S, BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY VOL. XVII 1968-1969 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) LONDON: 1970 DATES OF PUBLICATION OF THE PARTS No. i. igth November . . . . . 1968 No. 2. igth November ..... 1968 No. 3. 1 7th January . . , . . 1969 No. 4. ryth January . . . . . 1969 No. 5. aist January ..... 1969 No. 6. 7th March ...... 1969 No. 7. I5th April ...... 1969 No. 8. i6th May ...... 1969 PRINTED IN GREAT BRITAIN BY ALDEN & MOWBRAY LTD AT THE ALDEN PRESS, OXFORD CONTENTS GEOLOGY VOLUME XVII No. i. On the Cretaceous age of the so-called Jurassic Cheilostomatus Polyzoa (Bryozoa). A contribution to the knowledge of the polyzoa-fauna of the Maastrichtian in the Cotentin (Manche). E. VOIGT I No. 2. The caudal skeleton in Mesozoic Acanthopterygian fishes. C. PATTERSON 47 No. 3. Non-calcareous microplankton from the Cenomanian of England, Northern France and North America. Part I : R. J. DAVEY 103 No. 4. A redescription of W. Carruthers ' type Graptolites. I. STRACHAN 181 No. 5. A revision of the English Wealden Flora, I Charales-Ginkgoales. J. WATSON 207 No. 6. Two new Dicynodonts from the Triassic Ntawere Formation, Zambia. C. B. Cox 255 No. 7. Lower Cambrian Archaeocyatha from the Ajax Mine, Beltana, South Australia. F. DEBRENNE 295 No. 8. Some Bathonian Ostracoda of England with a revision of the Jones 1884, and Jones & Sherborn 1888 Collections. R. H. BATE 377 Index to Volume XVII 439 ON THE CRETACEOUS AGE OF THE SO-CALLED JURASSIC CHEILOSTOMATUS POLYZOA (BRYOZOA) A CONTRIBUTION TO THE KNOWLEDGE OF THE POLYZOA-FAUNA OF THE MAASTRICHTIAN IN THE COTENTIN (MANCHE) E. VOIGT BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 17 No. i LONDON: 1968 ON THE CRETACEOUS AGE OF THE SO-CALLER ' JURASSIC CHEILOSTOMATOUS POLYZOA (BRYOZOA) A CONTRIBUTION TO THE KNOWLEDGE OF THE POLYZOA-FAUNA OF THE MAASTRICHTIAN IN THE COTENTIN (MANCHE) BY EHRHARD VOIGT Pp. 1-45; 8 Plates; 2 Text-figures BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 17 No. i LONDON: 1968 THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted in 1949, is issued in five series corresponding to the Departments of the Museum, and an Historical series. Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year. In 1965 a separate supplementary series of longer papers was instituted, numbered serially for each Department. This paper is Vol. 17, No. I of the Geological Palaeontological series. The abbreviated titles of the Periodicals cited follow those of the World List of Scientific Periodicals. World List abbreviation Bull. Br. Mus. nat. Hist. (Geol.). Trustees of the British Museum (Natural History) 1968 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) Issued 19 November 1968 Price 2 ON THE CRETACEOUS AGE OF THE SO-CALLED JURASSIC CHEILOSTOMATOUS POLYZOA (BRYOZOA) A CONTRIBUTION TO THE KNOWLEDGE OF THE POLYZOA-FAUNA OF THE MAASTRICHTIAN IN THE COTENTIN (MANCHE) By EHRHARD VOIGT SYNOPSIS The supposed Jurassic cheilostomatous Polyzoa described by J. W. Gregory (1894) as Mem- branipora jurassica and Onychocella bathonica from the Bathonian of Ranville (Calvados), are of Cretaceous age and must have come from the Maastrichtian of the Cotentin (Manche, France). The matrices of the type specimens yielded thirty-three further species of Maastrichtian Polyzoa of which nine are Cyclostomata and twenty-four Cheilostomata. Three new species Radulopora minor n. sp., Rosseliana thomasi n. sp. and Frurionella fertilis n. sp., and the new genus Radulo- pora are described. CONTENTS I. INTRODUCTION AND ACKNOWLEDGMENTS ..... 3 II. CHEILOSTOMATA OF CRETACEOUS AGE DESCRIBED FROM JURASSIC BEDS .......... 4 III. POLYZOA FROM JURASSIC BEDS WHICH HAVE BEEN ERRONEOUSLY REFERRED TO THE CHEILOSTOMATA ...... y IV. ON THE ORIGIN OF GREGORY'S " JURASSIC " CHEILOSTOMATA . . IO V. PALAEOZOIC CHEILOSTOMATA? . . . . . . . 13 VI. THE ACCOMPANYING FAUNA OF THE ROCK-MATRIX OF GREGORY'S " JURASSIC " CHEILOSTOMATA . . . . . . . 14 VII. REFERENCES .......... 42 I. INTRODUCTION AND ACKNOWLEDGMENTS THE predominant orders of Polyzoa (Bryozoa) of the Cretaceous and Cainozoic periods are the Cyclostomata and the Cheilostomata. The Cheilostomata have been increasing ever since an explosive development in the Upper Cretaceous, whereas the Cyclostomata have decreased during the Tertiary and Quaternary following a flourishing period in the Cretaceous. Today there exists only a comparatively small relic of this cyclostomatous fauna stem which, in Jurassic times, was the only living group of Polyzoa apart from some rare species of boring Ctenostomata. This last statement contradicts all textbooks of palaeontology in which the Cheilostomata are being erroneously recorded as beginning at the latest in the Jurassic, an opinion repeated by R. S. Bassler (1953). It is the purpose of this paper to prove that the so-called Jurassic Cheilostomata described by Gregory (1894) from the Bathonian of Ranville (Calvados), are in fact GEOL. 17, I. I 4 UPPER CRETACEOUS POLYZOA FROM COTENTIN Upper Cretaceous in age and must have their origin in the Maastrichtian of the Cotentin (Manche) in Normandy. Other Jurassic Polyzoa described as Cheilostomata by several authors do not belong to that order but are true Cyclostomata. In any case it must be emphasized that, if they are undoubtedly Jurassic, they are not Cheilostomata, or if they belong to that order they are not Jurassic. I wish to record my warm appreciation of the late Dr. Dighton Thomas of the British Museum (Natural History) to whom I am very much indebted for studies in the Museum collection, for the loan of specimens, for discussion, and for his help and advice. At the time of his death he had begun tidying the English of the manuscript, and this was kindly finished by Dr. J. M. Hancock. I am also grateful to Dr. A. B. Hastings for interesting discussions about the subject. Grateful thanks are due to Prof. J. Lehman, Dr. J. Sornay and Dr. E. Buge of the Musee d'Histoire Naturelle (Paris) for permission to study and to photograph the types of Bryozoa in the d'Orbigny-Collection. I also wish to thank Dr. J. Roger, Paris, for kindly accompanying me and guiding me in the Maastrichtian territory of the Cotentin and other French classic localities. Finally my special thanks are due to the Deutsche Forschungsgemeinschaft for financial help to further this research and their support for my studies on Cretaceous Polyzoa. II. CHEILOSTOMATA OF CRETACEOUS AGE DESCRIBED FROM JURASSIC BEDS Considering first the ages of undoubted Cheilostomata which have been described as Jurassic forms, we can enumerate the following three species: Flustra flabelliformis Lamouroux 1821, Onychocella bathonica Gregory 1894, and Membranipora jurassica Gregory 1894. i. " Flustra " flabelliformis Lamouroux 1821 1821 Flustra flabelliformis Lamouroux: 113, pi. 76, figs. 11-13. This species, described in a footnote and figured by Lamouroux, apparently from Ranville, is a fanlike fragment of a bilaminar Onychocella. The very simple figure shows pyriform zooecia with round orifices increasing in size to the periphery of the frond. An avicularium, rounded distally, seems to be indicated. Below the broken cryptocyst the rectangular or box-like pattern of the zooecia is visible. The original diagnosis of Lamouroux is: "Flustra en forme d'eventail, fossile, epaisse, a bords entiers, composee de deux membranes, une superieure mince, un peu translucide, divisee en alveoles profonds, a bords irreguliers avec un oscule rond dans le centre, qui communique a une cellule en forme de carre long, tres regulier, avec des cloisons epaisses et solides, les transversales alternant entre elles, les longitudinales se prolongeant sans interruption de la base aux extremites; grandeur, 2 a 3 centimetres; epaisseur, environ i millimetre." UPPER CRETACEOUS POLYZOA FROM COTENTIN 5 Unfortunately there is no possibility of restudying the typespecimen because, according to Sherborn, Lamouroux's collection has been lost. Nevertheless, there is no doubt that the specimen is not from the Bathonian of Ranville as supposed by Gregory, but must be regarded as a Cretaceous Onychocella, likely to have come from one of the Maastrichtian localities of the Cotentin (Manche), as is demonstrated for the two following species. Lamouroux himself gives as the origin only the neighbourhood of Caen, which perhaps includes not only Ranville but also the Cotentin. Gregory (1896 : 214) in his Catalogue of the Jurassic Bryozoa in the British Museum suppressed his specific name bathonica for an Onychocella, regarded by him in 1894 as a new species from the Bathonian of Ranville, in favour of Lamouroux's species because he believed that the two forms were identical. As will be shown, Gregory's Onychocella bathonica is the well-known Maastrichtian Onychocella piriformis Goldfuss 1826 first described from the " Maastrichter Tuff- kreide " in the Netherlands. If Gregory was correct in identifying his Onychocella bathonica with Flustra flabelliformis Lamouroux, the latter name has priority. On the other hand Lamouroux's figure is rather unsatisfactory, as Gregory himself stated, for a conclusive identification of Flustra flabelliformis Lamouroux with Onychocella piriformis Goldfuss. There are so many similar species of Onychocella in the Upper Cretaceous that it seems to be quite impossible to find out which species was intended by Lamouroux under this name. Gregory notes that Lamouroux's figure is so indefinite that Pictet gave a figure of a form, which he referred to this species, which was really Diastopora lamellosa Michelin. It is probable that Gregory, in identifying his Onychocella bathonica with Flustra flabelliformis, was influenced by his supposition that they are of Bathonian age, and as Jurassic species of Onycho- cella must be very scarce, he believed that they must be identical. 2. " Onychocella bathonica " Gregory 1894 ( = Onychocella piriformis Goldfuss 1826) (PL 5, figs. 11-12) 1826 Eschar a piriformis Goldfuss: 23, pi. 8, fig. 10. 1851 Eschara piriformis (Goldfuss) von Hagenow: 75, pi. 9, fig. 6 and pi. n, fig. 6. 1894 Onychocella bathonica Gregory : 63, fig. 2 1896 Onychocella flabelliformis (Lamouroux) ; Gregory : 214, fig. 22. 1930 Onychocella piriformis (Goldfuss); Voigt: 454, pi. 16, figs. 1-2. This is the supposed Jurassic cheilostomate which was described and figured by Gregory under this name in 1894 and, as stated above, later treated by him as a synonym of " Flustra "flabelliformis Lamouroux. The first mention of a " Jurassic Onychocella " was made by Gregory (1893 : 239). It is represented in the British Museum Collection by two specimens, D.iSi (type) and 0.480, both from the Tesson-Collection. As previously suggested by the present author (Voigt 1930 : 454), these two specimens are not from the Bathonian " Calcaire a polypiers " as recorded 6 UPPER CRETACEOUS POLYZOA FROM COTENTIN by Gregory, but both are Upper Cretaceous in age. This is proved firstly by a Cheilostomatous Polyzoan, Stamenocella marginata (d'Orbigny) which is visible in the matrix of specimen D.i8i (PL 3, fig. i) and which was overlooked by Gregory, and secondly by a large fauna of other Maastrichtian Polyzoa which has been isolated from the adherent matrix of the block D.iSi and which is described in section VI. This study shows clearly that all these supposed Bathonian Polyzoa are in fact of Upper Cretaceous age as already indicated for " Membraniporajurassica " by Lang (1916 : 96, 97 and 1922 : 197-198) and Larwood (1962 : 223). Onychocella bathonica Gregory is the same species as Onychocella piriformis (Gold- fuss) 1826 from Maastricht. It is quite commonly found in various localities of the French Cotentin (Manche). Although Gregory has discussed Goldfuss's species and written that it has a lower zooecial aperture, while the avicularian aperture is larger and the front wall occurs only above and not on both sides of this, a comparison with a specimen from Chef du Pont indicates that they are synonyms (pi. 5, fig. n). The supposed differences are not significant and they are not found when material from Maastricht or from the Cotentin localities is used for comparison. Some of the opesiae 1 in the figured (pi. 5, fig. 12) British Museum specimen D.iSi are a little smaller than those of the figured specimen from Chef du Pont (pi. 5, fig. n), but there are variations within the same specimen. Gregory's figure is deceptive because it shows a small quadrangular fragment with only five whole zooecia and one avicu- larium; it does not correspond in size to his two originals. The Holotype of Onycho- cella bathonica Gregory is a large bent unilaminar frond of nearly 3 cm. length (pi. 5, fig. 12). It contains a dozen avicularia and shows, on some zooecia, the very small characteristic endozooecial ovicells which are just visible as minute swellings at the distal ends of the zooecia figured here (pi. 5, fig. 12). The other specimen 0.480 is a large unilaminar fragment of n x 8-5 cm. size with four avicularia. In discussing the affinities of Onychocella flabelliformis (Lamouroux) Gregory maintains that its nearest ally may be von Hagenow's Cellepora (Discopora) konincki- ana (1851 : 95, pi. n, figs. 11-12) from Maastricht, a species which, with its straight rows of avicularia and small zooecia, is very differently shaped. In discussing Onychocella ( = Cellepora} koninckiana (1896 : 215) he distinguished the form figured by von Hagenow in his fig. n as a new species Onychocella hagenowi. He assigned it to a separate species because he believed that it has larger elliptical opesiae with the longer axis longitudinal, an entire lower margin of the opesia and much larger avicularia. This example shows how dangerous it is to judge the variability of species on the evidence of figures alone. Von Hagenow was quite correct in considering the two forms as only one species because they can often be observed in the same zoarium as confirmed by Voigt (1930 : 460). The first known species of Onychocella are from the Cenomanian, and they are small and of a lower level of evolution. Judging from this point of view it would be very odd if the oldest species should have the largest zooecia of the genus, more than i mm. long, and big avicularia 1-3-1 -4 mm. in length as these are otherwise developed only at the acme of the Onychocellids in the late Upper Cretaceous. 1 I have followed the advice of Dr. Hastings in using " opesia ", plural " opesiae ", rather than " opesium ", plural " opesia ". UPPER CRETACEOUS POLYZOA FROM COTENTIN 7 3. Castanopora jur assica (Gregory) 1894 (PI. 7, figs. 4-7) 1894 M embranipora jur assica Gregory: 62, text-fig, i. 1896 Membranipora jurassica Gregory: 212, text-fig. 21, p. 213. 1916 Rhiniopora jurassica (Gregory) Lang: 96. 1922 Rhiniopora scabra Lang: 196, pi. 4, fig. 7, text-fig. 62. 1922 Rhiniopora jurassica (Gregory) Lang: 197. 1962 Castanopora jurassica (Gregory) Larwood: 223, pi. 17, figs. 3-5, text-figs. 108-109. Holotype D . 180, large bilaminar fragment of damaged zoarium partly embedded in matrix. Upper Maastrichtian, Cotentin, Manche, France (not Maastricht, Netherlands) . Lang recognized the cribrimorph nature of this species. It was overlooked by Gregory that the cribrimorph structure of the frontal-shield was preserved quite well in some zooecia of the type specimen D.iSo (Tesson Coll.) which was figured by Larwood (1962 : 224, text-fig. 108, pi. 17, fig. 5). Although Gregory described it correctly as " erect foliaceous, bilaminate ", Lang defined this form as " encrusting unilaminar " as did Larwood (1962). Dr. Dighton Thomas and Dr. A. Hastings were kind enough to confirm my first observation made in the Museum collection that Gregory's type specimen is bilaminar and not encrusting. Gregory, misled by the erroneous data of Tesson's label, recorded the species as coming from the Bathonian of Calvados. Lang (1922 : 197) recognizing the Maas- trichtian age, supposed that it originated from the Dutch locality Maastricht itself, and recorded the distribution of this species as, " Senonian, Maastrichtian, Maastricht, Limburg, Holland ", specimen 0.3313 being labelled " Maastricht, Old collection ". Larwood (1962, pi. 17, fig. 3) followed Lang and gave a photograph of this specimen. I am much indebted to Dr. Dighton Thomas for lending me 0.3313, for there is no other record of Castanopora jurassica from Maastricht, and the specimen may have come from the same locality as D.iSo. Dr. Dighton Thomas compared the matrices of these two specimens: they are very similar in colour, grain size, and in size of fossil debris, and they could have come from the same locality in the Cotentin. As will be shown in section IV of this paper, the type locality cannot be Maastricht in Holland, but must be the same as that of the last species, a locality in the Cotentin. The matrix of the type specimen D . 180 is a hard, Polyzoan-bearing, detrital lime- stone, resembling very much the " Craie a Baculites " or " Craie a Thecidees " of the Cotentin. Lang has already recognized some other Cheilostomata in the matrix of this block. They are described and figured here under the name of Multicrescis laxata d'Orbigny, Rosseliana thomasi n. sp. and Pliophloea sp. These species are unknown from Maastricht itself, and combined with the results above on " Onycho- cella bathonica" , there can be no doubt that the two species both come from the Cotentin. All this is now confirmed by the fact that I have found four fragments of " Casta- nopora " jurassica (Gregory) in my material from Chef du Pont (Cotentin, Manche). 8 UPPER CRETACEOUS POLYZOA FROM COTENTIN Dr. Dighton Thomas kindly compared them with the holotype, and he has no doubt that they represent the same species. They agree in measurements, in number of costae, and all other characteristics. All are bilaminar with the exception of a young zoarium which is unilaminar and whose zooecia are only 0-09-1-00 mm. long. In the others the very large zooecia are about 1-2-1-4 mm. long. There are about twenty-six to thirty-four costae with eight lateral costal fusions as stated by Larwood. PL 7, fig. 5 shows an instructive view with some zooecia, one of which has preserved its undamaged cribrimorph front wall; the others are broken and show the " membranimorph " habit of Gregory's figure. On the reverse side of the specimen (PI. 7, fig. 4) are some zooecia with the characteristic ovicells of the genus which were previously not known in full preservation. They are hyper- stomial and globular, prominent, and overlapping the distal zooecium as presumed by Larwood. They are perforated by isolated fine pores (invisible in my figure) as in Castanopora guascoi Ubaghs from Maastricht. It must be emphasized that Castanopora jurassica (Gregory) has never been found near Maastricht, although I have studied the Polyzoa of the Maastrichtian Tuffkreide for 40 years. Near Maastricht another allied bilaminar form is repre- sented, Cnstanopora guascoi (Ubaghs) (1865 : 51, pi. 2, fig. 3), whose zooecia in some cases reach a length of 1-5 mm. but which has fewer costae (fifteen to twenty-two). Specimens like these were described by von Hagenow as Cellepora (Dermatopora) faujasi (von Hagenow : 1851, p. 99, pi. 10, fig. 19). It is impossible to mistake Castanopora guascoi for Castanopora jurassica (Gregory), but it is of interest that, if the costae are broken down, this species gives the appearance of Membranipora bipunctata (Goldfuss 1826 : 26, pi. 9, figs. 7a-b), traces of costae being no longer visible. The Cotentin fauna contains further allied bilaminar and unilaminar species which could be mistaken for Castanopora jurassica, but there is little room to discuss them. Rhiniopora scabra Lang 1916 from Riigen is, according to Larwood, a synonym of Castanopora jurassica Gregory, because it agrees in general with the number of costae, and in having 8 lateral costal fusions and pelmatidia. In a recent work (Voigt 1968 : 65) finished before the issue of the present paper, the genus Rhiniopora Lang 1916, united with Castanopora by Larwood (1962), is maintained at least as a subgenus of Castanopora. SPECIMENS : 0.3313. Large fragment partly embedded in matrix, recorded by Lang (1922 : 197) and figured by Larwood (1962, pi. 17, fig. 3), labelled " Maastricht, Old col- lection ", but certainly from a locality of the Cotentin-Maastrichtian, Manche, France. 0.49724. Small bilaminar fragment with three ovicelled zooecia. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Collection E. Voigt. 0.49725. Young unilaminar zoarium with some damaged zooecia. Horizon and locality as above. Collection E. Voigt. Some fragments. Horizon and locality as above. In Collection E. Voigt, Ham- burg, Nr. 3585 and 3924. UPPER CRETACEOUS POLYZOA FROM COTENTIX g III. POLYZOA FROM JURASSIC BEDS WHICH HAVE BEEN REFERRED TO THE CHEILOSTOM AT A It is now evident that the above mentioned Cheilostomata are not Jurassic. There are scattered records in the literature of Cheilostomatous Polyzoa whose Jurassic age is undoubted. In all such cases suspicion rises that they are not Cheilostomata. The fallibility of many of these records depends on the progress of science and today nobody can take seriously mention of Cheilostomatous genera like Hippothoa, Eschara, Cdlaria or Cellepora etc. from the Jurassic by early authors. However, Gregory (1894 : 61) in his note on the Jurassic Cheilostomata considered Eschara ranvilliana Michelin from the Bathonian of Ranville, and Cellar ia smithi Phillips from the Cornbrash of Scarborough, as true Cheilostomata and cited them in support of his theory of the existence of Jurassic Cheilostomata. In his Catalogue of the Jurassic Bryozoa (1896 : 56 and 127) Cellaria smithi Phillips is assigned to Stomatopora and Eschara ranvilliana Michelin to Diastopora with no mention that they were regarded as Cheilostomes by himself two years earlier. F. D. Longe (1881) was still convinced that Eschara ranvilliana Michelin from the Bathonian of Ranville belonged to the Cheilostomata when he wrote: " It is perfectly clear, however, that some of the Oolitic Escharoids themselves possess the charac- teristic cell features of the Cheilostomata in a marked degree; and their affinity to the Cheilostomatous Escharidae has been recognized by no less authorities than d'Orbigny and Michelin ". He gave a detailed analysis of the supposed Cheilostomatous features of this species and related forms, and tried to derive the Cheilostomata from certain Oolitic Diastoporids whose zooecia remind one, by their oval or polygonal shape, of the Cheilostomata. (" The decumbent cells in Diastopora may be regarded as ancestral Cheilostomatous cells, and Diastopora itself as the parent stock from which many, if not all, of the families of the Chalk and subsequent periods, grouped as Cheilo- stomata, have been derived ".). Walford (1894) published a note " On Cheilostomatous Bryozoa from the Middle Lias ". Under the new generic name Cisternifera he described some species of cyclo- stomatous Bryozoa with large heterozooecia so-called " cistern-cells " whose relationship with the ovicells of the Cyclostomata was already assumed by Walford. They were thought to bear, occasionally, minute avicularia on the upper lip of the zooecia. Apart from the question of whether these structures are avicularia or not, the different forms attributed to Cisternifera are true Cyclostomata, and Gregory (1896) himself has distributed them amongst the genera Diastopora and Entalophora. Cellepora davaiacensis Lissajous 1923 from the Bathonian of the Macon district (France), from whose generic name one would expect it to be a Cheilostomatous Polyzoan, and which Lissajous has compared with the Cheilostome Cellepora poly- thele Quenstedt, has been recognized by David (1952) as belonging to the Calcispongia, probably of the genus Synopelta Zittel. The list of literature on Jurassic Polyzoa published by David (1960) contains many references to records of Cheilostomata in other works. From these investigations it is now certain that all supposed species of Jurassic Cheilostomata prove to be mistaken identifications or erroneous stratigraphic records. 10 UPPER CRETACEOUS POLYZOA FROM COTENTIN IV. ON THE ORIGIN OF GREGORY'S " JURASSIC CHEILOSTOM AT A ' In the Cotentin (Manche) Upper Maastrichtian a lithology similar to the " Tuff- kreide " from Maastricht itself has long been known. The fades in the Cotentin is a " tuffaceous " detrital limestone with remains of many Foraminifera, Polyzoa, Echinoderms, Brachiopods (Craie a Thecidees ") or hard limestones (Craie a Bacu- lites). It is true that certain blocks of these rocks can be easily mistaken for genuine " Maastrichter Tuffkreide ", because many of the small fossils which make up the rock are common to both strata. An important difference is the absence of any larger foraminifera like Orbitoides, Lepidorbitoides , Siderolites or Omphalocyclus etc. which are distinctive of the Upper Maastrichtian in Holland and Belgium. Hofker (1959) in his monograph on the Foraminifera of the Cotentin Maastrichtian has shown that these beds must be intermediate in age between horizons Cr 4 and Mb in the terminology of Uhlenbroeck for the Upper Cretaceous in South Limburg. This would indicate a stratigraphic position between the phosphatic chalk of Ciply and the base of the Tuffaceous chalk of St. Symphorien in the Mons basin, or an equivalent of the higher beds of Folx-les- Caves and Orp-le-Petit in northern Belgium. This could explain the lack of larger Foraminifera in this region which did not invade the northern regions before the higher horizons of the Maastrichtian. The Maastrichtian age is based upon the occurrence of Scaphites constrictus J. Sowerby in the area of Valogne. For the other fauna see Vieillard & Dollfus 1875. In 1957 I visited the Cotentin region in order to study the classic Senonian localities of d'Orbigny (1850-54) who had described from here, chiefly from Nehou and from Sainte Colombe, some hundred species of Polyzoa. Although there were no extant exposures in the immediate neighbourhood of these villages, some larger quarries near Fresville and Port Filiolet and an excavation for the dairy of Chef du Pont supplied material which furnished examples of a considerable part of the Maas- trichtian Polyzoa which were described and figured by d'Orbigny. (For the site Valognes Nthou zter Fresville * ^& S* Colombe Rauvilte - la Place *j Rue-Tourville 24 6 d 10 ) Chef du Pont Port Filiolet km FIG. i. Map of the outcrops (black) and the localities of the Maastrichtian in the Cotentin area (Normandy). (After Carte geologique detaillee de la France (i : 80000) Feuille 28, St. L6 (2 erne Ed.) 1926. UPPER CRETACEOUS POLYZOA FROM COTENTIN n of these localities see Text-fig, i.) This material was augmented by some samples given by Dr. F. Schmid (Hannover) and Polyzoa from Fresville presented by Dr. P. Marie (Paris). It is interesting that many different and new forms not mentioned by d'Orbigny were found at these localities, and by contrast, a large number of d'Orbigny's species could not be found again. Perhaps there are some horizons with a different fauna, and the beds of Sainte Colombe and Nehou may belong to such levels, whose exposures are abandoned today, or the different localities represent heterogeneous ecologic biotopes. The Polyzoan fauna of this region is the most important one in the northern area and its knowledge is fundamental to our knowledge of the Maastrichtian Polyzoa. With the exception of the inadequate revision of the Cretaceous Polyzoa of d'Orbigny as a whole, which was undertaken by Pergens (1889) and by Canu (1900), nobody has concerned himself with the Polyzoan fauna of the Cotentin for no years. An up-to-date revision of this fauna is therefore an urgent task, but it cannot be under- taken without a re-study of d'Orbigny's types which are preserved in the Musee d'Histoire naturelle in Paris. The most obvious obstacle for a successful revision is the fact that many of d'Orbigny's type specimens are missing, or, if specimens are present, it is often impossible to say with certainty whether they are true " types " or not. On the other hand many of the drawings do not agree well with the originals ; they may be stylized, restored or improved and consequently it is often rather difficult to give a satisfactory identification of d'Orbigny's species (see Canu 1900 : 335). His omission to specify the exact locality of the figured specimen in cases when there is more than one such locality, increases the difficulty of identifying the types. There are about 300 photographs of d'Orbigny's polyzoan types of the Paleontologie Fran9aise in the author's collection, and this stock, together with a rich collection of French Cretaceous Polyzoa, are the basis for the following discussion. The proofs for a Cotentin origin for Gregory's " Jurassic Cheilostomata " are: 1. The matrix of sediment adhering to Gregory's originals is identical to that of the above-listed Maastrichtian localities of Cotentin. The matrix detached from block D.i8i with Onychocella piriformis Goldfuss has yielded two specimens (66.42981-82) of the characteristic fossil Thecidea papillata V. Schlottheim (pi. i, figs. 9-10) and it is evident that it has been collected from the so-called " Craie a Thecidees " of the Cotentin. 2. The objection that this brachiopod is also common at Maastricht and in Belgium in a similar facies is weakened by the fact that the accompanying fauna in this block, consisting of thirty-one species of Polyzoa, contains some species which are confined to the Cotentin fauna and have never been found in Holland and Belgium, although indeed both strata have many species in common. Examples of species not known from these countries are: Multicrescis laxata d'Orbigny, Membranipora unipora Marsson, Stamenocella cf. marginata (d'Orbigny), Castanopora jurassica (Gregory), Onychocella bellona (d'Orbigny), Onychocella cf. cepha d'Orbigny, Onycho- cellaria caecilia sp., Semiescharinella complanata d'Orbigny, Lunulites sp., Rosseliana thomasi n. sp., Frurionella fertilis n. sp. and others. 12 UPPER CRETACEOUS POLYZOA FROM COTENTIN 3. A Cotentin origin is further suggested by another Cretaceous Polyzoan in the museum collection which is embedded in a typical sample of " Craie a Thecidees ". It is a large specimen of a Reticrisina, determined as the Jurassic " Reticulipora dianthus (Blainville) " (6.4569 Old collection) and labelled " Bathonian Fresville ". This example gives a good idea of how the error could have arisen : Bathonian rocks being absent near Fresville, Cretaceous Polyzoa were mistaken for Bathonian fossils. Gregory's " Jurassic Cheilostomata " both belong to the Tesson-Collection, which according to Gregory's Catalogue contains many Bathonian Polyzoa from the famous locality of Ranville (Calvados). The two localities are about 100 km. apart, and it is probable that the labels were confused or that the Cretaceous Polyzoan limestone was mistaken for the not dissimilar Coral and Polyzoan-limestone (" Calcaire a polypiers ") of the Bathonian. The Tesson-Collection to which Gregory's types belong was acquired in 1857 (Gregory 1896 : 35) and the fossils were probably collected many years ago at a time when geological mapping of this region was still in its infancy. It is not impossible that Gregory's types both came from Fresville like the Reti- crisina mentioned above. The rock matrix of D.iSi is very like that of the " Craie a Thecidees " of Fresville, but as Dr. Dighton Thomas informed me, that from Chef du Pont is even closer in appearance. But such lithological differences may occur in the same sequence, and are of no great importance. At least one other possible explanation of this error should be mentioned. Between the Maastrichtian outcrops of Sainte Colombe and Rue Tourville is the village Rauville-la-Place (see Text-fig, i). Possibly a label " Rauville " was misunderstood for " Ranville ", Calvados. It must be reckoned that errors like these are more common in old collections, and it may be recalled that a considerable number of Reuss' " Cenomanian Polyzoa of Saxony" came from the Vincentown Limesand in New Jersey (U.S.A.), of Dano- Paleocene age (Voigt 1942). Another mistake is the supposedly new " Polyphyma bulbosa " Hamm (1881) from Maastricht which is a Hauterivian form from northern Germany, and which was described in 1839 by Roemer as Alveolites heteropora (Voigt 1953 : 57). Certain suspicious anachronisms of phylogenetic level may be explained in this manner as has been shown by the present author with some of Reuss' types. 4. I have collected much material of Bathonian Polyzoa at the classic locality of Ranville (Calvados), but I have never found any trace of Cheilostomata. The objection, that the Cheilostomata may originate in Jurassic times cannot be con- tested; but we have no remains of them earlier than the Lower Cretaceous. It is true that in the last decade several important groups of fossils have been proved to be older than was formerly assumed. Today we know of Cambrian bivalves, Carboniferous belemnites, Triassic frogs, and these examples could be augmented. It is possible that Jurassic Cheilostomata may be found one day. But in such a case we should expect primitive forms like Membrammorphs as are found in the lower Cretaceous and not highly developed forms like Onychocella, or highly special- ized forms like the Cribrimorph Castanopora. Lower Cretaceous Cheilostomata are of the greatest phylogenetic interest, and all UPPER CRETACEOUS POLYZOA FROM COTENTIN 13 forms described to date are rare. Most of them are Albian and the oldest is recorded from the Neocomian. Their number is so small that they can be listed in a few lines : 1. Membranipora neocomiensis d'Orbigny 1853, Neocomian, Saint-Sauveur, Yonne, France. The only specimen was too bad to be figured by d'Orbigny. It must be regarded as a nomen nudum. 2. "Membranipora" constricta d'Orbigny 1853, Aptian, Les Croutes (Yonne). Remarks: I have photographed the supposed " type " of this species (pi. 7, fig. 8). It is without doubt a " Membranipora " s.L, but it does not cor- respond with the original figure. 3. Rhammatopora (?) johnstoniana Mantell 1844, Aptian, Lower Greensand, Shanklin-Sand, Kent. 4. Rhammatopora gaultina Vine 1890, with his synonyms Rhammatopora vinei Lang and Rhammatopora pembrokiae Lang (c.f. Thomas & Larwood 1960) Albian-Cenomanian, England. 5. Charixa vennensis Lang 1915, Albian, Dorset. 6. Pyripora texana Thomas & Larwood 1956, Albian, Texas (U.S.A.). 7. Wilbertopora mutabilis Cheetham 1954, Albian, Texas (U.S.A.). Vine (1890) mentioned Membranipora fragilis d'Orbigny, Membranipora ? obliqua d'Orbigny, Membranipora elliptica v. Hagenow and Hippothoa simplex d'Orbigny from the Red Chalk of Hunstanton. The identification of these forms must be revised, but nevertheless all the named species are primitive encrusting membrani- morphs of the Division Malacostega, suborder Anasca, and half of them are mono- or oligoserial, and lack either ovicells or avicularia as we must theoretically presume if our ideas about the evolution of the early Cheilostomata are correct. They are followed in the Cenomanian by the first primitive Cribrimorphs and Coilostega with the families of Onychocellidae and Microporidae. Therefore it is hard to under- stand why Canu & Bassler (1920 : 318) stated that following the Membraniporae, the Acroporidae Canu 1913 (= Porinidae d'Orbigny 1852) of the suborder Asco- phora are the most ancient Cheilostome fossils. The main evolution of this family is in the later Upper Cretaceous, and the only described species of Porina from the Cenomanian is P. cenomana Lecointre (1912) whose origin and inner structure needs revising. V. PALAEOZOIC CHEILOSTOMATA? In this connection the question of the systematic position of the north American families Worthenoporidae Ulrich 1893 (Carboniferous) and Palescharidae Miller 1889 (Ordovician-Devonian) cannot be neglected. These were established for the single genera Worthenopora Ulrich 1889 and Paleschara Hall 1874 which resemble Cheilo- stome Polyzoa in some features. Ulrich (1890) stated that the affinities of that genus are nearer to the Membraniporidae than any other and that his present views would admit it being placed in the Cheilostomata. Nickles & Bassler (1900) regarded both families without any restriction as Cheilostomata. This classification was followed by many authors, but in recent times this opinion seems to have been I 4 UPPER CRETACEOUS POLYZOA FROM COTENTIN abandoned. Bassler (1953) has placed them among the Cryptostomata, but con- siders Worthenopora " may belong among cheilostomes ". Worthenopora, indeed, has the outward aspect of a Cheilostome with its triangular or semielliptical apertures, with posterior raised margin and spine bases. Paleschara with its simple short polygonal zooecia is like a Membranipora which possesses completely opened opesiae without any trace of a gymnocyst. Dr. Dighton Thomas has kindly lent me some specimens of both genera from the museum collection. Although there is no space here to go into details, further studies of this problem are intended. I have got the impression that they cannot be attri- buted to the Cheilostomata. Paleschara must be regarded as a very primitive Cryptostome, and Worthenopora, in contrast, as a specialized one. The latter has reached a level of evolution which reminds one of certain Cheilostomata. It has not been found in beds younger than Mississippian, and there are no intermediate forms between it and the Cretaceous Cheilostomata. Consequently it seems to be impossible to regard it as an ancestor of the true earliest membranimorph Cheilo- stomata of the Lower Cretaceous. Paleschara, with its network of rather simple polygonal zooecia, may perhaps be primitive enough to give rise to Cheilostomata-like forms. But it could be better regarded as a cryptostomatous form corresponding to a membranimorph level of primitive Cheilostomata. We have no palaeontological evidence for a descent of the Cheilostomata from the Cryptostomata or from the Ctenostomata. As Borg (1930 : 54) and Cori (1941) regard the Ctenostomata as emanating from primitive Cheilostomata, it must be emphasized that the Ctenostomata are the older group, represented by fossils from the Ordovician onwards and that there is no possibility of deriving the Ctenostomata from the Cheilostomata. Silen (1942) regards the Cheilostomata and the Cteno- stomata as closely allied (" Cheilo-Ctenostomata ") and believes that both have common ancestors. Silen has established an interesting theory about the origin of the Cheilostomata from hypothetical primitive forms like the recent Labiostomella which he calls Protocheilostomata. These have erect zoaria with frontal budding and other primitive features, and it would mean that the encrusting growth of many Cheilo- stomata and the lateral budding as existing in all other Cheilostomata, are secondary. It should be emphasized that the oldest known Cheilostomata from the Lower Cretaceous do not show this primitive character. All are encrusting, and they must already have passed the evolutionary level of the " Protocheilostomata ", which according to Silen were feebly or not at all calcified. If this is admitted, there remains no possibility of regarding Worthenopora or Paleschara as early Cheilo- stomata of the Palaeozoic. VI. THE ACCOMPANYING FAUNA OF THE MATRIX OF GREGORY'S " JURASSIC CHEILOSTOMATA " It is satisfactory that in the matrix of Gregory's " Jurassic Cheilostomata " are enclosed other Bryozoa which were overlooked by Gregory. In block D.iSo, containing Castanopora jurassica, Multicrescis laxata d'Orbigny, Rosseliana thomasi n. sp. and Pliophoea sp. could also be observed. UPPER CRETACEOUS POLYZOA FROM COTENTIN 15 I am much indebted to Dr. Dighton Thomas for his permission to remove about 2 cm. 3 of the matrix of block D.iSi with Onychocella bathonica. This piece was cracked under a press and the washing of the residue has yielded a fauna of thirty- three species of Polyzoa and one Brachiopod. This fauna is described in detail in the following part of this paper. Most of the treated specimens are figured although the state of preservation is rather poor because the hard rock is unfavourable for clean preparation. All specimens are somewhat damaged, or rolled and, unfortu- nately, recrystallized or covered with minute calcite crystals as is often observed in the Cotentin material, and it is nearly impossible to stain them with colour. This explains any mediocre photos. In many cases figures of comparable specimens of the same species from the Cotentin are given for comparison. If they do not always present exactly the same picture as the specimens from the original French localities, this is because there often exists a great variability in size or growth-stage, and in preservation, and that the identification is based upon a vast quantity of material. It is surprising that in only 2 cm. 3 of rock from D . 181 some small fragments of new species were found, although these have long been known to the author from the Maastrichtian of the Cotentin. The opportunity is taken here to describe them. But it should not be forgotten that this small fauna from only a few cm. 3 matrix must represent a very small part of the rich Polyzoan fauna of the Cotentin Maas- trichtian, and that its composition is purely accidental. The list given in table i, p. 41, contains thirty-three Polyzoan species of which twenty-four are Cheilostomata. It is not complete because in some cases a com- plete identification could not be made, and for a few forms no identification was possible. Excepting Pliophloea sp. all forms have been found by the author in the Maastrichtian of the Cotentin. Twenty-six species were found at Chef du Pont, eighteen at Port Filiolet and nine at Fresville. But it should be taken into account that the investigated material is too poor for important deductions. Many species very common at all Cotentin localities are not represented, and others are new for the Cotentin. The affinities with the classic locality of Maastricht are proved by fourteen species but this figure may be too large because several forms of both regions may be identical. Von Hagenow's types from Maastricht having been lost during the second world war, it is not yet possible to decide this question with certainty. I. POLYZOA A. CYCLOSTOMATA Genus BERENICEA Lamouroux 1821 i. Berenicea sp. One small incomplete zoarium which is indeterminable, encrusts Reteporidea lichenoides Goldfuss. SPECIMENS. A minute incomplete zoarium encrusting Reteporidea lichenoides Goldfuss (=0.49569). 16 UPPER CRETACEOUS POLYZOA FROM COTENTIN Genus ENTALOPHORA Lamouroux 1821 2. Entalophora benedeniana (von Hagenow) 1851 Pustulopora benedeniana von Hagenow: 17, pi. i, fig. 6. 1899 Entalophora madreporacea Goldfuss var. benedeni von Hagenow; Gregory: 239. 1964 Entalophora benedeniana (von Hagenow) Voigt: 422, pi. i, figs. 1-7. One small fragment embedded in matrix has been referred to this well-known species from Maastricht which is represented in the author's collection from the Maastrichtian of Chef du Pont (Manche). For morphological details and ovicells see Voigt (1964). SPECIMENS. 0.49560. One poorly preserved fragment in matrix from D.iSi. Upper Maastrichtian, Cotentin, Manche, France. Genus IDMIDRONEA Canu & Bassler 1920 3. Idmidronea macitenta (von Hagenow) 1851 Idmonea macilenta von Hagenow: 29, pi. 2, fig. 4. ?i85i Idmonea ramosa d'Orbigny: 736, pi. 6n, figs. 11-15. 1899 Retecava ramosa (d'Orbigny) Gregory: 192, pro parte. 1951 Idmidronea macilenta Voigt: 38, pi. 4, figs. 14-17. One small distal fragment of this abundant species of the Upper Maastrichtian belongs to /. macilenta. These thin distal branches do not have the numerous firmatopores which are well developed on the reverse side of the broader and older stems. Therefore they are very similar to Idmonea (Tubigera) antiqua Defrance (d'Orbigny 1853 : 722, pi. 613, figs. 11-15, figured under the name of Stichopora regularis d'Orbigny). Idmonea disticha Goldfuss (sensu von Hagenow 1851 : 30, pi. 2, fig. 8) may be mistaken for this form also, but always it is not so flat as the distal ends of Idmidronea macilenta. The species cited by d'Orbigny from Sainte- Colombe under the latter name may be Idmonea macilenta. If this species is con-specific with Idmonea ramosa d'Orbigny, the latter name has priority and must be preferred. It is represented at Fresville (Cotentin, Manche). STRATIGRAPHICAL RANGE. Upper Maastrichtian. SPECIMENS. 0.49561. A worn fragment from matrix of D.iSr. Upper Maastrichtian, Cotentin, Manche, France. 0.49843. A fragment, upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. E. Voigt Collection. Genus HETEROCRISINA Gabb & Horn 1860, em. Voigt 1964 4. Heterocrisina communis (d'Orbigny) 1853 Idmonea communis d'Orbigny: 745, pi. 750, figs. 6 10. 1887 Idmonea pseudodisticha (non von Hagenow) Marsson: 28, pi. 2, fig. 8. 1860 Heterocrisina abbottii Gabb & Horn: 404, pi. 69, figs. 45-47. 1899 Retecava abbottii (Gabb & Horn) Gregory: 205. non 1907 Idmonea abbottii (Gabb & Horn) Ulrich & Bassler: 321, pi. 22, figs. 3-4. 1964 Heterocrisina communis (d'Orbigny) Voigt: 432 pi. 3, figs. i-io. UPPER CRETACEOUS POLYZOA FROM COTENTIN 17 Two small worn fragments belong to this species, which is very distinctive because of its large frontal ovicell, although it has often been mistaken for other species. For further information about its generic position and morphological details see Voigt (1964). This form is common at all Maastrichtian localities in the Cotentin (Fresville, Port Filiolet and Chef du Pont). STRATIGRAPHICAL RANGE. Campanian-Maastrichtian. SPECIMENS. D . 49562-63. From matrix of D . 181. Upper Maastrichtian, Coten- tin, Manche, France. Genus CRISISINA d'Orbigny 1847 5. Crisisina carinata (Roemer) (PL i, figs. 4-5) 1840 Idmonea carinata Roemer: 21, pi. 5, fig. 20. 1964 Crisisina carinata (Roemer) Voigt: 429, pi. 4, figs. 1-7. (See full references.) Three worn fragments belong to this very common and widespread species of the Upper Cretaceous. The synonymy is very confused see Voigt (1964) where all known synonymies are given and the generic classification is discussed. Very abundant at all Maastrichtian Cotentin localities. STRATIGRAPHICAL RANGE. Cenomanian-Paleocene. SPECIMENS. 0.49564-66. From matrix of D.iSr. Upper Maastrichtian, Coten- tin, France. 0.49567. A fragment for comparison with 0.49566. Upper Maastrichtian, Port Filiolet, Cotentin, Manche, France, Voigt Collection. Genus OSCULIPORA d'Orbigny 1847 6. Osculipora truncata (Goldfuss) 1826 Retepora truncata Goldfuss: 28, pi. 9, fig. 14. 1851 Truncatula truncata (Goldfuss) v. Hagenow : 35, pi. 3, fig. 2. 1851 Truncatula tetrasticha von Hagenow: 34, pi. 3, fig. 3. 1909 Osculipora truncata (Goldfuss) Gregory: 58. 1922 Osculipora truncata (Goldfuss) Canu & Bassler: 57, pi. 23, figs. 1-6. One small worn fragment may be referred to 0. truncata; it is in the condition of Truncatula tetrasticha von Hagenow, which represents highly worn branches of 0. truncata (Goldfuss.) Further investigations are necessary to check whether the older citations of this form from the Cenomanian by Reuss (1872) are correct. In the author's collection from Chef du Pont, Port Filiolet and Fresville are numerous fragments of this species, which was not recorded by d'Orbigny from any of his Cotentin localities. STRATIGRAPHICAL RANGE. Cenomanian (?) to Maastrichtian. SPECIMENS. 0.49568. A very poorly preserved worn fragment from the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. GEOL. 17, I. 2 i8 UPPER CRETACEOUS POLYZOA FROM COTENTIN Genus RETEPORIDEA d'Orbigny 1849 7. Reteporidea lichenoides (Goldfuss) 1826 Retepora lichenoides Goldfuss: 29, pi. 9, figs, isa-b. 1851 Idmonea lichenoides (Goldfuss) von Hagenow: 28, pi. 2, fig. 6. 1899 Retecava lichenoides (Goldfuss) Gregory: 194, fig. 16, p. 195. One very poor fragment which is encrusted by a small young Berenicea, has been recognized as this species which is very abundant in the Upper Maastrichtian. It is represented in the author's collection from Port Filiolet by some specimens. It has not previously been recorded from the Cotentin Maastrichtian. STRATIGRAPHICAL RANGE. Maastrichtian. SPECIMENS. 0.49569. A small worn fragment with an encrusting Berenicea from the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. Genus PETALOPORA Lonsdale 1850 8. Petalopora sp. (PI. i, figs. 1-3) There are a few fragments of a badly preserved Petalopora which are conspecific with similar specimens from Chef du Pont (pi. i, fig. 2) . Their identification involves some difficulties because it is impossible to identify them from published figures. They are allied to Heteropora reiiculata Marsson (1887 : 26, pi. 2, fig. 4) in the size of the branches and diameter of the apertures (about 0-14-0-15 mm.), but the mesopores are much less conspicuous, and on the figured fragment (pi. i, fig. i) they look a little like longitudinally-oriented lines which are straight or sinuously bent and interspersed between fine ribs. On the specimen from Chef du Pont the mesopores are larger and they show a more longitudinally-oriented inconspicuous network (pi. i, fig. 3). I have no doubt that these three specimens are conspecific in spite of this difference but I dare not identify them with any known species. SPECIMENS. 