areas. Missouri Geol. Surv. Wat. Resour. Rept. Invest. 45, 3-263, pis. 1-9. - — — FORD, N. s. and sweet, w. 1971. Conodonts from the Rushville Formation (Mississippian) of Ohio. J. Paleont. 45, 704-712, pi. 83. VOCES, A. 1959. Conodonten aus dem Untercarbon I and II (Gattendorfia- und Pericyclus-Stufe) des Sauerlandes. Palaont. Z. 33, 266-314, pis. 33-35. 1960. Die Bedeutung der Conodonten fur die Stratigraphic des Unterkarbons I und II (Gatten- dorfia- und Pericyclus-Stufe) im Sauerland. Fortschr. Geol. Rheinld. Westf. 3 (1), 197-288. ZIEGLER, w. 1959. Conodonten aus Devon und Unterkarbon Sudwesteuropas und Bemerkungen zur bretonischen Faltung (Montagne Noire, Massiv von Mouthoumet, Span. Pyrenaen). N. Jb. Geol. Palaont. Mh. 1959, 289-309. — — 1962. Taxionomie und Phylogenie oberdevonischer Conodonten und ihre stratigraphische Bedeutung. Abh. Iiess. Landesamt. Bodenforsch. 38, 11-166, pis. 1-14. 1963. Conodonten aus dem Unterkarbon der Bohrung Miinsterland 1. Fortschr. Geol. Rheinld. Westf. 11, 319-328, pis. 1, 2. - — — 1971. Conodont stratigraphy of the European Devonian. Mem. Geol. Soc. Am. 127, 227-284. ziKMUNDOVA, J. 1967. Konodontova zona Scaliognatlius anchor alls Branson and Mehl v ponikevskych bfidlicich Nizkeho Jeseniku. Vest, listred. Ust. geol. 42, 449-452, pis. 1-4. S. C. MATTHEWS, P. M. SADLER Geology Department, University of Bristol, Queen’s Building, University Walk, Bristol, BS8 ITR. E. B. SELWOOD Geology Department, University of Exeter, North Park Road, Exeter. Revised typescript received 16 March 1972 NEW UPPER CARBONIFEROUS CHELICERATA (ARTHROPODA) FROM SOMERSET, ENGLAND by T. AMBROSE and m. romano Abstract. Well preserved arthropods from the Farrington Group (Westphalian D) of the Somerset coalfield are described. Ten specimens of Euproops kUmersdonensis sp. nov., of which seven occur on a single bedding plane, permit a range of intraspecific variation and effects of deformation to be taken into account in erecting a new species. A reconstruction of E. kUmersdonensis does not show the presence of long ophthalmic spines. It resembles the type species, E. danae (Meek and Worthen 1865), differing in the shape of the cardiac lobe and ornament of the opisthosoma. Eophrynus jugatus sp. nov. is the third recorded species of the genus Eophrynus (which is known only in Britain) differing in dorsal and ventral ornament from the type species, Eophrynus prestvici (Buckland 1837). The specimens of Euproops kUmersdonensis sp. nov. and Eophrynus jugatus sp. nov. described in this paper occur in laminated and current bedded grey, silty mudstones with comminuted plant debris associated with the No. 9 coal seam, which is the lowest seam of the Farrington Group in the Somerset coalfield. The Farrington Group is assigned to the tenuis Zone, of Westphalian D age. The material described was found during a field excursion to the Bristol district and was discovered on the mine tip of Kilmersdon Colliery, near Radstock by Messrs. H. Bailey and R. Aldous, students in the Department of Geology, University of Shefireld, who kindly made the material available to the authors. The genus Euproops, Upper Carboniferous to Permian, is widely known from rocks of Westphalian age, and has been recorded from North America and Europe. It is often found associated with fossil arachnids, and is generally regarded as being non-marine. Although specimens of Euproops are fairly common in comparison with numbers of other Xiphosura, the occurrence of seven exoskeletons of one species on a single slab (see Plate 112, fig. 1) is unusual and provides an opportunity for investigating intra- specific variation. Eophrynus jugatus, is, as far as the authors are aware, only the third species of this genus to be described from Britain. Specimens of the first, Eophrynus prestvici (Buckland 1837) have been made available to the authors by Birmingham University and the British Museum. The second, Eophrynus warei (Dix and Pringle 1930) is referred to below. TERMINOLOGY The morphological terms for the Merostomata proposed by Stormer in Moore (1955) have been generally adopted. Some modifications to existing terms are used and other new terms are proposed to facilitate description. Menotes proposed new terms ^denotes modified terms Prosoma Text-fig. 1a Border — Synonymous with that defined by Harrington et at. in Moore (1959, p. 0119). ^Preophthalmic field — Area lying between the anterior border and cardiophthalmic region and limited laterally by the line running anteriorly (exsag.) from the eye to the border. [Palaeontology, Vol. 15, Part 4, 1972, pp. 569-578, pis. 112, 113.] 570 PALAEONTOLOGY, VOLUME 15 ^Cardiac ridge — Median ridge running from the posterior margin of the cardiac lobe to the mid-point of the anterior branches of the ophthalmic ridges. It need not be continuous. ^Cheek — Area of the prosoma lying lateral to the cardiophthalmic region and preophthalmic field, excluding the border and genal spine. Opisthosoma Text-fig. 1b ^Axis — Medial region of the opisthosomal dorsal exoskeleton, bordered by axial furrows. This proposal thus excludes the term cardiac lobe when referring to the opisthosoma. ^Pleural rib — Dorsal transverse segment of the lateral portion of the opisthosoma bounded by the axial furrow and excluding the opisthosomal rim (see below). ^Pleural ridge — Ridge separating adjacent pleural ribs. ^Opisthosomal rim — Flat flange surrounding the central raised portion on all but the anterior margin and crossed by ridges in direct continuation with the pleural ridges. ^Opisthosomal spine — Laterally to posteriorly directed pointed extension of the opisthosomal rim with a ridge along the posterior margin. Class MEROSTOMATA Dana, 1852 Order xiphosurida Latreille, 1802 Family euproopidae Eller, 1938 Genus euproops Meek, 1867 Type species. Euproops danae (Meek and Worthen 1865) Euproops kihnersdonensis sp. nov. Text-fig. 2; Plate 112, figs. 1-3; Plate 113, fig. 1 Diagnosis. Euproops with narrow, flat to roll-like border to prosoma; anterior part of cardiac lobe constricted; very short ophthalmic spines. Axis of opisthosoma with tubercle on first and third segment, short spine on terminal segment. Opisthosomal rim moderately narrow with long rim spines. Telson long. Type material. Elolotype. It. 61012 (PI. 112, fig. 1; PI. 113, fig. 1). Internal and external mould of nearly complete specimen. Internal mould figured. Paratype. It. 61013 (PI. 112, figs. 1-3). Internal and external mould of nearly complete specimen. Internal mould figured. Other figured material. It. 61014, external mould; It. 61015, internal mould; It. 61016, external mould; It. 61017, external mould; It. 61018, internal mould. Counterpart of material figured in Plate 112, fig. 1 is retained in the reference collections of the Department of Geology, University of Sheffield. The specimens figured are in the collections of the British Museum (Natural History) and their numbers are prefixed It. Horizon and locality. No. 9 seam, Farrington Group {tenuis Zone, Westphalian D). Upper Carboni- ferous (Moore and Trueman 1937, p. 228). Mine tip at Kilmersdon Colliery (Grid reference ST 681536) near Radstock, Somerset. Derivation of name. After the colliery at Kilmersdon where the specimens were found. Measurements. Dimensions in mm. (e) denotes estimated. EXPLANATION OF PLATE 112 Fig. 1. Slab with seven specimens oi Euproops kilmersdonensis sp. nov. Specimen 1 Holotype (It. 61012) Internal mould; 2 Paratype (It. 6101 3) Internal mould; 3-7 Not type material. 3 (It. 61014) External mould; 4 (It. 61015) Internal mould; 5 (It. 61016) External mould; 6 (It. 61017) External mould; 7 (It. 61018) Internal mould. Eigs. 2, 3. Euproops kilmersdonensis sp. nov. Paratype (It. 61013). Internal mould. 2, dorsal view of detail of cardiophthalmic region, x4. 3, dorsal view of detail of left side of opisthosoma, X3. Specimens whitened with ammonium chloride before photographing. Palaeontology, Vol. 15 PLATE 112 AMBROSE and ROMANO, Euproops r v> I i ^ TEXT-FIG. 1. Diagram showing measurements taken and new or modified terms for E. kilmersdoneusis sp. nov. Key to Symbols : Ps Prosoma! length measured in the sagittal line from the posterior margin of the cardiac lobe to the anterior margin. Pt Prosomal width measured in a transverse direction along the posterior margin of the cardiac lobe. Cs Length of cardiac lobe measured in sagittal line. Ct Width of cardiac lobe measured in a transverse direction along the posterior margin. ID Interocular distance measured in a transverse direction between the eyes. G 1 Length of genal spine measured along the length of the spine. Os Opisthosomal length measured in the sagittal line from the posterior margiit of the cardiac lobe to the posterior margin of the opisthosoma. Ot Opisthosomal width measured in a transverse direction through the greatest width. At Axial width measured in a transverse direction at the anterior end. Ts Length of telson measured along its length. Abbreviations : C. ridge Cardiac ridge pi. ridge Pleural ridge op. spine Opisthosomal spine pi. rib Pleural rib op. rim Opisthosomal rim 572 PALAEONTOLOGY, VOLUME 15 TABLE 1. Some measurements taken on E. kilmersdonensis sp. nov. For explanation of symbols see key to text-fig. 1 . Specimen Symbol 1 2 3 4 5 6 7 8 9 10 Ps 14 14-5 17 13 12-5 12(e) 10-5 — 10-5 21 Pt 36 48 32 41-5 48 53(e) 48(e) — 34(e) 36-5 Cs 5-5 4-5 6 5-5 5(e) 5-5 6 — 4-5 7(e) Ct 5-5 6-5 4-5 5(e) — — 5 5(e) 4 — ID 15(e) 18 13(e) 17-5(e) 19 22 18-5 ■ — 12(e) 14 G1 + 12 — 15 — + 11 — + 11 — — + 10 Os 12 10-5 14 10-5 11-5 12-5 — 18 11 -5(e) — Ot 20 24 27(e) 25(e) 25 27 22(e) 16 17(e) — At 5-5 6-5 4 6 6-5 7 — 3-5(e) — — . Ts + 17 + 15 — + 10 39 — — — — — TABLE 2. Some ratios of measurements taken on E. kilmersdonensis sp. nov. Ratio Specimen Ps'.Pt Os'.Ot Ps'.Os ID-.Pt 1 1:2-6 1:1-7 1-2:1 1:2-4 2 1:3-3 1:2-3 1-4:1 1:2-7 3 1:1-8 1:1-9 1-2:1 1:2-5 4 1:3-2 1:2-4 1-2:1 1:2-4 5 1:4-6 1:2-2 1-1:1 1:2-5 6 1:4-4 1:2-2 1-0:1 1:2-4 7 1:4-6 — — 1:2-6 8 — 1:0-9 — — 9 1:3-2 1:1-5 0-9:1 1:2-8 10 1:1-7 — 1:2-6 Specimens in Tables 1 and 2 numbered 1-7 refer to those in Plate 112, fig. 1 ; specimens 8-10 from the same mudstone fragment are not figured and are in the reference collections of the Department of Geology, University of Sheffield. Deformation of the sediment may be approximated to a compression of the mudstones parallel to the transverse axis of specimen 3 (Plate 1 12, fig. 1) as length to width ratios show greatest variation in specimens 3 and 6 which lie parallel and at right angles respectively to the direction of elongation. Original right angles are distorted in specimens (1 and 2) lying oblique to this direction. Tire absolute measurements (Table 1) and ratios of measurements taken either along the length or width of the specimens (Table 2) suggest a range of variation consistent with the assemblage being of adults of the same species. The number of specimens available for study does not warrant statistical analysis. Description. The prosoma is approximately semi-circular in outline with a straight margin anterior to the precardiophthalmic field. The prosomal border is narrow, widest {sag.) anteriorly and flat to gently roll-like in cross section. The narrow doublure is widest anteriorly and continues with the border along the genal spine. The cardioph- thalmic region is gently convex dorsally; the narrow {sag) precardiophthalmic field slopes steeply down to anterior border. The cheeks are fairly wide {trans.) and slope gently outwards. There is a narrow {exsag) border and shallow furrow along the posterior margin of the prosoma. AMBROSE AND ROMANO: CARBONIFEROUS CHELICERATA 573 The cardiophthalmic region is delimited by prominent ophthalmic ridges. Anterior to the eyes the ophthalmic ridges form a double arch; posterior to the eyes the ridges are concave outwards and continue backwards to protrude just beyond the posterior { c TEXT-FIG. 2. Reconstruction of Euproops kibnersdonensis sp. nov. based on holotype and paratype. Dorsal view: x2 (length of telson taken from specimen It. 61016, see Plate 112, fig. 1). margin as short spines (less than 1 mm in length). The distance between the ophthalmic spines is slightly less than the interocular distance. The cardiophthalmic region shows considerable wrinkling of the exoskeleton in all specimens but there is a fairly con- tinuous transverse ridge in most which crosses the area anterior to the cardiac lobe and 574 PALAEONTOLOGY, VOLUME 15 just posterior to the eyes, and dies out laterally before reaching the ophthalmic ridges. In a few specimens traces of other ridges within the cardiophthalmic region make an ‘H’ shaped pattern with the transverse ridge described above as the connecting ridge. The cardiac lobe is sub-triangular in outline with a nearly straight anterior margin and approximately as wide as long. The sides of the cardiac lobe converge quite strongly at about midlength and anterior to this the sides are more parallel sided. The cardiac lobe is bounded by a shallow furrow except anteriorly where the cardiac ridge crosses the furrow and continues to the anterior branch of the ophthalmic ridge. The cardiac ridge terminates at a distance of approximately one-third the length of the cardiac lobe from the anterior end and is represented by a small tubercle at the posterior end of the lobe. At about the midlength of the lobe a broad (sag.) shallow transverse furrow may occasionally be seen (PI. 112, fig. 2). The eyes are rarely preserved, and occur as small convex tubercles (0-5 mm across). Long, slightly curved genal spines are directed posteriorly or posterolaterally reaching to the posterior margin of the opisthosoma. The prosomal border continues along the genal spines. The central raised portion of the opisthosoma is oval in outline with maximum width at just under one-third the distance from the anterior margin. The axis is widest anteriorly where it is about the same width as the cardiac lobe along the posterior margin, narrowing slightly posteriorly to the expanded terminal segment. The axis is gently convex {tram.) and bordered by shallow furrows. Five axial rings of about the same length (5'ug.) are developed and separated by shallow furrows. The terminal segment is globular in shape, standing above the axial rings and not quite reaching the opisthosomal rim. The first and third axial rings bear a small median tubercle; the terminal segment has a short centrally placed spine. The pleurae are fairly flat except near margin where they are steeply downturned. In some specimens (PI. 113, fig. 1) a distinct geniculation on the pleurae exists along a line in continuation with the ophthalmic ridges (see Discussion). Up to seven pleural ribs are present, becoming narrower (exsag.) and directed more backwards towards the posterior end. The opisthosomal rim is moderately narrow with seven long, slender, gently curved opisthosomal spines. The more posterior spines are shorter and directed more back- wards. Pleural ridges separating the pleural ribs pass along the posterior edge of the spines (PI. 112, fig. 3). The telson is long, slightly carinate, and reaches to nearly twice the length of the body. The telson is ornamented with numerous fine, discontinuous, wavy ridges running oblique to the length. Discussion. Euproops kilmersdonensis sp. nov. differs from E. rotundatus (Prestwich) in several important characters. The cardiac ridge of E. rotundatus continues along the length of the cardiac lobe to the posterior margin whereas in E. kilmersdonensis the ridge is interrupted by a transverse furrow at about the midlength of the lobe. The major differences however are on the opisthosoma where in E. rotundatus the opisthosomal rim EXPLANATION OF PLATE 1 13 Fig. 1. Euproops kilmerscioneusis sp. nov. Holotype (It. 61012). Internal mould of nearly complete individual. Dorsal view, x2-4. Specimen whitened with ammonium chloride before photographing. Fig. 2. Eophryuus jugatus sp. nov. Holotype (It. 61019). Internal surface of dorsal part of abdomen with part of ventral exoskeleton attached. Ventral view, X 4. Specimen whitened with ammonium chloride before photographing. Palaeontology, Vol. 15 PLATE 113 AMBROSE and ROMANO, Euproops, Eophrynus AMBROSE AND ROMANO; CARBONIFEROUS CHELICERATA 575 is relatively wider and the opisthosomal spines much shorter than in the species from Kilmersdon. According to a reconstruction of E. rotundatus by Stormer {in Moore, 1955, fig. 12) the prosoma does not show a border. However, two syntypes of E. rotundatus (figured in Woodward 1878, pi. XXXI, fig. 5, and Prestwich 1840, pi. 41, fig. 7) have a narrow roll-like border to the prosoma. Stormer’s reconstruction also illustrates the presence of long ophthalmic spines that extend posteriorly to the terminal segment. Woodward though, in his drawing of that species (op. cit. pi. XXXI, fig. 5) does not put in any such spines and an investigation of the syntypes of E. rotundatus suggests that the spines are not present. However, there is in the syntype described by Woodward a geniculation in the opisthosoma which is in line with the opisthosomal ridges (see also in E. kilmersdonensis) and it is this which probably led to the misconception of long ophthalmic spines. Ophthalmic spines are not present in any specimen of E. rotundatus studied by the authors although they do occur in E. orbicularis Van der Heide (1951, p. 62, pi. 8, figs. 1-4). It is possible that the new material available to Stormer did show ophthalmic spines, but he does not state on what material his reconstruction was based. Another discrepancy between Woodward’s description and the reconstruction in Moore (1955) concerns the ornament on the opisthosomal axis. According to Woodward (although not shown too well on his figure) ‘each somite is ornamented on the median line with a single blunt tubercle, a somewhat larger tubercle marks the coalesced and rudimentary abdominal portion’ (op. cit., p. 247) while in Stormer’s reconstruction tubercles are shown only on the first and third segments with an elevated boss or short spine on the hindmost segment. The syntypes show that Woodward was mistaken in stating that each somite was ornamented. The type species for the genus, E. dauae (Meek and Worthen) (Raymond 1944, p. 484; 1945, p. 4, pi. 1, figs. 1, 2, pi. 2, figs. 1, 2: Langford 1963, p. 25, figs. 27-38) a well known and documented species, is common in the Pennsylvanian of the Mazon Creek area, Illinois, U.S.A. The prosomata of E. danae and E. kilmersdonensis are very similar; both show the border, ‘H’ shaped structure or ridges within the cardiophthalmic region and interruption of cardiac ridge on the cardiac lobe (see especially Raymond, 1945, pi. 1, fig. 1). However, the anterior margin of the cardiac lobe is much straighter in the present species (compare the reconstruction of Langford’s 1963, fig. 27a, with the specimen in PI. 112, fig. 2). In Langford’s reconstruction of E. danae ophthalmic spines reach backwards to the posterior end of the fourth axial segment. On examination of figured specimens of E. danae only five segments appear to be present (excluding terminal segment) and on none are opisthosomal spines visible. Differences on the opisthosoma are more marked between the two species. E. danae has a tubercle on each of the axial segments and in most figured specimens the opisthosomal rim is relatively narrower and the spines shorter than in E. kilmersdonensis. Raymond’s species E. tliomsoni (Raymond 1944, p. 486, fig. 1) is stated by that author as being similar to E. danae but ‘relatively more narrow headed’. Langford (op. cit., p. 29) commented on the uncertainty of E. thomsoni as a valid species based on such a character which depends to a large extent on preservation. The authors agree with Langford’s conclusions, especially since the present material shows widely varying length to width ratios depend- ing on the orientation of the specimens in the deformed sediment. E. amiae Woodward (1918, p. 465, figs. 2-4) from the Coal Measures of Glace Bay Mines, Nova Scotia, is difficult to compare with the present species from Woodward’s 576 PALAEONTOLOGY, VOLUME 15 drawings, but it differs mainly in shape of cardiac lobe, narrowing of the axis posteriorly and absence of a swollen terminal segment. Dix and Pringle (1929) recorded a number of species of Euproops from the South Wales Coalfield but unfortunately no photographs were included in their paper. Their inadequate descriptions and line drawings do not allow a close comparison to be made with the present material. Class ARACHNiDA Lamarck 1801 Order trigonotarbida Petrunkevitch 1949 Family eophrynidae Karsh 1882 Genus eophrynus Woodward 1871 Type species. Eophrynus prestvici (Buckland 1837) Eophrynus jugatus sp. nov. Plate 113, fig. 2 Diagnosis. Eophrynus with length of abdomen slightly greater than width. Dorsal exoskeleton of abdomen with one pair of large median tubercles on each sclerite, and ornament equigranular across entire width. Paired median ridges on ventral abdominal sclerites. Type material Holotype. It. 61019 (PI. 113, fig. 2). Internal and external mould of nearly complete specimen. Material figured shows internal surface of dorsal part of abdomen with part of ventral exoskeleton attached. The counterpart is retained in the reference collections of the Department of Geology, University of Sheffield. Horizon and locality. No. 9 seam, Farrington Group {tenuis Zone, Westphalian D), Upper Carboni- ferous (Moore and Trueman 1937, p. 228). Mine tip at Kilmersdon Colliery (Grid reference ST 681536) near Radstock, Somerset. Derivation of name. From the Latin jugum, meaning a yoke or ridge, referring to the ornament on the ventral exoskeleton. Description. The abdomen is 14-5 mm wide and 17 mm long. It has an evenly rounded posterior margin with greatest width at the midlength. Four conical smooth posterior spines are present, the median pair being slightly longer (2-5 mm) than the lateral ones. Each dorsal sclerite is ornamented by three pairs of tubercles, circular in outline and evenly covered by small granules. The ventral exoskeleton is composed of slightly over- lapping sclerites, each of which is ornamented by a pair of transverse median ridges, more prominent towards the posterior end of each sclerite, and less regularly spaced lateral paired ridges. Part of the margin of the carapace is preserved and is ornamented by large tubercles and small granules like those of the abdomen. The legs are poorly preserved, but appear similar to those of Eophrynus prestvici. The first right walking leg is the most completely preserved appendage, 18-5 mm in length and 2 mm wide at the proximal end. Ornament on the appendages varies from randomly placed granules on the distal segments to a rough ordering into transverse rows on the proximal segments. AMBROSE AND ROMANO; CARBONIFEROUS CHELICERATA 577 Discussion. Apart from the type and present species, E. ward (Dix and Pringle 1930, p. 142) is the only other species assigned to Eophrynus. The holotype of E. ward is far from complete and according to Petrunkevitch (1953, p. 77) is of doubtful generic affinities. The present species can be assigned to Eophrynus (Petrunkevitch, in Moore 1955, p. PI 12) and is discussed with reference to Buckland’s figured material of E. prestvid (Buckland, 1837, p. 77, fig. 2) and a cast of the specimen figured by Woodward (1871) and elsewhere (see Petrunkevitch 1953, p. 76). The dorsal ornament of E. prestvid in the median area is composed of coarse granules which may be one-quarter to one-half of the diameter of the smaller tubercles. The border and area around the lateral tubercles is ornamented by a larger number of much smaller granules. The granular ornament of E. jugatus is equal in size over the whole dorsal surface. The median tubercles of E. jugatus are developed as a single pair on each sclerite. They are not, as in E. prestvid, partly coalesced with a second anterior pair of smaller sub-triangular tubercles. E. jugatus shows on the ventral surface pairs of rounded (trans.) median ridges, more prominent at the posterior end and developed on each of the four sclerites preserved. No such ornament occurs in the ventral sclerites of E. prestvid. Petrunkevitch (1953, p. 77) rightly remarks that the ventral ornament indicated on Pocock’s figure (1902, fig. 1, p. 490) is an impression of the dorsal tubercles. This is also seen on the present material. The presence of lateral tubercles in addition to the granular tubercles on the carapace of E. jugatus is a further difference from E. prestvid. The abdomen of E. jugatus differs from that of E. prestvid in that it is slightly narrower {trans.) relative to its length. The shape of the abdomen is different, the widest part of E. jugatus being approximately at the midlength whereas in E. prestvid it is nearer the posterior margin. Although the four posterior spines in E. jugatus are 50% longer than those of Woodward’s specimen, Buckland’s specimen of E. prestvid referred to above has spines slightly longer even than those of E. jugatus. CONCLUSIONS The finding of seven randomly orientated specimens of Euproops kilmersdonensis on one bedding plane illustrates well the inadvisability of defining a species solely on length and width ratios when the rocks may have undergone deformation. The great majority of fossil arthropoda from the upper Carboniferous have been found preserved in nodules. Most of the figured American and British specimens of Euproops, and other specimens studied by the authors from the British Museum (Natural History) and the Sheffield City Museum are preserved in this way. The nodules range in size from 3 cm to 15 cm in diameter, and often in the case of the smaller ones the circular shape of a nodule has preserved only an incomplete telson ; in others the telson was evidently detached and not preserved with the body of the fossil. The specimen of Euproops kilmersdonensis which is preserved with a complete telson (pi. 112, fig. 1, specimen 5) shows a greater relative length of the telson than has been observed in other material or reconstruction. The Kilmersdon fossils are unusual in that they all occur on bedding planes in a grey silty mudstone, which has preserved the original material of the exoskeleton, and shows 578 PALAEONTOLOGY, VOLUME 15 finer structures than can normally be distinguished on casts preserved in nodules. The single exception to this is the very perfect preservation of Eophrynus prestvici described by Woodward 1871. Euproops kilmersdonemis appears to be more closely related to the American species of Euproops danae than to any previously described British species. This similarity to American fossils is also reflected in the associated floras which, during Westphalian D times bear close comparison with those of the Eastern United States (Dr. R. H. Wagner, pers. comm). Acknowledgements. The authors wish to record their thanks to Professor L. R. Moore for information on the horizon from which the specimens were obtained, and for use of facilities in the Department of Geology, University of Sheffield. They also wish to thank Mr. S. F. Morris, Palaeontology Depart- ment, British Museum (Nat. Hist.) and Dr. I. Strachan, Birmingham University for the loan of specimens in their care. Dr. E. G. Spinner kindly read the manuscript. REFERENCES BUCKLAND, w. 1837. Bridgewater Treatise. Geol. Miner., 2nd ed., 2. Dix, E. and PRINGLE, j. 1929. On the Fossil Xiphosura from the South Wales Coalfield with a note on the myriapod Euphoberia. Summ. Prog. geol. Snrv. G.B., 1928, Pt. II, 90-114, 16 figs. • 1930. Some Coal Measure Arthropods from the South Wales Coalfield. Ann. Mag. nat. Hist. [10], 6, 136-144, 4 figs. HEiDE, s. VAN DER, 1951. Lcs arthropodcs du terrain Houiller du Limbourg Meridional (excepte les scorpions et les insectes). Meded. Geol. Sticht., c-IV-3, 5, 1-84, 10 plates. LANGFORD, G. 1963. The Wilmington Coal Fauna and additions to the Wilmington Coal Flora from a Pennsylvanian Deposit in Will County, Illinois. Esconi Associates, 2nd ed., 280 pp. MEEK, F. B. 1867. Notes on a New Genus of Fossil Crustacea. Geol. Mag. 4, 320-321. MOORE, L. R. and trueman, a. e. 1937. The Coal Measures of Bristol and Somerset. Q. Jl geol. Soc. Fond. 93, 195-240. MOORE, R. c. 1955. Treatise on Invertebrate Palaeontology, Part P. Arthropoda 2. Lawrence, Kansas. 1959. Treatise on Invertebrate Palaeontology, Part O. Arthropoda 1. Lawrence, Kansas. PETRLTNKEViTCH, A. 1953. Palacozoic and Mesozoic Arachnida of Europe. Mem. Geol. Soc. Am. S3. pococK, R. I. 1902. On Eophrynus and Allied Carboniferous Arachnida. Geol. Mag. [4] 9, 439-448, 487-93. PRESTWICK, J. 1840. Memoir on the Geology of Coalbrook Dale. Trans. Geol. Soc. 2nd ser., v. RAYMOND, p. E. 1944. Late Paleozoic Xiphosurans. Bull. Mus. Comp. Zool. 94, 475-508, 2 plates. 1945. Xiphosura in the Langford Collection, in ‘Coal Age Fossils from Mazon Creek’. III. State Mus. Sci. Papers, 1-10, plates 1 and 2. WOODWARD, H. 1871. On the discovery of a new and very perfect Arachnide from the Ironstone of the Dudley Coalfield. Geol. Mag. 8, 385-388, plate 11. 1878. Monograph of British fossil Crustacea belonging to the Order Merostomata. Palaeontogr. Soc. [Monogr.], pt. v, 181-263, pis. XXXI-XXXVI. 1918. Notes on some Fossil Arthropods from the Carboniferous Rocks of Cape Breton, Nova Scotia. Geol. Mag. 55, 462-471, 5 figs. T. AMBROSE M. ROMANO Department of Geology, University of Sheffield, Mappin Street, Sheffield, SI 3JD Revised typescript received 13 December 1971 MONOGRAPTIDS FROM THE UPPER SILURIAN AND LOWER DEVONIAN OF YUKON TERRITORY, CANADA by D. E. JACKSON and a. c. lenz Abstract. Monograptids recorded and described from Arctic Canada for the first time comprise Morwgraptus cf. cormilus (Urbanek), M. cf. balticus Teller and M. helicoides sp. nov. from probable late Lndlovian; M. bouceki Pfibyl and Saetograptus pilosiis sp. nov. from mid-Pfidolian; M. uniformis angiistidens Pfibyl and M. uniformis parangiist ideas subsp. nov. from the latest Pfidolian angiistidens Zone; M. aequabilis aequabilis (Pfibyl) and M. cf. hercynicus subhercynicus Willefert from the lower Lochkovian uniformis Zone; and M. aequabilis notoaequabilis Jaeger and M. cf. craigensis Jaeger from the Pragian. In addition, the existence in Yukon Territory of M. uniformis uniformis Pfibyl is confirmed. The number of species and subspecies known from the Pfidolian and Lower Devonian strata in Yukon Territory is herein increased to eleven and nine respectively. The question of the youngest known species of Monograptus is also discussed. Existing knowledge of post-Ludlow graptolite biostratigraphy in Yukon Territory has unfolded in three stages. A beginning was made when we published an account of Monograptus yukonensis and stated it was the youngest Monograptus known in North America (Jackson and Lenz 1963). Dating the new species was, however, problematical because it was without graptolitic association and lay several hundreds of feet above the M. nilssoni Zone. Very shortly afterwards it became widely recognized throughout Yukon Territory and was shown to range through hundreds of feet of strata. The next significant development was the discovery of upper Budnanian and Lower Pfidolian graptolites on Porcupine River (Jackson and Lenz 1969); more recently we have been able to fill in many of the gaps in the zonal sequence above the Lower Pfidolian. Thus Lenz and Jackson (1971) documented the existence of the uppennost Pfidolian zone of M. transgrediens, the Lochkovian zones of M. uniformis and M. hercynicus, and the basal Pragian M. thomasi Zone directly below M. yukonensis. To these zones are now added the M. bouceki and M. uniformis angiistidens Zones. During the last few years, proof that species of Monograptus range upwards across the Siluro-Devonian boundary well into the Lower Devonian is considered to be one of the more significant contributions to biostratigraphy of this decade. Because agreement on the boundary problem is in sight and there appears to be widespread acceptance of a workable graptolite zonation within the Lower Devonian, it is natural that the tailing of the Monograptid Fauna should become a focal point of interest. At the present time there are no fewer than twelve species of Monograptus known from post-Pfidolian strata in various parts of the world, but whereas the successional array of the Loch- kovian species has a sound basis in observed stratigraphic occurrences, there is no general agreement on the relative ages of the Pragian species because most of them appear to have a restricted geographic distribution (see Table 1). At the present time it seems possible to assign these Pragian species to three lineages, namely, the monospecific lineages of M. atopus, and M. aequabilis, and the large and variable M. yukonensis lineage. The taxonomic status of some of these species in Table 1 [Palaeontology, Vol. 15, Part 4, 1972, pp. 579-597.] 580 PALAEONTOLOGY, VOLUME 15 is problematical due to poor preservation and to the extremely narrow range of variation in morphology upon which the speciation is based. Even more important, knowledge of the stratigraphic ages of some of the species has dramatically changed since their original description so that at this moment we have three contenders for the youngest known Momgraptus. They are M. atopus, M. yukouensis and M. pacificus. This state TABLE 1. Geographic distribution of Pragian species of Monograptus arranged in order of publication. Date Species ami author Occurrence 1963 M. yukouensis Jackson and Lenz Canada, U.S.A., Czechoslovakia 1964 M. belketaiefensis Planchon Algeria 1965 M. anguerensis Legrand Algeria 1966 M. atopus Boucek Czechoslovakia 1969 M. aequabilis notoaequabilis Jaeger Thailand, Australia, Russia, and Stein Czechoslovakia, U.S.A., Canada M. yukouensis fangensis Jaeger and Stein Canada, Thailand 1970 M. craigensis Jaeger Alaska, U.S.A. M. pacificus Jaeger Alaska, U.S.A. 1971 M. telleri Lenz and Jackson Canada of affairs stems from methods of dating as well as the circumstances of occurrence so let us digress one moment to examine both factors. The various methods used in dating these monograptids are: 1 . With reference to associated monograptids whose ranges elsewhere are known, e.g. M. telleri (see Lenz and Jackson 1971) and M. thomasi (see Jaeger 1967). 2. With reference to associated or contiguous shelly faunas, conodonts or plants, e.g. M. yukouensis (see Lenz 1967). 3. By stage of evolution achieved by monograptid in question, e.g. M. pacificus (see Churkin et ah, 1970). Concerning the relative ages of M. pacificus and M. yukouensis, we do not share the view expressed by Churkin, Jaeger and Eberlein (1970) that in Alaska M. pacificus post- dates and forms a mappable zone above M. yukouensis. If it is assumed that the two forms are distinct species, we believe that the existence of a M. pacificus-htanng horizon above another horizon containing M. yukouensis at a single locality carries little signi- ficance. Furthermore, there is little to be gained from using a presumed evolutionary stage of development until such time that a definitive phylogenetic scheme has been formulated which is firmly based upon stratigraphic succession. As far as the ages of the three contenders for the last monograptid is concerned, we simply believe it is an open question. It seems probable to us that M . pacificus becomes extinct before M. yukouensis because in Alaska both are associated with M. aequabilis notoaequabilis whereas in Yukon we find notoaequabilis associated with M. craigensis but not w\{h M. yukouensis sensu-stricto at the type locality. All specimens with GSC designation have been deposited with the Geological Survey of Canada, Ottawa. JACKSON AND LENZ: YUKON MONOGRAPTIDS 581 SYSTEMATIC DESCRIPTIONS Suborder monograptina Lapworth 1 880 Family monograptidae Lapworth 1873 Genus monograptus Geinitz 1852 Monograptus aeqiiabilis aequabilis (Pfibyl 1941) Text-fig. 1 A, C, K, L 1941 Pristiograptus aequabilis Pfibyl, p. 8, pi. 1, figs. 6-8. 1959 Monograptus aequabilis (Pfibyl 1941); Jaeger, pp. 102-105, pi. 1, fig. 8, pi. 4, fig. 3, pi. 5, figs. 1-5, text-fig. 17a, b. 1967 Monograptus aequabilis (Pfibyl 1941); Boucot, Cumming and Jaeger, p. 10, pi. 3, figs. 5-9. 1970 Monograptus aequabilis aequabilis (Pfibyl 1941); Churkin, Jaeger and Eberlein, fig. 9B, I. Material. A dozen specimens compressed on black mudstone from three stratigraphic levels in the Road River Eormation on Hart River. They include figured specimens; GSC 30083 and 30084 at 494 feet, GSC 30085 at 490 feet, and GSC 30087 from 500 feet stratigraphically above base of measured section. All material collected by D. E. Jackson and A. C. Lenz in 1969. Dimensions of figured specimens in mm GSC No. Length Rhabdosome width Thecae in tfii th^ th“ max. 1st 10 mm 2nd 10 mm 30083 55 0-8 (0-5) 0-9 (0-7) 1-2 (0-85) 1-4 thi®(l-0) 10 8 30084 8 0-95 (0-56) 1-0 (0-85) — — — — 30085 + 35 1-0 (0-7) M (0-85) 1-5(M) 1-8 tl+Ml-2) 11 10|- 30087 55 0-95 (0-5) ?M (0-7) 1-2 (10) 1-7 th"' 10 8 Description. Largest rhabdosome is 55 mm long exclusive of virgula and is 0-9-1 -0 mm across the hood of th^ widening to 0-9-1 -1 mm across th^ and attaining a maximum width of 1-7-1 -8 mm in first 2 cm, narrowing slightly distally. Dorsal edge of rhab- dosome is straight except for slight ventral incurvature at proximal end. Sicula is 1-5- 1-7 mm long, straight, apex extends distally to a level between apertures of thl and 2, sicula aperture furnished with fine virgellar spine and short dorsal tongue. Thecae biform. Thecae 1, 2 and occasionally 3 of uncinalus type possess down-curved supra-apertLiral hoods; more distal thecae have climacograptid profile and hoods are replaced by supra-apertural selvages about 0-2 mm long, the free ventral wall above the selvage is concave and slightly inclined. Inter-thecal septa are straight, inclined at 30 degrees to dorsal edge. Apertural excavations are approximately normal to interthecal septa and not at right angles to dorsal edge of rhabdosome. At the proximal end the length of the excavations occupy J-J- stipe width, increasing distally to where they are 0-5 mm long and 0-4 mm high. There are 10-112 thecae in first 10 mm reducing to 8-10| in second 10 mm and beyond. Remarks. The Hart River specimens fall within the limits of variation in width quoted by Jaeger (1959, p. 105) for the European material and these similarities can be extended TEXT-FIG. 1. (See opposite.) JACKSON AND LENZ; YUKON MONOGRAPTIDS 583 to include the length of the sicula and biform nature of thecae. The Yukon specimens have 8 to 10| thecae in 10 mm distally whereas Jaeger’s measured specimens never exceeded 9f in 10 mm. Occurrence and Distribution. According to Jaeger (1959), Monograptus aequabilis aequabilis is confined to the upper part of the M. uniformis Zone in Thuringia but Churkin, Jaeger and Eberlein 1970, p. 195, reported that this species ranged throughout the Lochkovian in the Carnic Alps. Willefert (1962) has also described M. aequabilis (Pfibyl) var. nov. from the Lochkovian of North Africa. In Yukon Territory, this sub- species is known from the Hart River section at four stratigraphic levels through the basal ten feet of the M. uniform is Zone, where it is associated with M. uniformis uniformis. On Peel River, it has been collected in the Upper Canyon (65° 52' N, 135° 45' 40" W) 235 feet above M. transgrediens praecipuus. Monograptus aequabilis notoaequabilis Jaeger and Stein 1969 Text-fig. 1 B 1966 Monograptus aequabilis (Pfibyl 1941); Jaeger, pp. 398-403, pi. 41. 1969 Monograptus aequabilis notoaequabilis Jaeger and Stein, pp. 182-184, text-fig. 1e-f, pi. 15, fig. A, B. 1970 Monograptus aequabilis notoaequabilis Jaeger and Stein; Churkin, Jaeger and Eberlein, pp. 194-195, fig. 9 c, J. 1970 Monograptus aequabilis notoaequabilis Jaeger sic, Keren and Enokyan, pi. IX, figs. 1^. Material. Two well preserved, and several poorly preserved specimens as films in uncleaved shale. Eigured specimen GSC 30086 from uppermost part of Road River Formation, Tetlit Creek; 66° 44' N, 135° 46' W, Yukon Territory; field designation DJ-66-8F; collected by M. C. Pick, Chevron Standard Ltd., 1966. Description. The figured specimen is 50 mm long, rhabdosome straight, widening from 0-8 mm across hood of th^ (0-5 mm immediately above hood), to 1-3 mm at th^® and a maximum of T7 at th^° diminishing to 1-5 mm at distal end. Thecae biform, th^~^ provided with down-curved apertural hoods. Subsequent thecae have climacograptid profile and hoods are replaced by selvages at point of geniculation. Apertural excavations horizontal, 0-5 mm long and 0-3 mm high in mature part of rhabdosome. Free ventral wall concave, inclined so that maximum width of thecae is at apertural lip. There are 9| thecae in first 10 mm; 8| in second 10 mm, and 41 in 50 mm. TEXT-FIG. 1. A, ic, L. Mouograptus aequabilis aequabilis (Pfibyl). A, K. GSC 30084 and 30083, 494 feet in Hart River section. X 24. l. GSC 30085 at 490 feet in Hart River section, x 24. b. M. aequabilis notoaequabilis Jaeger and Stein. GSC 30086 80 feet below top of Road River Formation in Tetlit Creek. X 2. c. M. aequabilis aequabilis Pfibyl. GSC 30087, 500 feet in Hart River section. X 24. D-i. M. uniformis parangustidens subsp. nov. d-f, h, i. Paratypes GSC 30088-30090, 30092-30093, 430 feet; and g, Paratype GSC 30091, 445 feet. Hart River section. x4. j, q. M. uniformis angustidens Pfibyl. GSC 30094 and 30095, 995 feet in tributary of Peel River. x24. p. M. uniformis uniformis (Pfibyl). GSC 30096, 509 feet in Hart River. x2. m-o. M. uniformis parangustidens subsp. nov. N. Holotype GSC 30098, 430 feet. x4; M, o. Paratypes GSC 30097 and 30099, 445 feet and 430 feet, respectively. Hart River section. x4. Note: two dots represent 5 mm. Qq C 9‘202 TABLE 2. Stratigraphic distribution of graptolite species from Road River Formation of Hart River section located at river level on west bank of Hart River, Ogilvie Mountains, latitude 65“ 34' N, longitude 136“ 55' W. Section 75% exposed; measured by authors June, 1969. X—positive occurrence; ? — possible occurrence at that interval. DISTANCE IN FEET ABOVE BASE OF SECTION SPECIES ?i- „ > , 5 2 1 s s 1 ; ? : s s s s s § s = s s 1 s 5 3 s ? S 1 s ? s S ; 1 ? • ■ >s X X X » ...... X ? X X V I8oiona«) X x,x X ? ' 1 i ' U. unei»a(i,« Tullb«r« 1 1 1 X 1 ' ' 1 «. r,p 1 rr 1 ! ! 1 M,oIt.fe»m.rl ISgrrgndtl i »' U.bon.li X i « et. TiiKr c.,«ggp./ K=bp. X r JotJitt. erPUrs X X Bog£t» X X "• X « c. T.M.r X « IT.W.rl ? ? ? ? X X X X X X X X X X X X ? X « X u cl Pt/r/«icuiw^^io^0Jufcpfl,epi X " «»"«'<■. 1 8r.b,i I < X - X > 1 1 1 1 1 1 1 1 1 PALAEONTOLOGY, VOLUME 15 TABLE 2. Stratigraphic distribution of graptolite species from Road River Formation of Hart River section located at river level on west bank of Hart River, Ogilvie Mountains, latitude 65° 34' N, longitude 136° 55' W. Section 75% exposed; measured by authors June, 1 969. X — positive occurrence ; ? — possible occurrence at that interval. 584 PALAEONTOLOGY, VOLUME 15 60S OOS X X X X DISTANCE IN FEET ABOVE BASE OF SECTION 06b X S8b X X Sbb X 0£b X Oib X SOb X 0l£ X CM. X Sb£ X X 0b£ X X IZ£ X X 02£ X X X 082 X X 092 X Sb2 X 9£2 X SC2 X b£2 X Ow. 022 CV. X SI2 X X (V . X 012 X X 0v_. S02 X X IZI X £2 1 X X 021 X X S II X 00 1 X X X 06 X X S8 X X X 08- Si (Y. X 89-S9 X X X 09 X X X SS X X X 0S-8b X X Sb X X X 0b-8£ Ow. S£ X 2£ X 0£ cv.. X 92 X il'SI X X X bl-£i X X X X 01 8 X X X X X b X X X 2 X X X X 1 X X X X X S 0- 0 X X GRAPTOLITE SPECIES S . .. .t> 55S6SS555SS SSS ^SSSS^SSSES 55^. JACKSON AND LENZ: YUKON MONOGRAPTIDS 585 The siciila is 2-2 mm long and 0-5 mm wide at aperture, apex lies at level of aperture of th^. Sicular aperture is furnished with virgella and prominent dorsal tongue. Remarks. Dimensions of rhabdosome, thecae and sicula agree with measurements given by Jaeger et al., except that the Yukon specimen is the longest specimen on record. Occurrence. These Yukon specimens occur 60 feet above M. cf. craigensis Jaeger et al., in the uppermost part of the Road River Formation on Tetlit Creek where the grapto- litic shales are overlain by shales with Siegenian conodonts. Apart from the Canadian occurrence, it is known to occur in Bohemia (Boucek 1966), Australia (Jaeger in Chur- kin et al. 1970), Thailand (Jaeger, Stein and Wolfart 1969) and Alaska (Churkin et al. 1970) and in all instances it is confined to the Pragian Stage. Monograptus cf. bait lens Teller 1966 Text-fig. 2 I, j, K cf. 1966 Monograptus balticiis Teller, p. 556, pi. 1, figs. 6-11, text-fig. 4a-b. Material. Thirty-two specimens, mostly comprising the proximal ends from the Road River Formation on Hart River, 85 feet above the base of the section. They include illustrated specimens GSC 30115 to 30117. All material collected by D. E. Jackson and A. C. Lenz 1969. Dimensions of figured specimens in mm GSC No. Length Rhabdosome width Thecae in 5 mm Sicula length 1-6 30115 16 tfii 0-7 (0-4) th® 1 05 (0-75) tlT» 1-2 max. 1*25 prox. 5i distal 5i 30116 19 0-7 (0-4) 10 (0-65) 1-3 1-3 5 1-6 30117 14 0-8 (0-4) 105 (0-7) 1-3 1-5 5i 5i 1-5 Description. Rhabdosome medium-sized, up to 37 mm long, with moderately weak dorsal curvature between th^ and th®~'^, thereafter straight, or occasionally having weakly developed ventral curvature. Width increases gradually from about 0-7 mm at th^, to a maximum of T3-T4 mm between th^‘^"^'‘. Thecae uniform, hooded. Free ventral wall subparallel to dorsal edge of rhabdosome except at proximal end. Thecal apertures unthickened, approximately at right angles to long axis of thecae, generally obscured by strongly recurved hoods. Thecal hoods occupy about half of rhabdosomal width proximally and about quarter width distally. Thecae occur at the rate of 1 1-12 in 10 mm proximally and 9|— 1 1 distally. Sicula about 1-6 mm long, apex reaching to about level of top of th^, aperture tilted toward ventral side. Well-developed virgella extends ventro-proximally and may reach 1 mm in length; dorsal tongue short. Remarks. The Flart River specimens resemble M. balticiis from Poland in the shape of the rhabdosome, number of thecae, length and shape of sicula and prominence of virgella. Our material differs in the lack of incurved thecal apertures and undulating interthecal septum shown in Teller (1966, pi. 1, figs. 8 and 9, and text-fig. 4), but more closely resemble Teller’s figures 6 (holotype), 10 and 11 of the same plate. Occurrence. Monograptus cf. balticiis lies 20 feet above the highest occurrence of M. boliemicus boliemiciis and 15 feet below M. paraformosus. It is associated on the same 586 PALAEONTOLOGY, VOLUME 15 TEXT-FIG. 2. A, B, D, F, H, L. Monograptus Iielicoides nov. sp. F, Holotype GSC 30122; Paratypes a, b, d, GSC 30123, 30124, 30127; h, l, GSC 30125 and 30126, 45 feet on Hart River, c, e, g, m. Monograptus cf. conmtus (Urbanek). c, GSC 30119; e, GSC 30121 ; g, GSC 301 18; m, GSC 30120, 60 feet on Hart River, i, j, k. Monograptus cf. balticus Teller, i, GSC 301 16; J, GSC 301 15; K, GSC 301 17, 85 feet on Hart River. All figures X 6. TEXT-FIG. 3. A-F, I. Moiiograptiis bouceki Pfibyl. a-f, GSC 30100-30105 respectively; i, GSC 30106, 550 to 555 feet in tributary of Peel River, g, h, j-l. Saetograptus pilosus sp. nov. k, Holotype GSC 30110 and Paratypes g, h, GSC 30109 and 30112; J and l, GSC 30113 and 30111. All figures x4. 