0.49570-72. Three small worn fragments from the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. 0.49573. A larger fragment from the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. 0.49574. A larger fragment. Upper Maastrichtian, Port Filiolet, Cotentin, Manche, France. Voigt Collection. One branched fragment. Upper Maastrich- tian, Chef du Pont, Cotentin, Manche, France. Voigt Collection, Hamburg, Nr. 3969. Genus MULTICRESCIS d'Orbigny 1854 9. Multicrescis laxata d'Orbigny (PL i, figs. 6-8) 1854 Multicrescis laxata d'Orbigny: 1077, pi. 800, figs. 10-11. HOLOTYPE. Upper Maastrichtian, Sainte Colombe, Cotentin, Manche, France. d'Orbigny Collection, Paris, Musee d'Histoire Naturelle Nr. 8416. UPPER CRETACEOUS POLYZOA FROM COTENTIN 19 There is a fragment of a " heteroporid " Polyzoan with an encrusting colony of Rosseliana thomasi n. sp. embedded in the matrix of Castanopora jurassica Gregory (D.i8i). It cannot be distinguished from d'Orbigny's type specimen of Multi- crescis laxata, whose apertures and the mesopores are exactly the same. Unfortu- nately there are no other specimens of this species in d'Orbigny's collection which would allow confirmation of the inner structures by sections. Specimens from Chef du Pont (Manche), which seem to be conspecific with d'Orbigny's species have a median lamella like Grammascosoecia Canu & Bassler 1922. Pergens (1889 : 373) and Canu & Bassler (1922 : 119) have included this species in the synonymy of Grammascosoecia dichotoma (Goldfuss) from Maastricht (see von Hagenow 1851 : 47, pi. 5, fig. 15). This might be correct; some specimens from Fresville can hardly be distinguished from the Maastricht species. On the other hand there are some fifty fragments from Fresville, and none show the characteristic pattern of small regular smooth quadrangles which grow from the calcified mesopores (cf. von Hagenow's fig. I5i and Voigt 1951 pi. 4, fig. i) and which can be observed in the majority of the Maastricht specimens. Therefore I still hesitate to unite it with Gramma- scosoecia dichotoma (Goldfuss). D'Orbigny attributed his species to his genus Multicrescis which is multilamellar. But this is not the case in M. laxata, although d'Orbigny noted two layers in his type-specimen from Sainte-Colombe, which is a basal fragment (pi. I, figs. 7-8). It is well known that the basal stems often develop more than one layer of zooecia, and therefore there is no reason to place this form in the genus Multicrescis. Many specimens from Fresville have radiating rows of peristomes as in Multicavea, which can also be observed in Grammascosoecia dichotoma (Goldfuss). But the ovicell of the Cotentin form has not yet been discovered. Therefore I prefer to leave this form provisionally under the name given by d'Orbigny. It should be noted that the median lamella in Grammascosoecia dichotoma (Gold- fuss) is not constant. It is lacking in many specimens from Maastricht, and there is no reason to assume a different species. STRATIGRAPHICAL RANGE. Upper Maastrichtian. SPECIMENS. 0.49575. A worn fragment with encrusting Rosseliana thomasi n. sp. embedded in matrix of Castanopora jurassica (Gregory) (D.iSo). Upper Maastrichtian, Cotentin, Manche, France. Labelled erroneously by Gregory as " Bathonian, Ranville, Calvados ". B. CHEILOSTOMATA MEMBRANIPORA Blainville 1830 (sensu lato) 10. " Membranipora " unipora (Marsson) (PL 2, figs. 7-8) 1852 Flustrella simplex d'Orbigny: 293, pi. 699, figs. 14-16. 1887 Biflustra unipora Marsson: 52. 1929 Membranipora genucia Brydone: 37, pi. 13, figs. 10-11. 20 UPPER CRETACEOUS POLYZOA FROM COTENTIN 1930 Membranipora unipora (Marsson) : Voigt 420, pi. 4, fig. 5. 1925 Membranipora unipora (Marsson): Levinsen 329, pi. 2, fig. 22. HOLOTYPE. Upper Maastrichtian, Nehou, Cotentin, Manche, France. d'Orbigny Collection, Paris. Musee d'Histoire Naturelle Nr. 8130. One unilaminar specimen embedded in matrix, showing some zooecia with the characteristic avicularium on the gymnocyst below the opesia and the narrow helmet-shaped hyperstomial ovicell, agrees very well with d'Orbigny's type from Nehou (pi. 2, fig. 8) and another well preserved specimen from the Maastrichtian of Port Filiolet (Manche). The zoarial length is from 0-80-1-20 mm. The spines of the well raised margin of the opesia, which is 0-45-0-50 mm. long, number about twenty, but they are inconspicuous and often hidden by recrystallization of calcite. The zooecia and the opesiae of my specimen of this species from Port Filiolet are a little larger than those of the Museum-specimen but this lies within the range of variation of this form. Membranipora genucia, described by Brydone from the upper Campanian of Meudon near Paris, is a synonym as shown by comparison with Cotentin and Rugen specimens with those from Meudon. Marsson who translated this species from Flustrella to Biflustra, has changed the species name simplex to unipora because there existed already a recent Biflustra simplex d'Orbigny 1839. Otherwise another Cretaceous Membranipora simplex d'Orbigny exists too. It is clear that the placing of this species in " Membranipora " is only provisional; a revision of the Cretaceous Membraniporae might put it in another genus. STRATIGRAPHICAL RANGE. Upper Campanian-Maastrichtian. SPECIMENS. 0.49576. Fragment with ovicelled zooecia from the matrix of D.iSr. Upper Maastrichtian, Cotentin, Manche, France. Genus AMPHIBLESTRELLA Prud'homme 1960 ii. Amphiblestrella elegans (von Hagenow) (PL 4, figs. 1-3) 1851 Siphonella elegans von Hagenow: 84, pi. 6, fig. 7. 1851 Flustrella baculina d'Orbigny: 291, pi. 699, figs. 4-6. 1930 Amphiblestrum elegans (von Hagenow) Voigt: 448, pi. 13, figs. 13-16. 1960 Amphiblestrella elegans (von Hagenow): Prud'homme: 949. 1962 Amphiblestrum elegans (von Hagenow) Berthelsen: 100, pi. 9, figs. 1-5. There is only one small fragment which is 1-7 mm. long and shows eight rows of zooecia; normally there are ten to sixteen. The zoarial dimensions are smaller than those of specimens from the type-locality of Maastricht, and from the Cotentin, where this species is very common. The axial canal of the hollow zoaria, clearly visible in the thicker zoaria, is much reduced in the slender branches, and may disap- pear almost completely as shown in d'Orbigny's type specimen of his Flustrella baculina (pi. 4, fig. 2) from Nehou. The zooecia from Danian material are mostly longer than those from the Maastrichtian (about 0-7 mm. instead of 0-6 mm.). The size of the opesiae is very variable in this species. Canu (1900) in his revision of UPPER CRETACEOUS POLYZOA FROM COTENTIN 21 d'Orbigny (1851-54) has incorrectly regarded this species as a synonym of Flustrella irregularis d'Orbigny. In the Cotentin localities this species is represented from Fresville and Chef du Pont and by d'Orbigny's type specimen of Flustrella baculina from Nehou. STRATI GRAPHICAL RANGE. Maastrichtian-Danian. SPECIMENS. D. 49577. A worn fragment from matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. D. 49578. A well preserved branched fragment. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection. Type-specimen of Flustrella baculina d'Orbigny. Upper Maastrichtian Nehou, Cotentin, Manche, France. In d'Orbigny Collection, Paris, Musee d'Histoire Naturelle Nr. 8127. Genus RADULOPORA nov. DERIVATIO NOMINIS. Derived from the species-name of Biflustra radula Marsson 1887. DIAGNOSIS. Zoarium bilaminar, dichotomously branched and probably radicelled at the base. Zooecia dimorphic, the marginal zooecia of the acute edges of the branches being larger, and having larger opesiae than the normal zooecia. Crypto- cyst well developed, finely granulated; opesiae small with straight proximal rim and occasionally developed lip. Distal interzooecial asymmetrical vibracula above the opesiae with long elliptical opening and a small thornlike process going out from the left or right inner margin. Ovicells inconspicuous exteriorly, deeply immersed, endozooecial. Type species: Biflustra radula Marsson 1887, Lower Maastrichtian Riigen (Germany) . REMARKS. This new genus comprises three characteristic species in the Maastrich- tian, which cannot be attributed to any other genus hitherto known. Biflustra radula Marsson was assigned incorrectly to Amphiblestrum by Voigt (1930). It differs from Amphiblestrum not only in its bilaminar and apparently radicelled zoarium with dimorphic zooecia and semicircular opesia in which a proximal lip is originally developed, but also by its avicularian-like vibraculum. This has no pivot for the articulation of the mandibula as in true avicularia, but asymmetrical curved thorn-like processes on the inner left or right edge of the vibracula (Text- fig. 2) and an ellipsoidal cavity in the proximal part adapted for the motion of the seta of the vibraculum. These structures are very well shown in the type-species, R. radula (Marsson) (pi. 3, figs. 11-12), while they are indicated in R. minor only in some better preserved specimens. Owing to the loss of the Marsson Collection during the last war, a neotype for R. radula Marsson should be erected. The specimen of Biflustra radula figured here could represent a good neotype, but it was collected from a chalk block in glacial drift and therefore a topotype from the Lower Maastrichtian of Riigen would be preferred. It is difficult to decide if this genus should be regarded as belonging to the Mala- 22 UPPER CRETACEOUS POLYZOA FROM COTENTIN costega or Coilostega. The cryptocyst is strongly calcified and the opesia is so small that the Malacostega level has probably been exceeded. Opesiules are absent as in many Coilostega. Nevertheless Marsson described it as Biflustra, and perhaps he was right to assign it to the Membranimorphs, although there are many inter- mediate forms between these two groups. This genus seems to be rather isolated and it is difficult to attribute it to one particular family of the Malacostega. FIG. 2. Radulopora radula (Marsson). Some zooecia and vibracularia. x6o. 12. Radulopora minor sp. n. (PI. 3, figs. 6-io) Holotype Upper Maastrichtian, St. Pietersberg near Maastricht (Netherlands). 0.49844. Voigt Collection. DERIVATIO NOMINIS. The name refers to the smaller zooecial dimensions in contrast to the type-species R. radula (Marsson) (pi. 3, figs. 11-12). DIAGNOSIS. Radulopora with slender dichotomous branches 0-7-1 -2 mm. wide, consisting of about three to seven alternating transverse rows of zooecia about 0-5-0 -6 mm. long. The edges of the zoarium are formed by the larger marginal zooecia: other zooecia not clearly distinct at their margins, elongate and pyriform with distal raised margin and a well developed cryptocyst deeply immersed proxi- mally. Opesiae rounded quadrangular or high-semicircular rounded distally and truncated proximally, showing in some specimens a well developed proximal lip. Distal vibracula small, the peak turned obliquely downwards but symmetrically oriented outwards from the median line toward the edges of the branches. Ovicells form inconspicuous slight swellings above the opesiae (pi. 3, fig. 9). DESCRIPTION. The bilaminar fronds have narrow cylindrical or prismatic basal rods which are pointed toward their proximal ends and suggest an articulated basal attachment of the zoarium. The zoarial and opesial dimensions are smaller in these proximal parts of the zoarium than in the more distal branches. The shape of the opesia varies from nearly semi-circular to high-oval, trapezoidal or oval: if it is oval it is narrowed proximally but it is never circular. The straight proximal edge of the opesia is deeply immersed in old zooecia and may disappear. The distinct UPPER CRETACEOUS POLYZOA FROM COTENTIN 23 proximal lip observed regularly in some specimens has its origin from this straight proximal edge. The opesia occupies about one quarter to one fifth of the length of the zooecium. The cryptocyst is highly calcified and often appears in the proximal part. The marginal zooecia appears to be larger than they actually are because they are not narrowed in their proximal part, and the opesiae are always distinctly larger than in the other zooecia, although there may be transitions between the size of the opesiae of the normal and those of the marginal zooecia. The ovicelled zooecia never have vibracula. The ovicells are very flat swellings above the opesiae. If their roof is broken away a large deep hollow is revealed showing that the ovicell is deeply immersed although it must be regarded as endozooecial. A minute fragment of eight zooecia belongs to this species which corresponds very well with the abundant material of the author's collection from the Cotentin localities and from Maastricht. MEASUREMENTS. Lz (= Distance between the opesia) : 0-50-0-67 mm. ho = 0-10-0-12 mm. ""I > normal zooecia. lo = 0-075-0-12 mm. J ho = 0-14-0-17 mm. ~\ . , , > marginal zooecia. lo = 0-15-0-16 mm. / STRATIGRAPHICAL RANGE. Upper Maastrichtian. SPECIMENS. 0.49579. A small worn fragment. From matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. 0.49580-81. Two fragments. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection. 0.49582. Fragment with ovicelled zooecia. Upper Maastrichtian, Cotentin, Manche, France. Voigt Collection. Genus HAGENOWINELLA Canu 1900 13. Hagenowinella cf. incrassata (d'Orbigny) (PI. 2, figs. 1-3) 1853 Flustrellaria incrassata d'Orbigny: 527, pi. 726, figs. 5-8. A large unilaminar fragment consisting of more than twenty zooecia may belong to this species, although the opesiae are oval and not so broad and truncated as in d'Orbigny 's type-specimen from Sainte-Colombe (Manche). A fragment of this form collected by the author near Chef du Pont (Manche), shows, however, that the shape of the opesiae can vary considerably and that in certain parts of the zoarium the opesiae are fairly oval as in 0.49583, in which two zooecia have the cryptocyst broken giving the false appearance of avicularia. The hyperstomial ovicells shown on the specimen from Chef du Pont are broad and low. The horse shoe-like lamella in the interior of the zooecia seems to be hidden by a fine calcitic crust, but it is indicated in pi. 2, figs. 2-3, by two small knob-like processes at the inner margin of the opesia. 24 UPPER CRETACEOUS POLYZOA FROM COTENTIN SPECIMENS. 0.49583. A small fragment of zoaiium, embedded in matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. 0.49845. Fragment of zoarium. Upper Maastrichtian, Fresville, Cotentin, Manche, France. Voigt Collection. Fragment with ovicelled zooecia. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection, Ham- burg, Nr. 3909. STRATIGRAPHICAL RANGE. Upper Maastrichtian. Genus BACTRELLARIA Marsson 1887 14. Bactrellaria rugica Marsson (PI. 2, figS. 9-14) 1887 Bactrellaria rugica Marsson 59, pi. 5, fig. 18. 1930 Bactrellaria rugica Marsson; Voigt: 444, pi. 12, figs. 25-26. The species is represented by one very poor fragment showing four zooecia of the frontal face with worn avicularia forming cavities between the opesiae. Although the specimen is very small and worn, the species is so characteristic that there is no doubt about the identification. It has been found by the author at Maastricht and Port Filiolet (Manche) where it is rare. The frontal avicularium below the opesia is, if well preserved, rather prominent, elongate beak-like and is proximally oriented with a raised spatulate peak. Marsson's figure showing only a round pore is inadequate, but he makes mention of the tube-like avicularia which occasion- ally occur on the lateral edges of the zoaria. The frontal avicularium below the opesia, is, if present and well preserved, rather prominent, tube-like, elongate and proximally oriented. Its beak is dilated and spatulate but is mostly broken away or damaged. It is rather well preserved on a specimen from an Upper Maastrichtian chalk-bearing erratic flint-boulder found in the gravel-pit of Wulmstorf near Harburg (pi. 2, fig. 13). If the dilated spatulate avicularian beak is destroyed there remains an oval ring or a scar like the ones in many specimens of the Cotentin (pi. 2, figs. 9 and 12). The small pit on Marsson's figure, however, is not the cicatrix of a destroyed avicularium, but corresponds to the spot where the avicularium is usually developed. Three pairs of oral spines are present in well preserved specimens. The ovicells are hyperstomial. Many fragments of the band-like zoaria from the Cotentin localities have four to six rows of zooecia instead of three; their diameter varies between 0-7-2-0 mm. The length of the zooecia is between 0-6-0-8 mm., that of the opesia c. 0-35 mm. In addition to the band-like specimens there occur, at Maastricht, prismatic vincularian rods with five or more rows of zooecia showing absolutely the same features and size of zooecia, avicularia and ovicells. They are very like Pithodella and it is possible that they represent another mode of growth of the same species. Eschara gaimardi von Hagenow (1851 : 82, pi. 12, fig. 10) is thought to be an Eschara-like stem development of this genus. STRATIGRAPHICAL RANGE. Maastrichtian. SPECIMENS. 0.49584. A small worn fragment. From matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. UPPER CRETACEOUS POLYZOA FROM COTENTIN 25 0.49585. A well preserved fragment. Upper Maastrichtian, Port Filiolet, Cotentin, Manche, France. E. Voigt Collection. A worn fragment. Upper Maas- trichtian, Port Filiolet, Cotentin, Manche, France. Voigt Collection, Hamburg Nr. 3914. A well preserved fragment. Upper Maastrichtian Chalk from Flint-boulder in Pleistocene Drift. Wulmstorf near Harburg-Hamburg, Germany. Voigt Collection Nr. 4146. Genus STAMENOCELLA Canu & Bassler 1917 15. Stamenocella marginata (d'Orbigny) (PI. 3, figs. 1-5) 1852 Flusirella marginata d'Orbigny: 295, pi. 700, figs. 7-9. 1852 Flustrella convexa d'Orbigny: 290, pi. 699, figs. 1-3. 1852 Biflustra tesselata d'Orbigny: 271, pi. 694, figs. 7-9. HOLOTYPE. A small fragment of Flustrella marginata d'Orbigny. Upper Maas- trichtian, Sainte Colombe, Cotentin, Manche, France. In d'Orbigny Collection, Paris, Musee d'Histoire Naturelle, Nr. 8134. A worn fragment of this species embedded in the matrix of Onychocella bathonica Gregory (D . 181) agrees in all essential characteristics with the two specimens which were described by d'Orbigny as Flustrella marginata from Sainte-Colombe and Biflustra tesselata from Nehou, and which are figured here for comparison. The latter is a worn specimen of Stamenocella marginata showing traces of ovicells and represents the same stage of preservation as figs. 3 and 8 of pi. 31 of Canu & Bassler (1920) where the ovicells and the avicularia of Stamenocella mediaviculifera and Stamenocella inferaviculifera cause two shallow cavities between the opesiae. PI. 3, fig. 5, shows a rather well preserved specimen which seems to be intermediate between Flustrella marginata d'Orbigny and Flustrella convexa d'Orbigny. The basal region, which tapers proximally and shows a radicelled base, has zooecia cor- responding to those of Flustrella convexa. They are closed by a calcareous lamella and pierced by elliptical or round openings as can also be observed in other species of Stamenocella. The variation in the size of the opesiae is enormous in this species as shown in pi. 3, fig. 5, and the following measurements: LZ = 0-50-0-63 mm. l z = 0-17-0-25 mm. L = 0-17-0-33 mm. 1 = 0-07-0-15 mm. STRATIGRAPHICAL RANGE. Maastrichtian. SPECIMENS. 0.47322. A fragment embedded in the matrix of Onychocella bathonica Gregory (D.iSi). Upper Maastrichtian, Cotentin, Manche, France. D . 49586. A fragment from the matrix of D . 181. Upper Maastrichtian, Cotentin, Manche, France. 26 UPPER CRETACEOUS POLYZOA FROM COTENTIN 0.49587. A small worn fragment. Locality and horizon as above. A well preserved fragment showing the stage of Flustrella convexa d'Orbigny at the base of the zoarium. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France, Voigt Collection, Hamburg, Nr. 4128. Type specimen of Biflustra tesselata d'Orbigny. Upper Maastrichtian, Nehou, Cotentin, Manche, France. d'Orbigny Collection, Paris, Musee d'Histoire Naturelle. Nr. 8053. Genus THYRACELLA Voigt 1930 16. Thyracella cf . meudonensis (d'Orbigny) (PI. 2, figs. 4-6) cf. 1851 Biflustra meudonensis d'Orbigny: 263, pi. 692, figs. 4-6. 1951 Thyracella cf. meudonensis (d'Orbigny) Voigt: 59, pi. 9, figs. 4-5. Three fragments of this bilaminar species are conspecific with a " Biflustra " which is very abundant in the Maastrichtian of Port Filiolet (Manche). It has a very prominent large avicularium as is typical of Thyracella (pi. 2, fig. 4). I think it could be identified with Biflustra meudonensis d'Orbigny recorded by him from Ne"hou and Meudon. In the catalogue of the d'Orbigny collection only one fragment from the Chalk of Meudon is registered under Nr. 8090 although I possess more than fifty specimens from this locality which would fit very well into that species. It is worn and does not correspond with the figure; it does not look like a chalk fossil from Meudon, but seems to have come from a more littoral facies resembling the Maastrichtian of the Cotentin. The worn specimens, described and figured by the present author from the Maastrichtian of Kunrade (S-Limburg) and Ilten (northern Germany) give a different impression from that of the well preserved material from the Cotentin localities in the author's collection. But there are all intermediate stages. For comparison a photograph of a characteristic specimen with a large avicularium, from Port Filiolet (Manche), is given (pi. 2, fig. 4). The length of the avicularium is c. 0-8 mm., and that of the zooecia c. 0-5-0-6 mm. The well preserved zooecia always have a sharp and distinct margin in the distal part as shown in the figure of d'Orbigny. STRATIGRAPHICAL RANGE. Maastrichtian and (fide d'Orbigny) Upper Campanian. SPECIMENS. D .49588. A damaged fragment from the matrix of D . 181. Upper Maastrichtian, Cotentin, Manche, France. 0.49846. A fragment embedded in matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. 0.49589-90. Two fragments from the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. A well preserved branched fragment with an avicularium. Upper Maastrichtian, Port Filiolet, Cotentin, Manche, France. Voigt Collection, Hamburg. Nr. 3919. A worn fragment labelled Meudon but likely. Upper Maastrichtian, locality uncertain (? Nehou, Cotentin, Manche). d'Orbigny Collection, Paris, Musee d'Histoire Naturelle Nr. 8040. UPPER CRETACEOUS POLYZOA FROM COTENTIN 27 Genus VINCULARIA (auct.) The genus Vincularia Defrance is here understood in the sense of the older authors, although this is not correct according to the rules of nomenclature. Vincularia is now restricted to those forms which Canu 1907 named Heterocella. But Vincularia fragilis Defrance 1829 from the Eocene is the type species of the genus of Vincularia (cf. Bassler 1953 : 157) and the name Vincularia must be reserved for this group. Heterocella, therefore, as a synonym of Vincularia, must be dropped. The consequence of this is that there is no name available for the many Cretaceous species of " Vincularia " of d'Orbigny, Marsson, Brydone and other authors. Canu (1900) has united most species of Vincularia under Smittipora Jullien (1881), which Bassler (1953) considered con-generic with Diplopholeos , Rectonychocella and Velu- mella Canu & Bassler. But the definition of Smittipora does not fit most " Vin- cularias ". Admitted that Vincularia has been established primarily as a zoarial growth-form for rod-like stems, there still exists a natural group of forms for which this name has been used and for which another name does not yet exist. On the other hand it seems to be necessary to distinguish the forms which are radicelled or articulated at their base from those which are attached by an encrusting base. This is not the place to give a new classification of Vincularia, and it is necessary to retain the old name provisionally. 17. Vincularia canalifera von Hagenow (PI. 5, figs. 7-10 and PI. 8, figs. 9-12) 1851 Vincularia canalifera von Hagenow: 61, pi. 6, fig. 14. ?i85i Vincularia flexuosa d'Orbigny: 76, pi. 656, figs. 16-18. 1930 Vincularia canalifera von Hagenow; Voigt, 467, pi. 17, fig. 18. More than thirty fragments of this species were found in the matrix of D.iSi. They are conspecific with the most common Vincularia-species of Maastricht and Kunrade (Netherlands) which was described by von Hagenow as Vincularia canali- fera. Von Hagenow's figure is not quite typical because the length of the opesia is relatively large, about one third or one quarter of the length of the zooecium. Among some hundred fragments there are only two which correspond to von Hage- now's figure, but this species can show great variation in size and shape of the zooecia and opesiae. The opesia varies between an oval (pi. 5, figs. 7, 9, 10) or more semicircular (pi. 5, fig. 8, pi. 8, figs. 9-11) opening which is truncated more or less proximally. In well preserved specimens it has a slight margin, but never a proximal lip. In the proximal basal region the zooecia are 0-25 mm. long, and when fully developed 0-5 mm. ; the length of the opesiae is 0-10-0-17 mm. The avicularia, which have not been figured until now, reach the length of the zooecia or exceed it (pi. 5, fig. 8, pi. 8, figs. 10, n). They are rare, straight, and appear at the beginning of a new row of zooecia, or are normally enclosed in these. They are broader than the autozooecia and have a flatly rounded prominent distal rim and a small elliptical 28 UPPER CRETACEOUS POLYZOA FROM COTENTIN opesia which is longitudinally oriented. Common to the zooecia and the avicularia is a furrow-like deepening in the median axis of the cryptocyst from which the specific name is derived. Ovicells have never been observed. The rods have a diameter of 0-5-0 -7 mm. and consist of eight to fourteen rows of zooecia. They taper proximally and were articulated or radicelled at their base. This is shown by some of the earliest zooecia which seem to be modified to lodge chitinous rootlets (pi. 5, fig. 7). Their opesiae are smaller and their upper half is covered by a bent calcitic lamella as is seen also in other articulated or radicelled forms. This species is also common in the Maastrichtian of the Cotentin near Port Filiolet, Fresville and Chef du Pont from which region d'Orbigny described some very similar vincularian species. I take it that Vincularia flexuosa d'Orbigny is conspecific with this species, which is cited by d'Orbigny from Nehou in the Cotentin and from the Santonian of Ven- dome. I have studied the type (Nr. 7752 in the d'Orbigny collection) and I cannot find any significant differences. It is figured on pi. 8, fig. 14 for comparison with specimens from Maastricht (pi. 5, figs. 7 and 8 and pi. 8, figs, n and 12). The reason why I hesitate to place this species under d'Orbigny's name is, that according to the label the type specimen is from the Santonian of Vendome (as stated also in his catalogue) although I suspect that it comes from Nehou in the Cotentin. Because it is not clear if the holotype comes from the Cotentin, I prefer the name Vincularia canalifem which is given by von Hagenow in the same year as d'Orbigny's name. Another form from the Cotentin which is very similar to Vincularia canalifera is V. concinna (d'Orbigny 1851 : 79, pi. 657, figs. 10-12). It may be another syno- nym. D'Orbigny's holotype is figured on pi. 8, fig. 13; this is the only specimen of this species in the d'Orbigny Collection of Nr. 7756). This specimen, recorded by Canu (1900 : 420) as "insuffisant ", has opesiae also which are about 0.11-0.17 mm. long, but the cryptocyst seems to be less deepened than in Vincularia canalifera von Hagenow. Unfortunately it shows no avicularia, knowledge of which is very import- ant for the discrimination of many species of Vincularia which are very similar and cannot be identified from the work of d'Orbigny. D'Orbigny never figured or mentioned in his species the avicularia, which are rather rare and similar to the autozooecia. STRATIGRAPHICAL RANGE. Maastrichtian. SPECIMENS. 0.49732. Branched fragment, Upper Maastrichtian, Geulem near Berg, Geul- Valley near Maastricht, Netherlands. Voigt Collection. D. 49733, 0.49735-36, 0.49737-66. Fragments from the matrix of D.iSi, Upper Maastrichtian, Cotentin, Manche. 0.49734. Fragment, Upper Maastrichtian Chef du Pont, Cotentin, Manche. Voigt Collection. 0.49840-41. Two fragments, Upper Maastrichtian, St. Pietersberg near Maas- tricht, Netherlands. Voigt Collection. 0.49842. Fragment, Upper Maastrichtian, St. Pietersberg near Maastricht, Netherlands. Voigt Collection. UPPER CRETACEOUS POLYZOA FROM COTENTIN 29 Genus QUADRICELLARIA d'Orbigny 1850 18. Quadricellaria elegans d'Orbigny (PI. 6, figs. 8-10) 1851 Quadricellaria elegans d'Orbigny: 33, pi. 652, figs. 1-5. 1900 Quadricellaria elegans d'Orbigny; Canu: 413. 1928 Quadricellaria excavata d'Orbigny; Voigt: 112, text-figs. 1-5. 1930 Quadricellaria excavata d'Orbigny ; Voigt : 489, pi. 25, figs. 12-14. Two small fragments of a Quadricellaria may be determined as Q. elegans d'Orbigny. The size of the opesia compared with the length of the zooecia varies considerably, and Canu may be right in uniting d'Orbigny's three species, Q. elegans, Q. excavata and Q. pulchella, under the name Q. excavata, as I did in 1928. My specimens agree largely with the figures of Q. elegans d'Orbigny which was first recorded from Nehou (Manche) . This species is represented in my material from Chef du Pont and Port Filiolet. One specimen from Chef du Pont has, on the narrow sides of the segments, two enormous avicularia which are I mm. long and have an opesia with a length of 0-3 mm. (pi. 6, fig. 10). SPECIMENS. 0.49591-92. Two worn fragments. From matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. A fragment with a large avicularium. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection, Hamburg, Nr. 4133. Genus COSCINOPLEURA Marsson 1887 19. Coscinopleura sp. Two minute indeterminable fragments of a Coscinopleura, showing only a few zooecia are present. They possibly belong to a species of Coscinopleura with small zooecia, like Coscinopleura lamourouxi von Hagenow, or to a similar form. Common in the Cotentin Maastrichtian near Port Filiolet and Chef du Pont (Manche). SPECIMENS. 0.49593-94. Two small worn fragments. From the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. Genus SEMIESCHARINELLA d'Orbigny 1852! 20. Semiescharinella complanata d'Orbigny (PI. 4, figs. 11-15) 1840 Cellepora ricata von Hagenow: 616. 1852 Semiescharinella complanata d'Orbigny: 427, pi. 714, figs. 1-4. 1 D'Orbigny published his genera Semiescharinella and Reptescharinella on p. 427 and p. 428 of his work. According to Sherborn (Geol. Mag. 1889 : 223-225) pages 187-472 appeared in 1852. I follow here the dates given by Sherborn, although Lang (1917 : 172) states 1853 for Reptescharinella and d'Orbigny himself writes 1851 for both genera. 3 o UPPER CRETACEOUS POLYZOA FROM COTENTIN 1852 Escharinella simplex d'Orbigny: 205, pi. 683, figs. 14-16. 1900 Rhagasostoma simplex (d'Orbigny); Canu: 433. 1930 Micropora subgranulata (von Hagenow) Voigt (pars): 476, pi. 24, fig. 18 (non 19). 1959 Semiescharinella complanata (d'Orbigny) Voigt: 54, pi. 6, fig. i. 1962 Semiescharinella complanata (d'Orbigny) Berthelsen: 134, pi. 14, figs. 2-3. Two very small unilaminar fragments agree very well with some specimens from Chef du Pont and with d'Orbigny's type-specimen from Sainte Colombe (Manche). The size of the zooecia is a little less in the type-specimen (about 0-56-0-60 mm.) compared with 0-70 mm. in 0.49595-96; but the abundant material in the author's collection shows that the size of the zooecia and opesiae is highly variable in this species. The " TYPE " is a fragment, with c. 18 zooecia, in which the relative length and width of the zooecia differ from d'Orbigny's figure as noted by Canu (1900 : 421), although in the catalogue of the d'Orbigny Collection only one speci- men from Sainte Colombe (Manche) is registered. The bilaminar Escharinella simplex d'Orbigny from Nehou (Manche) belongs to the same species. The distal pore which was interpreted by Canu as the trace of the ovicell is a true avicularium. Nevertheless, there are some real ovicells which are developed in place of the avi- cularia (see pi. 4, fig. 15). Comparison of the photographs (pi. 4, figs. 13-15) shows that there is no variation in the shape or size of the zooecia, nor of the opesiae, nor amongst the distal avicu- laria. I have figured the only poor fragment from the d'Orbigny collection which must be regarded as the type specimen (pi. 4, fig. 15). This species is congeneric with Cellepora (Discopora) subgranulata von Hagenow (1851 : 91, pi. n, fig. 15) which was chosen by Lang (1917 : 172) as the type species for the genus Repteschari- nella d'Orbigny 1852. D'Orbigny had united under this name eight Cretaceous, one Tertiary and two Recent species, of quite distinct systematic differences, said to be characterized by an " ouverture mediocre ", a " pore special " and encrusting zoaria. The genus Semiescharinella was not discussed by Lang, being represented only by Semiescharinella complanata d'Orbigny. There is no doubt what d'Orbigny meant by this name, and I prefer it, proposing to drop Reptescharinella as a synonym of Semiescharinella, the mode of growth being no generic criterion. This form was first described by von Hagenow 1840 under the name Cellepora ricata without any figure (Voigt 1959 : 54). STRATIGRAPHICAL RANGE. Maastrichtian-Danian. SPECIMENS. 0.49595-96. Two small worn fragments. From the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. 0.49847. A small fragment. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection. A small fragment showing an ovicelled zooecium. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection Hamburg Nr. 39i8b. Bilaminar fragment, Holotype of Escharinella simplex d'Orbigny. Upper Maas- trichtian, Nehou, Cotentin, Manche, France. d'Orbigny Collection, Paris, Musee d'Histoire Naturelle Nr. 7942. UPPER CRETACEOUS POLYZOA FROM COTENTIN 31 Genus ROSSELIANA Jullien 1888 21. Rosseliana thomasi sp. n. (PL i fig. 6, and PL 4, figs. 4-6) TYPE SPECIMEN. 0.49597. Zoarium encrusting a branched fragment of Radu- lopora minor n. g. n. sp. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Collection E. Voigt. DERIVATIO NOMINIS. In honour of Dr. Dighton Thomas, London, who first detected this species in the matrix of Castanopora jurassica (Gregory) . DIAGNOSIS. A Rosseliana with encrusting zoarium. Zooecia very small, oval, only 0-33 mm. long, with smooth cryptocyst and sharply raised rim in the distal region of the zooecia. Opesiae semicircular, occupying nearly a third of the length of the zooecia, sometimes with very inconspicuous lateral processes, but never trifoliate in shape; its lower rim is straight without opesiules. Ovicells globular swellings above the opesia and occupying the proximal part of the distal zooecium. DESCRIPTION. In addition to the British Museum specimen, which encrusts a branch of Multicrescis laxata d'Orbigny, there are two others from Chef du Pont (Manche). They show, although they are more fragile, the same essential specific characters, having no avicularia and no trifoliate opesiae. In some zooecia, very minute lateral processes might indicate a Floridina-like opesia, but this is so negli- gible that it can be ignored. They are therefore classified as Rosseliana and not Floridina or Floridinella. Nevertheless, it is clear that the difference between Rosseliana and Floridinella, based only on the existence of broad opesiular indenta- tions in the latter, is slight. The specimens are very similar to that figured by Bassler (1953, fig. 130.4), but he only records Rosseliana from the Oligocene to Recent. There are very few species with which this form could be compared. The shape and size of zooecia are very similar in Floridina (or better Floridinella) scutata Levinsen (1925 : 345, pi. 4, fig. 39) from the Danish Maastrichtian Chalk and Danian, and Semieschara complanata d'Orbigny (1852 : 369, pi. 708, figs. 5-8). Apart from the free unilaminar fronds of the latter, it is very difficult to find any constant difference between these two forms, which are distinguished from our new Rosseliana thomasi only by their opesiae, always markedly trifoliate as in Floridina or Floridinella. PL 4, fig. 7, shows a specimen of Floridinella scutata Levinsen from a Maastrichtian flint drift boulder from northern Germany. The difference of this species from Rosseliana thomasi n. sp. is clearly seen in the opesia which has significant lateral indentations. STRATIGRAPHICAL RANGE. Upper Maastrichtian. SPECIMENS. 0.47323. Encrusting zoarium with some ovicelled zooecia on Multicrescis laxata d'Orbigny (= 0.49575). In the matrix of Castanopora jurassica (Gregory) (D . 180) Upper Maastrichtian, Cotentin, Manche, France. Zoarium encrusting an echinoid fragment. Upper Maastrichtian, Port Filiolet, Cotentin, Manche, France. Voigt Collection, Hamburg, Nr. 3932. 32 UPPER CRETACEOUS POLYZOA FROM COTENTIN Genus MICROPORA Gray 1848 (non Eichwald 1855) 22. Micropora transversa (d'Orbigny) (PI. 7, figs. 9-12) 1851 Vincularia transversa d'Orbigny: 78, pi. 657, figs. 7-9. 1887 Vincularia rugica Marsson: 65, pi. 6, fig. 8. 1930 Micropora rugica (Marsson) Voigt: 472, pi. 21, fig. 20. 1951 Micropora rugica (Marsson) Voigt: 63, pi. 9, fig. 10 and pi. 10, fig. 4. HOLOTYPE. Two minute fragments evidently 2 pieces of one specimen. Upper Maastrichtian, Nehou, Cotentin, Manche, France. d'Orbigny Collection, Paris, Musee d'Histoire Naturelle Nr. 7755. There is only a very small fragment corresponding to the length of a single zoo- ecium of this species in the matrix of D . 181. The two symmetrical opesiules below the opesia show clearly that it must belong to Vincularia transversa d'Orbigny, the type of which, from Nehou, (pi. 7, figs. 10-11) was studied by the author. There is only one poorly preserved fragment in d'Orbigny's collection which is recorded in his catalogue. It is now broken into two pieces. It does not correspond to d'Orbigny's figure and description, because the paired opesiules, which are clearly visible, are neither mentioned in his text nor figured in his drawing ; they seem to be indicated by the deep furrows accompanying the thick margin of the zooecia, but they do not correspond to what d'Orbigny called " une depression lanceolee " which is situated below the aperture. This depression must correspond to the lanceolate cryptocyst of d'Orbigny's figure. I found some well preserved fragments at Port Filiolet and Chef du Pont (Cotentin), see pi. 7, fig. 12. This form is conspecific with Marsson's Vincularia rugica as is proved by many specimens from the Maastrichtian Chalk of Rugen and other localities in northern Germany. Marsson who published the first good description and figure of this species, was therefore unable to recognize that his species was the same as d'Orbigny's, and it is understandable that, since Marsson's description, this form has been re- corded only under the name rugica. Another similar species is Vincularia undata d'Orbigny from the Santonian of Vendome, which also has true opesiules below the opesia, not indicated in d'Orbigny's figure (d'Orbigny 1851 : 75, pi. 656, figs. 10-12) but which can be seen on the holotype. The appearance of Micropora transversa can vary considerably as shown in Voigt's figures (1951). The diameter of the rods is from 0-5-0 -7 mm., and the number of zooecial rows varies between five and ten. The length of the normal zooecia is o -49-0 -54 mm. The basal attenuated part of some rods shows clearly that this species was articulated or radicelled (cellarif orm) . This means that this form is not congeneric with all true Micropora, and it could be suggested that it belongs to a new genus. But I hesitate to erect a new genus for it because it is not impossible that this form is conspecific with Dimorphostylus tetrastichtis Voigt 1928. The genus Dimorphostylus was established for articulated rods, in which the zooecia are only developed on one side. But the size and other features of the UPPER CRETACEOUS POLYZOA FROM COTENTIN 33 zooecia in Micropora rugica and in Dimorphostylus tetrastichus , which are often associated together in the same localities, are identical, and there is the suspicion that Dimorphostylus tetrastichus might represent specimens with the zooecia de- veloped only on one side, and that the differentiation in a frontal and reversed side may be pathological. The two forms have not yet been found united in one speci- men, but if they were, the generic name Dimorphostylus must be applied to Micropora transversa d'Orbigny. Therefore it is provisionally here referred to Micropora, which also indicates that it may belong to the Microporidae. STRATIGRAPHICAL RANGE. Maastrichtian ; Lower Maastrichtian of Riigen and Denmark and Upper Maastrichtian of the Cotentin (Manche), Maastricht and Ilten near Hanover. SPECIMENS. 0.49768. A minute fragment. From matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. 0.49769. A well preserved fragment. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection. Genus PUNCTURIELLA Levinsen 1925 23. Puncturiella cf. superba Brydone (PL 4, figs. 8-10) 1936 Puncturiella superba Brydone: 84, pi. 40, fig. 18. The small unilaminar fragment with about a dozen poorly preserved zooecia, shows the cryptocyst pierced by two outer and two inner rows of pores. These are barely visible because they are partly obscured by recrystallization of the calcite, and the distal avicularium which is directed obliquely is very obscure. However the identity of this specimen with a few unilaminar fragments from Fresville and Chef du Pont is evident (pi. 4, figs. 8-9), and they are probably identical with some very similar specimens from Maastricht which are free or encrusting, but whose dimensions are a little larger. The average length of zooecia from the Cotentin localities is about 0-65-0-70 mm., that from Maastricht 0-80-1-00 mm., although some zooecia from the two localities are the same size, and it may be that the difference is ecological. The only form with which 0.49598 can be identified is Puncturiella superba Brydone from the Lower Maastrichtian Chalk of Trimingham (Norfolk), in which the zooecia are about 0.65-1 mm. long. STRATIGRAPHICAL RANGE. Maastrichtian. SPECIMENS. 0.49598. A small worn fragment embedded in matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. 0.49599. A small fragment. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection. GEOL. 17, I. 3 34 UPPER CRETACEOUS POLYZOA FROM COTENTIN 0.49600. A small well preserved fragment in matrix. Upper Maastrichtian Md, St. Pietersberg near Maastricht. Voigt Collection. Genus LUNULITES Lamarck 24. Lunulites sp. One very small fragment which is worn and shows a few zooecia is indeterminable. It seems to be conspecific with one of the numerous species of Lunulites represented in the Maastrichtian of the Cotentin. SPECIMENS. 0.49601. A very small indeterminable fragment. In matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. Genus ONYCHOCELLA Jullien 1881 25. Onychocella nysti (von Hagenow) (PI. 6, figs. 3-4) 1851 Eschar a nysti von Hagenow: 78, pi. 9, figs. 15-17. 1930 Onychocella nysti (von Hagenow) Voigt: 459, pi. 16, figs. 14-16. One small fragment belongs to this species which is very common at all European Maastrichtian localities and which I found at Port Filiolet and Chef du Pont (Manche) . It has not previously been recorded from the Cotentin Maastrichtian. STRATIGRAPHICAL RANGE. Campanian-Maastrichtian. SPECIMENS. 0.49602. A worn fragment. From the matrix of D.i8i. Upper Maastrichtian. Cotentin, Manche, France. 0.49603. A well preserved fragment. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection. 26. Onychocella cf. cepha (d'Orbigny) (PL 6, figs. 1-2) 1851 Eschara cepha d'Orbigny: 143, pi. 670, figs. 8-10. Two fragments of a narrow Onychocella belong to a species which is very abundant near Chef du Pont (Manche) and which is not distinguishable from an Onychocella from the Maastrichtian of Archiac (Gironde). Of the numerous species of Onycho- cella described by d'Orbigny from the French Cretaceous, the only one which can be compared with this form is Eschara cepha d'Orbigny from Roy an, although the type- specimen has more slender zooecia and thinner rims surrounding the zooecia. Com- parison with the type alone would suggest a different species, but when compared with all the material from Archiac and Chef du Pont, these differences are much diminished and identity is more justified than the foundation of a new species. STRATIGRAPHICAL RANGE. Maastrichtian. UPPER CRETACEOUS POLYZOA FROM COTENTIN 35 MEASUREMENTS. L z = 0-48-0-55 mm. l z = 0-30-0-37 mm. h = 0-10-0-17 mm. 1 = 0-10-0-13 mm. L av = 0-70-0-75 mm. SPECIMENS. 0.49604. A branched fragment. From the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. D . 49605 . A small fragment . From the matrix of D . 1 8 1 . Upper Maastrichtian , Cotentin, Manche, France. 0.49606. A fragment. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection. 27. Onychocella bellona (d'Orbigny) (PI. 6, figs. 5-7) 1851 Eschara bellona d'Orbigny: 134, pi. 668, figs. 7-9. 1900 Rhagasostoma bellona (d'Orbigny) Canu (pars): 431. Two fragments of this broad bilaminar species agree in all details with the material collected by the author at Chef du Pont (Manche) and with as pecimen in d'Orbigny's collection from Nehou (Manche). (d'Orbigny Collection Nr. 7812.) The small difference between the diameter of the apertures in our figured specimen and that of d'Orbigny is irrelevant because it is also shown in the specimens from Chef du Pont. The opesiae of the fertile zooecia are a little longer than the others. This species belongs to the group of Onychocella lamarcki von Hagenow, which is closely allied to it; but the latter always shows more slender branches (1-2-2-0 mm. diameter), and shorter and thicker zooecia, and the difference in size of the fertile and non fertile zooecia is much more evident than in Onychocella bellona d'Orbigny. It is the same with the fertile zooecia of Onychocella lamarcki von Hagenow from Maastricht, but the zoaria of that species are always flat and lamellar. MEASUREMENTS. L z = 0-60-0-70 mm. l z = 0-33-0-45 mm. h = 0-12-0-18 mm. 1 = 0-12-0-19 mm. L av = 0-75-1-00 mm. l av = 0-22-0-27 mm. This species has been found only in the Cotentin Maastrichtian. STRATIGRAPHICAL RANGE. Upper Maastrichtian. SPECIMENS. 0.49607. A worn fragment. From the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. GEOL. 17, I. 3 36 UPPER CRETACEOUS POLYZOA FROM COTENTIN 0.49608. A well preserved fragment. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection. A well preserved fragment with ovicelled zooecia. Upper Maastrichtian, Nehou, Cotentin, France. d'Orbigny Collection, Paris, Musee d'Histoire Naturelle Nr. 7812. Genus ONYCHOCELLARIA Voigt 1957 28. Onychocellaria caecilia (d'Orbigny) (PL 5, %s. 1-6) 1851 Eschara caecilia (d'Orbigny): 138, pi. 669, figs. 4-7. The zoarium was cellariiform in growth. This is proved by the tapering proximal ends of the segments and by pits occasionally found on the cryptocyst in which rootlets are inserted. This criterion, combined with the straight avicularia, puts it in the genus Onychocellaria, although the endozooecial ovicell, which is characteristic for this genus, has not yet been observed. The zooecia, the length of which is 0-4-0-5 mm., are nearly rectangular and are very often distinguished by a small horizontal band above the distal rim of the opesia. The slit-like pit between the opesia and this band is very characteristic, and is shown in d'Orbigny's figures. The opesiae, which vary greatly in size, may be rounded or oval to high-semicircular. In some segments the whole opesia or its upper half is closed by a calcareous lamella. MEASUREMENTS. L z = 0-42-0-51 nun. l z = 0-20-0-25 mm. h = o-io-0'70 mm. 1 = o-io-o-n mm. LAV = 0-55-0-58 mm. 1 AV = 0-25-0-30 mm. This species is represented by many small fragments. It is characteristic of the Maastrichtian of the Cotentin, and has also been found in abundance by the present author near Port Filiolet (Manche). The specimens agree entirely with the original specimens of d'Orbigny from the Maastrichtian of Nehou. These are not " use " as stated by Canu (1900 : 420). STRATIGRAPHICAL RANGE. Upper Maastrichtian. SPECIMENS. 0.49726. A small fragment with avicularia. From the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. 0.49727. Another fragment with avicularium. From the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. D. 49728-29. Two worn fragments. From the matrix of D.iSi. Upper Maas- trichtian, Cotentin, Manche, France. 0.49767. A small fragment corresponding in preservation to 0.49730. From the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. UPPER CRETACEOUS POLYZOA FROM COTENTIN 37 0.49849-54. 6 fragments from the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. D . 49848. Eight very worn small fragments from the matrix of D . 181. Locality and horizon as above. D. 49730. A well preserved fragment with some avicularia. Upper Maastrich- tian, Chef du Pont, Cotentin, Manche, France. Voigt Collection. D. 49731. A well preserved fragment. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection. Genus PLIOPHLOEA Gabb & Horn 1862 29. Pliophloea sp. (PI- 7, ng. 3) Of this species there is but one encrusting young zoarium, which is composed of the ancestrula and a dozen zooecia. The ancestrula is 0-20 mm. long and the other zooecia have a length up to 0-50 mm. The smooth intraterminal front wall shows about twelve to thirteen costae which are barely visible. The orifice is cribri- line to slightly pliophloean according to Lang's definition. There is only one avicularium if the interpretation as avicularium is correct which is c. 0-20 mm. long and which is close to the ovicelled zooecium on the right of our figure. Details of the spines are not visible. Pliophloea gluma Lang (1921 : 188, pi. 6, fig. 3) from the Danian shows some affinities in the shape of the zooecia and the apertures, but has clearly distinct costae, and visible intercostal fusions, and the zooecia are only 0-40 mm. long. There is no other species hitherto described with which this form can be identified and therefore it may be supposed that it is a new one. But this one poor specimen does not suffice to found a new species. STRATIGRAPHICAL RANGE. This form has not previously been observed in the Maastrichtian of the Cotentin. SPECIMENS. 0.47324. Small encrusting zoarium with ancestrula and an ovi- celled zooecium. In the matrix of Castanopora jurassica (Gregory) D . 180. Upper Maastrichtian, Cotentin, Manche, France. Genus DECURTARIA Jullien 1886 (= Prosoporella Marsson 1887) 30. Decurtaria cf. cornuta (Beissel) (PI. 7, figs. 1-2) 1865 Semiescharipora cornuta Beissel: 58, pi. 7, figs. 77-81. 1887 Prosoporella cornuta Marsson : 100. 1922 Decurtaria cornuta (Beissel) Lang: 388, text-fig. 124, pi. 8, fig. 9. 1925 Barroisina trifossata Levinsen: 387, pi. 8, fig. 6. 1930 Decurtaria cornuta (Beissel) Voigt: 516, pi. 32, fig. 6. 38 UPPER CRETACEOUS POLYZOA FROM COTENTIN This species is represented by a fairly large zoarium composed of about thirty zooecia. Unfortunately it is covered by a film of minute calcite-crystals which hides many of the characteristic minute details of the costae and the orifices. The small number of costae (six to seven) and the very stout distal shield indicate it to be Decurtaria cornuta, although the shape of the orifice is more like that of Decurtaria allecta Lang (1922 : 386, pi. 8, fig. 8) from the Upper Maastrichtian of Maastricht. D . 49609 corresponds exactly to a fragment from Chef du Pont (Manche) . In some respects that form might be regarded as intermediate between the two species, which are, indeed, more alike than is stated by Lang who had only one specimen of his D. allecta from Maastricht. Decurtaria cornuta is not confined to the Lower Maas- trichtian as might be concluded from Lang's work. I found a small but typical fragment of this species in the Upper Maastrichtian of Biebosch (South Limburg, Netherlands) . It seems that the shape of the orifice is not so constant as was assumed by Lang. According to him, in D. cornuta the orifice should be " super-cribriline " and in D. allecta " supernormal ". But Beissel had already stated that the different types of orifice shape may be observed in the same zoarium, and he has figured both forms in his figs. 77 and 78. The size of the zooecia is between 0-57 and 0-65 mm. Barroi- sina trifossata Levinsen is a synonym of this species which I concluded from the examination of Levinsen's type specimen. STRATIGRAPHICAL RANGE. Maastrichtian. SPECIMENS. 0.49609. A complete zoarium with ancestrula and some ovicelled zooecia. From the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. An incomplete zoarium. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection, Hamburg, Nr. 4137. Genus FRURIONELLA Canu & Bassler 1926 30. Frurionella fertilis sp. n. (PI. 8, figs. 5-7) HOLOTYPE. 0.49610. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Collection E. Voigt. DERIVATIO NOMINIS. Derived from fertilis = generative, on account of the numerous ovicells. DIAGNOSIS. A slender species of Frurionella, with bilaminar branches consisting of three to four alternating rows of zooecia. Zooecia not distinct, small, with quadrangular to high triangular or semicircular opesiae whose proximal margin is straight and generally provided with an inconspicuous proximal lip caused by the swelling of the median avicularium below the opesia. Elliptical or linear indistinct pores which may be avicularia are developed in the median axis of the zooecium. Ovicells numerous, forming deep characteristic cavities above the apertures if the covering lamina has been destroyed. UPPER CRETACEOUS POLYZOA FROM COTENTIN 39 MEASUREMENTS. L z = 0-50-0-62 mm. h = 0-10-0-12 mm. 1 = o -10-0 -12 mm. REMARKS. A small fragment, showing a scarcely constricted opesia with a minute avicularium below it, has been recognized as belonging to this new species which is represented from Chef du Pont by two other similar specimens. This form cannot be identified with any of the few known species of Frurionella. It is smaller in its zoarial and zooecial dimensions than the other species of this genus. At first it was supposed that it might be the slender distal branch of another new species of Frurionella which is much larger and very common at Fresville (Cotentin, Manche). This is improbable because there are no intermediate stages between these two very different forms and no distal branches smaller than 2-1-5 mm. seem to exist in the second species. This other species of Frurionella from Fresville (pi. 8, fig. 8) is conspecific with Frurionella europaea Voigt (1951 : 60, pi. 9, figs. 1-3) from the Upper Maastrichtian of Ilten (Hannover) and Kunrade (Netherlands). It has now also been found at Maastricht (Netherlands). STRATIGRAPHICAL RANGE. Maastrichtian. SPECIMENS. 0.49610. Holotype see above. D . 49611. A small worn fragment in matrix. From the matrix of D . 181. Upper Maastrichtian, Cotentin, Manche, France. 0.49612. A fragment with broken ovicelled zooecia. Horizon and locality as above. Genus BEISSELINA Canu 1913 32. Beisselina striata (Goldfuss) (PI. 8, figs. 3-4) 1826 Eschara striata Goldfuss: 25, pi. 8, fig. 16. 1960 Beisselina striata (Goldfuss) Wiesemann: 27, pi. i, figs. 1-3, pi. 2, figs. 3-4; pi. 12, figs 3-4; text-figs. 36, 4 9-10; 5, Nrs. 12-13 (with additional synonymy). 1967 Beisselina striata (Goldfuss) Voigt: 72, pi. 25, fig. i. One specimen belongs to this common Maastrichtian species which is, according to Wiesemann, represented near Chef du Pont (Manche) . Nevertheless the diameter of the stem (1-5 mm.) and the orifices (peristomicia) are a little smaller (0-08-0-12 mm.) than in most specimens from the type locality of Maastricht, and Kunrade, although the length of the zooecia (distance between proximal and distal apertures) is the same. Specimens with such small orifices are not lacking, and all intermediate sizes occur at these localities, but they are rare. For comparison see pi. 8, fig. 4, showing the surface of a worn specimen from Maastricht itself. These small forms are more common in the " Tuffeau de St. Symphorien " in the Basin of Mons (Bel- gium.) 4 o UPPER CRETACEOUS POLYZOA FROM COTENTIN STRATIGRAPHICAL RANGE. Maastrichtian. SPECIMENS. 0.49721. A worn fragment. From the matrix of D.i8i. Upper Maastrichtian, Cotentin, Manche, France. 0.49722. A worn fragment. Upper Maastrichtian, Geulem, near Berg, Geul valley near Maastricht (Netherlands). Voigt Collection. Genus BEISSELINOPSIS Voigt 1951 33. Beisselinopsis flabellata (d'Orbigny) (PL 8, figs. 1-2) 1852 Escharifora flabellata d'Orbigny: 460, pi. 715, figs. 10-12. non 1930 Beisselina flabellata (d'Orbigny) Voigt: 525, pi. 34, fig. n. One incomplete young zoarium represents d'Orbigny's species which is common at Sainte Colombe (Manche) and the type of which from this locality was studied by the present author. There are some specimens from Chef du Pont (Manche) which agree very well with Beisselinopsis flabellata d'Orbigny although they are more elongate than the flabelliform zoarium of d'Orbigny's type specimen. The species identified as Porina flabellata (d'Orbigny) by Marsson (1887 : 85) is Beisselin- opsis marginata v. Hagenow 1839 (cf. Voigt 1959 : n, pi. 9, figs. 1-2). The similar Danian and Montian form determined as Porina or Beisselina flabellata (d'Orbigny) by Levinsen (1925, pi. 7, fig. 83) and Voigt (1930 partim : 525, pi. 34, fig. n) is neither conspecific nor congeneric and corresponds to Eschara oblita Kade (1852 : 29, pi. i, fig. 18) as stated by Berthelsen (1962 : 201, pi. 24, fig. 6) and Voigt (1964 : 458, pi. 8, fig. 8 and pi. 14, figs. 1-3) and belongs to the genus Pavobeisselina Voigt 1964. It was formerly assigned to Beisselinopsis Voigt 1951, but this attribution was incorrect because Beisselinopsis has no ascopore which can always be observed in Pavobeisselina. The inner structure of Beisselinopsis flabellata d'Orbigny does not show any ascopore in the frontal wall as is always the case in Beisselina and flabelli- form Pavobeisselina. Therefore it must be regarded as a true Beisselinopsis. STRATIGRAPHICAL RANGE. Upper Maastrichtian. SPECIMENS. 0.49723. A young zoarium. From the matrix of D.iSi. Upper Maastrichtian, Cotentin, Manche, France. An adult zoarium. Upper Maastrichtian, Chef du Pont, Cotentin, Manche, France. Voigt Collection, Hamburg, Nr. 3908. II. BRACHIOPODA Genus THECIDEA Def ranee 1832 Thecidea papillata (von Schlottheim) (PI. i, figs. 9-10) 1959 Thecidea papillata von Schlotheim; Backhaus: 21, pi. i, figs. 1-4 (see for all references). UPPER CRETACEOUS POLYZOA FROM COTENTIN 41 TABLE I. List of the species from the matrix of " Castanopora " jurassica (Gregory) (D . 180) and Onychocella piriformis (Goldfuss) (= Onychocella bathonica (Gregory) (D.iSi). I. POLYZOA 1. Berenicea sp. 2. Entalophora benedeniana (von Hagenow) 1851 3. Idmidronea macilenta (von Hagenow) 1851 4. Heterocrisina communis (d'Orbigny) 1853 5. Crisisina carinata (Roemer) 1840 6. Osculipora truncata (Goldfuss) 1826 7. Reteporidea lichenoides (Goldfuss) 1826 8. Petalopora sp. 9. Multicrescis laxata (d'Orbigny) 1854 10. Membranipora unipora (Marsson) 11. Amphiblestrella elegans von (Hagenow) 1851 12. Radulopora minor n. sp. 13. Hagenowinella cf. incrassata (d'Orbigny) 1853 14. Bactrellaria rugica Marsson 1887 15. Stamenocella marginata (d'Orbigny) 1852 1 6. Thyracella cf. meudonensis d'Orbigny 1851 17. Vincularia canalifera von Hagenow 1851 18. Quadricellaria elegans (d'Orbigny) 1951 19. Coscinopleura sp. 20. Semiescharinella complanata d'Orbigny 1852 21. Rosseliana thomasi n. sp. 22. Micropora transversa (d'Orbigny) 1851 23. Puncturiella cf. superba Brydone 1936 24. Lunulites sp. 25. Onychocella nysti (von Hagenow) 1851 26. Onychocella cf. cepha (d'Orbigny) 1851 27. Onychocella bellona (d'Orbigny) 1851 28. Onychocellaria caecilia (d'Orbigny) 1851 29. Pliophloea sp. 30. Decurtaria cf. cornuta (Beissel) 1865 31. Frurionella fertilis n. sp. 32. Beisselina striata (Goldfuss) 1926 33. Beisselinopsis flabellata (d'Orbigny) 1852 II. BRACHIOPODA 34. Thecidea papillata (von Schlottheim) 1813 * After d'Orbigny. 3 +3 HH o _g '-J3 ,4 d ^_^ o cn OJ (D 4-J o M Maastri localitic 4-> cS 0) 5 s ! J3 o 1 oo M 00 M JL 1 i ^ Q Q * 0) O ,! $ $ g +-> rt "o 3 P-i 2 1 o | * 3 ~3 f3 3 _o 3 3 .3 , 1 l-i 0> rt a h- 1 a i i rt CA1 % O A PLH O g 4 2 UPPER CRETACEOUS POLYZOA FROM COTENTIN Two small valves of this very common and characteristic brachiopod of the " Craie a Thecidees " were found in the matrix of D.iSi. They are indistinguish- able from those of the Cotentin Maastrichtian collected by the author. The collec- tion of the " Geologisches Staatsinstitut " Hamburg possesses about 850 specimens from Chef du Pont, Fresville and Port Filiolet, which were studied in the monograph by Backhaus. The locality " Port Fiolet " (Backhaus 1959 : 27 and Hofker 1959 : 369, 380, 381) should be " Port Filiolet ". STRATIGRAPHICAL RANGE. Maastrichtian, mainly Upper Maastrichtian. SPECIMENS. 66.42981. Pedicle valve. From the matrix of D.iSi. Upper Maastrichtian, Craie a Thecidees, Cotentin, Manche, France. 66.42982. Small dorsal valve. Horizon and locality as above. REFERENCES BACKHAUS, E. 1959. Monographic der cretacischen Thecideidae (Brach). Mitt. Geol. Staat- sinst. Hamburg, 28 : 1-125, pis. i-io, 14 figs. BASSLER, R. S. 1953. In R. C. Moore: Treatise on Invertebrate Paleontology, Lawrence, Bryozoa Part G: 01-253, 175 figs. BEISSEL, J. 1865. Uber die Bryozoen der Aachener Kreidebildung. Natuurk. Verh. Holland. Maatsch. Wet., Haarlem, 22 : 1-92, pis. i-io. BERTHELSON, O. 1962. Cheilostome Bryozoa in the Danian Deposits of East Denmark. Geol. Surv. Denmark, Copenhagen, II. Ser., 83 : 1-290, pis. 1-28. BLAINVILLE, H. M. D. DE. 1830. Zoophytes. Diet. Sci. Nat. Paris, 50 : 1-546. BORG, F. 1930. Moostierchen oder Bryozoen (Ectoprocta) . Tierwelt Dtschl., Jena, 17 : 25- 142, 193 figs. BRYDONE, R. M. 1929. Further notes on new or imperfectly known Chalk Polyzoa. London, 1-40, pis. 1-14. 1930. Further notes on new or imperfectly known Chalk Polyzoa, Pt. II (Vincularia, Onychocella, Rhagasostoma, Porina etc.), London, 39-60, pis. 15-29. 1936. Further notes on new or imperfectly known Chalk Polyzoa, Pt. Ill (Semieschara, Micropora, Cryptostoma) , London, 61-88, pis. 30-42. CANU, F. 1900. Revision des Bryozoaires du Cretac6 figures par d'Orbigny, Pt. II Cheilostomata. Bull. Soc. geol. Fr., Paris, (3), 28 : 334-463, pis. 4-6. 1913. Etudes morphologiques sur trois nouvelles families de Bryozoaires. Bull. Soc. geol. Fr. Paris, Ser. 4, 13 : 132-147, figs. i-io. & BASSLER, R. S. 1917. A synopsis of American early Tertiary cheilostome Bryozoa. Smiths. Inst. Bull. U.S. Nat. Mus. Washington, 96: 1-87, pis. 1-6. 1920. North American early Tertiary Bryozoa. Smiths. Inst. U.S. Nat. Mus. Bull. Washington, 106 : 1-879, pis. 1-162. 1922. Studies on the cyclostomatous Bryozoa. Proc. U.S. Nat. Mus., Washington, 61, 22 : 1-160, pis. 1-28. 1925. Les Bryozoaires du Maroc et de Mauritanie. Mem. Soc. Sci. nat. phys. Maroc. Rabat & Paris, 10 : 1-79, pis. 1-9. CHEETHAM, A. H. 1954. A new early Cretaceous cheilostome bryozoan from Texas. /. Paleont. Chicago, 28, 2 : 177-184, pi. 20. CORI, C. J. 1941. "Bryozoa". Handbuch der Zoologie, Berlin, 3, 2. Halfte, Liefg. 15-16; Teil 5 : 263-502, figs. 273-611. DAVID, L. 1952. A propos de Cellepora davaiacensis Lissajous, pseudo-Bryozoaire du Batho- nien maconnais. C. r. Seanc. Soc. geol. Fr., Paris, 13 : 267. 1960. Bibliographic des Bryozoaires jurassiques. Colloque sur le Lias fran9ais. Mem- oires du Bureau de Rech. geol. et minieres C. r. Congr. soc. sav. Paris et d. Departm. Sous- section de geologic, Chambery, 4 : 205-208. UPPER CRETACEOUS POLYZOA FROM COTENTIN 43 EICHWALD, E. VON. 1855. Beitrag zur geographischen Verbreitung der fossilen Thiere Russ- lands. Bull. Soc. Imp. Nat. Moscow, 28 (4) Polyzoa : 448-462. GABB, W. M. & HORN, G. H. 1860. Descriptions of new species of American Tertiary and Cretaceous fossils. /. Acad. nat. Sci. Philadelphia, (2) 4 : 375-404, pi. 69. 1862. Monograph of the fossil Polyzoa of the secondary and tertiary formation of North America. /. Acad. Nat. Soc. Philadelphia Ser. 2, 4 : 111-179, pis. 19-21. GOLDFUSS, G. A. 1826. In: Petrefacta Germaniae, Diisseldorf, I, Bryozoa: 22-39, 94-100, pis. 8-1 o. GRAY, J. E. 1848. List of specimens of British Animals in the Collections of the British Museum. Part I. Centroniae or Radiated Animals, i-xiii: 1-173. British Museum, London. GREGORY, J. W. 1893. On the British Palaeogene Bryozoa. Trans, zool. Soc. Lond., London, 13, VI : 219-279, pis. XXIX-XXXII. 1894. On some Jurassic species of Cheilostomata. Geol. Mag., London, (4), 1 : 61-64, 2 figs. 1896. The Jurassic Bryozoa. Catal. Brit. Mus. (Nat. Hist.), London: 1-239, pis. i-n. 1899-1909. The Cretaceous Bryozoa. Cat. of the foss. Bryoz. in the Dept. of Geol. Brit. Mus., London, I, I-XIV; 1-45, pis. I-XVII, 64 figs., 1899. II, I-XLVII: 1-346, pis. I-IX, 75 figs., 1909. HAGENOW, K. F. VON. 1839. Monographic der Riigen'schen Kreide-Versteinerungen, I, Phytolithen und Polyparien. Neues. Jb. Min. Geol. Paldont., Stuttgart, 7 : 253-96 (Poly- zoa: 263-296, pis. 4, 5). 1840. Monographic der Riigen'schen Kreideversteinerungen, II. Abt. : Radiarien und Annulaten nebst Nachtragen zur I. Abt. N. Jb. f. Min. Geol. Paldont., Stuttgart (Bryozoa: 639-649, pi. 9)- 1846. (in H. B. GEINITZ) Grundrifi der Versteinerungskunde. Dresden u. Leipzig, (Bryozoa: 586-631, pi. 23 b.). 1851. Die Bryozoen der Maastrichter Kreidebildung. Kassel, i-iu, pis. 1-12. HALL, J. 1874. Descriptions of Bryozoa and Corals of the Lower Helderberg Group. Rep. New York State Mus. Nat. Hist. Albany, 26 : 94-115. HAMM, H. 1881. Die Bryozoen des Maastrichter Obersenon. I. Die cyclostomen Bryozoen. 1-48, Berlin. HOFKER, J. 1959. Les Foraminiferes du Cretace sup6rieur du Cotentin. Colloque sur le cretac6 sup6rieur francais. C. r. Congr. Soc. sav. Paris et des Dept. tenu a Dijon en 1959, Paris : 369-397, text-figs. 1-66. JULLIEN, J. 1881. Remarques sur quelques especes de Bryozoaires Cheilostomiens. Bull. Soc. Zool. Fr. Paris, 6 : 163-168. 1886. Les Costulid6es, nouvelle Famille des Bryozaires. Bull. Soc. Zool. Fr. Paris, 11 : 601-620, pis. 17-20. 1888. Bryozoaires. Mission scientifique du Cap Horn 1882-1883, 6, Zoologie : 1-92, pis. 1-15. Paris. KADE, A. 1852. Die losen Versteinerungen des Schanzenberges bei Meseritz. Ber. d. Oberrealschule zu Meseritz. : 1-36, i pi. LAMARCK, J. B. P. A. DE M. DE. 1816. Histoire naturelle des Animaux sans vertebres . . . First Ed., 2 (1816) : 1-568. Paris. LAMOUROUX, J. 1821. Exposition methodique des genres de 1'Ordre des Polypiers. Paris 4 to VIII + 115 p., 84 pis. LANG, W. D. 1915. On some Uniserial cretaceous cheilostome Polyzoa. Geol. Mag., London (6), 2, Nr. 617 : 496-504, pi. 17. 1916. A Revision of the " Cribrimorph " Cretaceous Polyzoa. Ann. Mag. nat. Hist., London (8), July 1916 : 6-112. 1917. The genotypes of certain Polyzoan Genera. Geol. Mag., London, VI, IV (No. 634) : 169-174. 44 UPPER CRETACEOUS POLYZOA FROM COTENTIN LANG, W. D. 1921 & 1922. Catalogue of the fossil Bryozoy (Polyzoa) in Dept. Geol. Brit. Mus. (nat. Hist.), The Cretaceous Bryozoa (Polyzoa), part III 1921, The Cribrimorphs, part I, London 1921 I-CX : 1-269, pis. I-VIII; part IV 1922, The Cribrimorphs, part II 1922 : 1-404, pis. I-VIII. LARWOOD, G. P. 1962. The Morphology and systematics of some Cretaceous Cribrimorph Polyzoa (Pelmatoporidae) . Bull. Brit. Mus. (nat. Hist.), Geology, London, 6, i : 1-285, pis. 1-23, 132 text-figs. LECOINTRE, G. 1912. Sur quelques Bryozoaires nouveaux ou peu connus du Cenomanien du Mans. Bull. Soc. ge'ol. Fr., Paris, (4), 12 : 349-355, pis. 13, 14. LEVINSEN, G. M. R. 1925. Undersogelser over Bryozoerne i den danske Kridtformation (Nach dem Tode des Verf. herausgegeben. v. K. Br. Nielsen und Th. Mortensen). Danske kgl. vidensk. Selsk. Skr., naturvid. og math. Afd., Copenhagen, Raekke 8, 7, 3 : 283-445, pis. 1-8. LISSAJOUS, M. 1923. Etude sur la faune du Bathonien des environs de Macon. Trav. Lab. Geol. Univ. Lyon, Lyon, fasc. V, mem. 3 : 113-272, pis. XXV-XXXIII. LONGE, F. D. 1 88 1. On the relation of the Escharoid Forms of Oolithic Polyzoa to the Cheilo- stomata and Cyclostomata. Geol. Mag., London, II, VIII : 23-34, pi- H- LONSDALE, W. 1850. Polyzoa, In: Dixon's Geology and Fossils of the Tertiary and Cretaceous formations of Sussex. Bryozoa : 159-161, 268-322. London. MANTELL, G. A. 1844. Medals of creation, ist Ed. i : 1-456; 2 : 457-1016. London. MARSSON, TH. F. 1887. Die Bryozoen der weifien Schreibkreide der Insel Riigen. Paldont. Abh., Berlin, 4, i : 1-112, pis. i-io. MILLER, S. A. 1889. Molluscoidea in North American Geology and Palaeontology. Cincin- nati : 289-330, 684-685. NICKLES, J. M. & BASSLER, R. S. 1900. A Synopsis of American fossil Bryozoa. Bull. U.S. Geol. Survey, 173 : 1-663. Washington. D'ORBIGNY, A. D. 1839-1847. Voyage dans I'Amerique Meridionale. 5, 4 Zoophytes : 7-28, pis. 1-13. Paris. 1849. Description de quelques genres nouveaux de Mollusques bryozoaires. Rev. et Mag. de Zool. (2) 1 : 499-504. 1850. Prodrome de Paleontologie stratigraphique, Paris, 1 : 1-394; 2 : 1-428. 1851-1854. Paleontologie franfaise, Terrain Cretace, Paris, 5, Bryozoaires : 1-1192, pis. 600-800. PERGENS, E. 1889. Revision des Bryozoaires du Cretace figur6s par d'Orbigny, I. Cyclo- stomata. Bull. Soc. beige Geol. Pal. Hydr., Bruxelles, 3 (Mems) : 305-400, pis. 11-13. PRUD'HOMME, J. 1960. Le Genre Amphiblestrella, nouveau genre de Bryozoaires cheilostomes, Bull. Soc. geol. Fr., Paris, (7), 2 : 947-950, i fig. REGUANT, S. 1959. Algunas consideraciones sobre las ideas actuales acerca de la filogenia de los Briozoos ectoproctos. Publ. Inst. Biol. apl. Barcelona, XXX : 87-103. REUSS, A. E. 1872. Die Bryozoen und Foraminiferen des unteren Planers in Geinitz " Das Elbthalgebirge in Sachsen ". Palaeontographica, Stuttgart, 20 : 97-144, pis. 24-33. ROEMER, F. A. VON. 1840. Die Versteinerungen des Norddeutschen Kreidegebirge. iv + J 45 pp., 1 6 pis. Hannover. 1841. Die Versteinerungen des Norddeutschen Kreidegebirges. Hannover, (Polyzoa : 11-25, pl- 6). SILEN, L. 1942. Origin and development of the Cheilo-Ctenostomatous stem of Bryozoa. Zool. Bidr. Uppsala, 22 : 1-59, 64 figs. THOMAS, H. D. & LARWOOD, G. P. 1956. Some "Uniserial" Membraniporine Polyzoan genera and a new American Albian species. Geol. Mag. London, 93 : 369-376. 1960. The cretaceous species oiPyripora d'Orbigny and Rhammatopora Lang. Palaeonto- logy, London, 3, 3 : 370-386, pis. 60-62, text-figs. 1-4. UBAGHS, J. C. 1865. Die Bryozoen-Schichten der Maastrichter Kreidebildung, nebst einigen neuen Bryozoen-Arten aus der MaastrichterTuff kreide. Verh. naturhist. Ver. preufi. Rheinl. u Westph., Bonn, 22, 3, 2 : 31-62, pis. 2, 2a, 3. UPPER CRETACEOUS POLYZOA FROM COTENTIN 45 ULRICH, E. O. 1889. Contributions to the Micropalaeontology of the Cambro-Silurian Rocks of Canada. Pt. 2, no. 4. On some Polyzoa (Bryozoa) ... of Manitoba. Geol. &> Nat. Hist. Survey of Canada, Montreal : 27-56, pis. 8, 9. 1890. Palaeozoic Bryozoa, III. Geol. Survey of Illinois, 8 : 285-688, pis. 29-78, figs. 1-18. 1893. On Lower Silurian Bryozoa of Minnesota. Rep. Geol. & Nat. Hist. Surv. of Minnesota, Minneapolis, 3 : 96-332, pis. 1-28 (1895). & BASSLER, R. S. 1907. Bryozoa in Weller, Geol. Survey of New Jersey, Palaeontology, 4, (Cretaceous faunas) : 307-356, pis. 20-26. VIEILLARD, E. & DOLLFUS, G. 1875. Etude geologique sur les terrains cretac6s et tertiaires du Cotentin. Bull. Soc. linn. Normandie, Caen, (2) IX : 5-181. VINE, G. R. 1891. Notes on the Polyzoa and Microzoa of the Red Chalk of Yorkshire and Norfolk. Proc. Yorksh. geol. (polyt.) Soc., New Ser., Halifax, XI, pt. Ill : 363-396, pi. XVII. VOIGT, E. 1928. Neue artikulierte cheilostome Bryozoeii aus einem Kreidegeschiebe ober- senonen Alters von Cothen in Anhalt. Z. Geschiebeforsch., Berlin, 4, 3 : 105-114, 17 figs. 1930. Morphologische und stratigraphische Untersuchungen uber die Bryozoenfauna der oberen Kreide. Leopoldina, Halle, Ber. d. Kais. Leop. Dtsch. Akad. d. Naturforsch., 6 : 379-579, pis. i-39- 1942. Kreidebryozoen aus New Jersey (U.S.A.) unter A. E. Reuss' Originalen zu seiner Monographic der Bryozoen und Foraminiferen des Unteren Planers (1872) in H. B. Geinitz' " Das Elbthalgebirge in Sachsen ". Z. dtsch. geol. Ges., Berlin, 94 : 326-338, pis. 18-19. 1953- Revision von: H. Hamm " Die Bryozoen des Maastrichter Obersenon (1881). Mitt. geol. Staats. Inst. Hamb., Hamburg, 22 : 32-75, pis. 1-14. 1957- Bryozoen aus dem Kreidetuff von St. Symphorien bei Ciply (Ob. Maastrichtian). Bull. Inst. roy. Sci. nat. Belg., Bruxelles, 33, 43 : 1-48, pis. 1-12. 1959- Revision der von F. v. Hagenow 1839-1850 aus der Schreibkreide von Rugen veroffentlichten Bryozoen. Geologie, Z. f. d. Gesamtgeb. d. Geol. u. Min. etc., Berlin, Beiheft 25 : 1-62, pis. i-io. 1964. A Bryzoan Fauna of Dano-Montian age from Boryszew and Sochaczew in central Poland. Acta Palaeont. Polon., Warszawa, 9, 4 : 419-498, pis. 1-16. 1967. Oberkreidebryozoen aus den asiatischen Gebieten der USSR. Mitt. Geolog. Staat- sinst., Hamburg, 36: 5-95, pis. 1-34. 1968 Uber Immurtian bei fossilen Bryozoen, dargestellt an neuen Fundenaus der Oberen Kreide. Nachricht. Akad. Wiss., Gottingen; Math.-Phys. Kl. : 47-63, pis. 1-4. WALFORD, E. A. 1894. On Cheilostomatous Bryozoa from the Middle Lias. Q. Jl. geol. Soc., London, 50 : 79-83, pis. 5-7. WIESEMANN, G. 1963. Untersuchungen an der Gattung Beisselina Canu 1913 und ahnlichen Bryozoen (Maastrichtien, Danien, Montien). Mitt. geol. (St) Inst. Hamb., Hamburg, 32 : 5-70, pis. 1-12, 22 figs. EXPLANATION OF PLATES Some of the specimens of Bryozoa and Brachiopoda referred to and figured in this publication are housed in the collections of the British Museum (Natural History) and are prefixed by the letters D and BB respectively. PLATE i Petalopora sp. (p. 18) FIG. i. Worn branch with narrow ribs between the small mesopores. From the matrix of D.iSi, Upper Maastrichtian (Cotentin, Manche). X25. 0.49573. FIG. 2. Branch in similar condition to fig. i with larger mesopores and peristomes. Upper Maastrichtian, Port Filiolet (Cotentin, Manche). Presented by the author. X25. 0.49574. FIG. 3. Well preserved branch. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). X25- Voigt Collection, Hamburg, Nr. 3969. Crisisina carinata (Roemer) (p. 17) FIG. 4. Worn fragment with damaged peristomes, lateral view. From the matrix of D. 181 Upper Maastrichtian (Cotentin, Manche). X25. 0.49566. FIG. 5. Worn fragment, frontal view. Upper Maastrichtian, Port Filiolet (Cotentin, Manche). Presented by the author. X25. 0.49567. Multicrescis laxata d'Orbigny (p. 18) (See also pi. 4 figs. 4-6) FIG. 6. Worn fragment 0.49575 with encrusting Rosseliana thomasi n. sp. (0.47323) in matrix of Castanopora jurassica (Gregory) (D.iSo). Upper Maastrichtian (Cotentin, Manche). X20. FIG. 7. Holotype. Upper Maastrichtian, Sainte Colombe (Cotentin, Manche) in coll. d'Orbigny Nr. 8416, Musee d'Histoire Naturelle, Paris, x 12. FIG. 8. Holotype. X25. Thecidea papillata von Schlottheim (p. 40) FIG. 9. Pedicle (ventral) valve with growth-facet on the apex, with the cast of a small lamellibranch on the right. From the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche), attesting the origin from the " Craie a Thecidees " of the Cotentin-region. c. X 12. BB. 42981. FIG. 10. Small dorsal valve slightly damaged. From the matrix of D.iSi. c. X 12. 86.42982. Bull. Br. Mus. nat. Hist. (Geol.) 17, i PLATE i 8 GEOL. 17, I. 4 PLATE 2 Hagenowinella cf. incrassata (d'Orbigny) (p. 23) FIG. i. Part of the encrusting zoarium showing an avicularium in the left upper corner and some damaged zooecia in the proximal region. From the matrix of D. 181. Upper Maastrich- tian (Cotentin, Manche). x 20. 0.49583. FIG. 2. Worn specimen with broken ovicells. The deeply immersed horseshoe-like processes in the interior of the zooecia can hardly be seen. Upper Maastrichtian, Fresville (Cotentin, Manche). Presented by the author. X2o. 0.49845. FIG. 3. Unilaminar fragment with large opesia. Some ovicells and an avicularium in the lower row of zooecia are visible. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). x 20 Voigt Collection, Hamburg, Nr. 3909. Thyracella cf. tneudonensis (d'Orbigny) (p. 26) FIG. 4. Branched narrow fragment with an avicularium, for comparison with fig. 5. Upper Maastrichtian, Port Filiolet, (Cotentin, Manche). x 20. Voigt Collection, Hamburg, Nr. 3919- FIG. 5. Damaged fragment of the bilaminar zoarium. From the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). X2O. 0.49588. FIG. 6. Worn fragment, d'Orbigny Collection, Paris Nr. 8040, Musee d'Histoire Naturelle. labelled Meudon but likely Upper Maastrichtian, locality uncertain (? Neliou, Cotentin, Manche). X2o. " Membranipora unipora " Marsson (p. 19) FIG. 7. Fragment with hyperstomial ovicells and median pores on the gymnocyst of the undamaged zooecia. The minute marks of spines are hidden by recrystallization. From the rock matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). x 20. 0.49576. FIG. 8. Holotype, of Flustrellaria simplex d'Orbigny Collection, Nr. 8130, Musee d'Histoire Naturelle Paris, Upper Maastrichtian, N6hou (Cotentin, Manche). x 20. Bactrcllaria rugica Marsson (p. 24) FIG. 9. Small worn fragment showing the pits of the avicularia below the opesium. From the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). X2O. 0.49584. FIG. 10. The same specimen, showing the backside, x 20. FIG. ii. Fragment with well preserved avicularia. Upper Maastrichtian, Port Filiolet (Cotentin, Manche). Presented by the author, x 20. 0.49585. FIG. 12. Worn fragment in similar condition as figs. 9-10. Upper Maastrichtian, Port Filiolet (Cotentin, Manche). x 20. Voigt Collection, Hamburg, ^.3914. FIG. 13. Well preserved fragment with raised avicularia whose rostra are damaged. From chalk of a flint-boulder of Upper Maastrichtian age, Wulmstorf near Harburg-Hamburg, Northern Germany, x 20. Voigt Collection, Hamburg, Nr. 4146. FIG. 14. The same as fig. 13, lateral view, showing the long tubular peduncles of the avicu- laria. x 20. Bull. Br. Mus. nat. Hist. (Geol.) 17, i PLATE 2 PLATE 3 Stamenocella marginata d'Orbigny (p. 25) FIG. i . Fragment of the bilaminar zoarium in the matrix of Onychocella piriformis (Goldf uss) (D.iSi). Upper Maastrichtian (Cotentin, Manche). X 20. 0.47322. FIG. 2. Similar fragment. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Pre- sented by the author. X2O. 0.49586. FIG. 3. Holotype of Biflustra tesselata d'Orbigny. Upper Maastrichtian, Nehou (Cotentin, Manche). d'Orbigny Collection, Nr. 8053, Paris Musee d'Histoire Naturelle. X2o. FIG. 4. Holotype of Biflustra marginata d'Orbigny. Upper Maastrichtian, Sainte Colombe (Cotentin, Manche). d'Orbigny Collection, Nr. 8134, Paris Musee d'Histoire Naturelle. X 20. FIG. 5. Well preserved large fragment showing the different aspect and size of zooecia and avicularia. The basal region represents Flustrella convexa d'Orbigny. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). X20. Voigt Collection, Hamburg, Nr. 4128. Radtilopora minor n. sp. (p. 22) FIG. 6. Small worn bilaminar fragment from the matrix of D.iSi. Upper Maastrichtian, (Cotentin, Manche). X2o. 0.49579. FIG. 7. Fragment showing the gradual tapering of the zoarium. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author. X2O. 0.49580. FIG. 8. Another fragment. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Pre- sented by the author. X2O. 0.49581. FIG. 9. Broad fragment, most zooecia with ovicells on the right side with anen crusting Foraminifer. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author. X2o. 0.49582. FIG. 10. Holotype, showing the larger marginal zooecia at the edge of the zoarium. Upper Maastrichtian, St. Pietersberg near Maastricht (Netherlands), x 20. Presented by the author. 0.49844. Radulopora radula (Marsson) (p. 21) FIG. ii. Fragment of the thick bilaminar zoarium for comparison with Radulopora minor n. sp., showing the larger zooecia and avicularia and some fertile zooecia with broken ovicells. Upper Maastrichtian Chalk in drift, Tornesch near Elmshorn (Schleswig-Holstein, Northern Germany). X2O. Voigt Collection, Hamburg Nr. 3974. FIG. 12. Singular vibraculum, x 120, showing the thornlike process which is curved inwards. Horizon, locality and collection as for fig. n. Bull. Br. Mus. nat. Hist. (Geol.) 17, i PLATEs 11 8 PLATE 4 Amphiblestrella elegans (v. Hagenow) (p. 20) FIG. i. Small worn fragment from the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). x 20. 0.49577. FIG. 2. Holotype of Flustrella baculina d'Orbigny. Upper Maastrichtian, N6hou (Cotentin, Manche). d'Orbigny Collection, Paris, Nr. 8127, Musee d'Histoire Naturelle. x 20. FIG. 3. Branched fragment. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Pre- sented by the author. X2O. 0.49578. Rosseliana thomasi n. sp. (p. 31) (see also pi. i, fig. 6) FIG. 4. Holotype, showing the encrusting zoarium upon a branched fragment of Radulopora minor n. sp. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author, x 20. 1X49597. FIG. 5. Same specimen, part of the zoarium x 40, showing the form of the opesium and some fertile zooecia with ovicells. FIG. 6. Zoarium encrusting an echinoid fragment. Upper Maastrichtian, Port Filiolet (Cotentin, Manche). X2o. Voigt Collection, Hamburg, Nr. 3932. Floridina scut at a (Levinsen) (p. 31) FIG. 7. Part of encrusting zoarium, for comparison with Rosseliana thomasi n. sp. Upper Maastrichtian chalk boulder in drift, Tornesch near Elmshorn (Schleswig-Holstein, Northern- Germany). X40. Voigt Collection, Hamburg, ^.4145. Puncturiella cf . superba Brydone (p. 33) FIG. 8. Small unilaminar fragment in matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). X2O. 0.49598. FIG. 9. Unilaminar fragment. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author, x 20. D . 49599. FIG. 10. Unilaminar fragment, well preserved. Upper Maastrichtian, St. Pietersberg near Maastricht (Netherlands) . Presented by the author, x 20. D . 49600. Setniescharinella complanata d'Orbigny (p. 29) FIGS. II-I2. Small badly preserved unilaminar fragment from the matrix of D. 181. Upper Maastrichtian (Cotentin, Manche). x 20. 0.49595 and 49596. FIG. 13. Fragment with well preserved zooecia and distal avicularia. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author. X2o. 0.49847. FIG. 14. Small fragment showing one zooecium with ovicell in the right lower corner. Upper Maastrichtian, Chef du Pont (Cotentin, Manche) . x 20. Voigt Collection, Hamburg, Nr. 39i8b. FIG. 15. Holotype of Escharinella simplex d'Orbigny. Upper Maastrichtian, N6hou (Coten- tin, Manche). d'Orbigny Collection, Paris, Nr. 7942, Mus6e d'Histoire Naturelle. x 20. Bull. Br. Mus. nat. Hist. (Geol.) 17, i PLATE 4 PLATE 5 Onychocellaria caecitia (d'Orbigny) (p. 36) FIG. I. Fragment of a well preserved segment with some avicularia on the edges of the slightly flattened rods. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). x 20. Presented by the author. 0.49730. FIG. 2. Same specimen seen from the narrow side with two avicularia. x 20. FIG. 3. Another fragment with closed zooecia in the proximal region and larger opesia. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author, x 20. 0.49731- FIG. 4. Small fragment with avicularium. From the matrix of D.iSi. Upper Maastrich- tian, Chef du Pont (Cotentin, Manche). X2O. 0.49726. FIG. 5. Another fragment with avicularium in the middle row of zooecia. From the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). X2o. Photographed under water. 0.49727. FIG. 6. Broad worn fragment. From the matrix of D . 1 81 . Upper Maastrichtian (Cotentin, Manche). X2O. Photographed under water. 0.49767. Vincularia canalifera (von Hagenow) (p. 27) (See also pi. 8, figs. 9-12.) FIG. 7. Proximal part of a well preserved segment, showing more oval opesia and some zooecia with the characteristic openings for radicell filaments at the lower end. Upper Maas- trichtian, Md., St. Pietersberg near Maastricht (Netherlands). X 20. Presented by the author. D . 49840. FIG. 8. Fragment with some avicularia in the distal and proximal part, showing more semicircular opesia than fig. 7. Upper Maastrichtian, Md., St. Pietersberg near Maastricht (Netherlands). X2O. Presented by the author. 0.49841. FIG. 9. Worn fragment. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). x 20. Presented by the author. 0.49734. FIG. 10. Worn fragment. From the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). x 20. 0.49733. Onychocella pirifortnis (Goldfuss) (p. 5) FIG. ii. Fragment of the unilaminar frond with an avicularium in the lower left corner for comparison with fig. 12. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). x 20. Voigt Collection, Hamburg, Nr. 3573. FIG. 12. Holotype of Onychocella bathonica Gregory, erroneously labelled as " Bathonian Ranville (Calvados, France) ", but certainly from the Upper Maastrichtian of the Cotentin (Manche). X2O. D.iSi. Bull. Br. Mm. nat. Hist. (Geol.) 17, i PLATEs PLATE 6 Onychocella cf. cepha d'Orbigny (p. 34) FIG. i. Branched fragment with some avicularia from the matrix of D.iSi. Upper Maas- trichtian (Cotentin, Manche). x 20. 0.49604. FIG. 2. Fragment for comparison with fig. i. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). X2o. Presented by the author. 0.49606. Onychocella nysti (v. Hagenow) (p. 34) FIG. 3. Worn fragment from the matrix of D . 181. Upper Maastrichtian (Cotentin, Manche). X20. 0.49602. FIG. 4. Well preserved specimen with two ovicells. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). X2O. Presented by the author. 0.49603. Onychocella bellona (d'Orbigny) (p. 35) FIG. 5. Well preserved fragment with several ovicells. Upper Maastrichtian, Nehou (Coten- tin, Manche). x 20. d'Orbigny Collection, Paris, Nr. 7812. FIG. 6. Fragment from the Upper Maastrichtian from Chef du Pont (Cotentin, Manche) intermediate between fig. 5 and fig. 7. Presented by the author, x 20. 0.49608. FIG. 7. Worn fragment from the matrix of D. 181. Upper Maastrichtian (Cotentin, Manche). X40. 0.49607. Quadricellaria elegans d'Orbigny (p. 29) FIG. 8. Fragment of a segment. Upper Maastrichtian (Cotentin, Manche). From the matrix of D.iSi. x 20. 0.49591. FIG. 9. Another worn fragment. Upper Maastrichtian (Cotentin, Manche). From the matrix of D.iSi. x 20. 0.49592. FIG. 10. Fragment with a large avicularium. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). X4O. Voigt Collection, Hamburg, Nr. 4133. Bull. Br. Mus. nat. Hist. (Geol.) 17, i PLATE6 10 PLATE 7 Decurtaria cf. cornuta (Beissel) (p. 37) FIG. i. Zoarium with ancestrula and one ovicelled zooecium. From, the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). x 20. 0.49609. FIG. 2. Incomplete zoarium with some ovicelled zooecia. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). X2O. Voigt Collection, Hamburg, ^.4137. Pliophloea sp. (p. 37) FIG. 3. Young zoarium with ancestrula and one ovicelled zooecium. In the matrix of Castanopora jurassica (Gregory) D.iSo. Upper Maastrichtian (Cotentin, Manche). X2o. D.47324. Castanopora jurassica (Gregory) (p. 7) FIG. 4. Bilaminar fragment with two ovicelled zooecia. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author. X2O. 0.49724. FIG. 5. Same specimen from the other side, showing one zooecium with preserved frontal wall, x 20. FIG. 6. Fragment with some zooecia. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). x 20. Voigt Collection, Hamburg, ^.3585. FIG. 7. Part of a young zoarium with smaller zooecia. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author. X2O. 0.49725. " Membranipora " constricta d'Orbigny (p. 13) FIG. 8. Badly preserved specimen encrusting a Brachiopod. Aptian les Croutes (Yonne, France) ? Holotype, d'Orbigny Collection, Paris, Mus6e d'Histoire Naturelle, Nr. 5691. X2O. Micropora transversa (d'Orbigny) (p. 32) FIG. 9. Small worn fragment showing the two opesiular pores below the opesium. From the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). x 20. 0.49768. FIGS. lo-n. Two fragments of the Holotype. Upper Maastrichtian, Neliou (Cotentin, Manche). X2O. d'Orbigny Collection, Paris, Musee d'Histoire Naturelle, Nr. 7755. FIG. 12. Fragment, Upper Maastrichtian, Chef du Pont (Cotentin, Manche). X2O. Pre- sented by the author. 0.49769. Bull. Br. Mus. nat. Hist. (Geol.) 17, i PLATE7 PLATE 8 Beisselinopsis flabellat a (d'Orbigny) (p. 40) FIG. i. Adult zoarium. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). x 20. Voigt Collection, Hamburg, Nr. 3908. FIG. 2. Young worn zoarium. From the matrix of D . 181. Upper Maastrichtian (Cotentin, Manche). X2o. D. 49723. Beisselina striata (Goldfuss) (p. 39) FIG. 3. Worn fragment. From the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). x 20. 0.49721. FIG. 4. Worn fragment for comparison with fig. 3. Upper Maastrichtian, Md., Geulem near Berg, Geul-valley near Maastricht (Netherlands). Presented by the author, x 20. D. 49722. Frurionella fer tilts n. sp. (p. 38) FIG. 5. Holotype, well preserved fragment with ovicelled zooecia. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author, x 20. 0.49610. FIG. 6. Fragment showing zooecia with damaged ovicells. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author. X2O. 0.49612. FIG. 7. Worn fragment, similar to fig. 6. From the matrix of D . 181. Upper Maastrichtian (Cotentin, Manche). x 20. 0.49611. Frurionella europaea (Voigt) (p. 39) FIG. 8. For comparison with Frurionella fertilis n. sp. Upper Maastrichtian, Chef du Pont (Cotentin, Manche). Presented by the author, x 20. 0.49720. Vincularia canalifera (v. Hagenow) (p 27) (See also pi. 5, figs. 7-10) FIG. 9. Small fragment. From the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). x 20. 0.49735. FIG. 10. Badly preserved branched fragment showing an avicularium in the middle of the right side. From the matrix of D.iSi. Upper Maastrichtian (Cotentin, Manche). x 20. 0.49736. FIG. 1 1 . Fragment showing an avicularium at the right upper corner. Upper Maastrichtian Md., Geulem near Berg, Geul-valley near Maastricht (Netherlands) . x 20. D . 49842. E. Voigt Collection. FIG. 12. Part of a branched stem, for comparison with fig. 9, showing relatively small zooecia and opesia. Upper Maastrichtian, locality as fig. n. 0.49732. Vincularia concinna (d'Orbigny) (p. 28) FIG. 13. Holotype. Branched fragment for comparison with Vincularia canalifera von Hagenow. Upper Maastrichtian, Nehou (Cotentin, Manche). x 20. d'Orbigny Collection, Paris, Musee d'Histoire Naturelle, Nr. 7756. Vincularia flexuosa (d'Orbigny) (p. 28) FIG. 14. Holotype. Fragment for comparison with Vincularia canalifera von Hagenow. Labelled as Santonian Venddme, Loir et Cher, France, x 20. d'Orbigny Collection, Paris, Mus6e d'Histoire Naturelle, Nr. 7752. Bull. BY. Mus. nat. Hist. (Geol.) 17, i PLATES PRINTED IN GREAT BRITAIN BY ADLARD & SON LIMITED BARTHOLOMEW PRESS, DORKING THE CAUDAL SKELETON IN MESOZOIC ACANTHOPTERYGIAN FISHES COLIN PATTERSON BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 17 No. 2 LONDON: 1968 THE CAUDAL SKELETON IN MESOZOIC ACANTHOPTERYGIAN FISHES BY COLIN PATTERSON British Museum (Natural History) Pp. 47-102; 28 Text-figures BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 17 No. 2 LONDON: 1968 THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted in 1949, is issued in five series corresponding to the Departments of the Museum, and an Historical series. Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year. In 1965 a separate supplementary series of longer Papers was instituted, numbered serially for each Department. This paper is Vol. 17, No. 2 of the Geological (Palaeontological) series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals. World List abbreviation: Bull. Br. Mus. nat. Hist. (Geol.) Trustees of the British Museum (Natural History) 1968 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) Issued 19 November, 1968 Price THE CAUDAL SKELETON IN MESOZOIC ACANTHOPTERYGIAN FISHES By COLIN PATTERSON CONTENTS Page I. INTRODUCTION ......... 49 II. SYSTEMATIC DESCRIPTIONS ....... 50 Order Ctenothrissiformes ....... 50 Order Beryciformes . . . . . . . . 56 Suborder Polymixioidei . . . . . . 56 Suborder Dinopterygoidei ...... 62 Suborder Berycoidei ....... 67 Order Lampridiformes ....... 78 Order Salmoniformes, Suborder Myctophoidei . . . 81 III. DISCUSSION .......... 83 (a) The relationships of Ctenothrissiformes, Myctophoidei and Beryciformes ........ 83 (b) The origin of Perciformes ...... 87 (c) Intermediate groups ....... 96 IV. CONCLUSIONS .......... 98 V. REFERENCES .......... 100 VI. ABBREVIATIONS USED IN FIGURES ...... 