588 PALAEONTOLOGY, VOLUME 15 bedding plane with M. aff. roemeri and M. cf. haupti. Its age is tentatively considered to be late Ludlovian as in Poland, but may range downward into the middle Ludlovian. Monograptus bouceki Pfibyl 1940 Text-fig. 3 A-F, I 1940 Monograptus (Poinatograptus) bouceki Pfibyl, pi. 1, figs. 7-8, text-fig. 1, no. 4. 1942 Monograptus bouceki Pfibyl, p. 6, pi. 1, figs. 1-3. 1964 Monograptus bouceki Pfibyl; Teller, pp. 56-57, pi. II, fig. 13, pi. V, fig. 5, pi. VI, figs. 1-3, pi. VIII, figs. 12, 13, pi. XV, figs. 4, 5, text-fig. 13a-d. Material. Seven adult or near adult specimens preserved as films in black shale are available, namely, GSC 30100 to 30106, Road River Formation on unnamed tributary of Peel River, latitude65° 53'45"N, longitude 135° 55' 25" W, field designation PW-2L at 550-555 feet; collected by A. C. Lenz and D. E. Jackson 1969. Description. Rhabdosome small, not seen to exceed 20 mm. Dorsal margin with dorsal curvature between apertures of th^ and th^^^, and straight distally. Width across the apertural hood of th^ is 0-75 to 1-0 mm (average 0-9 mm), across th® it is 1-0-1 -2 mm and at th^** only 1-3-1 -5 mm. A maximum width of about 1-5 mm is attained between th8-“. Thecae uniform, free ventral wall inclined up to 30° toward dorsal edge of rhabdosome proximally, becoming subparallel distally. Thecal apertures at proximal end are obscured by down-curved apertural hoods which distally become slightly smaller exposing apertural margin. There are 41 to 6 thecae in first 5 mm and 11 in first 10 mm. SicLila 2 mm long, apex reaching midway between apertures of th^ and th^, axis curved so that aperture deflected ventrally. Aperture carries prominent virgella and dorsal tongue. Dinieusious of figured specimens in nun GSC No. Length Rhabdosome width Thecae in tip th^ thi“ max. 1st 5 mm 1st 10 mm 30100 14 0-9 1-0 1-3 1-3 tfii" 5| 11 30101 +20 0-9 1-2 1-3 1-5 th“ 5^ 11 30102 14 0-9 1-2 1-5 1-5 tip 5 10|- 30103 17-5 0-9 1-2 1-5 1 -5 th« 4.1 10 30104 14 0-9 M 1-4 l-4th* 5 10.1 30105 15 1-0 1-2 1-4 1-5 th’’ 5^- 10 30106 14 0-75 1-2 1-4 1 -5 th“ 6 10. 1 Remarks. This material is identical with that described by Teller (1964) and differs from Pfibyl’s original description only in the slightly greater length of the sicula. Rhabdo- somes of the Yukon specimens have a distinct dorsal curvature as mentioned by Pribyl and illustrated by well-preserved material from Australia (Jaeger 1967, pi. 14, fig- a). Occurrence. The species occurs on a single bedding plane associated with Saetograplus pilosiis sp. nov. and fragments of Linograptus 20 feet above M. ex gr. M. iransgrediens and 350 feet below the yukonensis Zone. We consider tentatively that this occurrence JACKSON AND LENZ: YUKON MONOGRAPTIDS 589 signifies the existence of the M. bouceki Zone in Yukon Territory and represents the first recognition of the zone on this continent. Monograptus cf. cornutus (Urbanek) 1970 Text-fig. 2 c, E, G, M cf. 1970 Bohemograptiis cornutus Urbanek, p. 310, pi. 20, fig. D, pis. XXV-XXVIII. Material. Fourteen specimens compressed in black mudstone from the Road River Formation of Hart River, at 60 feet above the base of the measured section. Figured specimens are GSC 30118 to 30121. All material collected by D. E. Jackson and A. C. Lenz 1969. Dimensions of figured specimens in mm GSC No. Length Rhabdosome width Thecae in Sicula 1st 5 mm length tip th® maximum at distal end 30118 8 0-65 1-25 1-25 — 30119 8 0-65 1-25 1-45 1 1-45 30120 8 0-8 M 1-4 1 + 1-0 30121 7 0-55 1-1 10 1 — Description. Rhabdosome short, up to 8 mm long with strong ventral curvature proxi- mally, weakening distally. Width increases gradually from about 0-7 mm exclusive of lappets at th^ to a maximum of T45 mm at distal end. Thecae simple, inclined about 30-40° to axis of rhabdosome, overlapping one-half to two-thirds, and about three times longer than wide. Apertural margins of thecae modified by paired lappet-like appendages which exceed T5 mm in length. Specimens not well enough preserved to determine whether apertural lips are thickened by peri- dermal tissue, such as illustrated by Urbanek (1970, pi. XXV, fig. Ai). Sicula conical, ventrally curved, maximum length 1-45 mm and width 0-4 mm. Apex of sicula about level with top of thfi Remarks. This species is distinguished from all other species of bohemograptids, including the closely related M. praecornutus (Urbanek), by the long, thecal lappets. The Yukon specimens appear to differ from the Polish material in the apparent lack of lappets in the first two or three thecae, and in having a shorter sicula (T45 mm vs. 1-61-2T4 mm). In the latter feature, the Yukon material overlaps the sicula measure- ment of M. praecornutus (Urbanek). Occurrence. The species is known from a single bedding plane where it is associated with Monograptus bohemicus bohemicus, M. bohemicus tenuis, Monograptus aff. roerneri and ILinograptus sp. indet. (see Table 2). In Poland M. cornutus occurs within the cornutus Zone of Late Ludlovian age (Teller 1969, p. 457). Monograptus cf. craigensis Jaeger 1970 Text-fig. 4 F, G, M cf. 1970 Monograptus craigensis Jaeger in Churkin, Jaeger and Eberlein, pp. 198-202, figs. 6, 7B, C, 8B, C, 9A, F, K. Material. Three immature compressed specimens are available comprising figured specimens GSC 30107-30109, Road River Formation, Tetlit Creek, latitude 66° 44' N, longitude 135° 46' W, field designation DJ-66-2F; collected by M. C. Pick, Chevron Standard Ltd., 1966. 590 PALAEONTOLOGY, VOLUME 15 Description. Largest rhabdosome 20 mm long, dorsal margin is dorsally curved reversing to ventral curvature beyond th^; rhabdosome widens gradually to about th^^. Width increases from 0-8 mm across hood of th^ (04 mm immediately above hood) to 1-2- 14 mm across hood of th^ and 1-5-1 -8 mm at th^“. The maximum observed width of 2-5 mm is at th^" (1-7 mm immediately above apertural hood). Thecae are of M. yiikonensis form provided with apertural hoods which increase in size slightly towards distal end. Subapertural free ventral wall of thecae is concave and inclined at 30° to 40° to axis of rhabdosome. There are 6 thecae in first and second 5 mm reducing to 5 per 5 mm thereafter. The distance between the aperture of th^ and the base of sicula is 1-5 to T6 mm. Length of sicula is unknown, sicular aperture does not apparently possess large expanded virgella. Remarks. This material resembles M. craigensis in the degree of recurvature of the proximal end, the gradual widening of the rhabdosome to the same maximum width, and in having the same thecal spacing over the first 10 mm. A minor difference involves the thecal hoods which enlarge slightly rather than diminish distally. The lappet-shaped expansion of the virgella which Jaeger (in Churkin et al. 1970) considers characteristic of adult representatives of the species was not observed on our immature specimens. Occurrence. This form from Tetlit Creek occurs 75 feet below M. aequabilis noto- aequabilis thus providing a striking parallel to the range of the two species at the type locality at Port St. Nicholas, Prince of Wales Island, Alaska. Monograptus helicoides sp. nov. Text-fig. 2 A, B, D, F, H, L Material. Twenty-six moderately well-preserved specimens compressed on black mudstone, are available. Holotype. GSC 30122 (text-fig. 2f). Paratypes. GSC 30123-30127. Type locality and horizon. Road River Formation, 45 feet above base of measured section on Hart River; collected by D. E. Jackson and A. C. Lenz 1969. Derivation of name. Helix, Greek for screw. Dimensions of figured specimens in mm GSC No. Length Rhabdosome width No. of thecae Sicula in 5 mm length tfii th® max. 30122 12 0-3 10 10 7? 1-4? 30123 H-12 04 0-8 0-9 7? — 30124 + 12 0-3 0-8 10 7 14 30125 15 0-3 0-8 M 6 1-6 30126 + 10 045 0-9 0-9 — 1-6 30127 19 0-3 0-8 10 7 1-6 JACKSON AND LENZ: YUKON MONOGRAPTIDS 591 Description. Rhabdosome small, maximum length 19 mm, coiled in one and one-half spirals. From the fact that the sicular region may be hidden by, or partially conceals succeeding portions of rhabdosome, it is probable that the spiral was helicoid rather than planispiral. Rhabdosome widens gradually and uniformly from 0-3-04 mm across th^ to a maximum of about 1-1 mm. Thecae simple, inclined 15-20° to the dorsal edge of the rhabdosome, overlapping about one-half, and 4-5 times longer than wide. Apertures approximately at right angles to axis of thecae. Thecae are spaced at the rate of 7 in 5 mm. Sicula generally straight about 1-6 mm in length and 0-3 mm wide at aperture. Apex of sicula apparently lies just below level of aperture of thh Remarks. The combination of simple, Monograptiis bohemiciis type thecae, and the tight spiral of the rhabdosome distinguishes this species from all known monograptid species of Ludlovian age. Indeed it is the only monograptid in the Upper Silurian which has a coiled stipe. Occurrence. This species occurs on a single bedding plane in association with Mono- graptiis bohemicus tenuis and Monograptiis cf. egregius Urbanek. It lies 15 feet strati- graphically below Monograptiis cf. corniitus and 1 5 feet above M. leintwardinensis primus suggesting a middle or late Ludlovian age (see Table 2). Monograptus cf. hercyniciis siibhercyniciis Willefert 1963 Text-fig. 4 A, B, D cf. 1963 Monograptiis hercyniciis siibhercyniciis Willefert, p. 75, pi. II, figs. 16, 18, 23, text-figs. 4a-d, 5a-e. Material. Two dozen specimens, only seven of which show reasonable details, compressed on black mudstone from Hart River. Figured specimens consist of GSC 30128 to 30130 from Hart River. All material collected by A. C. Lenz and D. E. Jackson in 1969. Description. The species is characterized by a length greater than 10 mm, width across th^ is 0-8-0-9, and a maximum width of 1-9 mm. The thecae are of the hercyniciis type and number 6-7 in the proximal 5 mm, and 5-6 in the distal portions. The rhabdosome is essentially straight throughout except for a very weak dorsal curvature in the region of the proximal thecae. The most distinctive characteristic, and one which distinguishes it from otherwise similar forms such as M. hercyniciis, M. angustidens and M. praeher- cyniciis, is the sicula. It is triangular to slightly flaring, up to 2 mm long, and possesses straight virgella and dorsal process. Dimensions of figured specimens in mm GSC No. Length Rhabdosome width No. of thecae Sicula length tfii th® max. 1st 5 mm 2nd 5 mm 30128 20 10 1-23 1-85 tfii® 6-5 5-6 1-8 30129 16 10 1-54 1 -85 th® 6 5-5 1-9 30130 18-5 M5 1-46 1-85 tfiii 7 60 1-4 Remarks. The Yukon material resembles this subspecies in possessing a straight rhabdo- some, same thecal rate, and particularly in the shape and length of the sicula. TEXT-FIG. 4. (See opposite.) JACKSON AND LENZ; YUKON MONOGRAPTIDS 593 Occurrence. This species was collected 485 feet above the base of the Road River Formation on Hart River and is considered to lie within the Gedinnian M. imiformis Zone (see Table 2). Monograptus uniformis august idens Pfibyl 1940 Text-figs. 1 J, Q, 4 C, E, K, L 1940 Monograptus angiistidens Pfibyl, p. 70, text-fig. I, 1,2, pi. 1, figs. 3, 4. 1964 Monograptus angustidens Pfibyl; Teller, p. 60, pi. 2, fig. 11, pi. 8, figs. 1-3, pi. 9, figs. 13-15, pi. 13, fig. 5, text -fig. 15a, b. Material. Several dozen specimens preserved as carbonaceous films on black shale, collected from two bedding planes 990 feet and 995 feet above the base of the section on an unnamed tributary of Peel River, latitude 65° 53' 45" N, longitude 135° 55' 25" W, field designation PW-8J, at 990 feet and 995 feet. Figured specimens comprise GSC 30094 and 30095, and GSC 30131-30134, collected by D. E. Jackson and A. C. Lenz 1969. Description. Rhabdosome up to 36 mm long, straight except for very weak ventral curvature between th^ and th"~^. Width of rhabdosome increases fairly rapidly from 0-8-1 T mm across th^ to 1-25-1-3 mm across th^, and attains maximum width of up to 2-4 mm beyond th^®. Sicula normally 1-7 mm long, curved ventrally, and possessing a distinct ventrally directed dorsal process. Apex of sicula lies at level of aperture of th^. Thecae provided with down-curved apertural hoods which increase in size proximally, then maintain constant size, or decrease slightly, distally. Thecae overlap such that a transverse section across the rhabdosome would cut through only two tubes. There are 6 to 7 thecae in 5 mm proximally, and 9-1 1 in 10 mm distally. Dimensions of figured specimens in mm GSC No. Length Rhabdosome width Thecae in tip tip tfii® max. 1st 5 mm distally in 5 mm 30094 13 0-9 1-3 1-75 2-15 tffi’ 6| 5| 30095 42 0-9 1-17 1-6 2-16 th« 6f 4| 30131 31 1-1 1-3 1-7 2-4 th^” 6 4f 30132 29 0-8 1-25 1-8 2-25 th^* 6i- 5J 30133 22 0-75 M 1-5 1 -8 till’ 6 5 30134 20 0-8 1-25 1-65 2-25 th“ 6 5i Remarks. Because of the inadequate nature of Pfibyl’s (1940) original description, comparison with the type specimens is difficult. However, our material is very similar in every respect to the material described by Teller (1964). In addition, measurements of additional Bohemian material made by Jaeger (privately circulated at the Second International Symposium on the Silurian-Devonian Boundary, Leningrad, 1968) con- form closely to our material. Specimens from the Peel River area are extremely similar to those from Porcupine TEXT-FIG. 4. A, B, D. Moiiograptus cf. hercynicus subhercynicus Willefert. A, GSC 30130; b, GSC 301238; D, GSC 30129, 485 feet on Flart River, c, e, k, l. Monograptus uniformis angustidens Pfibyl. c, GSC 30132; k, GSC 30131, 990 feet on Peel River tributary; e, GSC 30134; l, GSC 30133, 995 feet on Peel River tributary, f, g, m. Monograptus cf. craigensis Jaeger, f, GSC 30109; g, GSC 30108; M, GSC 30107, 20 feet below top of Road River Formation in Tetlit Creek, h-j. Monograptus uniformis uniformis Pfibyl. h, GSC 30135; i, GSC 30136; j, GSC 30137, 509 feet in Hart River. All figures X3. 594 PALAEONTOLOGY, VOLUME 15 River described as M. aff. august idens by Jackson and Lenz (1969). The major difference between the forms is that the latter are consistently broader. Further collecting and studies will be necessary to determine if this difference is meaningful. Occurrence. Monograptus uniformis angustidens is, to date, known from an unnamed tributary on the north side of the Peel River, about six miles upstream from the beginning of the Upper Canyon. It lies 45 feet stratigraphically above the highest occurrence of M. transgrediens praecipuus, and 15 feet below a form morphologically intermediate between M. uniformis angustidens and M. hercynicus subhercynicus. The occurrence of this species undoubtedly represents the latest Pfidolian angustidens Zone, and as such is the first recognition of the zone in western Canada. Monograptus uniformis parangustidens subsp. nov. Text-fig. 1 D-i, M-o Material. A dozen specimens preserved as pyritized films in black shale and collected from two bedding planes. Holotype. GSC 30098 (text-fig. 1 n). Paratypes. GSC 30088-30093, 30097-30099. Type locality and horizon. Road River Formation on Hart River, 430 feet and 445 feet above base of measured section; collected by D. E. Jackson and A. C. Lenz 1969. Description. Largest rhabdosome 28 mm exclusive of virgula, straight except for dorsal curvature between th^“^ and ventrally deflected sicula. Width of rhabdosome is 0-6- 0-8 mm across hood of th^ increasing to L0-T4 mm across th^ and has a maximum width of T5-1-7 mm beyond th^°. Sicula is 1-7 mm long, distinctly curved ventrally with prominent virgellar spine, apex lies at a level slightly below th^ aperture. All thecae provided with apertural hoods which sometimes increase in size distally. Proximal thecae slightly isolate with strongly down-curved hoods, distal thecae have climacograptid profile and with hoods above horizontal apertural excavations, and a transverse section across the rhabdosome would cut through only two tubes. Thecal rate is 11-12 in 10 mm. Dimensions of figured specimens in mm GSC No. Length Rhabdosome width Thecae in tfii th^ tip® max. 1st 5 mm 1st 10 mm 30088 5-7 0-8 (0-4) 1-4 (0-8) — — — — 30089 14 0-75 (0-45) 1-25 (0-9) 1-4 (1-2) 1-5 th“ 6 11 30090 180 0-8 (0-5) 1-2 (0-75) 1-5 (1-2) 1-5 tip® 6 — ■ 30091 10-5 0-75 (0-6) 1-0 (0-75) 10 (0-9) 1 -3 th® 64 — 30092 13-5 0-8 (0-4) +0-8 (0-8) 10— 1-5 tlP» 5| 12 30093 17-3 0-7 (0-3) 10 (0-75) 1-2(10) 1-4 th'® 64 12 30097 25-8 0-6 (0-3) 10 (0-6) 1-2 (10) 1-7 tlr^ 64 12 30098 28 0-75 (0-4) 1-2 (0-7) 1-25 (0-8) 1-5 tip® 64 12 30099 23-3 0-9 (0-5) 1-25 (0-9) 1-4 (1-25) 1-75 tip* 11 Remarks. This subspecies is clearly related to the Monograptus uniformis group, and particularly to M. uniformis angustidens and stratigraphically appears to be a chrono- species of uniformis. It differs in its more distinct proximal dorsal curvature, more JACKSON AND LENZ: YUKON MONOGRAPTIDS 595 gradual increase in width, and in general by being more slender (average 1-5 or 1-6 mm for parangustidens vs. greater than 2 mm for angustideus). M. imiformis paragiistidens differs from M. telleri Lenz and Jackson in its more pronounced proximal dorsal curva- ture, and more gradual increase in width. Occurrence. The specimens were collected 430 feet and 445 feet above the section base, on the Hart River. The lower collection lies 20 feet above the highest occurrence of M. transgrediens praecipuus, and the higher occurrence lies 40 feet below M. cf. hercynicus subhereynicus. Their occurrence almost certainly represents the highest Pfidolian angustideus Zone. Monograptus imiformis uniformis Pfibyl 1940 Text-figs. 1 p, 4 H-J 1940 Monograptus (Poinatograptus) imcinatus uniformis Pfibyl, p. 71, pi. 1, fig. 1. 1959 Monograptus uniformis Pfibyl; Jaeger, p. 94, pi. 1, fig. 3, pi. 3, figs. 9-10, pi. 4, figs. 4-15, ffs. 16d-h. For complete synonymy up to 1959, see Jaeger 1959, p. 94. 1971 Monograptus cf. uniformis Pfibyl; Lenz and Jackson, p. 16, text-fig. 3C, G. Material. Three specimens from 494 feet, and nine specimens from 509 feet above the base of the section on Hart River. All compressed on black shales. Figured specimens comprise GSC 30096 and 30135-30137. All material collected by D. E. Jackson and A. C. Lenz 1969. Diseussion. This species was tentatively identified by us (1971) also from the Hart River section, but its fragmentary condition prevented certain identification. Subsequent collecting has yielded complete and better preserved material. Our material resembles that described by Jaeger (1959) by lengths of 50 mm or more, maximum width of up to 3 mm, strongly overlapping thecae, and thecae numbering up to 7 in 5 mm proximally and 5-6 in 5 mm distally. Occurrenee. M. uniformis is associated with M. aequabilis aequabilis at 494 feet, and is without association at 509 feet in the Road River Formation of Hart River. These two species characterize the Gedinnian uniformis Zone. Genus saetograptus Pfibyl 1943 Saetograptus pilosus n. sp. Text-fig. 3 G, H, j-L 1947 Monograptus chimaera var. alaskaensis Ruedemann, p. 475, pi. 85, figs. 17-22. Material. A dozen specimens preserved as silvery films on black shale are available. Holotype. GSC 30110, text-fig. 3 H. Paratypes. GSC 30111-30114. Type locality and horizon. Road River Formation, on unnamed tributary of Peel River; latitude 65° 53' 45" N, longitude 135° 55' 25" W, field designation PW-2L at 550-555 feet collected by A. C. Lenz 1969. Derivation of name. Pilosus, Latin for hairy. Description. Rhabdosome up to 21 mm long exclusive of virgula, dorsal margin exhibits slight ventral curvature in proximity of sicula and strong dorsal curvature between th^ and th^, becoming nearly straight distally. The amount of dorsal curvature varies between 596 PALAEONTOLOGY, VOLUME 15 10 and 35°. Width of rhabdosome across aperture of th^ is 1-0-1 -3 mm increasing to 1-3-1 -7 mm at th^ the maximum width not exceeding 1-7 mm. Thecae increase in size rapidly so that a maximal width is often achieved by th^. Thecal apertures provided with conspicuous processes which we interpret as lateral lappets and spines. The lappets are presumed to consist of monofusellar tissue whereas the spines may be formed from enrolled periderm as in Saetograptus chimaera cervicornis (see Urbanek 1958, pp. 54-61) and Saetograptus variaus (Wood) (see Hutt 1969, pp. 361-8). The spines appear to be longest in the proximal thecae although this may result from a change in orientation as Hutt found in S. variaus. There are 10 to 11 thecae in the first 10 mm. Sicilia is 1-5 to 1-7 mm long, curved, sicular aperture deflected toward ventral side, apex extends to between th^ and th^. Aperture furnished with a conspicuous virgella and dorsal tongue. Dimensions of figured specimens in mm CSC No. Length Rhabdosome width In first In first 5 mm 10 mm tN th® thi« max. 30110 200 1-5 (10) 1-5 (10) 1-7 6 10 30111 210 1-3 (0-7) 1-5 (0-8) — — 6 — 30112 6-3 10 (0-5) 1-7 (1-0) — — 6 — 30113 120 10 (0-5) 1-4 (10) 1-7 (10) 1-7 11 30114 60 1-0 (0-3) 1-3 (0-9) — — 5 — Remarks. The species is undoubtedly conspecific with the undated Monograptus chimaera alaskaeusis Ruedemann 1947, collected on the Porcupine River in Alaska. A comparison of the apertural processes of S. pilosus with Urbanek’s (1958, pi. II, fig. 4a) illustration of the thecal profile of S. chimaera cervicornis leaves no doubt of the affinities of the Yukon species with the chimaera group. However, because of the distinct dorsal curvature of the rhabdosome, and ventrally recurved sicula of the North American specimens, we prefer to assign them to a new species. Occurrence. The species is known from a single bedding plane associated with M. bouceki and fragments of Linograptus posthumus tenuis, 25 feet above M. sp. ex gr. M. transgrediens, and 400 feet below M. yukonensis. It is therefore tentatively dated as mid-Pfidolian, and possibly occurs within the bouceki Zone. Acknowledgements. We are indebted to V. Zay Smith and Associates, Calgary, Alberta, for release of stratigraphic information, and for the opportunity to collect the material herein described. Ruede- mann’s type specimens of Monograptus chimaera alaskaeusis were made available for comparative study through the help of Michael Churkin, Jr., U.S. Geological Survey. Expenses involved in the initial preparation and publication of this study were borne by National Research Council of Canada Grants A4236 to A2631 to A. C. Lenz and to D. E. Jackson respectively. REFERENCES BOUCOT, A. J., GUMMING, L. M. and JAEGER, H. 1967. Contributions to the age of the Gaspe Sandstone and Gaspe Limestone. Pap. geol. Surv. Can. 67-25. CHURKIN, M., JAEGER, H. and EBERLEiN, G. D. 1970. Lower Devonian Graptolites from southeastern Alaska. Letliaia, 3, 183-202. Hurr, j. 1969. The development of the Ludlovian graptolite Saetograptus varians. Lethaia, 2, 361-368. JACKSON AND LENZ; YUKON MONOGRAPTIDS 597 JACKSON, D. and lenz, a. c. 1963. A new species of Monograptus from the Road River Formation, Yukon. Palaeontology, 6, 751-753. 1969. Latest Silurian graptolites from Porcupine River, Yukon Territory. Bull. geol. Surv. Can. 182, 17-29, pis. 3-5. JAEGER, H. 1959. Graptolithen und Stratigraphie des jiingsten Thiiringer Silurs. Abli. dt. Akad. fVlss. Berl. (Chem. Geol. Biol.), 1959, No. 2, pp. 1-197, pis. 1-14. — — 1966. Two late Monograptus species from Victoria, Australia, and their significance for dating the Baragwanathia flora. Proc. R. Soc. Viet. 79, 393-413. 1967. Preliminary stratigraphical results from graptolite studies in the Upper Silurian and Lower Devonian of southeastern Australia. J. geol. Soc. Aust. 14, 281-286. 1970. Remarks on the stratigraphy and morphology of Praguian and probably younger mono- graptids. Lethaia, 3, 173-182. STEIN, V. and wolfart, r. 1969. Fauna (Graptolithen, Brachiopoden) der unterdevonischen Schwarzschiefer Nord-Thailands. Neues Jb. Geol. Paldont. (Abh.) 133, 171-190. KOREN, T. N. and ENOKYAN, V. s. 1970. Silurian and Lower Devonian deposits from northwestern Yugorsky Peninsula and islands of the Pechora Sea. Naiik-Issled. Inst. Geol. Arktiki, Uclien'e Zapiski, Palaeont. Biostrat. 30, 5-25. [In Russian, with English summary]. KUHNE, w. G. 1955. Unterludlow-Graptolithen aus Berliner Geschieben. Neues Jb. Geol. Paldont. (Abh.) 100, 350-401. LEGRAND, p. 1965. Quelques nouveaux graptolites a la limite des systenies silurien et devonien au Sahara algerien. Mem. Bur. reck. geol. min. 33, 53-56. LENZ, A. c. 1967. Upper Silurian and Lower Devonian biostratigraphy. Royal Creek, Yukon Terri- tory, Canada. Int. Sympos. Devonian System, 2, 589-597. and JACKSON, d. e. 1971. Latest Silurian (Pfidolian) and Early Devonian Monograptus of North- western Canada. Bull. geol. Surv. Can. 192, 1-21, 2 pis. PRiBYi,, A. 1940. Die Graptolithenfauna des mittleren Ludlows von Bohem (oberes e/S). Vest. st. geol. List. 16. 1941. Uber eine neue Graptolithenarten aus dem bohmischen Obersilur. Mitt, tschech. Akad. 1L755. 51, 9 pp., 2 pis. 1942. Uber die stratigraphischen Verhaltnisse des Silurs und des Devons in der Podoler Zement- fabrik bei Prag. Rozpr. Ceske Akad. 52, no. 27, 14 pp., 1 pi. TELLER, L. 1964. Graptolite fauna and stratigraphy of the Ludlovian deposits of the Chelm borehole. Eastern Poland. Stadia Geol. Pol., 13, 1-88. • 1966. Two new species of Monograptidae from the Upper Ludlovian of Poland. Bull. Acad. pol. Sci. cl. II, 14, 553-558, 1 pi. 1969. The Silurian biostratigraphy of Poland based on graptolites. Acta geol. poL 19, no. 3, 393-501. URBANEK, A. 1958. Monograptidae from erratic boulders of Poland. Palaeont. pol. 9, 105 pp., pis. I-V. 1970. Neocucullograptinae n. subfam. (Graptolithina) — their evolutionary and stratigraphic bearing. Acta palaeont. pol. 15, 163-377, pis. 1-45. wiLLEFERT, s. 1962. Quelques graptolites du silurien superieur du Sahara septentrional. Bull. Soc. Geol. Fr. [7] 4, 24-40. • 1963. Graptolites du silurien et du Lochkovien de Touchchent (anticlinorium de Kasba-Tadla- Azrou, Maroc central). Notes et mem. Serv. geol. Maroc. 23, 69-98, 2 pis. DENNIS E. JACKSON Department of Earth Sciences Open University Bletchley, Bucks. ALERED C. LENZ Department of Geology University of Western Ontario Typescript received 12 November 1971 London, Canada LATE TRIASSIC PLANTS FROM THE CHINLE FORMATION IN NORTH-EASTERN ARIZONA by SIDNEY R. ASH Abstract. Three plants based on megafossils are described from the Late Triassic Chinle Formation at a new locality in north-eastern Arizona. They are the leafy shoot and cone of Selagmella anasazia sp. nov., the leafy and fertile branches of Dechellyia gormani gen. et sp. nov., a conifer of uncertain affinities, and Masculostrobus clathmtus sp. nov., a male coniferous cone. The cone is noteworthy because it contains pollen closely resembling the Late Triassic grains called Equisetosporites chinleana by Daugherty (1941) and later referred to the genus Ephedra by Scott (1960). This paper contains the results of a study of the recognizable species of plant mega- fossils recently obtained from the Late Triassic Chinle Formation at a locality near the mouth of Canyon de Chelly (locally pronounced de-shay) in north-eastern Arizona. Prior to this investigation little was known about the Late Triassic plants in the area. The first Late Triassic plant fossil to be reported from north-eastern Arizona and the entire south-west as well, was discovered on 4 September 1849 by Lieutenant J. H. Simpson (1850) of the U.S. Army. The fossil was a piece of petrified wood which Simpson found in a bed of conglomerate (now called the Shinarump Member of the Chinle Formation) exposed in the wall of a tributary canyon of Canyon de Chelly. Although the specimen was taken back to Washington, D.C. for study, it was never described and its present whereabouts is unknown. The only other reference to Late Triassic plant fossils in the immediate area is contained in a recent review of the Chinle flora (Ash 1972n). In that report a sketch was given of the fertile branch shown in Plate 118, figure 7 and a few remarks were made about it. Fossil plants have been found at many localities in the Chinle Formation since 1849 and a history of the search for them has been given recently (Ash 19726). As shown else- where (Ash 1972a), most of the localities have yielded only a few specimens of just a small number of species. Until now, the largest and most significant collections have been described from three localities — Arroyo del Cobre, New Mexico (Newberry 1876), Fort Wingate, New Mexieo (Ash 1970a) and Petrified Forest National Park, Arizona (Daugherty 1941, Ash 1970a, 19706). Publication of this report increases the number of significant localities to four; their locations are shown on the index map (text-fig. 1). LOCALITY The fossils described in this report were collected from the Monitor Butte Member of the Chinle Formation on the north bank of Chinle Wash at the mouth of Canyon de Chelly in north-eastern Arizona (see text-fig. 2). They occur in beds of mudstone and sandstone exposed near the west boundary of Canyon de Chelly National Monument in the N I, SW f, sec. 15, T. 32, R. 10 W. and about 70 meters west of the bridge over Chinle Wash. The fossil-bearing beds have been assigned U.S. Geological Survey (USGS) fossil plant locality number 10093 and Museum of Northern Arizona (MNA) locality number 200. [Palaeontology, Vol. 15, Part 4, 1972, pp. 598-618, pis. 114-119.] 113° ASH: LATE TRIASSIC PLANTS 109° 599 I -37° 33° 113° 109° 105° TEXT-FIG. 1 . Index map of part of the south-western United States showing the location of Canyon de Chelly National Monument in north-eastern Arizona where the plant fossils described in this report were collected. Also shown on the map are the other three places in the southwest where the Chinle Formation has yielded large numbers of leaf fossils which have been described in the past. The fossils described here include the leafy shoot and cone of Selaginella anasazia sp. nov., the leafy and fertile branches of Dechellyia gonnani gen. et sp. nov., a conifer of uncertain affinities, and Mascidostrobiis dathratus sp. nov., a male cone of possible coniferous affinities. The cone is of par- ticular interest because it contains pollen grains which are very similar to the Late Triassic grains that were first called Equisetosporites chinleaiia by L. H. Daugherty (1941) and then transferred to the genus Ephedra by R. A. Scott (1960). The beds also contain the remains of several other plants, but they are too poorly preserved and fragmentary to describe. One may be the pinwheel structure of Dinophyton spinosiis which commonly occurs in the Chinle Formation (see Ash 1970c). Others are fragments of large, linear, single- and multi-veined leaves. Portions of stems of several sizes are also present. The fossils described in this report have been deposited in the U.S. National Museum (USNM), Washington, D.C. Duplicates have also been deposited in the Museum of Northern Arizona, Flagstaff, Arizona. STRATIGRAPHY The Chinle Formation is about 400 meters thick in most of north-eastern Arizona (Repenning and others, 1969, fig. 6). In the vicinity of the fossil locality, however, erosion has removed much of the formation and only the lower 60 m have been preserved. It consists of the Shinarump Member (at the base) and the Monitor Butte Member at the top (text-figs. 2 and 3). The Chinle Formation is Rr C 9202 600 PALAEONTOLOGY, VOLUME 15 R. 10 W. EXPLANATION O) 4> C C 0) O o O O *53 *0 0- n \ Q°i I Alluvium UNCONFORMITY Chinie Formation "Rem, Monitor Butte Member "fics, Shinarump Member UNCONFORMITY 5 E Pde I De Chelly Sondstone Contact Dashed \«here approximately located Strike and dip of beds Number and location of measured stratigraphic sections shown on figure 3 TEXT-FIG. 2. Geologic map of the extreme western end of Canyon de Chelly National Monu- ment, Arizona, showing the location, near the mouth of Canyon de Chelly, of the beds containing the fossil plants described here. The base map was adapted from the U.S. Geo- logical Survey Chinie 4 NE topographic quadrangle and the geology is modified from Cooley and others (1969, pi. 1, sheet 5). underlain unconformably by the Dechelly Sandstone of Early Permian age and locally both formations are overlain unconformably by Quaternary alluvium. The Shinarump Member of the Chinie is about 1 6 m thick near the locality and usually forms rounded ledges and cliffs. It is composed mainly of reddish-brown, coarse-grained, crossbedded conglomeratic sandstone (see text-fig. 3) and is thought to have been deposited by ancient streams and rivers. In places, such as in nearby Cottonwood Canyon, the Shinarump contains petrified logs as much as a meter in diameter. Elsewhere the Shinarump also contains fossil leaves (Ash 1972fl). In the vicinity of the fossil locality the Monitor Butte Member is about 40 m thick. It consists mainly of dark red mudstone and several discontinuous beds of ripple-marked, fine to medium grained, sandstone that are often dark-red to light-grey in colour. The sandstone beds generally form promi- nent ledges while the mudstone usually weathers into steep, irregular slopes. Plant fossils occur in the lower part of this member in a unit that is approximately 2-3 m above the Shinarump and 20 m above the Dechelly Sandstone (see text-fig. 3). The plant-bearing unit is about 2 m thick and is com- posed of grey, fine-grained, flaggy sandstone and an underlying bed of massive grey mudstone. Rocks in the fossil bearing unit are relatively soft and usually form a slope below a prominent ledge of hard, brown sandstone. Fossils seem to occur throughout the unit wherever it is exposed over a horizontal distance of at least 50 m. ASH; LATE TRIASSIC PLANTS 601 EXPLANATION Mouth of Canyon de Chel I y TEXT-FIG. 3. Stratigraphic sections of the pre-Cenozoic rocks exposed near the mouth of Canyon de Chelly, Arizona. The localities where the sections were measured are indicated on text-fig. 2. SYSTEMATIC DESCRIPTIONS PTERIDOPHYTA Class LYCOPSIDA Family selaginellaceae Selaginella anasazia sp. nov. Plate 114; text-fig. 4 Holotype. USNM 168945. Paratypes. USNM 168903, 168909. 602 PALAEONTOLOGY, VOLUME 15 Distribution. This species occurs in the lower part of the Monitor Butte Member of the Chinle Forma- tion at the mouth of Canyon de Chelly, Arizona at USGS fossil plant locality 10093 and MNA locality 200. Derivation of name. The name is derived from the Navajo word ‘anasazi’, old or ancient people. Diagnosis. Shoots all (as far as known) leafy, axis dichotomous, but one branch often stronger and tending to continue in nearly the same direction, lateral branches continue to fork but axis becoming thinner and leaves smaller and more crowded. Angle of dichotomy 50°-80°. Larger stems 1 mm thick, substance dense, smaller ones less dense. Stems showing two closely placed steles. Each stele having 4-8 tracheids. Tracheids 8-18 /xm wide and up to 450 /xm long, walls scalariform, spiral tracheids also present. Ordinary epidermal cells of stem rectangular to wedge-shaped, about 8-16 gm wide, up to 60 /xm long, anticlinal walls usually straight, about 1 /xm thick. Stomata not noted on stems. Leaves in four ranks, two ventral ranks of large spreading leaves, two dorsal ranks of small leaves on top of stem, one dorsal leaf being at the same node as a ventral leaf. Large leaves ovate, 0-6-1 -0 mm wide, 1 -5-2-3 mm long, apices acute to mucronate, arising alternately at interval of about 1 -7-2-0 mm and at an angle of about 45°, spreading, a similar leaf also occurring in fork of each dichotomy. Leaf slightly asymmetrical with the midrib nearer the lower margin. Small leaves lanceolate, 0-4-0-6 mm wide, 1-0- 1-2 mm long, apices acute to mucronate, arranged in two ranks along dorsal side of stem, arising alternately at intervals of 1 -7-2-0 mm and at an angle of about 30°, some- what adpressed to stem, apices often overlapping large leaves. Shape slightly bent with the midrib curving to be parallel to the stem. Leaves of both sizes containing a conspicuous midrib up to 35 /xm wide. Midrib con- taining a vascular bundle composed of 1-5 rows of tracheids with scalariform thicken- ings. Vascular bundle departing from stele of stem without forming a leaf gap, extending to within 30-50 /xm of leaf apex, narrow and inconspicuous in leaf base, containing 1-2 rows of tracheids, becoming broad above leaf base, containing as many as 5 rows of tracheids. Margins of leaves flat entire, bordered with 4-10 rows of narrow, rectangular cells in a single layer, anticlinal cell walls thick (3-5 /xm), cells about 10 /xm wide, up to 50 /xm long. In apical region ordinary epidermal cells of one side nearly rectangular, about 4-25 /xm wide, 25-60 /xm long, anticlinal walls about EXPLANATION OF PLATE 114 Figs. 1-10. Selaginella anasazia sp. nov. 1-3, transfer, USNM 168945, apical regions of three large lower leaves. The several rows of thick-walled cells along the margins and the midveins show as dark zones in the photographs. Note that the epidermal cells generally are rectangular and occur in longitudinal rows. All X 30. 4, transfer, USNM 168946, basal portion of a large lower leaf show- ing the spindle-shaped ordinary epidermal cells that prevail in this region. Compare the straight side walls of these cells with the sinuous side walls shown by the ordinary cells in the apical regions of the leaves in figures 1-3, X 30. 5-7, leafy shoots showing the growth habit of this plant, all X 1. 5, USNM 168910. 6, USNM 168909. 7, USNM 168908. 8-10, Stomata. The stomata in figures 8-9 are beside the midvein in the central part of a leaf whereas the stomata in figure 10 is near the upper end of a midvein in the apical region of a leaf. Note that the guard cells clearly overlap the ordinary epidermal cells in these examples. Compare the thick anticlinal walls of the marginal cells to the left in figure 10 which the thin anticlinal walls of the other cells elsewhere in the figure and in figures 8-9. Tracheids of the midveins are visible in places in figures 8-9. All X400. 8-9, transfer, USNM 168947; 10, transfer, USNM 168945. Palaeontology, Vol. 15 PLATE 114 ASH, Selaginella f . ' ■ \ i \ ? c • •-■ Vir- 'j 5 ■mm •> ASH: LATE TRIASSIC PLANTS 603 1-2 jjim thick, side walls markedly sinuous (maximum sinus amplitude about 3 /xm) end walls straight to slightly sinuous, anticlinal walls becoming progressively less sinuous towards leaf base. Ordinary cells of the other side of leaf elongated with straight to wavy (rarely sinuous) walls. Ordinary epidermal cells of both sides gradually become spindle-shaped with more or less straight anticlinal walls in basal part of leaf, remaining about the same size as those in apical part. Periclinal walls of all epidermal cells flat, smooth. All epidermal cells arranged in distinct longitudinal rows, leaf widening by addition of rows of ordinary cells interior to marginal cell rows. Stomata few, occurring in rows a single stoma wide on either side of midrib, or directly under midrib, stomatal rows not sunken, individual stoma separated by 2-6 ordinary epidermal cells, about 6-10 per leaf. Guard cell pair oval, usually longitudinally oriented, 140-160 p.m long, 80-100 iu.m wide, aperture elliptical about 70 pm long, guard cells overlapping epidermal cells to a small extent. Mesophyll probably represented by darker strips beside the midrib. Ligule forming a dark area at leaf base about 0-3 by 0-2 mm broad (outline and cells not observed clearly), base of ligule dense. Cone known to be at least 3 mm long, 1-5 mm wide, sporophylls not forming an enlarged ventral set but all alike, megasporophylls at base, ovate, about 1 mm wide, 2 mm long, pointed, texture resembling a leaf. Megasporangia not preserved but still containing at least 3 of the spore tetrad. Megaspores about 200 pm in diameter, wall thick apparently pitted but details not clear. Microsporophylls not observed. Discussion. S. anasazia is not very common and occurs only in the lower part of the fossiliferous zone at locality 10093. It is preserved in a fine grained slightly calcareous mudstone and the plant substance is rather dark. Although the fossil shows the epidermal cells rather indistinctly to direct observation, transfers show them very clearly in both the leaves and the stems. Remains of four cones were recognized; all are poorly pre- served but clearly are attached to the ends of leafy shoots. Some leaves in transfer show one set of cells (with coarsely sinuous walls) clearly and the other set of the opposite epidermis (with straight to wavy walls) only obscurely or scarcely at all. Other leaves, however, show both sets clearly. I am not sure which is the upper epidermis in any of the specimens. However, in certain living species of Sela- ginella, such as S. marteusii and S. prodiicta, the upper sides of the leaves have straight to wavy walled epidermal cells and lack stomata, whereas the lower sides have sinuously walled epidermal cells and bear a few sparse stomata over the midrib. Thus, I think that the side of the leaf in the fossil that bears stomata and shows sinuously walled epidermal cells is the lower. I cannot explain why only one epidermis is preserved on some of these leaves but a similar type of preservation was shown by some specimens of the fern Cladophlebis daughertyi (Ash 1970^?, p. D43) which was also described from the Chinle Formation at other localities in the south-western United States. If S. anasazia has a cuticle at all it must be exceedingly delicate, and this is ordinarily taken as there being none. Although the epidermis is beautifully preserved and excellent acetate transfers were obtained and gently macerated in HNO3+KCIO3 for half an hour, all organic matter dissolved when the transfer was placed in NH4OH. The broad, spreading leaves of the species described here are similar to the leaves of many living selaginellas, such as S. krausiana and S. abyssinica, which typically live in a fairly moist, and shady habitat. On the other hand, they contrast with the tiny, 604 PALAEONTOLOGY, VOLUME 15 TEXT-FIG. 4. (See opposite). ASH; LATE TRIASSIC PLANTS 605 Strongly appressed leaves of several species, such as S. pilifera and S. mutica which inhabit a much more arid environment. This suggests that S. anasazia lived in a moist environment. In most places the stems show two steles clearly but below a dichotomy there seems to be three and at some places only one was seen. The vascular bundles in a few leaves clearly appear separate from stem vascular bundles in transfers (see text-fig. 4A, 4B) but the gap is rather small (perhaps OT mm). In others, however, it is continuous. In these the veins are quite narrow and delicate in the leaf base and expand upward as commonly occurs in many living species of Selaginella (Harvey-Gibson 1897, p. 152). New features for a fossil Selaginella presented by S. anasazia are firstly the cellular details of the leaf and stem, and secondly the ligule, although this is seen imperfectly. The details of the cone are not very good. We have, however, apart from the missing cone details, remarkably close agreement with many living species. This species raises the question of the separation of Selaginellites from Selaginella acutely. If all fossil plants or at any rate all Mesozoic and Paleozoic ones are to be separated (on age) as a convention, there is no problem. But if a genus is only held to be valid if it has clear morphological distinctions from other genera then the fossil described here would be placed in Selaginella. It is not differentiated by any known character and agrees in a large number with Recent species. Admittedly the cone is only poorly known, but so far as it goes, it is typical and a better specimen may well remove any doubt and also provide good spores. We do not have its rhizophore (a part which is likely to be missing from the detached fragments found as fossils) and of course we have no information about gametophytes and other soft parts and this we may never have. Certain authors, however, have held that knowledge of a fossil can never be complete and this, in effect, is a difference. Thus Seward (1910), has used Selaginellites in pre- ference to Selaginella, Lycopodites instead of Lycopodium, etc. More recently Chaloner (1967) followed this principle and transferred several fossil species from Selaginella to Selaginellites in volume two of the Traite de Paleobotanique. Florin (1936) considered the problem when he was working with fossil ginkgos. He took the position that while Ginkgo was suitable for fossil leaves which in his TEXT-FIG. 4. Selaginella anasazia sp. nov. A, Portion of a leafy shoot showing a leaf in the fork of the bifurcated axis. Transfer, USNM 168947, X 5. B, Portion of a leafy shoot showing three large lower leaves, three small upper leaves, and the distribution of vascular tissue in the leaves and shoot axis. The remains of the ligules are indicated by the dark oval areas in the leaf bases, transfer, USNM 168948, X 20. C, Epidermal cells on the shoot axis, transfer, USNM 168947, X 100. D, Leafy shoot, transfer (holotype), USNM 168945, X 5. E, Epidermal cells in the apical region of a leaf. The dense stippling along the margin shows the distribution of thick-walled cells in this part of the leaf. The end of the mid-vein and a stoma is also shown. Transfer, USNM 168945, X 100. F, Irregular epidermal cells between the midrib (to the left) and the margin (to the right) in the basal portion of a leaf, transfer, USNM 168946, X 100. G, Epidermal cells near the edge of leaf. The rectangular, thick-walled cells at the left are along the margin and contrast with the irregular, thin-walled cells elsewhere on the lamina of the leaf, transfer, USNM 168947, x400. H, Apex of a leaf, transfer preparation, USNM 168945, X400. I, One of the spores in K showing some of the pitting on the walls. Transfer, USNM 172282, X 200. J, Reconstruction of a leaf bearing three spores. Based on a transfer, USNM 172283, X 10. K, The remains of three spores. The stippled area shows pitting on the spore wall. Transfer, USNM 172282, X 10. L, Two stomata on the epidermis of a leaf. The midrib is to the right, transfer, USNM 168947, X400. 