102 SYNOPSIS This paper contains descriptions of the caudal skeleton in all Ctenothrissiformes and Mesozoic acanthopterygians, comparisons with living relatives being made where possible, and a brief account of the caudal skeleton in Mesozoic Myctophoidei. These groups have a basically similar caudal skeleton with the first ural and pre-ural centra fused, the second ural centrum free, six hypurals, three epurals, a stegural and a second uroneural. Caudal scutes persist in Ctenothris- siformes and primitive Myctophoidei and there is a single urodermal in some Cretaceous myctophoids. The bearing of caudal structure on the origin and early evolution of acanthopterygians is discussed and variations in the second pre-ural neural spine are shown to be important. This structure is primitively short and slender: two opposite developments from this condition are elongation of the spine to support procurrent fin-rays (as in Polymixioidei) and reduction to a low crest (as in Berycoidei and Perciformes). Wherever a full second pre-ural neural spine occurs in perciform and higher groups it is produced secondarily by fusion with the first epural. The evidence of the caudal skeleton does not support a polyphyletic origin of Perciformes from different groups of Beryciformes. The Berycoidei appear to be the beryciform group most closely related to the Perciformes. The Danian Bathysoma is shown to be a lampridiform : the lampridiforms probably arose from near Aipichthys and Pharmacichthys. The Zeiform.es are close relatives of the Berycoidei. I. INTRODUCTION IN an earlier paper (Patterson 1964) I gave detailed accounts of the skeletal anatomy of the known Mesozoic acanthopterygian fishes (see also Patterson 1967) but without describing the caudal skeleton. This was omitted because when the bulk GEOL. 17, 2 5 50 THE CAUDAL SKELETON IN of the work was done there existed no comprehensive description or even terminology of the caudal skeleton of living teleosts. But during the last few years, following from the work of Gosline (1960), the caudal skeleton has become recognized as an important structure in tracing the relationships of teleost fishes and we now have both a satisfactory terminology (Gosline 1960; Nybelin 1963; Monod 1967; Patterson 1968) and a reasonable idea of the composition of the caudal skeleton in most teleostean groups (Gosline 1960, 1961, ig6ia, 1963, 1965; Norden 1961; Rosen 1962, 1964; Greenwood, Rosen, Weitzman & Myers 1966; Monod 1967; Greenwood 1967; Weitzman 1967). In the light of this new information it seemed that an investigation of the caudal skeleton in Mesozoic acanthopterygians might prove a useful check on hypotheses of the origins of acanthopterygians and par- acanthopterygians (Greenwood et al. 1966) . Also included are accounts of caudal structure in the Ctenothrissiformes and, more briefly, in the Mesozoic Myctophoidei. The terminology used here is that of Nybelin (1963) with the distinction between " urodermal " and " uroneural " introduced by Patterson (1968) and the addition from Monod (1967) of the terms " parhypural ", for the haemal arch of the first pre-ural centrum, the terminal structure perforated by the caudal vein and artery and which bears the hypurapophysis (Nursall 1963), and " stegural " for the paired structure which articulates with the dorso-lateral surface of the first pre-ural centrum and has a slender shaft extending postero-dorsally, lateral to the nerve cord. Monod considers the stegural to be an element sui generis, not a compound structure. In my opinion (Patterson 1968), the stegural represents the first uroneural fused with neural arch material from the first pre-ural and first ural vertebrae. The term stegural is nevertheless useful, since it obviates repetition of the cumbersome term " first uroneural fused with the neural arches of the first ural and pre-ural centra ". The material used is mainly in the collections of the British Museum (Natural History) : these specimens are referred to by a registered number without prefix or with the prefix " P ". Specimens from the American University, Beirut, the American Museum of Natural History, New York, the Royal Scottish Museum, Edinburgh, and the United States National Museum, Washington, are referred to with the prefixes " AUB ", " AMNH ", " RSM " and " USNM " respectively. The illustrations of fossil species are not reconstructions but camera lucida drawings of single specimens. I am grateful to Drs. P. H. Greenwood and D. E. Rosen, who have read and criticized parts of this paper in manuscript, and for the loan of specimens to Prof. T. Raven, American University, Beirut, Drs. C. D. Waterston and R. S. Miles, Royal Scottish Museum, Dr. Bobb Schaeffer, American Museum of Natural History, and Drs. D. H. Dunkle and D. M. Cohen, United States National Museum. II. SYSTEMATIC DESCRIPTIONS Order CTENOTHRISSIFORMES (Patterson 1964 : 218) According to Marshall (1961) the living Macristium chavesi is a surviving cteno- thrissiform (see also Greenwood, Rosen, Weitzman & Myers 1966; Patterson 1967 : 86). Unfortunately, the only extant specimen of Macristium is a post-larval MESOZOIC ACANTHOPTERYGIAN FISHES 51 individual and the caudal skeleton is not ossified. But there appear to be six hypurals, as in Ctenothrissiformes. Berry & Robins (1967) have described a second macristiid, Macristiella perlucens, known only by a single post-larval specimen. In this fish there are six hypurals, as in Ctenothrissiformes, but Berry & Robins think it unlikely that the Macristiidae are close to the Ctenothrissiformes. Family AULOLEPIDIDAE Patterson (1964 : 247) Genus PATEROPERCA Smith Woodward (1942 : 543) The type and only species, P. libanica Smith Woodward from the Middle Ceno- manian of Hajula, Lebanon, is known only by two specimens, AUB 108906 (the holotype) and AUB 108904. In 108904 the caudal region is badly preserved and yields no useful information, but in the holotype the caudal skeleton is very well preserved (Fig. i). The second pre-ural centrum (pu2] has a fully developed neural spine (npu2) and an autogenous haemal arch. It is not possible to see with certainty whether the haemal arch of the third pre-ural centrum (Pu3) is autogenous. The first pre-ural and first ural centra are fused (pu i + ui) , these fused centra bearing the parhypural (ph) and the first and second hypurals (hi, h2) in the normal way. The first hypural is rather slender, only a little broader than the parhypural and u2 npu2 pu1+u1 FIG. i. Pateroperca libanica Smith Woodward. Caudal skeleton of the holotype, AUB 108906, standard length 86 mm., Middle Cenomanian, Hajula, Lebanon. For explanation of lettering see p. 102. Arrows mark the outermost (unbranched) principal fin-rays. 52 THE CAUDAL SKELETON IN the second. Dorsally, the fused first pre-ural and ural centra bear a large stegural (st) and there is a second uroneural (d2) ending, as usual, in front of the base of the first unbranched principal ray of the caudal fin. There are three epurals (ei j), the third small and short. There is a small independent second ural centrum (u2) bearing three hypurals (hj 5) decreasing in size upwards, and above these there is a small sixth hypural (h6) which probably failed to make contact with the ural centrum. The foremost procurrent rays of the caudal fin, which are unsegmented lepidotrichia, not spines, are inserted in front of the neural and haemal spine of the third pre-ural centrum. The uppermost principal ray of the fin (unbranched) is preceded by nine rays, only the last two or three segmented, and the lowermost by seven rays, the last two segmented. FIG. 2. Aulolepis typus Agassiz. Caudal skeleton of 47932, standard length c. 165 mm., Lower Chalk, Burham, Kent. /. r, foremost procurrent fin-ray (displaced) ; for explana- tion of other lettering see p. 102. The upper hypurals are displaced ventrally, the epurals and second uroneural are missing. MESOZOIC ACANTHOPTERYGIAN FISHES 53 Genus AULOLEPIS Agassiz (1844 : 109) In the type and only species, A. typus Agassiz from the Upper Cenomanian of the English Chalk, only two specimens show any details of the caudal skeleton, 4033 (the holotype, Fig. 3) and 47932 (Fig. 2), and in both the bones are disturbed to some extent. The caudal skeleton of Aulolepis agrees with that of Pateroperca in most respects : fusion of the first pre-ural and first ural centra (pui + ui) , presence of a free second ural centrum (u2], three epurals (ei 3), six hypurals (hi 6) of which the first and the third are largest, an autogenous haemal arch on the second pre-ural centrum (pu2), a stegural (st) and a free second uroneural (d2). The haemal arch of the third pre-ural centrum (pu^] appears to be autogenous in 47932 but fused with the centrum in 4033. Aulolepis differs clearly from Pateroperca in having the neural spine of the second pre-ural centrum developed as a low, broad crest (npu2) whose hind edge fits between the front edges of the stegurals. In my earlier des- criptions of Aulolepis and Pateroperca (Patterson 1964 : 247) I discussed the difficulty of separating the two genera and considered the possibility that P. libanica is merely npu2 d2 pu1+u1 FIG. 3. Aulolepis typus Agassiz. Caudal skeleton of the holotype, 4033, standard length c. 135 mm., Lower Chalk, Lewes, Sussex. For explanation of lettering see p. 102. The epurals and upper hypurals are displaced ventrally, the stegural is displaced forwards. 54 THE CAUDAL SKELETON IN a species of Aulolepis, but the presence on the second pre-ural centrum of a fully developed neural spine in Pateroperca and of a low crest in Aulolepis clearly separates the two genera. The caudal of 47932 shows one other point of interest, a slender, elongated caudal scute (/. s, Fig. 2) in front of the upper lobe of the fin. There is no conclusive evidence of a similar structure in front of the lower lobe of the fin, but it was probably present since there is a caudal scute in front of each lobe of the fin in Ctenothrissa (see below) and in most living teleosts which still retain these structures (Elops, Tarpon, Albula, Aulopus and Chanos, Gosline 1965 : 192; Argen- tina, Weitzman 1967 : 532). There is no sign of caudal scutes in either specimen of Pateroperca, but again it is probable that they were present. No specimen of Aulol- epis has the fin-rays of the caudal fin well preserved but 4033 shows that the foremost procurrent rays of the lower lobe articulated with the haemal spine of the third pre-ural centrum, and both 4033 and 47932 show that these foremost rays were true spines. Family CTENOTHRISSIDAE Smith Woodward (1901) Genus CTENOTHRISSA Smith Woodward (1899 : 490) In Ctenothrissa, the only genus of the Ctenothrissidae, the caudal skeleton and fin are well exposed in the two species from the Cenomanian of the Lebanon, C. vexillifer e3 u2 pu1+u1 FIG. 4. Ctenothrissa signifer Hay. Caudal skeleton of P. 47524, standard length 83 mm., Middle Cenomanian, Hajula, Lebanon. For explanation of lettering see p. 102. Arrows mark the outermost (unbranched) principal fin-rays. MESOZOIC ACANTHOPTERYGIAN FISHES 55 (Pictet 1850), the type species, and C. signifer Hay (1903) (Fig. 4), and the caudal skeleton is preserved in one or two specimens of C. microcephala (Agassiz) (Fig. 5) and C. radians (Agassiz) from the Upper Cenomanian of the English Chalk. There seem to be no significant differences in caudal anatomy between these four species. As in Aulolepididae, there is a free second ural centrum (u2), the first ural and first pre-ural centra are fused (pui + ui], there are six hypurals (hi 6), the first and third the largest and the sixth failing to articulate with the second ural centrum, there are three epurals (ei j), a stegural (st) and a second uroneural (2) and the haemal arch of the second pre-ural centrum (pu2) is autogenous. The haemal arch of the third pre-ural centrum (pu>3) appears partially or completely fused to the centrum. As in Aulolepis, but in contrast to Pateroperca, the neural spine of the second pre-ural centrum (npu2) is represented by a broad crest, about half as high as the preceding spine. There is a slender caudal scute (/. s.) in front of both the upper and lower lobes of the caudal fin. In C. vexillifer and C. signifer, the only species in which the caudal fin-rays are well preserved, both the upper and lower principal rays are preceded by six rays, the last two segmented, and the foremost fin-rays are inserted on the neural and haemal spines of the third pre-ural centrum. e3 e1 npu2 2mm FIG. 5. Ctenothrissa microcephala (Agassiz). Caudal skeleton of 49881, standard length c. 1 15 mm., Lower Chalk, zone of Holaster subglobosus, Reigate, Surrey. For explanation of lettering see p. 102. The upper hypurals are displaced ventrally. Of the two elements labelled ? h6, that on the left may be the base of a fin-ray. 5 6 THE CAUDAL SKELETON IN In these small species from the Lebanon, the foremost procurrent caudal rays are unsegmented lepidotrichia, with separate right and left halves, but in the larger species from the English Chalk they appear to be true spines. Order BERYCIFORMES Suborder POLYMIXIOIDEI Patterson (1964 : 433) Family POLYMIXIIDAE Gill (1862) The only living genus of this family and suborder is Polymixia. The caudal skeleton of Polymixia nobilis has been figured by Regan (1911, fig. i) and briefly discussed by Gosline (1961 : 14). The specimen illustrated here (Fig. 6) is almost certainly that used by Regan. The haemal arches of the second and third pre-ural centra (pu2, pus) are autogenous, the second pre-ural centrum has a fully developed neural spine (npu2) , the first pre-ural and first ural centra are fused (pui + ui) and there is a free second ural centrum (14,2). There is a large stegural (st) and a free second uroneural (d2). There are three slender epurals (ei j) and six hypurals (hi 6) of which the first and the fourth are the largest, the third being excavated posteroventrally to give a notch between the hypurals supporting the upper and lower lobes of the fin. The second ural centrum has a long posterior process and makes contact with the sixth hypural. The first rays of the caudal fin articulate with the neural and haemal spines of the third pre-ural centrum. In the upper lobe of the fin the first unbranched principal ray is preceded by four true spines and two e1-3 npu2 pu1+u1 u2 FIG. 3mm 6. Polymixia nobilis Lowe. Caudal skeleton of a dried skeleton, 1895.5.28.1, standard length 220 mm., Madeira. For explanation of lettering see p. 102. MESOZOIC ACANTHOPTERYGIAN FISHES segmented rays, in the lower lobe by three spines and two segmented rays, caudal skeleton of P. japonicus does not differ from that of P. nobilis. 57 Genus BERYCOPSIS Dixon (1850 : 372) In the type species, B. elegans Dixon from the Upper Cenomanian and Turonian of the English Chalk, the caudal skeleton is well preserved in 25881 (Fig. 7) and P. 6465. The caudal skeleton of B. elegans agrees with that of Polymixia in almost every detail except that the third hypural (Aj) is not excavated postero-ventrally so that it is as large as the fourth, the haemal arch of the third pre-ural centrum, probably autogenous in P. 6465, is fused with the centrum in 25881, and the third and fourth hypurals, separate in 25881 (standard length c. 13 cm.), are completely fused in the larger P. 6465 (standard length c. 16 cm.). In 25881 the lateral surface of the second ural centrum is covered by a small, triangular plate (x, Fig. 7). P. 6465 is not sufficiently well preserved to see whether this plate is present. The plate appears to be part of the caudal skeleton, not a dermal element which has become apposed to the centrum post mortem. The only record of a similar structure is in the living myctophoid Synodus foetens, where Hollister (193 ja, figs. 5, 12) found a " uroneural " in this position, ventral to the normal uroneurals. The foremost h6 h5 npu2 pul+ul FIG. 7. Berycopsis elegans Dixon. Caudal skeleton of 25881, standard length c. 130 mm., Chalk, Sussex, x, unidentified plate covering second ural centrum; for explanation of other lettering see p. 102. 5 8 THE CAUDAL SKELETON IN unsegmented caudal fin rays of B. elegans are soft rays, not spines as they are in Polymixia. Of P. 1047 1, a small fish lacking the head, I rashly stated that it is " certainly B. elegans " (Patterson 1964 : 278), but on re-examining this specimen I find that the scales are larger than in B. elegans, the dorsal and anal spines are longer, there are five anal spines (not four as stated earlier) and in the caudal skeleton the neural spine of the second pre-ural centrum is much reduced. This specimen is clearly not B. elegans and the reduced neural spine of the second pre-ural centrum shows that it is not even a polymixioid, but there is insufficient evidence to decide on its real position. The second species of Berycopsis, B. germanus (Agassiz) from the Campanian of Westphalia, shows nothing in the caudal skeleton to distinguish it from B. elegans except that the haemal arch of the third pre-ural centrum is clearly autogenous and there is no sign of fusion between the third and fourth hypurals in the two specimens where this region is visible. B. germanus is so preserved that it shows the details of the caudal fin-rays much better than B. elegans. The foremost rays of the caudal fin are arranged asymmetrically, the upper rays articulating with the neural spine of the second pre-ural centrum, the lower with the haemal spine of the third pre-ural centrum. The first principal ray (unbranched) of the upper lobe is preceded by four unsegmented and one segmented ray, the lower by three unsegmented and one segmented ray. It is not possible to see whether the foremost rays are spines (as in Polymixia) or soft rays, as in B. elegans and Omosoma. Genus OMOSOMA Costa (1857 : Io6 ) Having now had the opportunity to make a direct comparison between the holo- types of Omosoma pulchellum (Davis 1887) (RSM 1891.59.72) and 0. intermedium Smith Woodward (1901) (48112) I find that the two are conspecific, the median fin counts (D V, 28-30; A III-IV, 24) being the same in both. The apparent differences in proportions of the two (Smith Woodward 1901 : 420) are due to longitudinal compression of the holotype of 0. pulchellum, the holotype of 0. intermedium showing the true proportions of the fish. In my earlier description of Omosoma (Patterson 1964 : 374) the median fin counts given for 0. pulchellum (D IV-V, 35 ; A III-IV, 29) were based on distorted specimens of 0. sahelalmae. 0. intermedium is therefore a synonym of 0. pulchellum. None of the specimens of Omosoma available, belonging to 0. sahelalmae Costa (the type species) and 0. pulchellum (Davis), from the Upper Santonian of Sahel Alma, Lebanon (Ejel & Dubertret 1966, have recently found evidence of the precise age of these beds), has the caudal skeleton sufficiently well preserved to be illustrated, but all give a picture of a caudal skeleton which does not differ significantly from that of Berycopsis. The foremost rays of the caudal fin are arranged asymmetrically, those of the upper lobe articulating with the neural spine of the second pre-ural centrum, those of the lower lobe with the haemal spine of the third pre-ural centrum, just as in B. germanus. In both the lobes of the fin there are 4 unsegmented lepidotrichia (not spines) and two segmented rays in front of the principal rays. MESOZOIC ACANTHOPTERYGIAN FISHES 59 It is impossible to see whether the haemal arch of the third pre-ural centrum is autogenous. Genus PYCNOSTERINX Heckel (1849 : 337) Pycnosterinx is known by four species, all from the Upper Santonian of Sahel Alma, Lebanon. Figure 8 shows the caudal skeleton of P. russeggerii Heckel, the type species: it is very similar to those of the other polymixiids and differs from Polymixia (Fig. 6) only in having the postero-ventral part of the third hypural com- pletely ossified, so that there is no gap between the hypurals supporting the upper and lower lobes of the fin. As in Polymixia the neural and haemal spines are strongly inclined and the epurals slender. As in Berycopsis and Omosoma, but in contrast to Polymixia, the foremost caudal fin-rays are arranged asymmetrically, the upper ones articulating with the neural spine of the second pre-ural centrum, the lower with the haemal spine of the third pre-ural centrum. In both the upper and lower lobes of the fin the principal rays are preceded by three unsegmented soft rays and three segmented rays. P. discoides Heckel and P. gracilis Davis do not differ in caudal structure from P. russeggerii except that in P. gracilis there appear to be only five procurrent rays. I have seen no specimens of P. dubius Davis in which the caudal skeleton and fin are well preserved, but in the holotype of Pycnosterinx latus Davis (1887 : 534, pi. 27, fig. 2), RSM 1891.59.77 (referred to as a "frag- mentary fish of indeterminable genus " by Smith Woodward 1901 : 395), the scales have the same spinous surface as in P. dubius (Patterson 1964 : 380) and it is probable that the specimen is a large, distorted individual of P. dubius, although the state of u2 pu1>u1 2mm FIG. 8. Pycnosterinx russeggerii Heckel. Caudal skeleton of 47820, standard length 78 mm., Upper Santonian, Sahel Alma, Lebanon. For explanation of lettering see p. 102. 6o THE CAUDAL SKELETON IN preservation of this and the holotype of P. dubius is such that this cannot be demon- strated with sufficient certainty to synonymize the two species. P. latus appears to have one or two more anal fin-rays and caudal vertebrae than the holotype of P. dubius. The caudal skeleton is completely exposed in the holotype of P. latus and does not differ from that of P. russeggerii in any way, but both in this specimen and in P. dubius the foremost fin-rays are true spines, a difference from the other species of Pycnosterinx and a resemblance to the living Polymixia. Genus HOMONOTICHTHYS Whitley (1933 : 146) All three species of this genus occur in the Upper Cenomanian Lower Turonian of the English Chalk. In the type species, H. dorsalis (Dixon), only one specimen, 43575, shows anything of the caudal skeleton and here it is only possible to see that there was a fully developed neural spine on the second pre-ural centrum and that the foremost rays of the upper caudal lobe are true spines, articulating with the neural spine of the third pre-ural centrum. In these last two features H. dorsalis e2 e3 d2 2mm FIG. 9. Homonotichthys pulchellus (Dixon). Caudal skeleton of P. 11112, standard length c. no mm., Lower Chalk, Amberley, Sussex. For explanation of lettering see p. 102. MESOZOIC ACANTHOPTERYGIAN FISHES 61 resembles Polymixia and differs from other Cretaceous polymixiids except Pycnosterinx dubius. In H. pulchellus (Dixon) the caudal skeleton is preserved in the holotype, 25886, and in two specimens only recently recognized as belonging to this species, P. 11112 (Fig. 9) and P. 10639: the latter are about no and 120 mm. respectively in standard length, the largest examples of this species yet recorded. The caudal skeleton of P.IIH2 (Fig. 9) is abnormal in the partial doubling of the neural spine of the second pre-ural centrum and in having the second epural smaller than the third. The caudal skeleton of H. pulchellus is very like that of Polymixia, with autogenous haemal arches on the second and third pre-ural centra (pu2, pu3) six hypurals (hi 6), three slender epurals (ei 3), a free second ural centrum (u2), a stegural (st) and a second uroneural (d,2}. The postero- ventral corner of the third hypural is truncated, but less so than in Polymixia. Broad flanges on the anterior margin of the neural spines of the second and third pre-ural centra are characteristic of H. pulchellus. All three specimens show that the foremost caudal rays are spines, as in H. dorsalis and Polymixia. In P . 10639 the caudal rays articulate with the neural and haemal spines of the third pre-ural centrum, as in Polymixia. In P . 11112 where the second pre-ural spine is abnormal, the foremost upper caudal rays articu- late with the neural spine of the second pre-ural centrum. The lowermost principal caudal ray (unbranched) is preceded by four spines and three segmented rays. P.iiii2 shows the anal fin of H. pulchellus, previously unknown. The fin contains five spines, the first very small and the fifth the longest and thickest, and about eleven soft rays. The fifth anal spine is equal in length to the longest dorsal spine, just over one-quarter of the maximum depth of the trunk. H. rotundus (Smith Woodward) is known only by the holotype, P . 315, and P . 5682. P . 5682 shows most of the details of the caudal skeleton : there is nothing to distin- guish it from the other polymixiids described here except that there is no flange on the anterior margin of the neural spine of the second pre-ural centrum as there is in H. pulchellus. P. 315 shows that the foremost caudal rays are soft rays, longitudinally divided, not spines as they are in H. dorsalis, H. pulchellus and Poly- mixia. These two points may be added to others (Patterson 1964 : 299) indicating that this species is incorrectly placed in Homonotichthys, but more specimens are necessary before its true position can be decided. Family SPHENOCEPHALIDAE Patterson (1964 : 383) The only member of this family is Sphenocephalus fissicaudus Agassiz from the Campanian of Westphalia. Of the six specimens in the British Museum (Natural History), three, P. 2100 (Fig. 10), P. 8772 and P. 9059 have the caudal skeleton well preserved. As in Polymixiidae, the second pre-ural centrum has a fully developed neural spine (npu2], there is a free second ural centrum (1*2), a stegural (st), a second uroneural (d2), and six hypurals (hi 6). As in Polymixia, the foremost procurrent rays articulate with the neural and haemal spines of the third pre-ural centrum. In contrast to all Polymixiidae there are only two epurals (ei, 2), the first curved forwards proximally and with a gap between it and the neural spine of the second 62 THE CAUDAL SKELETON IN FIG. 10. Sphenocephalus fissicaudus Agassiz. Caudal skeleton of P. 2100, standard length 92 mm., Campanian, Sendenhorst, Westphalia. For explanation of lettering see p. 102. pre-ural centrum, there is a wide gap between the upper and lower hypurals, and there is a large number of procurrent rays, apparently nine in each lobe, six unsegmented soft rays and three segmented in the upper, five unsegmented and four segmented in the lower. Suborder DINOPTERYGOIDEI Patterson (1964 : 434) This suborder contains four monotypic Uppper Cretaceous familes, probably not closely related. As the caudal skeleton in the type family, the Dinopterygidae, is poorly known, the best known family, the Aipichthyidae, will be described first. Family AIPICHTHYIDAE Patterson (1964 : 303) Genus AIPICHTHYS Steindachner (1860 : 763) I have seen no material of the type species, A. pretiosus Steindachner from the Lower Cenomanian of Dalmatia, and the caudal region is not preserved in A. MESOZOIC ACANTHOPTERYGIAN FISHES 63 nuchalis (Dixon) from the English Chalk. The other two species, A. minor (Pictet) and A. velifer Smith Woodward, both from the Cenomanian of Hakel, Lebanon, are well represented in the British Museum (Natural History). They show a remarkable range of variation in the structure of the caudal skeleton. In most speci- mens of Aipichthys the distal parts of the hypurals, epurals and uroneurals are obscured by the deeply cleft bases of the caudal fin-rays, but in P. 82 (A. minor, Fig. uA) the fin-rays are displaced, exposing these bones. In this specimen and in all others the neural and haemal spines of the third pre-ural centrum (pu3) are broad and elongate and the first procurrent rays of the fin articulate with them. The haemal arches of the second and third pre-ural centra are autogenous. In contrast to all Polymixioidei, the neural spine of the second pre-ural centrum (npu2) is normally only about half as long as its predecessor, which makes contact with the first epural distally. Though somewhat expanded, this neural spine is more like those of Elops (Nybelin 1963, figs, i, 4) and Nematonotus (Fig. 25) than the shorter, broader spine in Aulolepis and Ctenothrissa (Figs. 2, 4, 5) : it does not resemble the very low crest on this centrum in Berycoidei and percoids. In one specimen of Aipichthys minor, P. 6183 (Fig. nC), the neural spine of the second pre-ural centrum is fully developed and supports epaxial fin-rays, as in Polymixioidei. This is clearly an individual variation, comparable to those found in certain individuals of Monocentris (Fig. 14) and Siniperca (Fig. 26), and is recognizable as such by the gap between the spine and the proximal part of the first epural, which is filled by flanges from the posterior face of the spine and from the anterior face of the epural. In all specimens of Aipichthys there are three epurals (ei 3) and the first ural and pre-ural centra are fused (pui + ui). There is normally a free second ural centrum (u2, Figs. 1 1 A, C), but in occasional individuals, like the large specimen shown in Fig. nB, the second ural centrum is fused into the preceding compound centrum (pui + ui + U2), although the line of fusion is visible in transparency under xylene. There is always a stegural (st) and a second uroneural (d2). In A. minor there are normally six hypurals (hi 6, Figs. nB, C), as in Polymixioidei, but in P. 82 (Fig. 1 1 A) the fifth hypural is partially divided distally suggesting that the sixth is fused into it. In all specimens of A. velifer in which the hypurals are visible (P. 4743, P. 4744, 49486, P. 47862) there appear to be only five hypurals. Preceding the principal rays of the caudal fin there are four unsegmented soft rays and three seg- mented rays in the upper lobe, three unsegmented and three segmented rays in the lower. Family PHARMAGICHTHYIDAE Patterson (1964 : 398) This family contains only Pharmacichthys venenifer Smith Woodward (1942) from the Middle Cenomanian of Hakel, Lebanon. None of the five known specimens of this species has the caudal skeleton sufficiently well preserved to be illustrated, but the holotype, AUB 104691/99, and AUB 101872 show most of the caudal anatomy. The suggestion (Patterson 1964 : 401) that the nearest relative of Pharmacichthys is Aipichthys is borne out by the structure of the caudal skeleton and fin. In almost every respect the caudal skeleton of Pharmacichthys is identical with that of Aipi- chthys (Fig. n). The foremost caudal fin-rays articulate with the neural and haemal GEOL. I 7, 2 6 6 4 THE CAUDAL SKELETON IN ,e1-3 d2 npu2 h5 pu1+u1+u2 MESOZOIC ACANTHOPTERYGIAN FISHES e1-3 d2. h6 puUul u2 FIG. ii. Aipichthys minor (Pictet). Caudal skeleton of A, P. 82, standard length 32 mm.; B, RSM 1881 .5.41, standard length c. 50 mm.; C, P. 6183, standard length 36 mm. All from Middle Cenomanian, Hakel, Lebanon. For explanation of lettering see p. 102. In B arrows mark the outermost (unbranched) principal fin-rays. spines of the third pre-ural centrum, the neural spine of the second pre-ural centrum is about half as long as its predecessor, as in Aipichthys, the haemal spine of the second pre-ural centrum has a broad flange on its anterior edge, there are three epurals and the principal rays of the fin are preceded by four unsegmented soft rays and three segmented rays above, three unsegmented and three segmented rays below. It is impossible to see whether there are five or six hypurals. A further resemblance to Aipichthys, not previously noticed, is that the bases of the caudal rays are deeply cleft, overlapping much of the hypurals (" hypurostegy ", Le Danois & Le Danois 1964). The only difference from Aipichthys, suggested by the holotype, AUB 101872 and AUB 102601, is that the first and second hypurals appear to be fused. Family PYGNOSTEROIDIDAE Patterson (1964 : 389) The only member of this family is Pycnosteroides levispinosus (Hay 1903) from the Middle Cenomanian of Hajula, Lebanon. The caudal skeleton and fin are present in two specimens, P. 13900 (Fig. 12) and AMNH 45190 (Hay 1903, pi. 32, fig. 3). Pycnosteroides differs from Aipichthys and Pharmacichthys in having the foremost rays of the caudal fin inserted on the first epural above and the haemal spine of the third pre-ural centrum below, and in having a fully developed neural spine on 66 THE CAUDAL SKELETON IN el npu2, st FIG. 12. Pycnosteroides levispinosus (Hay). Caudal skeleton of P. 13900, standard length c. 50 mm., Middle Cenomanian, Hajula, Lebanon. For explanation of lettering seep. 1 02. the second pre-ural centrum (npu2). This neural spine differs from those of the Polymixioidei in being expanded distally. The haemal arches of the second and third pre-ural centra (pu2, puj) are autogenous. In the fusion of the first ural and first pre-ural centra (pui -f- ui), the free second ural centrum (u2), and the form of the stegural (st} and second uroneural (missing in the figured specimen but present in AMNH 45190;) Pycnosteroides agrees with Aipichthys and the Polymixioidei. There are three epurals (ei 3). In P. 13900, as in Aipichthys velifer, there are only five hypurals (hi 5), but here the first and third are the largest, the third being much larger than the fourth. In AMNH 45190 there are six hypurals. In contrast to Aipichthys and Pharmacichthys there are only three soft rays, all unsegmented, in front of both the upper and lower principal rays, and the fin-rays are not deeply cleft basally. Family DINOPTERYGIDAE Jordan (1923 : 173) This family, the type of the suborder, contains only Dinopteryx spinosus (Davis) from the Upper Santonian of Sahel Alma, Lebanon. The caudal region is very imperfectly preserved in the holotype, MESOZOIC ACANTHOPTERYGIAN FISHES 67 but is more or less completely visible in USNM 22217 and 22219. As in other respects (Patterson 1964 : 392), Dinopteryx resembles Pycnosteroides in the caudal skeleton. The second pre-ural centrum has a fully developed neural spine and both this and the preceding neural spine are expanded distally, as in Pycnosteroides (Fig. 12). As in Pycnosteroides there are three epurals, the first ural and pre-ural centra are fused, there is a free second ural centrum, a stegural and a second uroneural. USNM 22219 shows that there were at least three upper hypurals, shaped as in Pycnosteroides, but it is impossible to be certain whether or not a small sixth hypural was present. The foremost procurrent caudal rays articulate with the neural spine of the second pre-ural centrum above and the haemal spine of the third pre-ural centrum below, further forwards than in Pycnosteroides. There are four spines and two segmented rays in front of the principal rays above, three spines and two segmented rays below, both the holotype and USNM 22217 showing that the foremost procurrent rays were true spines, a difference from other Dinopterygoidei. Suborder BERYCOIDEI Patterson (1964 : 433) This suborder, containing eight living families, is represented in the Cretaceous only by two families, the Trachichthyidae and Holocentridae. A detailed discussion of the relationships between these two families in the Cretaceous will be found in npu3 pu1+u1 npu2 FIG. 13. Hoplostethus mediterraneus Cuvier & Valenciennes. Caudal skeleton of a dried skeleton, 1878.4.5.8, standard length 190 mm., Japan. For explanation of lettering seep. 102. 68 THE CAUDAL SKELETON IN Patterson (1967). Among living Berycoidei, the caudal skeleton of Centroberyx affinis (Berycidae) has been illustrated by Regan (1911, fig. 2), those of Hoplostethus (Trachichthyidae), Holocentrus and Myripristis (both Holocentridae) are briefly discussed by Gosline (1961 : 14) and a caudal skeleton of Holocentrus ascensionis is figured by Rosen (1964, fig. 23D). Fig. 13 shows the caudal skeleton of the living Hoplostethus mediterraneus (Trachichthyidae). In most respects this is typical of primitive Berycoidei, with autogenous haemal spines on the second and third pre-ural centra, the second pre-ural centrum without a neural spine, bearing only a low, broad crest (npu2) which lies below the proximal end of the first epural, three epurals (ei 3), six hypurals (hi 6), a free second ural centrum (u2), a stegural (st) and a second uroneural (d2) . In the specimen illustrated the neural spine of the third pre-ural centrum (npuj) is double distally and the third and fourth hypurals are fused proximally: these features are individual abnormalities. The neural crest on the second pre-ural centrum is autogenous this is a primitive feature which also occurs in primitive myctophoids (N ematonotus and Aulopus, p. 81). In Hoplo- stethus the foremost procurrent fin-rays articulate with the neural and haemal spines of the third pre-ural centrum. In living trachichthyids (Hoplostethus, Trachichthys, Paratrachichthys, Gephyroberyx) the nineteen principal caudal rays are normally preceded by six spines and one segmented ray above and below. The caudal skeleton in living holocentrids is described below (p. 75, Fig. 20). In Berycidae (Regan 1911, fig. 2) the caudal skeleton is advanced over the trachi- chthyid condition in having both the second ural centrum and the stegural fused into the compound first ural and pre-ural centrum. Nothing is known of the caudal skeleton in Korsogasteridae and Anomalopidae. Dissection of single specimens of Diretmus (Diretmidae) and Anoplogaster (Anoplogasteridae) shows that both resemble Berycidae in having the stegural and second ural centrum fused into the preceding compound centrum, while in Anoplogaster the second uroneural is lost and in Diretmus the sixth hypural is lost and there is fusion between the first and second hypurals and between the third and fourth hypurals. In the Stephano- berycoidei, which appear to be only specialized derivatives of the Berycoidei, Gibberichthys (Gibberichthyidae) agrees with trachichthyids such as Hoplostethus in the caudal skeleton but Melamphaes (Melamphaeidae) has both the stegural and second ural centrum fused with the preceding centrum, a much reduced second uroneural, only five hypurals and fusion within the upper and lower hypurals. All living berycoids and stephanoberycoids seem to be characterized by the presence of spinous procurrent caudal rays. In Monocentris (Monocentridae) the caudal skeleton normally shows the same major features as the figured specimen of Hoplostethus (even to the partial doubling of the neural spine of the third pre-ural centrum), but in one of the available skeletons (Fig. 14) there is a fully developed neural spine on the second pre-ural centrum (npu2) and the neural spine of the third pre-ural centrum is single. There are three epurals (ei~3), so that in this individual the neural spine on the second pre-ural centrum has apparently developed instead of the normal doubling of the neural spine of the preceding centrum. The caudal skeleton of this specimen resembles those of polymixioids and the dinopterygoids Pycnosteroides and Dinopteryx, but MESOZOIC ACANTHOPTERYGIAN FISHES 69 FIG. 14. Monocentris japonicus (Houttuyn). Caudal skeleton of a dried skeleton showing a neural spine on the second pre-ural centrum, 1862. u . i .47, standard length 125 mm., Japan. For explanation of lettering see p. 102. it is questionable whether this is significant. However, Monocentris seems to be the only living berycoid in which the procurrent caudal rays are sometimes not spines but unsegmented soft rays : of four specimens examined, one has no spines and one has only one spine in front of the lower caudal lobe and none above. Monocentris is a fish of highly specialized appearance and the skull suggests that the Monocent- ridae are derivatives of the Trachichthyidae, but there is a patch of teeth on the endopterygoid (Starks 1904 : 618), a character otherwise unknown in Berycoidei, and this and the occasional absence of procurrent caudal spines suggest that the Monocentridae may have had a long independent history. Family TRACHICHTHYIDAE Bleeker (1859) The caudal skeleton of the living Hoplostethus mediterraneus is described above (Fig. 13). Genus HOPLOPTERYX Agassiz (1838 : 4) Hoplopteryx, with eight species ranging from the Middle Cenomanian to the Upper Senonian, is the longest-ranging and largest genus of Cretaceous Trachichthyidae. 7 o THE CAUDAL SKELETON IN h6 h5 pu1+u1 FIG. 15. Hoplopteryx lewesiensis (Mantell). Caudal skeleton of P. 19486, standard length c. 175 mm., Chalk, Sussex. For explanation of lettering see p. 102. In the type species, H. antiquus Agassiz from the Campanian of Westphalia, I have seen no specimens in which the caudal skeleton is preserved. The best known species is H. lewesiensis (Mantell) which ranges throughout the English Chalk: several specimens show the caudal skeleton (Figs 15, 16) which is almost identical with that of the living Hoplostethus (Fig. 13), with the neural arch of the second pre-ural centrum (npu2) reduced (though not so much as in Hoplostethus}, the haemal arches of the second and third pre-ural centra autogenous, three epurals (ei~3) of which the first is especially large, a stegural (sf) and a second uroneural (d2), a free second ural centrum (u2) and six hypurals (hi-6), the uppermost very small. The neural arch of the second pre-ural centrum is not autogenous as it is in Hoplostethus. The foremost caudal fin-rays are inserted on the first epural above and on the haemal spine of the third pre-ural centrum below. There are only three spines and one segmented ray in front of the upper principal rays and two spines and one segmented ray in front of the lower. The other species of Hoplopteryx in the English Chalk are H. simus Smith Wood- ward, H. macr acanthus Patterson and H. gephyrognathus Patterson. The caudal MESOZOIC ACANTHOPTERYGIAN FISHES U2 2mm FIG. 16. Hoplopteryx lewesiensis (Mantell). Caudal skeleton of P. 5421, standard length c. 135 mm., Lower Chalk, Lewes, Sussex. For explanation of lettering see p. 102. The second uroneural is displaced ventrally and the fifth and sixth hypurals are missing. skeleton of H. simus is exposed in P. 11202, that of H. macmcanthus in P. 30186: neither appears to differ from H. lewesiensis in any way. The caudal region is not preserved in the two known specimens of H. gephyrognathus. The earliest species of Hoplopteryx is H. lewisi (Davis) from the Middle Ceno- manian of Hakel, Lebanon. In this species the caudal skeleton and fin are exposed in P. 10709 and partially shown in the holotype, P. 4758. H. lewisi seems to agree with H. lewesiensis in every detail, even to the insertion of the foremost upper fin rays on the first epural, except that there are four spines and one segmented ray in front of the upper principal rays, three spines and one segmented ray in front of the lower. The remaining two species of Hoplopteryx, H. syriacus (Pictet & Humbert) and H. spinulosus Smith Woodward, are from the Upper Santonian of Sahel Alma, Lebanon. In H. spinulosus I have seen no specimen in which the caudal skeleton 7 2 THE CAUDAL SKELETON IN is preserved. In H. syriacus parts of the caudal skeleton are preserved in 49553 and they show nothing to distinguish the species from H. lewesiensis. Genus LISSOBERYX Patterson (1967 : 73) The type species, L. dayi (Smith Woodward 1942), is from the M. Cenomanian of Hakel and Hajula, Lebanon. I have briefly described the caudal skeleton (Pat- terson 1967 : 78) which is preserved in AUB 108926 (Fig. 17) and AUB 101997. The caudal skeleton agrees well with those of Hoplostethus (Fig. 13) and Hoplopteryx (Figs 15, 16). The neural spine of the second pre-ural centrum (npu2) is reduced to about the same extent as in Hoplopteryx, there are three epurals (21-3), the d2 st npu2 pu3 FIG. 17. Lissoberyx dayi (Smith Woodward). Caudal skeleton of AUB 108926, standard length 34 mm., Middle Cenomanian, Hajula, Lebanon. For explanation of lettering see p. 102. Arrows mark the outermost (unbranched) principal fin-rays. first long and closely applied distally to the neural spine of the third pre-ural centrum, a stegural (st) and a second uroneural (d,2), and six hypurals (hi-6). The only significant differences from Hoplopteryx and Hoplostethus are that the fused first ural and pre-ural centrum is longer, with clear signs in the surface sculpture of its origin from two centra, and that the second ural centrum is also longer, so that the caudal skeleton appears more upturned. As in Hoplopteryx the foremost caudal rays articulate with the first epural above and the haemal spine of the third pre-ural centrum below. The principal rays are preceded by four spines and one unseg- mented ray above, three spines and one segmented ray below. MESOZOIC ACANTHOPTERYGIAN FISHES 73 Genus ACROGASTER Agassiz (1838 : 5) I have seen no specimens of A. parvus Agassiz, the type species, or of A. brevi- costatus von der Marck, both these species from the Campanian of Westphalia being poorly known. The remaining species, A. heckeli (Pictet) and A. daviesi (Davis), from the Upper Santonian of Sahel Alma, Lebanon, are common and several specimens in the British Museum (Natural History) show the caudal skeleton clearly: I can find no differences between these two species in caudal anatomy. A specimen of u2 FIG. 18. Acrogaster heckeli (Pictet). Caudal skeleton of P. 4155, standard length c. 55 mm. Upper Santonian, Sahel Alma, Lebanon. For explanation of lettering see p. 102. A. heckeli is illustrated in Fig. 18. As in the other trachichthyids described above, the neural spine of the second pre-ural centrum is reduced (npu2), there are three epurals (ei-j), a stegural (st), a second uroneural (d2) and six hypurals (hi-6). The fused first ural and pre-ural centra (pui + ui) and the second ural centrum (u2) are elongate, as in Lissoberyx, and the caudal skeleton appears strongly upturned. As in Hoplopteryx and Lissoberyx, the foremost caudal rays articulate with the first epural above and the haemal spine of the third pre-ural centrum below. The principal caudal rays are preceded by four or five spines and one segmented ray above, three or four spines and one segmented ray below. 74 THE CAUDAL SKELETON IN Genus TUBANTIA Patterson (1964 : 413) The only species is T. cataphractus (von der Marck), from the Campanian of Westphalia, in which the caudal skeleton is well exposed in P. 21984 (Fig. 19). Tubantia agrees with other trachichthyids in the reduction of the neural spine of the second pre-ural centrum (npu2), the three epurals (ei-3), stegural (st), second uro- d2 h6 npu2 pul+ul FIG. 19. Tubantia cataphractus (von der Marck). Caudal skeleton of P. 21984, standard length 130 mm., Campanian, Baumberg, Westphalia. For explanation of lettering seep. 102. neural (d,2], and six hypurals (hi-6). As in Hoplostethus, the fused first ural and pre- ural centrum (pui + ui] is only as long as the preceding centrum. Tubantia differs from other Cretaceous trachichthyids and resembles living forms in having the number of procurrent rays increased to nine spines and two segmented rays above, six spines and two segmented rays below, these small rays extending forwards in front of the tips of the neural and haemal spines of the third pre-ural centrum. Genus GNATHOBERYX Patterson (1967 : 81) The type and only species, G. stigmosus Patterson (1967 : 82), from the Upper Santonian of Sahel Alma, Lebanon, is known by two specimens and the caudal skeleton is preserved only in the holotype, AUB 100402, where it is compressed and MESOZOIC ACANTHOPTERYGIAN FISHES 75 distorted. So far as can be seen, the caudal skeleton and fin agree with other Creta- ceous trachichthyids such as Lissoberyx, Hoplopteryx and Acrogaster, with the neural spine of the second pre-ural centrum reduced, the stegural free, a free second ural centrum, and the upper principal rays preceded by four spines and one segmented ray, the foremost articulating with the first epural. Family HOLOCENTRIDAE Richardson (1846) The caudal skeleton of the living Myripristis adustus is shown in Fig. 20 : it shows no significant differences from those of several species of Holocentrus. The neural and haemal spines of the fourth (puj) and fifth pre-ural centra are expanded but short. The neural spine of the second pre-ural centrum (npu2) is greatly reduced 2mm FIG. 20. Myripristis adustus Bleeker. Caudal skeleton of a dried skeleton, 1858 .4.21. 