606 PALAEONTOLOGY, VOLUME 15 personal opinion showed morphological points of agreement with the leaves of G. biloba’ he used Ginkgoites for species where the form, or the fine structure, were different enough to leave him (again personally) in doubt. He also used it where he had doubt of another kind, that is where some sort of the necessary evidence, such as the fine structure, was not available. He attempted no definition of how great a difference should be or what sort to cause him to doubt generic identity. At one time Harris (1941, 1945) followed Seward and used Equisetites for Jurassic fossils he considered to be like Equisetum. Recently (1961), however, he has changed his opinion and reassigned some of the fossils he had called Equisetites to Equisetum because he could not show a real morphological difference between them and the Recent genus. Since I cannot show any real morphological differences between the fossils described here and the Recent Selaginella, 1 refer them to that genus. Comparisons. A considerable number of Paleozoic and Mesozoic fossils have been described as species of Lycopodites, Selaginella, and Selaginellites. Many are probably twigs of conifers and are not considered further, but others show important points of resemblance to Selaginella. Not one of these shows a full set of characters — form of shoot, leaf, cone, mega- and microspores, cellular details of leaf and stem, the ligule and rhizophore (the two last organs are not clearly shown in any description of a fossil species known to me). The better known species are not satisfactorily comparable, for in some the shoot, cone, and spores are well known but there is little if any infonnation on the fine structure of stem or leaves. S. anasazia on the other hand has a poorly known cone but the fine structure of the stem and leaves is exceptionally well known. One species that resembles S. anasazia in general form is L. macrophyllous Golden- berg, 1855 (see also Halle 1908). It, however, can be easily distinguished by its large leaves (twice the size of those of S. anasazia). L.falcatus Lindley and Hutton from the Middle Jurassic of Yorkshire (see Harris 1961) agrees in size of the foliage and in its anisophylly, but the small leaves are not arranged as in Selaginella and there also are differences in the leaf epidermis. For example, the stomata are wide spread on the leaves of L. falcatus, whereas they are highly localized in S. anasazia (just over and near the midrib). Selaginellites nosikovii Kryshtofovich and Prinada (1932) from the Jurassic of Siberia also agrees in some features but it has lower leaves that are much longer than in S. anasazia and that point at right angles to the stem, not at a much lower angle as in the Arizona fossil. The most similar species known to me are L. scanicus Halle (see Lundblad 1950) from the Rhaetic of Sweden. The shoot and leaves, however, are more crowded and their margins are microscopically denticulate instead of entire. Three species of what are believed to be Lycopods are already known from the Chinle Formation, namely Lycostrobus chinleana Daugherty, Chinlea campii Daugherty em. Miller (1968), and Lycopodites (?) sp. Daugherty. All of these represent far larger plants with nothing in common with the delicate little fossil described here. Spores named Lycospora sp. by Peabody and Kremp (1964) have been described from the Chinle Formation but there is no reason to identify them with S. anasazia. Full comparison with the species of the huge living genus Selaginella would be a formidable task as a large number of the species growing in the dark and shade have shoots looking much like S. anasazia. The familiar S. krausiana for instance agrees in ASH: LATE TRIASSIC PLANTS 607 the appearance of its shoots in the general form of the leaves and even in several of the microscopic characters of the leaves, but there are also small differences. For example, the dorsal leaves are of slightly different shape and S. anasazia leaves lack the marginal teeth of the leaf of S. krausiana. GYMNOSPERMOPHYTA Class CONIFEROPSIDA Order coniferales. Family? Genus Dechellyia Ash, gen. nov. Type species — Dechellyia gormani Ash, sp. nov. Diagnosis. Leafy shoot and foliage shed in one piece, main axis bearing opposite branches. Branches bearing linear foliage leaves and small clasping scale leaves. Foliage leaves persistent, opposite, arising in two decussating pairs but flattened into the horizontal plane, confined to upper two-thirds of axis, bases evenly contracted but not forming distinct petioles, lamina containing a broad midrib and several narrow parallel strands. Scale leaves persistent, apparently opposite, abruptly giving place to foliage at end of lower third of branch axis. Fertile branches borne on same shoot as foliage. Fertile axis often elongated, other forms with one or more elongated lateral branches or with several very short lateral branches. Fertile axis covered with small outgrowths (presumed to be reduced leaves) along its margins (and surface?), at its end bearing a pair or two pairs of sporophylls. Each sporophyll comprising a single basal seed sur- mounted by a lanceolate lamina. Lamina containing two broad parallel ribs and two narrow strands, all apparently ending separately at apex. Sporophylls with their seeds readily detached. Derivation of name. The name is derived from Canyon de Chelly, Arizona, at whose mouth the type species was collected. Dechellyia gormani Ash, sp. nov. Plates 115-118; text-figs. 5, 6 A-C 1967 Samara, Ash, p. 130, fig. 3a. 1972fl New genus A, Ash, p. 19, fig. 4. Holotype. USNM 168919. Paratypes: USNM 168915, 168907. Distribution. Dechellyia gormani occurs in the lower part of the Monitor Butte Member of the Chinle Formation at the mouth of Canyon de Chelly, Arizona at USGS fossil plant locality 10093 (also MNA locality 200) and in the Fort Wingate area. New Mexico at USGS fossil plant locality 10060. Derivation of name. The species is named in honour of Mr. David Gorman of Canyon de Chelly National Park who told me about the presence of fossil plants at the mouth of the canyon. Diagnosis. Main axis 2-5 mm wide, repeatedly branching. Leafy branches 9-25 cm long, foliage 2-10 cm wide, axis 1-2 mm wide, straight, furrowed, bearing many spreading, outward directed, typically straight to curving, linear persistent foliage leaves above and small clasping leaves below. Foliage leaves projecting at an angle of 40°-70° to the stem, often at a lower angle near apex and a higher angle near base, L2-4 mm wide, 1-7-5 cm long, margins nearly parallel except near base and apex, apex acutely pointed. 608 PALAEONTOLOGY, VOLUME 15 basal 2-4 mm evenly contracted to a width of 0-5-1 mm. Midrib about 200-300 wide. Strands four (rarely five), about 50/i.m wide, typically two on either side of midrib, more or less parallel except in leaf base where they originate. One member (midrib marginal strand) of each pair is very close to vein, the other (marginal strand) is about 100 /Ltm from the margin. Each pair arising by the forking of a single basal strand which enters the leaf base beside the midrib, each strand apparently ending separately near leaf apex. Occasionally marginal strands bifurcate once near leaf base, outer member of the pair running parallel to other veins for a short distance and then disappearing near leaf margin. Scale leaves acutely pointed, about 4 mm long, 1 mm wide at base, extending parallel to axis and situated close to it. Fertile axis bearing along its length spine-like reduced leaves about 1 mm long. Sporophyll consisting of seed attached to axis at its base and continued above into a symmetrical lanceolate lamina. Seed oval, about 7 mm long, 3 mm wide, firmly attached to lamina, made up of a (hollow) interior about 2-5 X 1-5 mm, enclosed in a shell nearly 0-5 mm thick and surrounded by a delicate flange 0-2-0-3 mm wide which continues the lamina downwards. Lamina 3-4-2 cm long, 7-10 mm wide, apex obtusely rounded, lower part narrowed gradually to the seed, containing two broad ribs and two narrow marginal strands. Ribs about 300 /xm wide, more or less parallel except in base where they are close and in apical region where they converge slightly, ending separately in apical region, generally about 1 -5-2-0 mm apart. One narrow strand about 50 jum wide arising from the basal area of the lamina on either side of the pair of ribs, running parallel and close to lateral margins of lamina, apparently ending separately in apical region. EXPLANATION OF PLATE 115 Figs. 1-5. Dechellyia gonnani gen. et sp. nov. illustrating the variation in the size of the leafy branches and the foliage leaves. All X 1. 1, USNM 168948, portion of a large leafy branch bearing very large foliage leaves. The two midrib marginal strands are clearly visible as narrow dark lines in many of the leaves. The midrib itself is represented by a wider clear band between the two strands. 2, USNM 168926, the upper portion of a leafy branch bearing foliage leaves of typical size. Here the midribs and associated strands are often represented by a clear band where the carbonaceous material has fallen off. 3, USNM 168949, the central portion of a leafy branch bearing several foliage leaves that show swellings. The portion of the leaves above the swellings is broken off in most of these cases. Exceptions are the two near the top of the branch and to the right which are bent downward at the swellings. Compare figs. 1, 2, 5, Plate 116. 4, USNM 168921, lower part of a leafy branch showing typical foliage leaves and several scale leaves on the lower part of the axis. 5, USNM 168941, the upper portion of a leafy branch. Here once again the midrib and associated strands are represented by a broad clear band. EXPLANATION OF PLATE 116 Figs. 1-5. Dechellyia gonnani gen. et sp. nov. All X 1 . 1 , USNM 1 68942, the base of a specimen show- ing unusually small foliage leaves. Scale leaves are visible along the axis below the foliage leaves. 2, USNM 168950, a leafy branch bearing a number of foliage leaves that show swellings. Note that only about half of the leaves show this feature and that the swellings all occur about 1 cm from the branch axis. 3, USNM 168951, portion of another leafy shoot that bears many leaves showing swellings. Note that several of the leaves on the right side of the stem are bent at the swellings whereas the others are broken off. 4, USNM 168952, bifurcated shoot axis. Scale leaves are visible along the axes of several of the leafy branches. 5, USNM 168953, a specimen in which about half the leaves have swellings. Note that the sixth leaf from the top on the right side of the axis has two swellings. Typically the swellings occur about 1 cm from the axis. Palaeontology, Vol. 15 PLATE 115 ASH, Dechellyia Palaeontology, Vol. 15 PLATE 116 ASH, Dechellyia ASH: LATE TRIASSIC PLANTS 609 Preservation. To the naked eye Dechellyia looks well preserved but unfortunately destruction has occurred at some stage. The leaves are black to dark brown and the fossil substance has often adhered to both sides of the rock (usually with more on one than the other) causing numerous gaps. Then the plant substance has often contracted into little polygonal blocks about 10 /xm broad and they are separated by narrow gaps. TEXT-FIG. 5. Dechellyia gonuani gen. et sp. nov. A, penultimate shoot. An axis bearing the remains of two sporophylls is to the left, the naked branch below the axis may be either another fertile axis or the base of an ultimate leafy branch. Drawn from USNM 168915, X 1. B, reconstruction of a fertile branch, xl. This, of course, precludes making a typical cuticle preparation. Even where this has not happened the substance adheres firmly to the rock and needs transfer treatment to remove large areas. The transfers give little additional information even when they are macerated or bleached. Thus practically all that is known about Dechellyia is what the specimens show to the naked eye or at a low magnification. We have no knowledge at all of the fine structure beyond what is shown by surface observation. It is particularly regrettable that we have no fine structure of the body here regarded as a seed. Discussion. The organization of the foliage and the fertile organ is by no means certain. For the purpose of description the following assumptions which seem to be the simplest possible have been made. I call the simple spreading organs of the foliage ‘leaves’ or 610 PALAEONTOLOGY, VOLUME 15 ‘foliage leaves’ and the smaller ones below ‘clasping’ or ‘scale leaves’. They are borne on what I term a ‘leafy shoot’. Admittedly, the spreading organs might have been described as ‘pinnae’ or ‘reduced pinnae’ on the rachis of a pinnate leaf. Whether leaf or leafy shoot it clearly is of strictly limited growth. It seems likely that the part which bears the leafy shoots is itself a stem. For purpose of this description I have assumed that we are dealing with deciduous leafy shoots because isolated leaves are rarely found. If this is so, then the shoots are rather like those of such conifers as Taxodium and Metasequoia although the fructi- fications are quite dilferent. On the other hand if they are simply pinnate leaves borne on rather slender stems then they would be like the Triassic Pteridosperms Stenopteris and Xylopteris. No little clasping scale leaves, however, are seen in those plants. Several forking stems are known (see PI. 117, figs. 13-15). The branches in most are short and do not bear foliage leaves. One large specimen, however, has branches bearing typical foliage leaves (PI. 117, figs. 13-14). Most of the forking stems also show a short stump in the angle of the fork. Probably they represent the uppermost portion of a main stem and the stump is the apex of the main axis as in many modern trees and shrubs. A small percentage (perhaps 5%) of the leafy shoots have leaves showing conspicuous swellings (see PI. 116, figs. 2, 5) and some show them on nearly all leaves (see PI. 115, fig. 3, pi. 1 16, fig. 3). Most leaves have only one swelling but a few have two (PI. 1 16, fig. 5). Very often the leaf is broken off or at least its direction changes at the swelling. Most swellings occur about 1 cm from the leaf base and maybe on the side of the leaf facing the stem apex or the side facing the stem base or in about the middle of the leaf. Many are oval and broader (about 3 mm) than the leaves on which they occur, but smaller and round ones are also known. All of the swellings bulge strongly and are composed of a good deal of brittle coaly matter. On one specimen there appears to be a series of EXPLANATION OF PLATE 1 17 Figs. 1-15. Dechellyia gonnaui gen. et sp. nov. All X 1. 1-11. Detached sporophylls. 1, USNM 168918, the largest specimen. 2, USNM 168943, the lamina and seed in this rather large specimen are not connected because the rock was broken in collecting. The two ribs in the lamina are clearly visible as dark lines in some places. 3, USNM 168917, a fairly complete example of a detached sporophyll in which the ribs, strands, and lamina apex are fairly well preserved. 4, USNM 168920, the remains of a sporophyll which has an unusually broad lamina. 5, USNM 168923, a typical specimen. 6, USNM 168939, a complete example of a small sporophyll with a narrow lamina and the basal portion of another sporophyll. 7, USNM 168911, lower part of a specimen in which the seed has not been compressed and stands about 0-5 mm above the level of the lamina. 8, USNM 168940, the apical constriction of the lamina in this very small example is probably due to an accident of preservation. 9, USNM 1689 12 A, basal part of a specimen in which the hollow interior of the seed is filled with sediment. The shell is preserved as a small ridge around the interior filling and a narrow band of lamina appears to enclose the seed. 10, USNM 168913, the remains of a fairly large megasporophyll. 1 1, USNM 168944, the marginal strands show fairly clearly as dark lines in this specimen. In places strands are preserved near the margins of the lamina. 12, USNM 168922, a fairly complete shoot. 13, USNM 168927A, the fork of a large shoot axis. There is a slight indication of the stump between the two leafy branches. 14, USNM I68927B, counterpart of the lower part of the specimen in 13. The stump is clearly visible between the branches. 15, USNM 168914, the fork of a small shoot axis. A low stump is visible between the two branches. Palaeontology, Vol. 15 PLATE 117 ASH, Dechellyia ASH; LATE TRIASSIC PLANTS 611 radiating ridges and grooves. Maceration of the coaly matter and of transfers of the swellings yielded no useful information. The swellings are here described as galls caused by some parasite rather than as reproductive structures. I have used the following terminology in referring to the several linear structures shown in the foliage leaves and the lamina of the structures I call sporophylls. The broad (about 200-300 fim) central structure in the foliage leaves is termed a midrib whereas the narrow (about 50 /xm) structures are called strands with the ones that are close to the midrib being designated midrib marginal strands and those near the lamina margins, marginal strands. The two broad structures in the laminae of the sporophylls are called ribs and the two narrow structures near the margins of the lamina are called marginal strands. The marginal strands in both the foliage leaves and the sporophyll laminae look just like the midrib marginal strands but may be slightly narrower. They are composed of a somewhat coherent brown material which occasionally can be detached, bleached and mounted on a glass slide. They are then seen to be made up of rectangular parenchyma cells in longitudinal files. Neither tracheid thickenings or elongated thick-walled cells such as fibres were seen. Frequently the midribs of the leaves and the ribs in the laminae of the sporophylls are almost black although in other cases they are only slightly darker than the lamina. Its substance has decomposed into little blocks separated by narrow gaps. These blocks tend to form longitudinal files which probably represent elongated thick-walled cells. Nothing like tracheid thickenings could be recognized in these structures. The lamina is the palest part and again its substance has disintegrated into separate little blocks. For the most part these are uniformly scattered but at some points they are broader than long and tend to form transverse files. These blocks are more visible under low magnification than high. A possibility is that; (1) The midrib in the leaves and the ribs in the sporophyll laminae are the only vascular strands. (2) The midrib marginal strands and the marginal strands are resin ducts, the resin forming the coherent brown matter. Or they may be supporting tissue but if so it is strange that nothing like a fibre was seen. It is even possible they represent air canals in a water plant but though the cells seen do seem to be of reasonable shape, they are unlikely to be preserved as a dark, coherent strand. Again, I cannot be sure that the exposed oval bodies looking like seeds are indeed seeds because they have not been shown to have the fine structure of gymnosperm seeds. Also, I cannot be sure of their exact relation to the sporophylls as I have not yet found a specimen showing the relations clearly. Nevertheless, the gross characteristics of the bodies suggests they are seeds. For example, the shell of the body is substantially thicker than the carbonaceous film of the sporophyll lamina indicating that it originally was much bulkier and thicker than the lamina. Several specimens show an oval gap in the carbonaceous material where the shell has fallen off (see PI. 118, fig. 9). In some there is either a depression or mound showing as much as 0-5 mm relief at the site of the body (see PI. 117, fig. 7). All of this evidence is consistent with the view that the body is truly a seed. 612 PALAEONTOLOGY, VOLUME 15 Comparisons. I know of no plant, fossil or living, that exhibits the combination of characteristics shown by Dechellyia. Probably the most outstanding are the winged structures thought to be sporophylls. Some fossil plants and many living trees have some sort of a wing attached to a seed or fruit that aids wind dispersed. For instance, Fraxinus, Acer, and Tilia have such an appendage although its morphology is entirely different. Among fossils there are Dioonitocarpidhim, Cycadocarpidium, Fraxinopsis, and Dinophyton. Dioonitocarpidiiim Lilienstern (1928) is from the Late Triassic of Bavaria. It has a narrow rachis bearing four rows of sickle-shaped pinnae and two basal bodies thought to be seeds. The fossil described here has no pinnae and a single seed. Cycadocarpidium Nathorst (1886) from the Late Triassic of Sweden, Russia, and Greenland is more similar. It has a broad lamina containing several (4-10) parallel veins and two seed-like bodies which are borne on a small separate, basal, ovuliferous scale. Evidently it is dis- tinct from the sporophyll of Dechellyia which has only two ribs and a single seed-like body imbedded in the base of the lamina. Fraxinopsis Wieland (1929) from the Middle Triassic of Argentina and Australia is still more similar. Here again the lamina is broad but it contains several (7 or more) parallel veins. There are, however, two bodies thought to be seeds imbedded in the base of the lamina not one as in Dechellyia. The fructification of Dinophyton, Ash 1970c (from the same flora as Dechellyia) differs greatly in having four wings but in view of a certain similarity in its shoot, is referred to again. It is possible to make some comparison between Deehellyia and the conifer Podo- zamites because the shoots of Podozamites often bear basal scale leaves and the foliage leaves may be in two lateral ranks but normally they are not opposite in decussate pairs. In all species of Podozamites, however, even ones with very narrow leaves, there are several veins and these converge towards the apex, as they do in other conifers with several veins. In an earlier report (Ash 1972a) Dechellyia (though not named) was tentatively com- pared with Podozamites arizonicus Daugherty because the leafy shoots are superficially similar. A recent study of the cotypes of P. arizonicus shows, however, that they actually EXPLANATION OF PLATE 118 Figs. 1-8. Dechellyia gormani gen. et sp. nov. All X 1. 1, USNM 168928, pedicel bearing the bases of several sporophylls. Two of the organs are attached to the end while the others are attached alternately to the sides of the structure. 2, USNM 168907A, sporophyllus organ composed of four sporophylls attached to an unbranched spiny pedicle. 3, USNM 168907B, counterpart of the specimen in 2. 4, USNM 168929, lower part of a fertile branch bearing two empty pedicels. 5, USNM 168916, upper portion of a leafy branch axis and the detached end of a fertile branch bearing the bases of two sporophylls which are attached to the end of a pedicel. 6, USNM 1 68924A, lower part of a fertile branch bearing a fragmentary pedicel on the left and the right. Excavation on the counterpart has shown that the central linear structure is connected to the left pedicel near the fork. That structure may be another pedicel or the axis of a leafy branch. 7, USNM 168919, holotype, a leafy branch bearing two pedicels. The pedicel on the right bears the seedlike body of a sporophyll; the other pedicel bears two complete sporophylls. One of these sporophylls (on the right) is flattened in the same plane as the leafy axis whereas the other is more or less perpendicular to the plane and shows as a dark line in the photograph. The line actually is the edge of a sporophyll as excavation on the specimen has shown. Scale leaves are fairly obvious on the branch axis between the foliage leaves and the pedicels. 8, USNM 168925 A, sporophyllus organ consisting of a branched pedicel bearing one sporophyll on one branch and two on the other. Palaeontology, Vol. 15 PLATE 118 ASH, Dechellyia ASH; LATE TRIASSIC PLANTS 613 are distinct. One of the outstanding differences involves the venation. In P. arizonicus the leaves contain 10 or more veins, as is characteristic of the genus, and thus contrast with the leaves of Dechellyia. No firm classification for Dechellyia is offered but since the shoots do recall those of several conifers I think it may be a conifer. The fructification is plainly not like that of any conifer family yet described. Genus masculostrobus Seward 1911, em. Barnard 1968 Masculostrobiis clatliratus sp. nov. Plate 119; text-fig. 6 D-K Holotype. USNM 168984. Paratypes. USNM 168956, 168955, 168954. Derivation of the name. The specific name is derived from the Latin, ‘clathratus’, latticed, and refers to the lattice-like appearance of the ektexine of the pollen that occur in these cones. Diagnosis. Cone cylindrical, shortly stalked; length about twice the diameter, length 9-18 mm, diameter 4-8 mm; stalk smooth, 2 mm long by about 1 mm wide. Cone axis bearing spirally arranged fertile appendages; axis containing 4-6 narrow strands of parenchyma (c. 50 gm wide) and associated tracheids. Appendage consisting of a broad head and a slender stalk. Stalk 0-5 mm wide by about 2-3 mm long, attached at 90° to the cone axis in the mid region of the cone; containing two narrow parenchyma strands. Head a depressed oval in outline, width exceeding the height, about 2-2-5 mm wide by T7-2-0 mm high; apex acuminate, pointing upwards; outer surface of head normally convex, inner surface concave, rounded into the expanded end of the stalk; stalk attached to the lower part of the inner surface of the head. Heads occasionally inverted with outer surface concave. Parenchyma strands in appendage arising from those of the cone axis by forking, disappearing in upper part of head. Head bearing 5 pollen sacs. Pollen sacs elliptical, attached to the lower edge and inner surface of the head, fused above, distal ends hanging free; walls poorly preserved. Pollen masses ellipsoidal, about 400-550 gm long, 200-300 pm wide, containing numerous pollen grains. Epidermal cells of head (outer surface?) isodiametric, irregularly rectangular, anticlinal walls thick, average with 8 ftm (range noted 6-10 pm). Pollen grains round to elliptical in distal view, about 48-60 pm X 24-60 pm, consisting of a silicate endexine and a ridged ektexine. Endexine thin (about 1 pm thick), smooth, sulcus extending full length of grain, greatest width near middle, narrowing at ends, lips narrowing at ends, lips narrow. Ektexinous ridges 12-25, alternating with narrow furrows, extending longitudinally, often with a slight spiral. Ridges about 2-5 ;Ltm wide, 1 pm thick, semicircular in cross section, converging near the ends of the major equatorial axis, usually forming four distinct areas of convergence. Areas of convergence opposite to sub-opposite at the ends of the equatorial axes, separated by 1-3 ridges that extend with a slight spiral almost around the grains in the direction of the equatorial axis. Ridges sometimes joined near the ends of the major equatorial axis, in some grains partially reflexed by extensions of sub-opposite convergence points past the midline of the grain. Ridges marked on their inner side by granules called ‘columellae’. Rarely all bands fuse at the ends of the major equatorial axis to form a single small area. 614 PALAEONTOLOGY, VOLUME 15 Discussion. M. dathratus is fairly common at locality 10093 being represented by 40 more or less good specimens. As far as can be seen, they are all very much alike. The heads of the appendages usually are preserved as a coaly structure showing no surface details of interest and no remains of any coherent surface membrane or cuticle. In these, the substance has cracked into tiny pieces and on maceration it gives no useful result. Transfers do give a little additional information and a dozen or so were made. Epidermal cells were observed on two heads in one cone that had been transferred. Some of these cells are shown in text-fig. 6K, but I do not know whether they come from the outer or inner surface of the heads. Other transfers showed masses of pollen and some of the internal structure of the cone axes and appendage stalks (see text-fig. 6H). Pollen sacs are preserved at or near the lower edges of several of the heads (see text- figs. 6D-6H) and these show many pollen grains but no definite details of the wall tissue. Pollen is fairly numerous in some of the sacs but does not form a dense mass. It looks as though the sac had opened but many grains had stuck to the wall. Although they often look well preserved in the transfers, the pollen grains contained in the cones were entirely destroyed by maceration. The cone axes contain several narrow coherent strands coipposed of parenchyma cells. The strands fork occasionally and a pair of branches from adjacent strands enters each appendage stalk (text-fig. 6H) and then passes into the appendage head where they disappear some distance below the upper margin (see text-figs. 6E, 6F). These strands closely resemble the strands that occur in the leaves and sporophylls of Dechellyia and as in that plant their function is uncertain. In contrast to DecheUyia, however, tracheids are preserved in the cone axes and appendage stalks in association with the strands. Comparisons. I find it difficult to compare this cone because I know so little about its gross organization. However, what little is known suggests that it does not compare very closely with any of the fossil cones which have been described in detail during the past. This is emphasized when the sacs on the appendage heads of M. dathratus and the pollen grains they contain are considered. The pollen grains contained in these cones agree with the microfossil that was described from the Chinle Formation as Equistoporites diinleana by Daugherty in 1941. It was originally interpreted as being a round spore wrapped in two elaters resembling the spores of the living Equisetum. I have examined Daugherty’s specimen and it seems to me that the supposed elaters are merely ektexine ridges somewhat similar to those of the present grains and of the pollen of some living species of Ephedra, as Scott noted EXPLANATION OF PLATE 119 Figs. 1-22. Masciilostrobus dathratus sp. nov. 1-8, pollen grains. The dark area in each specimen is the endexine of the grain and the narrow indistinct, light coloured bands are the ektexinoiis ridges. In figure 1 the ektexinoiis ridges are some distance from the endexine and a convergence area is clearly shown. In the other grains the convergence areas are not as prominent and the ektexinoiis ridges frequently cross each other. Note the prominent sulcus in the elliptical grains in figures 6 and 8 and the apparent absence ofthis feature in the oval grains in figures 2-5. All x400. 1, 5-8, transfer, USNM 168957. 2-4, transfer, USNM 168954. 9-22,cones. 9, USNM 168932, X 2. 10, USNM 168955, X 1 . 1 1, USNM 168955, X 2. 12, USNM 168956, X2. 13, USNM 168931, X2. 14, USNM 168930, x2. 15, USNM 168933, x2. 16, USNM 168936, X 2. 17, USNM 168938, X 2. 18, USNM 168937, X 1. 19, USNM 168935, X2. 20, USNM 168937, X2. 21, USNM 168936, X2. 22, USNM 168934, x2. Palaeontology, Vol. 15 PLATE 119 ASH, Masculostrobus ASH; LATE TRIASSIC PLANTS 615 TEXT-FIG. 6. Decliellyia gormani gen. et sp. nov., (A-C), and Masculostrobus clathratus sp. nov., (D-K). A-B, bases of two leaves showing the midrib (heavy stippling) and several strands (broad solid or broken lines). Note that two strands enter the bases of the leaves on either side of the midrib and that each strand bifurcates a short distance above the leaf base. The branch which becomes the marginal strand on the right in B is exceptional as it bifurcates once a short distance above its base so that there are five strands in the lower part of this leaf. A dashed line is used where the strands are poorly preserved. A, transfer, USNM 172286, X 5. B, transfer, USNM 172285, x5. C, the apical region of a leaf showing the midrib and four strands. The apex is very dark and it is impossible to determine whether the strands join or are free at the margins. Transfer, USNM 172287, X 5. D-G, do rsi ventral ly compressed heads of the appendages of M. clathratus. The black oval structures below several of the heads are sacs containing pollen grains. In some specimens (such as E) the sacs are covered by the appendage stalk or are not preserved (the centre specimen in D). The two strands that frequently occur in the heads are fairly prominent in E and E. The three heads in G are from the same cone and are shown in the position in which they were fossilized. D, transfer, USNM 172288, X 5. E, transfer, USNM 172289, x 5. F, transfer, USNM 172290, X 5. G, transfer, USNM 172291, X 5. H, portion of the cone axis and the remains of five appendages. Dashed lines represent the strands in the axis and appendages. Each of the black, oval structures consists of a mass of pollen, the contents of one sac. Holotype. Transfer, USNM 168954, X 5. I, pollen grain showing ektexine bands and a slightly shrunken endexine which is clearly sulcate. Note that the bands converge near the ends of the equatorial axis where they fuse. Holotype. Transfer, USNM 172284, x 1000. J, a series of drawings showing variations in the form of the endexine of the pollen grains in a single cone. The ektexine bands are preserved on each grain but for simplification they are not shown here. Holotype. Transfer, USNM 172284, x200. K, epidermal cells and possible trichome bases (small circles) from the head of an appendage. Transfer, USNM 172284, x400. C 9202 ss 616 PALAEONTOLOGY, VOLUME 15 (1960). The fossil called Eq. chinleana resembles the pollen described here in most characters. It falls within the size range shown by them and all have a smooth, thin- walled endexine and a ridged ektexine. The ridges and furrows of the ektexine are about the same size and have the same arrangement and converge at the ends of the major equatorial axes of the grains where they are typically united. In 1960 Scott reported the discovery in the Chinle Formation of pollen grains that fairly closely resemble the grains of the living Ephedra. He demonstrated that his fossils were also very close to Eq. chinleana and concluded (1960, p. 276) they were con- specific. Scott proposed, therefore, a new combination of names for both Daugherty’s fossil and the grains he had discovered, calling them Ephedra chinleana (Daugherty) Scott. I have examined several examples of the grains Scott described and agree that they are indeed similar to the fossil Daugherty called Eq. chinleana and are probably conspecific. The pollen grains described here are also similar to those described by Scott in 1960. They are about the same size and have a smooth endexine and a ridged ektexine. The ridges have conspicuous granules or columellae on their inner faces and converge to two points near the ends of the major equatorial axis (called polar areas by Scott) where they typically are united. The only noticeable difference is that in the grains found in the cones of M. clathratus, the endexine occasionally is elliptical with pointed ends and has a sulcus extending the entire length of the grain. Scott figured some grains which show only the endexine and none of these are sulcate. Only one of Scott’s grains consists of both the endexine and ektexine and although it is elliptical, it is not sulcate. All the other grains figured by him have lost their endexine. M. clathratus shows no resemblance to either the cones of Equisetum, or to the male cones of Ephedra. So although Daugherty had referred the grains now known to be from these cones to the first genus and Scott to the second, this is not supported by the mor- phology of the cones. Although I see no resemblance to Ephedra, I admit that further knowledge might disclose some. If the cone should prove to belong to the same plant as Dechellyia, with which it is so closely associated, then the number of differences from Ephedra would be considerably increased. Some of the grains of M. clathratus also resemble certain of the microfossils described by Wilson (1962) from the Upper Permian rocks of Oklahoma. The ones described by him under the names Vittatina lata, V. costabilis, V. sp. and Ephedripites corrugatus are all somewhat similar in size, shape and in their external ribs. The ribs, however, differ in being united to the endexine and not separate as here. Furthermore, none of these Permian grains have been shown to have a sulcus. The pollen grains in the cones described here should be referred to Equisetosporites chinleana Daugherty if found dispersed. Although the holotype of that genus is difficult to photograph as Scott (1960) indicated, it is identifiable contrary to the opinion of Balme (1970). Thus, Equisetosporites has priority over Ephedripites Bolkhovitina (1953) ex Potonie (1958), if they are congeneric as some suspect. Acknowledgements. I am grateful to Mr. David Gorman of Chinle, Arizona, for telling me about the locality that yielded the fossils described here. The co-operation of Mr. Leslie Cammack, Superintendent of Canyon de Chelly National Monument, who authorized me to collect fossils from the area under his supervision is appreciated. I am also indebted to Mr. Victor Means, Bureau of Indian Affairs of the Chinle Agency for his comments about the geology of the fossil locality, to Professor T. M. ASH: LATE TRIASSIC PLANTS 617 Harris, University of Reading, for his many helpful suggestions about this report, to Mr. Howard Schorn, University of California Museum of Palaeontology for loaning me the type specimen of Equisetosporites chinleana and to Dr. Richard A. Scott, U.S. Geological Survey for loaning me speci- mens of the pollen he called Ephedra chinleana and for interesting discussions about these grains. This work was supported by National Science Foundation Grant GA-25620. REFERENCES ASH, s. R. 1967. The Chinle (Upper Triassic) megaflora, Zuni Mountains, New Mexico. New Mexico Geol. Soc. 18th Ann. Field Conf. Guidebook, 125-131. 1970n. Ferns from the Chinle Formation (Upper Triassic) in the Fort Wingate area. New Mexico. Prof. Pap. U.S. Geol. Surv. 613-D, 1-52, pis. 1-5. 19706. Pagiophyllum simpsonii, a new conifer from the Chinle Formation (Upper Triassic) of Arizona. J. Paleont. 44, 945-952, 4 text-figs. 1970c. Dinophyton, a problematical new plant from the Upper Triassic of the south-western United States. Palaeontology, 13, 646-663, pis. 122-124. 1972n. Plant megafossils — The Chinle Formation, in Breed, C. S. and Breed, W. J. Investigations in the Triassic Chinle Formation. Mas. Northern ,4riz. Bull. 47, 23-44. 19726. The search for plant fossils in the Chinle Formation. Ibid., 45-58. BALME, B. E. 1970. Palynology of Permian and Triassic strata in the Salt Range and Surghar Range, West Pakistan, in Kummel, B., and Teichert, C. Stratigraphic boundary problems: Permian and triassic of West Pakistan. Univ. Kansas Dept. Geol. Spec. Pub. 4, 305-453, pis. 1-22, figs. 1-21. BARNARD, p. D. w. 1968. A new species of Masculostrobus Seward producing Classopollis pollen from the Jurassic of Iran. J. Linn. Soc. (Bot.) 61, 167-176, 1 pi. BOLKHOViTiNA, N. A. 1953. Sporovo-pyl’tsevaia kharakteristika melovykh otlozhenii tsentral’nykh oblastel. Trudy Inst. Geology Nank. 143, Geol. ser. No. 61, 184 pp. CHALONER, w. G. 1967. Lycophyta, in Boureau, E., Trade de Paleobotaniqne, 2, 437-802, figs. 310-507. COOLEY, M. E., and others. 1969. Regional hydrology of the Navajo and Hopi Indian Reservations, Arizona, New Mexico and Utah. Prof. Pap. U.S. Geol. Surv. 521-A, 1-61, pis. 1-5. DAUGHERTY, L. H. 1941. The Upper Triassic flora of Arizona. Pnbls Carnegie Instn Wash. 526, 1-108, pis. 1-34. FLORIN, R. 1936. Die Fossilen Ginkgophyten von Franz-Joseph-Land Nebst Erorterungen Uber Vermeintliche Cordaitales Mesozoischen Alters. Palaeontographica, B, 81, 71-173, pis. 11-52. GOLDENBERG, F. 1855. Flora Saraepontana fossilis. pi. I. Saarbruck. HALLE, T. G. 1908. Einige Krautartige Lycopodiaceen palaozoischen und mesozoischen Alters. Ark. Bot. 7, (5), 1-17, pis. 1-3. HARRIS, T. M. 1941. On some species of Equisetites columnaris Brongn. Ann. Mag. nat. Hist. [11], 9, 393-401, figs. 1-2. 1945. Notes on the Jurassic flora of Yorkshire, 16-18. 16. Baiera furcata (L. & H.) Braun. 17. Sphenopteris pecten sp. n. 18. Equisetites lateralis (Phillips) and its distinction from E. columnaris (Brongn.). Ann. Mag. nat. Hist. [11], 12, 213-234, 7 figs. 1961. The Yorkshire Jurassic flora. I. Thallophyta-Pteridophyta. Brit. Mus. (Nat. Hist.), 212 pp., 71 figs. 1964. The Yorkshire Jurassic flora. II. Caytoniales, Cycadales and Pteridosperins. Brit. Mus. (Nat. Hist.), 191 pp., 7 pis. HARVEY-GiBSON, J. 1897. Contributions toward a knowledge of the anatomy of the genus Selaginella, Spr. Ann. Bot. 11, 123-155. KRYSHTOFOvicH, A. N. and PRiNADA, V. 1932. Contributions to the Mesozoic flora of the Ussuriland. Bull. Geol. Prosp. Serv. USSR. 51, 363-373, pis. 1-2 [In Russian, with English summary]. LiLiENSTERN, H. R. VON, 1928. "Diootutes peiuiaeformis Schenk’, eine fertile Cycadee aus der Lettenkohle. Paldont. Z. 10, 91-107. LUNDBLAD, B. 1950. On a fossil Selaginella from the Rhaetic of Hyllinge, Scania. Svensk. bot. Tidskr. 44, 477^87, pis. 1-2. MILLER, c. N., JR. 1968. The Lepidophytic affinities of the genus Chinlea and Osmundites walkeri. Am. J. Bot. 55, 109-115, 14 figs. 618 PALAEONTOLOGY, VOLUME 15 NATHORST, A. G. 1886. Floran vid Bjuf, Hafte 3. Sverig. geol. Unders. C, 85, 85-131, pis. 19-26. NEWBERRY, j. s. 1876. Geological report, in Macomb, J. N., Report of the exploring expedition from Santa Fe, New Mexico, to the junction of the Grand and Green Rivers of the Great Colorado River of the West, in 1859, under the command of Capt. J. N. Macomb. U.S. Army Eng. Dept., pp. 9-22. PEABODY, D. M. and KREMP, G. o. w. 1964. Preliminary studies of the palynology of the Chinle Forma- tion, Petrified Forest, in Roadifier, J. E., and others. Preliminary investigations of the microenviron- ment of the Chinle Formation, Petrified Forest National Park, Arizona. Geochronology Lab., Univ. Ariz., Interim Res. Rept. 3, 11-26, pis. 1-5. POTONIE, R. 1958. Synopsis der Gattungen der Sporae dispersae. II. Teil: Sporites (Nachtrage), Saccites, Aletes, Praecolpates ; Polyplicates, Monocolpates: Beih. geol. Jb. 31, 114 pp. REPENNING, c. c. and others. 1969. Stratigraphy of the Chinle and Moenkopi Formations, Navajo and Hopi Indian Reservation, Arizona and New Mexico, and Utah. Prof. Pap. U.S. Geol. Surv. 521-B, 1-34. SCOTT, R. A. 1960. Pollen of Ephedra from the Chinle Formation (Upper Triassic) and the genus Equisetosporites. Micropaleontology, 6, 271-276. SEWARD, A. c. 1910. Eossil plants. Vol. II. Cambridge Univ. Press, 624 pp., 376 figs. SIMPSON, J. H. 1850. Journal of a military reconnaissance from Santa Fe, New Mexico to the Navaho country, made in 1849. U.S. 31st Cong., 1st Sess., Sen. Ex. Doc. 64, 56-138, 146-148. wiELAND, G. R. 1929. Antiquity of angiosperms. Proc. Int. Congr. Plant. Sci. 1, 429-456. WILSON, L. R. 1962. Plant microfossils from the Flowerpot Formation, Greer County, Oklahoma. Circ. Oklahoma Geol. Surv. 49, 1-50, pis. 1-3. SIDNEY R. ASH Department of Geology and Geography, Weber State College, Typescript received 31 December 1971 Ogden, Utah 84403 TRINOCLADUS EXOTICUS, A NEW DASYCLADACEAN ALGA FROM THE UPPER CRETACEOUS OF BORNEO by GRAHAM F. ELLIOTT Abstract. A new species of the dasycladacean alga Trinocladus is described as T. exoticiis sp. nov. ; it comes from the Upper Cretaceous of Borneo. Other Trinocladus spp. are Cretaceous and Eocene of the circum- Mediterranean, Middle East, and Caribbean. The Chert-Spilite Series of Sabah (North Borneo) shows a variety of rocks including spilite, basalt, agglomerate, tuff, chert, sandstone, shale, limestone, and marl : it is con- sidered to be of Upper Cretaceous-Lower Eocene age (Fitch 1955, 1961). Thin-sections of six limestone samples were sent to me by Mr. Leong Khee Meng of the Geological Survey, Borneo Region of Malaysia. They proved on examination to be richly algal, containing a new species of the dasycladacean genus Trinocladus, associated with crustose solenoporacean algae referable to Parachaetetes and Petrophyton or a very similar genus. These three algae are considered to indicate an Upper Cretaceous age for these samples. The Trinocladus, abundant and well-preserved, is described below. Of the occurrence, Mr. Leong Khee Meng writes : The Trinocladus specimens occur in the matrix of the calcareous conglomerate which occurs as lenses admixed with amygdaloidal basalt typical of the Chert-Spilite Formation. The calcareous conglomerate also contains subangular to subrounded, varied sized fragments of similar amygdaloidal basalt. The basaltic rocks apparently formed islands, around which the conglomerate was deposited in very shallow marine waters more or less contemporaneously with the basaltic eruption. (Unpublished letter, November 1971). The locality is shown on Text-fig. 1. algae: chlorophyta Family dasycladaceae Kutzing 1843 orth. mut. Hauck 1884 Tribe thyrsoporelleae Pia 1927 Genus trinocladus Raineri 1922 Diagnosis (after Elliott). Calcified tubular dasyclad showing successive verticils of radial branches, each branch showing outwardly widening primaries giving rise to several similar-shaped secondaries, and these in turn to bunches of tertiaries: branches of the lower verticils may not show the full detail. Branches usually not alternate in position from verticil to verticil. Type-species: T. tripolitanus Raineri, Upper Cretaceous of North Africa. [Palaeontology, Vol. 15, Part 4, 1972, pp. 619-622, pi. 120.] 620 PALAEONTOLOGY, VOLUME 15 Trinocladus exoticus sp. nov. Plate 120, figs. 1-6 Diagnosis. Large Trinocladus in which terminal branch-thickening is only conspicuous in the primaries, and with conspicuous annular waxing and waning of proportionally narrow stem-cell diameter. Description. This is a thick-walled dasycladacean, and a large species of its genus. The maximum length observed was 4T6 mm, but this is incomplete and in the living plant this dimension may well have been double or more. The maximum observed external diameter was T38 mm, with an internal diameter varying from OT 82-0-234 mm, giving a d/D ratio of 13-16%. Smaller examples are common. The stem-cell cavity in longi- tudinal section shows a regular ‘waxing-and-waning’ in diameter, widening at verticil- levels and constricting between. The verticils occur regularly at about 0-286 mm apart: each verticil shows eight branches, set at right angles to the long (vertical) axis of the stem-cell. Each branch shows a flask-shaped primary, paddle-shaped in section, com- mencing with a thin insertion at the stem-cell boundary and swelling rapidly to a rounded-triangular termination: these primaries occupy the inner half of the wall thick- ness. From each termination a clump of about eight thin secondaries diverges outwards: each secondary thickens very slightly outwards, but regularly and not normally to a swollen termination. The secondaries extend through most of the outer half of wall- thickness. At their terminations, near the outer surface, each secondary divides into six to eight short, very thin tertiaries, whose terminations occasion the pore-pattern on the outer surface of the fossil. The thin branch-systems of each verticil are clearly spaced apart in the wall-thick- ness, so that the only zone of overlap is peripherally with the tertiary branchlets. Holotype. The specimen figured in pi. 120, fig. 6, from the Chert-Spilite Formation (‘calcareous lenses apparently intermixed with volcanic rocks typical of this formation’); Upper Cretaceous. Kuamat- Malabuk Malua area, Darvel Bay, Sabah, Malaysia. BMNH, V. 56285. Paratvpes. The specimens figured in pi. 120, figs. 1-5, same locality and horizon. BMNH, V. 56282, V. 56283, V. 56286. Discussion. T. exoticus is closest in structure (detail and proportions) to the Upper Cretaceous type-species T. tripolitanus Raineri (Pia 1936, Elliott 1968), rather than to the Maestrichtian T. radoicicae Elliott or to T. perplexus Elliott from the Palaeocene. In size, however, T. exoticus is much larger and approaches that of T. pinarensis Keijzer (Upper Cretaceous of Cuba). This is the largest species, diff'ering however in some EXPLANATION OF PLATE 120 Figs. 1-6. Trinocladus exoticus sp. nov.; Chert-Spilite Formation, Upper Cretaceous; Kuamat- Malabuk Malua area, Darvel Bay, Sabah, Malaysia. Reg. nos. are those of British Museum (Natural History), Dept, of Palaeontology, London. 1. Oblique-vertical section, paratype, Xl7; V. 56286. 2. Transverse section, paratype, x28; V. 56283. 3. Vertical-tangential section, paratype, X22; V. 56282. 4. Transverse cut of broken example to show branch-structure, paratype, X42; V. 56282. 5. Fragment to show branch-structure, paratype, X64; V. 56283. 6. Longitudinal section of large example, holotype, x 19; V. 56285. Palaeontology, Vol. 15 PLATE 120 ELLIOTT, Trinocladus ELLIOTT: TRINOCLADUS EXOTICUS 621 dimensions, proportions and detail from the other two Upper Cretaceous species. Dimensions and structures of these three species are set out below for comparison. In T. tripoJitanus the lower branches may not show the full branch detail: this may well have been the case in the Borneo fossil, but has not been noted by me. The swollen II7°33'E II7°35'E 1 1 1 LOCATION OF r/?//V£?CL.4Z>f/S - BE ARING SAMPLES, SABAH, MALAYSIA Scale 0 '/2 I MILE LEGEND MIOCENE I Kuamut Formation : slump 1 breccia, mudstone OLIGOCENE ? Lobong Formation : J sondstone, mudstone UPPER CRETACEOUS- EARLY tertiary I Chert-Spilite Formation: 1 sandstone, omygdoloidal basalt with admixed colcareous conglomerate lenses LOWER TRIASSIC AND/OR OLDER ^757^ Crystalline Basement: V Geological boundary Fault Strike ridges Location of outcrop of Trinoc/adus-beQv\T\q samples J7832 Sample number River or stream Logging road / /V '^^^yP^Trinoclodus~bea^'\nq sampleri { J7832-J7837A) II6®£ / J 117®£ lie°E II9»E LOCATION MAP 7® Scole in Miles 7® SOUTH 40 0 40 CHINA SEA / Cy '*>B Area Illustrated N 6® _Koto Kinabalu SULU SEA - N 6® SABAH gfidon a ^ \ p Kinabat N ^Ji /J N 5® J II 1 \ H 4® Sl*' KALIMANTAN^ ♦(1 1 'v ‘-Rr;— , N 4® TEXT-FIG. 1. Geological map of sample locality. primaries probably served as sporangia in life, and sueh later parts of the plant are usually the most heavily calcified and would stand the best chanee of being fossilized. T. exoticus is abundant in the thin-sections examined : in life it would have grown in warm very shallow marine waters as do the living Neomeris and Bornetella in the Indo- Pacific. The Borneo fossil shows that in Upper Cretaceous times Thnocladus probably had either a circum-global distribution in the right environment: Caribbean, Mediterranean, 622 PALAEONTOLOGY, VOLUME 15 Trinocladus spp. L D d d/D V P s t pinarensis 3-8 + T5-20 0-35 or less 18-23% 0-2 12-13 short, swollen 10-12 elongate 10 short, slim tripolitanus 3-36 + 0-47-0-68 01 6-0- 19 28-34% 0-1 8 club- shaped 5-6 club- shaped 6 short, slim exoticus 4-16 + 1-38 0-182- 0-234 13-16% 0-86 8 club- shaped 8 elongate 6-8 short, slim L = length (incomplete), D = outer diameter, d = internal diameter (stem-cell cavity), V = distance apart of successive verticils, p = number and description of primary branches, s = number and description of secondary branches, t = number and description of tertiary branches. Dimensions in mm. These fossils are plants and therefore the detail set out above may vary. Middle East and Indonesia, or alternatively, if one accepts the concept of continental drift, that all the species were relatively adjacent in a much smaller Tethyan area. The genus survived into the older Tertiary, when along with other cladospore forms such as the related Thyrsoporella and Belzungia, and also Broeckella, it became extinct. REFERENCES ELLIOTT, G. F. 1968. Permian to Palaeocene Calcareous Algae (Dasycladaceae) of the Middle East. Bull. Br. Mus. nat. Hist. (GeoL), Siippl. 4, 101 pp., 24 pis. FITCH, F. H. 1955. The geology and mineral resources of part of the Segama Valley and Darvel Bay area. Colony of North Borneo. Geol. Surv. Dept. Brit. Terr. Borneo, Mem. 6. (New edn. in prepara- tion, 1971, Geol. Surv. Borneo Region Malaysia). 1961. Geological Map of North Borneo, Brunei and part of Sarawak', 1, 2,000,000. Surv. Dept., Fed. Malaya. KEiJZER, F. G. 1945. Outline of the geology of the eastern part of the province of Oriente, Cuba. Geogr. geol. Meded., Utrecht, (2) 6, 1-238, 11 pis., map. PIA, J. 1936. Calcareous green algae from the Upper Cretaceous of Tripoli (North Africa). J. Paleont. 10, 3-13, pis. 1-5. G. F. ELLIOTT Department of Palaeontology British Museum (Natural History) London, SW7 5BD Typescript received 17 November 1971 THE ORIGIN OF THE SILURIAN CLARKEIA SHELLY FAUNA OF SOUTH AMERICA, AND ITS EXTENSION TO WEST AFRICA by L. R. M. COCKS Abstract. The brachiopod Arataiiea monodi Schmidt 1967 from Mauretania (West Africa) is put into the synonomy of Clarkeia antisiensis, chief constituent of the Malvinokaffric Clarkeia fauna of Central South America of Silurian age. The recently described fauna of latest Ordovician age from South Africa is considered to have been the ancestor of the Clarkeia fauna. In particular Marklandella africana is reassigned to Heteror- thella, and considered the ancestor of Helerortliella freitana \ Plectothyrella haughtom is considered the ancestor of Clarkeia itself. Eostropheodonta, Orbicidoidea and homalonotid trilobites are common to both faunas. The ecology of both the South African and Clarkeia faunas is discussed. They are considered to represent the shallower end of the benthic depth spectrum, and may have been parallel communities to the contemporaneous Hirnantia, Cryptotliyrella and Eocoelia Communities of Eurasia and North America. They may also have lived in colder water than the northern communities. Over the whole of the central part of South America there occurs commonly a fauna characterized by large numbers of a few genera of brachiopods, and dominated by the ribbed genus Clarkeia Kozlowski 1923. The age of this fauna has long been unsettled; it was originally thought of as Devonian, but now most workers agree that it is Silurian, and the age is discussed further below. This Clarkeia fauna has been described from Argentina (Kayser 1897, Thomas 1905, Castellaro 1959, 1967), Brazil (Clarke 1899, 1913) and Bolivia (d’Orbigny 1842, Kozlowski 1923). However, bearing in mind the exploratory state of geological knowledge in much of the South American hinterland, it may be expected also elsewhere in that general area. No other shelly Silurian faunas are known from this area. The chief constituents of the fauna are brachiopods: Clarkeia antisiensis (d’Orbigny) in overwhelming numbers, together with scarcer Heterorthella freitana (Clarke), Eostropheodonta fascifer (Kayser), Aiistralina jachalensis Clarke, Strophochonetes fuer- tensis (Kayser), Leptocoelia acutiplicata Kayser, Leptaena argentina (Thomas) and various inarticulates, notably Orbiculoidea sp. Among other phyla Tentacidites sp., homalonotid trilobites, bivalves and gastropods have all been recorded sporadically. Material available in the British Museum (Natural History) includes a collection from Argentina made by the Shell Petroleum Company in 1961 and a collection from Bolivia recently presented by Professor Boucot. One of the problems in using some of the older literature cited above lies in disentangling the Silurian Clarkeia fauna from the genuine Devonian fauna which was in many cases simultaneously described from adjacent deposits. All these brachiopod species, as well as the genera Clarkeia and Australina, are probably endemic to South America and West Africa, although more work is necessary on the Leptaena; and the fauna as a whole is noteworthy as being one of the few examples of possible provincialization in Silurian shelly faunas. As with the later Devonian province in the same general area, the term Malvinokaffric (Berry and Boucot 1972) [Palaeontology, Vol. 15, Part 4, 1972, pp. 623-630, pi. 121.] 624 PALAEONTOLOGY, VOLUME 15 may be used. It is probable that the recently described Ancillotoechia cooper ensis of Amos and Noirat (1971), from Jujuy Province, Argentina, is synonymous with Clarkeia autisiensis; their figured specimens appear identical to the Bolivian mould material figured in this paper. The Clarkeia fauna has not previously been reported from outside South America. However it can now be recorded from West Africa; in Mauretania in the western Sahara. These are the specimens described by Schmidt (1967) as Aratanea monodi gen. et sp. nov. A comparison of her excellent descriptions and figures leave no doubt that this form is identical with typical Clarkeia autisiensis (compare Schmidt 1967, pi. 1, figs. 1-6 with Plate 121, figs. 1-8 of the present paper). The stratigraphy of the Majabat al-Koubra area was briefly described by Monod (1958) and consists of a sandstone containing large numbers of Clarkeia in an otherwise unfossiliferous sequence of clastic rocks. No other brachiopods are recorded. This form of occurrence is very similar to many in South America. The precise age of the fauna in Mauretania is not known; Monod (1958, p. 113) records the presence of graptolites further east, but these are not identified, and in any case their stratigraphical position relative to the ‘Gres tendres a Brachiopodes’ is not yet established. The known occurrences of Clarkeia and its allies are plotted on a map (Fig. 1 — • redrafted after Bullard, Everett, and Smith) which shows South America and Africa in their relative positions before the Mesozoic to Recent continental drift occurred. Also shown on the map is the position of the fauna described from South Africa by Cocks, Brunton, Rowell, and Rust (1970). These brachiopods, from the Cedarberg Formation of the Table Mountain Group, were described as ?Plectoglossa sp., Trematis taljaardi Rowell, Orbiculoidea sp., Marklaudella africana Cocks and Brunton, Eostropheodoiita discumbata Cocks and Brunton and Pleclothyrella haiightom Cocks and Brunton. At the time that paper was written Marklandella Harper, Boucot, and Walmsley was an unpublished genus, but discussion with Dr. Walmsley and a brief look at their manu- script lead Dr. Brunton and myself to place the South African enteletacean within it. However, on fuller consideration of the published paper (Harper et al. 1969) and further discussion with Professor Boucot, I am now convinced that the Table Mountain Group species africana is better placed within another genus described in the same paper, Heterortliella. In particular Heterorthella africana (re- figured here, pi. 121, figs. 14-16) is considered to be a close relative of PI. freitana (Clarke), one of the constituents of the Clarkeia fauna, particular emphasis being placed on the similarity of the shape of the muscle field in the pedicle valves, and the musculature and general valve form of the brachial valves. In the South African paper (Cocks et al. 1970, p. 594) we had remarked on the similarity of africana to '’Ortliis' tacopayana Kozlowski 1923; this latter species is now considered to be a junior synonym of Heterorthella freitana (Clarke 1899) (Harper et al. 1969, p. 80, pi. 16, figs. 1-6). As further discussed in the systematic notes below, Clarkeia itself is very probably a close relative, and potential descendant, of Plectothyrella Temple 1965; in particular C. autisiensis is thought to be descended from P. haughtoni of the South African fauna (pi. 121, figs. 10-13). In addition Eostro- pheodonta fascifer of the Clarkeia fauna belongs to the same species group as E. discumbata from the Table Mountain Group, and the genus Orbiculoidea is common to both the Clarkeia and South African faunas. As to the other inarticulate brachiopods, it is possible that the form described as Lingula cf. oblata Hall by Clarke (1899, p. 8, pi. 1, fig. 3) may be comparable with the ?Plectoglossa sp. of the South African fauna, but the inarticulates of the Clarkeia fauna are badly in need of expert taxonomic revision and redescription. The age of the South African fauna is thought to be very close to the Ordovician- Silurian boundary, with the probability of an age just Ordovician, i.e. high Ashgill in European terms. The age of the Clarkeia fauna, discussed more fully below, is certainly Silurian. It is postulated here that the direct ancestor of this endemic Clarkeia fauna COCKS: SILURIAN CLARKEIA FAUNA 625 TEXT-FIG. 1. Reconstruction of Africa and South America, prior to continental drift. Occurrences of Clarkeia (C) and South African (S) faunas are shown. might very well have been the South African fauna from the Table Mountain Group, the latter being the starting point of the later, Silurian, minor provincialization. AGE OF THE CLARKEIA FAUNA The age of the Clarkeia fauna has given rise to much discussion. In the Los Espejos Formation of San Juan Province, Argentina, Cuerda (1965) has illustrated Mono- graptus leintwardinensis var. incipiens Wood from beds with Australina jachalensis, and 600 metres above beds with A. jachalensis, Clarkeia antisiensis, and Strophochonetes 626 PALAEONTOLOGY, VOLUME 15 fiiertensis. This graptolite, which seems correctly identified but should be termed Saetograptus incipiens, appears for the first time elsewhere in the basal Ludlow nilssoni- scanicus Zone, and certainly suggests that the upper range of A. jachalensis extends to that age. However the fuller Clarkeia fauna 600 metres below is associated with what Cuerda identifies (but does not figure) as Monograptus aff. vomer inus (Nicholson). This suggests an age most probably in the lower half of the Wenlock for the Clarkeia fauna in that particular succession. This is the only reliable graptolite data associated with a Clarkeia fauna which have yet been published, although the presence of graptolites has been recorded elsewhere, some of which also suggest a Wenlock age. However, Helge L. Hansen has recently discovered graptolites in Paraguay of the convohitus or sedgwickii Zones of the Llandovery (M. lobiferus and M. aff. sedgwickii, identified by Dr. Rickards). These are associated with a shelly fauna, a very small sample of which includes poorly preserved ribbed atrypoids or rhynchonellides. Lingula and Tentaculites. Internal age evidence is imprecise since all of the species, and the genera Clarkeia and Australina are endemic. Clarkeia is probably derived from the Ashgill and early Llandovery Australina is close to the Llandovery to Ludlow Glassia. As to the other genera, Heterorthella occurs only in the Ashgill of South Africa and the Wenlock of Canada, Eostropheodonta occurs from the Ashgill to the late Wenlock, although most commonly in the Ashgill; Strophoclionetes occurs from the Ashgill to the Ludlow, although rare outside the Wenlock and Ludlow, and the remaining fauna suggest no more than a general Silurian age. Perhaps the Clarkeia fauna may best be assessed as Llandovery to Wenlock in age, with a few elements such as Australina persisting into the Ludlow. The age span of the fauna is not clear; whether it persisted in different areas through the Llandovery and EXPLANATION OF PLATE 121 Figs. 1-8. Clarkeia antisieusis (d’Orbigny). Figs. 1, 4 from San Juan Province, Argentina, collected Shell Petroleum Company 1961, Figs. 2, 3, 5-8 from Arquillos, near Tomina, Chuquisaca Depart- ment, Bolivia. Figs. 1, 4, BB 51599, external views showing ribbing, high shoulders and fold and sulcus, X 2 0. Figs. 2, 7, BB 34263, oblique view of internal mould of both valves, and latex cast of the umbonal area showing erect crural lobes, X 3-0. Fig. 3, BB 34264, internal moulds of several valves illustrating rock-forming abundance, x 2-0. Fig. 5, BB 34265, latex cast from internal mould of conjoined valves, showing cruralia; the shadow reveals their form, x3 0. Fig. 6, BB 34266, latex cast from Internal valve of conjoined valves, oblique view showing massive cardinal process projecting posteriorly, X 3 0. Fig. 8, BB 34268, latex cast of pedicle valve interior, showing muscle field and laterally flaring rims from the teeth, X 3-0. Fig. 9. Plectothyrella crassicosia (Dalman), BB 31867, the conjoined valves figured by Wright (1968, fig. 4) from Ashgill mudstones within Kildare Limestone, knoll west of Chair Farm, County Kildare, Ireland, X 1-5. Figs. 10-13. Plectothyrella haiightoiii Cocks and Brunton, from Cedarbcrg Formation, Wellington Sneeukop, Cape Province, South Africa, collected by 1. C. Rust. Fig. 10, BB 31528, posterior view of brachial valve internal mould, X 1-5. Fig. 11, BB 31566, holotype, latex cast of a brachial valve internal mould, showing anteriorly directed crural lobes, X 2-0. Fig. 12, BB 31527, posterior view of brachial valve internal mould, X 2 0. Fig. 1 3, BB 31529, latex cast of brachial valve internal mould, showing vertical and stubby crural lobes, X 2 0. Figs. 14-16. Heterorthella africana (Cocks and Brunton), from Cedarberg Formation, Langvlei, near Porterville, Cape Province, South Africa, collected by I. C. Rust. Figs. 14, 15, BB 31601, internal mould of brachial valve and latex cast of it, X2 0. Fig. 16, BB 31582, internal mould of pedicle valve, showing shape of muscle field, X 2 0. Palaeontology, Vol. 15 PLATE 121 COCKS, Clarkeia fauna COCKS: SILURIAN CLARKEIA FAUNA 627 Wenlock or whether all the fauna occurred at a similar time is unknown. As can be seen from the lists of Castellaro (1967, pp. 13-20), not all the elements of the Clarkeia fauna occur at all the localities; perhaps future work may enable further separation to be made. Berry and Boucot (1972) also discuss the age problem. ECOLOGY In the European Upper Llandovery five shelly assemblages have been defined, which are considered to reflect the marine benthic communities of that time (Ziegler, Cocks, and Bambach 1968). They were distributed according to some direct function of depth, chiefly because they map out as bands subparallel to palaeo-shorelines. As with their modern marine level-bottom counterparts, the diversity of animals within the com- munities increases from the shallow to the deeper part of the bathymetric range. Thus the fairly shallow-water Eocoelia Community usually has no more than six or seven species of brachiopod, in addition to some molluscs, a few bryozoa, TentacuUtes and rather rare trilobites (usually either Dalmanites or Encrimirus). In contrast the Clorinda Com- munity, at the other end of the depth spectrum, may have as many as thirty brachiopod species represented, as well as many other animals. In the lower part of the Llandovery Eocoelia itself does not occur, and the place of the Eocoelia Community is taken by the Cryptotliyrella Community, such as occurs in abundance in the Mulloch Hill Formation of Girvan, Scotland. In the late Ashgill there occurs another most distinctive fauna, once again dominated by brachiopods, called the Hirnantia fauna, after a common enteletacean. This occurs in Poland (Temple 1965), Bohemia (Marek and Havlicek 1967), Sweden (Bergstrom 1968), Great Britain and Ireland (Wright 1968) and in North America (Boucot and Johnson 1970). The Hinumtia fauna has several elements in common with the South African fauna, in particular Plectothyrella haughtoui from the Table Mountain Group is related toP. n‘o.s'5/cos'7<7(Dalman) (pi. 121, fig. 9), and Eostropheodonta discumbata from Africa is close to E. himantensis of the Hirnantia fauna. Orbicidoidea and homalonatid trilobites occur in both faunas. Although the composition and structure of late Ashgill shelly communities has not yet been fully evaluated, it is reasonable to postulate that the low- diversity Hirnantia fauna represents a community filling a comparable ecological niche to the Eocoelia and Cryptothyrella Communities of the lower Silurian. Another relevant assemblage, just described from the upper Ashgill ‘Gres du deuxieme Bani’ of Morocco (Havlicek 1971), consists of Hirnantia sagittifera, Eostro- pheodonta squamosa (very close to, if not conspecific with E. discumbata), Destombesium zagoraensis and D. ellipsoides (new enteletacean taxa), and a new species of Plectothyrella named P. chauveli. Unfortunately the material is scarce and rather poor, but this last species appears to be morphologically intermediate between P. crassicosta and P. haugli- toni. This assemblage is of interest as being the first Hirnantia fauna described from Africa, and one providing suggestive links between the typical Hirnantia fauna and the fauna from South Africa. The South African and Clarkeia faunas also have low diversity indices. Two explana- tions are possible; firstly, that the two southern hemisphere faunas lived in shallow water, in a similar way to the European and North American committees, and can therefore be considered as parallel communities to the northern ones; and secondly, that the low 628 PALAEONTOLOGY, VOLUME 15 diversity of the southern faunas was due to cold-water conditions, perhaps correspond- ing to some Arctic faunas today. There is certainly good evidence of a late Ordovician- early Silurian glacial period in both north and south Africa (Destombes 1968, Cocks et al. 1970). However, no elements of the South African fauna are more than specifically distinct from European faunas, and it seems most likely that that fauna at least was a relatively shallow-water one, parallel to the Eocoelia and Hirnantia Communities. Whether the same is true of the Clarkeia fauna remains at the moment equivocal : there is evidence of a glacial period in Argentina and Bolivia, but not from Brazil or the Sierra de la Ventana of Argentina (Berry and Boucot 1972); but a shallow- water sea over the area populated by the Clarkeia fauna seems a fair hypothesis in our present state of knowledge. SYSTEMATIC NOTES Several authors have discussed the problems in distinguishing ribbed atrypidines and rhynchonellides during the Ordovician and Silurian. The only objective difference be- tween the two groups is the presence of calcareous spiralia supported by a jugum in the atrypidines and their absence in the rhynchonellides. The situation is further complicated by the rhynchonellides’ probable position as the parent stock of the atrypidines. The practical difficulty lies in ascertaining whether a particular brachiopod genus had spiralia ; their absence in any given specimen being so often due to the vagaries of preservation. Fortunately perfect Clarkeia material was available from San Juan province, Argen- tina (pi. 121, figs. 1, 4), some specimens of which were serially sectioned after calcining. Long rhynchonellide crura, which ended in very fine subparallel blades, were completely preserved. There was no trace of spiralia or any jugal apparatus, which would certainly have been seen in this fine preservation. This confirms the conclusion suggested by the Bolivian mould material (pi. 121, figs. 2, 3, 5-8), and by Schmidt’s work (1967), that Clarkeia is a rhynchonellide. Flectothyrella material completely preserved in calcite and suitable for sectioning has not been available, and so the presence of spiralia in that genus is still in doubt. Bergstrom (1968) and Boucot and Johnson (1970) both suggest that it is a rhynchonellide, as opposed to Temple (1965), Boucot and Johnson earlier (see 1970, p. 894) and Cocks et al. (1970) who believed it to be an atrypidine, perhaps related to the morphologically similar genus Cliiitonella Hall, from the Lower Silurian of North America, which has spiralia. However, it is here proposed that Flectothyrella is more closely related to Clarkeia than to any other genus; they seem best united in the subfamily Plectothyrellinae which was erected by Bergstrom (1968), and included in the rhynchonellides. Schmidt (in Williams et al. 1956, p. H570) includes Clarkeia in the Eatoniidae. Dr. G. A. Cooper has very kindly sent on exchange some specimens of Eatonia meclialis (Vanuxem), the type species of the family, as well as some Costellirostra peculiaris (Conrad) for comparison. It seems reasonable to include the Plectothyrellinae as a subfamily within the Eatoniidae, rather than in the Ancistrorhynchidae as suggested by Bergstrom (1968), although Silurian rhynchonellides as a whole need systematic reappraisal. Clarkeia antisiensis and Flectothyrella haughtoni are closely related and probably descended from one another. The differences between them are as follows. Clarkeia has a massive cardinal process, usually projecting posteriorly and sometimes bifurcating at the end; F. haughtoni has almost no cardinal process, and the diductor muscles must COCKS: SILURIAN CLARKEIA FAUNA 629 have been attached almost directly to the floor of the valve in the brachial umbo. Clarkeia's crural lobes are nearly always erect and tapering, only sometimes slightly directly anteriorly as well as ventrally, whilst those of P. haughtoni always project anteriorly except in the most gerontic specimens, and in general they do not taper. Clarkeia has a thickened rim of shell material laterally from the teeth in the pedicle valve, whilst in P. haughtoni this rim is absent. Clarkeia has small dental plates; these are absent in P. haughtoni, although sometimes developed in Plectothyrella crassicosta. The ornament of Clarkeia is more massive, with approximately 12 costae, four of which are nearly always on the fold, whilst P. haughtoni has approximately 30 costae, although these are difficult to count accurately laterally. Of course there are many features which are very similar in the two forms; the dorsal median septum is extremely variable in both species, sometimes bladelike up to 3 mm high, sometimes a rounded myophragm, sometimes hardly detectable. The general valve shape is very similar, as is the adductor muscle field on both valves. The denticulations are closely comparable. The foramen of Clarkeia was probably functional throughout life; the data for P. haughtoni are less certain, the pedicle may have been sealed in large individuals, but in small ones it was similar to that of Clarkeia. Acknowledgements. I am most grateful to Professor A. J. Boucot, who first suggested the reassignment of africana to Heterorthella, who kindly supplied some Clarkeia from Bolivia, and who read the manu- script. I also thank Dr. R. B. Rickards for discussion on some graptolite faunas. The photographs were taken by Mr. T. Parmenter and myself, and the specimens (BB numbers) are in the British Museum (Natural History). REFERENCES AMOS, A. J. and noirat, s. 1971. A new species of Ancillotoeclna from the Zapla Formation, Northern Argentina. Smithson. Contr. Paleobiol. 3, 139-142, pi. 1. BERGSTROM, J. 1968. Upper Ordovician brachiopods from Vastergotland, Sweden. Geol. et Palaeont. 2, 1-21, pis. 1-7. BERRY, w. B. N. and BOUCOT, A. J. 1972. Correlation of the South American Silurian Rocks. Spec. Pap. geol. Soc. Am. BOUCOT, A. J. and Johnson, j. g. 1970. Redescription of Clintonella Hall (Silurian, Brachiopoda). J. Paleont. 44, 893-897, pi. 128. CASTELLARO, H. A. 1959. Braquiopodos gotlandicos de la Precordillera de San Juan. Rev. Assoc. Geol. argent. 13, 41-65, pis. 1-5. 1967. Giiia Paleontologica Argentina. Parte I: Paleozoico. Seccion III — Faunas Siluricas. Buenos Aires, 57 pp. CLARKE, J. M. 1899. A fauna Siluriana superior do Rio Trombetas, Estado do Para, Brazil. Archos Mils. nac. Rio de J. 10, 1-174, pis. 1-8. • 1913. Fosseis Devonianos do Parana. Monografias Serv. Geol. Min. Brasil, 1, 1-353, pis. 1-27. COCKS, L. R. M., BRUNTON, c. H. c., ROWELL, A. J. and RUST, I. c. 1970. The first Lower Palaeozoic fauna proved from South Africa. Q. Jl. geol. Soc. Lond. 125 (for 1969), 583-603, pis. 39-41. CUERDA, A. J. 1965. Monograptiis leintwardinensis var. incipiens Wood en el Silurico de la Precordillera. Amegliiniana, 4, 171-178, pi. 1. DESTOMBES, J. 1968. Sur la nature glaciare des sediments du groupe du 2e Bani, Ashgill superieur de I’Anti-Atlas (Maroc). C. r. liebd. Seanc. Acad. Sci. 267, 684-686. HARPER, c. w., BOUCOT, A. J. and WALMSLEY, V. G. 1969. The rhipidomellid brachiopod subfamilies Heterorthinae and Platyorthinae (new). /. Paleont. 43, 74-92, pis. 15-17. HAVLiCEK, V. 1971. Brachiopodes de I’Ordovicien du Maroc. Notes Mem. Serv. Mines Carte geol. Maroc, 230, 1-135, pis. 1-26. 630 PALAEONTOLOGY, VOLUME 15 KAYSER, E. 1897. Beitrage zur Kenntniss einiger palaozoischer Faunen Sild-Amerikas. Z. dt. geol. Ges. 49, 274-317, pis. 7-12. KOZLOWSKi, R. 1923. Faune devonienne de Bolivie. Annls Paleont. 12, 1-110, pis. 1-10. MAREK, L. and havlIcek, V. 1967. The articulate brachiopods of the Kosov Formation (Upper Ash- gillian). Vest, listfed. tlsf. geol. 42, 275-284, pis. 1-4. MONOD, T. 1958. Majabat al-Koubra, contribution a I’etude de ‘I’Empty Quarter’ ouest-saharien. Mem. Inst.fr. Afr. noire, 52, 1-407, pis. 1-81. d’oRBiGNY, A. 1842. Voyage dans I'Ameriqne meridionale. Vol. 3, 4th part. Paleontologie, 188 pp., 22 pis. Paris. SCHMIDT, H. 1967. Aratanea monodi n. g, n. sp. (Brachiopoda; W-Sahara). Senckenberg. letli. 48, 91-97, pi. 1. TEMPLE, J. T. 1965. Upper Ordovician brachiopods from Poland and Britain. Acta palaeont. pol. 10, 379-427, pis. 1-21. THOMAS, I. 1905. Neue Beitrage zur Kenntnis der devonischen Fauna Argentiniens. Z. dt. geol. Ges. 57, 233-290, pis. 11-14. WILLIAMS, A. et al. 1965. Treatise on Invertebrate Paleontology. Part H, Brachiopoda, (ed.), R. c. moore, 927 pp. Kansas. WRIGHT, A. D. 1968. A westward extension of the upper Ashgillian Hirnantia fauna. Lethaia, 1, 352-367. ZIEGLER, A. M., COCKS, L. R. M., and BAMBACH, R. K. 1968. The Composition and structure of Lower Silurian marine communities. Lethaia, 1, 1-27. L. R. M. COCKS Department of Palaeontology British Museum (Natural History) Typescript received 30 November 1971 London, SW7 5BD MOULTS OF DAKOTICANCER OVERANUS, AN UPPER CRETACEOUS CRAB FROM THE PIERRE SHALE OF SOUTH DAKOTA by GALE A. BISHOP Abstract. Of 4,000 specimens of fossil decapods collected from three stratigraphically different localities in the Pierre Shale (Upper Cretaceous, Maestrichtian) of South Dakota, twenty specimens of the brachyuran Dakoticancer overaiuis Rathbun were found preserved in Salter’s position. Several of these specimens have the portion of the carapace which lies outside pleural sutures still attached to the sternal plastron, proof that they are exuviae. The part of the carapace between the pleural sutures is completely overturned on each specimen and rotated on most specimens. The escaping movements of the moulting crab probably caused water currents which in turn caused the central part of the carapace to slip backward over the moulting foundation resulting in the new type of Salter’s position. Decapod crustaceans, encased in non-growing exoskeletons, must moult or periodically shed their exoskeleton in order to grow. Moulting is a continuous process in the life of decapods. Drach (1939) has defined several stages in this process (Table 1). TABLE 1 . Brachyuran moulting Cycle, summarized from Passano 1 960 (after Drach). Stage Name Characteristics Activity level Feeding Water 0/ /o Duration 0/ /o A Newly Moulted Water absorption, mineralization slight none 86 2-0 B Paper Shell Endocuticle formation, chelae hard, tissue growth begins full starts 85 8-0 C Hard Main tissue growth, accumulation of organic reserves full yes 60 66+ D Peeler Epicuticle formed. Exocuticle secretion, skeletal resorption pleural sutures open full to reduced some to none 60 to rise 24 E Moult Rapid water uptake and exuviation none none rapid rise 0-5 The shedding of the exoskeleton (Stage ‘E’) takes only a short time. The animal stands in its normal position. The carapace splits along lines of weakness called the pleural sutures (text-fig. 1 a, b). The part of the carapace between the pleural sutures remains hinged to the internal skeleton and flips upward and forward (into Salter’s position) as the animal escapes from the old exoskeleton. The escape is made possible by a drastic water loss in soft tissues and throbbing muscular contractions. After moulting there is a rapid uptake of water in the soft tissues which causes the increase in size of the animal. [Palaeontology, Vol. 15, Part 4, 1972, pp. 631-636, pis. 122-123.] C 9202 T t 632 PALAEONTOLOGY, VOLUME 15 C E F TEXT-FIG. 1. Preservation of decapods, a, b, dorsal and lateral views showing lineae (pleural sutures) of Paromola ciivieri (Risso) (from Glaessner 1969, p. R406); c, position of crab (Card aides maetias) buried alive ; d, e, Salter’s position of moults of Potamou cjuenstedti as is shown by lower part of carapace remaining attached to the sternum (after Schafer 1951);/, position of lobster moult (Hoploparia longimana (Sow.) from Glaessner 1929). Fossilization. The moult, or shed exoskeleton, is capable of being fossilized, just as is the living brachyuran. Schafer (1951) has summarized the conditions and processes of the fossilization of brachyurans. The position of the fossil in the sediment and the arrangement of the skeletal elements are very important in the determination of the condition of the animal at the time of its burial. EXPLANATION OF PLATE 122 Figs. UlO. Dakoticancer overaiiiis Rathbun, specimen 4-369, Mobridge Locality, Mobridge, South Dakota. 1-5, x I -5, stereo, Figs. 6-10, x 1. 1,6. Dorsal. 2, 7. Anterior. 3, 8. Right side, notice crack along pleural suture. 4, 9. Ventral, anterior-lateral carapace margin, sternal plastron, maxillipeds, and pleural sutures visible. 5, 10. Oblique of mouth frame and anterior of crab showing triangular notch on mouth frame margin which gives rise to the pleural suture. Palaeontology, Vol. 15 PLATE 122 BISHOP, moults of Dakoticancer BISHOP: MOULTS OF DAKOTICANCER OVERANUS 633 Brachyurans buried alive attempt to escape from the entombing sediment. The escape attempt consists of a rowing or lifting movement of the appendages. If the escape attempt is unsuccessful, the crab is usually preserved with the appendages frozen in the position of lifting; the legs are commonly raised above the carapace and the chelae are open (text-fig. Ic). Dead crabs and moulted exoskeletons act as passive objects within the limits of their articulation. Physical, chemical, and biological forces act on the crab remains to break them down. After soft tissues decay, corpses and moults react to the forces of destruction in a similar manner. It takes only a few days to decompose the soft tissues. The decomposition of the exoskeleton takes a longer time. The calcareous middle layer breaks down first, within about four weeks (Schafer 1951). The chitinous inner and outer layers are more resistant to destruction and last for many months. Corpses and moults can both be found preserved in a characteristic position, Salter’s position, in which the carapace is lifted and flipped forward, making an angle of about 90° with the sternum (text-fig. 1 d, e). Moults are distinguished from corpses by noting that the carapace has split along the pleural sutures, and the parts of the carapace lying outside of the pleural sutures are still attached to the sternal plastron (marked by an ‘X’ in text-fig. 1 d). Fossilized moults of Ranina, Nolopocorystes, Coeloma, Polamon, and Macrophthalnms have been described (Glaessner 1969, p. R431). DAKOTICANCER Origin of the specimens. Abundant fossil decapods were collected from the Pierre Shale (Upper Cretaceous, Maestrichtian) from three localities in South Dakota; Creston, Thomson Butte, and Mobridge. The Pierre Shale is a thick body of fine-grained clastic rocks, mostly shales, siltstones, and calcareous shales interbedded with thin bentonite beds. Fossils are numerous and well preserved in calcite and siderite concretions but rare in the shale itself. Ammonites provide the basis for a biostratigraphic zonation (Gill and Cobban 1966). Decapods are pre- served in apatite concretions distributed through in- tervals of shale ten to twenty-five feet thick and continuous over areas of six to six hundred square miles. Fossil assemblages from the three localities are similar, and are dominated by the brachyuran Dakoti- cancer over anus Rathbun 1917, the epifaunal bivalve Inoceranius, the nektonic ammonite Baculites, and the feces of soft-bodied burrowing organisms. Many other organisms are present but less common. Dakoticancer overanus is the most abundant fossil, and this recurring suite of fossils is termed the Dakoticancer Assemblage. The consistent faunal composition, similar mode of preservation, the age distribution of the decapods, and the lack of scavenging suggests that the fossils are the remains of mass killings of a recurring decapod community. Morphology of Dakoticancer. The squarish-oval carapace of Dakoticancer overanus (text-fig. 2) is dominated by longitudinal ridges ahead of the cervical furrow and on the sagittal ridge and by transverse ridges and grooves on the branchial regions. Two TEXT-FIG. 2. Line drawing of the carapace of Dakoticancer overanus Rathbun (Drawn by R. Cries). 634 PALAEONTOLOGY, VOLUME 15 transverse branchial ridges are separated by a broad furrow. A sharp line lies just behind the anterior ridge. The two ridges swing forward as they approach the lateral wall of the carapace. The anterior ridge continues on the lateral wall forward and downward then swings upward to the lower edge of the orbit. The posterior branchial ridge splits at the edge of the lateral wall of the carapace. One short branch runs anteriorly to the sharp line just behind the anterior branchial ridge. The other branch runs backward and down- ward about one-third of the distance to the posterior-lateral margin. The second ridge begins again on the lateral wall of the carapace, paralleling the carapace edge until it disappears near the lower end of the mouth frame. A third ridge is found between these two ridges on the anterior-lateral wall of the carapace. It extends to near the upper end of the mouth frame. Between the lower edge of the orbits, where the anterior branchial ridge terminates, and the upper part of the buccal frame, where the middle ridge terminates, there is a small triangular notch in the buccal frame margin. The pleural suture begins at the posterior apex of this notch and runs in the depression between the upper and middle ridges, through the furrow between the two branchial ridges until it crosses the trend of the posterior branchial ridge on the smooth area below the point where the ridge splits, and continues on to the posterior-lateral margin. Moults of Dakoticancer. Of the more than 4,000 specimens of Dakoticancer collected twenty specimens of moults were found. All were drawn from 2,500 decapods collected near Mobridge, and are from the Zone of Baculites grandis Hall and Meek (Maestrichtian). Each moult is preserved in Salter’s position, but with the carapace flipped upward and completely overturned (text-fig. 3). This position has not been described as the Salter’s position before and therefore expands the meaning of the term for decapods. In most specimens the carapace is also rotated about a vertical axis and about the sagittal axis. That these specimens are indeed moults is proved by several specimens on which the part of the carapace outside the pleural sutures still rests on the sternum (PI. 123, figs. 8-13). The carapaces are apparently rotated because the mode of attachment of the middle part of the carapace to the moulting foundation was non-existent or very weak. As the crab escaped from its old exoskeleton the carapace was probably flipped forward, stopping in a vertical position as is normal in Salter’s position. Any water currents made by the escaping animal were enough to disturb the equilibrium and the middle part of EXPLANATION OF PLATE 123 Figs. 1-13. Exuviae of Dakoticancer overanus Rathbun. All figures X 1, stereo. Fig. 1. Ventral view of specimen showing left side of sternum and carapace which has flipped over and rotated 80°, specimen 4-1039 (USNM 173394). Figs. 2-3. Female with carapace flipped over and rotated about 33°. 