239, standard length 155 mm., Amboina. hap, hypurapophysis ; for explanation of other lettering see p. 102. (and not autogenous as it is in Hoplostethus), with the tip of the first epural lying above it. The haemal arches of the second and third pre-ural centra (pu2, pu3) are autogenous and there are three epurals (ei~3) . In contrast to the trachichthyids the two ural centra, the first pre-ural centrum and the stegural are all fused into a single structure (pui + UI + U2-}- st) and there are only five hypurals (hi-5) the uppermost hypural present in trachichthyids having been lost. The second uroneural (d2) is free and fits proximally into a notch in the stegural rather than lying below and behind it. In Myripristis the foremost caudal rays articulate with the first epural above (as in most Cretaceous trachichthyids) and with the haemal 7 6 THE CAUDAL SKELETON IN spine of the third pre-ural centrum below. In Holocentrus the foremost rays arti- culate with the neural and haemal spine of the third pre-ural centrum. In Myrip- ristis the principal caudal rays are preceded by four spines and one segmented ray above and below. Rosen (1964, fig. 230) has figured a caudal skeleton of Holo- centrus ascensionis which differs from all the Recent holocentrid skeletons that I have seen in having a free second ural centrum. Rosen does not say how large his specimen was, but if it was an alizarin-stained juvenile this difference can be explained. Genus CAPROBERYX Regan (1911 : 8) In the type species, C. superbus (Dixon) from the Turonian of the English Chalk, the caudal skeleton is exposed in P. 3979 (Fig. 21). The neural and haemal spines of the fourth (Puj) and fifth pre-ural centra are normal, not expanded as in living holocentrids. The haemal arches of the second and third pre-ural centra (pu2, PUJ) are autogenous, the neural spine of the second pre-ural centrum (npu2) is reduced and there are three epurals (ei-3), all as in living holocentrids. In contrast to living holocentrids, the stegural (st) is not fused to the underlying centrum (pui + ui) and the second ural centrum (u2) is free. The second uroneural (d2) is fused to the stegural (st) distally, but this is perhaps only a consequence of the very large size (standard length c. 40 cm) of this specimen. As in living holocentrids there are only five hypurals, but the distribution of the branched principal fin-rays (one on e1-3 d2 npu2 10mm ph FIG. 21. Caproberyx superbus (Dixon). Caudal skeleton of P. 3979, standard length c. 400 mm., Chalk, Sussex. For explanation of lettering see p. 102. MESOZOIC ACANTHOPTERYGIAN FISHES 77 the uppermost hypural, two on the one below and six on the next) shows that the three upper hypurals are the third and fourth fused (hj + 4), the fifth (/5) and the sixth (h6), for in living holocentrids the three upper hypurals, the third, fourth and fifth, bear two, six and one branched principal rays respectively. Caproberyx therefore agrees with the trachichthyids in retaining the small sixth hypural, and the fusion of the third and fourth hypurals shown by P. 3979 is again probably merely a consequence of the large size of the fish. As in Myripristis and most Cretaceous trachichthyids, the foremost caudal fin-rays articulate with the first epural above and the haemal spine of the third pre-ural centrum below, and there are probably four spines and one segmented ray in front of the principal caudal rays above and below. Characters of the skull and fins suggest that Caproberyx is the most primitive holocentrid known, lying near to the common stock of the Holo- centridae and Trachichthyidae (Patterson 1964 : 359; 1967 : 103). This is con- firmed by the structure of the caudal skeleton in C. superbus, for in the six hypurals, the free stegural and second ural centrum, and the unexpanded neural and haemal spines of the posterior caudal vertebrae, this species resembles the trachichthyids rather than other holocentrids. The other species of Caproberyx are C. polydesmus (Arambourg 1954) from the Lower Cenomanian of Jebel Tselfat, Morocco, and C. pharsus Patterson (1967 : 97) from the Middle Cenomanian of Hakel, Lebanon. C. pharsus is known only by a specimen lacking the caudal region. C. polydesmus is known only by the holotype: Arambourg's figure (1954, pi. 19, fig. i) shows that the neural and haemal spines of the fourth and fifth pre-ural centra are not expanded, as in C. superbus, and he described the principal rays as having five or six small rays in front of them in each lobe. Genus STICHOCENTRUS Patterson (1967 : 88) The type and only species is S. liratus Patterson from the Middle Cenomanian of Hajula, Lebanon. The caudal skeleton is well preserved in AUB 108923 (Fig. 22) and is partially shown in AUB 108927 and 108929. The neural and haemal spines of the fifth pre-ural centrum are unmodified but those of the fourth (puj) are expanded, though not so strongly as in living holocentrids. The autogenous haemal arches on the second and third pre-ural centra, the reduced neural spine on the second pre-ural centrum (npu2) and the three epurals (ei~3) are as in Caproberyx and living holocentrids. The stegural is fused with the underlying centrum (pui + ui + U2 -\- st) anteriorly. The second uroneural (d2) is free and lies below and behind the first, not notched into the first as it is in living holocentrids. The second ural centrum is fused to the compound first pre-ural and ural centrum in AUB 108923 (standard length c. 75 mm.) although the line of fusion is clearly seen, but in AUB 108929, a much smaller specimen (standard length c. 35 mm.) the centrum appears free. As in living holocentrids, there are only five hypurals (hi-5), the small sixth hypural present in Caproberyx having been lost. As in Myripristis, the foremost caudal rays are inserted on the first epural above and the haemal spine of the third pre-ural centrum below, and there are four spines and one segmented ray in front of both the upper and lower principal rays. THE CAUDAL SKELETON IN e3 2mm FIG. 22. Stichocentrus liratus Patterson. Caudal skeleton of AUB 108923, standard length c. 75 mm., Middle Cenomanian, Hajula, Lebanon. For explanation of lettering see p. 102. Arrows mark the outermost (unbranched) principal fin-rays. As in the skull and fins, Stichocentrus is more advanced towards the living holocentrids than Caproberyx in the expanded neural and haemal spines of the fourth pre-ural centrum, the five hypurals and the partial fusion of the stegural and second ural centrum with the preceding centrum. The remaining Cretaceous holocentrids are Trachichthyoides ornatus Smith Wood- ward (1902), known only by an isolated head from the English Chalk, and Kansius sternbergi Hussakof (1929) known by the two syntypes from the Niobrara Formation, Gove Co., Kansas. Nothing is known of the caudal anatomy of these forms. Order LAMPRIDIFORMES Suborder LAMPRIDOIDEI Berg (1940 : 463) This suborder is used to contain both the Lampridoidei (Lampris only) and the Veliferoidei (Velifer, etc.] of Berg (Bonde 1966). ? Family VELIFERIDAE Bleeker (1860) Genus BATHYSOMA Davis (1890 : 424) The type and only species is B. lutkeni Davis from the Danian stage of southern Sweden. The caudal skeleton is partially preserved in two specimens in the British Museum (Natural History). P. 9947 (Fig. 23 A) shows that the upper hypurals are MESOZOIC ACANTHOPTERYGIAN FISHES 79 fused into a triangular plate and that this plate is fused basally with the second ural centrum (u2 -\- uh). Above the second pre-ural centrum (pu2) and the fused first pre-ural and ural centra (pui + ui) there are two elongate bones: the second of these (e) is certainly an epural but it is impossible to be certain whether the first is an epural or the neural arch and spine of the second pre-ural centrum. Above the upper hypural plate there is a third slender bone of uncertain nature. P. 9948 B pu1+u1 u2+uh 2mm FIG. 23. Bathysoma lutkeni Davis. Caudal skeleton of A, P. 9947, standard length c. 95 mm.; B, G, P. 9948 (part and counterpart), standard length 78 mm. Both from Danian, Limhamn, southern Sweden. U2 + uh, second ural centrum fused with one or more upper hypurals; for explanation of other lettering see p. 102. (in counterpart, part and counterpart shown in Fig. 236, C) shows the second pre- ural centrum (pu2) with an autogenous haemal arch and no sign of a neural spine, the fused first pre-ural and ural centra (pui + ui) bearing the parhypural (ph) and the first hypural (hi) ,the distal part of the second hypural (h2), and the upper hypural plate with a fragment of the second ural centrum fused to it (u2 + uh). As in P. 9947, there is a slender bone lying above the upper hypural plate. The bones above the first and second pre-ural centra are shattered and displaced. Although GEOL. 17, 2 7 8o THE CAUDAL SKELETON IN these two specimens are far from complete, they show that in Bathysoma the haemal arch of the second pre-ural centrum was autogenous, the first pre-ural and ural centra were fused, the parhypural and the first and second hypurals were separate and autogenous, there was at least one epural, and the upper hypurals were fused with each other and with the second ural centrum. On the available material it is difficult to interpret the upper hypural plate and the slender bone above it, which may be an epural, a stegural or a free hypural. In the upper hypural plate at least three hypurals can be recognized in transparency under xylene, presumably hypurals 3-5, but the uppermost part of the plate is of a different texture, suggesting that the stegural may also be fused into the structure, a most unusual condition which can only be confirmed on more complete material. e1-3. st>d2 pti W hp FIG. 24. Mene maculata (Bloch & Schneider). Caudal skeleton of a dried skeleton, 1866.6.8.59, standard length 95 mm., Taiwan, hp, hypurapophysis ; for explanation of other lettering see p. 102. The caudal skeleton of Bathysoma is very different from that of Mene, the only genus of the Menidae, in which Bathysoma was previously placed (Patterson 1964 : 423). In the living Mene maculata (Fig. 24) the first pre-ural centrum, both the ural centra, and all but the uppermost hypural are fused into a symmetrical, fan-shaped plate (pui + ui + U2 + hi-4). The parhypural (ph) has a very large hypura- pophysis (hp) and lies free below this plate. The uppermost hypural (h^) is also free, articulating with a hook on the upper edge of the hypural plate. There are three normal epurals (ei-3) and the neural arch of the second pre-ural centrum (pu2) is reduced to a very low crest. The stegural (st + d2) is autogenous, articu- lating with the underlying compound centrum by a large and clearly mobile joint. The distal part of the shaft of the stegural is grooved longitudinally in M. maculata ; in the Middle Eocene species M. rhombeus (Volta) and M. oblongus (Agassiz), in which the caudal skeleton is otherwise identical with that of the living species, there MESOZOIC ACANTHOPTERYGIAN FISHES 81 is a free second uroneural, and the groove on the shaft of the stegural in M. maculata clearly marks the line of fusion between the first and second uroneurals. Mene has seventeen principal caudal rays with fifteen branched, the outermost two or three rays unsegmented and the inner ones only sparsely segmented, preceded by five unsegmented but divided (in the median plane) rays above and four below. The bases of the caudal rays are deeply cleft, covering much of the hypural plate. The caudal skeleton of Mene can be derived from the basal perciform type (p. 87) by fusion of the first four hypurals with each other and with the supporting centra, and the caudal fin has the perciform number of rays. The caudal skeleton of Bathysoma is of a much more primitive type, differing from that of Beryciformes mainly in the fusion of the upper hypurals with the second ural centrum. A caudal skeleton of this type occurs in the Lampridiformes Velifer (Gosline 1961; fig. 3D), Palaeocentrotus (Kiihne 1941, fig. 2 ; Bonde 1966) and Lampris, and the known skele- tal features of Bathysoma (Patterson 1964, fig. 90) agree as well with Palaeocentrotus and Velifer (Regan 1907; Smith 1951) as they do with Mene. The holotype of Bathysoma lutkeni, in Copenhagen, shows that the supraoccipital crest is attached to the skull roof only at the posterior end, with a gap between it and the frontal crest (Bonde, personal commn) : an exactly similar supraoccipital crest occurs in Bonde's (1966) ? veliferid from the Lower Eocene [Mo-clay], while in Palaeocentrotus there is a large foramen between the supraoccipital and frontal crests (Kiihne 1941, fig. 3). Provisionally Bathysoma may be placed in the Veliferidae but, like Bonde's Eocene form, it may well prove to be closer to Palaeocentrotus. Order SALMONIFORMES (Greenwood et al. 1966) Suborder MYCTOPHOIDEI The most generalized of living myctophoids is Aulopus (Aulopodidae, Regan 191 la : 121 ). The caudal skeleton of Aulopus is briefly discussed by Gosline (1961 : 10), who notes that this genus is one of the few living teleosts retaining large caudal scutes in front of the caudal lobes. The caudal skeletons of more advanced myctophoids have been figured by HoUister (Synodus, Trachinocephalus , 1937^, figs 1-14), Gosline (C hlorophthalmus , 1961, fig. 26), Rosen (Myctophum, 1964, fig. 236) Greenwood et al. (Neoscopelus, 1966, fig. 36) and Weitzman (Parasudis and Saurida 1967, figs 17, 18). In Aulopus the caudal skeleton is almost identical with that of the Cretaceous Ctenothrissa (Figs 4, 5), with a free second ural centrum, a stegural and a second uroneural, three epurals and six hypurals, of which the first and third are the largest. The only differences from Ctenothrissa are that the haemal arch of the third pre-ural centrum and the neural arch of the second pre-ural centrum are auto- genous, the neural spine of the second pre-ural centrum is less expanded, more spine-like, and just over half as long as its predecessor, and the second ural centrum has a long posterior process. In all these characters, Aulopus appears to be more primitive than Ctenothrissa. In other living myctophoid families conditions are much as in Aulopus, but the caudal scutes are lost, the neural spine of the second pre-ural centrum ceases to be autogenous and becomes shorter and expanded, the first epural tending to move forwards above it, there is often fusion within the upper 82 THE CAUDAL SKELETON IN and lower hypurals, the sixth hypural and one epural may be lost (Synodontidae), and the second ural centrum and stegural may fuse with the compound first ural and pre-ural centrum. These trends are very like those seen within the Berycoidei (p. 68; cf. figs. 236, D in Rosen 1964). In the Upper Cretaceous myctophoids were abundant, the best known genera being Sardinioides, Acrognathus, Cassandra ( = Leptosomus) and Nematonotus. Nematonotus appears to be the most primitive of these and will serve as an example. Figure 25 shows specimens of Nematonotus bottae (Pictet & Humbert), from the Cenomanian of Hakel, Lebanon, and N. longispinus (Davis), from the Cenomanian e1-3 d2- e1-3. h5 pu1*u1 pu1*u1 ph ud 2mm FIG. 25. The caudal skeleton in A, Nematonotus bottae Pictet & Humbert, 49563, stand- ard length c. 90 mm., Middle Cenomanian, Hakel, Lebanon; B, Nematonotus longispinus (Davis), P. 13882, standard length 114 mm., Middle Cenomanian, Hajula, Lebanon. In G are the bases of the upper caudal rays in P. 48825, N. longispinus, standard length 97 mm., to show the urodermal, ud. For explanation of other lettering see p. 102. In B the second uroneural is missing, in A and C arrows mark the outermost (unbranched) principal fin-rays. MESOZOIC ACANTHOPTERYGIAN FISHES 83 of Hajula, Lebanon. There are no significant differences between these two species in the caudal region. In the specimen of N. bottae illustrated, the haemal arches of the third and fourth pre-ural centra are fused and the haemal spine of the third pre-ural centrum is partially doubled. This is an individual abnormality. As in Aulopus there is a large caudal scute (/. s) above and below the caudal skeleton and the neural spine of the second pre-ural centrum (npu2) is autogenous, spine-like and just over half as long as its predecessor. The autogenous haemal spines on the second and third pre-ural centra, three epurals (ei-3), six hypurals (hi-6), long second ural centrum (u2, Fig. 256) and second uroneural (d2) are as in Aulopus. The stegural (st) is forked proximally, with a process extending forwards on to the second pre-ural centrum. This process, absent in Aulopus and Ctenothrissiformes, is a primitive feature present in Flops and some clupavids (Patterson 19670, fig. n) which indicates the double origin of the first uroneural (Regan 1910 : 355 ; Patterson 1968 : 226). In Nematonotus there is a single urodermal (Fig. 256), a structure absent from living myctophoids and Ctenothrissiformes but also present in the Cenomanian Sardinioides attenuatus. In Nematonotus the foremost procurrent caudal rays articulate with the neural and haemal spine of the third pre-ural centrum and the nineteen principal rays are preceded by four unsegmented and four segmented rays above, four unsegmented and two segmented rays below. In myctophoids the procurrent caudal rays are normally longitudinally divided, not spinous, but pro- current caudal spines appear in advanced members of the family Myctophidae (Fraser-Brunner 1949 : 1033). III. DISCUSSION (a) The relationships of Ctenothrissiformes, Myctophoidei and Beryciformes. Monod (1967 : 118) has remarked that the structure of the caudal skeleton is rarely of value in discriminating between taxa at the generic level and below, but that it becomes increasingly valuable at the familial, subordinal and ordinal level. This observation is fully borne out by the forms described here, the only exception being the occurrence of two types of caudal skeleton in the ctenothrissiform family Aulolepididae, Pateroperca having the neural spine of the second pre-ural centrum fully developed, Aulolepis having it short and expanded, as it is in Ctenothrissidae. This suggests that Aulolepis and Pateroperca belong to different families, but Pateroperca is as yet so poorly known that no conclusion can be reached on this point until more specimens are discovered. Among the Ctenothrissiformes, Myctophoidei and Beryciformes caudal anatomy is very similar. The basal type of caudal skeleton in these groups has the following features: the first pre-ural and ural centra fused, a free second ural centrum, three epurals, a stegural, a second uroneural and six hypurals (two lower and four upper). Caudal scutes above and below the caudal skeleton are present in Aulopus, most Cretaceous myctophoids (Nematonotus, Acrognathus, Sardinioides) and Ctenothriss- iformes (Aulolepis, Ctenothrissa) : they are absent in Beryciformes and higher groups. A single urodermal is present in Nematonotus and Sardinioides attenuatus among Cretaceous myctophoids. There is no urodermal in Ctenothrissiformes or Bery- ciformes. The main variations encountered in the caudal skeletons of myctophoids, GEOL. 17, 2 7 84 THE CAUDAL SKELETON IN ctenothrissiforms and beryciforms involve reductions in the number of caudal elements by fusion or suppression and the condition of the neural spine of the second pre-ural centrum, which may be fully developed (Pateroperca, Polymixioidei, Pycno- steroides, Dinopteryx), spine-like and about half as long as its predecessor (Nematono- tus, Aulopus, Aipichthys, Pharmacichthys) , short and expanded (Ctenothrissa, Aulolepis, many myctophoids) or reduced to a low crest (Berycoidei) . The basal teleostean caudal skeleton, seen in such genera as Leptolepis, Allothris- sops, Thrissops, Ichthyodectes, Hiodon, Flops, Salmo, etc., contains two free ural centra, three epurals, two lower hypurals and five upper hypurals (there may be six or seven upper hypurals in Leptolepis; Patterson 1968 : 220; there are only four in salmonids; Norden 1961 : 738), up to seven uroneurals (Patterson 1968) which extend forwards to the fourth pre-ural centrum in Thrissops and Ichthyodectes, to the third in Leptolepis and Allothrissops and to the second pre-ural centrum in Hiodon, Elops, Salmo, etc., and one or two urodermals (Patterson 1968 : 230). The condition of the neural arches and spines in the caudal region is often complicated by doubling of the segmental structures, usually as individual variations, but in Leptolepis, Allothrissops, Flops and some salmonids there is normally a neural arch and spine on the first pre-ural centrum, and in Leptolepis, Hiodon, osteoglossoids (Greenwood 1967) and Alepo- cephalus (Patterson 1968, fig. 12) there may be a more or less well developed arch and spine on the first ural centrum. It is usually assumed (e.g. Gosline 1961 : 14; Patterson 19670 : 104) that the second pre-ural centrum primitively bears a complete neural spine, supporting epaxial fin-rays. But among the forms described here the most primitive (Aulopus, Nematonotus) have the neural spine of the second pre-ural centrum about half as long as that of the third, and this is also true of Elops (Nybelin 1963, figs, i, 4), which in other respects seems to be the most primitive living teleost. This suggests that the short second pre-ural neural spine may be primitive for some teleost groups. There are three possible conditions of the second pre-ural neural spine. It may be fully developed, as in Allothrissops (Patterson 19670, fig. 6), Ichthyodectes (Cavender 1966, fig. i), Tarpon (Nybelin 1963, fig. 7), and many other primitive telosts, normally supporting epaxial fin-rays but in Tarpon ending just in front of the foremost epaxial fin-ray; it may be about half as long as its prede- cessor, as in Elops, " Clupavus " (Patterson 19670, fig. n), Nematonotus and Aulopus; or it may be represented only by a low crest, as in Berycoidei and generalized percoids (p. 87). Intermediates between the second and third of these conditions occur in Aulolepis, Ctenothrissa and many myctophoids. Intermediates between the first two conditions seem to occur only in primitive protacanthopterygian groups (salmonids, as in the specimen of Cristivomer illustrated by Vladykov 1954, fig. 2; characinids, as in the specimen of Brycon illustrated by Weitzman 1962, fig. 15) in which the pre-ural neural spines are very variable and both conditions may occur in a single species. The third condition, the spine reduced to a low crest, is undoubtedly advanced and may be left out of consideration here. In the ancestors of the teleosts, the pholidophorids, the neural spines of the last three pre-ural verte- brae decrease in size progressively so that all three end on approximately the same oblique plane : the first pre-ural neural spine is very short, the second is both shorter and more slender than the third (Patterson 1968, figs. 1-4). In pholidophorids MESOZOIC ACANTHOPTERYGIAN FISHES 85 these neural spines do not reach the dorsal edge of the trunk and do not support fulcra or fin-rays. This condition of the pre-ural neural spines seems to be primitive for the teleosts as a whole and it persists in the Lower Jurassic Leptolepis cory- phaenoides and L. normandica (Nybelin 1963, figs. 9, 10) and the Upper Jurassic L. dubia (Nybelin 1963, fig. 8; Patterson 1968, fig. 10). In the Upper Jurassic two of the three modern types of second pre-ural neural spine were already in exis- tence. In Allothrissops and Thrissops (Nybelin 1963, figs, u, 12; Patterson 19670, fig. 6) both the second and third preural neural spines have elongated so that they reach the dorsal edge of the trunk, ending just in front of the foremost procurrent fin-rays: this is essentially the condition in living Tarpon. In the Upper Jurassic E lops-like fish illustrated by Nybelin (1963, fig. 6) the second pre-ural neural spine remains short but the third and fourth pre-ural neural spines are elongated, reaching the dorsal edge of the trunk and supporting the foremost procurrent fin-rays. This is essentially the condition in living Flops. There is no a priori reason to regard either of these two conditions as more primitive, both are a response to a new need, the necessity to support the epaxial procurrent rays as they extend forwards to increase the dorso-ventral symmetry of the tail. However, we know that the Elops type, with a short second pre-ural neural spine, has persisted unchanged in elopids since the Upper Jurassic and that this type of second pre-ural neural spine is primi- tive for the teleosts as a whole, and there is no reason for regarding the short neural spine of Elops and its Jurassic relative as a secondary regression from a long neural spine of Tarpon type. I conclude, therefore, that when one finds a second pre-ural neural spine resembling that of Elops in a generalized teleost one should regard it as a primitive feature unless there is good evidence to the contrary. In support of this interpretation is the occurrence of a second pre-ural neural spine of this type only in teleosts in which the caudal skeleton retains such primitive features as a free second ural centrum and nineteen principal caudal rays. A generalized teleost having an elongate second pre-ural neural spine is to be regarded as having developed this from a short spine of leptolepid or elopid type : this development can apparently take place spontaneously (see Aipichthys, Fig. n). From the basal type of teleostean caudal skeleton, the most primitive members of the myctophoid-ctenothrissiform-beryciform assemblage (such as Nematonotus} differ in the loss of one hypural, the seventh, whether by suppression or by fusion with the sixth is as yet unknown, have reduced the number of uroneurals to two, principally by loss of the small posterior uroneurals (ural neural arches 6-8), since the forked first uroneural of Nematonotus (Fig. 25) is clearly homologous with that of Elops, representing the second and third ural neural arches, while the second uroneural is probably homologous with the second uroneural of Elops, representing the fourth and fifth ural neural arches (Patterson 1968 : 226), the first pre-ural and ural centra have fused, and the first uroneural has fused with the first ural and pre- ural neural arches to produce a stegural, a development which took place very early in the protacanthopterygian lineage (salmonids, " Clupavus ", etc.). These changes raise the question of the origin of the myctophoids and ctenothrissiforms. Green- wood et al. (1966 : 371) wrote of the ctenothrissiforms " we link them with some early group of myctophoid-like salmoniform fishes in which the supramaxillae were 86 THE CAUDAL SKELETON IN not reduced, the premaxilla had not excluded the maxilla from the gape, and in which the adipose fin had disappeared. " By extension, the ancestor of the mycto- phoids would be such a fish with an adipose fin. On the other hand, Gosline (1961 : 35 ; also Gosline, Marshall & Mead 1966 : 5) points out that the large caudal scutes of Aulopus make it impossible to derive the myctophoids from any living teleost except the elopoids. Weitzman (1967 : 532) discusses this point and notes that caudal scutes are present in Argentina, but the structures he describes in the salmonoid Plecoglossus and the galaxioid Retropinna do not seem to resemble caudal scutes. The absence in all known salmonoids of large caudal scutes and of a forked first uroneural of the type found in Elops and Nematonotus, together with the absence of recognizable salmonoids from pre-Tertiary rocks, make it difficult to envisage any direct relationship between salmonoids and myctophoids (Greenwood et al. 1966, fig. i) . A short second pre-ural neural spine resembling those of Aulopus and Nematonotus occurs in some salmonoids (Coregonus, which also has a urodermal), but the last few neural arches and spines and the epurals are apparently very variable, and no clear pattern emerges from Norden's (1961 : 738) analysis. In Argentina caudal scutes are present (though they are reduced) and the first pre-ural and ural centra are fused (Gosline 1960, fig. 10), as they must have been in the common ancestor of Myctophoidei and Ctenothrissiformes, but the neural spine of the second pre-ural centrum is elon- gate and supports procurrent fin-rays : in this character Argentina is advanced over the basal myctophoids and ctenothrissiforms. In the Clupavidae, a family with a fossil record extending back to the Upper Jurassic, large caudal scutes are usually present in the tail, the skull seems primitive enough to have given rise to both myctophoids and Ctenothrissiformes, and the caudal skeleton may be strikingly like that of Nematonotus (Patterson 19670, fig. n). It is unlikely that the known clupavids were ancestral to the myctophoids and ctenothrissiforms because of their reduced dentition and clupeid-like jaws, with a high coronoid process on the dentary, but they suggest a possible source for this type of caudal skeleton. Taking the caudal skeleton of Nematonotus as the primitive condition for the mycto- phoids and ctenothrissiforms, the myctophoids are characterized by shortening and broadening the neural spine of the second pre-ural centrum, and in more advanced forms by fusion of the hypurals with each other and with the supporting centra. Within the ctenothrissiforms two distinct types of caudal skeleton occur: in Ctenothrissa and Aulolepis the neural spine of the second pre-ural centrum is expanded, as in myctophoids, tending towards the condition in Berycoidei and Percoidei; in Pateroperca the second pre-ural neural spine is fully developed, as it is in Polymixioidei. In Aulolepis and some species of Ctenothrissa there are pro- current spines in front of the caudal fin : this is specific evidence of evolution towards the Berycoidei and like the reduction of the second pre-ural neural spine in these fishes it distinguishes them from the Polymixioidei and Dinopterygoidei, in which procurrent caudal spines occur only in advanced forms. Within the Beryciformes, the structure of the caudal skeleton gives some support to the division of the order into three suborders. In Polymixioidei there are always six hypurals and the neural spine of the second pre-ural centrum is fully developed. In Berycoidei the second pre-ural spine is reduced to a low crest (except in one MESOZOIC ACANTHOPTERYGIAN FISHES 87 individual of Monocentris, Fig. 14), the procurrent rays are spinous, and there are trends towards loss of the sixth hypural and fusion of the stegural and second ural centrum with the preceding centrum. In the Dinopterygoidei, already known to be a heterogeneous group, the neural spine of the second pre-ural centrum retains the primitive short condition (Aipichthys, Pharmacichthys) or is fully developed (Dinop- teryx, Pycnosteroides) and there are trends towards loss of the sixth hypural. The polymixioid caudal skeleton (also found in Dinopteryx and Pycnosteroides) resembles that of the ctenothrissiform Pateroperca, differing only in having no caudal scutes and in having one less principal ray. The caudal skeleton of Berycoidei resembles those of Aulolepis and Ctenothrissa, differing only in having lost the caudal scutes and further reduced the second pre-ural neural spine. Aipichthys and Pharma- cichthys seem to have the most primitive caudal skeletons known in Beryciformes, with nineteen principal rays, the primitive short second pre-ural spine, and, at least in some specimens of Aipichthys, six autogenous and separate hypurals and a free second ural centrum. Although both Aipichthys and Pharmacichthys are too specialized in other characters to have given rise to Beryciformes, their caudal skele- ton could give rise to both the polymixioid condition (by elongation of the second pre-ural neural spine, which occurs spontaneously in some individuals of Aipichthys, Fig. nC) and the berycoid condition (by shortening of the second pre-ural neural spine). (b) The origin ofPerciformes. Gosline (i96ia) discussed the caudal skeleton of Perciformes and found that the most generalized type contains fifteen branched principal rays, no neural spine on the second pre-ural centrum, three epurals, two free uroneurals, no free ural centra, five autogenous hypurals, and the haemal arches of the second and third pre-ural centra autogenous (see also Monod 1967, fig. 3). Gosline mentioned that this type of caudal skeleton occurs in Kuhlia (Kuhliidae), Chaetodon (Chaetodontidae), Polydactylus (Polynemoidei) and juvenile Sphyraena (Sphyraenoidei) . Monod (1967) refers to this type of caudal skeleton as " sciaeno-sparidien banal " and states that it occurs in many Perciformes, mentioning Sciaena (Sciaenidae), Pagrus, Sparus (Sparidae) and Gaterin (Pomadasyidae) . I find that this generalized type of caudal skeleton also occurs in Centropomus (Centropomidae) , Lateolabrax, Polyprion, Dicentrarchus, M or one, Acanthistius (Percichthyidae, sensu Gosline 1966), Branchio- stegus (Branchiostegidae), Pomatomus (Pomatomidae), Brama (Bramidae), Arripis (Arripidae), Lutjanus (Lutjanidae), Nemipterus, Scolopsis (Nemipteridae), Lobotes (Lobotidae), Xenocys, Xenistius, Xenichthys (Pomadasyidae), Lethrinus, Sphaerodon (Lethrinidae), Monodactylus (Monodactylidae) , Kyphosus, Medialuna (Kyphosidae) , Ephippus, Drepane, Platax (Ephippidae), Chelmo, Heniochus, Pomacanthus (Chaeton- ontidae), Histiopterus (Pentacerotidae), Cirrhitus (Cirrhitidae) and Schedophilus (Stromateoidei). Many other groups, among them the Serranidae (sensu Gosline 1966), Cichlidae, Percidae, Acanthuroidei, etc. differ from this basal type only in the loss or incorporation in the stegural of the second uroneural. The occurrence of an apparently identical type of caudal skeleton in such a wide range of perciform groups, including forms with lunate, forked, emarginate and rounded caudal fins 88 THE CAUDAL SKELETON IN in habitats ranging from pelagic to lacustrine, suggests that caudal anatomy is unlikely to contribute much to the unravelling of lineages among generalized Perciformes. A caudal skeleton very similar to the basal perciform type occurs in some advanced Berycoidei (living Holocentridae and Diretmidae differ only in having the stegural fused with the underlying centrum (Fig. 20), Berycidae have the stegural fused with the centrum and also retain the sixth hypural (Regan 1911, fig. i)) but here the hypurals support nineteen principal rays. In Myctophoidei a slightly different sequence of fusion is followed in which the second ural centrum partially retains its individuality. So far as I know, the basal perciform caudal skeleton is not precisely duplicated elsewhere. The differences between the caudal skeleton and fin of basal Perciformes and those of generalized Beryciformes (Polymixia, Aipichthys, Pycnosteroides, Monocentris) are: (i) Reduction of the neural spine of the second pre-ural centrum. This has already taken place in all Berycoidei. (ii) Fusion of the second ural centrum with the preceding two centra. This has taken place in most living Berycoidei and occurs in some individuals of Aipichthys. (iii) Loss of the sixth hypural (the development of the caudal skeleton in Mugil and Sphyraena (Hollister 1937) suggests that Perciformes have lost the sixth hypural, not incorporated it in the fifth). This has already occurred in some individuals of Aipichthys and Pycnosteroides, and takes place during the evolution of the Berycoidei (Holocentridae, Diretmus] . (iv) Reduction of the number of principal caudal rays from nineteen (Berycoidei, Aipichthys and Pharmacichthys] or eighteen (Polymixioidei, Dinopteryx and Pycnos- teroides) to seventeen. (v) In all living Beryciformes, the foremost procurrent rays in each caudal lobe are true spines, but in Perciformes they are usually (? always) unsegmented lepido- trichia, with the right and left halves separate, and this is true of the earliest Perci- formes (Prolates). In this character Perciformes are more primitive than living Beryciformes. Among Cretaceous Beryciformes, all Berycoidei, like their living relatives, have spines in front of the caudal fin; in Polymixioidei procurrent spines occur only in Homonotichthys and Pycnosterinx dubius, already known to be evolving towards the living Polymixia (Patterson 1964 : 301, 380), and in Dinopterygoidei they occur only in Dinopteryx. It is striking to find that spinous procurrent caudal rays occur only in those Cretaceous genera (except Dinopteryx} already known to be closely related to riving Beryciformes. The first of these five differences, the condition of the neural spine of the second pre-ural centrum, is the most interesting. As discussed above (p. 84) the primitive condition of this structure in teleosts seems to be as in Flops, Aulopus and Nematono- tus, where the spine is slender and about half as long as its predecessor. This type of spine may elongate so that it supports procurrent fin-rays, as in Polymixioidei and many primitive telosts, or it may become reduced to the percoid condition (Fig. 28) . But apart from these two simple alternatives there are other possibilities which complicate the issue. First, the fully developed spine might become detached as an epural, producing the percoid condition direct. Secondly, from a low neural crest MESOZOIC ACANTHOPTERYGIAN FISHES 89 of percoid type an apparent full neural spine might develop secondarily by fusion with the first epural. Thirdly, an apparent neural spine on the second pre-ural centrum might be produced by fusion between the second and third pre-ural centra. The last of these possibilities can normally be recognized by the partial or complete doubling of the neural or haemal spine on the compound centrum, as in Pleuronectes (Harrington 1937, fig. i) and the specimen of Saurida illustrated by Weitzman (1967, fig. 18). The abnormal specimen of Monocentris illustrated in Fig. 14 is evidently a special case of this type of fusion, where the neural and haemal spines of the third pre-ural centrum are normally double and the posterior half of the neural spine has become attached to the succeeding centrum. This type of fusion does not seem of general significance in the present discussion. Fusion of an epural with the neural crest of the second pre-ural centrum to produce a secondary neural spine (Fig. 28E) is a common occurrence in acanthopterygians : this process appears to account for the complete neural spine on the second pre-ural centrum in such groups as the Nandidae (but not Pristolepis ; Gosline 1968, fig. 2b) among Percoidei, the Channiformes, Anabantoidei, Luciocephalus , some scombroids, pleuronectoid and soleoid pleuronectiforms, tetraodontiforms, etc. (Monod 1967; Liem 1963, 1967; Gosline 1968). In Psettodes, the most primitive living pleuronecti- form, the caudal skeleton (Monod 1967, fig. 13) is of basal perciform type, with five autogenous hypurals and two uroneurals, but there is only one free epural and there appears to be a neural spine on the second pre-ural centrum. Monod identifies this spine as the first epural, for the element is partially or completely autogenous and the suture at the base lies not between the arch and the centrum but between the arch and the spine. Psettodes demonstrates clearly that the neural spine of the second pre-ural centrum in Pleuronectoidei and Soleoidei is an epural which has secondarily regained contact with and fused with a neural arch. That this has also occurred in scombroids such as Neothunnus can be seen by comparing figs. 15 and 16 of Monod (1967) . A further peculiarity of the caudal skeleton of pleuronectoids and soleoids is that the parhypural tapers proximally and fails to make contact with the centrum (Monod 1967 : 117). The effect of this is to give dorso- ventral sym- metry to the caudal skeleton, the free parhypural opposing the single epural just as the neural and haemal spines of the second pre-ural centrum oppose one another. A free parhypural, tapering proximally, also occurs in acanthopterygians such as the Channiformes (Monod 1967 : 117; Gosline 1968, fig. 20), most Anabantoidei (Liem 1963 : 32), Luciocephalus (Liem 1967 : 114) and balistoids (Whitehouse 1910, pi. 50, fig. 33; Monod 1967 : 117), all forms with a neural spine on the second pre- ural centrum. Since none of these fishes has more than two epurals, all these groups appear to be cases of secondary fusion between the first epural and the second pre-ural neural arch in order to increase the dorso-ventral symmetry of the caudal skeleton. A complete neural spine on the second pre-ural centrum also occurs occasionally in basal percoids: Fig. 26 shows such a structure in a large specimen of Siniperca (Percichthyidae). In this individual there is a perfectly formed neural arch and spine (npu2] fully fused to the second pre-ural centrum, and there are only two epurals (ei, 2] compared with the three of normal Siniperca, most percichthyids and serranids. This specimen is best regarded as an abnormality foreshadowing the go THE CAUDAL SKELETON IN fusion of the first epural with the second pre-ural neural arch in nandids, pleuro- nectiforms, etc. It is thus well established that in many perciform groups and perciform derivatives the first epural can fuse with the second pre-ural neural arch. This raises the question of the homology of the perciform first epural: does the frequent fusion of this bone with the second pre-ural centrum indicate that these two structures were h5. FIG. 26. Siniperca chuatsi (Basilewsky). The caudal skeleton of a dried skeleton showing a complete neural spine on the second pre-ural centrum, 1888.3.23.3, standard length 340 mm., Kiu Kiang, China. For explanation of lettering see p. 102. originally part of the same segment? If so, the perciform first epural may have appeared by detachment of a fully developed neural spine in a caudal skeleton of polymixiid type. Rosen (1964 : 244) suggested that this took place in the evolution of the exocoetoids : that the ancestral exocoetoid had only two epurals and a fully developed neural spine on the second pre-ural centrum which became detached as the foremost of the three epurals in such a fish as Dermogenys. If this neural spine became detached in a fish which still retained the original three epurals one would expect there to be four epurals. Such a condition occurs very occasionally in MESOZOIC ACANTHOPTERYGIAN FISHES .d2 h5 puUuUu2 5 mm FIG. 27. Pentaceropsis recurvirostris (Richardson). Caudal skeleton of a dried skeleton showing four epurals, 1869.2.24.24, standard length 405 mm., Tasmania. For explana- tion of lettering see p. 102. the generalized percoid caudal skeleton, as in the individual of Pentaceropsis (Pentacerotidae) shown in Fig. 27. The epurals of teleosts are the remnants of a more numerous set of bones in primi- tive actinopterygians which are serial homologues of the supraneurals above the anterior vertebrae, and like these bones they were primitively metameric (Patterson 1968 : 221), but the metameric arrangement of the epurals has been lost in living chondrosteans. In the most primitive teleosts, such as the Jurassic Leptolepis and Allothrissops (Patterson 19670:, figs. 3, 6), the three epurals still show an apparent metamery, suggesting that they represent the neural spines of the first pre-ural neural arch and two ural neural arches : this seems to be supported by conditions in osteoglossoid fishes, where there are often fully-developed neural spines on both the first pre-ural and the first ural centra, and where there is never more than one epural (Greenwood 1967). But in higher teleosts, as in living chondrosteans, a metameric arrangement of the epurals is no longer recognizable. At the percoid level, for example, one can find fishes with all the epurals behind the neural crest on the second pre-ural centrum (Gosline ig6ia, fig. i, Kuhlia; 1968, fig.^c.Bathymaster), or with one epural above the crest (Gosline ig6ia, fig. 2, Parupeneus; Hollister 1937, 92 THE CAUDAL SKELETON IN figs. 12-14, Sphyraena) or with two above it (Hollister 1937, fig. 8, Mugil). Further, in groups such as the Berycoidei and Percoidei there is good evidence that in primi- tive forms the first epural lies behind the neural crest of the second pre-ural centrum, moving forwards above it in more advanced forms (in berycoids cf. Figs. 17, 21 with Figs. 13, 20). Since there is no evidence that strict metamery of the epurals is maintained in acanthopterygians there is no reason to believe that the perciform first epural is the detached neural spine of the second pre-ural centrum. The individual of Pentaceropsis shown in Fig. 27 is best interpreted merely as exhibiting a supernumary epural: such a condition is already known to occur in the salmonid Oncorhynchus, where Vladykov found four epurals in three out of 1,020 specimens (1962, tableS). The conclusions drawn from this discussion of the second pre-ural neural arch and spine are that the low neural crest in basal percoids is to be regarded as having evolved by reduction of the primitive short neural spine, and that the ancestors of the perciforms are to be found among fishes having a low crest or short neural spine on the second pre-ural centrum, not among those with a full neural spine. Where there is a full neural spine on the second pre-ural centrum in Perciformes and perci- form derivatives, it is to be regarded as having arisen by fusion of an epural with a low neural crest. These conclusions are illustrated in Fig. 28. With this background on the perciform caudal skeleton, we can now consider the evidence of caudal structure in the various beryciform-perciform lineages that have been suggested (Patterson 1964). These were, in decreasing order of confidence, Polymixiidae (Omosoma Berycopsis lineage) > Scorpididae, Monodactylidae and Kyphosidae Aipichthyidae > Carangidae Pharmacichthyidae > Acanthuroidei Sphenocephalidae > basal Percoidei (Serranidae, etc.) Pycnosteroididae > Chaetodontidae Dinopterygidae > Centrarchidae Detailed study of the caudal skeleton cannot be said to give support to these lineages. Of the various perciform groups mentioned, the scorpidids, monodacty- lids, kyphosids, basal percoids (Centropomidae, Percichthyidae), Chaetodontidae and Centrarchidae have the generalized percoid type of caudal skeleton, differing from those of Beryciformes in the characters listed on p. 88. The acanthuroids are more advanced only in having lost the second uroneural, while the carangids have enlarged the first epural and show fusion between the first and second hypural and between the third and fourth hypurals. Among the beryciform groups, the polymixiids, sphenocephalids, pycnosteroidids and dinopterygids all have a com- plete neural spine on the second pre-ural centrum : for reasons given in the discussion above, it is unlikely that a percoid caudal skeleton can be derived directly from this condition. In Sphenocephalus there are only two epurals, suggesting the possibility MESOZOIC ACANTHOPTERYGIAN FISHES 93 of arriving at a percoid arrangement of three epurals and a low neural crest on the second pre-ural centrum by detachment of the neural spine on this centrum, but further study of Sphenocephalus has yielded strong evidence (to be discussed in a forthcoming paper by the author and D. E. Rosen) that far from being an ancestral percoid it is related to the percopsiforms. In the aipichthyids and pharmacichthyids FIG. 28. Diagrams to show changes in the second pre-ural neural spine (black) in the evolution of the myctophoids, ctenothrissiforms and acanthopterygians. A, the primi- tive short neural spine, as in Nematonotus, Aulopus, Aipichthys and Pharmacichthys; B, elongation of the neural spine to support procurrent fin-rays, as in Pateroperca, Poly- mixiidae, Dinopteryx and Pycnosteroides ; C, expansion of the neural spine into a plate, as in most myctophoids, Ctenothrissa, A ulolepis and some primitive Berycoidei ; D, reduction to a low crest, with the first epural (stippled) moving forwards above it, as in most Berycoidei and Perciformes (a berycoid is illustrated) ; E, production of a secondary neural spine by fusion of the first epural in fishes with no free second ural centrum and five hypurals, as in Zeiformes (except Caproidae), Nandidae, Channiformes, Anabantoidei, Pleuronectiformes, etc. 94 THE CAUDAL SKELETON IN the neural spine of the second pre-ural centrum is still short, as in the most primitive teleosts, and could give rise to the percoid condition by reduction. Further, in Aipichthys (Fig. n) there is a tendency to reduce the number of hypurals to five and also to fuse the second ural centrum into the preceding centrum, as in percoids. There seems to be nothing in the caudal skeleton of Aipichthys to oppose the postu- lated link with the carangids. In Pharmacichthys the caudal skeleton is like that of Aipichthys, so far as it is known, and the fin-rays are deeply cleft basally, as in carangids: this character reduces the possibility of a relationship between Phar- macichthys and the acanthuroids and balistoids, in which the caudal fin-rays are unmodified. Except for the Aipichthys-carangid lineage, the evidence of the caudal skeleton indicates that Gosline's (19660) criticism of the polyphyletic scheme of perciform origins which I proposed are well founded. But the evidence on which Gosline based these criticisms, the structure of the supraoccipital crest, does not in fact oppose the various beryciform-perciform lineages. Gosline (19660 : 412) contrasted the type of supraoccipital crest seen in the polymixiids Homonotichthys and Poly- mixia, and in the carangids, priacanthids, etc., which extends forwards between the frontals, is knife-edged and buried in musculature, with the type of crest seen in Antigonia, acanthuroids, chaetodontids, etc., which is short, high and thickened anteriorly, extending " up and back over the nape as a sort of protective shell ". Gosline finds that these two types of supraoccipital crest are " structurally and functionally ... far apart ", and writes " nor does it appear that one could be developed from the other except by going all the way back through some inter- mediate form with a relatively small, unspecialized occipital crest ". Gosline draws the conclusion that Aipichthys and Sphenocephalus , both with a short crest, thickened anteriorly, could not have given rise respectively to the carangids and serranids, which have a long, knife-edged crest. But the evolution of the Berycoidei shows (as clearly as such processes can be shown by the fossil record) that fishes with a short " Aipichthys-type " crest can give rise to fishes with a long " carangid-type " crest. The most primitive Berycoidei have an " Aipichthys-type, " crest, moderately high and thickened anteriorly : this is true of both the trachichthyid lineage (Lisso- beryx, Patterson 1967, fig. 2) and the holocentrid lineage (Caproberyx, Patterson 1964, fig. 67; 1967, fig. 10, Stichocentrus, Patterson 1967, fig. 8). From this basal type there are in berycoids two divergent trends in the evolution of the supra- occipital crest. In Holocentridae the frontals grow backwards, partially covering the parietals, eliminating the supratemporal fossa, and producing a low supra- occipital crest which secondarily comes to resemble the small supraoccipital crest of primitive teleosts. In Trachichthyidae (Hoplopteryx, Hoplostethus) the supra- temporal fossa extends forwards and the supraoccipital crest becomes elongated, thickened centrally and knife-edged. This trend continues further in the Berycidae, and Beryx has a long knife-edged crest, continued forwards by the frontals above the orbit, which resembles those of Polymixia and the carangids. If a carangid- type " supraoccipital crest developed from an " Aipichthys-type. " within the Berycoidei, there is no reason why these changes could not have occurred in other lineages. MESOZOIC ACANTHOPTERYGIAN FISHES 95 It should also be mentioned here that Gosline (19660 : 410) has indicated that Beryciformes differ from Perciformes in the number of infraorbitals and the extent of the subocular shelf. He finds that in Beryciformes (except Holocentridae) there are four circumorbitals behind the lachrymal whereas in Perciformes there are five, and that in Beryciformes the subocular shelf extends over more than one infra- orbital while in Perciformes it is confined to the second infraorbital. The subocular shelf extends along all the infraorbitals in Holocentridae (living and fossil) and in Polymixiidae (Polymixia, Homonotichthys) , but in Trachichthyidae (living and fossil), Berycidae, Monocentridae, etc., the shelf is restricted to the second infraorbital, as in Perciformes. The subocular shelf also extends along all the infraorbitals in Anabantidae and Belontiidae (Liem 1963) : probably a subocular shelf on all the infraorbitals, as in polymixiids, holocentrids and anabantids, is the primitive condition of the structure (Smith & Bailey 1962 : 3). In the number of infraorbitals the basic beryciform condition is undoubtedly as in percoids, with a total of six bones, a lachrymal, four infraorbitals, and a dermosphenotic overlying the autosphenotic : this condition occurs in Polymixiidae, living and fossil, and in living Holocentridae. In Trachichthyidae, Berycidae, Monocentridae and some Cretaceous Holocentridae (Caproberyx,Stichocentrus, Patterson 1964 : 347; 1967 : 89) the apparent reduction in number of circumorbital bones is caused by fusion of the dermo- and autosphenotics, which had already occurred in the Cretaceous trachichthyid Hoplopteryx (Patterson 1964, fig. 55). In summary, of the various beryciform-perciform lineages which have been proposed, only the Aipichthys carangid lineage emerges unscathed from a detailed examination of caudal structures. The caudal skeletons of Polymixiidae, Sphen- ocephalus, Dinopteryx and Pycnosteroides differ fundamentally from those of Perci- formes in having a full neural spine on the second pre-ural centrum. In Aipichthys and Pharmacichthys there are tendencies towards the Perciformes in the occasional fusion of the second ural centrum with the preceding centrum and the occurrence of five hypurals, but it is only among the Berycoidei that the basal perciform caudal skeleton is duplicated. All known Berycoidei, even the very generalized Lissoberyx, are more specialized than Perciformes in having procurrent caudal spines, and this and characters of the skull (Patterson 1964 : 467) show that no perciform could have evolved from any known berycoid. But if the Perciformes and cognate groups (Channiformes, Scorpaeniformes, Pleuronectiformes, Tetraodontiformes) should prove to be a monophyletic group, an alternative to the traditional method " an attempt first to define orders and other higher taxa and then to speculate upon their origin, albeit in the light of the known fossils " (Greenwood et al. 1966 : 346) is to use the criteria recommended by Hennig (1966 : 88, 120) and to search among the living fauna for the sister group (Hennig 1966 : 139; see also Brundin 1966 : 17) of this assemblage. A preliminary analysis suggests that the Berycoidei, not the Beryciformes as a whole, may fill this role. This is indicated not only by features of the caudal skeleton but by the fact that the Berycoidei is the only beryciform group showing such perciform features as a subocular shelf confined to the second infraorbital, the absence of epineurals, the pelvic girdle firmly joined to the cleithra (in Berycidae especially), pelvic fins containing a spine and five soft rays (in Anomalo- 96 THE CAUDAL SKELETON IN pidae and Gibberichthyidae, for example), partially separate soft and spinous dorsal fins, etc., and in the stephanoberycoids, which appear to be merely specialized offshoots of the trachichthyid lineage, loss of the orbitosphenoid. While it is clear that many of these perciform characters in Berycoidei have arisen independently within the group and were not inherited from a common ancestor of Berycoidei and perciforms, they appear to be true parallelisms (Simpson 1961 : 78), and are indicative of relationship. (c) Intermediate groups. Between the Beryciformes and the basal Percoidei, Greenwood et al. (1966 : 398) place the Zeiformes, Lampridiformes, Gasterosteiformes, Channiformes, Synbranchi- formes, Scorpaeniformes, Dactylopteriformes and Pegasiformes. Of the Gastero- steiformes, Scorpaeniformes, Dactylopteriformes and Pegasiformes I have nothing to say. The Channiformes (see Gosline 1968) and Synbranchiformes are probably derived from the percoid level rather than from the beryciform or pre-beryciform. In Zeiformes the caudal skeleton resembles those of basal Perciformes in having no free second ural centrum and in having only five hypurals, and is more advanced than basal percoids in having lost the second uroneural. In Zeidae (Zeus, Cyttus], Oreosomatidae (Neocyttus) and Grammicolepidae (Xenolepidichthys] there is a com- plete neural spine on the second pre-ural centrum, but this never occurs in conjunc- tion with three epurals, and in Caproidae (Capros, Antigonia; Gosline 1961, fig. 4A), which in other respects appear to have the most primitive caudal skeletons of the group (the hypurals autogenous, the stegural autogenous in Antigonia) there are three epurals and there is a low crest on the second pre-ural centrum, as in percoids. Conditions in the Caproidae indicate that the neural spine on the second pre-ural centrum in Zeidae, Oreosomatidae and Grammicolepidae has arisen secondarily by fusion of the first epural (see above, p. 89, Fig. 28), and that as Gosline (1961) has already said, there is nothing in the caudal skeleton to distinguish Zeiformes from Perciformes. It has long been recognized that the Zeiformes are probably related to the Beryciformes, principally because of the pelvic ray count, but they also show many perciform features (Gosline 1961 : 36) and no conclusion has yet been reached on whether they are more closely related to the beryciforms or the perciforms, although in most recent classifications they are placed directly after the Beryci- formes. Stinton (1967) has recently shown that there is a remarkable resemblance between the otoliths of Antigonia and those of Berycoidei (Berycidae, Trachichthyi- dae, and especially Monocentridae) . Stinton interprets this as indicating that Antigonia is a berycoid, for he finds that the otoliths of Capros resemble those of the zeids rather than Antigonia. However, the evidence that Antigonia and Capros are related can hardly be ignored, and in Stinton's illustrations of zeiform otoliths it seems possible to recognize a trend in reduction and specialization of the otolith in the sequence Antigonia-Capros-Cyttus-Zeus: the same sequence of increasing specialization is also shown by fusion within the caudal skeleton. In my opinion Stinton's otolith evidence indicates not that Antigonia is a berycoid, but that as the most primitive living zeiform it retains the clearest evidence of a common ancestry with the Berycoidei. The percoid-like caudal skeleton, pelvic spine, etc., of the MESOZOIC ACANTHOPTERYGIAN FISHES 97 Zeiformes do not oppose such a relationship. In Hennig's terminology, the Zei- formes appear to be the apomorph sister group of the Berycoidei, these two groups together forming the plesiomorph sister group of the perciform assemblage (see above, p. 95). The Lampridiformes, previously unknown before the Oligocene, have recently acquired a respectable fossil record with Bonde's (1966) preliminary description of a ? veliferid from the basal Eocene Mo-clay of Denmark, his opinion that Palaeo- centrotus Kuhne (1941), from the same beds, is a lampridoid, not a zeiform, and the suggestion (p. 81) that the Danian Bathysoma is a lampridiform, not a menid. It appears that the deep-bodied Lampridiformes of the suborder Lampridoidei were an important element of early Tertiary faunas. The caudal skeleton of Lampridi- formes (known in Velifer, Lampris, Palaeocentrotus and Bathysoma) is characterized by fusion of one or more of the upper hypurals with the second ural centrum, but in other respects it does not differ from that of Beryciformes (there are six hypurals in Velifer). In Lampris and Velifer the neural spine of the second pre-ural centrum is reduced, but in Palaeocentrotus (Kiihne 1941, fig. 2) it is about half as long as its predecessor, a primitive condition only found among Beryciformes in Aipichthys (Fig. n) and Pharmacichthys. Further, in Lampris and Velifer there are seventeen branched caudal rays and the caudal rays are deeply cleft basally, covering much of the hypurals. These points tend to confirm the suggestion (Patterson 1964 : 473) that the Lampridiformes are an offshoot of the Dinopterygoidei. Within the Dino- pterygoidei, the second pre-ural neural spine, the seventeen branched principal rays, the " hypurostegy " and the absence of a pelvic spine all point to the Aipich- thyidae and Pharmacichthyidae as ancestral forms. Transference of Bathysoma to the Lampridiformes leaves unsettled the position of Mene, to which I thought Bathysoma was related (Patterson 1964 : 424). The similarities between Mene and Bathysoma are numerous, and extend to the form of the fin-rays, which are preserved in the middle part of the anal fin of the holotype of B. lutkeni in Copenhagen, and are short, broad and unbranched (Bonde, personal commn). The skull of Mene is also strikingly like that of Velifer (Regan 1907, figs. 167, 169). But the caudal skeletons of Mene and the Lampridiformes are very different. The six hypurals (in Velifer}, fusion of the second ural centrum with the upper hypurals rather than with the preceding centrum, and the seventeen branched principal rays of Lampridiformes can only be derived from beryciform or pre-beryciform ancestors, but the caudal skeleton of Mene, with fifteen branched principal rays and fusion of the second ural centrum and first four hypurals with the preceding centrum could have evolved from the caudal skeleton of basal percoids (p. 81). Nor can I find anything in the skull and vertebral column of Mene which is against perciform ancestry. Nevertheless, Mene has an unusually long fossil record, extending back to the Lower Palaeocene. The earliest recorded species, M. phosphaticus Astre (1927) from the Montian of Tunisia, seems to agree with Mene rather than with the Lampridiformes in caudal structure. Mene is very com- mon in the Middle Eocene of Monte Bolca: with Mene at Monte Bolca there occur other deep-bodied fishes of similar structure such as Exellia ( = Semiophorus) . These fishes, though poorly known, seem to agree with Mene in the absence of spines 9 8 THE CAUDAL SKELETON IN in the dorsal and anal fins (see Blot 1967 on Exellia} and also in the deeply cleft bases of the caudal fin-rays. It is difficult to know how much significance can be attached to this last character. E. & Y. Le Danois (1964), who coined the term " hypurostegy " for it, give great importance to it and use it to unite in an " ordre des Scombres " fishes as diverse as the holostean Pachycormidae, the Cretaceous Tselfatiidae, the lampridiform Veliferidae and Lampridae, the carangids, scombrids and others. In my opinion this assemblage is entirely spurious; the occurrence of hypurostegy in such varied groups indicates not that the fishes are related but that hypurostegy has arisen independently in a number of lines, for reasons as yet unknown. E. & Y. Le Danois made a new family Vomeridae to include Mene, the Eocene Vomeropsis, and the deep-bodied carangids Vomer, Selene, Alectis and Hynnis, grouping this family with the Lampridae, Veliferidae, Ephippidae and Exellia. Although the reasoning on which this grouping is made is doubtful, there may be some truth in it. It seems possible that Mene and Exellia could represent an independent attainment of the perciform grade from the Palaeocene lampridiform stock. But the possibility that the resemblances between these Eocene forms and the Lampridiformes are due to convergence is by no means ruled out: revisionary studies on the Monte Bolca fauna now in progress (Blot 1967) may settle this question. IV. CONCLUSIONS Study of a single structural complex such as the caudal skeleton is unlikely to produce firm conclusions on matters of phylogeny and relationships. Rather it will serve as a means of checking existing hypotheses and will raise questions to be settled by more comprehensive work. The main points arising from this paper are as follows. An asterisk indicates that the genus or group is extinct. 1. Ctenothrissiformes,* Myctophoidei and Beryciformes have a basically similar caudal skeleton, with the first ural and pre-ural centra fused, a free second ural centrum, a stegural (the first uroneural fused with neural arch material from the first ural and pre-ural centra), a second uroneural, three epurals and six hypurals. The primitive nineteen principal caudal rays are retained in all ctenothrissiforms and myctophoids, and in all beryciforms except the Polymixiidae, Sphenocephalidae,* Dinopterygidae * and Pycnosteroididae,* which have eighteen. 2. Ctenothrissiformes * (Aulolepis, Ctenothrissa) and Myctophoidei (Aulopus, Nematonotus,* Sardinioides* Acrognathus *) have a large caudal scute above and below the caudal skeleton. The myctophoids Nematonotus * and Sardinioides attenuatus * have a single urodermal on the base of the upper caudal rays. Both caudal scutes and urodermals are relict structures absent in all higher groups. 3. The neural spine of the second pre-ural centrum in teleosts is primitively slender and about half as long as its predecessor, as in Elops and Leptolepis.* This type of second pre-ural neural spine persists in the myctophoids Aulopus and Nematonotus * and in the beryciforms Aipichthys * and Pharmacichthys.* In Ctenothrissiformes,* Ctenothrissa and Aulolepis have this short spine expanded into a plate, as it is in most myctophoids, but in Pateroperca the spine is elongate and supports procurrent fin-rays. A fully developed second pre-ural neural spine also occurs among Beryci- formes in all polymixioids (Polymixiidae, Sphenocephalidae*) and in the dinop- MESOZOIC ACANTHOPTERYGIAN FISHES 99 terygoids Dinopfayx* and Pycnosteroides.* In Berycoidei, as in generalized Perci- formes, the second pre-ural neural arch and spine are reduced to a low crest. 4. The differences between the caudal skeletons of Myctophoidei and Cteno- thrissiformes* and those of the most primitive teleosts are minor. Among primitive teleosts the Clupavidae * come closest to the Ctenothrissiformes and Myctophoidei in caudal anatomy. 5. The foremost procurrent caudal fin-rays are spinous in all living Beryciformes. This is a feature peculiar to Beryciformes: Perciformes appear to be without procurrent caudal spines. Procurrent caudal spines also occur in some species of Ctenothrissa and in Aulolepis (Ctenothrissiformes*), probably an indication of rela- tionship between Ctenothrissiformes and Beryciformes, although procurrent caudal spines have also developed in advanced members of the myctophoid family Mycto- phidae. Among Cretaceous Beryciformes, procurrent caudal spines occur only in forms already known to be closely related to living Beryciformes (all Berycoidei and the polymixiids Homonotichthys* and Pycnosterinx*) and in Dinopteryx.* 6. Within the Beryciformes, the principal variations in the caudal skeleton and fin (apart from those in the second pre-ural neural spine, principal fin-ray count and procurrent fin-rays already mentioned) are the presence of only two epurals in Sphenocephalus* (Polymixioidei), the presence of only five hypurals in Aipichthys velifer* some specimens of Pycnosteroides* (both Dinopterygoidei) and in many Berycoidei (living holocentrids, Diretmus), and the fusion of the second ural centrum with the preceding centrum in some specimens of Aipichthys* and in many Bery- coidei (Berycidae, Diretmidae, Anoplogasteridae, living Holocentridae) . A full neural spine on the second pre-ural centrum has been found in one individual of Aipichthys* and one of Monocentris, simulating the polymixoid condition. 7. The basal perciform caudal skeleton differs from the basal beryciform type in having only five hypurals, no free second ural centrum and only seventeen principal rays. Perciformes are primitively characterized by having a low neural crest on the second pre-ural centrum. Where a complete second pre-ural neural spine occurs as a normal feature in perciform or higher groups the condition is secondary, the spine representing an epural which has secondarily fused with the neural arch. 8. Evidence from the caudal skeleton does not support the various independent beryciform perciform lineages which have been proposed. The polymixioids and the dinopterygoids Dinopteryx* and Pycnosteroides* differ fundamentally from the various perciform groups which they otherwise resemble in having a fully developed neural spine on the second pre-ural centrum. Pharmacichthys* which resembles the acanthuroids and balistoids in many ways, differs from them in having the bases of the caudal fin-rays deeply cleft, and is therefore unlikely to have been ances- tral to these groups. Only the postulated link between Aipichthys* and the caran- gids is not opposed by evidence from the caudal skeleton and fin. 9. Although no known berycoid could have been ancestral to any perciform, the Berycoidei is the only beryciform group in which the caudal skeleton evolves towards the percoid condition. There are many other characters and evolutionary trends in which the Berycoidei is the only beryciform group to resemble the percoids : these suggest that the Berycoidei is the sister group of the Perciformes and cognate groups. TOO THE CAUDAL SKELETON IN 10. The caudal skeleton shows that the Danian Bathysoma* is a lampridiform (the earliest yet known), not a member of the Menidae. It is suggested that the Lampridiformes originated from near the beryciform families Aipichthyidae* and Pharmacicthyidae.* In the Eocene there are fishes (Mene, Exellia*} which have reached the perciform grade but resemble the Palaeocene and Eocene Lampridi- formes : possibly these forms represent an independent attainment of the perciform grade from lampridiform ancestors. 11. The caudal skeleton of the Zeiformes is basically of percoid type, but the evidence of zeiform otoliths indicates that they are closely related to the Berycoidei. The Zeiformes is evidently the sister group of the Berycoidei, these two groups together being the sister group of the perciform assemblage. V. REFERENCES ARAMBOURG, C. 1954. L GS poissons cretaces du Jebel Tselfat (Maroc). Notes Mem. Sew. Mines Carte geol. Maroc, Rabat, 118 : 1-188, 20 pis. ASTRE, G. 1927. Le carangide des phosphates Tunisiens. Bull. Soc. Hist. nat. Toulouse, 56 : 501-504, pi. 5. BARRINGTON, E. J. W. 1937. The structure and development of the tail in the plaice (Pleuro- nectes platessd) and the Cod (Gadus morrhua). Q. Jl microsc. Sci., London, 79 : 447-469, 25 figs. BERG. L. S. 1940. Classification of fishes, both recent and fossil. Trav. Inst. Zool. Acad. Sci. URSS, Leningrad, 5 : 87-517, 190 figs. BERRY, F. H. & ROBINS, C. R. 1967. Macristiella perlucens, a new clupeiform fish from the Gulf of Mexico. Copeia, Washington, 1967 : 46-50, 4 figs. BLOT, J. 1967. Quelques remarques preliminaires concernant la faune ichthyologique du Monte Bolca (Italic). Collogues int. Cent. Natn. Rech. scient., Paris, 163: 133-138, i pi. BONDE, N. 1966. The fishes of the Mo-clay Formation (Lower Eocene). Meddr. dansk geol. Foren., K0benhavn, 16 : 198-202. BRUNDIN, L. 1966. Transantarctic relationships and their significance, as evidenced by chironomid midges. K. svensha VetenskAhad. Handl., Stockholm, (4) 11, i : 1-472, 30 pis. CAVENDER, T. 1966. The caudal skeleton of the Cretaceous teleosts Xiphactinus, Ichthyo- dectes and Gillicus, and its bearing on their relationship with Chirocentrus. Occ. Pap. Mus. Zool. Univ. Mich., Ann Arbor, 650 : 1-15, i pi. DAVIS, J. W. 1887. The fossil fishes of the Chalk of Mount Lebanon, in Syria. Sci. Trans. R. Dublin Soc., (2) 3 : 457-636, 25 pis. 1890. On the fossil fish of the Cretaceous formations of Scandinavia Sci. Trans. R. Dublin Soc., (2) 4 : 367-434, 19 pis. EJEL, F. & DUBERTRET, L. 1966. Sur 1'age precis du gisement de Poissons et de Crustaces cretaces de Sahel Alma (Liban). C. r. Seanc. Soc. geol. Fr., Paris, 1966 : 353-354, i fig. FRASER-BRUNNER, A. 1949. A Classification of the Fishes of the Family Myctophidae. Proc. Zool. Soc. Lond., 118 : 1019-1106, 167 figs. GOSLINE, W. A. 1960. Contributions toward a classification of modern isospondylous fishes. Bull. BY. Mus. nat. Hist. (Zool.), London, 6 : 325-365, 15 figs. 1961. Some osteological features of modern lower teleostean fishes. Smithson. Misc. Coll., Washington, 142, 3 : 1-42, 8 figs. 19610. The Perciform Caudal Skeleton. Copeia, Ann Arbor, 1961 : 265-270, 3 figs. 1963. Considerations regarding the relationships of the percopsiform, cyprinodontiform and gadiform fishes. Occ. Pap. Mus. Zool. Univ. Mich., Ann Arbor, 629 : 1-38, 11 figs. 1965. Teleostean Phylogeny. Copeia, Washington, 1965 : 186-194, i fig. 1966. The limits of the fish family Serranidae, with notes on other lower percoids. Proc. Calif. Acad. Sci., San Francisco, (4) 33 : 91-112, io figs. MESOZOIC ACANTHOPTERYGIAN FISHES 101 GOSLINE, W. A. 19660. Comments on the Classification of the Percoid Fishes. Pacific Sci., Honolulu, 20 : 409-418, 2 figs. 1968. The Suborders of Perciform Fishes. Proc. U.S. Nat. Mus., Washington, 124: 1-77, 12 figs. GOSLINE, W. A., MARSHALL, N. B. & MEAD, G. W. 1966. Order Iniomi. Characters and synopsis of families. Mem. Sears Fdn Mar. Res., New Haven, 1, 5 : 1-18, 6 figs. GREENWOOD, P. H. 1967. The caudal skeleton in osteoglossoid fishes. Ann. Mag. nat. Hist., London, (13) 9 : 581-597, 12 figs. GREENWOOD, P. H., ROSEN, D. E., WEITZMAN, S. H. & MYERS, G. S. 1966. Phyletic studies of Teleostean fishes, with a provisional classification of living forms. Bull. Amer. Mus. Nat. Hist., New York, 131 : 339-456, pis. 21-23, 3 2 charts. HAY, O. P. 1903. On a collection of Upper Cretaceous fishes from Mount Lebanon, Syria, with descriptions of four new genera and nineteen new species. Bull. Amer. Mus. Nat. Hist., New York, 19 : 394-452, 14 pis. HENNIG, W. 1966. Phylogenetic Systematics. 263 pp., 69 figs. Translated by D. D. Davis & R. Zangerl. Univ. of Illinois, Urbana. HOLLISTER, G. 1937. Caudal skeleton of Bermuda Shallow Water Fishes. II. Order Per- comorphi, Suborder Percesoces: Atherinidae, Mugilidae, Sphyraenidae. Zoologica, N.Y., 22 : 265-280, 14 figs. 19370. Caudal Skeleton of Bermuda Shallow Water Fishes. III. Order Iniomi: Syno- dontidae. Zoologica N.Y., 22 : 385-399, 1 8 figs. HUSSAKOF, L. 1929. A new teleostean fish from the Niobrara of Kansas. Amer. Mus. Novit., New York, 357 : 1-4, 2 figs. KUHNE, W. G. 1941. A new Zeomorph Fish from the Paleocene Moler of Denmark. Ann. Mag. nat. Hist., London, (n) 7 : 375-386, 3 figs. LE DANOIS, E. & LE DANOIS, Y. 1964. L'Ordre des Scombres. Mem. Inst. fr. Afr. noire, Dakar, 68 : 153-192, 19 figs. LIEM, K. F. 1963. The comparative osteology and phylogeny of the Anabantoidei (Tele- ostei, Pisces). Illinois biol. Monogr., Urbana, 30 : 1-149, 104 figs. 1967. A morphological study of Luciocephalus pulcher, with notes on gular elements in other recent teleosts. /. Morph., Philadelphia, 121 : 103-134, 23 figs. MARSHALL, N. B. 1961. A young Macristium and the Ctenothrissid fishes. Bull. Br. Mus. nat. Hist. (Zool.), London, 7 : 353-370, 4 figs. MONOD, T. 1967. Le complexe urophore des Teleosteens: typologie et evolution (note pre- liminaire). Colloques int. Cent. natn. Rech. scient., Paris, 163: 111-131, 16 figs. NORDEN, C. R. 1961. Comparative osteology of representative Salmonid Fishes, with particu- lar reference to the Grayling (Thymallus arcticus] and its Phylogeny. /. Fish. Res. Bd. Can., Ottawa, 18 : 679-791, i6pls. NURSALL, J. R. 1963. The hypurapophysis, an important element of the caudal skeleton. Copeia, Ann Arbor, 1963 : 458-459. NYBELIN, O. 1963. Zur Morphologie und Terminologie des Schwanzskelettes der Actinop- terygier. Ark. Zool., Stockholm, (2) 15 : 485-516, 22 figs. PATTERSON, C. 1964. A review of Mesozoic acanthopterygian fishes, with special reference to those of the English Chalk. Phil. Trans. R. Soc., London, (B) 247 : 213-482, pis. 2-5. 1967. New Cretaceous berycoid fishes from the Lebanon. Bull. Br. Mus. nat. Hist. (Geol.), London, 14 : 67-110, 4 pis. 19670. Are the teleosts a polyphyletic group? Colloques int. Cent. natn. Rech. scient., Paris, 163:93-109, n figs. 1968. The caudal skeleton in Lower Liassic pholidophorid fishes. Bull. Br. Mus. nat. Hist. (Geol.), London, 16:201-239, 5 pis. REGAN, C. TATE. 1907. On the anatomy, classification and systematic position of the teleo- stean fishes of the suborder Allotriognathi. Proc. Zool. Soc. Land., 1907 : 634-643, figs. 166-171. 102 THE CAUDAL SKELETON IN AC ANTHOPTERYGI AN FISHES REGAN, C. TATE 1910. The caudal fin of the Elopidae and of some other teleostean fishes. Ann. Mag. nat. Hist., London, (8) 6:354-358, 2 figs. 191 1. The anatomy and classification of the teleostean fishes of the Orders Berycomorphi and Xenoberyces. Ann. Mag. nat. Hist., London, (8) 7 : 1-9, pi. i. 191 1. The Anatomy and Classification of the Teleostean Fishes of the Order Iniomi. Ann. Mag. nat. Hist., London, (8) 7 : 120-133, 7 figs. ROSEN, D. E. 1962. Comments on the Relationships of the North American Cave Fishes of the Family Amblyopsidae. Amer. Mus. Novit., New York, 2109 : 1-35, 24 figs. 1964. The relationships and taxonomic position of the halfbeaks, killifishes, silversides, and their relatives. Bull. Amer. Mus. Nat. Hist., New York, 127 : 217-268, pis. 14, 15. SIMPSON, G. G. 1961. Principles of Animal Taxonomy, xii + 247 pp., 30 figs. Columbia Univ., New York. SMITH C. L. & BAILEY, R. M. 1962. The Subocular Shelf of Fishes. /. Morph., Philadelphia, 110 : 1-18, 3 pis. SMITH, J. L. B. 1951. The Fishes of the Family Veliferidae from South Africa. Ann. Mag. nat. Hist., London, (12) 4 : 497-510, pis. 10-12. STARKS, E. C. 1904. The osteology of some Berycoid fishes. Proc. U.S. Nat. Mus., Washing- ton, 27 : 601-619, 9 figs. STINTON, F. C. 1967. The otoliths of the Teleostean fish Antigonia capros and their taxonomic significance. Bocagiana, Funchal, 13 : 1-7, 2 pis. VLADYKOV, V. D. 1954. Taxonomic characters of the eastern North American chars (Salveli- nus and Cristivomer] . J. Fish. Res. Bd. Can., Ottawa, 11 : 904-932, 12 figs. 1962. Osteological studies on Pacific salmon of the genus Oncorhynchus . Bull. Fish. Res. Bd. Can., Ottawa, 136 : 1-172, 89 figs. WEITZMAN, S. H. 1962. The osteology of Brycon meeki, a generalized characid fish, with an osteological definition of the family. Stanford ichthyol. Bull., Palo Alto, 8 : 1-77, 21 figs. 1967. The origin of the stomiatoid fishes with comments on the classification of salmoni- form fishes. Copeia, Washington, 1967 : 507-540, 18 figs. WHITEHOUSE, R. H. 1910. The caudal fin of the Teleostomi. Proc. zool. Soc. Lond., 1910 : 590-627, pis. 47-50. WOODWARD, A. SMITH. 1901. Catalogue of the fossil fishes in the British Museum (Natural History}. 4. xxxviii + 636 pp., 19 pis. Brit. Mus. (Nat. Hist.), London. 1942. Some new and little-known Upper Cretaceous fishes from Mount Lebanon. Ann. Mag. nat. Hist., London, (n) 9 : 537-568, 5 pis. VI. ABBREVIATIONS USED IN FIGURES dz second, uroneural e 1-3 epurals f.s caudal scute h 1-6 hypurals hpu2 haemal spine of second pre-ural centrum npu2, npu3 neural spines of second and third pre-ural centra ph parhypural (haemal spine of first pre-ural centrum) pui + ui centrum formed by fusion of first pre-ural and ural centra puz, pus, pu4 second, third and fourth pre-ural centra st stegural (first uroneural fused with pre-ural neural arch material) U2 second ural centrum Combinations of symbols linked by plus signs indicate compound elements formed by fusion of the bones indicated. PRINTED IN GREAT BRITAIN BY ADLARD & SON LIMITED BARTHOLOMEW PRESS, DORKING NON-CALCAREOUS MICROPLANKTON FROM THE CENOMANIAN OF ENGLAND, NORTHERN FRANCE AND NORTH AMERICA PART I R. J. DAVEY BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 17 No. 3 LONDON: 1969 NON-CALCAREOUS MICROPLANKTON FROM THE CENOMANIAN OF ENGLAND, NORTHERN FRANCE AND NORTH AMERICA PART I BY ROGER JACK DAVEY Pp. 103-180; ii Plates; 16 Text-figures BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 17 No. 3 LONDON: 1969 THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted in 1949, is issued in five series corresponding to the Departments of the Museum, and an Historical series. Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year. In 1965 a separate supplementary series of longer papers was instituted, numbered serially for each Department. This paper is Vol. 17, No. 3 of the Geological (Palaeontological] series. The abbreviated titles of periodicals cited follow those of the World List of Scientific Periodicals. World List abbreviation : Bull. Br. Mus. nat. Hist. (Geol.) Trustees of the British Museum (Natural History) 1969 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) Issued 17 January, 1969 Price 2 i6s, NON-CALCAREOUS MICROPLANKTON FROM THE CENOMANIAN OF ENGLAND, NORTHERN FRANCE AND NORTH AMERICA PART I By ROGER JACK DAVEY Manuscript accepted May 1967 CONTENTS SYNOPSIS I. INTRODUCTION Page 107 107 Acknowledgments . . . . . . . . no II. STRATIGRAPHIC LOCATION OF SAMPLES . . . . . 112 Fetcham Mill Borehole . . . . . . . 112 Compton Bay ......... 112 Speeton .......... 112 Hunstanton ......... 115 Devon .......... 115 Escalles Borehole . . . . . . . . 115 Saskatchewan ......... 115 Texas .......... 115 III. SYSTEMATIC DESCRIPTIONS . . . . . . . 120 Genus Gonyaulacysta Deflandre . . . . . . 120 Gonyaulacysta cassidata (Eisenack & Cookson) . . . 120 whitei Sarjeant . . . . . . 120 fetchamensis Sarjeant . . . . . 120 exilicristata sp. nov. . . . . . 121 delicata sp. nov. . . . . . 123 sp. A. 124 Genus Cribroperidinium Neale & Sarjeant . . . . 125 Cribroperidinium intricatum sp. nov. . . . . 125 Other species ......... 128 Cribroperidinium orthoceras (Eisenack) . . . . 128 Genus Carpodinium Cookson & Eisenack . . . . 129 Carpodinium obliquicostatum Cookson & Hughes . . . 129 Genus Ellipsodinium Clarke & Verdier . . . . . 129 Ellipsodinium rugulosum Clarke & Verdier . . . . 130 Genus Apteodinium Eisenack . . . . . . 130 Apteodinium granulatum Eisenack . . . . . 130 Genus Trichodinium Eisenack & Cookson . . . . 131 Trichodinium castaneum (Deflandre) . . . . . 131 Genus Microdinium Cookson & Eisenack . . . . 132 Microdinium cf. ornatum Cookson & Eisenack . . . 132 setosum Sarjeant . . . . . . 133 distinctum sp. nov. . . . . . 133 variospinum sp. nov. . . . . . 135 veligerum (Deflandre) . . . . . 136 crinitum sp. nov. . . . . . 137 GEOL. 17, 3 i 106 CENOMANIAN NON-CALCAREOUS MICROPLAN KTON, i Genus Histiocysta nov. . . . . . . . 138 Histiocysta palla sp. nov. . . . . . . . 138 Genus Fromea Cookson & Eisenack . . . . . 140 Fromea amphora Cookson & Eisenack . . . . 140 Genus Chytroeisphaeridia Sarjeant . . . . . 140 Chytroeisphaeridia euteiches sp. nov. ..... 141 Genus Cassiculosphaeridia nov. . . . . . . 141 Cassiculosphaeridia reticulata sp. nov. ..... 142 Genus Epelidosphaeridia nov. . . . . . . 142 Epelidosphaeridia spinosa (Cookson & Hughes) . . . 143 Genus Hystrichosphaeridium Deflandre . . . . . 143 Hystrichosphaeridium tubiferum (Ehrenberg) . . . 143 deanei Davey & Williams . . . 144 readei Davey & Williams . . . 144 radiculatum Davey & Williams . . 144 mantelli Davey & Williams . . 145 bowerbanki Davey & Williams . . 145 difficile Manum & Cookson . . . 145 Genus Oligosphaeridium Davey & Williams . . . . 146 Oligosphaeridium complex (White) . . . . . 146 reticulatum Davey & Williams . . . 147 prolixispinosum Davey & Williams . . 147 anthophorum (Cookson & Eisenack) . . 147 reniforme (Tasch) . . . . . 148 Genus Litosphaeridium Davey & Williams . . . . 148 Litosphaeridium siphoniphorum (Cookson & Eisenack) . . 148 Genus Polysphaeridium Davey & Williams . . . . 151 Polysphaeridium pumilum Davey & Williams . . . 151 laminaspinosum Davey & Williams . . 151 Genus Tanyosphaeridium Davey & Williams . . . . 151 Tanyosphaeridium variecalamum Davey & Williams . . 151 Genus Callaiosphaeridium Davey & Williams . . . . 151 Callaiosphaeridium asymmetricum (Deflandre & Courteville) . 152 Genus Cleistosphaeridium Davey, Downie, Sarjeant & Williams 152 Cleistosphaeridium heteracanthum (Deflandre & Cookson) . 152 multifurcatum (Deflandre) . . . 152 armatum (Deflandre) . . . . 153 polypes (Cookson & Eisenack) . . . 154 var. clavulum nov. . . . . . 154 huguonioti (Valensi) . . . . 155 var. pertusum nov. . . . . 156 flexuosum Davey, Downie, Sarjeant & Williams ...... 157 parvum sp. nov. ..... 157 aciculare sp. nov. . . . . . 158 Genus Surculosphaeridium Davey, Downie, Sarjeant & Williams 158 Surculosphaeridium longifurcatum (Firtion) . . . . 158 Genus Hystrichokolpoma Klumpp . . . . . . 159 Hystrichokolpoma ferox (Deflandre) . . . . . 159 Genus Prolixosphaeridium Davey, Downie, Sarjeant & Williams 160 Prolixosphaeridium conulum sp. nov. . . . . . 160 Genus Coronifera Cookson & Eisenack ..... 161 Coronifera oceanica Cookson & Eisenack . . . . 162 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 107 Genus Exochosphaeridium Davey, Downie, Sarjeant & Williams 162 Exochosphaeridium phragmites Davey et al. . . . . 163 pseudohystrichodinium (Deflandre) . . 163 striolatum (Deflandre) . . . . 164 var. truncatum nov. . . . . 164 Other species ......... 166 Genus Cyclonephelium Deflandre & Cookson . . . . 166 Cyclonephelium distinction Deflandre & Cookson . . . 166 membraniphorum Cookson & Eisenack . . 167 vannophorum sp. nov. . . . . 168 paucispinum sp. nov. . . . . 170 eisenacki sp. nov. . . . . . 170 Genus Adnatosphaeridium Williams & Downie . . . 171 Adnatosphaeridium chonetum (Cookson & Eisenack) . . 171 Genus Hystrichosphaera O. Wetzel . . . . . . 172 Hystrichosphaera ramosa (Ehrenberg) . . . . . 172 var. ramosa (Ehrenberg) . . . . 172 var. gracilis Davey & Williams . . . 172 var. multibrevis Davey & Williams . . 173 var. reticulata Davey & Williams . . 173 cingulata (O. Wetzel) . . . . 173 var. reticulata Davey & Williams . . 174 crassimurata Davey & Williams . . . 174 crassipellis Deflandre & Cookson . . 174 Genus Achomosphaera Evitt. . . . . . . 174 Achomosphaera ramulifera (Deflandre) . . . . 174 sagena Davey & Williams . . . . 174 Genus Hystrichodinium Deflandre . . . . . . 174 Hystrichodinium voigti (Alberti) . . . . . . 175 dasys sp. nov. . . . . . . 175 IV. REFERENCES . . . . . . . . . .176 SYNOPSIS This paper, which will appear in two parts, presents the results of a detailed study of some non-calcareous microplankton from the Upper Cretaceous, and in particular of assemblages of Cenomanian age. The stratigraphical potentialities of the fossil microplankton are briefly assessed by the analysis of samples from five localities in England and one in France. To assess the potentialities of long-range correlation, assemblages from Saskatchewan and Texas have also been examined. Both quantitative and qualitative methods have been employed and the correlations, both intra- and inter-regional, are promising. Seven new genera and thirty-five new species and varieties are described. I. INTRODUCTION FOSSIL non-calcareous microplankton consist mainly of cysts of dinoflagellates, together with various forms of unknown affinity placed in the Group Acritarcha Evitt (1963). The majority of dinoflagellates are free-living, oceanic and plank- tonic. They have a complex life-cycle, usually composed of four stages, in one of which (the motile stage) they are capable of limited vertical movement by the use of two flagella. During the life cycle, if the organism is subject to adverse con- ditions, a resting cyst is formed. This, most palynologists believe, is the only stage io8 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i in the dinoflagellate life-cycle preserved in the fossil state. Fossil non-calcareous microplankton are useful as stratigraphic indices because they are planktonic, of relatively resistant composition, abundant in most marine sedimentary samples, and easy to extract. Hence the principal object of the study was to assess how accurate dinoflagellate cysts are for intra- and inter-regional stratigraphic correlations. The order of description of the dinoflagellate cyst-families follows that in Sarjeant FIGURE ONE MAP of SPECIMEN LOCALITIES in ENGLAND and FRANCE FIG. i. Map of specimen Localities in England and France. CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 109 & Downie (1966). The adjectives used when describing the frequency of a species in an assemblage have been defined as follows: Very Common 10-100% Common 1-10% Infrequent 0-1-1% Rare 0-01-0-1% Very Rare under o o i % The first fossil microplankton were described and figured by Ehrenberg (1838, 1843, 1854) and included forms embedded in flakes of Upper Cretaceous flint from Germany and Denmark. These observations stimulated English microscopists in the mid-nineteenth century into finding these organisms in English flints of the same age. Little further research in this sphere was published until 1933 when O. Wetzel described assemblages from Upper Chalk flints of the Baltic region. In 1934 Deflandre published the first of a number of well illustrated papers dealing with fossil microplankton from the flints of the Paris Basin. Unfortunately the strati- graphic horizons of the flints are unknown, many being picked up in the streets of Paris. His two most important papers describing Upper Cretaceous microplankton were published in 1936 and 1937. They contain accurate figures and descriptions of many new species and genera mainly from the Senonian, although some of the flints are probably of Cenomanian and Turonian age. Between 1936 and 1952 few papers dealing with microplankton from the Upper Cretaceous were published. Firtion (1952) described the first definite Cenomanian assemblage, the material having been obtained from the Lower Cenomanian of France. All his species have subsequently been observed in the Lower Chalk of England and France except for Pareodina sp. which is unidentifiable. Firtion's Hystrichosphaeridium cf . salpingophorum may correspond to H. mantelli, and Micrhy- stridium ambiguum is probably Cleistosphaeridium huguonioti. Since 1955 a number of publications have been produced dealing mainly with the systematics of Cretaceous microplankton. Assemblages have been described from Germany by Gocht (1957, 1959), Eisenack (1958), and Alberti (1959, 1961), but only the latter author records Cenomanian species. Alberti (1961) described a series of assemblages from the Valanginian to Turonian of northern Germany. Eight species were recorded from the Cenomanian and all of these, except for Korojonia dubiosa, have been found during the present study. Hystrichodinium pulchrum, identified by Alberti from the Cenomanian, is here included within Hystrichodinium voigti. Gony- aulax orthoceras, figured by Alberti, is undoubtedly Cribroperidinium intricatum sp. nov., and Palaeohystrichophora cf. paucisetosa is probably a form of P. infusorioides possessing fewer spines than usual. Lower Cretaceous assemblages were described by Neale & Sarjeant (1962), Pocock (1962), and Tasch, McClure & Oftedahl (1964). Gorka (1963) described nine species of microplankton from the Cenomanian of Poland. All of these species have been found in the Cenomanian deposits of England and France, although some are con- sidered to be incorrectly identified. Gonyaulax orthoceras, illustrated by Gorka no CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i appears to be Gonyaulacysta exilicristata sp. nov. and her Gonyaulax sp. is probably Cribroperidinium intricatum ; Hystrichosphaeridium asterigerum is probably equivalent to Oligosphaeridium complex and Hystrichosphaeridium polytrichum possibly to Cleistosphaeridium armatum. Her Hystrichosphaeridium cf. striolatum may also belong to the latter species. Baltes (1963) described eight species of microplankton from the Cenomanian deposits of Roumania. Of these seven have been recorded in the present study. The species not recorded, Ascodinium hialinium, probably belongs to the genus Deflandrea. The identities of three other species are in doubt: some of the specimens illustrated as Hystrichosphaeridium longifurcatum probably belong to this species (transferred to Surculosphaeridium by Davey et al. 1966), but one specimen (pi. 7, fig. 12) probably belongs in Hystrichosphaera. Histrichosphaeridium sp. 22 resembles Exochosphaeridium striolatum var. truncatum nov. and Hystrichosphaeridium sp. 23 is possibly Cleistosphaeridium multifurcatum. Manum & Cookson (1964) describe species of supposed lower Upper Cretaceous age from Arctic Canada and of these, eight have also been recorded from the Lower Chalk of England and France; however, the Arctic Canadian assemblages are more comparable to those obtained from Saskatchewan. Cookson & Hughes (1964) gave the first account of microplankton from the deposits of Upper Albian and basal Cenomanian age in England. Thirty- three species were described from the Cenomanian and of these only six have not been recorded from the basal Cenomanian of Fetcham Mill (sample FM 840) and Compton Bay (CB i). A number of papers have been published (between 1954 and 1965) dealing with the microplankton of Australia and, in part, of New Guinea and Papua Deflandre & Cookson (1954, 1955), Cookson (1956, 1965), Cookson & Eisenack (1958, 19600, b, 1961, 19620, b) Eisenack & Cookson (1960), and Cookson & Manum (1964). Un- fortunately it is usually only possible to give the approximate age of the samples and, therefore, stratigraphic conclusions are not as meaningful as one would have hoped. A number of species from the Cenomanian deposits of England were described by Davey, Downie, Sarjeant & Williams (1966) and the present paper published in two parts forms a natural continuation of that work. Clarke & Verdier (1967) describe microplankton assemblages of Cenomanian to Senonian age from the Isle of Wight, southern England. The Cenomanian samples examined by them were obtained from a locality approximately sixteen miles to the east of Compton Bay. Their results indicated that the Upper Cretaceous could be divided into 5 zones and 5 subzones, and also into 7 " intervals " based on extinction points. The results concerning the Cenomanian are mainly substantiated in the present study. ACKNOWLEDGMENTS The research has been carried out during the tenure of a research studentship in the Department of Geology, The University of Nottingham. CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 646' 1 0) 1 s -8 0) _Q 72 7'* i ^ subglobosus Zone ' i 151-CB21 - 137- 19 -116- 17 - 107- 15 -92- 13 76'- 11 Ligh Ch / Gre) C t grey alk t / marly :halk l i l l ES l I I I l I l I i i 1 l I l 1 1 i i l l 1 I l i 1 1 1 i i \ varians Zone - 59 - 9 h44- 7 - 28'- 5 15'- 3 l l 1 I l I l i l 1 l l 1 1 i i i i i i ^_ Phosphatic nodules i i I i i i Blue-grey Chalk .11' A" i i i i i l Chloritic Marl i i l l V ' Albion 23' 21' 6" FIG. 3. Section showing the Location of the Samples analysed from Compton Bay, Isle of Wight. Scale i inch to 20 feet. 114 CENOMANIAN NON-CALCAREOUS MICROPLAN KTON, i 122- Zone subg ** 78'. varians Zone Turanian T , I II ^1 II TTI "115-Sp7 100- 6 78- 5 52'- 4 28-3 12'- 2 0' - 1 plenus Marls Grey -white Chalk - 80 Hard grey Chalk -. 78 ' Grey Chalk 8' Marly Chalk 28' Pink Chalk U Green -white Chalk V Aibian FIG. 4. Section showing the Location of the Samples analysed from Speeton, Yorkshire. Scale i inch to 20 feet. CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i 115 4. Hunstanton, Norfolk. (TF 675420) The varians and subglobosus zones are exposed in the Hunstanton cliffs (Peake & Hancock 1961). Three samples were processed, two from the lower zone and one from the base of the Totternhoe Stone (Fig. 5). 5. Devon, South Coast The Cenomanian is represented between Salcombe and Lyme Regis by isolated patches composed of a few feet of sandy Cenomanian Limestone. These deposits were divided into four beds by Jukes-Browne (1903) Ar, A2, B and C. Bed C is probably the Actinocamax plenus Marls. The samples were collected from four localities by Dr. W. E. Smith (Fig. 6) : (i) Maynards Cliff (see Smith 1961 : 114) (ii) Beer Head (see Smith 1957 : 123) (iii) Whitediff (see Smith 1957 : 118) (iv) Humble Point (see Smith 1965 : 126) 6. Escalles Borehole, Cap Blanc-Nez, Pas de Calais The borehole, drilled by the Bureau de Recherches Geologiques et Minieres in 1958 has been described by Destombes (1961). Eleven samples were obtained for analysis at about 20 ft. (6m.) intervals (Fig. 7). 7. Saskatchewan, south-east The borehole, from which the Saskatchewan samples were obtained, was drilled for the Department of Mineral Resources, Saskatchewan and is called " International Yarbo, no. 17 ". It is located east of Regina at Lsd. i, Sec. 24, Twp. 20, Rg. 33, Wrst Meridian. All depths are measured from the Kelly Bushing which is at an elevation of 1,690 ft. above sea level. Six samples of Albian/Cenomanian age were analysed for their microplankton content (Fig. 8). 8. Texas, north Two samples were obtained from the Upper Cenomanian of north Texas (Tarrant County) for the author by Dr. J. D. Powell. The lower sample (T5) was obtained from the Upper Woodbine Formation (Acanthoceras wintoni Zone) and consists of a yellowish, slightly calcareous clay. The higher sample (T4) is from 35 ft. above the base of the Eagle Ford Formation (Eucalycoceras Zone). This is a thin-bedded yellow limestone containing shelly fossils and plant debris, in particular leaf fragments. n6 CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i r20' JS C o N -13' 10" - H2 18' 6" - H3 HI -18' Totternhoe Stone Creamy -white Chalk 4 1 1 5 Inoceramus Bed Paradoxica Bed Red Rock FIG. 5. Section showing the Location of the Samples analysed from Hunstanton, Norfolk. Scale i inch to 2 feet. CENOMANIAN NON-CALCAREOUS MICROPLAN KTON, i 117 n8 CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i 502 N /> 2 o i 60f- 0) c O M 716 Turonian ' - E153 -521' 6" - 159 -541 - 165 -561' - 171 -5806- 177 -600' - 183 620 - 189 639V- 195 659'6"- 201 679' - 207 698' 6'- 213 Grey- white Chalk Grey marly Chalk Chloritic Marl 712' 6* Albian FIG. 7. Section showing the Location of the Samples analysed from Escalles, France. Scale i inch to 30 feet. CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 119 Q_ ID o O o o O u 1 tt Shale / Sas UJ Q_ Q. ID Second White Speckled Shale 007' -835 V -QO/ -890' Shale c960 ex. \* f \j \J -967' UJ Fish Scale Zone o -c1030' -1023 Shale ~l-IVJjw -1084' Viking FIG. 8. Section showing the Location of the Samples analysed from Saskatchewan, Canada. Scale i inch to 50 feet. GEOL. 17, 3 120 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i III. SYSTEMATIC DESCRIPTIONS Class DINOPHYCEAE Pasher Subclass DINIFEROPHYCIDAE Bergh Cyst-FamUy GONYAULACYSTACEAE Sarjeant & Downie 1966 Genus GONYAULACYSTA Deflandre emend. Sarjeant 1966 Gonyaulacysta cassidata (Eisenack & Cookson) emend. Sarjeant 1960 Gonyaulax helicoidea subsp. cassidata Eisenack & Cookson : 3, pi. i, figs. 5, 6. ig66a Gonyaulacysta cassidata (Eisenack & Cookson) Sarjeant : 125, pi. 14, figs. 3, 4, text-fig. 31 (see also for earlier references). 1967 Gonyaulacysta cassidata (Eisenack & Cookson) Clark & Verdier: 29, pi. 4, figs. 4-6. DIMENSIONS. Range of observed specimens: overall length 59 (67-5) 78/4, overall width 40 (46-4) 60 fi. Number of specimens measured, 14. REMARKS. The Cenomanian specimens examined are very similar to the Austra- lian Aptian-Cenomanian forms of Eisenack & Cookson (1960) and Cookson & Eisenack (19626) except that the former are slightly smaller in size. OCCURRENCE. G. cassidata is an infrequent species at all horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It has not been recorded in the North American samples. Gonyaulacysta whitei Sarjeant ig66a Gonyaulacysta whitei Sarjeant : 126, pi. 14, fig. 2, text-fig. 32. DIMENSIONS. Range of observed specimens: overall length 55-62/4, overall width 42-50/4. Number of specimens measured, 5. REMARKS. The shape of the shell, the tabulation and the crests make G. whitei an easily recognizable and distinctive species. G. cf. ambigua Cookson & Eisenack (19606) from the Upper Jurassic is of the same shape and possesses similar crests and tabulation; plate i"" is absent. However, the apical horn of G. cf. ambigua is rudimentary or absent, thus making differentiation easy. OCCURRENCE. Of five specimens of G. whitei observed, four are from sample FM 770 and one from sample FM 750. The restricted distribution of this species and its similarity to G. cf . ambigua from the Upper Jurassic suggest the possibility that this is a derived form. Gonyaulacysta fetchamensis Sarjeant 19660 Gonyaulacysta fetchamensis Sarjeant : 128, pi. 15, figs, i, 2, text-fig. 33. REMARKS. G. fetchamensis has a rather unusual tabulation two posterior inter- calary plates and a seventh postcingular plate and, as pointed out by Sarjeant (19660), may subsequently form the basis of a new genus. As yet only two specimens have been studied and this species has, therefore, been placed in Gonyaulacysta. It CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i 121 appears to be a transitional type from the normal Gonyaulacysta to forms now placed in Cribroperidinium Neale & Sarjeant. The latter forms are of similar overall appear- ance but the shell is divided by low crests into an unusually large number of areas. OCCURRENCE. G. fetchamensis is a rare species recorded only from the Chalk of H.M. Geological Survey borehole, Fetcham Mill, Surrey, at 840 ft. depth. Upper Cretaceous (Lower Cenomanian). Gonyaulacysta exilicristata sp. nov. (PI. i, figs, i, 2; Figs. gA, B) DERIVATION OF NAME. Latin, exilis, thin or poor; cristatus, crested with re- ference to the poorly defined sutural crests. DIAGNOSIS. Shell subspherical ; moderately well developed apical horn. Shell wall thick, finely but densely granular. Reflected tabulation 3', la, 6", 6c, 6'" (7"'?), i p, i"". Plate boundaries marked by low, poorly defined crests which sometimes form small spines at crestal nodes. Cingulum narrow, weakly laevo- rotatory; sulcus of moderate width, widening slightly posteriorly. HOLOTYPE. G.S.M. slide PF 3987 (i). Lower Chalk, H.M. Geological Survey Borehole, Fetcham Mill, Surrey at 730 ft. depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Holotype: overall length 83/4, overall width 68 ju,, length of horn I2/*. Range: overall length 70 (81-6) 98^, overall width 58 (64-5) 71/4. Number of specimens measured, 24. DESCRIPTION. The shell wall measures between 2 and 3/4 in thickness, the endophragm being approximately twice as thick as the periphragm. The latter forms the apical horn which is triangular in cross-section due to the sutural ridges which delimit the apical plates extending along it. Lines of ornamentation, some- times similar to the sutural crests, are present on some of the plates, particularly those in the postcingular series. Occasionally it appears that plate 4'" is subdivided by a low ridge so giving seven postcingular plates. The cingulum is fairly narrow (3 to 47*) and only slightly laevo-rotatory. The sulcus often possesses a posterior ventral plate. In the medial-posterior part of the furrow there is usually an elongate depression, as seen in the holotype (Fig. gA). A precingular archaeopyle is typically present. REMARKS. G. exilicristata sp. nov. is distinguished from all previously described species by its overall shape, the type of plate boundaries and the tabulation. Apteodinium granulatum Eisenack (1958) is similar but has a stouter apical horn and a tabulation appears to be absent; the cingulum is only rarely visible. Cribroperidinium orthoceras (Eisenack) comb. nov. is also similar but possesses a longer apical horn, and the tabulation differs and is more clearly defined. OCCURRENCE G. exilicristata is a rare species recorded from a number of horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. At only one horizon is this species common, in sample FM 730 from Fetcham Mill. Two specimens have been recorded from Saskatchewan, both from sample Sas 1084. 