2. Dorsal view looking down into inside of carapace. 3. Ventral view, specimen 10-120 (USNM 173395). Figs. 4-5. Female with carapace flipped over and rotated about 10°. 4. Dorsal. 5. Ventral, specimen 4-395 (USNM 173396). Figs. 6-7. Specimen with carapace flipped over and rotated 95°, specimen 4-2002 (USNM 173397). Figs. 8-10. Female with carapace flipped over and rotated 10° clockwise. 8. Dorsal. 9. Ventral. 10. Anterior showing portion of carapace outside pleural suture still on sternum, specimen 4-1131 (USNM 173398). Figs. 11-13. Male with carapace flipped over and rotated 106° counterclockwise. 11. Dorsal. 12. Ventral. 1 3. Anterior showing portion of carapace outside the pleural suture still on the sternum, specimen 4-824 (USNM 173399). Palaeontology, Vol. 15 PLATE 123 BISHOP, moults of Dakoticancer '"‘'4 . ' ^'' '. ''h:' ,rt "I. •(> . {i> ‘4. ... . j;-,v' i. ii ^ .rl4 .'.. BISHOP: MOULTS OF DAKOTICANCER OVERANUS 635 the carapace tipped over and slipped back over the empty moulting foundation. This accounts for the carapace being reversed as well as being upside-down. The rotation around the vertical and around the sagittal axes can be accounted for by the anterior part of the moulted gastric lining catching on the moult, by water currents caused by the escaping crab, or by non-moulting water currents. TEXT-FIG. 3. Orientation of moults of Dakoticancer overainis Rathbun. The short-heavy arrow is the anterior end of the medial axis of the sternal plastron. Each moult is oriented with this arrow pointing upward. The long- heavy arrow is the medial axis of the carapace and points toward the anterior. The short-light arrow perpendicular to the medial axis points the direction the carapace is tilted downward. CONCLUSIONS All the specimens described are preserved in Salter’s position. They are not like other specimens of brachyurans found in Salter’s position because the central part of the carapace is flipped completely over and rotated, not held and fossilized at right angles to the sternum (Schafer 1951, p. 233; Glaessner 1969, p. R431). Twenty of 2,500 specimens were found preserved in the Salter’s position. In contrast, most of the other specimens in this collection are preserved with the carapace in its normal position (but split along the pleural sutures) and with the appendages spread out in a relaxed position around the body, conditions interpreted as evidence for burial 636 PALAEONTOLOGY, VOLUME 15 of corpses killed in recurring mass killings of the Dakoticancer community. The con- sistent mode of preservation of specimens in the Salter’s position and the demonstration of some of these being moults lends credibility to the hypothesis that most of the speci- mens are not moults and were corpses when buried (or a greater number would have been found in Salter’s position). What remains unknown is how many of the specimens with the carapace split along the pleural sutures but with the carapace in its normal position are moults in which the carapace simply flopped back to its normal position after the crab escaped. Acknowledgements. The specimens were collected while working at the University of Texas. Aid was received from the South Dakota Geological Survey; the Department of Geology, South Dakota School of Mines and Technology; the Department of Geological Sciences, University of Texas, and the Geological Society of America. Mrs. Sue Colson typed the paper at Georgia Southern College, the plates were assembled by Mrs. Ruth Gries, and the manuscript was critically read by Dr. P. U. Rodda, Curator, California Academy of Science. All of the above are thanked for their aid. Plate and text contributions were provided by the Geology Foundation, University of Texas. REFERENCES ORACH, p. 1939. Mue et cycle d’intermue chez les crustaces Decapodes. Ann. Inst. Oceanogr. (Paris). 19, 103-391. GILL, J. R. and COBBAN, w. A. 1966. The Red Bird section of the Upper Cretaceous Pierre Shale in Wyoming. Prof. Pap. U.S. geol. Surv. 393-A, 1-73, 12 pis. GLAESSNER, M. 1929. Zur Kenntnis der Hiiutung bei Fossilen Krebsen. Palaeobiologica, 2, 49-56, pi. 6. ■ 1969. Decapoda in Treatise on Invertebrate Palaeontology (Part R), 2, 399-566, 626-628, Geol. Soc. Am. and Univ. of Kansas Press. Lawrence (Kansas). PASSANO, L. M. 1960. Molting and its control in The Physiology of Crustacea (H. T. Waterman, ed.), 1, 473-536. Academic Press, New York. RATHBUN, M. J. 1917. New species of South Dakota Cretaceous crabs. Proc. U.S. natn. Mas. 52, 385-391, pis. 32-33. SCHAFER, w. 1951. Fossilisations-Bedingungen brachyurer Krebse. Abh. senckenb. natnrforsch. Ges. 485, 221-238, pis. 53-54 GALE A. BISHOP Department of Geology Georgia Southern College Statesboro Georgia 30458, U.S.A. Typescript received 18 November 1971 OBSERVATIONS ON THE SHELL STRUCTURE OF TRIASSIC AMMONOIDS by E. T, TOZER Abstract. Despite mineralogical alteration Triassic ammonoids provide significant data on the layers forming the shell wall and umbilical plug. Two layers (outer and inner test) are recognized in the outer wall. Outer test incorporates growth lines, ornament and colour markings, defines the fundamental architecture, was evidently secreted only on the flanks and venter, probably only at the mantle edge. Inner test deposits are secondary, modifying in various ways the chamber interior, and were secreted both dorsally and ventrally, probably at the mantle surface. Discotropites has a dorsal secondary layer within the phragmocone described as dorsal shield and interpreted as a manifestation of the inner test. Nathorstites, in contrast, has no dorsal deposit in this position. Secondary deposits secreted within the flanks and venter tpreseptal layer of Guex) occur in both phragmocone and in part, but not all, of the body chamber of many Triassic ammonoids. The position of this layer may have exercised buoyancy control. In Ceratitida wrinkle-layer (Runzelschicht) with fingerprint pattern was deposited only on the dorsum and above the umbilical plug and is thus comparable with the Nautilus black layer in position, although ditferent in composition and texture. Like the outer test it was probably a secretion of the mantle edge. This kind of wrinkle-layer evidently characterizes Ceratitida and Palaeozoic Ammonoidea but not Phylloceratida, Lytoceratida and Ammonitida. Internal moulds of the flanks of Nathor- stites have markings (ritzstreifen) with a pattern unlike that of the wrinkle-layer. Accordingly, with Mojsisovics, they are interpreted as impressions of the inner surface of the shell wall, not of wrinkle-layer. Maclearnoceras enode sp. nov. is described. Observations made in the course of a survey of the Triassic Ammonoidea (Tozer 1971dr) provide new information on some of the less well-known features of the ammo- noid shell. This work raises problems of terminology and interpretation and has yielded data which may have phylogenetic significance. Study of the shell structure of Triassic ammonoids is hampered by the mineralogical alteration that has affected most of the material known from North America and the classical localities of Europe and Asia. Most of the features to be described are relatively gross, and amenable to study under reflected light with a binocular microscope. A few thin sections were examined but as a result of mineralogical alteration they proved unrewarding. The features considered in this paper are: 1. The outer test. 2. Secondary deposits, consisting of the Preseptal layer. Dorsal shield, and Umbilical deposits. 3. Wrinkle- layer (Runzelschicht) and Ritzstreifen. The structure of the septa and siphuncle are not considered. OUTER TEST Following Casey (1961, p. 178) the term ‘outer layer of test’ (abbreviated here to ‘outer test’) is applied to the layer of the shell wall which preserves the growth lines and ornament and defines the fundamental architecture of the shell. An example of Owenites koeneni Flyatt and Smith (PI. 126, fig. 3) shows colour markings in this layer. The presence in the outer test of these three features — growth lines, ornament and colour bands — leaves little doubt that the outer test of ammonoids corresponds to the porcel- lanous ostracum of Nautilus, a secretion of the apertural edge of the mantle (Stenzel, in Moore 1964, p. K77). This layer is also known as the outer prismatic layer (Erben [Palaeontology, Vol. 15, Part 4, 1972, pp. 637-654, pis. 124-128.] 638 PALAEONTOLOGY, VOLUME 15 et al. 1968); spheriilitic prismatic layer (Mutvei 1964, p. 241) and outer porcelaiious [5/c] layer (Flower 1964, p. 9). Casey (1961) and Birkelund and Hansen (1968, p. 75) have shown that in some Cretaceous ammonoids (Roloboceras and Saghalinites) the outer test wedges out against the flank or venter of the preceding whorl, and was not secreted on the dorsum. TEXT-FIG. 1. Nathorstites macconnelli (Whiteaves). Camera lucida drawings (x 3) of polished sections, (a) GSC No. 28027 (GSC loc. 42333). (b) GSC No. 28028 (GSC loc. 68264). Liard Formation, Liard River, 3|- miles west of Hell Gate, British Columbia. Solid black indicates thickness of test. Interrupted line is whorl outline defined by internal mould, be, body chamber. Stipple pattern, umbilical deposit. Arrows indicate surfaces on which wrinkle-layer is preserved. Lines traversing umbilical deposit are interpreted as sectional view of wrinkle- layer. Test and umbilical deposits clearly differentiated in (a), less so in (b). Sections of shell walls of Triassic ammonoids, with rare exceptions, do not show discrete layers. Nevertheless, as mentioned below, there is clear evidence that the struc- ture was composite. The surface with ornament and growth lines is referred to as outer test although its thickness, in relation to that of the whole shell wall, cannot generally be determined. Specimens of Nathorstites maccoimelli (Whiteaves) show the shell wall to be much thicker in the umbilical area than on the flanks and venter (text-fig. 1) and also that the septa are attached to the wrinkle-layer of the preceding whorl (PI. 124, fig. 2), indicating that within the phragmocone no layer continuous with the outer shell wall secreted on the dorsum. Other specimens (PI. 124, fig. 4; PI. 125, figs. 3, 4) show a com- TOZER; SHELL STRUCTURE OF TRIASSIC AMMONOIDS 639 parable situation within the body chamber, with the outer wall of the body chamber in the umbilical area wedging out against the wrinkle-layer covering the outer test of the penultimate whorl (text-fig. Zb). Whether or not the wedge near the umbilicus includes a layer of outer test has not been determined. Despite the nature of the Nathorstites shell wall one should not conclude, with Palframan (1967, p. 1130), that planispiral ammo- noids in general did not secrete a dorsal wall. The dorsal shield layer, described below, and interpreted as a manifestation of the inner test, refutes this generalization. SECONDARY DEPOSITS Triassic ammonoids show deposits with surface textures unlike that of the outer test and which are secondary in that they appear to be moulded to structures already defined by the outer test. From comparison with Nautilus and well preserved Jurassic and Cretaceous ammonites this is to be expected, for in Nautilus the porcellanous ostracum is lined by two secondary layers: the nacreous and inner prismatic layers (Mutvei 1964, Erben et al. 1968), these, unlike the porcellanous ostracum, being secreted at the surface of the mantle. Mineralogical alteration obscures comparable layers in Triassic ammo- noids in which the shell wall is mineralogically and texturally uniform in section, pre- sumably due to recrystallization. Cladiscites toniatus (Bronn), according to Mojsisovics (1873, p. 73) provides an exception, showing two layers: the outer with sculpture, the inner being smooth, transparent, and nacreous. Other observations, described below, indicate that at least two layers form the outer wall. For the inner, the name ‘inner layer of test’ (abbreviated here to ‘inner test’) may be used (Casey 1961, p. 178). Casey used the term, not only for the portion lining the flanks and venter, but also for the material secreted on the dorsum. In Nautilus the most substantial layer inside the porcellanous ostracum is the nacreous layer, and the inner test presumably corresponds with this. Nathorstites macconnelli shows that the texture of the inner and outer surfaces of the shell wall are different, and suggests that the inner is lined with a secondary deposit. The outer surface (i.e. surface of the outer test) shows the characteristic growth lines (e.g. PI. 125, fig. 4). The inner surface is reflected by internal moulds which may be more or less smooth (PI. 125, figs. 1,2) or pitted with the ritzstreifen (PI. 125, fig. 4) discussed below. The actual inner surface of the shell wall has been observed only near the umbilicus. Two specimens, GSC Nos. 28016 (PI. 125, figs. 3, 4) and 28231 (PI. 124, figs. 3, 4) show this surface particularly clearly. Both have a band of smooth shell material restricted to the innermost part of the flank, overlying the wrinkle-layer on the underlying whorl. Towards the axis of the umbilicus this smooth material merges with the mass of crystal- line calcite sealing the umbilicus (PI. 125, fig. 4E). The smooth material is clearly the inner surface of the shell wall of the body chamber, wedging out against the underlying whorl, a relationship shown diagrammatically by text-figure Zb. The first specimen, a complete phragmocone with most of the body chamber broken off, shows the relation- ship within the body chamber. The second specimen, which is nearly complete, shows the original limit of the smooth wedge. Interpretation of no. 28231 is facilitated by reference to No. 28232, which although presumably not fully grown, is nevertheless complete because it preserves the peristome with a notched rostrum (PI. 124, figs. 5, 6). The length of the body chamber is IJ whorls. The smooth layer and length of the 640 PALAEONTOLOGY, VOLUME 15 body chamber, although not shown by the illustrations, were observed by breaking the specimen. The extent of the smooth band on the inner flank corresponds with that of the rostrate peristome, confirming that the limit of the smooth band shown by no. 28231 (PI. 124, fig. 4) indicates the terminal point of attachment of the body whorl. These specimens thus show the surface of both the outer and the inner test. The outer, with growth lines, compares with that of the Nautilus porcellanous layer, and the smooth inner surface is like the inside of the nacreous layer. Triassic ammonoids also show secondary material in three situations: (i) on the flanks and venter, where they form the preseptal layer (Guex 1970); (ii) on the dorsum, as a layer obliterating or modifying the ornament of the outer test of the preceding whorl, the ‘dorsal shield’ of Casey (1962, p. 264); (hi) in the umbilicus, where they form a callus, or plug. Preseptal layer. On the flanks and venter diflFerentiation of inner and outer test is based partly on direct observation, but mainly on the differences shown by comparing the ornament of internal moulds with that of the outer test. A direct observation is provided by Maclearnoceras enocle n. sp. (PI. 128, figs. 3, 4), described in the appendix. The only specimen is so preserved as to leave little doubt that the shell wall, at least on the initial part of the body chamber, is composed of these two layers. The outer test, attaining a maximum thickness of 0-2 mm is corrugated to form ribs on both the inner and outer surface (PI. 128, fig. 3). The inner test (maximum thickness 0-5 mm) has a corrugated outer surface, moulded to that of the outer test, but the inner surface, lining the body chamber, and the surface to which the septa were attached, is smooth. This specimen provides grounds for interpreting many Triassic ammonoids in which the ornament and whorl section, as preserved on an internal mould, differs con- spicuously from that shown by the outer surface. These discrepancies are shown in Frank! tes sutherlandi (McLearn) (PI. 128, figs. 5-9) in which the periphery, where the EXPLANATION OF PLATE 124 Specimens coated with ammonium chloride. Figs. 1-7. Nathorstites maccowielli (Whiteaves). 1, (Xl6), GSC No. 28028, Liard Formation, 323 feet below Triassic-Cretaceous contact, Liard River, 3|- miles west of Hell Gate, British Columbia (GSC loc. 68264), Upper Ladinian, Sutherlandi Zone, wrinkle-layer in umbilical area of sectioned specimen (text-fig. 16). 2, (x4), GSC No. 28026, Liard Formation, 280 feet below Triassic- Cretaceous contact, Liard River, 3-| miles west of Hell Gate, British Columbia (GSC loc. 42335), Upper Ladinian, Sutherlandi Zone, lateral view of portion of incomplete phragmocone with remnants of last eight septa visibly attaehed to the wrinkle-layer of the preceding whorl, arrow indicates position of last septum. 3, 4, GSC No. 28231, horizon and locality as fig. 1 (GSC loc. 68264), 3 (x 1 ) lateral view of whole specimen, mostly preserved as internal mould except in umbilical area, body chamber length 1 whorl, 4 (x 16) oriented as 3, shows detail of umbilical area, described in text. 5, 6 (x 1), GSC No. 28232, Liard Formation, 310 feet below Triassic-Cretaceous contact, Liard River, 3f miles west of Hell Gate, British Columbia (GSC loc. 42334), Upper Ladinian, Sutherlandi Zone, ventral view (5) shows notched rostrum, lateral view (6) the side of the peristome, body chamber length If whorls. 7 (x8), 8 (xl6), GSC No. 28230, Liard Formation, about 40 feet below Triassic-Cretaceous contact, Liard River, 2f miles west of Hell Gate, British Columbia (GSC loc. 42351), Upper Ladinian, Sutherlandi Zone. 7 illustrates lateral view of final portion of phragmocone, from umbilicus to venter (7, top), wrinkle-layer cover sentire dorsum, on greater part ridges are radial but at umbilicus they sweep into a spiral (8). Palaeontology, Vol. 15 PLATE 124 TOZER, ammonoid shell structure TOZER: SHELL STRUCTURE OF TRIASSIC AMMONOIDS 641 outer test is preserved, has rounded ventral shoulders, abrupt rib terminations at the shoulder, and a sulcus on the siphonal line (PI. 128, fig. 8). The mould, in contrast, on both the phragmocone and the initial part of the body chamber (PL 128, figs. 6, 9) has well defined ventral shoulders and an almost smooth venter, with no trace of a sulcus. Near the aperture, however, there is an abrupt change to a condition where the ornament of the mould and the outer test nearly correspond (PI. 128, figs. 6, 9). Obviously the inside and outside of the Frankites sutherlandi shell wall were very different, except near the aperture. The interpretation of this discrepancy (text-fig. 2) follows that provided by Guex (1970) to account for features shown by internal moulds of Jurassic Dactylio- ceratidae, in which the initial part of the body chamber has a zone of smooth ornament followed abruptly by a zone of sharp ornament (Guex 1970, p. 2, PI. 2, fig. 3). Guex named the material responsible for suppressing the ornament in the smooth zone as the preseptal layer. In his interpretation the preseptal layer and ‘conotheca’ (i.e. outer test) are only partly in contact with a hollow space occupying the summits of ribs and spines. Macleanioceras enode gives no indication that a hollow space existed, on the contrary, the preseptal layer was apparently firmly cemented to the outer test with the outer surface of the preseptal layer faithfully reproducing the rib pattern of the inner surface of the outer test (PI. 128, fig. 3). The lateral ribs of Frankites sutherlandi are more sharply defined on the test than the mould, indicating that the preseptal layer was deposited on the flanks as well as the venter. A development of preseptal layer on the flank is clearly shown by Muensterites glaciensis (McLearn), the ribs on the outer test bearing tubercles which are obliterated on the internal mould (PI. 127, fig. 6). The function of the preseptal layer is unknown. It may have been merely a layer to strengthen and make smooth the inside of the body chamber. But if this alone was its function, why was it not secreted throughout the whole length of the body chamber ? Or it may have served some role in muscle attachment, like the annular ridges and eleva- tions known at the posterior end of the body chamber of some nautiloids (Teichert, in Moore 1964, p. K 27) and ammonoids (Jordan 1968). There is also the possibility that it may have contributed to bouyancy control. Being fairly thick, its mass, in relation to the mass of the whole animal, must have been appreciable. The fact that it terminates abruptly, instead of merely tapering off, suggests that its limit in the body chamber, may have been rigidly defined, and that its extent may be related to the regime of the animal. In other words the preseptal layer of the body chamber may have represented a layer of ballast, precisely positioned in relation to the adult aperture, and thus exercising a degree of control over the position of the animal in life. Dorsal shield. The term ‘dorsal shield’ was introduced by Casey (1962, p. 264) for the thick layer of shell secreted on the dorsum of Douvilleiceras, serving to render smooth what would otherwise be a very rough roof for the successive body chambers, rough because the venter of Douvilleiceras, from an early stage, bears strong tubercles. Douvilleiceras from the Queen Charlotte Islands show this feature well (McLearn 1972). One specimen (GSC No. 5014 d) shows that the dorsal shield is not merely a layer secreted in the adult body chamber, but extends back behind the last septum. This specimen also shows lamination of the test, and it appears that the dorsal shield represents a part of the inner test. 642 PALAEONTOLOGY, VOLUME 15 The Triassic ammonoid DiscotropUes has a structure comparable in position but with spiral sculpture. The presence of this layer on DiscotropUes was noted by both Suess (1870, p. 315) and Mojsisovics (1893, pp. o 283, 287). Suess considered it a form of wrinkle-layer (runzelschicht) ; Mojsisovics regarded it as a form of ‘Epidermiden’, but did not employ the name runzelschicht. The thickness varies considerably. Mojsisovics’ illustrations (1893, PI. 130, fig. 13; PI. 131, figs. 1,4, 10) show thin developments of the layer. Of the specimens he studied, a thick development is shown by the original of his Plate 130, fig. 12, for which he illustrated only the suture line. I have seen this specimen at the Geologische Bundesanstalt, Vienna. A published illustration of the h thick development of this layer has been provided by Smith (1927, pi. 11, fig. 8) for DiscotropUes laurae (Mojsisovics). A plaster cast of Smith’s specimen shows the layer to be about 1 mm thick at the venter. Another specimen from British Columbia (PI. 128, figs. 1, 2) shows a mere film, OT mm thick or less. As noted by Suess (1870, p. 315) this spirally sculptured layer has a close counter- part in Jurassic Amaltheus. Most writers who have discussed the Amaltheus layer (e.g. Suess 1870, Zittel 1895, p. 407, fig. 1112; Walliser 1970, pi. 4, fig. 5) have con- sidered it to be wrinkle-layer (runzelschicht). The properties of true wrinkle-layer are discussed below. Two differences distinguish the spirally sculptured layer of DiscotropUes and Amaltheus from wrinkle-layer. First, the Diseotropites layer has a wholly spiral sculpture, unlike the dominantly radial, fingerprint pattern of true wrinkle-layer; second, this layer is of variable thickness, and when thick, much thicker than wrinkle- TEXT-FiG. 2. Diagrammatic median section {a) and whorl section of phragmocone (b) of Frankites siitherlandi (McLearn) based on GSC Nos. 18903 and 28025 (PI. 128, figs. 5-9). O (heavy line) represents the outer test, I (ruled), the preseptal layer, B the place where the pre- septal layer ends as indicated on the internal mould by the change from the zone of smooth ornament to zone of sharp ornament. X marks position of last septum. Thicknesses of layers schematic. EXPLANATION OF PLATE 125 Specimens coated with ammonium chloride. Arrow marks position of last septum. Figs. 1-4. Nathorslites macconnelli (Whiteaves). Lateral views. 1 (xl), 2 (X4), GSC No. 28014, GSC loc. 4235 1 (see explanation, Plate 1 24). Body chamber > 1 whorl, probably complete and adult. 2 shows detail of umbilical area with wrinkle-layer covering outer test and extending about f whorl beyond aperture. 3 (X 1), 4 (x20), GSC No. 28016, GSC loc. 42335 (see explanation, PI. 124). 4 shows detail of inner flank and umbilical area with ritzstreifen pits on internal mould of body chamber (A), surface of outer test with growth lines (B), overlain by wrinkle-layer (C). Near umbilicus wrinkle-layer is overlain by wedge of smooth shell (D), the inner margin of the shell wall of the body chamber. (E) is broken face of body chamber shell wall where parallel to axis of coiling. Boundary between shell wall and umbilical plug not discernible owing to recrystallization. Small granules on (D) are of ammonium chloride. Ritzstreifen certainly present only on body chamber. Palaeontology, Vol. 15 PLATE 125 TOZER, ammonoid shell structure TOZER: SHELL STRUCTURE OE TRIASSIC AMMONOIDS 643 layer, which invariably seems to be a single lamina. ‘Dorsal shield’ seems a more appro- priate term for the Discotropites layer. Because the Discotropites dorsal shield may be thick, thin, or absent it may be interpreted as a secretion of the mantle surface, not of the mantle edge. If so it was probable originally nacreous and a structure of the inner test, comparable with the preseptal layer, but dorsal instead of ventral in position. Umbilical deposits. A number of Triassic ammonoids have the umbilicus sealed by a callus or plug, as does Nautilus pompilius Linne. The Nautilus callus is a milk-white porcellanous substance (Gregoire 1962, p. 9), different in appearance from the nacreous material forming the thick inner layer of the shell wall. Examples of Nathorstites macconnelli preserving shell in the umbilical area have much in common with Nautilus pompilius but owing to mineralogical alteration exact homologies cannot be established. In section, material of one texture forms the shell wall, of another the callus, and in places the boundary between the two is clearly defined (text-fig. 1). Polished and thin sections have failed to resolve the structure of the shell wall. Part is outer test but some, or all, where it is thickest in the umbilical area may have been secondary nacreous material. The position and overall appearance of the callus deposit invites comparison with that of Nautilus. The helicolateral deposits (Nassichuk 1967) described on the Carboniferous goniatite Clistoceras globosum Nassichuk may also be comparable. Adult Clistoceras is im- perforate, like Nautilus pompilius and Nathorstites (e.g. Nassichuk 1967, pi. 28, fig. 10). Sections made by Nassichuk show that Clistoceras, in terms of the architecture of the shell wall, has an unusually undercut umbilicus (Nassichuk 1967, text-figs. 1,2). Although I am unaware of any ammonoid showing comparable undercutting there are other ammonoids with distinctly undercut umbilical walls, e.g. Tropites subquadratus Silber- ling (Silberling 1959, pi. 3, fig. 5). In Clistoceras deposits fill most of the umbilical cavity, including the undercut portion. This material, named the ‘helicolateral deposit’ by Nassichuk, extends well towards the middle of the flank. Specimens from which the outer whorl has been stripped (e.g. Nassichuk 1967, pi. 28, figs. 2, 6) present a unique appearance, particularly as the deposits bear what seem to be growth lines. These features led Nassichuk to conclude that the helicolateral deposits are primary (i.e. secreted in front of the aperture, before secretion of the next whorl), and are not homo- logous with the umbilical callus of Nautilus. On the other hand the helicolateral deposits may be homologous with the callus of Nautilus and Nathorstites since : — firstly, secretion in front of the aperture has not been demonstrated (indeed the example of Clistoceras with a body chamber (Nassichuk 1967, p. 28, fig. 10) shows no deposit beyond the aperture) ; secondly, the presence of helicolateral deposits has been demonstrated only between whorls; and thirdly, the growth lines that suggest a primary deposit may be interpreted as impressions from the outer test of the succeeding whorl. Undercut umbilical walls are not unique ; this feature of Clistoceras is one of degree, not of kind. Therefore the helicolateral deposits of Clistoceras are not unique but owe their extra- ordinary appearance to the unusually undercut nature of the umbilical wall. WRINKLE-LAYER (RUNZELSCHICHT) AND RITZSTREIFEN The terms ‘Runzelschicht’ and ‘Ritzstreifen’ were introduced by the Sandbergers (1850, pp. 58, 93, 121). The English equivalent term for Runzelschicht is ‘wrinkle-layer’ 644 PALAEONTOLOGY, VOLUME 15 (Foord and Crick 1897, p. xx). For Ritzstreifen (Scratch-grazes) there seems to be no wholly satisfactory English term. Foord and Crick’s (1897, p. xx) use of ‘Epidermids’ invites confusion because Barrande (1877) coined the term ‘Epidermides’ to include both Runzelschicht and Ritzstreifen. Clausen (1969, p. 104) treats Ritzstreifen as a synonym of Runzelschicht; House (1971) and Senior (1971) use the term ‘ventral wrinkle-layer’. The nomenclature of Clausen, House, and Senior would be satisfactory if it was entirely certain that Ritzstreifen and Runzelschicht are manifestations (the impression and layer respectively) of one and the same thing, as held by Barrande, Clausen, House, Senior, and others. But it will be shown below that Ritzstreifen may not be the impressions of wrinkle-layer. Accordingly for descriptive purposes it seems more appropriate to adopt the more objective terminology employed by the Sandbergers (and also Mojsisovics), which was to use the different names. Definition of wrinkle-layer {Runzelschicht). Runzelschicht had been recognized by Keyserling (1846, pp. 274-275) before the publication of the Sandbergers work. The feature recognized by Keyserling and the Sandbergers is a thin layer superimposed on the outer test of Devonian goniatites. There is no doubt that Keyserling and the Sandbergers were describing a distinct layer which they identihed only on the dorsum. To Keyserling (1846, p. 275) it was a deposit laid down ‘an der sogenannten Bachseite des Umganges’; to the Sandbergers (1850, p. 58, footnote), a deposit on the ‘Ruckenoberflache’, i.e. a deposit on that part of the outer test of the earlier formed whorl which forms the dorsum of the later chamber. Since the time of publication of these early papers the names Runzelschicht or wrinkle- layer have been applied to a number of features. Some are certainly closely comparable with the layer named by the Sandbergers, others less certainly so, some certainly not. In order to avoid ambiguity a definition, and reference to a typical example, is essential. Keyserling (1846, p. 275) recognized wrinkle-layer on ammonoids now referred to Tornoceras, Beloceras and Ponticeras (?). In addition to these three the Sandbergers (1850, p. 58) described wrinkle-layer on representatives of what are now known as Maenioceras, Manticoceras, and Pharciceras. Eor present purposes the layer as de- veloped on Tornoceras may be taken as typical and representative. This seems entirely EXPLANATION OF PLATE 126 Specimens, except 3, coated with ammonium chloride. Arrows mark position of last septum. Figs. 1 (x 1), 2 (x 16). Nordopliiceras spatlii (Kummel and Steele). GSC No. 28018. North side Mill Canyon, about 2 miles north-east of Crittenden Ranch, Elko County, Nevada (GSC loc. 64685). Lower Triassic, Smithian (Silberling and Tozer 1968, p. 29). Most of specimen preserves outer test with wrinkle-layer (fig. 2) superimposed, extending for 8 mm, measured at venter, beyond aperture. Figs. 3 (x 1), 4 (x 16). Oweiiites koeiieni Hyatt and Smith. GSC No. 28017. Locality as above. Mostly preserved with outer test retaining reddish-brown colour bands (dark in figure 3). Small patches of wrinkle-layer (fig. 4) preserved for f volution beyond aperture. Figs. 5 (x2), 6 (x 16). Juvenites septentrioualis Smith. GSC No. 28019. Locality as above. Phragmo- conc with outer test and wrinkle-layer. Figs. 7 (x I), 8 (x 16). Proarcestes sp. GSC No. 28020. Bluff 10 miles south-east of Mount Mary Henry, British Columbia (GSC loc. 68284). Lower Ladinian, Poseidon Zone. Complete phragmocone preserving outer test and wrinkle-layer (fig. 8), removed from complete specimen with If whorls of body chamber. Palaeontology, Vol. 15 PLATE 126 TOZER, ammonoid shell structure TOZER; SHELL STRUCTURE OF TRIASSIC AMMONOIDS 645 justified : it is a genus on which the layer was studied by Keyserling and the Sandbergers; it is clearly shown on some of the old illustrations (e.g. Sandberger and Sandberger 1850, pi. 10, fig. 14); and excellent photographic illustrations have been provided by House (1965, pi. 6, fig. 42, pi. 8, figs. 71, 72, pi. 9, figs. 78, 79). The features of wrinkle-layer are: it is thin, apparently a single lamina, encrusting the surface of the outer test on the dorsum; it is also characterized by a very distinctive texture of small ridges and furrows, resembling human fingerprints. Walliser (1970) and House (1971) have distinguished several kinds of pattern and suggested that a more refined classification may have systematic significance. Keyserling’s comparison with fingerprints nevertheless remains apt to characterize, at least in a gross sense, the most distinctive feature of wrinkle-layer, which has for long been recognized on Triassic and Carboniferous ammonoids in addition to those of Devonian age. Very different from the layer with fingerprint pattern are a number of features which have been described in the literature as wrinkle-layer. First there is the spirally sculp- tured layer of Discotropites and Amaltheus, which does not closely resemble true wrinkle-layer. There is also the granular wrinkle-layer, the ‘grbbkornig runzelschicht’ of Mojsisovics 1873, p. 69, pi. 24, fig. 2 described on Sageceras haidiugeri (Hauer). Mojsisovics described this as a form of wrinkle-layer, unlike the typical variety, and as resembling in texture the black layer of Nautilus. Examination of the specimen described by Mojsisovics (illustrated here PI. 127, figs. 4, 5) as preserving this kind of layer, has failed to convince me that it is other than an inorganic crust, deposited on the outer test. Wrinkle-layer as interpreted by Miller (1947, pi. 3), and Teichert {in Moore (ed.), 1964, p. K15) is a feature revealed by the etching or weathering of the nacreous layer and is certainly not the same as the thin discrete layer discriminated by the Sandbergers. On the other hand the dorsal wrinkle-layer identified by Senior (1971) on Jurassic Graphoceratidae appears to be closely comparable, in terms of thickness and position, with true wrinkle-layer. The texture, however, is not the same, the wrinkling being much finer, and the fingerprint pattern absent. Definition of Ritzstreifen. The Sandbergers (1850, p. 121) introduced the term Ritz- streifen to describe markings preserved on moulds providing an impression of the sur- face of the inside of the lateral and ventral parts of the ammonoid whorl. In the same work (1850, p. 93) they also use the term ‘Einritzung des Manteleindrucks’ for this feature. What they were describing is clearly shown by their illustration of ‘‘Goniatites lamed var. cordatus' (1850, pi. 8, fig. 6b) (a representative of Manticoceras) on which the internal mould shows more or less radial striae indicating that a patterned surface lined the lateral and ventral parts of the whorl. Photographic illustrations of Manticoceras ritzstreifen have been provided by Clausen (1969, pi. 26, figs. 10, 12, 13, 14). The pattern displayed by the ritzstreifen of Manticoceras resembles, at least in a gross way, that of wrinkle-layer. In other ammonoids, however, the impressions derived from surface of the lateral and ventral interior take the form of pits instead of ridges, indicating a surface wholly unlike that of wrinkle-layer. Judging from the observations of Walliser (1970, p. 121, text-fig. 5D, pi. 2, figs. 1, 6) this is the case in Tornoceras and Maenioceras, both of which have wrinkle-layer with fingerprint pattern on the dorsum, and ritzstreifen in the forms of pits on the flank. Ritzstreifen, unlike wrinkle-layer, seem to be known only from impressions and in consequence their interpretation is the more difficult. The 646 PALAEONTOLOGY, VOLUME 15 a b TEXT-FIG. 3. Diagrammatic restored side view (a) and section (b) of Nathorstites maccoimelli (Whiteaves) about x 1 but thickness of shell layers and relief of surface responsible for ritzstreifen schematic. (a) based only on GSC No. 28014 (PI. 125, fig. 1); black area indicates where wrinkle-layer is actually preserved; patterned area its presumed original extent in relation to the aperture. (b) shows distribution of wrinkle layer (dotted line) on outer and inner whorls. Solid black indicates shell wall, with small elevations at posterior end of body chamber forming ritzstreifen (PI. 125, fig. 4). Stipple pattern indicates umbilical deposit, be, the start of the body chamber. Septa of inner whorls not shown. Wrinkle-layer of outer whorl encrusting umbilical deposit is shown on Plate 124, fig. 1 ; the layer passing beneath the wedged out portion of the body chamber on Plate 124, fig. 4 and Plate 125, fig. 4. The layer is shown to traverse the umbilical area of inner whorls from interpretation of GSC No. 28028 (text-fig. lb), but the evidence that it does so is not conclusive. relationship between the two will be discussed after description of wrinkle-layer and ritzstreifen in Triassic ammonoids. Wrinkle-layer of Triassic Ammonoids. The occurrence of wrinkle-layer closely resembling that of Tornoceras is well known in Triassic ammonoids (Mojsisovics 1873-1875). A review of the literature and my own studies have established the presence of wrinkle- layer with fingerprint pattern in the following families of Ceratitida: Ophiceratidae (Nordopliiceras); Melagathiceratidae (Juvenites); Paranannitidae {Paranannites, Owen- ites); Aspenitidae (Aspenites); Megaphyllitidae (Megap/iyllites); Gymnitidae (P/tic/to); TOZER: SHELL STRUCTURE OF TRIASSIC AM MONOIDS 647 Ptychitidae {Ptychiles)’, Carnitidae (Caniites); Pinacoceratidae (Pompeckjites)', Daniibi- tidae (Arctolnaigarites); Longobarditidae (Intomites); Nathorstitidae {Natlwrstites); Cladiscitidae (Paracladiscites); Arcestidae (Proarcestes, Arcestes); Sphingitidae (Spliingites) ; Joannitidae {Joaimites) ; Cyrtopleuritidae {Drepanites, Hauerites) ; Lobitidae {Lobites)', Didymitidae {Didymites). Of these I have examined well preserved representa- tives of all except Canutes, Paracladiscites and Sphiugites. In every case the presence of wrinkle-layer has been confirmed, or observed for the first time, on the dorsum. None, with the possible exception of MegaphyUites (discussed below) provide any indication that a comparable layer lined the ventral and lateral parts of the chambers. Good preser- vation is necessary in order to see wrinkle-layer. This limits assessment of its systematic significance in that it is often impossible to decide whether absence is true, or due to un- favourable preservation. Most that show it clearly are smooth forms. None of the really rough-shelled Ceratitida are known to have wrinkle-layer. It is therefore of some significance that Drepanites (PI. 127, figs. 1, 2), a relatively smooth member of a rough- shelled family, shows wrinkle-layer, indicating that the ability to secrete this layer was not lost in all rough-shelled families of Ceratitida. The wrinkle-layer of the ammonoids listed above has the fingerprint pattern. There is considerable variation in the spacing of the ridges. They are closest in forms like Juvenites (PI. 126, fig. 6) (about 20 ridges to the mm); most widely separated in Proar- cestes (PI. 126, fig. 8) (about 4). Natlwrstites (about 10) falls in between and has a similar pattern to Tornoceras. Further work on the lines proposed by Walliser (1970) and House (1971) will perhaps result in a more refined classification. Natlwrstites macconnelli (Whiteaves) shows the pattern and relationships of the wrinkle-layer particularly clearly. It is a layer of calcareous material, about 0-04 mm thick, superimposed on the outer test (PI. 125, fig. 4). GSC No. 28230 (PI. 124, figs. 7, 8) shows the arrangement of the ridges to be radial over the greater part of the flank, and sweeping into a spiral in the immediate umbilical area. The details of the pattern in the umbilical area are particularly clearly shown by No. 28028 (PI. 124, fig. 1). No. 28026 (PI. 124, fig. 2) shows the layer within the phragmocone, with the septa of the next whorl attached. The layer has also been observed on nuclei 9 mm in diameter. Nos. 28231 and 28016 show the relationships of the wrinkle-layer in the umbilical area, both at the aperture (PI. 124, figs. 3, 4) and within the body chamber (PI. 125, figs. 3, 4). On these specimens the wrinkle-layer near the umbilicus is covered by smooth (originally nacreous?) shell material representing the inner test of the wedged out portion of the succeeding volution. Furnish and Glenister (1971) have described a specimen of Mescalites discoidalis (Bose) with wrinkle-layer similarly overlain by smooth material. Polished sections of Natlwrstites (text-fig. 1) show a dark line traversing the umbilical deposit. These are at the same ‘level’ as the wrinkle-layer and it is suggested that they indicate its position in the umbilical area. No. 28014 (PI. 125, figs. 1, 2), probably complete and fully grown, shows wrinkle-layer over- lying the umbilical plug and on the inner flank, where it extends about one-quarter volution beyond the aperture. On this specimen the layer ends abruptly and the terminal boundary where preserved on the inner flank, is radial. This boundary may indicate the true edge and limit of the layer. If so, and if the radial course was followed to the venter, the complete specimen would have had the appearance indicated by text-figure 3a. Many specimens of Natlwrstites, preserved as internal moulds, show an impression u u C 9202 648 PALAEONTOLOGY, VOLUME 15 of the inside of the flanks and venter. None shows a trace of the pattern of the wrinkle- layer but some show minute pits which may be described as ritzstreifen. Ritzstreifen of Triassic Ammonoids. The best known occurrence of ritzstreifen among Triassic ammonoids is that of Megaphyllites (Mojsisovics 1873, pp. 45-46). Mojsisovics described them as ‘Ritzstreifen des Manteleindruckes’, presumably implying that the markings provide an impression of the inner surface of the shell wall. Walliser (1970) has described these markings, interpreting them as impressions of wrinkle-layer. On the internal mould of Megaphyllites humilis (Mojsisovics) the ritzstreifen form ridges and pits (PI. 127, fig. 3). Near the umbilicus they are spirally arranged ridges; at mid-flank more or less radial pits and ridges; radially arranged pits, alone, form the outer part; and on the outer third all markings disappear. They are most prominent near the aperture. Behind the aperture they fade, and a quarter-volution back, those on the flank have disappeared. The spiral marks near the umbilicus persist further, to an unknown extent. Mojsisovics (1873, p. 47) mentions that Megaphyllites humilis shows wrinkle- layer as well as ritzstreifen but it is not shown on his illustrations, nor is it visible on the specimens 1 have examined. The markings have an appearance not unlike the impression of wrinkle-layer. However there is no evidence that they are impressions of a thin dis- crete layer encrusting the inner wall of the body chamber in the way that wrinkle-layer encrusts the test on the dorsum. The ritzstreifen of Nathorstites maceonnelli (PI. 125, fig. 4), like those on the outer flank of Megaphyllites humilis, form pits in the internal mould. On the illustrated speci- men they are most prominent on the inner third of the flank at the posterior part of the body chamber. Towards the venter and also in an adoral direction they fade and eventually disappear. Comparable, but less prominent, ritzstreifen are present on GSC No. 28231 (although not visible on PI. 124, fig. 3). Most specimens of Nathorstites mac- connelli show no ritzstreifen; whether this indicates that their presence is rare, or depends on the preservation, is not known. Relationship between wrinkle-layer and Ritzstreifen. The evidence provided by Triassic ammonoids suggests that wrinkle-layer was deposited only on the dorsum and in the EXPLANATION OF PLATE 127 Specimens coated with ammonium chloride. Arrows mark position of last septum. Figs. 1 (x 1), 2 (x 16). Drepanites hyatti rutherfordi McLearn. Topotype, GSC No. 28021. Pardonet Formation, McLay Spur, Peace River, British Columbia (GSC loc. 9146). Middle Norian, Ruther- fordi Zone. Partly internal mould, partly with outer test. Small patch of wrinkle-layer (fig. 2) present about one-quarter volution beyond aperture. Fig. 3 (X4). Megaphyllites Immilis (Mojsisovics). GSC No. 28022. Hallstatt Limestone, Sandling, Austria. Upper Carnian, Tropites subbullatus Zone. GSC loc. 19548 (Exchange from Stanford University, LSJU loc. 28187). Last half whorl is internal mould with impression of mouth border and ritzstreifen. Figs. 4 (X 1), 5 (x 16). Sageceras haidingeii (Hauer). Geol. Bundesanstalt, Vienna, No. 2467. Hallstatt Limestone, Raschberg, Austria. Lower Carnian, Trachyceras aonoides Zone. Phragmocone with slightly abraded outer test. Fig. 5 shows surface at beginning of outer whorl and is part illustrated in Mojsisovics (1873, pi. 24, fig. 2). Fig. 6 (X 1). Miiensterites glaciensis (McLearn). Topotype, GSC No. 9536. Talus, west slope East Glacier Spur, Peace River, British Columbia (GSC loc. 9797). Upper Ladinian, Sutherland! Zone. Portion with smooth ribs is internal mould; with tubcrculate ribs, outer test. Palaeontology, Vol. 15 PLATE 127 TOZER, ammonoid shell structure TOZER: SHELL STRUCTURE OE TRIASSIC AMMONOIDS 649 umbilical area. In Nathorstites (as in Maenioceras and Tornoceras) the ritzstreifen on the flanks form a pattern wholly unlike that of the wrinkle-layer. Everything about the wrinkle-layer and ritzstreifen of these ammonoids, at least, suggests that the two are different things. In Megaphyllites and Mcmticoceras, on the other hand, the ritzstreifen patterns are not unlike those of wrinkle-layer. Many authors: Suess (1870), Barrande (1877), Nassichuk (1967), Clausen (1969), Walliser (1970), House (1971), and Senior (1971) have, for one reason or another, concluded that wrinkle-layer lines the whole (dorsal, umbilical, lateral, and ventral parts) of the ammonoid chamber. Implicitly or explicitly they interpret ritzstreifen as the impressions of wrinkle-layer. Before con- sidering the evidence that wrinkle-layer may line the whole chamber I will recapitulate the relevant data provided by Nathorstites. This seems particularly significant because it admits of only one interpretation, which readily accommodates the data provided by Maenioceras and Tornoceras, although not that from Manticoceras. Firstly, the close similarity between the wrinkle-layer on the dorsum of Tornoceras with that of Nathor- stites leaves no doubt that the two are strictly comparable. To describe the layer on the dorsum of Nathorstites as wrinkle-layer is fully justified. Secondly, inside the body chamber and the phragmocone the wrinkle-layer passes beneath the wedged out portion of the succeeding whorl. This is the most significant feature revealed by Nathorstites. Preservation is good, the relationship is clear. If wrinkle-layer had lined the whole of the chamber it would overlie this wedge; in Nathorstites it clearly does not. This poses a problem in that it seems unlikely that such a distinctive layer would line the entire chamber in some ammonoids and be restricted to the dorsum and umbilical area in others. Now to consider the evidence for the alternative interpretation. Nassichuk bases his conclusion on study of Clistoceras globoswn Nassichuk. One specimen (GSC No. 19965, Nassichuk, 1967, text-figs. 1, 2) is described as having wrinkle-layer lining the inside of the ultimate whorl. This conclusion is based on the interpretation of a black layer, seen in section, as wrinkle-layer, but no part of the specimen actually shows the characteristic wrinkle-layer surface. Nassichuk (1967, p. 239) describes the wrinkle-layer of Clistoceras as black, and as preserved on GSC No. 19966 (1967, pi. 28, fig. 6) this description is wholly accurate. But GSC No. 19968 (1967, pi. 28, fig. 1) shows wrinkle-layer composed of grey calcareous material. From this I suggest that the black layer discriminated by Nassichuk on the sectioned specimen (No. 19965) is not necessarily wrinkle-layer. I interpret this black layer as a bituminous film lining the interior of one chamber, an interpretation suggested by the presence on the same specimen of black bituminous coating on a septal surface, where it certainly is not wrinkle-layer. Also, GSC No. 19968 (Nassichuk 1967, pi. 28, figs. 1, 2, 8) seems to show the wrinkle-layer passing beneath the helicolateral umbilical deposit, not above it, as would be the case if the interpretation provided by Nassichuk (1967, fig. 2) were correct. Similarly the wrinkle-layer of GSC No. 19966 (1967, pi. 28, figs. 5, 6) does not demonstrably cover the umbilical deposit. The black crust above the deposit is devoid of fingerprint pattern. Clausen (1969) and House (1971) base their conclusion that wrinkle-layer lined the whole chamber essentially on data provided by Manticoceras. In order to avoid nomen- clatural confusion it should be recalled that although Manticoceras may be regarded as showing typical and representative ritzstreifen, it cannot be claimed that the markings on the lateral and ventral parts of the whorl are typical, representative wrinkle-layer. 650 PALAEONTOLOGY, VOLUME 15 As mentioned by House ( 1 97 1 , p. 26) pyritic internal moulds of Manticoceras from Budes- heim commonly show wrinkle impressions on the lateral and ventral parts of the whorl. Biidesheim specimens in the collections of the Geological Survey of Canada show these markings, and when broken clearly show that the radial wrinkled impressions extend right around the whorl, from the dorsum, across the umbilical wall, to the venter, as was shown by Sandberger (1851, pi. HI, figs. 21, 22, 24). House (1971, p. 24, pi. 1, figs. 1, 2) has described and figured the interior of the body chamber of a specimen of Mantico- ceras sinuoswn (Hall), preserved as a barytic replacement, showing the same relationship. Internal moulds illustrated by Clausen (1969, pi. 26, figs. 10, 12, 13, 14) provide further documentation. From what these workers have shown there seems to be no doubt that a wrinkled surface lined the whole Manticoceras chamber. House and Clausen interpret this surface to be the impression of wrinkle-layer. If this interpretation is correct it indi- cates that the Manticoeeras wrinkle-layer is distributed differently compared with that of Nathorstites. It would also indicate a significant difference in comparison with Tornoceras and Maenioceras, in which the lateral patterns are unlike those on the dorsum. This poses an interesting problem in that the evidence from Manticoeeras is hard to reconcile, not only with that from Triassic ammonoids, but also with that from some {Tornoceras and Maenioeeras) of Devonian age. Without suggesting that this problem has been resolved there may be reasons to question that the wrinkled surfaces of Manticoceras are impressions of wrinkle-layer. In order to establish that wrinkle-layer lines the whole chamber is it not necessary to show that a thin crust or lamina is in this position? The internal moulds certainly do not establish this and even the barytic replacement, as illustrated, does not show the relationship between the wrinkled surface on the dorsum to the outer test. Because the presence of a crust lining the whole chamber has not been clearly demonstrated, is it possible that the wrinkled surfaces of Mantico- eeras, despite the similarity in pattern, are impressions of a surface other than wrinkle- layer? Sandberger (1851, p. 301) certainly thought so. Making a distinction between a wrinkle-layer and a wrinkled surface may seem a play on words of little significance, but EXPLANATION OF PLATE 128 Specimens coated with ammonium chloride. Arrows mark position of last septum. Figs. 1 (xl), 2 (x2). Discotropites theron (Dittmar). GSC No. 28023. Pardonet Formation, near Mile Post 428, Alaska Highway, British Columbia (GSC loc. 42389). Upper Carnian, Welleri Zone. Phragmocone, mostly preserved as internal mould with some test. Beginning of outer whorl (fig. 2) has spirally ridged dorsal shield superimposed on outer test. Figs. 3, 4 (x2). Macleanioceras enode n. sp. Holotype, GSC No. 28024. Liard Formation, British Columbia. Upper Ladinian, Sutherlandi Zone. Smooth part of outer whorl is internal mould showing nature of inside of preseptal layer. Last half whorl preserves two ribbed layers, the outer shows outer surface of outer test; the inner, outer surface of preseptal layer. Figs. 5-9. Frankites sutherlandi (McLearn) 5, 6 (xl), GSC No. 28025, Liard Formation, about 20 feet below Triassic-Cretaceous contact, Liard River, 2|- miles west of Hell Gate, British Columbia (GSC loc. 42352). Upper Ladinian, Sutherlandi Zone. Body chamber, partly internal mould partly with test. Venter (Fig. 6), wholly internal mould except at aperture where small piece of test remains, shows abrupt change from zone of smooth ornament to zone of sharp ornament, at place where preseptal layer evidently ended. 7-9 (x2) GSC No. 18903, GSC loc. 42351 (see explanation, PI. 124, fig. 7). Most of phragmocone preserves test, body chamber is internal mould. Ornament of internal mould and test different except near aperture, where they correspond. Palaeontology, Vol. 15 PLATE 128 TOZER, ammonoid shell structure i TOZER: SHELL STRUCTURE OF TRIASSIC AMMONOIDS 651 it seems to have been a distinction very much in the minds of the Sandbergers and Mojsisovics in their application of the terms Runzelschicht and Ritzstreifen over a cen- tury ago. The distinction should be made to draw attention to the fact that surface texture alone may not provide sufficient evidence to establish that a layer, comparable with the wrinkle-layer on the dorsum of Tornoceras and Nathorstites, encrusted the whole interior of the Mcmticocems chamber. That caution should be exercised in drawing conclusions from surface texture alone seems to be warranted by considering the evi- dence from Nautilus, in which a similar granular surface may be developed on both the black layer and the nacreous layer. Senior (1971) interprets sections and casts of Jurassic ammonites {Staufenia siuon Bayle, Leioceros lineatum (S. Buckman)) to indicate that the ‘ventral wrinkle-layer’ is continuous with the dorsal. But what he interprets as indicating a continuous layer is described as largely unornamented except in the immediate venter area. Senior provides no illustrations of markings on the inside of the flank, or on the umbilical shoulder. Although he has shown that a wrinkled surface may line the ventral part of the ammo- noid chamber he does not seem to have produced evidence that a discrete layer lines the whole chamber. Comparison of wrinkle-layer with the Nautilus black and nacreous layers. Several authors (e.g. Sandberger 1851, p. 303, 1856; Zittel 1895, p. 389; Foord and Crick 1897, p. xx; Nassichuk 1967, p. 240) have suggested that the wrinkle-layer is comparable with the black layer of Nautilus. The evidence provided by Nathorstites seems to confirm that this comparison may have significance for it shows that the position of the wrinkle-layer in relation to the outer test, umbilical deposit and inner test is the same as that of the black layer in relation to, respectively, the porcellanous ostracum, umbilical plug, and nacreous layer. The wrinkle-layer of Nathorstites does not line the lateral and ventral parts of the chambers. This is also true of the Nautilus black layer, although a narrow black deposit does line the ventral part of the adult Nautilus (Stenzel, in Moore (ed.), 1964, p. K72). The black layer is apparently a secretion of the mantle edge, not of the mantle surface (Senior 1971, p. 111). The wrinkle-layer may be interpreted as a comparable secretion, being a thin sheet and devoid of the lamination that characterizes the nacreous, mantle-surface secreted deposits of Mollusca in general. Flouse (1971) and Senior (1971) rightly point out that the wrinkle and black layers cannot be regarded as exactly the same because the wrinkle-layer is calcareous; the black layer, in contrast, is carbonaceous, composed of ‘conchiolin’. As noted by Senior, carbonaceous material (e.g. the siphuncle tube) is not uncommonly preserved in a carbonaceous state in ammonoids. If the wrinkle-layer had been carbonaceous similar preservation would be expected. Nassichuk (1967) has described one specimen of Clistoceras giobosum Nassichuk in which the wrinkle-layer is preserved (or coated) with black carbonaceous material and has tentatively interpreted this to indicate that the original material may have been carbonaceous. But on another specimen the layer is preserved in the normal fashion, as grey calcite, so the evidence of Clistoceras is not decisive. An unquestionable difference between the wrinkle and black layers is that of texture, that of the black layer being granular, unlike the fingerprint pattern of the layer preserved on Palaeozoic and Triassic ammonoids. This was long ago recognized by Sandberger (1856, p. 184) and Mojsisovics (1873, pp. 9, 69). 652 PALAEONTOLOGY, VOLUME 15 As an alternative to comparison with the black layer, House (1971), with Suess (1870), concludes that wrinkle-layer finds a closer homologue in the nacreous layer, but when he made this comparison the new information provided by Mescalites and Nathorstites, in which smooth nacre-like material overlies wrinkle-layer, was not available. House’s comparison is based, mostly if not wholly, on the data provided by Manticoceras, inter- preted to indicate that wrinkle-layer lines the whole chamber, as does the nacreous layer, but not the black layer. House also notes that the inside of the Nautilus nacreous layer in places has a granular surface, impressions of which are not unlike some ritzstreifen. House’s interpretation of the Manticoceras wrinkled surfaces would seem to resolve the whole problem, were it not for the fact that it is hard to reconcile with the evidence derived from other ammonoids, of both Palaeozoic and Triassic age. For this reason it is questioned, with the tentative suggestion that the Manticoceras wrinkles are impressions of a surface other than wrinkle-layer. In Nautilus nacreous material is deposited on the black layer, over the porcellanous material forming the umbilical plug, and on the whole of the inside of the chamber, but nowhere does it directly overlie the outer surface of the porcellanous ostracum. This combination of characteristics is unlike that of wrinkle-layer, which at least from the evidence provided by most Palaeozoic and all Triassic ammonoids, directly overlies the outer test and lines only the dorsal portion of the chamber. For these reasons it is con- cluded that wrinkle-layer and the nacreous layer are in no way comparable. Walliser’s (1970) conclusion that wrinkle-layer corresponds to the inner prismatic layer, for the same reasons, has little to recommend it. Wrinkle-layer seems to be comparable with the black layer in position, and was probably secreted by a comparable part of the mantle, although the two are different in composition and texture. The ritzstreifen preserved on the flanks and ventral parts of the whorl, on the other hand, almost certainly represent impressions derived from the nacreous surface. On this point there seems to be universal agreement. Systematic and Phylogenetic significance of wrinkle-layer. Wrinkle-layer with fingerprint pattern seems to characterize only a limited group of ammonoid taxa. House (1971) has listed the known occurrences in Palaeozoic Ammonoidea and finds it to be present in all Palaeozoic Orders (Anarcestida, Clymeniida, Goniatitida, Prolecanitida). As shown above it also characterizes many Ceratitida, including Ophiceratidae (PI. 126, figs. 1, 2), a family close to Xenodiscidae, which provide the principal link between the Palaeo- zoic and Mesozoic ammonoids. On the other hand it would appear that wrinkle-layer with fingerprint pattern does not occur in Phylloceratida, the only group certainly known to link the Triassic and Jurassic Ammonoidea (Tozer 1971/?). Suess (1870, p. 316) and Mojsisovics (1873, p. 31), in discussing Lytoceras (which as then interpreted included all Triassic Phylloceratida) mention that no trace of wrinkle-layer had been found. Examination of well preserved examples of Leiophyllites, Ussurites, Mono- phyllites, Discophyllites and Rhacophyllites has revealed no sign of wrinkle-layer. It would appear that wrinkle-layer with fingerprint pattern characterizes only the earlier Ammonoidea — the Palaeozoic Orders and the Ceratitida — and that it is not present in Phylloceratida, Lytoceratida, and Ammonitida. It would also appear that nothing exactly like this wrinkle-layer is definitely known in other Cephalopods. TOZER: SHELL STRUCTURE OF TRIASSIC AMMONOIDS 653 APPENDIX Family trachyceratidae Haug 1894 Genus Maclearnoceras Tozer 1963 Macleanwceras enode sp. nov. Plate 128, figs. 3, 4 Material. Known only from holotype, GSC No. 28024, from loose block derived from Liard Formation, Boiler Canyon, Liard River, British Columbia (GSC loc. 68364). Dimensions. Maximum D, 39 mm. At = 36 mm: 0-30, 0-32, 0-40. Age. Lobites ellipticus Hauer and Daonellae elegans McLearn, associated in same block, indicate Upper Ladinian, Sutherlandi Zone. Diagnosis. Evolute Maclearnoceras with ribs dividing on flank at beginning of outer whorl, at ventral shoulder on last half; 15-20 ribs per half whorl at umbilical margin; about 30 at periphery. Periphery bituberculate on inner whorls, faintly sulcate at beginning of outer whorl, with continuous and un- interrupted ribs on last half whorl. Preseptal layer thick. Description and Discussion. The character of the periphery of the inner whorl is shown by the dorsum of the penultimate whorl. No closely similar ammonoids are known. Assignment to Maclearnoceras is based on tuberculate nature of inner whorls, close resemblance in suture lines between M. enode and M. niaclearni and roughly similar style of ribbing. However, M. enode is more evolute, has much more distant ribbing, and less frequent rib division compared with M. niaclearni, the only other described species. Acknowledgements. I wish to thank Professors W. M. Furnish and B. F. Glenister, of the University of Iowa, for donating the specimen of Owenites koeneni with colour markings to the Geological Survey of Canada; Professor Dr. R. Sieber, of the Geologische Bundesanstalt, Vienna, for lending material for study; Drs. W. T. Dean and W. W. Nassichuk, Geological Survey of Canada, for helpful criticism and Dr. L. R. M. Cocks, of the British Museum (Natural History) for providing information on a paper by Sandberger (1851). REFERENCES BARRANDE, J. 1865-77. Svsteme silurien du centre de la Boheme, Premiere Partie: Recherches paleonto- logiques. 2, Classes des Mollusques, Ordre de Cephalopodes. Prague. BiRKELUND, T. and HANSEN, H. J. 1968. Early Shell Growth and Structures of the Septa and the Siphuncular Tube in some Maestrichtian Ammonites. Medd. Dansk. Geol. Foren. 18, 71-78, pis. 1-4. CASEY, R. 1960-6. The Ammonoidea of the Lower Greensand. Palaeontogr. Soc. [Monogr.], 1-582, pis. 1-92. CLAUSEN, c. D. 1969. Obcrdevonische Cephalopoden aus dem Rheinischen Schiefergebirge. II. Gephuroceratidae, Beloceratidae. Palaeontographica Abt. A, 132 (4-6), 95-178. ERBEN, H. K., FLAJS, G., and siEHL, A. 1968. Ammonoids! Early Ontogeny of Ultra-microscopical Shell Structure. Nature {London), 219, 396-398. FLOWER, R. H. 1964. Nautiloid Shell Morphology. Mem. N. Mex. Bur. Mines. 13, 1-79, pis. 1-6. FOORD, A. H. and crick, g. c. 1897. Catalogue of the fossil Cephalopoda in the British Museum (N.H.). Part 3. London. FURNISH, w. M. and GLENISTER, B. F. 1971. Permian Gonioloboceratidae (Ammonoidea). Smithson. Contr. Paleobiol. 3, 301-312, pis. 1, 2. GLENISTER, B. F. 1958. Upper Devonian Ammonoids from the Manticoceras Zone, Fitzroy Basin, Western Australia. J. Paleont. 32, 58-96, pis. 5-15. GREGOiRE, c. 1962. On submicroscopic structure of the Nautilus shell. Bull. Inst. r. Sci. nat. Belg. 38, 1-71, pis. 1-24. 654 PALAEONTOLOGY, VOLUME 15 GUEX, j. 1970. Sur les moules internes des Dactylioceratides. Bull. Lab. Geol. Mineral. Geophys. et Mas. Geol. Univ. Lausanne, 182, 1-7, pis. 1, 2. HOUSE, M. R. 1965. A Study in the Tornoceratidae: the Succession of Tornoceras and related genera in the North American Devonian. Phil. Trans. R. Soc. B., 250, 79-130, pis. 5-1 1. 1971. The Goniatite Wrinkle-Layer. Smithson. Contr. Paleobiol. 3, 23-32, pis. 1-3. JORDAN, R. 1968. Zur Anatomie mesozoischer Ammoniten nach den Strukturelementen der Gehiiuse- Innenwand. Beih. geol. Jb. 77. KEYSERLING, A. 1846. WissenschoftUche Beobacht ungen anf einer Reise in das Petschora-Land im Jahre 1843. St. Petersburg. MCLEARN, E. H. 1972. Amuionoids of the Lower Cretaceous Sandstone Member of the Haida Formation, Skidegate Inlet, Queen Charlotte Islands, Western British Columbia. Bull. Can. geol. Snrv. 188. MILLER, A. K. 1947. Tertiary Nautiloids of the Americas. Mem. geol. Soc. Am. 23, 1-234, pis. 1-100. MOJSisovics, E. 1873-75. Die Mollusken-Faunen der Zlambach und Hallstatter-Schichten. Abh. geol. Reichsanst. Wien 6(1), 1-174, pis. 1-70. 1882. Die Cephalopoden der mediterranen Triasprovinz. Abh. geol. Reichsanst. Wien 10, 1-322, pis. 1-94. • 1893. Die Cephalopoden der Hallstiitter Kalke. Abh. geol. Reichsanst. Wien 6(2), 1-835, pis. 71- 200. MOORE, R. c. (ed.), 1957. Treatise on Invertebrate Paleontology. Part L, Mollusca 4. Univ. Kansas and Geol. Soc. Amer. • 1 964. Treatise on Invertebrate Paleontology. Part K, Mollusca 3. Univ. Kansas and Geol. Soc. Amer. MUTVEi, H. 1964. On the shells of Nautilus and Spirula with notes on the shell secretion in non- cephalopod mollusca. K. Svenska Vetensk. Akad. Handl., Ark. Zool. 16, 221-278, pis. 1-22. NASSiCHUK, w. w. 1967. A morphologic character new to Ammonoids portrayed by Clistoceras gen. nov. from the Pennsylvanian of Arctic Canada. J. Paleont. 41, 237-242, pi. 28. PALER AMAN, D. F. B. 1967. Mode of early shell growth in the ammonite Promicroceras marstonense Spath. Nature {London), 216, 1128-1130. SANDBERGER, G. 1851. Beobaclitungen iiber mehrere schwierigere Puncte der Organisation der Goniatiten. Jahrb. d. Nassau. Verein. f. Natiirk. 7, 292-304, pis. 2, 3. ■ 1856. Beitrag zur vergleichenden Naturgeschichte lebender und vorweltlicher polythalamer Cephalopoden. Palaeontographica, 4, 184-201, pi. 36. and SANDBERGER, F. 1850-1856. Systematische Beschreibung und Abbildiing der Versteineriingen des Rheinischensystems in Nassau. 1-136, pis. 1-18. SENIOR, J. R. 1971. Wrinkle-layer Structures in Jurassic Ammonites. Palaeontology, 14, 107-113, pis. 13-14. siLBERLiNG, N. J. 1959. Pre-Tertiary Stratigraphy and Upper Triassic Paleontology of the Union District Shoshone Mountains Nevada. Prof. Pap. U.S. Geol. Surv. 322, 1-67, pis. 1-9. and TOZER, e. t. 1968. Biostratigraphic Classification of the Marine Triassic in North America. Spec. Pap. geol. Soc. Am. 110, 1-63. SMITH, J. p. 1927. Upper Triassic Marine Invertebrate faunas of North America. Prof. Pap. U.S. Geol. Surv. 141, 1-135, pis. 1-121. suess, e. 1870. Uber Animoniten. Sber. Akad. Wiss. Wien Abt. I, 61, 305-322. TOZER, E. T. 1971a. Triassic Time and Ammonoids: Problems and Proposals. Can. J. Earth Sci. 8, 989-1031. 1971/5. One, two or three connecting links between Triassic and Jurassic Ammonoids? Nature {London), 232, 565-566. WALLiSER, o. H. 1970. Uber die Runzelschicht bei Ammonoidea. Gottinger Arb. Geol. Paldont. 5, 115-126, pis. 1-4. ziTTEL, K. 1895. Grundzuge der Palaeontologie {Palaeozoologie). Munich and Leipzig. E. T. TOZER Geological Survey of Canada 601 Booth St. Ottawa Canada, K 1 A OE8 Typescript received 28 October 1971 GYMNOSPERMOUS WOOD FROM THE KIMMERIDGIAN OF EAST SUTHERLAND AND FROM THE SANDRINGHAM SANDS OF NORFOLK by G. T. CREBER Abstract. A description is given of the types of gymnospermous wood occurring in material collected from the Kimmeridgian of the East Sutherland coast. These types are compared with those collected from the Sandringham Sands of Norfolk. All of the material is described in terms of biorecords and events', references are made to previously described species in the form of comparison records. In traditional taxonomy some of the specimens are referable to Cedroxylon Kraus and the remainder to Filyoxylon Kraus. A close similarity is found between the two collections of wood providing a useful rough correlation of the base of the Sandringham Sands with the Kimmeridgian. The present work is based on two collections of gymnospermous wood; one from the Kimmeridgian of the East Sutherland coast and the other from the base of the Lower Portlandian of Norfolk (Sandringham Sands). The Scottish material consists of large calcareous petrifications from the shore at Helmsdale. The material from Norfolk is calcified in a highly ferruginous matrix and was obtained during the excavation of a drainage channel near West Dereham. The differing natures of the materials led to the use of two distinct techniques. Cellulose peels were made from the Helmsdale material but rock sections were prepared in studying the Sandringham Sands specimens as they required excessive etching time in attempts at making peels. STRATIGRAPHY The Sutherland material was collected mainly from the shore near Helmsdale where it is abundant; specimens occur with decreasing frequency northwards and southwards from Helmsdale along most of the outcrop of the Kimmeridgian from Kintradwell to Dun Glas. These wood specimens are most frequently encountered where there are outcrops of the boulder beds ( PI. 129, fig. 1). These remarkable beds consist of ‘immense angular blocks and smaller waterworn stones of Middle Old Red Sandstone rocks embedded in a matrix of gritty, shelly limestone’ (Phemister 1960). This matrix yields plant megafossil compression remains, a fauna of a littoral marine type with typical Kimmeridgian fossils and also some ammonites. Bailey and Weir (1933) considered that the breccias or boulder beds collected at the base of a submarine fault-scarp, the littoral shells and plant debris being swept down from the upthrow side together with Old Red Sandstone material by tsunamis or tidal waves energized by earthquakes along the fault. It is only in comparatively recent years that the exact age of the Sandringham Sands of Norfolk has been worked out. Various excavations in the county have been of great assistance in providing exposures with ample fossil material. Larwood (1961) mentions the site where the Norfolk material described in this paper was found. Excavations [Palaeontology, Vol. 15, Part 4, 1972, pp. 655-661, pis. 129-131.] 656 PALAEONTOLOGY, VOLUME 15 (TL 655997) immediately south-west of Abbey and West Dereham Station, about two miles to the east-north-east of Fordham, exposed part of the Sandringham Sands. The section exposed was about 6-5 m deep. The lower beds consisted of a conglomerate passing down into very pebbly dark grey sands; these rested on T5 m of blue-grey silty sands which formed the bottom of the section. Contractors’ bore-logs for this locality show that the blue-grey silty sands, with occasional beds of harder cemented sandstone, continue below the visible base of the excavated section for at least 8 m, giving a thick- ness of about 16 m for the sands at this point. The blue-grey silty sands contained a layer of nodules ; in the latter were large fragments of wood and phosphatized casts and moulds of many bivalves, some belemnites and occasional ammonites. Of the fossils Casey (1961) said that this part of the excavation revealed a suite indicative of a Berriasian (Infra-Valanginian) fauna new to Britain. He described the occurrence of the fossils in nodular masses of hard grey-brown glauconitic sandstone with carbonized plant debris about 10 m above the base of the formation. Included were species of Hartwellia, Isocypriua and Isodonia which have their closest parallels in the Upper Jurassic. Further evidence for the age of the Sandringham Sands was provided by a section exposed by an excavation for a North Sea Gas pipeline near King’s Lynn (TF 65 15). Casey and Gallois (1968) describe a previously unrecorded sequence of ammonite faunas in a facies of glauconitic sand and phosphate nodules. The genus Subcraspedites was succeeded by forms of the group of Garniericeras tolijeuse (Nikitin), diagnostic of the Uppermost Volgian of the Northern U.S.S.R., in turn succeeded by similar ammo- nites accompanied by an undescribed genus ancestral to the basal Cretaceous (Ryaza- nian) Hectoroceras; the sequence continued with Hectoroceras and Surites. In addition, Ager (1971) has strengthened the correlation of the Sandringham Sands with the Volgian of the U.S.S.R. in studies on the brachiopod genus Roudleria. The present work therefore lends considerable support to an Upper Jurassic age for the base of the Sandringham Sands, by showing that there is a very high degree of corre- lation between the gymnospermous woods of these beds and those of the Kimmeridgian of East Sutherland. The possibility has to be faced that in both the Sandringham Sands and in the Helmsdale Kimmeridgian the wood material might have been re-worked. However, in the case of the Norfolk material one piece of wood (MGC/T) shows borings by xylophagoLis Crustacea (or mollusca) which are fresh and do not appear to have suffered the abrasion likely to have occurred in re-working. The largest pieces of wood seen at the Helmsdale locality measured 40 cm by 25 cm (PI. 129, fig. 3). Associated with EXPLANATION OF PLATE 129 Fig. 1. About eight square metres of the boulder beds on the shore at Helmsdale, polished by the sea and showing numerous clasts. Figs. 2, 6. Biorecord 1 tracheidoxyl MS. 2, Specimen MGC/T, a portion of a tangential longitudinal section with transverse sections of a number of rays, some of which are partially biseriate; X 150. 6, Specimen MGC/A, cross-field pits in a radial longitudinal section, the ray cells are filled with resin; X 1 50. Fig. 3. Two large pieces of calcified wood, drilled by Recent rock-boring molluscs, on the storm beach at Helmsdale; length of hammer 30 cm. Fig. 4. cfA 2 TRACHEIDOXYL RC. Specimen B 34/17, part of a tangential longitudinal section showing a fusiform ray with resin canal; X 150. Fig. 5. cfA I TRACHEIDOXYL MS. Specimen B 34/9, large cross-fields pits seen in a radial longitudinal section passing along a ray; X 150. Palaeontology, Vol. 15 PLATE 129 CREBER, Late Jurassic petrified wood r 1 1 CREBER: G YMNOSPERMOUS WOOD 657 the fossil corals referred to above is a large quantity of carbonized wood fragments up to 3 cm in diameter; connection between this wood and the petrified wood has not yet been demonstrated. SYSTEMATIC SECTION For description of the material use is made of the biorecord system of Flughes and Moody-Stuart (1969). A biorecord is defined as a conceptual taxon based on a specimen from a stated locality. The biorecord is not in essence different from a palaeontologic species at the stage of description by its originator, but differs in the use that can sub- sequently be made of it; virtually no literature search is involved and description priorities are not considered. The title heading and reference line for a biorecord (e.g. 1 TRACHEiDOXYL MS) consists of {o) a serial number which outside this paper would be preceded by an identifier such as author’s initials, {b) an informal (but stored) classification guide, and (c) an author’s working reference printed in italics to indicate that it is a ‘non-search’ item. The term tracheidoxyl is used to indicate that the speci- men involved is a detached portion of wood characteristically composed of tracheids with only a minor proportion of other tissues. Additional specimens are listed as comparison records; those which cannot readily be distinguished from the biorecord are designated ‘cfA’. Specimens of progressively lower grades of comparability then receive the prefix ‘cfB’ if there is one quantitative difference and ‘cfC’ in the case of further divergence. BIORECORDS JUR 28 27 GB XYL The above heading, in a form suitable for data storage, indicates that the biorecords are wood specimens from localities in the British Jurassic. The Ages/Stages are those used in the Fossil Record (Harland et al. 1967), numbered consecutively back from Recent. Thus 28 and 27 are the Oxfordian and Kimmeridgian, respectively. All specimens are lodged at the Sedgwick Museum, Cambridge. I TRACHEroOXYL MS Diagnosis. Growth rings well marked, 0-5 cm broad. Early wood tracheids cross-section 50 ^tmx 50 /j-m, reducing only slightly through the season which terminates with a few rows of very small elements. Bordered pits, 15 /xm diameter, on the radial walls of the tracheids, uniseriate arrangement, sometimes contiguous and compressed. Rays uni- seriate or partially biseriate, 2 to 20 cells in height, average 1 1 (PI. 129, fig. 2). Cross-field with a solitary, large, oblique pit. Record Specimens. MGC/A, MGC/B, MGC/Q and MGC/T from the Flood Relief Channel cut in the Sandringham Sands to the south and east of Abbey and West Dereham Station (TL/655997). Description. This biorecord is readily distinguished by its highly characteristic cross- field pitting (PI. 129, figs. 5 and 6), the ‘eiporen’ of the German authors. Comparison Records. cfA 1 tracheidoxyl MS\ (1) MetacedroxyJon scoticum described by Holden (1915) from the Kimmeridgian of Loth, East Sutherland; Sedgwick Museum Specimen No. K613. (2) Specimens B34/9 and B34/20 from Helmsdale, East Suther- land. 658 PALAEONTOLOGY, VOLUME 15 2 TRACHEIDOXYL RC Diagnosis. Growth rings very variable in breadth, 1-3 mm. Early wood tracheid cross- section 40 f(.mx40 jj.m. Marked zone of late wood, 12 rows of smaller dense elements. Bordered pits on tracheid radial walls 10-12 /xm in diameter, bars of Sanio. Pits generally iiniseriate but smaller and biseriate at ends of tracheids. Vertical (PI. 130, fig. 1) and horizontal resin canals. Rays iiniseriate or fusiform with resin canal, 2-20 cells in height, average 8. Cross-field with one to two small bordered pits. Record Specimen. MGC/R from the Sandringham Sands (TL/655997). Description. This biorecord is of considerable interest in that it possesses rays with fusiform cross-sections (PI. 129, fig. 4) and bars of Sanio (PI. 130, fig. 2) on the radial walls of the tracheids. It would appear that there is no other undoubted record of such a wood as early as the Kimmeridgian. Comparison Records. cfA 2 tracheidoxyl RC: Specimen B34/17 from Helmsdale, East Sutherland and 3 tracheidoxyl RD (Specimen MGC/0) from the Sandringham Sands (TL/655997). cfB 2 tracheidoxyl RC: Piceoxylon scleromeduUosum, described by Shimakura (1937) from the Senonian of Sakhalin has many similar features to 2 tra- cheidoxyl RC; a much closer comparison could be made if the preservation of 2 tra- cheidoxyl RC would allow the certain determination of the existence of ray tracheids (there is some evidence of their presence but insufficient to include as an item in the diagnosis). 3 TRACHEIDOXYL RD Diagnosis. Growth rings uniform, 0-5 cm broad. Early wood tracheids cross-section 60 |um (Rad) X 50 jum (Tan). Gradual transition to late wood through the ring. Bordered pits on tracheid radial walls 17 pm in diameter. Bars of Sanio. Pits generally iiniseriate but occasionally biseriate at ends of tracheids. Vertical (PI. 130, figs. 3 and 4) and hori- zontal resin canals. Rays iiniseriate or fusiform with resin canal, 2-17 cells in height, average 9 (PI. 130, fig. 5). Cross-field with one to two small bordered pits. Record Specimen. MGC/O from the Sandringham Sands (TL/655997). Distinction. This is very similar to biorecord 2 above. In order to avoid conflicting items in the diagnosis they are being maintained separately for the time being. It is fully realized that the differences may only be such as one might expect between pieces of wood from the inner and outer parts of the trunk or between branch and trunk. Collection of further material may elucidate the matter. EXPLANATION OF PLATE 130 Figs. 1, 2. cfA 2 TRACHEIDOXYL RC. 1, Specimen B 34/17, a vertical resin canal seen in transverse section; X 150. 2, Specimen B 34/17, a portion of a radial longitudinal section showing bordered pits with bars of Sanio; X 150. Figs. 3-5. Biorecord 3 tracheidoxyl RD. Specimen MGC/O. 3, 4, vertical resin canals seen in transverse section; x 150. 5, part of a tangential longitudinal section showing a fusiform ray with resin canal; X 150. Fig. 6. cfA TRACHEIDOXYL CH. Specimen K 571, a portion of a tangential longitudinal section showing one of the deep rays; X 150. I ) Palaeontology, Vol. 15 PLATE 130 CREBER, Late Jurassic petrified wood CREBER: GYMNOSPERMOUS WOOD 659 4 TRACHEIDOXYL CH Diagnosis. Growth rings well marked, 0- 1-0-5 cm broad. Rings sometimes subdivided by bands of smaller tracheids. Bordered pits on tracheid radial walls 17 fim in diameter, generally uniseriate but not infrequently biseriate and opposite (PL 131, fig. 1). Rays uniseriate, sometimes very large (PI. 130, fig. 6), one to 34 cells high, average 16. Cross- field pitting one to four small pits, most frequently two. Record Specimens. MGC/H and MGC/U from the Sandringham Sands (TL/655997). Comparison Records. cfA 4 tracheidoxyl CH\ (1) CedroxyJon hornei described by Seward and Bancroft (1913) from Helmsdale shows clearly the characteristic cross-field pitting (PI. 131, figs. 2 and 5) and curious variability in the growth rings (PI. 131, fig. 4) which are such special features of this biorecord. (2) Sedgwick Museum Specimen No. K 571 from Helmsdale. (3) Specimens B 41 /2 and B 34/13 from Helmsdale, East Suther- land. 5 TRACHEIDOXYL GR Diagnosis. Very broad growth rings, 0-9 cm wide. Early wood tracheid cross-section 50 /xmx 50 pm. Wide zone of late wood, about 1/3 of the ring, distorted and collapsed (PI. 131, fig. 3). Bordered pits on radial tracheid walls 13 pm in diameter. Uniseriate arrangement, sometimes contiguous. Rays uniseriate 6-29 cells in height, average 15. Description. This material is not very well preserved but it is represented in the Helmsdale collection by such striking hand specimens that it demands recognition. The most noticeable feature is the width of the growth rings some of which are nearly one centi- metre wide. The most likely explanation is that the material originated from the heart- wood of a very large trunk; sections of the modern Sequoia gigantea show at the centre growth rings of the calibre of this biorecord. Such fine detail as can be observed in the specimens would indicate that they are probably the central portions of trunks of 4 TRACHEIDOXYL CH. Record Specimens. B34/1 and B41/1 from Helmsdale, East Sutherland. GENERAL REMARKS Although a number of attempts have been made to systematize the very large number of published species of fossil gymnospermous wood, exquisitely preserved material is often required for successful identification in the complicated keys involved. For example Krausel (1949) published a key separating the various woods into 25 genera, many of the 27 dichotomies requiring the presence in the specimens of minute details. Further keys follow, dividing the genera into species: in some genera there are upwards of 20 species. Many of the latter have names which imply supposed relationships with modern species; the evidence for these relationships is frequently very scanty. The aim of the present work is to depart as far as possible from the purely taxonomic approach and to attempt to use the material as a means of geological correlation. In this initial essay there would appear to have been considerable success. By using the bio- record system of Hughes and Moody-Stuart (1969) it has been possible, with material not especially well preserved, to show an essential similarity between the assemblages of gymnospermous wood from two widely separated localities. Further support has been provided for the correlation previously carried out by other workers on these localities. 660 PALAEONTOLOGY, VOLUME 15 However, if this technique is to be applied to a number of localities a certain measure of taxonomy may be desirable in order to appreciate the relative proportions of broadly different types of wood occurring in each. The following is proposed : A. Wood with resin ducts normally present. This would comprise the species pre- viously described under the generic names Pityoxylon, Piceoxylon, Pinuxylon, and further subdivisions of these as proposed by subsequent authors. The modern genera that fall unto this group are: Pinus, Larix, Picea and Pseiidotsuga. B. Wood with resin ducts only of traumatic origin, pits on the radial walls of the tracheids separate and circular and, if in two or more rows, opposite. Contiguous and more or less flattened pits may occur but never as a general rule. This would comprise the species previously described under the generic names Cedroxylon, Cupressinoxylon, Mesembrioxylon, and the various later sub-divisions of these genera. Modern genera in this category are: Cedrus, Abies, Juniperus and Sequoia. C. Wood in which the pits on the radial walls of the tracheids are normally contiguous with their mutual boundaries flattened to form a polygonal pattern. These pits are usually multiseriate and alternate. This would comprise Dadoxylon and its later subdivisions. Examples of modern genera: Agathis and Araucaria. This classiflcation implies a considerable amount of ‘lumping’ as compared to the more complicated ones extant in the literature but for present purposes it would have a num- ber of advantages. First, virtually all the available material could be classified; only the very lowest grade of preservation would fail to preserve the relatively major features necessary to place a specimen in its proper category. The relative proportions of the numbers of specimens in each group would serve as useful data in the process of corre- lating strata. A further advantage lies in the fact that it may prove possible to apply this method to wood preserved other than by petrification; that is to say, lignitic fragments and fusain. EVENTS AND TIME-CORRELATION Event: H — Helmsdale. Woods of Type A: cfA 2 tracheidoxyl RC (B 34/17). Woods of Type B: Biorecord 5 tracheidoxyl GR (B 34/1 and B 41/1). cfA 1 TRACHEIDOXYL MS (B 34/9, B 34/20 and K613). cfA 4 TRACHEIDOXYL CH (B 41/2, B 34/13 and K571). Woods of Type C: None. Woods indeterminate: 16 specimens. EXPLANATION OF PLATE 131 Figs. 1, 2, 4. Biorecord 4 tracheidoxyl CH. 1, Specimen MGC/U, biseriate pitting on the radial walls of a group of tracheids; X 150. 2, Specimen MGC/U, cross-field pitting seen in a radial longi- tudinal section passing through a ray; X 150. 4, Specimen MGC/H, a transverse section showing three ‘false rings’ produced in one season; the section shows the effect of drilling by xylophagous animals; x90. Fig. 3. Biorecord 5 tracheidoxyl GR. Specimen 41/1, a transverse section showing a sector of one complete growth ring with the late wood collapsed and distorted; x45. Fig. 5. cfA 4 tracheidoxyl CH. Specimen K 571, cross-field pitting seen in a radial longitudinal section; xl50. Palaeontology, Vol. 15 PLATE 131 CREBER, Late Jurassic petrified wood CREBER; G YMNOSPERMOUS WOOD 661 Event: N — Norfolk. Woods of Type A: Biorecord 2 tracheidoxyl RC (MGC/R). Biorecord 3 tracheidoxyl RD (MGC/0). Woods of Type B: Biorecord 1 tracheidoxyl MS (MGC/A, MGC/B, MGC/Q and MGC/T). Biorecord 4 tracheidoxyl CH (MGC/H and MGC/U). Woods of Type C: None. Woods indeterminate: 5 specimens. In both events woods of Type A are rare, especially at Helmsdale where only one out of 25 specimens is of this type. Of the 16 indeterminate specimens the preservation was at least sufficient to show that they were not of Type A. The proportion in the Norfolk event seems to be slightly larger with two out of 13 specimens. It will be seen that these two wood-events are very similar but that there is no evidence to indicate which is earlier in time. However, the external evidence of time-correlation suggests that the Norfolk locality is Volgian whilst the Scottish one is probably Lower- Middle Kimmeridgian. The situation as regards the latter is much confused by consider- able folding and faulting but the northernmost part of the outcrop (i.e. the Helmsdale area) is usually regarded as containing the higher beds. Acknowledgements. I wish to thank Mr. N. F. Hughes for considerable assistance in the preparation of the manuscript. 1 am very grateful to the Governing Body of Emmanuel College, Cambridge, for the award of a Schoolmaster Fellow-Commonership which made possible my work at the Sedgwick Museum. I am also indebted to the Royal Society’s Committee for Research in Schools for a grant to aid this work. REFERENCES ACER, D. V. 1971. The Brachiopods of the Erratic Blocks of Spilsby Sandstone in Norfolk and Suffolk. Proc. Geol. Ass. 82, 393—401. BAILEY, E. B. and WEIR, J. 1933. Submarine Faulting in Kimmeridgian Times: East Sutherland. Trans. Roy. Soc. Edin. 57, 429. CASEY, R. 19616. The Geological Age of the Sandringham Sands. Nature, London, 190, 1100. 