122 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i FIG. 9. Gonyaulacysta exilicristata sp. nov., A. Ventral Surface of Holotype (X 900), B. Dorsal Surface of Holotype, (X 900). Gonyaulacysta Sp. A., C. Ventral Surface slide PF. 3987 (2) (X 900). D. Dorsal Surface slide PF. 3987 (2) (X 900) CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 123 Gonyaulacysta delicata sp. nov. (PI. i, figs. 7, 8; Figs. loA, B) DERIVATION OF NAME. Latin, delicatus, delicate with reference to the delicate nature of the shell. DIAGNOSIS. Shell subspherical, epitract and hypotract of similar size. Shell wall thin, smooth. Apical horn absent, there being a small circular apical plate in this position surrounded by three large apical plates. Reflected tabulation 4', la, 6", 6c, 6"', ip, i"". Plate boundaries well defined by low crests. Cingulum wide, strongly laevo-rotatory ; sulcus broad. HOLOTYPE. B.M. (N.H.) V. 51979(1). Lower Colorado, Second White Speckled Shale, International Yarbo Borehole No. 17, Saskatchewan at 835 ft. depth. Upper Cretaceous (Cenomanian). PARATYPE. B.M. (N.H.) V. 51979(2). DIMENSIONS. Holotype: length of shell 57/4, width 51/4. Paratype: length of shell 55/4, width 47/4. Range: length of shell 52-60/4, width 47-51/4. Number of specimens measured, 4. DESCRIPTION. The shell wall is very thin (less than 0-5 ju, thick) and only attains a thickness of 0-5 /* when forming the plate boundaries. Due to the thinness of the shell wall specimens are easily distorted. The tabulation has, however, been formulated after the examination of a number of specimens. A B FIG. 10. Gonyaulacysta delicata sp. nov., A. Ventral surface of Holotype (x 15)- B. Dorsal surface of Holotype (X 1500). 124 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i Anteriorly three apical plates (2', 3' and 4') abutt against a small circular plate, plate i', which is in the position normally occupied by the apical horn. The pre- cingular and postcingular series of plates are large and more or less pentagonal. Plate 3" is always lost in archaeopyle formation. The first postcingular plate is elongate and has a poorly defined sulcal border. The cingulum is wide (4 to 7 fi) and tends to be constricted at the cingular plate boundaries. The sulcus is broad and widens slightly towards the posterior. The anterior end of the sulcus has a flat margin from which arise plates i' and la. At the posterior end of the sulcus there may be developed a posterior ventral plate. REMARKS. The presence of very low sutural crests, the lack of an apical horn and the tabulation differentiate G. delicata sp. nov. from all previously described species. G. ambigua Deflandre, from the Kimeridgian of France, is of similar appearance but possesses a small apical horn and differs in tabulation detail. OCCURRENCE. G. delicata has been found in only one sample, Sas 835, and it is there infrequent. This restricted distribution may well indicate that this is a derived species. Gonyaulacysta sp. A. (PL i, figs. 9, 10; Figs. gC, D) DESCRIPTION. Only one well preserved specimen of this species has so far been observed. It possesses a subspherical shell, bearing a moderately well developed horn with a trifid termination. The shell wall (c. 2/1 thick) is irregularly studded with granules of varying shapes and sizes. The sutural crests are quite well defined, but low, and indicate a reflected tabulation of 3', la, 6", 6c, 6'" (7'"?), ip., i"". The crests, demarcating the three apical plates, extend along the apical horn and give rise to three small spines at its distal termination. Plates 2", 3" and 4" are relatively large, plates i" and 5" rather elongate and plate 6" is reduced due to the anterior intercalary plate. In the postcingular series, plates i"' and 2'" are reduced and plate 7'" does not have a clearly marked plate boundary. Plate 2'" possesses a curved line of ornamentation which is characteristic of Cnbroperidinium Neale & Sarjeant. There is a single posterior intercalary plate and a large antapical plate. The cingulum is strongly laevo-rotatory and varies considerably in width (2-5 (JL), being constricted at the cingular plate boundaries and expanding on either side. The sulcus is broad and possesses a central depressed area of elongate shape. A archaeopyle is present. FIGURED SPECIMEN. G.S.M. slide PF. 3987, specimen 2. Lower Chalk, H.M. Geological Survey Borehole, Fetcham Mill, Surrey at 730 feet depth, Upper Cretaceous (Cenomanian) . DIMENSIONS. Overall length 75/11, overall width 65^, length of horn 14 p.. REMARKS. The form of the apical horn, the type of granulation and the shape of the cingular plates distinguish this species from all described forms. The shape CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 125 of the shell and the tabulation are most comparable to G. exilicristata suggesting a relationship between the two species. Genus CRIBROPERIDINIUM Neale & Sarjeant emend. EMENDED DIAGNOSIS. Proximate cysts, subspherical to ovoidal, thick-walled. Crests numerous and well developed. Tabulation ?6', (i-5a), 8-9", oc, g"', ip, i-3p.v. (5-7 p.c.), o"" ( ?2""). Cingulum laevo-rotatory. Archaeopyle precingular. TYPE SPECIES. Cribroperidinium sepimentum Neale & Sarjeant 1962. REMARKS. The diagnosis has been emended to draw attention to the fact that the anterior intercalary series and the posterior circle series of plates are not always readily distinguishable. The tabulation attributed to this genus by Neale & Sarjeant was 6', i~5a, 8", 9'", ip, i-2p.v. 6-Pyp.c., o"". In determining the tabula- tion they numbered all the delimited areas on the shell surface in the normal manner. This numbering procedure has been used in the study of C. intricatum sp. nov. with certain reservations. Difficulties arose due to the large number of delimited areas and also to some variation on the dorsal surface of the hypotract. Some of the crests are rudimentary and others, although appearing fairly normal, are unusual in their position. It was discovered that by the removal of these unusual crests a normal Gonyaulax-type tabulation could be reconstructed. This is clearly shown in Figs. nA, B, of the ventral surface of the holotype of C. intricatum. Thus Cribro- peridinium is basically a form of Gonyaulacysta which possesses additional crests. These additional crests may correspond to an increase in the number of thecal plates composing the motile dinoflagellate, but this is thought to be unlikely since: (a) the plates formed would be of an extremely unusual shape ; (b) these crests subdivide detached opercula demonstrably composed of a single precingular plate (3") ; (c) many of these crests are poorly developed and show considerable positional variation on the dorsal surface of the hypotract; and (d) that by their removal a normal Gonyaulax-type tabulation remains. Thus, these additional crests may be regarded as an ornamentation or perhaps a rather, superficial cyst strengthening device. It is, therefore, considered more practical and correct to use a different numbering system for the Cribroperidinium cyst tabulation, vis, roman numerals. Three species, Gonyaulacysta orthoceras (Eisenack), G. muderongensis (Cookson & Eisenack) and G. edwardsi (Cookson & Eisenack), are here transferred to Cribroperi- dinium. This genus appears to be very limited in stratigraphic range (Hauterivian Lower Turanian) and, as such, is a useful stratigraphic indicator. Cribroperidinium intricatum sp. nov. (PL 2, figs. 1-3; Figs, n, 12) DERIVATION OF NAME. Latin, intricatus, complicated with reference to the complex crest arrangement. 126 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i DIAGNOSIS. Shell subspherical, epitract and hypotract of similar size. Apical horn of moderate length, subconical. Shell wall moderately thick, granular and bears a few, randomly arranged tubercles. Crests usually in form of low ridges, well denned, sometimes membranous, delimiting a large number of plates on shell surface. Sutural spines absent. Plate IF" crossed diagonally by low crest. Operculum possessing semi-circular crest. Sulcus possessing posterior ventral plates. Cingulum narrow, plates not defined. HOLOTYPE. B.M. (N.H.). V. 51980 (i). Upper Lower Colorado, Fish Scale Zone, International Yarbo Borehole No. 17, Saskatchewan at 1,023 feet depth. Lower Cretaceous (Albian). DIMENSIONS. Holotype: overall length 120 /*, overall width 114/1,, length of horn 20 p. Range: overall length 107 (125-1) 142/11, overall width 101 (108-2) 126/1.. Number of specimens measured, 17. DESCRIPTION. The shell wall is i to 1-5 p in thickness and densely granular. The crests are typically low thickenings of the periphragm (2-3 p wide), but in the ant- apical region, and occasionally elsewhere, the crests take the form of high flanges The latter (up to 6 /z. in height) are membranous, thin and always perforate. The number of apical plates always appears to be six. Plate I', equivalent to the first apical plate, is elongate and abuts against the anterior end of the sulcus. The crest arrangement on the ventral surface appears to be practically constant and is characteristic of this species. The crests limiting the plates F, I", II", III", IV", VII" and VIII" are always constant in position. Plates F" and IF" are reduced due to the presence of a posterior intercalary plate. Plate IF" always possesses a crest passing diagonally across it and the crest between plates IIF" and IV" is of a A B FIG. ii. Cribroperidinium intricatum sp. nov. A. Ventral surface of Holotype, (X 700). B. Reconstruction to show Gonyaulax-type tabulation (x 700). CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 127 characteristic right angle shape. The positions of the crests on the dorsal surface of the epitract are fairly constant, only varying in minor details. The large plate V" is usually detached in archaeopyle formation and possesses a semi-circular crest from which radiate a small number of other crests (PI. 2, fig. 3). These pass over the boundary of plate V" to continue on adjacent parts of the epitract. The posi- tions of the crests on the dorsal surface of the hypotract appear to be less constant in position. Their predominent direction is parallel to the longitudinal axis of the shell; sometimes a series of posterior circle plates may be present. Antapical plates, if present at all, are very reduced and obscured by the crestal membranes. The cingulum is narrow (c. 6/z in width), slightly laevo-rotatory, and tends to possess a thicker wall than is usual for the remainder of the shell. The sulcus only projects onto the epitract for a short distance, being considerably larger and wider on the hypotract. It is always divided into a number of posterior ventral plates by reduced crests. REMARKS. Figs nB, I2B have been prepared from the holotype and one other specimen, omitting the additional crests, to show the basic Gonyaulax-type tabulation. C. intricatum may be differentiated from C. orthoceras (Eisenack), C. muderongensis (Cookson & Eisenack), Gonyaulacysta apionis and G. diaphanis by its more spherical form, details of crest arrangement and the absence of spines. C. edwardsi (Cookson & Eisenack) is most similar, being almost spherical, but possesses a very well developed, stiff apical horn and the crests on the ventral surface are arranged differ- ently. In particular the diagonal crest on plate II'" is absent. Gonyaulax sp. (Gorka 1963) from the Cenomanian of Poland is very similar and may be conspecific with C. intricatum. A B FIG. 12. Cribroperidinium intricatum sp. nov., A. Dorsal surface of Holotype with archaeopyle (X 700). B. Reconstruction to show Gonyaulax-type tabulation (X 700). 128 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i OCCURRENCE. C. intricatum is very rare in the British Cenomanian and has been observed only in samples FM 840, CB I and CB 7. At Escalles it is infrequent in sample E 201, very rare in sample E 195 and lacking elsewhere. In the Saskatche- wan sample Sas 1023 it is common, in sample Sas 1084 rare and is lacking in all the other North American samples. It is interesting to record that this species was observed in the Albian sample FM 886 from Fetcham Mill. Thus C. intricatum has a range from Albian to Lower Cenomanian in the examined material. OTHER SPECIES The following species are here attributed to the genus Cribroperidinium on the basis of overal shape, and number and position of periphragm crests: Cribroperidinium orthoceras (Eisenack 1958) comb, nov., 1958 Gonyaulax orthoceras Eisenack, Neues. Jb. Geol. Palaont., Abh., 106 (3) 388: pis. 21, figs. 3-14; 24, fig. i; text-figs. 2, 3. Cribroperidinium edwardsi (Cookson & Eisenack 1958) comb, nov., 1958 Gonyaulax edwardsi Cookson & Eisenack: Proc. R. Soc. Viet., 70 (i), 32-33; PL III, figs. 5, 6, text-fig. 7. Cribroperidinium muderongensis (Cookson & Eisenack 1958) comb, nov., 1958 Gonyaulax muderongensis Cookson & Eisenack: Proc. R. Soc. Viet., 70 (i), 32; PI. Ill, figs. 3, 4, text-fig. 15. Gonyaulacysta apionis (Cookson & Eisenack 1958) and G. diaphanis (Cookson & Eisenack 1958), both from the Lower Cretaceous of Australia, are of similar appear- ance to forms included in Cribroperidinium and may at a later date be transferred to this genus. Eisenack (1958 text-figs. 2, 3) figured the ventral surface tabulation of C. ortho- ceras omitting, or dotting in, some of the crests which are seen to be present on the photographs of the same specimens. Thus a true representation of the crestal positions was not given, somewhat misleading later workers. These figures have been redrawn (Figs. I3A, B) from the photographs and show the remarkable similarity between the ventral surface of C. orthoceras and that of C. intricatum. For this reason the former species has been transferred to Cribroperidinium and the diagnosis emended. Eisenack does in fact compare and contrast his species with the reattributed Australian forms above, also with Gonyaulactysa wetzeli (Lejeune Carpentier 1939) and G. obscura (Lejeune-Carpentier 1946), all of which he con- siders to be in the same group. The latter two species, however, have a typical Gonyaulacysta tabulation and must remain in that genus. Cribroperidinium orthoceras (Eisenack) emend. (Figs. I 3 A, B) 1958 Gonyaulax orthoceras Eisenack : 388, pi. 21, figs. 3-14, pi. 24, fig. i ; text-figs. 2, 3. !959 Gonyaulax orthoceras Eisenack; Gocht : 54, pi. 5, figs. 12, 13. 1961 Gonyaulax orthoceras Eisenack; Alberti : 6, pi. n, figs. 1-3. 1963 Gonyaulax orthoceras Eisenack; G6rka : 30, pi. 3, figs. 1-4. 1965 Gonyaulax orthoceras Eisenack; Baltes : 12, pi. 3, figs. 95-99. EMENDED DIAGNOSIS. Shell ovoidal, moderately thick-walled, bearing strong, thorn-like apical horn constituting approximately one quarter of shell length. Shell CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 129 wall granular bearing tubercles. Distinct tabulation marked by strong, low crests delimiting large number of plates. Plate IF" crossed diagonally by low crest and operculum possessing semi-circular crest. Cingulum narrow, devoid of plate boundaries. HOLOTYPE. The specimen illustrated by Eisenack (1958, pi. 21, fig. 5) from Pre- paration Ob. Apt. No. 32. Aptian glauconitic limestone, Deutschen Erdol A. G., Erdolwerke Holstein boring Marne, Feld Heide, North Germany, at 7617 metres depth. REMARKS. The emended diagnosis excludes those forms described by Sarjeant (1966) from the Speeton Clay which are to be described elsewhere under a new specific name. Genus CARPODINIUM Cookson & Eisenack, 1962 Carpodinium obliquicostatum Cookson & Hughes (PL i, figs. 3, 4) 1964 Carpodinium obliquicostatum Cookson & Hughes : 48, pi. 6, figs. 1-6. 1967 Carpodinium obliquicostatum Cookson & Hughes; Clarke & Verdier: 23, pi. 2, figs. 4, 5. DESCRIPTION. The shell is elongate-ovoidal and bears relatively high sutural crests. The latter are smooth or slightly granular and the distal margin may be entire or spinous. The short apical horn is a prolongation of one of the larger crests and is typically asymmetrically placed. The periphragm of the shell possesses an unusual ornamentation small areas of triangular to polygonal shape, composed of thick periphragm, are separated by narrow anastomosing " canals " where the periphragm is unthickened or perhaps absent (PL r, fig. 4). The cingulum is marked by indentations in some of the crests and is not apparent on the shell surface. The sulcus, delimited by crests, is occasionally observable and extends the length of the shell, being narrow near the apex and widening pos- teriorly. The precingular and postcingular plates, probably six in each series, are very elongate and difficult to discern because of the obscuring nature of the high crests. A precingular archaeopyle is commonly present. Four apical plates and a single antapical plate appear to be present. Intercalary plates were not observed. DIMENSIONS. Range of observed specimens: overall length 56(69-5) 82 //,, overall width 33(43-8) 57 /*, height of crests 6-15 ^u,. Number of specimens measured, 16. REMARKS. The Cenomanian specimens studied resemble the type material from the Upper Albian and Lower Cenomanian of Cambridgeshire in all respects. OCCURRENCE. C. obliquicostatum is a rare to very rare species found at all horizons throughout the Cenomanian of Fetcham Mill and Compton Bay and in three samples from Escalles (E 195, E 189 and E 159). Genus ELLIPSODINIUM Clarke & Verdier 1967 REMARKS. A number of microplankton genera have been described as possessing a reticulate shell wall, occasionally with an outer membrane but only rarely with any 130 CENOMANIAN NON-CALCAREOUS MI CROPL AN KTON, i signs of tabulation. In all the described forms possessing a cingulum the archae- opyle is apical. Hence the combination of numerous crests, a cingulum and a precingular archaeopyle differentiates Ellipsodinium from all previously described genera. Ellipsodinium rugulosum Clarke & Verdier (PI. 3. fig. i ; Figs. 140, D) 1967 Ellipsodinium rugulosum Clarke & Verdier: 69, pi. 14, figs. 4-6, text-fig. 29. DIMENSIONS. Range of observed specimens : shell length 30 (37 . 7) 46 ju., shell width 25 (33-6) 40 p, maximum height of crests 1-5 (2-4) 3-5 ju,. Number of specimens measured, 20. DESCRIPTION. The crests are thin, occasionally perforate lamellar structures which thicken slightly before joining the shell surface. The cingulum may be delimited by a pair of crests, or in the absence of cingular crests, crests may terminate abruptly at its borders. Rarely crests traverse the cingulum. The sulcus is not obvious because of the nature of the elongate areas outlined by the crests. Apical and antapical processes or horns are absent. The precingular archaeopyle is sub- triangular in outline. OCCURRENCE. E. rugulosum is a rare to fairly common species at all horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It was not observed in the North American material. Cyst-Family PAREODINIACEAE Gocht emend. Sarjeant & Downie 1966 Genus APTEODINIUM Eisenack 1958 REMARKS. Members of this genus have occasionally been observed in the Euro- pean Cenomanian and have been placed in A . granulatum. Rarely plate boundaries may be discernable. This fact together with the overall shape and the well developed precingular archaeopyle indicate that this genus is closely related to Gonyaulacysta and at a future date may be transferred to the Cyst-Family Gonyaulacystaceae. Apteodinium granulatum Eisenack (PI. 3, figs. 5, 6) ?I935 Palaeoperidinium ventriosum O. Wetzel; Deflandre : 228, pi. 5, fig. 5; pi. 6, figs. 9, 10. ?i936& Palaeoperidinium ventriosum O. Wetzel; Deflandre : 27, pi. 5, figs. 1-4. ?i936a Palaeoperidinium ventriosum O. Wetzel; Deflandre: fig. 100. 1958 Apteodinium granulatum Eisenack : 386, pi. 23, figs. 8-14, text-fig, i. 1958 Apteodinium granulatum Eisenack; Gocht : 64, pi. 5, fig. 2. 1961 Apteodinium granulatum Eisenack; Alberti : 24, pi. 4, figs. 4-6. 1963 Apteodinium granulatum Eisenack; Baltes. : 584, pi. 4, fig. u. DESCRIPTION. Shell subspherical with a moderately thick wall (1-2-5 //,) possessing a stout conical horn. The wall is densely granular, the granules often being elongated into fine, short processes which are linked laterally thus covering CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 131 the shell surface with an intricate network of a matted furry appearance. The cingular boundaries are always visible as slight thickenings of the periphragm. Very rarely similar but more reduced thickenings indicate the presence of plate boundaries. A precingular archaeopyle is commonly developed. DIMENSIONS. Range of observed specimens: overall length 42 (48-2) 53 /z, width 31 (38-6) 44 p., length of apical horn 5 (6-1) 8 /*. Number of specimens measured, 9. REMARKS. The Cenomanian specimens are very similar in appearance to the type material described by Eisenack from the Aptian of Germany. The matted, furry appearance was not described but when this feature is only slightly developed the shell wall merely appears to be very granular. The Aptian and Cenomanian examples of A . granulatum appear to be comparable to Palaeoperidinium ventriosum O. Wetzel as illustrated by Deflandre (1935, 1936^, b). The holotype of this species has been re-examined by Lejeune-Carpentier (1946) and a distinct tabulation des- cribed. Deflandres' forms which do not possess a tabulation are hence here ten- tatively reattributed to A . granulatum. OCCURRENCE. Only two specimens have been recorded from the English Ceno- manian, both from sample FM 690. At Escalles two specimens have been recorded from both samples E 183 and E 153, and three from sample E 165. A. granulatum is fairly common in sample FM 886 (Albian) from Fetcham Mill but is absent in the North American material and from sample FM 520 (Turanian). Genus TRICHODINIUM Eisenack & Cookson emend. Clarke & Verdier REMARKS. This genus differs from Exochosphaeridium Davey, Downie, Sarjeant & Williams (1966) by the presence of a well developed cingulum and by the shorter spines. Trichodinium castaneum (Deflandre) PI. n, figs. 1-3 J 935 Palaeoperidinium castanea Deflandre : 49, pi. 6, fig. 8. 19366 Palaeoperidinium castanea Deflandre; Deflandre : 25, pi. 16, figs. 1-4. 19360 Palaeoperidinium castanea Deflandre; Deflandre: fig. 99. 19526 Palaeoperidinium castanea Deflandre; Deflandre: fig. 96. 19626 Palaeoperidinium castanea Deflandre; Cookson & Eisenack : 489, pi. 3, figs. 9-11. 1964 Palaeoperidinium castanea Deflandre; Cookson & Hughes : 49, pi. 5, fig. 14. 1967 Trichodium castanea (Deflandre) Clarke & Verdier: 19, pi. i, figs, i, 2. DESCRIPTION. The shell is subspherical with occasionally a small apical horn or a tuft of apical spines. The shell wall is slightly punctate and bears numerous small spines. These are solid, often bifurcate either proximally or distally, and typically terminate in a small bifurcation. Rarely they may be acuminate. The cingulum (3-5 fj, in width) is marked by two parallel lines of thickening along which spines are concentrated. Lines of similar thickening sometimes occur perpendicularly to the cingulum and are probably sutural. A sulcus has not been observed. A precingular archaeopyle is commonly present. 132 CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i DIMENSIONS. Range of observed specimens: shell diameter 35 (49-0) 64 /n, length of spines I (3-3) 5 /x. Number of specimens measured, 16. REMARKS. The Cenomanian specimens resemble the type material from the Upper Cretaceous of France in all respects. The presence of a precingular archaeo- pyle, an apical prominence or apical spines, and a well developed cingulum indicate that this species should be placed in Trichodinium. T. intermedium Eisenack & Cookson, from the Aptian to Lower Albian of Australia, is very similar but is larger (shell diameter 69-90 /A). OCCURRENCE. T. castaneum is a rare species occurring at most horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It has not been recorded from the North American samples. This species has a recorded strati- graphic range from the Aptian to the Senonian (Deflandre). Cyst-Family MIGRODINIACEAE Eisenack emend. Sarjeant & Downie 1966 Genus MICRODINIUM Cookson & Eisenack emend. Sarjeant 1966 REMARKS. Microdinium differs from Gonyaulacysta primarily in possessing a single apical plate which is detached in archaeopyle formation. Gorka (1965), in des- cribing Upper Jurassic assemblages, erected the genus Tetrasphaera which is diagnosed as having a feebly marked tabulation and short spines on the plate boundaries. Gorka has since stated (personal communication with Dr. W. A. S. Sarjeant) that this genus differs from Microdinium only in these two respects. However, it follows that if the tabulation is not clear it is difficult to compare this new genus with Microdinium or any other genus possessing a tabulate cyst, since the diagnosis of these genera is primarily based on the observed tabulation. Also, spines have been observed on the plate boundaries of Microdinium, even in the type species. Thus it would appear that Tetrasphaera at the moment is not precisely defined. Microdinium cf. ornatum Cookson & Eisenack. (PI. 4, fig. 5; Figs. I3C, F) ig66fl Microdinium cf. ornatum Cookson & Eisenack; Sarjeant : 149, pi. 16, figs. 3-6, text-fig. 38. 1967 Microdinium ornatum Cookson & Eisenack; Clarke & Verdier: 66, pi. 5, figs. 11-14. DESCRIPTION. The shell is subspherical to ovoidal possessing a smooth body wall ornamented by a few large tubercles. The latter may be flat or distinctly concave distally. In two specimens small tubercles were seen to delimit a plate in the posterior portion of the ventral area. The plates are bordered by short, broad projections (Fig. I3F), which may be isolated or united distally. DIMENSIONS. Range of observed specimens: shell length 31-34 //,, width 23-31 /z, height of crests 1-2 /*. Number of specimens measured, 6. REMARKS. M . cf. ornatum differs from M. ornatum Cookson & Eisenack (19600) in that (i) the small cingular plate, ventral to plate 6c, is absent or has a very reduced ventral suture, and (ii) the plates are not bordered by ledges, which are sometimes perforate, but by isolated tubercles and spines. CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i 133 OCCURRENCE. M. cf. ornatum is a rare species in the material examined being found only in the following samples: FM 810, FM 670, FM 650, FM 520 (Turonian), E 189 and E 153. In Australia M. ornatum has a stratigraphic range from the Albian to the Lower Turonian. Microdinium setosum Sarjeant. (PI. 2, fig. 4; Fig. i 3 H) 1966 Microdinium setosum. Sarjeant : 151, pi. 16, figs. 9, 10; text-fig. 39. 1967 Microdinium echinatum Clarke & Verdier: 64, pi. i, figs. 9, 10, text-fig. 26. DESCRIPTION. The shell is spherical to ovoidal and possesses a well developed tabulation. The shell surface is either lightly or coarsely granular. The two plates observed by Sarjeant in the medial region of the sulcus are not always delimited. The sutural crests are often relatively high and give rise to numerous, well developed thorn-like spines (Fig. I3H). DIMENSIONS. Range of type material: shell length 25 (29-3) 37 p., width 21 (26-2) 31 p, maximum height of crests 1-5 (3-8) 7 /x. Number of specimens measured, 22. REMARKS. In general form M. setosum is similar to M. ornatum but differs in the presence of spiny crests, a dense granulation and, slightly, in the tabulation exhibited. OCCURRENCE. M. setosum is an infrequent to fairly common species at most horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It has also been recorded from the Albian of Fetcham Mill. Microdinium distinctum sp. nov. (PI. 2, figs. 9-11 ; Figs. 130, E. I) 1967 Microdinium ornatum Cookson & Eisenack; Clarke & Verdier: pi. 5, figs. 11-12. DERIVATION OF NAME. Latin, distinctus, different with reference to the distinctive appearance of this species. DIAGNOSIS. Shell subspherical, thick-walled, smooth. Sutural crests well developed bearing short, broad, flat-topped spines. Reflected tabulation i, oa, 6" 6"', ip, i""; plates i" and 6" very reduced and cingular plates absent. Cingulum broad, weakly laevo-rotatory. HOLOTYPE. G.S.M., slide PF 3989, (i). Lower Chalk, H.M. Geological Survey Borehole, Fetcham Mill, Surrey at 730 ft. depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Holotype: shell length 36^, width 37 p, height of crests c. 2,\i. Range: shell length 29-36^, width 30-37 //., height of crests 2-2-5^. Number of specimens measured, 7. DESCRIPTION. The shell wall is relatively thick, c. i /z, and develops pronounced ridges at the plate boundaries. The sutural spines are closely set, very broad and flat distally. 134 GENOMANIAN NON-CALCAREOUS MICROPLANKTON, i H FIG. 13. Cribroperidinium orthoceras (Eisenack). A. Specimen illustrated by Eisenack (1958, text-fig. 3) redrawn to show full crestal arrangement. B. Holotype (Eisenack 1958, text-fig. 2) redrawn to show full crestal arrangement. Microdinium cf. ornatum Cookson & Eisenack, C. Ventral surface of Holotype (X 700). Microdinium distinctum sp. nov., D. Ventral surface of Holotype (x 700), E. Dorsal surface of Holotype (X 700). Microdinium cf. ornatum Cookson & Eisenack, F. Sutural spines. Microdinium vario- spinum sp. nov., G. Sutural spines. Microdinium setosum Sarjeant, H. Sutural spines. Microdinium distinctum sp. nov., I. Sutural spines. CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 135 The precingular plates are generally smaller than the postcingular plates. Plates i" and 6" are small and appear as slight projections in the ventral area. This is widest posteriorly and is open anteriorly. The apical plate is six-sided and is lost in archaeopyle formation. The archaeopyle possesses slits extending posteriorly between the precingular plates. REMARKS. M . distinctum may be distinguished from all other species of Micro- dinium by the thick, smooth shell wall, the form of the precingular plates and the absence of cingular plates. It is considered that the absence of the latter does not, at present, warrant the erection of a new genus. OCCURRENCE. This species is very rare, being recorded only seven times, from the following samples: FM 790, FM 730, FM 710, FM 690, 153 and 617. It has never been observed in the lower horizons of the Cenomanian. Microdinium variospinum sp. nov. (PI. 2, figs. 5, 6; Fig. i 3 G) DERIVATION OF NAME. Latin, varius, different; spinosus, spine with reference to the variable appearance of the spines. DIAGNOSIS. Shell subspherical to ovoidal; shell wall thin, smooth, granular or lightly reticulate. Sutural crests low, bearing small number of variably shaped spines. These may be simple tubercles to complex bifurcating protrusions. Re- flected tabulation i', oa, 6", (6c), 6'", rp, i""; cingular plate boundaries very faint or absent. Cingulum weakly laevo-rotatory. HOLOTYPE. B.M. (N.H.) V. 51981 (i). Lower Chalk, Bureau de Recherches Geologiques et Minieres Borehole, Escalles, Pas de Calais at 165 metres depth. Upper Cretaceous (Cenomanian) . DIMENSIONS. Holotype: shell length 24^1, width 23/4, length of spines 1-1*5/4 Range: shell length 20 (23-6) 27/4, width 17 (20-0) 23/i, length of spines 1-3 /u,. Number of specimens measured, 9. DESCRIPTION. The cingular plate boundaries are usually absent, however, one Saskatchewan specimen had them lightly defined. The sutural spines are few in number and tend to be concentrated at the posterior of the shell (PI. 2, fig. 5). They are stout and may be either pointed or flattened distally (Fig. 130). REMARKS. The characteristic features of M. variospinum are the nature of the shell wall, the form of the spines and the lack of well defined cingular plate bound- aries. These characters together differentiate this species from all other described species. OCCURRENCE. M. variospinum is a rare species found at all horizons, save the lower three, at Escalles, and in sample CB i, CB 7 and Sas 1084. It has not been recorded from Fetcham Mill. GEOL. 17, 3 IO 136 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i Microdiniutn veligerum (Deflandre) comb. nov. (PI. 3, fig- 4i PL 4, fig- 4) 1937 Micrhystridium veligerum Deflandre : 81, pi. 12, fig. 9. 1943 Ceratocorys veligera (Deflandre) Lejeune-Carpentier : 22, text-figs. 1-6. 19520 Ceratocorys veligera (Deflandre) Deflandre : 120, text-fig. 102. i952b Ceratocorys veligera (Deflandre) Deflandre: text-figs. 304 A-C. 1967 Eisenackia crassitabulata Deflandre & Cookson; Clarke & Verdier: 64, pi. 8, figs. 4-6. DESCRIPTION. The shell is ovoidal, densely granular and bears a number of high crests delimiting a tabulation. The crests are distinctive, being i to 5 /u in height, with typically a smooth outer margin. They consist of two membranes, joined distally and diverging proximally to form a broad base, I to 3 p, wide to the crest. There is between the two membranes a crestal cavity which is occasionally sub- divided by septa, particularly where two crests diverge. Here a conical chamber is usually found. The reflected tabulation appears to be i', ?5", 6c, 6'", ip, i"". The hypotract is considerably larger than the epitract, the latter being devoid of crests and usually possessing a pentagonal apical archaeopyle. The shape of the latter is the only indication that there are five precingular plates. The cingulum is broad and does not appear to be spiral. Plate i'", and to a lesser extent 2'", are reduced to accommodate the posterior intercalary plate. The remaining four postcingular plates are large and there is a single large antapical plate. The sulcus is very narrow just posterior to the cingulum and then widens rapidly towards the antapex. The sulcus extends onto the epitract where it sometimes bears five small sulcal plates. DIMENSIONS. Range of observed specimens: shell length 28 (31-5) 38 /*, width 25 (28-2) 32 //,. Number of specimens measured, 13. REMARKS. Lejeune-Carpentier (1943) placed this species, originally observed in the Upper Cretaceous of France, in the genus Ceratocorys Stein (1883) on the basis of its similarity to motile dinoflagellates contained in this genus. However, it is a cyst possessing an apical archaeopyle and should not be attributed to a motile dinoflagellate genus. Thus this species is here transferred to Microdinium on the basis of the tabulation, apical archaeopyle, reduced size of epitract compared with the hypotract and overall small size. M. veligerum does, however, differ slightly from the other species in this genus by the apparent absence of crests on the epitract and the probable presence of five precingular plates instead of six. Eisenackia crassitabulata as illustrated by Clarke & Verdier (1967) is undoubtedly M. veligerum. The former, as originally described from the Australian Lower Tertiary, is of different overall form and is considerably larger (72-78 by 55-67 /*) . The size of the specimen illustrated by Clarke & Verdier is approximately 30 by 32 / and is thus comparable to the Cenomanian specimens of M. veligerum. E. crassi- tabulata has been recorded from the Lower Tertiary and also from the Maestrichtian of South Africa by the present author. M. irregulare Clarke & Verdier (1967) appears to be very similar to M. veligerum and any definite distinction is not apparent. OCCURRENCE. The Cenomanian forms examined resemble the specimens illus- trated by Lejeune-Carpentier in all respects. M. veligerum is an infrequent to CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 137 common species at all horizons, save two, at Fetcham Mill, Compton Bay and Escalles. These two horizons are basal Cenomanian, FM 840 and CB i, and it was not recorded by Cookson & Hughes (1964) from the Upper Albian/basal Cenomanian of Cambridgeshire. It is also absent in the North American material and has not been described from Australia. This species is present in sample FM 520 of Turonian age. M. veligerum thus appears for the first time just above the base of the Ceno- manian and extends into the Turonian but is, apparently, of restricted geographical distribution. fMicrodinium crinitum sp. nov. (PL 2, figs. 7, 8) 1967 Cometodinium obscurum Deflandre & Courteville; Clarke & Verdier: pi. 10, fig. 3: pi. n, fig. 9- DERIVATION OF NAME. Latin, crinitus, hairy with reference to the numerous hair-like spines. DIAGNOSIS. Shell subspherical, periphragm granular and giving rise to numerous, fine, flexuous spines. Sutural crests low, bearing numerous spines. Cingulum wide, composed of elongate plates. Epitract smaller than hypotract. Archaeopyle not normally visible. HOLOTYPE. G.S.M. slide PF 3990(1). Lower Chalk, H.M. Geological Survey Borehole, Fetcham Mill, Surrey at 690 feet depth. Upper Cretaceous (Cenomanian.) DIMENSIONS. Holotype: shell diameter 27 by 28 p, length of spines c. 12 /JL. Range: shell diameter 24 (30-1) 38/1,, length of spines 6-19 /x. Number of specimens measured, 18. DESCRIPTION. The periphragm granules, which are evenly spaced on the shell surface, are up to 0-5 /j, high, and often form the bases of the hair-like spines. The latter tend to be especially concentrated along the sutures, and because of this tendency and the spherical form of the shell, it has not been possible to fully for- mulate a tabulation. However, precingular, cingular and postcingular plates are quite obvious when the orientation is favourable, the precingular plates being smaller than the postcingular plates. The cingulum is broad, c. 5 /x. The archaeopyle, although it has not been observed, is probably apical. REMARKS. The numerous hair-like spines and the tabulation easily distinguish 1M. crinitum sp. nov. from all previously described forms of dinoflagellate cysts. The overall shape, the small size and the fact that the epitract is smaller than the hypotract all indicate that this species is closely related to the genus Microdinium. However, plate spines have not been recorded in Microdinium, although a granula- tion has, and since the tabulation has not been elucidated in the present species it is only placed tentatively in this genus. OCCURRENCE. 1M. crinitum is infrequent at all horizons throughout the Ceno- manian of Fetcham Mill, Compton Bay and Escalles and is also present in the Albian sample from Fetcham Mill, sample FM 886. It has only once been recorded at Saskatchewan, in sample Sas 835 (Cenomanian). 138 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i Genus HISTIOCYSTA nov. DERIVATION OF NAME. Greek, histos, mesh or network; kystis, sac or cell with reference to the reticulate ornamentation on the shell surface. DIAGNOSIS. Proximate cysts, spherical to subspherical ; shell wall composed of two layers; outer layer giving rise to low crests. Crests reflecting Gonyaulax type tabulation and coarse, subsidiary reticulation within plate boundaries. Sutural crests better denned than crests of subsidiary reticulation. Archaeopyle apical with angular margin. Operculum probably single apical plate. TYPE SPECIES. Histiocysta palla sp. nov. Lower Chalk (Cenomanian) ; England. REMARKS. The reasonably well denned tabulation, the plate ornamentation and the apical archaeopyle easily distinguish Histiocysta from all previously described genera. The most similar genera are Ellipsoidictyum Klement (19606) and Dic- tyopyxidia Eisenack (1961), both from the Upper Jurassic. Both genera possess an apical archaeopyle and a cingulum but a distinct Gonyaulax-type tabulation is absent. However, it seems probable that Histiocysta is genetically related to these two genera. Although the precise tabulation of Histiocysta cannot be elucidated the presence of precingular, cingular and postcingular series of plates together with an apical archaeopyle indicate that this genus belongs to the Family Microdiniaceae. Histiocysta palla sp. nov. (PI. i, figs. 5, 6; Figs. I4A, B) 1939 Micrhystridium sp.? Deflandre & Courteville : pi. 3, fig. 4. DERIVATION OF NAME. Greek, palla, ball with reference to the more or less spherical shape of this species. DIAGNOSIS. Shell spherical to subspherical, thin-walled, periphragm smooth and forming a reticulate network of crests. Network consisting of reflected dinoflagel- late tabulation with central region of each plate occupied by coarse but simple reticulation. HOLOTYPE. G.S.M. slide PF 3052 (2). Lower Chalk, H.M. Geological Survey Borehole, Fetcham Mill, Surrey at 650 feet depth. Upper Cretaceous (Cenomanian) . PARATYPE. G.S.M. slide PF 3991(1). Lower Chalk, H.M. Geological Survey Borehole, Fetcham Mill, Surrey at 710 feet depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Holotype : shell diameter 30 by 33 p, height of crests 4-5 /*. Para- type: shell diameter 26 by 29^, height of crests c. 2-5 /A. Range: shell diameter 25 (31-8) 38 /z, height of crests 1-5 /z. Number of specimens measured, 19. DESCRIPTION. The precingular and postcingular plates may be observed on most specimens but the exact tabulation has not, as yet, been elucidated. The central region of each plate is occupied by a coarse, subpolygonal reticulation which occasion- ally extends to the plate boundaries (Figs. I4A, B). The cingular region is clearly defined by crests and encircles the shell. Cingular plates are not usually discernible, CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 139 the entire region being occupied by a coarse reticulation. An apical archaeopyle with an angular margin is typically present, the six-sided operculum often remaining attached to the shell. REMARKS. The combination of apical archaeopyle, tabulation and reticulation makes H. pdlla an easily recognizable species distinct from all previously described B C D FIG. 14. Histiocysta palla sp. nov., A. Lateral view of Holotype showing attached oper- culum, apical archaeopyle, precingular plates and cingulum (X 13)- B. Lateral view of Holotype showing well defined plate boundaries (X 13)- Ellipsodinium rugulosum Clarke & Verdier, C. Lateral view showing partially detached operculum and cingulum (X 1300). D. Lateral view (X 1300). i 4 o CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i forms. This species was figured, but not described, by Deflandre & Courteville (1939) as Micrhystridium sp.? from the Senonian. Membranilarnax cf. pterosper- moides Deflandre (1937) is of similar form but does not possess a reticulation. OCCURRENCE. H. palla is found throughout the European Cenomanian, except for the lowermost horizons. It is fairly common at Fetcham Mill but rare at Comp- ton Bay and Escalles. It has not been observed in the North American samples. Cyst-Family FROMEACEAE Sarjeant & Downie 1966 Genus FRO ME A Cookson & Eisenack 1958 REMARKS. A number of specimens referable to the type species, F. amphora, do not appear to possess a cingulum. The absence of a cingulum makes this genus similar to Chytroeisphaeridia Sarjeant (1962). They differ, however, in that the archaeopyle of Fromea has a rounded margin, and in the elongate shape typical of the latter genus. Fromea amphora Cookson & Eisenack (PI. 3, figs. 2, 3) 1958 Fromea amphora Cookson & Eisenack : 56, pi. 5, figs. 10, u. 19666 Fromea amphora Cookson & Eisenack; Sarjeant : 209, pi. 22, fig. 4; pi. 23, fig. 3 (see also for earlier references). DESCRIPTION. The shell is ovoidal, thick-walled (2-3 p,) and typically possesses an apical archaeopyle with a rounded margin. In one specimen (PL 3, fig. 3) the apical region is still attached and may be seen to be perfectly rounded. A cingulum was not observed in any of the specimens. DIMENSIONS. Range of observed specimens: shell length 56 (72-5) 85^, width 47 (60) 71 /A. Number of specimens measured, 6. REMARKS. The Cenomanian specimens are identical with the type material from the Aptian-Cenomanian of Australia except that the cingulum is absent. Cookson & Eisenack (1958) state, however, that the cingulum may be rather faint and it is probable that the European forms fall within the range of variation for this species. F. amphora has been recorded from the Barremian of England by Sarjeant (19666). The specimens described by Sarjeant, like the Cenomanian forms, do not possess a cingulum. Maliavkina et al. (1961) describes some very similar, but rather smaller, forms from the Maestrichtian of Siberia, calling them Chrysomonadinael . These forms possess the typical rounded archaeopyle of this genus. OCCURRENCE. Five specimens have been recorded from Fetcham Mill, from samples FM 810, 770, 750 and 650, and one specimen from Escalles, sample E 177. Genus CHYTROEISPHAERIDIA Sarjeant 1962 REMARKS. Chytroeisphaeridia and Canningia Cookson & Eisenack (19606) are similar and probably fairly closely related. The latter is usually more polygonal, has an apical horn, and sometimes the vestiges of a cingulum. CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 141 Chytroeisphaeridia euteiches sp. nov. (PL 3, figs. 8, 9) DERIVATION OF NAME. Greek, euteiches, well- walled with reference to the stout wall of this species. DIAGNOSIS. Shell subspherical; shell wall thick and densely granular. Angular apical archaeopyle typically present. HOLOTYPE. B.M. (N.H.) V. 51982 (2). Lower Chalk, Bureau de Recherches Geo- logiques et Minieres Borehole, Escalles, Pas de Calais, at 159 metres depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Holotype: shell length 53^, width 59 /*. Range: shell length 48-60 jit, width 49-59 p. Number of specimens measured, 6. DESCRIPTION. The shell wall is thick (2-3 /*) and may be composed of two layers. If the wall is bipartite then the inner layer is thin, the outer layer making up almost the entire wall thickness. This layer appears to be composed of minute cellular elements and is densely granular on the surface. The apical archaeopyle, when developed, is angular with small slits passing posteriorly from its margin between each precingular plate. A sulcal notch is also present (PI. 3, fig. 8). Plate boundaries and cingulum are not discernible. REMARKS. This species is a simple, subspherical shell possessing an apical archaeopyle. One species, C. chytroeides Sarjeant (1962), from the Upper Jurassic of England, differs from C. euteiches in having a fairly thin and only slightly granular shell wall. Chytroeisphaeridia sp. Sarjeant (19656) is similar in size and also granular but does not possess a thick shell wall. Canningia rotundata Cookson & Eisenack (1961) is also similar but tends to have a polygonal shell with a small apical horn. OCCURRENCE. C. euteiches has been recorded from two horizons only, E 165, & E 159 from Escalles, where it is infrequent. Genus CASSICULOSPHAERIDIA nov. DERIVATION OF NAME. Latin, cassiculus, hunting-net; sphaera, ball with reference to the surface reticulation of the shell. DIAGNOSIS. Proximate cysts; shell spherical to subspherical, composed of two layers, without apical or antapical protuberances. Periphragm giving rise to low crests or membranes which form a reticulate pattern. Tabulation absent. Archae- opyle apical with angular margin. TYPE Species. Cassiculosphaeridia reticulata sp. nov. Lower Chalk (Ceno- manian); France. REMARKS. The surface reticulation of this genus is very similar to that of Ellipsoidictyum cinctum Klement (1960). Both possess an apical archaeopyle, but whereas in Cassiculosphaeridia all signs of a tabulation are absent, in Ellipsoidictyum there is an obvious cingulum. Dictyopyxidia Eisenack (1961) is also very similar but possesses a cingulum and sulcus. 