1967. in Annual Report (for 1966). Inst, of Geol. Sciences, London, 90-91. and GALLOis, r. w. 1968. in Annual Report (for 1967). Inst, of Geol. Sciences, London, 104-105. HARLAND, w. B. et al. (Eds.). 1967. The Fossil Record. London (Geological Society), xii + 828 pp. HOLDEN, R. 1915. A Jurassic Wood from Scotland. New Phytologist, 14, 205-209. HUGHES, N. F. and moody-stuart, j. c. 1969. A method of stratigraphic correlation using early Creta- ceous spores. Palaeontology, 12, 84-111. KRAUSEL, R. 1949. Die fossilen Koniferenholzer; Teil II. Palaeontographica B 89, 83-203. LARWOOD, G. p. and FUNNELL, B. M. (Eds.). 1961. The Geology of Norfolk. Trans. Norfolk and Norwich Nat. Hist. Soc. 19, Pt. 6, 270-375. PHEMiSTER, J. 1960. Scotland: The Northern Highlands. Inst, of Geol. Sciences, London, 89-90. SEWARD, A. c. and BANCROET, N. 1913. Jurassic plants from Cromarty and Sutherland, Scotland. Trans. Roy. Soc. Edin. 48, 867. SHiMAKURA, M. 1937. Studies on fossil woods from Japan and adjacent lands. Sci. Rep. Tohoku Imp. University, 2 Geol., 19, No. 1, 28. G. T. CREBER University College School Hampstead London NYMPHS OF PALAEODICTYOPTERA (INSECTA) FROM THE WESTPHALIAN OF ENGLAND by ROBIN J. WOOTTON Abstract. The remains of three nymphal Palaeodictyoptera from the English Coal Measures (Upper Carboni- ferous) are described: Rochdalia parkeri Woodward 1913; Idoptiliis onisciformis n. gen., n. sp.; and a wing pad and prothoracic fragment situated in the pelvis of the holotype of the amphibian Eugyrinus wildi. The two named nymphs are onisciform, with prothoracic and true abdominal paranotal lobes, and posterolaterally directed wing pads; they show no aquatic adaptions; and the relief of their pronota is complex, and comparable with that of the meso- and metanota. The fossils are not assigned to families; but Idoptiliis is compared with Eugereo- nidae and Breyeriidae, and Rochdalia and the third nymph with Dictyoneuridae. The prothoracic paranota and wing pads may have been capable of movement. In the course of work for the Treatise on Invertebrate Palaeontology, Dr. W. D. I. Rolfe discovered that Rochdalia parkeri Woodward 1913, from the Lower Coal Mea- sures of Lancashire, previously supposed to be a branchiopod crustacean, was really an insect nymph. In 1967 he published illustrations and a brief account of Rochdalia, and of a similar nymph from the Middle Coal Measures of Barnsley, Yorkshire (British Museum (Natural History) In 44654), suggesting that both might belong to the extinct Order Palaeodictyoptera. He then most courteously invited me to undertake their formal description. Dr. Rolfe’s discovery was of particular interest, as in 1967 no authentic juvenile stages of the extinct palaeopterous insect orders had been properly described as such. At about that time, however, several relevant fossils were in process of study: nymphal Megase- coptera and a palaeodictyopterous nymph from the Upper Carboniferous of Illinois; mayfly nymphs from the Lower Permian of Czechoslovakia and Kansas; and nymphal and adult Palaeodictyoptera from the Upper Carboniferous of Siberia. Moreover Dr. Kukalova was revising the Stephanian Palaeodictyoptera of Commentry, France, and it was evident that this work would greatly clarify the structure and classification of the Order. It seemed sensible, therefore, to delay publishing the account of the English nymphs until some of this material was in print; the more so since doubt was expressed (Carpenter and Richardson 1968) as to whether these insects were indeed Palaeodicty- optera. The new work (Carpenter and Richardson 1968; Kukalova 1968, 1969«, b, 1970; Sharov 1971) has in fact helped to confirm the position of Rochdalia and the Barnsley nymph in the Palaeodictyoptera. Their description follows; and the opportunity is taken to describe a wing pad and part of the thorax of another nymph, also apparently palaeodictyopteran, which is located in the pelvis of the holotype of the amphibian Eugyrinus wildi (Woodward 1891) from the Lower Coal Measures of Lancashire. This latter specimen is considered too fragmentary to merit formal naming. The Barnsley nymph is named, however, as its completeness and preservation are comparable with those of Rochdalia. [Palaeontology, Vol. 15, Part 4, 1972, pp. 662-675, pi. 132.] WOOTTON: PALAEODICTYOPTERA NYMPHS 663 SYSTEMATIC DESCRIPTIONS Order palaeodictyoptera Family incertae sedis Genus rochdalia Woodward 1913 A partial synonymy was given by Rolfe (1967). The generic characters are at present indistinguishable from those of the type species. Type species. Rochdalia parkeri Woodward 1913. Rochdalia parkeri Woodward 1913 Plate 132, figs. 1-2; text-fig. 1 Holotype. L 11464, Manchester Museum: the obverse and reverse impressions of a nymph, situated in an ironstone nodule from above Arley Mine Seam, communis Zone, Lower Coal Measures, Sparth Bottoms, Rochdale, Lancashire. Description. The specimen lacks most head appendages, legs, the distal parts of the cerci, and much of the left side. Only the dorsal aspect is visible. Length, excluding cerci: 22 mm. Reconstructed width of third abdominal segment, including paranota: 10 mm. Head : relatively broad (3 • 5 mm) ; apparently preserved in hypognathous position ; with an anterior median lobe. Eyes small, prominent, widely spaced. No clypeus visible. Anteriorly, the basal parts of two appendages can be seen diverging on either side of the slight anterior lobe of the head. These show no segmentation; they may be the maxillary palps. Thorax : prothorax, mesothorax and metathorax approximately equal in length, and showing a median keel which is continued on the abdominal nota. Pronotum bearing paranotal lobes, anteriorly arising high on the notum, close to the mid-line, but con- tinuous posteriorly with the hind margin of the notum. Paranota rounded-triangular; the hind margin more or less perpendicular to the longitudinal body-axis, the fore margin running anteromesally to a rounded peak, level with the front of the head, thence posteromesally in an oblique concave arc. Three straight veins diverge from a point near the base of the paranota, about two thirds of the distance from the anterior peak to the posterior margin. The pronotum extends anteriorly beyond the insertion of the paranota as a plate, laterally emarginated, which overhes the back of the head. Behind the plate, the pronotum shows considerable relief, which seems similar in detail to that on the other thoracic nota; although on a smaller scale, being confined to the relatively narrow region between the paranota. Detailed description and interpretation of the thoracic relief must await later study, including reference to adult insects. Mesothorax and metathorax are similar, being emarginated posteriorly, and about as broad as they are long. The relief appears to show scutal and scutellar regions, together with a broad posterior band which is continuous with the anal region of the wing pads. The wing pads are directed posterolaterally, and are broadly attached to the terga, there being a wide trough in the area of junction. The mesothoracic pad has a broad C 9202 XX 664 PALAEONTOLOGY, VOLUME 15 costal area, making the outer margin of the pad nearly parallel to the body axis, and bringing it into alignment with the outer edge of the prothoracic paranotum. The costal margin of the metathoracic pad is more curved, and the costal area narrower than in the mesothoracic pad. The pads are equal in length (7 mm), and, behind Sc, in breadth (c. 3 mm). Venation is poorly preserved and seems similar, if not identical, in the two pads. Sc and Ri are long and more or less parallel. Parts of Rs and M may be seen in both pads, and CuA in the hind pad only. Abdomen; tapering towards the rear, with ten recognizable terga. The posterior margin of each is concave; the degree of concavity increasing from segments 2-9. Segment 1 is laterally obscured by the metathoracic wing pad. The remaining segments each bear a pointed posterolaterally directed paranotal lobe, slightly raised above the plane of the tergum. The lobes appear to be in homologous series with the prothoracic paranotum and the wing pads. Each lobe shows a slight ridge, with a groove behind, nearly parallel to the anterior margin, and which appears to correspond in position with the anterior margin of the lobe of the succeeding segment where it underlies its predeces- sor. On each tergum, too, a line is visible which seems to represent the position of the front edge of the tergum succeeding; suggesting that each overlapped the one behind to a marked degree. Near this line, and about two thirds of the distance from the keel to the origins of the paranota, there is visible on each side of each segment a shallow depression, whose function is not apparent. Segment 10 bears two broad cerci, of which c. 5 mm are preserved. Affinity. In our present ignorance it would have been hard to identify Rochdalia con- clusively as a palaeodictyopteran were it not for the Barnsley nymph, whose better preserved venation places it assuredly in this Order; and which Rochdalia closely resembles. Rochdalia, however, displays some characters more clearly than the Barnsley nymph, and these support its allocation to the Palaeodictyoptera. They are as follows. 1. The form of the prothoracic paranota. Prothoracic paranota which there is good reason to believe to be primary structures homologous with the wings occur in two insect groups; the Order Palaeodictyoptera; and in some members of the complex of primitive neopterous forms grouped by Carpenter (1966) in the Protorthoptera, and by Sharov (1968) in the Protoblattodea. The pronotal shield of the related Blattodea may also be homologous, in part at least, with the primary paranota. Only in Palaeodictyoptera and in the protorthopterous Lemmatophoridae are para- nota known to occur as two large separate lobes, set high on either side of the pro- notum. In Rochdalia the lobes appear to arise in this position anteriorly; but they are broadly attached, and their posterior margin is continuous with that of the pronotum. The hind edges of the wing pads, however, are also continuous with their respective segments, and it is probable that this is in both cases a condition of the nymph, and that the paranota of the adults had narrow bases, like the wings. EXPLANATION OF PLATE 132 Figs. 1 , 2. Rochdalia parkeri. Stereoscopic photographs. Specimens whitened with ammonium chloride. 1, Convex impression, X 2. 2, Concave impression, x2. Fig. 3. Idopliliis oniscifonnis, convex impression, X 1-5. Stereoscopic photograph. Specimen whitened with ammonium chloride. Palaeontology, Vol. 15 PLATE 132 WOOTTON, Palaeodictyoptera nymphs WOOTTON: PALAEODICTYOPTERA NYMPHS 665 The anterior peak of the paranota is odd, and nothing similar is known m adult Palaeodictyoptera. Paranota extending forward on either side of the head do, though. occur in this Order ; most notably, perhaps, in the Dictyoneuridae (Kukalova 1970, figs. 50, 55). The head of Rochdalia, however, appears rather broad for this family. Rather similar paranota occur on the undescribed nymph of uncertain affinity (Hunterian Museum, Glasgow University A 2680a) from the same horizon and locality as IdopliJus onisciformis, figured by Rolfe (1967, pi. 50, fig. 8; text-fig. 2d). This nymph too is onisciform; as is Euryptilodes hoiridus Sharov 1961, placed by that author in the Protoblattodea, from the L. Permian of the Kuznetsk Basin, Siberia. Both show superficial overall re- semblance to Rochdalia and Idoptilus, which may be convergent. 2. The abdominal paranota. These are known in few adult Palaeodictyoptera, but are arguably present in the Dictyoneuri- dae. Kukalova (1970) has denied that the lateral abdominal structures in Stenodictya are true paranota, homologous with those of the prothorax; and this will be discussed later. It is worth noting here that the form and relationships in Rochdalia and in Dr. Kukalova’s figures of Stenodictya appear very similar, when allowance has been made for the post-mortem distension of the Stenodictya abdomens; and further- more that both show oblique ridges run- ning to the apices of the lobes, which seem to mark the position of the underlying anterior edges of the nota following. 3. The cerci. Strong filiform cerci are widespread in Palaeodictyoptera. They also, however, occur in Ephemeroptera, TEXT-FIG. 1. Rochdalia parkeri Woodward, Man- chester Museum L. 11464, x6-6. Drawn by tracing photographs taken both by oblique light and with the specimens immersed in alcoholic glycerin. The drawing is based on the convex im- pression, but includes details, particularly of the head and wing pads, derived from the counterpart. 666 PALAEONTOLOGY, VOLUME 15 Megasecoptera, Diaphanopterodea, Archodonata and Plecoptera, and were probably present in many Protorthoptera. Further, but entirely inconclusive suggestions of dictyoneurid affinity come from the apparent homonomy of the fore and hind wings; and from the posterior emargination of the abdominal tergites, which is far more marked in adult Stenodictya abdomens than in those of other known Palaeodictyoptera. Homonomous fore and hind wings occur, however, in many palaeodictyopterous families, and the structure of the abdomen is known in few. It is therefore impossible confidently to assign Rochdalia to any family; the more so since one cannot be certain how many characters are nymphal adaptations absent from the adult. Genus idoptilus gen. nov. The generic characters are at present indistinguishable from those of the type species. Type species. Idoptilus oniscifonnis sp. nov. Derivation of name, Idas — needle; ptilon — wing. Idoptilus oniscifonnis sp. nov. Plate 132, fig. 3; text-figs. 2, 3a, 3b 1967 ‘Rochdalia-like nymph’, Rolfe, pp. 310-311, pi. 50, figs. 5-7, text-fig. 2b Holotype: In 44654, British Museum (Natural History), the dorsal aspect of a nymph in an ironstone nodule from above the Barnsley Coal, simiiis-pidchra Zone, Middle Coal Measures. Round Green open cast mine, Barnsley, Yorkshire. Description. The specimen lacks the head, appendages, anterior margin of the pro- notum, most of abdominal segments 7-9, segment 10, and much of the right side of the entire body. Some traces of the cuticle remain, particularly along the margins of the wing pads and paranota, which are fringed throughout with short spines. The surface of the insect is covered with papillae, which are much coarser on the terga than on the wings or paranota. Length of specimen 45 mm. Reconstructed width of second abdominal segment including paranota 20 mm. Thorax : Prothorax, mesothorax and metathorax approximately equal in length, and showing a median keel which is continued on the abdominal terga at least as far as segment 5. Pronotum bearing large triangular paranotal lobes marked off from the notum by a wide, crescent-shaped furrow. The anterior edge of the lobe lies at c. 65°, the pos- terior at c. 110° to the longitudinal body axis. Venation of paranotum obscure, but a single vein-like groove, nearly perpendicular to the body axis, appears to give rise at about one third of its length to two oblique branches running respectively antero- laterally and posterolaterally. The pronotum shows considerable relief which, as in Rochdalia, appears comparable to, though smaller than that on the meso- and metanota. Mesothorax and metathorax about as broad as long, and similar in size and detail. The posterior margin of the mesonotum is slightly emarginated; that of the metanotum is not preserved. The relief is marked, and as in Rochdalia appears to show scutal and scutellar regions and a broad posterior band. Both segments bear large posterolaterally WOOTTON: PALAEODICTYOPTERA NYMPHS 667 TEXT-FIG. 2. Idoptilus oiiisciformis n. gen., n. sp. British Museum (Natural History) In 44654, X 3-6. Drawn by tracing photographs taken by oblique light. 668 PALAEONTOLOGY, VOLUME 15 directed wing pads, marked off from the notum by broad grooves; that of the meta- thorax being particularly deep. The area of junction between the mesonotum and its pad is partly destroyed, and the pad appears slightly displaced from its proper position. The wing pads are equal in length, but differ in breadth and proportion. Both show the venation in some detail. Fore wing pad. Length: 15 mm; breadth at base: 8 mm, at halfway: 5 mm. Distance from Sc to hind margin at one-third the wing length from base: 4-5 mm. The anterior part of the pad-base is missing, and the pad is distorted by a transverse crease at about one-quarter its length from the base. Costal field broad and triangular, the basal third of the anterior margin lying perpendicular to the body axis, the remainder at 20°. The posterior margin of the pad is straight, at c. 50° to the body axis. Neither precostal strip nor postcostal vein is visible, although their absence cannot be assumed. Sc concave, long, unbranched, curving in the distal part of the pad to run parallel to the anterior margin almost to the wing-tip. R markedly convex, unbranched, nearly parallel to Sc throughout its length. Rs departing from R at about half the pad-length from the base, concave, faintly preserved; probably pectinately 4-branched, although it is possible that the fourth branch forks again distally. MA and MP separate just distally to the diver- gence-point of R and Rs. MA markedly convex, strongly arched, unbranched. MP con- cave, faint, five- branched; dividing initially into 3, with the first and third branches dividing again. CuA convex, strongly arched, unbranched. CuP concave, faint, branch- ing obscured. Anal veins obscure. Hind wing pad. Length: 15 mm; breadth at base: 7-5 mm, at half-way: 6 mm. Distance from Sc to hind margin at one-third the wing-length from base: 5-5 mm. Anterior margin curved, the proximal half appearing slightly arched to form a long low anterior lobe. Costal field much narrower than that of fore wing pad. Precostal strip doubtfully present: an anterior band is visible in the basal half of the wing, but its posterior edge is faint and concave, which is not in keeping with the costa. No postcostal vein visible, but its absence cannot be assumed. Sc long, unbranched, as fore wing. R markedly convex, unbranched, parallel to Sc. Rs departing from R slightly more distally than in fore wing, concave, faint, branching probably as in fore wing. Base of M very faint, apparently concave. MA markedly convex, strongly arched, unbranched. MP concave, faint, branching obscure but probably as fore wing. Cu basally convex; CuA and CuP separating at about one-third the wing-length from the base. CuA strongly convex, unbranched. CuP concave, two-branched. Anal veins pectinately 7-branched ; the 6th forking to make 8 branches in all. At the base of the pad is a line of raised areas which appear to be associated with the longitudinal veins. These may be the precursors of the axillary plates of the wing; and they correspond fairly closely with the pattern described by Kukalova (1960) in Ostrava nigra, although the sub-costal plate seems less closely associated with the radial than in that species, and the anal veins appear to arise from a broad raised area associated with the cubital plate, rather than from the plate itself. Abdomen: tapering towards the rear, with traces of nine abdominal terga; the last four being progressively incomplete and poorly preserved. The posterior margin of each segment is concave. Segment 1 is laterally obscured by the metathoracic wing pad, although a trace of the posterior margin may be seen underlapping the pad. Each remaining segment bears a pointed posterolaterally directed paranotal lobe, not, as in WOOTTON: PALAEODICTYOPTERA NYMPHS 669 A B C TEXT-HG. 3. A, B, Idoptihis onisciformis, wing pads, reversed and drawn by tracing photographs. X 5. A. Fore wing pad. B. Hind wing pad. 3 C. Nymph 3. Manchester Museum W1222, x 17. Drawn with camera lucida. 670 PALAEONTOLOGY, VOLUME 15 Rochdalia, noticeably raised above the level of the tergum. Each tergum shows a faint line, which may represent the anterior margin of the tergum succeeding. Affinity. While less of the body of IdoptiJus onisciformis is preserved than is the case with Rochdalia parkeri, the surface details of the preserved parts are far clearer. In particular the wing pad venation provides good evidence for the assignment of the insect to the Palaeodictyoptera . The venation of many polyneopterous insects resembles to some extent that of Palaeodictyoptera. Two characters, however, indicate that Idoptilus was certainly palaeopterous; and these are probably of general value in distinguishing between Palaeoptera and Polyneoptera. Firstly MA and MP are clearly convex and concave respectively, whereas in most if not all Neoptera the median veins are flat; and secondly CuP and lA of the hind wing are both strongly curved. In all Polyneoptera with fully developed hind wings these veins are straight or nearly so, being aligned with the necessarily straight vannal fold. Unless, therefore, Idoptilus belongs to a primitive and otherwise unknown neopterous order which lacks a vannus, it may be taken to belong to the Palaeoptera. Of known Palaeoptera, the venation most nearly resembles that of two groups of Palaeodictyoptera, which are believed by Kukalova (1969Z?) not to be particularly closely related the one to the other. These are the Breyeriidae, whose members show simpler venation than any of the other three families, Homoiopteridae, Lycocercidae and Graphiptilidae with which Kukalova groups them; and the assem- blage of families including the Eugereonidae, the Megaptilidae and the Archaemegapti- lidae. Were the cross-veins or the area at the base of the costa of the fore wing preserved it would be relatively easy to decide to which of these groups the nymph should be referred. In their absence it is difficult. Dr. Rolfe (1967) correctly quotes my having tentatively compared this nymph, whose photographs he had sent me, with the Breyerii- dae. I did not and do not, however, believe there to be any evidence for referring Rochdalia to this family. Idoptilus shares with all these families the following characters; 1. At least the median, cubital and anal veins strongly and evenly curved to the hind margin. 2. Sc, R, MA and CuA unbranched. 3. R separating from Rs and MA from MP at nearly the same level. 4. The anterior product of the first forking of Rs branching pectinately several times. In Idoptilus the posterior product of the first forking branches once, as in known Eugereonidae and Breyeriidae. In Archaemegaptilus kieffevi and Megaptilus blanchardi it branches pectinately 2 and 3 times respectively. The mode of branching of Rs in Dictyoneurella perfecta Laurentiaux, placed by Kukalova in the Archae- megaptilidae, is different from that of all others in the group of families. 5. The anal veins are pectinately branched. The number of branches of Rs, MA and CuP falls within the range observed in alt the mentioned families, except that no breyeriid is known with more than 4 branches to MA. Idoptilus further resembles Breyeriidae in having the hind wing broader than, but similar in length and venation to, the fore wing; and in possessing a marginal fringe of WOOTTON: PALAEODICTYOPTERA NYMPHS 671 macrotrichia (which may, however, be only a juvenile character in Idoptiliis). It differs in having Sc long and distinct from R throughout its length, whereas in Breyeriidae it is relatively short, and ends on R. The venation of the fore and hind wing pads of Idoptilus most closely resembles that of the fore wing of Dictyoptihis (F. Eugereonidae). The hind wings of this family, how- ever, differ from the fore wings in being shorter, and in the form of the cubital and anal veins (Carpenter 1964, Kukalova 1969h). The fore wings of Megaptilidae and Archae- megaptilidae were probably similar to the hind wings, but the venation of the latter is less like that of Idoptilus; and the same is true of the related families Protagrionidae and Calvertiellidae. The only other family of Palaeodictyoptera whose wing venation is comparably simple is the Dictyoneuridae. In this group, however, MP never has more than 3 branches ; and the fore and hind wings are similar in breadth. The precise position of Idoptilus must therefore remain in doubt. The proportions of the wings and spacing of the veins recall the fore wings of Eugereonidae more than Breyeriidae, and it may well be that the insect is closest to the former family, although excluded from it by the form of the hind wing. It should be noted, though, that the pronotum and paranota of Eugereon boeckingi Dohrn do not particularly resemble those of Idoptilus. Nymph 3 Text-fig. 3c Description. The specimen consists of a nearly complete mesothoracic wing pad, the tip of another pad, and part of the notum of a thoracic segment; all situated between the bones of the pelvic girdle of the holotype of Eugyrinus wildi (Woodward) W 1222, Manchester Museum, from the Soapstone bed. Mountain Four-feet Mine, Carre Heys, Trawden, near Colne, Lancashire (L. Coal Measures, lenisulcata zone). The pads are preserved as a concave impression; but the notum is convex, and its sagittal plane is nearly perpendicular to the plane of the wing pads. The relative position of the wing pads and notum clearly does not reflect that in life. The relief of the notum resembles the pronota rather than the mesonota of Rochdalia and Idoptilus, and it is probably the pronotum of the insect. The more fragmentary wing pad is poorly preserved, and its venation appears similar to the more complete one. The latter is 6 mm long and c. 2-5 mm wide; slightly smaller than those of Rochdalia. Its surface is coarsely punctate, especially in the costal area. The latter is basally very broad, and tapers to the wing tip; the costal margin being strongly arched in the basal third of the wing, and nearly straight for the remaining two thirds. This condition clearly recalls the mesonotal pads of Rochdalia and Idoptilus. Sc is unbranched, long, and evenly curved to the wing tip. R is unbranched, deeply con- cave (in this impression) and runs almost parallel with Sc. Rs departs from R near the mid point of the wing, and is pectinately 3-branched. MA departs from MP level with the point of separation of R and Rs, and is unbranched, evenly curved, and deeply concave. MP is pectinately 3-branched. CuA is unbranched and concave. CuP appears unbranched, but may be incomplete. It reaches the hind margin at about 2/3 the wing length from the base, indicating that the anal area was long. No anal veins are preserved. The left half of the pronotum is preserved, and shows strong relief, which is com- parable with that of Rochdalia and Idoptilus, particularly the former. 672 PALAEONTOLOGY, VOLUME 15 Affinity. The venation of the wing pad is remarkably simple. In possessing a long Sc; R, MA and CuA unbranched and evenly curved; Rs and MP with few branches, pectinately arranged; and a long anal area, it recalls the Dictyoneuridae and to a lesser extent the Breyeriidae. All known members of these families, however, have more than three branches to Rs; and indeed, unless further branches were added at subsequent moults, this insect had one of the simplest venation patterns of all Palaeodictyoptera. Comparable simplicity occurs in Carboniferous Megasecoptera and in Carboniferous and Permian Ephemeroptera (cf. the wing pads of Protereisma described by Kukalova (1968) from the L. Permian of Oklahoma and Moravia); but the high costal lobe is typical of neither order, and is, as we have seen, associated in Rochdalia and Idoptilus with the presence of pronotal paranota, continuing the line of these down the side of the thorax. The evidence supports the assumption that this nymph, like the others, belongs to the Palaeodictyoptera. As such it provides one of the earliest records of the Order. It is uncertain whether the situation of the fragment in the pelvis of Eugyrinus is due to the amphibian having eaten a living or dead nymph, or to chance superimposition of the remains. Watson (1941) suggests that the marine origin of the Soapstone bed may indicate that the insect had been eaten. DISCUSSION The three English palaeodictyopterous nymphs are of interest for the following reasons: 1 . While not demonstrably terrestrial, they show no aquatic adaptations. 2. They confirm that the wing pads of juvenile Palaeodictyoptera were postero- laterally directed as in Megasecoptera (Carpenter and Richardson 1968) early Ephemeroptera (Kukalova 1968) and many Neoptera; but unlike those of extant Palaeoptera. This is also apparent from other material, undescribed at the time of writing but mentioned by Carpenter and Richardson (1968, p. 308) and by Sharov (1971). 3. They show that some families, at least, had onisciform nymphs, whose pronotal and abdominal paranota formed with the developing wing pads a series of lateral plates, complete from the head, which they partly shielded, to the end of the abdomen. 4. They show remarkable development of the pronotum, with relief which recalls that of the meso- and metathorax. 5. Their wing pads show precocious development of some adult structures — notably venation {Idoptilus and nymph 3) and axillary plates {Idoptilus). Moreover both the pads and the pronotal paranota are marked off from the nota by deep furrows, which may have been lines of flexure, and which have also been observed in palaeodictyopterous nymphs from Siberia (Sharov 1971). The significance of these features requires further investigation, particularly of the thoracic structure of adult forms; but it is beginning to seem possible that these nymphs were capable of moving both wing pads and pronotal paranota, perhaps in association with parachuting and limited gliding (see also Sharov 1971). Smart (1971) has suggested 673 WOOTTON: PALAEODICTYOPTERA NYMPHS that the ancestors of Pterygota may have fed on the fruiting bodies of early land plants, but the presence of a muscular clypeal region indicates a liquid feeding habit in Palaeo- dictyoptera (Carpenter 1971). In either case the selective advantage associated with the development, enlargement and refinement of paranotal lobes could have been increased control of the manner of falling from the food plants, so that the insect tended to reach the ground the right way up. Gliding and direction control would, one imagines, follow. B TEXT-FIG. 4. Diagrammatic representations, from a posterodorsal viewpoint of two pairs of abdominal segments, showing alternative interpretations of the same dorsal view. In A the lateral lobes are the acute posterior angles of the segments ; in B they are true paranotal plates. It may well be that this way of life continued in the nymphs of some early fully-winged insects, and provided reasons for the retention of the pronotal and abdominal paranota. Many adult Palaeodictyoptera in fact lack abdominal paranota. They have usually been supposed to be present in Dictyoneuridae, but Kukalova (1970) believes that the apparent lateral abdominal expansions in this family are not true paranota but just the strengthened and rather acute posterior angles of the terga. From a dorsal impression it is not easy to be certain whether a segment fits wholly into the one preceding (text-fig. 4A), and occupies almost all the width of the notum, or whether the notum is extended laterally in paranotal plates (text-fig. 4B). Kukalova (1970) believes the former to be the situation in Stenodictya. In Idoptilus, and certainly in Rochdalia, the latter interpretation is correct, and there can be no doubt that true paranotal lobes are present. The detailed resemblance between the abdominal segments 674 PALAEONTOLOGY, VOLUME 15 of Rochdalia and those in Kukalova’s figures of Stenodictya has already been indicated and, while I have not seen the specimens, it seems to me possible that Stenodictya too may after all have true paranota. It should be noted, though, that the abdomens of Rochdalia and Idoptilus are relatively far broader than those of Stenodictya species. While it cannot be proved that the abdominal paranota of Rochdalia and Idoptilus are not a nymphal adaptation, it seems more probable that they are a primitive character. If so, and if they are indeed present in Stenodictya, they are relevant to Kukalova’s conclusions (1970) on the position of the Dictyoneuridae within the Palaeodictyoptera. She has concluded that, far from being the most primitive Family in the Order, they are a relatively advanced group, with secondarily simplified venation ; and whose only out- standingly primitive character is a uniform archedictyon. The presence of abdominal paranota would be another such character, although it would in no way contradict Kukalova’s view of the position of the Family. It is interesting, however, that nymph 3 comes from the Lower Westphalian A, and demonstrates that some lines with extremely simple venation were already established at this time, alongside the Spilapteridae with their far more primitive venation. Acknowledgements. I am indebted to Dr. W. D. I. Rolfe for inviting me to describe Rochdalia parkeri and Idoptilus onisciformis, and to Professor F. M. Carpenter, who first drew my attention to the nymph in the pelvis of Eiigyrinus. Dr. R. M. C. Eagar of the Manchester Museum, and Dr. H. W. Ball and Mr. S. F. Morris of the British Museum (National History) have kindly allowed me to examine and borrow specimens for description. My thanks are also due to Professor F. M. Carpenter, Dr. Jarmila Kukalova and Dr. A. G. Sharov for their comments on the material. REFERENCES CARPENTER, F. M. 1964. Studics Oil Carbonifcrous Insects of Commentry, Erance: Part VI. The Genus Dictyoptihis (Palaeodictyoptera). Psyche, Cainb. 71, 104-116. 1966. The Lower Permian Insects of Kansas, Part 1 1, The Orders Protorthoptera and Orthoptera. Psyche, Camb. 73, 46-88. ■ 1971. Adaptations among Palaeozoic insects. Proc. N. Anier. paleont. Convention, 2, Part I, 1236-1251. CARPENTER, F. M. and RICHARDSON, E. s., jR. 1968. Mcgasecopterous Nymphs in Pennsylvanian Con- cretions from Illinois. Psyche, Camb. 75, 295-309. KUKALOVA, j. 1960. New Palaeodictyoptera (Insecta) of the Carboniferous and Permian of Czecho- slovakia. Sb. ustred. TJst. geol. 25, 239-251. 1968. Permian Mayfly Nymphs. Psyche, Camb. 75, 310-327. 1969u. Revisional Study of the Order Palaeodictyoptera in the Upper Carboniferous Shales of Commentry, Prance, Part I. Psyche, Camb. 76, 163-215. 19696. Revisional Study of the Order Palaeodictyoptera in the Upper Carboniferous Shales of Commentry, France, Part II. Psyche, Camb. 76, 438-486. 1970. Revisional Study of the Order Palaeodictyoptera in the Upper Carboniferous Shales of Commentry, France, Part III. Psyche, Camb. 77, 1-44. ROLFE, w. D. I. 1967. Rochdalia, a Carboniferous insect nymph. Palaeontology, 10, 307-313. SHAROV, A. G. 1961. Order Protoblattodea in rohdendorf, b. b. et al. Palaeozoic insects of the Kuznetsk Basin. Trudy Paleont. Inst. 85, 157-164. In Russian. 1968. The phylogeny of Orthopteroid insects. Trudy Paleont. Inst. 118, 1-243. In Russian. • 1971. Habitat and relationships of Palaeodictyoptera. Proc. Xlllth Internat. Congr. Ent., Moscow, 1968, 1, 300-301. SMART, J. 1971. Palaeoecological factors affecting the origin of winged insects. Proc. Xlllth Internat. Congr. Ent., Moscow, 1968, 1, 304-306. WOOTTON: PALAEODICTYOPTERA NYMPHS 675 WATSON, D. M. s. 1941. The origin of frogs. Tram. Roy. Soc. Edinburgh, 60, 195-231. WOODWARD, H. 1891. On a Microsaurian (Hylonomiis ivildi, sp. nov.) from the Lancashire Coal- field. Geol. Mag. 28, 211-213. — — 1913. Roclidalia parkeri, a new branchiopod Crustacean from the Middle Coal Measures of Sparth, Rochdale. Geol. Mag. 50, 352-356. Manuscript received 19 November 1971 ROBIN J. WOOTTON Biological Sciences Department Exeter University C 9202 Yy PALAEONTOLOGICAL ASSOCIATION Notes for Authors submitting papers for publication in PALAEONTOLOGY Scope of journal. Papers will be published on any aspect of palaeontology, including palaeoecology and stratigraphical palaeontology. Papers on Recent material may be acceptable if their palaeontological purpose is explicit. Review papers are particularly welcome. A high standard of illustration is a feature of the journal. Submission. Two complete copies of the typescript (including all explanations of illus- trations) should be submitted; photocopies of all text-figures (correctly reduced) and of all plates should accompany each typescript. The original figures (if they are small) may if necessary serve as illustrations for one copy. Although failure to provide two copies will usually result in slower handling, no paper will be rejected for this reason. Submissions should be made to the Secretary of the Publications Committee at the address given on the inside cover of the current issue of ‘Palaeontology’. PUBLICATION POLICY OF THE PALAEONTOLOGICAL ASSOCIATION ^Palaeontology'' provides full opportunity for illustration, but to compensate for the cost of this, special attention to brevity is required in the composition of the text. The style should be simple and care should be taken to avoid long complicated sentences so that papers may be read easily by those unaccustomed to the English language. Classification of papers. Papers will be classified and handled as follows: (a) Normal papers — less than 25 printed pages of text. Present target for publication is not more than 12 months from submission; this time does not include any delays due to major referee criticism, {b) Short communications- — these can often be published more quickly than normal length papers. (c) Long papers — authors are advised to consult the Editors before submission. See also notes on ‘Special Papers in Palaeontology’, below. This statement is intended only as an explanation of intention, and should not be taken as any kind of guarantee; the ultimate control is the available finance of the Association. Editorial work, correspondence, and delay will be reduced to a desirable minimum if authors will follow these notes carefully. THE TEXT OF THE PAPER Style and arrangement. Authors should consult published volumes of ‘Palaeontology’ and construct their papers in accordanee with the practices used there, particularly in [Palaeontology, Vol. 15, Part 4, 1972, pp. 676-681.] NOTES FOR AUTHORS 677 the arrangement of headings and in the explanations of plates and of text-figures. Numerals are printed in arabic figures. Typing. Papers should be typed on good quality paper, preferably of International A4 size (297x210 mm). All pages of typing should be numbered consecutively. Double spacing of lines (not 1 is required throughout (including references and explanations of plates and text-figures), and there should be a wide margin. Headings should not be underlined, and nothing should be underlined in running text without good reason ; generic and specific names should be underlined for italic type-setting. Lower-case letters are used in English words, except for proper names. In the text, references should be cited by the author’s name, followed by the date in brackets and by page references where necessary. Examples : Cox (1963); Cox (1963, p. 20); (Cox 1963). The Title should be short and should include fossil group, age, and locality. An Abstract of not more than 200 words is required at the beginning of all papers. It should summarize results (rather than contents) of the paper and should mention new systematic names erected ; it should be carefully prepared so that it will serve satis- factorily for international use in abstracting journals. Footnotes. These will only be allowed in exceptional circumstances. Permission to publish, for instance, should be included with other acknowledgements. References should be arranged in alphabetical order of authors’ names at the end of the paper. The author’s name should be followed by the year of publication, and the title of the paper in full. The name of the journal (which will be printed in italics) should be abbreviated in the style of the 4th edition of the ‘ World List of Scientific Periodicals', London, 1963-5. New titles and abbreviations are listed in 'British Union-catalogue of Periodicals', Butterworths, London (quarterly), 1964 onwards. Volume number, part or fascicule number in brackets (only if really necessary), pagination, and number of plates, should be given in arabic figures with the items separated by commas only. For books the title is underlined for printing in italics, and the place (town of publication) should be given. When a reference has been translated or transliterated, the language of origin should be stated in square brackets at the end. Examples: FISCHER, w. 1961. tJber die Lias/Dogger-Grenze in Siiddeutschland. Neiies Jb. Geol. Paldont. Mh. 8, 394-400. LYDEKKER, R. 1885. Catalogue of Fossil Mammalia in the British Museum. Part 1. London, xxx+268 pp., 33 figs. WEBER, V. N. 1951. Upper Silurian trilobites of the U.S.S.R. Trudy naucimo-issled. geol. Inst. 2, 1-55, Pis. 1-7 [In Russian.] Synonynnes should be in a commonly accepted form; the recommended form is: 1947 Beltanella gilesi Sprigg, p. 218, pi. 6, fig. 1. 1949 Beltanella gilesi Sprigg; Sprigg, p. 81, pi. 10, fig. 1. Authors may use symbols to indicate the degree of confidence with which particular items in the list may be referred to the species under discussion. Such symbols are 678 PALAEONTOLOGY, VOLUME 15 given on p. 53 in richter, r. 1948. Emfiilmmg in die zoologische Nomenklatur, 252 pp. Kramer Verlag, Frankfurt am Main, 1948 (2nd edition). e.g. vl937 = the author has seen the cited material and agrees with the reference. v*1937 = the author has seen the type of the species. Units and symbols. As far as possible the recommendations contained in Quantities, Units and Symbols (1971, The Royal Society) should be followed; in particular the International System of Units (SI) should be used whenever it is practicable to do so (e.g. /xm, etc.). LINE ILLUSTRATIONS Text-figures. Whenever possible original drawings, in black on good quality white card such as ‘Bristol board’, should be submitted. They should preferably be made twice the size finally required, and when reduced must not exceed the type area of a page, 190 x 140 mm (7|x5| inches). If the caption (explanation) of a full-page figure is long, allowance for its inclusion on the same page should be made by reducing the height of the figure. In composing smaller text-figures, space on the page is best used if the figure is wide rather than high, and the full width of the page (140 mm) should be used if possible. All lettering should be inserted by the author, and must be readable when reduced. Authors using materials other than white card should assess the risk of distor- tion. Good quality photocopies (unglazed on photostat paper) may be suitable for reproduction and are preferable to excessively large originals; originals however may have to be called for by the editor. Tables are expensive to print especially if vertical ruling is required. As far as possible all tables of more than two or three lines should be prepared, by careful stencilling or by use of a varitype machine, for line reproduction. Two copies at correctly reduced size must be supplied. Fold-out tables are not acceptable because of prohibitive cost. If a table of larger than page size is essential it should be arranged for two facing pages. Explanations of text-figures, tables, and plates should be typed (double-spaced) and placed at the end of the typescript. Position in text. The approximate position desired for insertion of text-figures and tables should be indicated on the typescript. PLATES Photographic illustrations will be reproduced by offset-lithography using 300 screen. Making-up. Originals should be submitted with the photographs printed on commercial bromide glossy paper, glazed and mounted on strong card which should be kept flat and clean. It is important that all photographs on any one plate should be of similar tone. Negatives are not acceptable. The photographer will reproduce what he is given and prints should therefore be made as they are intended to appear. Prints mounted on a tough thin card survive transit by mail better tlian on thick card; the original plates should be protected by thin paper overlays. The protective wrapping round the whole package should be thick. Plates should be assembled after careful study of the methods and detailed notes given below; no plate-space should be wasted. If there is any doubt about method. NOTES FOR AUTHORS 679 a mock-up should be sent to the Editors for comment before submission. Figures should always be arranged as close together as they will reasonably go. Lines should not be drawn around the plate, or between figures to indicate groupings. Size. The size of full-page illustrations is 203 X 146 mm (8x5f inches), and these are referred to as Plates. Where authors have photographs which fill less than a full page such illustrations will be referred to as text-photographs and treated as text-figures. They will be reproduced by the same screen process and the photographs should be prepared in the same way as for plates but to text-figure dimensions (see above). Every effort should be made to ensure that no page-space is wasted whichever size is submitted, particularly in respect to the width in the case of text-photographs. Originals should be submitted as x 1 ; there is normally no advantage in reduction and the Editors should be consulted before submission of originals which exceed X 1|. Lighting. Where possible, the convention of lighting the fossils from the top left should be followed. Method of mounting prints. One of the following four methods should be followed. 1. Aligned rectangular prints. Mount accurately trimmed rectangular prints close together with edges parallel and carefully aligned. Prints should be cut very accurately to ensure that they are true rectangles. The plate-maker will {a) only block out the intervening narrow strips on his negative so that they appear white on the plate, and {b) insert figure numbers in type either below the prints in these strips (leave sufficient space), or in a small cut-out (normally in the bottom right-hand corner). This method is best for rectangular photomicrographs, and in cases where further print trimming is undesirable. It avoids the possibility of misunderstanding in blocking- out. Examples : Vol. 14, Part 4 (1971), PI. 136. Vol. 15, Part 1 (1972), PI. 28 2. Fossils on cleared background. Mount the prints (any shape) reasonably close together to avoid waste of space, but do not trim prints too close to nor parallel to the edges of the fossils; do not draw a thin black line round the trimmed outline. The plate-maker will block out to the actual edges of all the fossils or rock matrix, and insert figure numbers m type. The plate background will thus be paper white. This method is suitable for many kinds of macro-fossils, but some responsibility for outline is left to the plate-maker. Examples ; Vol. 14, Part 3 (1971), PI. 99 and Part 4, PI. 135 3. Black backgrounds. This method should be avoided if possible, but if it is used it is important that the background is uniformly black. Edges of the fossil illustrated should be made distinct. White figure numbers and lettering should be stencilled on the black background. Example : Vol. 13, Part 2 (1970), PI. 52 4. Untouched clean card background. Mount the prints on very white card kept absolutely clean (no pencil marks, traces of adhesive, etc.). Prints should be trimmed 680 PALAEONTOLOGY, VOLUME 15 very carefully up to the actual edges of the fossils, or as required. No blocking-out will be attempted, and the original white background will reproduce as a faint grey; figure numbers will be overprinted as in methods 1 and 2. This method should be used where there is any labelling with directing lines and lettering around the fossil; setting labelhng correctly in type is time-consuming and expensive. Examples: Vol. 6, Part 1 (1963), PL 22; Vol. 7, Part 4 (1964), PI. 87': Combination of methods on one plate may be acceptable, if suitably arranged. Examples: Vol. 9, Part 2 (1966), PI. 32; Vol. 10, Part 1 (1967), PI. 10, and Part 4, PI. 105- Figure numbers will be overprinted on the plates (except in Method 3). They should be inserted clearly, in the final positions (not opposite the centres of the figures) on the protective overlay, and if desired (in methods 1 and 2 only) lightly in pencil on the correct position on the original plate. Magnification. The magnification of each figure should be stated in the plate explanation. Linear scales are difficult to set in type in the blocked-out areas in Methods 1 and 2 and should be either {a) drawn on a suitable part of the print, or {b) drawn separately on very white card which will be treated as a print (resulting background — faint grey). In Method 4, linear scales may be inserted where required. Explanations of all plates should be typed (double-spaced) in journal style and placed at the end of the typescript text; repetition of names, etc., should be avoided. Magnifi- cations should be stated. e.g. Fig. 1. Agenus beta (Author). Repository. Description, locality, magnification. Figs. 2-5. Begemis gamma (Author). 2, Repository. Description ... 3, Repository. Descrip- tion . . ., etc. ADMINISTRATION Preservation of types and other specimens. In accordance with recommendations of the International Codes of Botanical and Zoological nomenclature, these should be in the collection of a reputable institution such as a museum. The registered numbers should be quoted. Proofs. Authors will normally receive one proof ; this proof is for the purpose of correct- ing printing errors and not for altering the wording or substance of the paper. Authors will be charged for excessive alterations. The Editors will only be responsible for author corrections notified by return of post. Plate proofs will not normally be sent to authors. Offprints. Fifty offprints of each paper will be sent free of charge, and authors may purchase further copies at prices shown on an order form which will be sent to the author (or corresponding author) when a paper is accepted. Deposition of data. The Association expects to make use of an arrangement with the National Lending Library, Boston Spa, Yorkshire, whereby some publication expense can be saved by depositing, rather than printing, tables of data and other reference material under the N.L.L. Scheme. The deposited material is stored on microfiches; either microfiche copies or full-size enlargements may be obtained from N.L.L. by any applicant using N.L.L. prepaid coupons on a standard scale of charges which allows for NOTES FOR AUTHORS 681 postages (details of these coupons will be published in Association circulars at regular intervals). The N.L.L. will only accept deposited material through the Publication Committee of the Association ; all such material will be refereed as part of a published paper. The de- posited copy must be prepared camera-ready by the author ; page-size 300 mm X 2 1 0 mm, each page with descriptive heading or column headings for tabulation, all pages clearly numbered, brief introductory preface on first page. The published paper will bear a reference to deposited copy with full details of its pagination. Association policy on deposition is likely to evolve but at present neither plates nor formal taxonomic data will be considered for deposition. Authors should indicate and separate as appendices those parts of their papers that they propose for deposition; the Pubhcation Committee may also decide after referee comments that part of a paper should be deposited rather than printed. SPECIAL PAPERS IN PALAEONTOLOGY This Series is for longer papers than are normally accepted for Palaeontology. Prepara- tions of papers should be in the same style as for the Journal. Submission. Prospective authors should consult the Editors well in advance of sub- mission, supplying at least title, abstract, synopsis of contents and sample page and illustrations, so that an estimate of cost may be prepared. Cost of publication. The Association’s funds for this Series are limited. Authors are asked to obtain grants wherever possible. When this cannot be done by the author the Association will require time to seek such funds. Offprints. A small number of free offprints will be supplied and a further small number at a special reduced charge. Details will be supplied at the time by the editor concerned. GRANTS IN AID OE PUBLICATION Authors are requested to seek grants in aid of publication from their institutions or from research funds, or to apply for publication costs in research grants ; this is particu- larly necessary for long papers. Although acceptance of a paper for publication will not be dependent on the receipt of such grants, authors will appreciate that the funds available to the Association are limited. Every grant or donation therefore will directly help the Association to publish more. Prepared by the Publications Committee The Palaeontological Association LONDON 1972 THE PALAEONTOLOGICAL ASSOCIATION Annual Report of the Council for 1971-2 Membership. On 31 December 1971 there were 1,236 members (740 Ordinary, 107 Student, and 389 Institutional), a net decrease of 42 members during the year. The number of subscribers through Blackwell’s agency was 300, a net decrease of 24 on last year’s figure. Finance. During 1971 the Association published volume 14 of Palaeontology at a cost of £12,368; Special Papers 8 and 9 which are expected to cost £4,650; and a consolidated index to volumes 1-12 of Palaeontology at a cost of £671. This expenditure of £17,689 is the greatest the Association has ever spent in one year on publication. In addition volume 1 of Palaeontology has been reprinted at a cost of £2,080. Total expenditure was £20,908, an increase of £7,362 on 1970. Although administrative costs have risen sharply by £509, your Council believes that most members appreciate the enlarged circulars which account for this. The bulk of the enormous rise in expenditure is due to the relatively ambitious publication programme of the Association at a time of staggering increases in the costs of printing. Your Council has raised the subscription rate to Special Papers for Institutional Members to £8 p.a. from 1 January 1972, and for Ordinary and Student Members to £4 p.a. from I January 1973. The discount on the price of individual numbers of Special Papers to Ordinary and Student Members is now reduced to 25 per cent. Total income in 1971 at £15,935 was £1,306 more than in 1970, largely from the increased cover price of Palaeontology, but only marginally higher than the income in 1969. The Association is grateful to the Queen’s University, Belfast, for a grant of £285 to pay for the plates in Special Paper 1 ; to the National Museum of Wales; to Birkbeck College, London; and to Trinity College, Cambridge, for grants to assist the publication of papers in Palaeontology. In 1971 the Association received the last annual donation under a seven-year covenant from the Burmah Oil Company. This company, together with British Petroleum, Shell Oil, and Texaco, have given us a total of £7,975 since the Association was founded. The Council of the Association takes this opportunity of thanking these companies for their generous support over the years. The Association’s investments at the end of 1971 had a market value of £18,637, but the publication reserve account has dropped to £8,531, the lowest figure since 1967. This would be barely sufficient to publish three parts of Palaeontology. Although the cost of Palaeontology itself is adequately covered by subscriptions and sales, it is essentia! to increase the number of Institutional subscribers to Special Papers if we are to maintain the series without a large grant for each number. Publications. Four parts of volume 14 of Palaeontology were published during 1971 ; they contained 49 papers and consisted of 722 pages and 1 38 plates. A Cnmiilative Index to Palaeontology volumes 1—12 (1957-1969), prepared by Dr. Isles Strachan, was also published during the year. Special Papers in Palaeontology 8 (for 1970), 9 and 10 (for 1971) were published. Experiments with the printing of plates by offset lithography had proved so successful that from volume 15 part 1 all plates would be printed by this process. The text of Special Papers 9 and 10 was filmset to enable this process to be evaluated. If adopted it would mean that our journals could be printed in one run, eliminating the expensive operation of tipping in plates by hand, ensuring greater uniformity of quality in the plates and resulting in small savings in cost. Meetings. Five meetings were held during 1971-2. The Association is grateful to the Council of The Geological Society of London, Professors F. H. Stewart and G. Y. Craig (Edinburgh University) and Professor H. B. Whittington (Cambridge University) for generously granting facilities for meetings, to Drs. J. M. Hancock and W. J. Kennedy for leading the field meeting and to the local secretaries for their efficient services. a. The Fourteenth Annual General Meeting was held in the rooms of The Geological Society of London on Wednesday, 3 March 1971. The Annual Report of the Council for 1970-1 was ac- cepted and the Council for 1971-2 elected. Professor David Nichols of the University of Exeter C 9202 zz 684 THE PALAEONTOLOGICAL ASSOCIATION delivered the Fourteenth Annual Address on ‘The water vascular system in living and fossil echino- derms’. b. A Field Demonstration Meeting on Cenomanian ammonite zonation was led by Drs. J. M. Hancock and W. J. Kennedy to the Isle of Wight on 8 May 1971. c. A Lecture on ‘The relationship of ontogeny and phytogeny (myth, metaphor, analogy, common causality or identity ?)’ was given by Dr. Stephen J. Gould of the Museum of Comparative Zoology, Harvard University in the rooms of the Geological Society of London on Wednesday, 19 May 1971. d. A Discussion Meeting on the ‘Palaeoecology of reefs’ was held in the Appleton Tower of Edinburgh University as part of the first co-ordinated meeting of British geological societies, on Thursday, 9 September 1971. About 100 people attended this meeting, which comprised thirteen papers, eleven exhibits, and a film of the Great Barrier Reef. The Local Secretary was Dr. E. N. K. Clarkson. e. A Symposium on ‘Organisms and continents through time’ was held at Cambridge University, 15-17 December 1971. Over 200 people attended this symposium which was organized jointly with The Geological Society and the Systematics Association. The Local Secretary was Dr. C. P. Hughes. Council. The following were elected members of Council of the Association for 1971-2 at the Annual General Meeting on 3 March 1971: President'. Dr. W. S. McKerrow. Vice-Presidents'. Professor M. R. House, Dr. Gwyn Thomas. Treasurer'. Dr. J. M. Hancock. Membership Treasurer. Dr. A. J. Lloyd. Secretary : Dr. W. D. I. Rolfe. Editors : Mr. N. F. Hughes, Dr. I. Strachan, Dr. R. Goldring, Dr. J. D. Hudson, Dr. D. J. Gobbett. Other Members'. Dr. E. N. K. Clarkson, Dr. L. R. M. Cocks, Dr. R. H. Cummings, Dr. W. J. Kennedy, Mr. M. Mitchell, Dr. Marjorie D. Muir, Dr. B. Owens, Dr. W. H. C. Ramsbottom, Dr. Pamela L. Robinson, Dr. E. P. F. Rose, Dr. C. T. Scrutton, Dr. V. G. Walmsley, Dr. A. D. Wright. Dr. Julia A. E. B. Hubbard was co-opted to Council to serve as Circular Reporter. Circulars. The Circular was redesigned during the year and considerably enlarged to provide more information for Members; four Cireulars (Nos. 65-68) were distributed to Ordinary and Student Members, and to Institutional Members on request. Other activities. Council has been actively furthering the following projects, many of them suggested by the two successive Stimulus Committees. A list of palaeontologists in the U.K., both members and non-members of the Association, has been compiled by Dr. R. H. Cummings and will shortly be distri- buted. More palaeontological news and information has been included in Circulars in the form of book reviews, lists of palaeontological newsletters, lists of key references relating to lectures given to the Association, and lists of British and Irish theses and doctoral research topics in palaeontology. Members have provided brief reports on relevant meetings of other societies that they have attended. An informal group of members interested in Carboniferous palaeontology has been formed, with Dr. W. H. C. Ramsbottom and Mr. Murrary Mitchell as co-ordinators. The Association also helped to buy a copy of Professor D. M. Raup’s colour film on the computer simulation of shell form, which is now available for hire from the British Film Institute. Two ‘Catalogues of fossil fishes in The Royal Scottish Museum, Edinburgh’, produced by that museum, were distributed by the Association to its Institutional Members. Other issues on which discussion continues within Council are the future of palaeontological pub- lication— practical proposals are anticipated shortly; the feasibility of handbooks on fossils of parti- cular formations and the initiation of teach-in (‘master-class') meetings on specialist topics as a service to the Association. In all these activities. Council exhorts members to come forward with suggestions to advance the aim of the Association : to further palaeontology. BALANCE SHEET AND ACCOUNTS FOR THE YEAR ENDING 31 DECEMBER 1971 Liabilities Publications Reserve: Balance at 31 December 1970 13,504-06 Excess Expenditure 1971 4,973-12 Subscriptions for 1 972 in advance ...... Provision for ^nnXmg Palaeontology vol. 14 as per Income and Expenditure A/c ' 11,983-58 Less Expenditure already incurred ...... 9,239-14 Provision for printing Special Papers as per Income and Expenditure A/c Special Paper 8 ........ . 3,000-00 Special Paper 9 ........ . 1,650-00 Loan from the Royal Society ....... Sundry Creditors: Auditor .......... 25-00 Oxford University Press ........ 1,530-30 Assets Office Equipment ...... Investments at cost : Chester City Loan ...... £1,000 Treasury 9% 1994 .... £860 L.C.C. 6|% 1974 £1,000 Kirkby U.D.C. Loan .... Industrial & Commercial Finance Corporation £500 Loan Stock A 1972 .... £200 Loan Stock B 1974 .... £2,000 Agricultural Mort. Corp. 9|% 1980-5 Wagon Finance Corp. Loan .... Equities Fund for Charities 2135 units Foreign & Colonial Inv. Trust, 750 shares Grattan Warehouses, 266 shares Westland Aircraft, 500 shares .... New Throgmorton Inv. Trust, 4000 income shares Sundry Debtors : Authors for Offprints ..... Income Tax refunds due: 1970 investment income .... 1971 investment income .... Covenant on gen. donation .... Advance payment for offprints .... Cash at Bank : Deposit Account ...... Current A/c — Sheffield ..... 1,000-00 955-00 751-47 1,000-00 487-75 195-25 1,937-60 3,000-00 3,047-07 1,073-93 465-79 361-64 1,116-86 496-44 216-29 344-29 38-75 130-77 2,460-18 8,530-94 183-40 2,744-44 4,650-00 1,440-00 1,555-30 19,104-08 5-00 15,392-36 1,095-77 20-00 2,590-95 19,104-08 686 THE PALAEONTOLOGICAL ASSOCIATION Income and Expenditure Account for the year ended 31 December 1971 Expenditure To provision for cost of publication oi Palaeontology, Vol. 14 Part 1 Part 2 Part 3 Part 4 3,119-60 2,917-44 2,846-54 3,100-00 11,983-58 extra cost of Vol. 13 pt. 3 ....... 153-63 Vol. 13 pt. 4 231-00 12,368-21 To reprinting Vol. 1 ........ . 2,080-00 To provision for cost of publication of Special Papers No. 8 3,000-00 No. 9 1,650-00 4,650-00 To printing Index to Vols. 1-12 of Palaeontology .... 670-65 sales of Index . ........ 243-00 427-65 To Administration: Postage Circulars Meetings Stationery Membership of Societies . Honorarium; Mrs. Lloyd Audit Fee Posters .... Distribution of fish catalogue Film by Prof. D. Raup . Copying membership lists Miscellaneous 1,377-48 5-00 20,908-34 806-53 77-85 28-61 3-07 70-00 25-00 53-67 64-10 43-57 18-68 60-40 To Depreciation THE PALAEONTOLOGICAL ASSOCIATION 687 Income By Subscriptions for 1971 . Subscriptions for 1970 Sales of Palaeontology Sales of Offprints . . . . Sale of shares . . . . Interest received (gross) : Chester City Loan . . . . Treasury 9% 1994 . . . . L.C.C. 6i% 1974 . . . . Kirkby U.D.C I.C.F.C. Loan Stock A 1972 I.C.F.C. Loan Stock B 1974 Agricultural Mortgage 1980-5 Wagon Finance Loan Charifund (Equities Fund for Charities) Foreign and Colonial Investment Trust Grattan Warehouses .... Westland Aircraft .... New Throgmorton Inv. Trust Deposit Account .... 6,544-60 498-08 4,822-32 516-02 51-67 105-65 81-94 58-05 92-50 50-00 20-00 185-00 255-00 227-62 21-57 18-79 15-00 97-50 4-92 Less holding charges ......... By Specific Donations ........ General Donation: Burmah Oil ..... . Special Papers : Sales + Subs. ......... Special Donations ......... Sale of Offprints ......... Miscellaneous Receipts ........ Excess of Expenditure over Income transferred to Publications Reserve .......... 1,203-54 10-50 1,193-04 159-50 100-00 1,677-85 285-00 25-00 1,987-85 42-14 4,973-12 20,908-34 Report of the Auditors to the Members of The Palaeontological Association. We have examined the above Balance Sheet and annexed Income and Expenditure Account which in our opinion give respectively a true and fair view of the state of the Association’s affairs as at 31 December 1971 and of its income and expenditure for the year ended on that date. Thornton Baker & Co. Chartered Accountants, Auditors J i A: ''■■ M .. ■ • '( ■ .i-zt'VtJ-. :! f* :.: w - ' « ■i •iL 3 r h ' .' --ii"' t INDEX Pages 1-185 are contained in Part 1 ; pages 187-380 in Part 2; pages 379A-518 in Part 3 ; pages 519-675 in Part 4. Figures in Bold Type indicate plate numbers. A Acanthocrania papillifera, 474, 86. Africa, East: Jurassic belemnite ^ Rhopaloteuthis’ somaliensis, 158. Africa, South : hardground horizons in Cretaceous of Zululand, 539; Texanites-Spinaptychus association, 394; Triassic dinosaur skeleton, 29. Africa, West: Silurian Clarkeia fauna in, 623. Algae: crystal development in discoasters, 476; dasy- cladacean from Cretaceous of Borneo, 619; possible alga from English Cretaceous, 501. Ambrose, T. and Romano, M. New Upper Carboni- ferous Chelicerata (Arthropoda) from Somerset, England, 569. Amnionoidea : affinities of Idiohamites, 400 ; Dorseten- sia from Skye, 504; from transgressive Cretaceous in Gemany, 445; Saltern Cove goniatite bed, 430; shell structure of Triassic, 637; Texanites-Spiiia- ptychus association, 394. AmpIiicUna amoena, 78. Anisoceras aff. picteti, 81. Arbacia pimctulata, 5. Arberia minasica, 116, 24-26. Archegocystis stelhdifera, 6. Aristocystites bohemicus, 6. Arthropoda: moults of Cretaceous crab, 631; new Carboniferous chelicerates from Somerset, 569; nymphs of Carboniferous insects, 662. See also Ostracoda, Trilobita. Ash, S. R. Marcouia gen. nov., a problematical plant from the Late Triassic of the southwestern U.S.A., 423. — Late Triassic plants from the Chinle Formation in north-eastern Arizona, 598. Astroides calycularis, 14, 18. Athyris vittata, 78. Australia: Lower Gondwana pteridosperms, 108; new Devonian crinoid, 326; Precambrian coelen- terates, 197; two new Cambrian trilobites from Tasmania, 226. B Baker, P. G. The development of the loop in the Jurassic brachiopod Zeilleria leckenbyi, 450. Banks, H. P. The stratigraphic occurrence of early land plants, 365. Bate, R. H. Fossil and living Hemicypris (Ostracoda) from Lake Rudolf, Kenya, 184. — Phosphatized ostracods with appendages from the Lower Cretaceous of Brazil, 379A. Bates, D. E. A new Devonian crinoid from Australia, 326. Belemnite: morphology and taxonomic status of ' Rlwpaloteuthis' somaliensis, 158. Belgium: Lower Tertiary ostracods from, 267. BeUerophon, 79. Bishop, Gale A. Moults of Dakoticancer overanus, an Upper Cretaceous crab from the Pierre Shale of South Dakota, 631. Bivalvia: mechanical properties of shell structures, 73; Ptychodiis predation on Cretaceous Inocerainus, 439; stunting in Cerastodenna, 61. Black, C. C. Review of fossil rodents from the Neogene Siwalik Beds of India and Pakistan, 238. Black, M. Crystal development in Discoasteraceae and Braarudosphaeraceae (planktonic algae), 476. Borneo: new Cretaceous dasycladacean alga, 619. 'Botuid’ sp., 106. Braariidosphaera ; africana, 94; bigelowi, 93; discula, 93; hoschultzi, 93. Brachina delicata, 207, 42. Brachiopoda: Acanthocrania in Welsh Ordovician, 473; development of loop in Zeilleria, 450; origin of Silurian Clarkeia fauna of S. America, 623; systematic position of Cadomella, 405. Bradleya dolabra, 299, 53. Brazil: phosphatized Cretaceous ostracods with appendages, 379A; pteridosperms from Lower Gondwana, 108. Brett, D. W. Fossil wood of Platanus from the British Eocene, 496. Brightonicystis gregarius, 7. Britain: regional environmental changes across a Jurassic stage-boundary, 125. Brunton, C. H. C. and MacKinnon, D. L. The systematic position of the Jurassic brachiopod Cadomella, 405. C Cadomella, 405; davidsoni, 76-78; moorei, 76. Calymene sp., 359, 64. Cambrian: two new trilobites from Tasmania, 226. Canada: Silurian and Devonian monograptids from Yukon, 579; Triassic ammonoid shell structure, 637. Candona (Pseudocandona), 282; fertilis fertiUs, 282, 49; sp., 283, 47. Carboniferous: compression structures in miospore, 121; conodonts from Devon, 550; insect nymphs, 662; new chelicerates from Somerset, 569. Caryocystites ; dubia, 3, 4 ; lagenalis, 3, 4. Cerastodenna edule, 61, 8-10. Ceratocephalal sp., 361, 64. Chelicerata, new Carboniferous from Somerset, 569. 690 INDEX Chiozoon, 355 ; cowiei, 356, 63. Chomintes, 29. Cladarocythere, 285 ; apostolesciii, 287, 50, 56 ; haiitonensis, 288, 50. Cladocora caespitosa, 11, 13, 16, 21, 22. Clarkeia antisiensis, 121. Clayton, G. Compression structures in the Lower Carboniferous miospore DictyotrUetes admirabilis Playford, 121. Cocks, L. R. M. The origin of the Silurian Clarkeia shelly fauna of South America, and its extension to West Africa, 623. Coelenterata ; from Precambrian of S. Australia, 197; skeletal microstructure of corals, 88. Coleoidea: morphology and taxonomic status of '' Rhopaloteuthis' somaliensis, 158. Conodonts; Lower Carboniferous fauna from Devon, 550. Cooksonia sp., 65. Corals: skeletal microstructure, 88. Crab: moults of Cretaceous, 631. Creber, G. T. Gymnospermous wood from the Kimmeridgian of East Sutherland and from the Sandringham Sands of Norfolk, 655. (l)Crepidolithus, 28. Cretaceous: affinities of Idiohamites, 400; ammonites from Rhenish Massif, 445; hardground horizons, 539; moults of crab, 631; new dasycladacean alga from Borneo, 619; phosphatized ostracods with appendages from Brazil, 379A; possible alga, 501; Pfychodiis predation on Inoceramus, 439; Texanites- Spinaptycluis association, 394. Cretacicriista, 501; diihiosa, 501, 100, 101. Crinoid, new Devonian from Australia, 326. Cyamocytheridea, 290 ; pimctateila producta, 290, 52 ; p. pimctateila, 290, 56. Cyclomediisa, 204; davidi, 41; radiata, 42. Cyphoproteiis sp., 349, 61. Cypridopsis soyeri, 275, 48. ^Cypris' tenidstriata, 274, 47. Cystoids: morphology and function of pore-struc- tures, 1. Cytherella, 212. Cytherelioidea jonesiana, 272; j. joiiesiana, 273, 45; j. crassata, 273, 45. Cytheretta, 306. Cy tiler idea, 289; eberti, 290; pernota, 289, 56. Cytlieroinorpha zirmdorfi, 310, 55. Cytlieriira, 306; porciiia, 307, 55, 56. D Dakoticancer overanus, 631, 122, 123. Decliellyia gormaiii, 607, 115-118. Devon: Saltern Cove goniatite bed — a slump, 430; Carboniferous conodonts, 550. Devonian: early land plants, 365; monograptids from Yukon, 579; new Australian crinoid, 326; Saltern Cove goniatite bed, 430. Dicraiiopeltisl sp., 360, 61, 64. DictyotrUetes, 121. Dinosaur: post-cranial skeleton of Fabrosaiiriis, 29. Diplocraterion, 29. Discoaster; adamanteus, 87-89; asymnietricus, 91; broiiweri, 90-92 ; chaileugeri, 89 ; lodoensis, 93 ; obtiisus, 87, 88. Discotropites theron, 128. Dollymae, 553; hassi, 558, 109. Dorseteiisia, 505; iiannoveraiia, 513, 105; liebridica, 516, 105 ; liostraca, 506, 102-104 ; piiigiiis, 510, 105 ; roinani, 508, 103, 104. Drepauites hyatti riitherfordi, 127. Dyoros sp., 78. E Eagar, R. M. C. Use of the pictograph, 378. Ecliinocorys scutatiis, 5. Echiiiocytliereis, 304; El hamsteadensis, 305, 55, 56. Echinodermata : new Devonian crinoid from Australia, 326; pore-structures in cystoids, 1; water vascular system, 519. Echiiiosphaerites', aiirantiiim s.l., 4; a. americaniim, 3, 4 ; a. aiirantiiim, 2 ; a. sueciciis, 2. Ediacara fliiidersi, 218, 43. Elliott, G. F. Cretacicriista gen. nov., a possible alga from the English Cretaceous, 501. — Trinocladiis exoticiis, a new dasycladacean alga from the Upper Cretaceous of Borneo, 619. England: Carboniferous insect nymphs, 662; Eocene Plataniis wood, 496 ; Jurassic gymnospermous wood, 655; Lower Tertiary ostracods, 267; new Carboni- ferous chelicerates, 569; new Jurassic ostracods, 187; possible Cretaceous alga, 501. Eocene: new whale from India, 490; wood of Plataniis from Britain, 496. Eocytheropteron, 307; plicatoreticiilatiim, 308, 55. Eopliryniis jiigatiis, 576, 113. Eoporpita medusa, 198, 40, 41. Eiicypris, 276; amygdala, 276, 48; peciieibronnensis, 277, 48 ; sp., 277, 48. Eiiproops kilmersdonensis, 570, 112, 113, Evolution: of water vascular system in echinoderms, 519. Exogyra, 107. F Eabrosaiiriis australis, 29. Farrow, G. E. Periodicity structures in the Bivalve shell: analysis of stunting in Cerastoderma ediile from Burry Inlet (South Wales), 61. Fish : Ptycliodus predation on Inoceramus, 439. France: Jurassic brachiopod Cadoniella, 405; Lower Tertiary ostracods, 267. Frank it es sutlierlandi, 128. Functional morphology: mechanical properties of Bivalve shell structures, 73; pore-structures in cystoids, 1 ; water vascular system in echinoderms, 519. Fiingia scut area, 11, 14, 18, 23. G Gastropoda: Lower Palaeozoic Bellerophontacea from N. America, 412. INDEX 691 Germany: Cretaceous ammonites on Rhenish Massif, 445. Giiathodus, 559; delicatus, 559, 110; pimctatus, 560, 109, 110; semiglaber, 562, 110. Goldillaenid gen. et. sp. indet., 346, 347, 60, 62. Gondwana: Arberia and related fructifications from Brazil and New South Wales, 108. Grawaimella, 192; apostolescui, 193, 39. Graptolites: Upper Silurian and Lower Devonian monograptids from Yukon, 579. Greenland, new Silurian trilobites from, 336. Gymnosperms: Arberia and related Gondwana female fructifications, 108; Marconia, a problematical Triassic plant, 423; Triassic plants from Arizona, 598; wood from British Jurassic, 655. H Hammatocythere, 300; hebertiana, 301, 53, 54, 56; trituberculata, 303, 53. Hancock, J. M., Kennedy, W. J. and Klaumann, H. Ammonites from the transgressive Cretaceous on the Rhenish Massif, Germany, 445. Haplosphaeronis\ oblonga, 6; sp. nov., 6. Heliocrinites', guttaeformis, 3, 4; ovalis, 2; stellatus, 2; sp. nov., 3. Hemicyprideis, 291; elougata, 295, 52; helvetica, 52; henisensis, 297, 52; montosa, 291, 51, 52, 56. Hemicypris, 184; klei, 184; posterotruncata, 184. HeterortheUa africana, 121. Holocystites', alternatus, 7; saitellatus, 7. Homoeospira evax, 78. Hydrozoa: Precambrian of S. Australia, 197. Hyphoplites aff. canipichei, 81. Hyrokobe, 357; pharanx, 358, 64. Hystrix, 246; cf. leucurus, 247; sivaleiisis, 246. I Idiohamites ellipticoides, 400, 74, 75. Idoptilus onisciformis, 666, 132. Ilyocypris, 283; boehli, 283, 45; cranmorensis, 285, 45. India: new Eocene whale, 490; Siwalik rodents, 238. Inoceramus, 439; tenuis, 81. Insecta: nymphs of Carboniferous Palaeodictyoptera, 662. J Jackson, D. E. and Lenz, A. C. Monograptids from the Upper Silurian and Lower Devonian of Yukon Territory, Canada, 579. Jago, J. B. Two new Cambrian trilobites from Tas- mania, 226. Jeletzky, J. A. Morphology and taxonomic status of the Jurassic belemnite ^Rhopaloteuthis' somaliensis Spath 1935, 158. Jurassic: ammonite Dorsetensia in Skye, 504; gymno- spermous wood, 655; loop in brachiopod Zeilleria, 450; morphology of belemnite 'Rlwpaloteiilhis', 158; new genera of ostracods, 187; regional environmental changes across a stage-boundary, 125; systematic position of Cadomella, 405. Jiivenites septentrionalis, 126. K Kanisamys, 258; indicus, 258; sivaleiisis, 259. Kauffman, E. G. Ptychodus predation upon a Cre- taceous Inoceramus, 439. Keen, M. C. The Sannoisian and some other Upper Palaeogene ostracoda from north-west Europe, 267. Kennedy, W. J. The affinities of Idiohamites ellipti- coides Spath (Cretaceous Ammonoidea), 400. — and Klinger, H. C. A Texanites-Spinaptychus association from the Upper Cretaceous of Zululand, 394. ■ Hiatus concretions and hardground horizons in the Cretaceous of Zululand (South Africa), 539. Kennedy, W. J. See also Hancock, J. M. Kenya: fossil and living Hemicypris from Lake Rudolf, 184. Kimberella qiiadrata, 214, 43. Klaumann, H. See Hancock, J. M. Klinger, H. C. See Kennedy, W. J. Koninckella triassina, 78. L Labeceras {L.) plasticum, 75. Lane, P. D. New trilobites from the Silurian of north- east Greenland, with a note on trilobite faunas in pure limestones, 336. Layman, M. See Taylor, J. D. Legiiminocythereis, 297 ; verricula, 298, 54, 56. Lenz, A. C. See Jackson, D. E. Lineocypris sp., 283, 49. Lithophaga, 106, 108. Lophelia prolifera, 11, 15, 19, 21, 23. Lophotocystis', araneiis, 2; granatum, 1; malaisei, 1. Lord, A. Wicherella and Gramannella, two new genera of Lower Jurassic Ostracoda from England, 187. Loxoconclia, 309; delemontensis, 309, 55; nystiana, 309, 55. M MacKinnon, D. L. See Brunton, C. H. C. Maclearnoceras enode, 653, 128. Marmnals: new Eocene whale from India, 490; review of Siwalik rodents, 238. Manicina areolata, 11, 18, 20-22. Marconia, 423; neuropteroides, 424, 80. Masculostrobus clathratns, 613, 119. Matthews, S. C., Sadler, P. M. and Selwood, E. B. A Lower Carboniferous conodont fauna from Chillaton, southwest Devonshire, 550. Megaphyllites humilis, 127. Megerlia truncata, 77. Meroperix ataphriis, 343, 60. Micrantholithus', concinniis, 96; pinguis, 95. Microcytherura, 308; delpbina, 309, 55. Microfossils: Carboniferous conodonts from Devon, 550; compression structures in miospore, 121; Hemicypris from Kenya, 184; new genera of Jurassic ostracods, 187; phosphatized ostracods with appendages, 379 A; Sannoisian ostracods from north-west Europe, 267. 692 INDEX Miospore, compression structures in Carboniferous, 121. Mishra, V. P. See Sahni, A. Moeiwcypris, 211 \ forbesi, 278, 46; inida, 280; reidi, 280, 46, 47; sherborni, 279, 46, 47. Mollusca. See Ammonoidea, Bivalvia, Coleoidea, Gastropoda. Monograptus, 581; aequabilis aeqiiabilis, 581; a. notoaeqiiabilis, 583; cf. balticus, 585; boiiceki, 588; cf. cornutiis, 589; cf. craigemis, 589; helfcoides, 590; cf. hercynicus subbercyniciis, 591; uniformis angustideus, 593; u. pamngustidens, 594; u. uni- formis, 595. Montastrea annularis, 12. Morton, N. The Bajocian ammonite Dorse tensia in Skye, Scotland, 504. Moults, of Cretaceous crab, 631. Muensterites glaciensis, 127. N Nathorstites macconnelli, 124, 125. Neocyprideis williamsoniana, 297, 52. Nesokia cf. hardwicki, 265. Nichols, D. The water-vascular system in living and fossil echinoderms, 519. Nordophiceras spathi, 126. North America : Lower Palaeozoic Bellerophontacea, 412; moults of Cretaceous crab, 631; Triassic plants, 423, 598; stratigraphic occurrence of early land plants, 365. O Opoa, 340; adainsi, 341, 59. Opsidiscus, 221 \ argusi, 229, 44; bilobatus, 44. Ordovician: brachiopod Acanthocrania in Wales, 473; some Bellerophontacea from N. America, 412. Ostracoda: fossil and living Heinicypris from Kenya, 184; phosphatized with appendages from Brazilian Cretaceous, 379; Sannoisian from north west Europe, 267; two new Jurassic genera, 187. Ostrea, 107. Otarion, sp., 351, 61. Owenites koeneni, 126. P Pakistan: review of Siwalik rodents, 238. Palaeoecology : hiatus concretions and hardground horizons, 539; trilobite faunas in pure limestones, 336. Palaeogeography: changes across a Jurassic stage- boundary, 125; Silurian Clarkeia fauna in S. America and W. Africa, 623. Palaeozoic: pore-structures in cystoids, 1. Paraulacodus, 243; indicus, 244. Pattersoncypris, 380A; micropapillosa, 381, 66-71. Paul, C. R. C. Morphology and function of exothecal pore-structurcs in cystoids, 1. Peel, J. S. Observations on some Lower Palaeozoic tremanotiform Bellerophontacea (Gastropoda) from North America, 412. Pemnia; papillatum, 94; rotundum, 95. Pentacystis; simplex, 7; sphaeroidalis, 1. Permo-Trias: Lower Gondwana female fructifica- tions, 108. Pernerocrinus, 326; discus, 328, 57, 58; sp., 329, 58. Pictograph, use of, 378. Pinna folium, 28. Plants: Late Triassic from Arizona, 598; stratigraphic occurrence of early land plants, 365. See also Algae, Gymnosperms, wood. Plataninium, 497 ; decipiens, 497, 99. Platanus, 496; sp., 496, 98, 99. Plectothyrella', crassicosta, 121; haiightoni, 121. Polygnathus communis carina, 563, 111. Pontocytliere therwilensis, 291. Porites porites, 12, 16, 23. Precambrian: coelenterates from Ediacara fauna, S. Australia, 197. Proarcestes sp., 126. Proliserpula sp., 106, 107. Protachyoryctes, 259; tatroti, 260. Protocetus, 490; sloani, 491, 97. Pseudopolygnathus triangulus, 564, 111. Ptychodus, 439, 81. Pimctatisporites; admirabilis, 122, 27; planus, 123, 27. R Rattus colberti, 264. Recent : coral skeletons, 88 ; Hemicypris from Kenya, 184; stunting in cockles, 61. Rhizocorallium, 29. Rhizomyoides, 249; nagrii, 254; pilgrimi, 255; pin- joricus, 251 \ punjabiensis, 252; sivalensis, 249. ‘Rhopaloteuthis’' somaliensis, 158, 30-38. Rigby, J. F. On Arberia White, and some related Lower Gondwana female fructifieations, 108. Rochdalia parkeri, 663, 132. Romano, M. See Ambrose, T. Rugoconites, 220; enigmaticus, 220, 43; tenuirugosus, 222, 43. S Sadler, P. M. See Matthews, S. C. Saetograptus pilosus, 595. Sageceras haidingeri, 127. Sahni, A. and Mishra, V. P. A new species of Proto- cetus (Cetacea) from the Middle Eocene of Kutch, Western India, 490. Salpingostoma buelli, 415, 79. Sayimys, 240; perplexus, 240; sivalensis, 241. Scliaryia, sp., 352, 61. Schizosphaereila, 28. Schloenbachia', varians subplana, 81; v. subtuberculata, 81; V. subvarians, 8) ; v. varians, 81. Schmalenseeia, 232; gostinensis, 233, 44. Sciponoceras roto, 81. Scotland: Bajocian ammonite Dorsetensia from Skye, 504; Jurassic gymnospermous wood, 655. Scyphozoa: Precambrian of S. Australia, 197. Selaginella anasazia, 601, 114. Selenoliarpes loma, 353, 62. INDEX 693 Sellwood, B. W. Regional environmental changes across a Lower Jurassic stage-boundary in Britain, 125. Selwood, E. B. See Matthews, S. C. Siderastrea radians, 12, 15, 17, 23. Silurian: early land plants, 365; monograptids from Yukon, 579; new trilobites from Greenland, 336; origin of South American Clarkeia fauna, 623; some Bellerophontacea from North Africa, 412. Sipbonodella, 564; cremdata, 564, 111; obsoleta, 565, 111. Sivancanthion complicatiis, 246. Skye: Bajocian ammonite Dorsetensia from, 504. Somalibelus, 158; somaliensis, 161, 30-38. Sorauf, J. E. Skeletal microstructure and micro- architecture in Scleractinia (Coelenterata), 88. South America: origin of Silurian Clarkeia shelly fauna, 623. See also Brazil. South Wales: analysis of stunting in cockles, 61. Spathognatbodus cf. stabilis, 565, 109. Spbaeronites', globulus, 5; litcbi, 5; pomum, 5; pyri- formis, 5; sp. nov. 5. Spiiiaptyclms, 394, 72, 73. Spiroserpula, 107. Stenopareia sp., 348, 61. Structure: coral skeleton, 88; mechanical properties of Bivalve shells, 73; periodicity structures in Bi- valve shells, 61 ; Triassic ammonoid shell, 637. T Tasmania: two new Cambrian trilobites, 226. Taylor, J. D. and Layman, M. The mechanical pro- perties of Bivalve (Mollusca) shell structures, 73. Technique: use of the pictograph, 378. Tertiary: crystal development in discoasters, 476; fossil rodents of India and Pakistan, 238; new whale from India, 490; ostracods from north-west Europe, 267 ; wood of Platanus, 496. Tetractinella trigonella, 78. Texanites, 394; soutoni, 72. Thalassinoides, 29. Thulborn, R. A. The post-cranial skeleton of the Triassic ornithischian dinosaur Fabrosaiiriis aust- ralis, 29. Tozer, E. T. Observations on the shell structure of Triassic ammonoids, 637. Trace fossils: Lower Jurassic in Britain, 125. Tracbypbyllia amarantum, 16, 17, 22. Tremanotus alpbeus, 417, 79. Treniatocystis; globosus, 7; rotundus, 7. Triamara sp., 6. Trias: ammonoid shell structure, 637; plants from U.S.A., 423, 598; post-cranial skeleton of Fabro- saurus, 29. Trilobita: new Silurian from Greenland, 336; two new Cambrian from Tasmania, 226. Trinocladus, 619; exoticus, 620, 120. Tucker, M. E. See Van Straaten, P. U Ulricbocystis exiniia, 1. V Van Straaten, P. and Tucker, M. E. The Upper Devonian Saltern Cove Goniatite Bed is an intra- formational slump, 430. Vecticypris jacksoni, 281, 48. Vertebrates: new Eocene whale from India, 490; post- cranial skeleton of Fabrosaurus, 29; Ptycbodus predation on Inoceramus, 439; review of Siwalik rodents, 238. W Wade, Mary. Hydrozoa and Scyphozoa and other medusoids from the Precambrian Ediacara fauna. South Australia, 197. Wales: Ordovician brachiopod Acantbocrania, 473. Warburgella sp., 350, 61. Wicberella, 187; semiora kirtonensis, 191, 39; s. semiora, 189, 39. Wood: gymnospermous from British Jurassic, 655, Platanus from British Eocene, 496. Wootton, R. J. Nymphs of Palaeodictyoptera (Insecta) from the Westphalian of England, 662. Wright, A. D. The brachiopod Acanthocrania in the Ordovician of Wales, 473. Z Zeilleria leckenbyi, 450, 82-85. Zululand: Cretaceous hiatus concretions, 539; Texanites-Spinaptycbus association, 394. PALAEONTOLOGY The journal Palaeontology is devoted to the publication of papers on all aspects of palaeon- tology. Review articles are particularly welcome, and short papers can often be published rapidly. A hi^ standard of illustration is a feature of the journal. Four parts at least are published each year and are sent free to all members of the Association. Members who join for 1972 will receive Volume 15, parts 1 to 4. All back numbers are still in print and may be ordered from B. H. Blackwell, Broad Street, Oxford, England, at £5 per part (post free). A complete set. Volumes 1-14, consists of 55 parts and costs £275. SPECIAL PAPERS IN PALAEONTOLOGY This is a series of substantial separate works published by the Association. The subscription rate is £8 (U.S. $22.00) for Institute Members and £4 (U.S. $1 1 .00) for Ordinary and Student Members. Subscriptions and orders by members of the Association should be placed through the Membership Treasurer, Dr. A. J. Lloyd, Department of Geology, University College, Gower Street, London, WCIE 6BT, England. The following Special Papers are available. Members may obtain them at reduced rates through the Membership Treasurer. Non-members may obtain them from B. H. Blackwell, Broad Street, Oxford, England, at the prices indicated. 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 Femvale Limestones of the Arbuckle Mountains, 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 Brachio- poda, by A. williams and a. d. weight. 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). 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). PALAEONTOLOGY VOLUME 15 • PART 4 CONTENTS The water-vascular system in living and fossil echinoderms. By david nichols 519 Hiatus concretions and hardground horizons in the Cretaceous of Zululand. By W. J. KENNEDY and H. C. KLINGER 539 A Lower Carboniferous conodont fauna from Chillaton, south-west Devonshire. By S. C. MATTHEWS, P. M. SADLER, and E. B. SELWOOD 550 New Upper Carboniferous Chelicerata (Arthropoda) from Somerset. By T. AMBROSE and M. ROMANO 569 Monograptids from the Upper Silurian and Lower Devonian of Yukon Territory, Canada. By dennis e. jackson and A. c. lenz 579 Late Triassic plants from the Chinle Formation in north-eastern Arizona. By SIDNEY R. ASH 598 Trinocladus exoticus, a new dasycladacean alga from the Upper Cretaceous of Borneo. By graham f. elliott 619 The origin of the Silurian Clarkeia shelly fauna of South America, and its extension to West Africa. By L. R. M. cocks 623 Moults of Dakoticancer overanus, an Upper Cretaceous crab from the Pierre Shale of South Dakota. By gale a. bishop 631 Observations on the shell structure of Triassic ammonoids. By E. T. tozer 637 Gymnospermous wood from the Kimmeridgian of East Sutherland and from the Sandringham Sands of Norfolk. By G. T. creber 655 Nymphs of Palaeodictyoptera (Insecta) from the Westphalian of England. By robin J. wootton 662 Notes for authors submitting papers for publication in Palaeontology 676 PRINTED IN GREAT BRITAIN AT THE UNIVERSITY PRESS, OXFORD BY VIVIAN RIDLER, PRINTER TO THE UNIVERSITY I O O xig^DC^ RIES SMITHSONIAN INSTITUTION NOIlfUliSNI “"NVINOSHil WS S 3 1 d V d 8 IT ”* L I B R A R I ES^ SMITHSONI CD ?o > IpS :d m x^osv^ ^ m ' "S^' ^ rn CO ± CO ± CO _ NVINOSHilWS S3ldVdan LIBRARIES SMITHSONIAN INSTITUTION NOHnillSNl NVINOSHil RIES SMITHSONIAN INSTITUTION NOIiniliSNI_ NVINOSHillMS S3ldVdaiT LIBRARIES SMITHSONI _i z _j 2 -j z liSNi NvmosHilWS SBiavaan libraries smithsonian institution NoiiniiiSNi nvinoshiii o) > :d m ^ XoujASX m A ' 2 X^OSV^ Z CO T LIBRARIES SMITHSONI, Z CO ^ ^ (pf' " -//J; c ' ■ ' / ij m VijMUir^Cy Q X^OS'i^X _ X>ASt^^X O _ z _i 2: _ liSNI NVINOSHilWS S3ldVdaiT LIBRARIES SMITHSONIAN INSTITUTION NOliniliSNl NVINOSHill 1 • «' sol = fri irX ;TITUTI0N^N0lini!iSNll”'NVIN^SHilWS^S3 laVdail^'uBRARI es^smithsonian”‘institution NOlir TITUTION NOliniliSNI NVINOSHil^^JS S3iavyan LIBRARIES SMITHSONIAN INSTITUTION NOlif Crt ^ ^ (f) ~ ^ avaan libraries smithsonian institution NoiiniiisNi nvinoshiiiais saiavaan libr r- z r- z t- z {/5 — ^ — in \ — in 5T1TUTION NOliniliSNI NVINOSHil/JS S3iaVaaiT LIBRARIES SMITHSONIAN INSTITUTION NOlif z in z in z .V. ■Sy/Mi ^ O //V. X CO ^ - " 'W' - t ^ > ■ 2 2 ^ . > X{0m5>iX 2 ^ z C/) 2 CO‘Z x avaan libraries smithsonian institution NoiiniiiSNi nvinoshtiws saiavaan libr — C/5 2 „ X ... cn — O XO\dC^ — O "CtD^ _ O ITITUTION NOliniliSNl'^NVlNOSHilWS S3iavaaiT libraries smithsonian institution NOlif lavaan l 2 < DO . o- I XI- rn 7 rn X^ostA^2 ^ m CO ' — CO _ CO BRARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHillNS S3iavaaiT LI CO z . CO 2 CO 2 s: < A. 2 -JSi!. < 2 .0^ . < cc^ vx CO CO 2 C/1 •- 2 CO Z TITUTION NOliniliSNI NVlNOSHilWS S3iavaaiT LIBRARIES SMITHSONIAN INSTITUTION NOlif CO \ CO — CO ~ CO BRARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3iavaai3 LIBR