142 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i Cassiculosphaeridia reticulata sp. nov. (PL 3, fig- 7: PL 4, % 3) DERIVATION OF NAME. Latin, reticulatus, net-like with reference to the reticulate pattern formed by the periphragm crests. DIAGNOSIS. Shell spherical to subspherical. Shell surface bearing low ridges, forming a coarse reticulation, from which arise fine membranous crests. Shell wall lightly to densely granular. HOLOTYPE. B.M. (N.H.) .51981 (4). Lower Chalk, Bureau de Recherches Geologiques et Minieres Borehole, Escalles, Pas de Calais, at 165 metres depth. Upper Cretaceous (Cenomanian) . DIMENSIONS. Holotype: shell diameter 38 by 38/11, height of crests 4 to 6 p. Range : shell diameter 33 (43-6) 55 p,, maximum height of crests 3 (7-3) II /n. Number of specimens measured, 15. DESCRIPTION. The areas delimited by the low ridges are typically subpolygonal but may be of irregular shape. They vary considerably in size, from 2 to io/u, in diameter. The membranous crests are very fine and tend to be flexuous since supporting structures are absent. REMARKS. The surface reticulation, the absence of any tabulation and the apical archaeopyle together distinguished this species from all previously described forms. OCCURRENCE. C. reticulata is rare to common in samples from the Middle and Upper Cenomanian of Fetcham Mill (not found below sample FM 750), and from the Lower, Middle and Upper Cenomanian of Escalles (not found below sample E 207). This species was absent from the samples from Compton Bay and from North America. Genus EPELIDOSPHAERIDIA nov. DERIVATION OF NAME. Greek, epelidos, cover or lid; sphaera, ball with reference to the conical apical operculum which sometimes remains attached to the shell. DIAGNOSIS. Shell subpolygonal; epitract conical with small apical protuberance, hypotract polygonal with small antapical horn on one side. Shell wall two layered, periphragm giving rise to a moderate number of spines, truncated or forked distally Cingulum and sulcus outlined by spines. Cingulum slightly laevo-rotatory. Apical archaeopyle. TYPE SPECIES. Palaeoperidinium spinosum Cookson & Hughes 1964. Cambridge Greensand (Cenomanian), England. REMARKS. The presence of an apical archaeopyle, together with a well developed cingulum and sulcus differentiate Epdidosphaeridia from all previously described genera. Doidyx Sarjeant (19666) is most similar but differs in that the shell is asym- metrical, the hypotract is conical and a sulcus is absent. CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 143 Epelidosphaeridia spinosa (Cookson & Hughes) comb. nov. (PI. 3, figs. 10-12) 1964 Palaeoperidinium spinosum Cookson & Hughes : 49, pi. 8, figs. 6-8. 1967 Palaeoperidinium spinosum Cookson & Hughes: Clarke & Verdier: 70, pi. 14, figs. 10-12. DESCRIPTION. The shell possesses convex sides, conical epitract and a hypotract which is more or less truncated posteriorly. A small apical prominence is commonly present. The periphragm is smooth or lightly granular and forms a moderate num- ber of small, stout spines. The spines appear to be hollow, closed proximally, parallel sided, and are oblate distally or terminate with a small fork. The spines widen slightly before joining the shell and they are sometimes joined proximally. This is particularly well developed in the antapical region where the processes are joined medially and form a slight projection on one side of the shell. The cingulum is clearly delimited by two parallel lines of closely set spines and is 5 to 8 p in width. It is only slightly helicoid and bears few spines on its surface. The sulcus in most specimens is clearly defined, being slightly hollowed and almost devoid of spines. The spines on the remainder of the shell surface are usually randomly arranged, but occasionally a vague alignment is present suggesting a tabulation. An apical archa- eopyle is constantly developed, the margin being only slightly angular. DIMENSIONS. Range of observed specimens: shell length 32 (43) 56 fj,, width 27 (42-1) 57 p, maximum length of spines 2-5 (3-7) 5ju,. Number of specimens measured, 21. REMARKS. The Cenomanian specimens examined strongly resemble the type material from the Cambridge Greensand and Chalk Marl (Lower Cenomanian) of Cambridgeshire, England. In the lower horizons of the Cenomanian the sulcus, although always present, is not so clearly defined as in higher horizons. OCCURRENCE. E. spinosa is a rare to common species in the lower and middle horizons of the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It is absent from samples FM 690, 670, 650; CB 13, 15, 17, 19, 21; and E 159, 153. It has only beeen observed in one North American sample sample Sas 1023 (Saskatchewan, Albian). Cyst-Family HYSTRICHOSPHAERIDIACEAE Evit emend. Sarjeant & Downie 1966 Genus HYSTRICHOSPHAERIDIUM Deflandre emend. Davey & Williams 1966 Hystrichosphaeridium tubiferum (Ehrenberg) (PI. 5, figs. 5, 8) 1838 Xanthidium tubiferum Ehrenberg : pi. i, fig. 16. 19666 Hystrichosphaeridium tubiferum (Ehr.) Davey & Williams : 56, pi. 6, figs, i, 2; pi. 8, fig. 5; pi. 10, fig. 2; text-fig. 13. (See also for earlier references). DIMENSIONS. Range of observed specimens: diameter of central body 28 (38-1) 51 /z, maximum length of processes 15 (25-0) 37 //,. Number of specimens measured, 26. i 4 4 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i REMARKS. H. tubiferum is an infrequent to common species at all horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It is also recorded from the Albian (sample FM 886) and Turanian (sample FM 520) of Fetcham Mill. This species was not recorded in the North American material. Thus the earliest recording of H . tubiferum is from the Albian ; it ranges throughout the Upper Cretaceous and has been recorded from the Eocene (Ypresian) by Davey & Williams (19666). It is a long-ranging species of little stratigraphic value. Hystrichosphaeridium deanei Davey & Williams (PI. 4, %. i) 19666 Hystrichosphaeridium deanei Davey & Williams : 58, pi. 6, figs. 4, 8. 1967 Hystrichosphaeridium stellatum Maier; Clarke & Verdier: 55, pi. 12, figs, i, 2. DIMENSIONS. Range of observed specimens: diameter of central body 37 (45-7) 54 fji, maximum length of processes 22 (35-8) 45 [i. Number of specimens measured, 10. REMARKS. One specimen, occurring in sample E 207 (Escalles), possesses broad processes and appears to occupy a position midway between H. deanei and H. tubiferum. OCCURRENCE. H. deanei is a rare species confined to the Middle and Upper Cenomanian of Fetcham Mill, Compton Bay and Escalles (Table 25). The samples in which it first occurs are FM 710 (Fetcham Mill), CB 13 (Compton Bay) and E 195 (Escalles).//. deanei has also been recorded from the Turonian sample FM520, from Fetcham Mill. It is absent from the North American material. Hystrichosphaeridium readei Davey & Williams 19666 Hystrichosphaeridium readei Davey & Williams : 64, pi. 6, fig. 3 (See also for earlier references) . DIMENSIONS. Range of observed specimens: diameter of central body 31 (42-1) 57 p, maximum length of processes 23 (29-7) 35 /t. Number of specimens measured, ii. OCCURRENCE. H. readei is very rare in samples FM 810, 790, 770 (Fetcham Mill) and E 207 (Escalles); and it is rare to infrequent in samples FM 690, 670, E 183, 177, X 65, 159 and CB 9 (Compton Bay). Two specimens were recorded in the Albian sample (FM 886) from Fetcham Mill. This species was not recorded in the North American samples. Hystrichosphaeridium radiculatum Davey & Williams (PI. 4, fig- 8) 19666 Hystrichosphaeridium radiculatum Davey & Williams : 65, pi. 7, fig. 9; pi. 9, fig. 6. DIMENSIONS. Range of observed specimens: diameter of central body 27 (38-8) 43 ju, maximum length of processes 12 (15-9) 20 /*. Number of specimens measured, II. CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 145 OCCURRENCE. H. radiculatum is a very rare to infrequent species occurring at a number of horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It has also been recorded in the Albian sample from Fetcham Mill (sample FM 886), but is absent from the North American material. Hystrichosphaeridium mantelli Davey & Williams (PI- 4. ng. 9) 19666 Hystrichosphaeridium mantelli Davey & Williams : 66, pi. 6, fig. 6. DIMENSIONS. Range of observed specimens: diameter of central body 32 (38-6) 48 fi, maximum length of processes 12 (21-4) 26^. Number of specimens measured, 15- REMARKS. The reticulate nature of the central body and the fibrous processes differentiate H. mantelli from most previously described species. H. radiculatum is the most similar but differs from H. mantelli by the more branched and deeply furcate processes and the tendency for the fibrils of the processes to continue across the surface of the central body. OCCURRENCE. H. mantelli is a very rare to infrequent species occurring in most samples throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It has also been recorded in the Turonian sample from Fetcham Mill (sample FM 520), but is absent from the North American material. Hystrichosphaeridium bowerbanki Davey & Williams (PI- 5, fig- 9) 19666 Hystrichosphaeridium bowerbanki Davey & Williams : 69, pi. 8, figs. 1,4. DIMENSIONS. Range of observed specimens: diameter of central body 25 (31-9) 40jLt, maximum length of processes 20 (25-5) 28 p. Number of specimens measured, 10. REMARKS. H. bowerbanki is rare to infrequent in six Middle Cenomanian samples FM 770, 750, 730 and 690 from Fetcham Mill and CB 9, and 17 from Compton Bay. It has also been recorded from the Albian sample (FM 886) and the Turonian sample (FM 520), both from Fetcham Mill. Hystrichosphaeridium difficile Manum & Cookson (PI. 4, figs. 2, 6, 7) 1964 Hystrichosphaeridium difficile Manum & Cookson : 12, pi. 3, figs. 1-3, 7. DESCRIPTION. The shell is subspherical, sometimes with a small apical prom- inence; shell wall thick (c. i /JL), smooth to lightly granular. The processes, approxi- mately 30 in number, are complex, usually broadly tubiform or buccinate in shape, of constant length on any specimen but variable in width (3 to 18/11). Distally the 146 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i larger processes have a rectangular opening, with a serrate margin which gives rise to four broad spines. Typically there are four bands of thickening extending along the length of the processes and passing onto the shell surface. There, each joins with a similar thickening from a neighbouring process, thus forming a coarse reticula- tion on the shell surface. The large tubular processes are arranged in a circular manner around the shell, reflecting the precingular, cingular and postcingular series of plates. Fine processes are uncommon and may be sulcal in position. An apical archaeopyle is typically developed and possesses an angular margin marked at intervals with V-shaped notches. Although often remaining attached, isolated operculae have been identified (PI. 4, figs. 6, 7) and bear four moderate-sized tubular processes. DIMENSIONS. Range of observed specimens: diameter of central body 49 (64-9) 79 p,, maximum length of processes 18 (24-5) 31 p. Number of specimens measured, 8. REMARKS. The specimens studied are extremely similar to the type material of Lower Cretaceous age described by Manum & Cookson (1964) from Arctic Canada. The only difference appears to be that in the type material the shell surface has a fine reticulation which is absent from the specimens studied. Two similar species are H. costatum Davey & Williams (19666) from the Oxford Clay of England and H. readei Davey & Williams (19666) from the Cenomanian of England. However, both species are considerably smaller than H. difficile and possess narrower and less complex processes. OCCURRENCE. H. difficile has only been recorded from the Saskatchewan material. It is infrequent in samples Sas 1084 (Albian) and Sas 890 (Cenomanian), and common in sample Sas 805 (Cenomanian). Genus OLIGOSPHAERIDIUM Davey & Williams 1966 Oligosphaeridium complex (White) (PI. 5, figs. 6, 7) 1842 Xanthidium tubiferum complex White : 39, pi. 4, div. 3, fig. n. 19666 Oligosphaeridium complex (White) Davey & Williams : 71, pi. 7, figs, i, 2; pi. 10, fig. 3; text-fig. 14 (See also for earlier references). 1967 Hystrichosphaeridium complex (White) Clarke & Verdier: 53, pi. n, figs. 10, n. DIMENSIONS. Range of observed specimens: diameter of central body 34 (41-1) 55 /A, maximum length of processes 22 (34*6) 43 //,. Number of specimens measured, 12. REMARKS. Davey & Williams described examples of 0. complex from the Speeton Clay (Barremian), Lower Chalk (Cenomanian) and London Clay (Ypresian), all from England. OCCURRENCE. 0. complex is a rare to common species in all samples from Fetcham Mill, Compton Bay and Escalles. It is also present in the Albian sample (FM 886) and the Turonian sample (FM 520) both from Fetcham Mill. This species is present CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 147 in the Albian and the lowermost Cenomanian sample from Saskatchewan Sas 1084 1023, 967, and 890 but is absent from the Texas samples. 0. complex thus has a known stratigraphic range from the Neocomian (Gocht 1959; Cookson & Eisenack 1958) to the Eocene, Ypresian (Davey & Williams 19666). Oligosphaeridium reticulatum Davey & Williams 19666 Oligosphaeridium reticulatum Davey & Williams : 74, pi. 7, fig. 10. DIMENSIONS. Range of observed specimens: diameter of central body 29-47/4, length of processes 14-26/4. Number of specimens measured, 5. OCCURRENCE. 0. reticulatum is an infrequent species occurring in the two lower samples from Fetcham Mill, samples FM 840 and 810. Oligosphaeridium prolixispinosum Davey & Williams (PI- 5, %. 4) 19666 Oligosphaeridium prolixispinosum Davey & Williams : 76, pi. 8, figs. 2, 3. DIMENSIONS. Range of observed specimens: length of central body 33 (39-0) 43 /A, width 20 (28-7) 34 /u, maximum length of processes 18 (25-0) 30/4. Number of specimens measured, 15. OCCURRENCE. 0. prolixispinosum is a rare species confined mainly to the Lower and Middle Cenomanian of Fetcham Mill and Escalles, although it does occur occas- ionally in the Upper Cenomanian. This species has only been recorded twice in the samples CB 17 and 21 from Compton Bay, both from the Upper Cenomanian. It has not been recorded in the North American material. Oligosphaeridium anthophorum (Cookson & Eisenack) (PI. 5, figs, i, 2, 3) 1958 Hystrichosphaeridium anthophorum Cookson & Eisenack : 43, pi. n, figs. 12, 13; text-figs. 16-18. 1958 Hystrichosphaeridium anthophorum Cookson & Eisenack; Eisenack, 402, pi. 26, figs, i, 2. 1961 Hystrichosphaeridium anthophorum Cookson & Eisenack; Alberti : 34, pi. 9, fig. 16. 19666 Oligosphaeridium anthophorum (Cookson & Eisenack) Davey & Williams : 77. DESCRIPTION. The shell is subspherical ; shell wall smooth to lightly granular. An apical archaeopyle is typically developed and possesses an angular margin. The processes are hollow, buccinate to infundibular, with the distal flared portion per- forate. Distally the margins of the processes are usually entire, but may bear one or two small spines. A complete specimen possesses 18 processes, reflecting a tabulation characteristic of this genus. The processes are all of equal size except for the first postcingular (i"') and the posterior intercalary (ip) which are often reduced. DIMENSIONS. Range of observed specimens: diameter of central body 38 (46-3) 57 /z, maximum length of processes 18 (34-1) 43 /*. Number of specimens measured, 8. 148 CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i REMARKS. The presence of complex perforate processes having an entire distal margin differentiate this species from all other similar forms. OCCURRENCE. 0. anthophorum has been recorded from the Upper Jurassic Lower Cretaceous (Aptian-Albian) of Australia (Cookson & Eisenack 1958), from the Aptian of Germany (Eisenack 1958) and from the Upper Barremian Albian of Germany (Alberti 1961). This species has been recorded from one sample, Sas 1023, from the Albian of Saskatchewan where it is common. Oligosphaeridium reniforme (Tasch) (PI. 6, fig. i) 1964 Hystrichosphaeridium reniforme Tasch : 193, pi. 2, fig. 6. 19666 Oligosphaeridium reniforme (Tasch) Davey & Williams : 77. DESCRIPTION. The shell is subspherical to ovoidal, shell wall lightly granular. The processes are hollow, tubiform, widening distally into a broad, flat-topped funnel. The distal margin of the funnel bears a small number of pointed and irregularly shaped spines. Distally the processes sometimes possess large, subcircular perfora- tions. An apical archaeopyle is typically developed. DIMENSIONS. Range of observed specimens: diameter of central body 31 (42-4) 49 fj,, maximum length of processes 20 (24-9) 30 p. Number of specimens measured, 7. REMARKS. The specimens appear to be very similar to the type material from the Albian of Kansas although Tasch did not describe the presence of distal perfora- tions. 0. reniforme differs from 0. anthophorum and 0. perforatum (Gocht 1959) by the presence of distal spines. It differs from 0. pulcherrimum Deflandre & Cookson (1955) in that the processes possess only a few spines, are not so complexly perforate and distally are flat-topped. OCCURRENCE. 0. reniforme is infrequent in all the Albian-Cenomanian samples from Saskatchewan. It has not been recorded elsewhere. Genus LITOSPHAERIDIUM Davey & Williams 1966 Litosphaeridium siphoniphorum (Cookson & Eisenack) (PI. 6, figs. 3, 4; Fig. 15) 1958 Hystrichosphaeridium siphoniphorum Cookson & Eisenack : 44, pi. n, figs. 8-10. 19666 Litosphaeridium siphoniphorum (Cookson & Eisenack) Davey & Williams : 80, pi. 7, figs. 7, 8; text-figs. 16, 17. (See also for earlier references). 1967 Hystrichosphaeridium siphoniphorum Cookson & Eisenack; Clarke & Verdier: 55, pi. n, figs, i, 2. DIMENSIONS. Range of observed specimens: diameter of central body 21 (34-0) 47 /u,, length of processes 4-25 //,. Number of specimens measured, 160. DESCRIPTION. A statistical study was made of this species using 20-30 specimens from each of six samples at 4O-foot intervals from the Fetcham Mill Borehole. The purpose of this study was to see if the variation in the mean size of L. siphoniphorum was directional, and if there was a significant difference in this measurement for 0) -Q E CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 149 40 30 B 50 60 50 60 70 40 50 60 70 40 50 60 40 50 60 40 50 60 70 Overall diameter FIG. 15. Overall diameter frequency histograms of Litosphaeridium siphoniphorum (Cookson and Eisenack) at six horizons from Fetcham Mill, Surrey. A. Sample FM 650, B. Sample FM 690, C. Sample FM 730, D. Sample FM 770, E. Sample FM 810, F. Sample FM 840. 150 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i successive samples. The measurement taken was the overall diameter. The position of the archaeopyle is always obvious, allowing easy specimen orientation. Thus to make all measurements strictly comparable the overall diameter was always taken in the plane of the archaeopyle. Histograms (Fig. 15) were drawn for each assemblage. The mean overall diameter for each assemblage varied for each horizon but, unfortunately, the variation was not directional and, therefore, was of little stratigraphic value. The Student's t-test was performed on successive pairs of assemblages to see whether or not they were significantly different (Table A). A probability of 0-05 or less was taken as being significant. FM650 FM6go FM6go FM730 FM 770 FM 770 FMSio FMSio FM8 4 o (x= 50-0 ft) x = (x= 56-6 fj) (x= 4 8- 3yM ) (x = 48-3/0 (x= 48-10) (X = 4 8-I ft) (X= 48-1 ft) (x= 48-1/0 (x= 58-0/0 TABLE A t = 3-08 t = 3-16 t = 0-084 t = o t=5'4 (significant difference) (significant difference) (no significant difference) (no significant difference) (significant difference) The results show that there is a significant difference at the 5 % level between some of the assemblages with respect to this character. However, all the specimens measured were apparently morphologically identical and differ only in size. Speci- mens from one sample vary considerably in size but were probably formed by one species of motile dinoflagellate. Thus the size of L. siphoniphorum appears to be quite variable and should, at the moment, not be used as a diagnostic feature for the subdivision of this species. The reason for the means in successive samples to be significantly different is probably because of palaeoecological changes in the environment. REMARKS. All the specimens of L. siphoniphorum examined agree fairly closely with the type material from Australia. The Surrey specimens appear to be smaller, but the range of the Australian forms was not given so no true size comparison can be made. OCCURRENCE. L. siphoniphorum is rare to common at all horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It is present in the Albian sample FM 866, but absent from the Turonian sample, FM 520. It is rare to infrequent in the lower four samples from Saskatchewan samples Sas 1084, 1023, 967 and 890. This species is also present in the Upper Woodbine Formation of Texas. Thus L. siphoniphorum has a wide geographical distribution and a fairly restricted range. It has been recorded from the Albian of Australia, Rumania, Canada and Britain and from the Cenomanian of Australia, Canada and Britain. CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 151 Genus POLYSPHAERIDIUM Davey & Williams 1966 Polysphaeridium pumilum Davey & Williams ?I955 Hystrichosphaeridium recurvatum White; Deflandre & Cookson : 269, pi. i, fig. 12. 19666 Polysphaeridium pumilum Davey & Williams : 93, pi. 7, figs. 3, 4. DIMENSIONS. Range of observed specimens : overall diameter 30-40 //,, diameter of central body 17-25/1,, length of processes 7-10/1,, width of processes 1-1-5 /A, number of processes 38-44. Number of specimens measured, 3. OCCURRENCE. Only three specimens of P. pumilum have been observed, one from sample FM 750 and two from sample FM 770. Polysphaeridium laminaspinosum Davey & Williams (PL 4, figs. 10, ii) 19666 Polysphaeridium laminaspinosum Davey & Williams : 94, pi. 8, fig. 8. DIMENSIONS. Range of observed specimens: diameter of central body 20 (26-8) 29 [L, maximum length of processes 9 (13-7) 17 p. Number of specimens measured, 8. OCCURRENCE. P. laminaspinosum is rare to very rare, occurring spasmodically throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It has not been recorded elsewhere. Genus TANYOSPHAERIDIUM Davey & Williams 1966 Tanyosphaeridium variecalamum Davey & Williams (PI. 6, figs. 2, 5) 19666 Tanyosphaeridium variecalamum Davey & Williams : 98, pi. 6, fig. 7; text-fig. 20. DIMENSIONS. Range of observed specimens : length of central body 27 (32-3) 43 /*, width 14 (20-1) 24 n, maximum length of processes 11 (15-0) 24/4. Number of specimens measured, 14. OCCURRENCE. T. variecalamum is a rare to infrequent species at most horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It is also present in the Albian sample, FM 886, and the Turonian sample, FM 520, both from Fetcham Mill. One specimen was located in the Saskatchewan material, from the Albian sample Sas 1084. Genus CALLAIOSPHAERIDIUM Davey & Williams 1966 REMARKS. Hexasphaera Clarke & Verdier (1967; 42) is a junior synonym of Callaiosphaeridium . GEOL. 17, 3 IX 152 CENOMANIAN NON-CALCAREOUS MICROPLAN KTON, i Callaiosphaeridium asymmetricum (Deflandre & Courteville) (PI. 6, fig. 6) 1939 Hystrichosphaeridium asymmetricum Deflandre & Courteville : 100, pi. 4, figs, i, 2. 19666 Callaiosphaeridium asymmetricum (Deflandre & Courteville) Davey & Williams : 104, pi. 8, figs. 9, 10; pi. 9, fig. 2. 1967 Hexasphaera asymmetrica (Deflandre & Courtville) Clarke & Verdier: 43, pi. 7, figs. 1-3 text-fig. 17. DIMENSIONS. Range of observed specimens: diameter of central body 37 (45-8) 58 ju, maximum length of cingular processes 10 (25-0) 32/1,. Number of specimens measured, 6. REMARKS. The author disagrees with the description of Clarke & Verdier (1967) in that the archaeopyle is epitractal, not apical, and that the antapical plate is five-sided, not six-sided, bearing a process at each corner. The position of the three sutural crests separating the large plates has, however, been verified. OCCURRENCE. C. asymmetricum is rare to infrequent at most horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It has been recorded in the Albian sample FM 886 and the Turonian sample FM 520, both from Fetcham Mill, but is absent from the North American material. Genus CLEISTOSPHAERIDIUM Davey, Downie, Sarjeant & Williams 1966 Cleistosphaeridium heteracanthum (Deflandre & Cookson) (PI. 7, fig- 8) 1955 Hystrichosphaeridium heteracanthum Deflandre & Cookson : 276, pi. 2, figs. 5, 6; text-figs. 40, 41. 1966 Cleistosphaeridium heteracanthum (Deflandre & Cookson) Davey, Downie, Sarjeant & Williams : 168, pi. 2, figs. 6, 7 (See also for earlier references). DIMENSIONS. Range of observed specimens: shell diameter 42 (53-2) 63^, maximum length of processes 9 (13-8) 17 ju. Number of specimens measured, 9. OCCURRENCE. C. heteracanthum is a rare species restricted to the Upper Ceno- manian samples FM 690, 670 and 650 from Fetcham Mill, samples CB 19 and 21 from Compton Bay and sample E 153 from Escalles. It was not recorded from North America. Cleistosphaeridium multifurcatum (Deflandre) (PI. 8, figs. 7, 10) 1937 Hystrichosphaeridium multifurcatum Deflandre : 76, pi. 16, figs. 1-3. 1939 Hystrichosphaeridium multifurcatum Deflandre; Deflandre & Courteville : 102, pi. 3, fig. 2. 1952 Hystrichosphaeridium multifurcatum Deflandre; W. Wetzel : 400, text-fig. 16. 1955 Hystrichosphaeridium multifurcatum Deflandre; Valensi : 588, pi. i, fig. 21; pi. 5, fig. 5. 1960 Baltisphaeridium multifurcatum (Deflandre) Klement : 59. CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 153 1963 Hystrichosphaeridium multifurcatum Deflandre; Gorka : 66, pi. 9, figs. 4-6; text-fig. 8, fig. i. 1966 Cleistosphaeridium multifurcatum (Deflandre) Davey, Downie, Sarjeant & Williams: 170. DIMENSIONS. Range of observed specimens: diameter of central body 39 (55-0) 68 fj,, maximum length of processes 8 (ii'2) 13 //.. Number of specimens measured, 8. OCCURRENCE. C. multifurcatum is rare at most horizons in the Cenomanian of Fetcham Mill, Compton Bay and Escalles. In a few Upper Cenomanian samples (FM 690, 670 and 650; CB 19, 21 and E 153) it is completely absent. In these samples it seems to be replaced by C. heteracanthum. C. multifurcatum was not observed in the North American samples. Cleistosphaeridium armatum (Deflandre) comb. nov. (PL 8, figs, i, 2, 12) 1937 Hystrichosphaeridium armatum Deflandre : 76, pi. 16, figs. 6, 7. 1947 Hystrichosphaeridium armatum Deflandre; Deflandre: fig. i, No. 10. 1952^ Hystrichosphaeridium armatum Deflandre; Deflandre: fig. 14. 1963 Baltisphaeridium armatum (Deflandre) Downie & Sarjeant : 91. 1967 Baltisphaeridium armatum (Deflandre) Clark & Verdier: 71, pi. 13, fig. 3. EMENDED DIAGNOSIS. Shell subspherical ; shell wall of moderate thickness, densely granular. Processes numerous, fairly broad, rigid, tapering gradually distally. Proximally processes possess longitudinal basal striations; distally simple or giving rise to variable number of small spines. Apical archaeopyle occasionally developed. HOLOTYPE. Slide AJ. 54, Laboratoire de Micropaleontologie, Ecole Practique des Hautes Etudes, Paris. (Figured by Deflandre 1937, pi. 16, fig. 6). Upper Cretaceous flint from the Paris Basin. DIMENSIONS. Holotype: shell length 20/11, width 18-20 /z, length of processes 10-15 p. Range of Cenomanian specimens: diameter of central body 19 (30-8) 42 /u,, maximum length of processes 5 (9-6) 16 //,. Number of specimens measured, 33. DESCRIPTION. The granules are elongate (c. 0-1-0-5 P i n height) so giving the shell surface the appearance of possessing a matting of short hairs. This ornamen- tation, commented on by Deflandre in the original description of this species, was verified by the present author when examining the type material in Paris. The processes are all of a similar length on any one specimen and may vary in width from i to 3 JM. They are hollow, always closed distally terminating either simply (the extremity may be recurved) or more commonly the distal one quarter of the processes bear a small number of stiff spines (pi. 8, fig. 2). The opening, when observable, possesses an angular margin characteristic of an apical archaeopyle. REMARKS. The specimens studied strongly resemble C. armatum as described by Deflandre from Upper Cretaceous flints, one of which was of Cenomanian age. The appearance of the shell surface and form of the processes make this an easily recognizable species. OCCURRENCE. C. armatum is common at all horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It is also present in the Upper Wood- 154 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i bine and Lower Eagle Ford formations of Texas, but has not been observed in the Saskatchewan samples. Cleistosphaeridium polypes (Cookson & Eisenack) comb. nov. (PI. 6, figs. 7, 8) 19626 Hystrichosphaeridium recurvatum subsp. polypes Cookson & Eisenack : 491, pi. 4, figs. 11-13. DESCRIPTION. The shell is spherical to subspherical and bears numerous slender processes. The shell wall is thin and may be smooth or slightly granular. The processes are narrow, usually less than i p in width for most of their length, broaden- ing slightly proximally. They are closed distally and terminate in a number of short, fine spines. The latter are not so numerous in the English and French speci- mens as in those from Australia and North America where the distal spines are numerous. A fairly large archaeopyle, probably apical, with an angular margin is often present. DIMENSIONS. Range of observed specimens : shell diameter 31 (38-4) 47 //,, maxi- mum length of processes 9 (13-4) 18 /A. Number of specimens measured, 20. REMARKS. C. polypes was originally described by Cookson & Eisenack from the Aptian-Cenomanian of Australia and was considered to be a subspecies of Hystrichosphaeridium recurvatum (White) . The resemblance between the two forms is very slight and they are not considered to be closely related. C. polypes is placed in this genus because of the presence of numerous closed processes and the probable apical location of the archaeopyle. OCCURRENCE. C. polypes is a very rare species restricted to the Middle and Upper Cenomanian of Fetcham Mill (samples FM 750, 730), Compton Bay (samples CB 9, 17, 21) and Escalles (samples E 189, 177, 165). It is common in the Upper Wood- bine formation of Texas, and is present in one sample from Saskatchewan (sample Sas 835). Cleistosphaeridium polypes var. clavulum nov. (PI. 6, figs. 9, 10) 1964 Hystrichosphaeridium recurvatum subsp. polypes Cookson & Eisenack; Cookson & Hughes : 47, pi. 9, fig. 14. DERIVATION OF NAME. Latin, clavulus, small nail with reference to the pin-like shape of the processes. DIAGNOSIS. A variety of C. polypes possessing fine, capitate processes. Processes terminating with fine spines are extremely rare. TYPE. G.S.M. slide PF 3995(1). Lower Chalk, H.M. Geological Survey Borehole, Fetcham Mill, Surrey at 840 feet depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Type: diameter of central body 29 by 32 /x, length of processes 12-13 /A. Range: diameter of central body 29 (31-5) 39 p, maximum length of processes 13 (14-7) 15 p. Numbers of specimens measured, 10. CENOMANIAN NON-CALCAREOUS MICROPL AN KTON. i 155 REMARKS. C. polypes var. davulum differs from C. polypes in that the processes are capitate, the terminal bulge being flattened and resembling the head of a pin. Two specimens have been observed each possessing one process of the type charac- teristic of C. polypes thus indicating a relationship between the two forms. C. polypes probably evolved from this variety, spiny processes replacing the capitate ones. OCCURRENCE. This variety has only been recorded from the lowermost Cenoman- ian samples at Fetcham Mill (sample FM 840) and Compton Bay (sample CB i). It was previously recorded by Cookson & Hughes (1964) from the Upper Albian basal Cenomanian of England and so appears to be of stratigraphic importance in England for indicating the base of this stage. It does not occur in the lowermost sample from Escalles. Cleistosphaeridium huguonioti (Valensi) comb. nov. (PI. 7, fig. 10) 1955 Hystrichosphaeridium huguonioti Valensi : 38, fig. 2a. ig6oa Hystrichosphaeridium ancoriferum Cookson & Eisenack : 8, pi. 2, fig. n. 1963 Hystrichosphaeridium ancoriferum Cookson & Eisenack; Baltes : 586, pi. 6, fig. 13. 1964 Hystrichosphaeridium ancoriferum Cookson & Eisenack; Cookson & Hughes : 47, pi. 9, fig. 7- 1964 Chlamydophorella nyei Cookson & Eisenack; Cookson & Hughes : 54, pi. 6, fig. 12. 1966 Cleistosphaeridium ancoriferum (Cookson & Eisenack) Davey, Downie, Sarjeant & Williams: 167, pi. 9, fig. i. 1967 Hystrichosphaeridium huguonioti Valensi; Clarke & Verdier: 54, pi. n, fig. 4, 5. DESCRIPTION. The shell is subspherical; the shell wall is smooth and gives rise to numerous bifurcating processes which are not aligned to any noticeable extent. Most of the specimens possess an apical archaeopyle, the shape of which is usually difficult to determine because of distortion. However, detached apical regions are common and are 6-sided. The processes are hollow, the central cavity often being constricted to some extent along its length, and closed distally and proximally. The sides of the processes are practically parallel, diverging slightly proximally before joining the shell. Distally they give rise to two slightly recurved spines. Cookson & Eisenack comment on the " transparent tips " of the processes. The extremities of the processes are in fact closed by a thin, transparent membrane. The processes may be isolate or a few may be linked together distally by their spines. The shell wall rarely forms a small rounded apical bulge. DIMENSIONS. Range of observed specimens: diameter of central body 20 (31-8) 45 ju, length of processes up to 8 p. Number of specimens measured, 30. REMARKS. Cookson & Hughes (1964) described C. huguonioti and Chlamydo- phorella nyei from the Upper Albian and Lower Cenomanian of England, distinguish- ing the latter, with difficulty, by the presence of an outer membrane and an apical bulge. After a detailed examination of numerous Cenomanian specimens it was concluded that Chlamydophorella nyei s.s., as described from Australia, does not occur in the English and French Cenomanian and that C. nyei as described by Cookson & 156 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i Hughes belongs in Cleistosphaeridium huguonioti. C. huguonioti does occasionally appear to possess a membrane linking some of the processes when their spines are joined distally, and an apical prominence may also rarely be present. Some speci- mens possess joined processes but apparently no apical prominence, whereas others possess an apical prominence but no joined processes. The presence of distinctive bifurcate processes and an apical prominence indicates that C. huguonioti is related to Chlamydophorella, the outer membrane of the latter being reduced to a small, transparent membrane at the distal end of each process. OCCURRENCE. C. huguonioti is common throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles (Fig. 22). It is absent from the uppermost sample from Escalles (sample E 153) and from the Turonian sample FM 520. In the Upper Cenomanian at these localities C. huguonioti tends to be replaced by its variety, C. huguonioti var. pertusum nov., which appears to become more abundant as C. huguonioti declines. It is present in the Albian sample FM 886 and is common in the Upper Woodbine and Lower Eagle Ford formations of Texas, but was not observed in the Saskatchewan material. This species was first recorded from the Albian-Cenomanian of Australia and is also present in the Albian-Cenomanian of Rumania (Balte, 1963). Thus C. huguonioti appears to be wide-spread geographically and restricted to the Albian and Cenomanian. Cleistosphaeridium huguonioti var. pertusum nov. (PL 7, figs. 6, 7, 9) DERIVATION OF NAME. Latin, pertusum, perforated with reference to the per- forate appearance of the processes. DIAGNOSIS. A variety of C. huguonioti possessing spherical to subspherical, smooth walled shell bearing numerous processes. Processes broad-based, tapering distally and terminating with two small recurved spines. Lumen of processes restricted by transverse septa. HOLOTYPE. G.S.M. slide PF 3040(2). Lower Chalk, H.M. Geological Survey Borehole, Fetcham Mill, Surrey at 670 feet depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Holotype : diameter of central body 36 by 36 p,, length of processes 7-9 IJL, maximum width of processes distally 2 /*. Range : diameter of central body 24 (34-1) 46 //,, length of processes 6 (8-7) II p, maximum width of processes distally i (2-7) 4-5 p. Number of specimens measured, 24. DESCRIPTION. The processes taper distally from a fairly broad base (2-5-4^ in width) to a narrow neck (c. 0-5 p in width) before bifurcating to give two short, recurved spines. The processes are hollow but the lumen is traversed by a number of small septa which thus subdivide it, giving the processes a " holey " or vacuolated appearance. An archaeopyle is only rarely observable. REMARKS. This variety differs from C. huguonioti in the form of its processes which are vacuolated, tend to be longer and bear relatively small distal spines. C. huguonioti found in the same samples possesses processes which are considerably CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 157 wider distally (5-5-7-5 /i). C. huguonioti var. pertusum appears to have evolved from C. huguonioti in the Middle to Upper Cenomanian. OCCURRENCE. C. huguonioti var. pertusum occurs only in the Upper Cenomanian, where it is common at all horizons. It first occurs in samples FM 710 (Fetcham Mill), CB 15 (Compton Bay) and E 171 (Escalles). At these horizons it is occasionally difficult to distinguish from C. huguonioti. Like C. huguonioti, this variety is absent from sample E 153 (Escalles) and FM 520 (Fetcham Mill, Turonian). It has not been observed in the samples from North America. fCleistosphaeridium flexuosum Davey, Downie, Sarjeant & Williams (PI. 7, figs. 4) 1966 ? Cleistosphaeridium flexuosum Davey, Downie, Sarjeant & Williams: 169, pi. 2, fig. 5. OCCURRENCE. This is a very rare species occurring at most horizons throughout the Cenomanian of Fetcham Mill. It has not been recorded elsewhere. ? Cleistosphaeridium parvum sp. nov. (PL 7, figs, ir, 12) DERIVATION OF NAME. Latin, parvus, little with reference to the small size of this cyst. DIAGNOSIS. Shell ovoidal, small; shell wall smooth, bearing numerous long, fine spines. Spines may be aligned along upper and lower boundaries of cingulum. Cingulum strongly laevo-rotatory. Apical archaeopyle typically developed. HOLOTYPE. B.M. (N.H.) .51981 (3). Lower Chalk, Bureau de Recherches Geologique et Minieres Borehole, Escalles, Pas de Calais at 165 metres depth. Upper Cretaceous (Cenomanian) . DIMENSIONS. Holotype: length of central body I2//,, width iiyu,, length of processes 6-n p. Range: length of central body 11-14 /x, width 10-13 /z, length of processes 6-12 /*. Number of specimens measured, 5. DESCRIPTION. The spines are long, very fine, terminate distally in a point and widen only slightly when joining the shell. The cingulum is not always observable but some alignment of the spines parallel to the archaeopyle margin is usually pres- ent. REMARKS. ?C. parvum resembles only one previously described species, that is Palaeostomocystis echinulata Deflandre (1937) from the Upper Cretaceous of France. This species differs in that there are fewer spines (c. 12), but is similar in its ovoidal shape, apical archaeopyle, cingulum and small size (6-7 //. long). The presence of a cingulum has not previously been observed in members of Cleistosphaeridium and might later be used as a character in generic subdivision. However, difficulty of observation makes it of dubious value. 158 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i OCCURRENCE. ?C. parvum is fairly common in three Upper Cenomanian samples from Escalles E 171, 165 and 159. It has also been observed in a single Lower Cenomanian sample from Compton Bay (CB 5). In all other samples it appears to be absent. fCleistosphaeridium aciculare sp. nov. (PL 6, figs, n, 12) DERIVATION OF NAME. Latin, adcularis, like a needle with reference to the acuminate shape of the processes. DIAGNOSIS. Shell spherical to subspherical ; shell wall of moderate thickness, densely granular. Processes numerous, finely to broadly acuminate, slightly flexuous, less than one-third of shell diameter in length. HOLOTYPE. B.M. (N.H.) slide V. 51979 (3). Second White Speckled Shale, International Yarbo Borehole No. 17, Saskatchewan at 835 feet depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Holotype : diameter of central body 43 by 50 /A, length of processes 12-14 /* Range: diameter of central body 32 (43-0) 54/^1, maximum length of processes 8 (13-9) 21 //.. Number of specimens measured, n. DESCRIPTION. The processes may be finely or broadly acuminate but on each individual their width is constant. On individuals bearing fine processes these are more densely packed than in individuals with broad processes. All intergradations exist between the fine and the broad processed forms. The processes are always pointed distally and occasionally bear small subsidiary spines near their extremities. An archaeopyle has never been observed. REMARKS. ?C. aciculare is only tentatively placed in this genus, for although it resembles other members in overall appearance, an apical archaeopyle has not been observed. The numerous acuminate processes and the densely granular shell sur- face differentiate ?C. aciculare from most previously described forms. Most similar seems to be Exochosphaeridium (Hystrichosphaeridiurri) cf. striolatum (Deflandre) as illustrated by G6rka (1963, pi. 10, fig. 6) from the Cenomanian of Poland. OCCURRENCE. 1C. aciculare is common in the following samples from Saskat- chewan Sas 1084, 1023 and 967 (all Albian) and Sas 835 (Cenomanian). It has not been recorded elsewhere. Genus SURCULOSPHAERIDIUM Davey, Downie, Sarjeant & Williams 1966 Surculosphaeridium longifurcatum (Firtion) (PI. 8, fig. 9) 1952 Hystrichosphaeridium longifurcatum Firtion : 157, pi. 9, fig. i ; text-fig, i, H, K, L and M 1963 Baltisphaeridium longifurcatum (Firtion) Downie & Sarjeant: 91. 1966 Surculosphaeridium longifurcatum (Firtion) Davey, Downie, Sarjeant & Williams : 163, pi. 8, figs. 7, ii ; text-figs. 43, 44. CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 159 DIMENSIONS. Range of observed forms: lateral view diameter of central body 30 (38-3) 47 /LI; apical view diameter of central body 36 (42-2) 50/1., maximum length of processes 14 (22-2) 29 /*. Mean diameter of archaeopyle, 20 //,. Number of specimens measured, 24. OCCURRENCE. This is a rare species at all horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. The only exception is sample FM 730 from Fetcham Mill, where this species is very common, composing 22-5 % of the total microplankton present. The reason for this unusual abundance is unknown. 5. longifurcatum is also present in the Lower Eagle Ford formation of Texas, but was not observed in the Saskatchewan samples. It has not been observed in the Albian sample (FM 886) or the Turonian sample (FM 520) from Fetcham Mill, and thus appears to be characteristically, Cenomanian. Genus HYSTRICHOKOLPOMA Klumpp emend. Williams & Downie 1966 Hystrichokolpoma ferox (Deflandre) (PL 9, figs. 5-7) 1937 Hystrichosphaeridium ferox Deflandre : 72, pi. 14, figs. 3, 4. ig66a Hystrichokolpoma ferox (Deflandre) Williams & Downie : 181. 1967 Baltisphaeridium ferox (Deflandre) Clarke & Verdier: 73, pi. 15, fig. 4 (see also for earlier references). EMENDED DIAGNOSIS. Shell subspherical, densely granular or reticulate. Pro- cesses thin walled, granular, often striated, of three kinds : (i) 6 large precingular and 4 large postcingular, possessing wide bases and tapering distally giving rise to 2 or more tubules, typically open; (ii) a single long tubular antapical process and (iii) cingular and sulcal processes, of moderate length, slender and tubular, only joining proximally if at all. Archaeopyle apical. HOLOTYPE. Slide AH 72, Laboratoire de Micropaleontologie, Ecole Practique des Hautes Etudes, Paris. (Figured by Deflandre 1937^, pi. 14, fig. 3). Senonian flint from the Paris Basin. DIMENSIONS. Holotype: length of shell 46 p, width 36/4, overall length 78^, length of processes 15-17 //.. Range of Cenomanian specimens : diameter of central body 39 (46-1) 56 /z, maximum length of processes 27 (30-2) 36/11. Number of specimens measured, 13. DESCRIPTION. H. ferox appears to be a fairly variable species. In some speci- mens the tabulation is not clearly indicated by the processes which tend to be smaller and may even be closed distally. However, the processes usually are well developed, their bases covering an area of the shell surface which is often slightly raised and of the same shape as a thecal plate. The precingular and postcingular processes are largest and give rise distally to as many as 10 tubules. Each cingular process divides proximally into 2 or 3 long slender tubules aligned along the cingulum. The sulcal processes may proximally divide into two tubules or may consist of a single tubule which is sometimes reduced and closed distally. The antapical process is long, tubular, open distally, and terminates with a smooth or serrate margin. 160 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i REMARKS. The Cenomanian specimens closely resemble the holotype of H . ferox, which was examined in Paris by the author, by kind permission of Professor Deflandre. The processes of the holotype are of the same form as, and similarly positioned to, those of the Cenomanian forms. The antapical process of the holotype, not shown in Deflandre's illustration (1937, pi. 14, fig. 3), is long and tubular. OCCURRENCE. H . ferox is rare to very rare at most horizons throughout the Ceno- manian of Fetcham Mill and Escalles, but has not been observed at Compton Bay or in the North American samples. It is also present in the Turonian sample (sample FM 520) from Fetcham Mill. Hence the stratigraphic range is from Aptian (Eisenack 1958) to Upper Cretaceous, probably Senonian (Deflandre 1937). Genus PROLIXOSPHAERIDIUM Davey, Downie, Sarjeant & Williams 1966 DIAGNOSIS. Shell elongate ovoidal to ellipsoidal, one pole (apical) typically lost in archaeopyle formation. Opposite pole occupied by one or two antapical processes. Remaining processes arranged in distinct rows, encircling shell and slightly offset at a position corresponding to sulcus. Number of processes exceeds 30. Processes closed proximally, typically but not constantly closed distally; their distal termina- tions simple, faring in varied fashion, or briefly furcate. Shell surface sometimes bears cover of coarse granules or very short, simple spinelets. REMARKS. The diagnosis has been changed slightly: ' typically but not constantly closed distally ' being inserted in place of ' closed or open distally ', with respect to the processes. The processes of Prolixosphaeridium usually taper distally and are closed. Specimens with open tubular processes belong to Tanyosphaeridium Davey & Williams (19666). Prolixosphaeridium conulum sp. nov. (PI. 8, figs. 5, 6) DERIVATION OF NAME. Latin, conulus, cone with reference to the rather conical shape of the processes. DIAGNOSIS. Shell elongate ovoidal; shell wall densely granular and bearing moderate number of processes. Processes acuminate to subconical, smooth walled, typically rigid, pointed distally. Processes tend to be aligned in circular manner around shell. Archaeopyle apical. HOLOTYPE. B.M. (N.H.) .51981 (5). Lower Chalk, Bureau de Recherches Geo- logiques et Minieres Borehole, Escalles, Pas de Calais, at 165 metres depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Holotype : shell length 47 //,, width 27 /u, length of processes 11-16 /*. Range : shell length 38 (43-8) 50 \i, width 20 (25-9) 29 //,, maximum length of processes ii (15-1) i8/i. Number of specimens measured, 8. DESCRIPTION. The length of the shell is slightly less than twice the width. The surface granules are relatively large (c. 0-2-0-4 /j, in width and height); they are CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 161 equidistant from one another. The processes (45-60 in number) have broad bases (up to 6 fj,) and taper distally to terminate in a point. They are approximately half the shell width in length, hollow and typically rigid, only occasionally being bent near the distal end. In the central region of the shell the processes are aligned, the rows encircling the shell. However, in one longitudinal portion in this region the processes are generally smaller and haphazard in arrangement. This region probably corresponds to the sulcus. REMARKS. The distinctive shape, number and size of the processes distinguish P. conulum. A rather similar but longer form was illustrated by Cookson & Eisenack (1958, pi. 8, fig. IT ) as Hystrichosphaeridium parvispinum Deflandre. This specimen comes from the Aptian of Australia and is said to grade into forms possessing more numerous processes similar to P. conulum. Deflandre's species was placed in Prolixosphaeridium by Davey, Downie, Sarjeant & Williams (1966). Two other species are rather similar. The holotype of P. granulosum (Deflandre) measures 18 by 33 /* and possesses 20-30 long processes, the latter being approxi- mately equal to the shell width. The number of processes and the relative lengths of the processes, therefore, differentiate this species from P. conulum. P. granulosum as described by Valensi (1955) from the Upper Cretaceous, and Sarjeant (1962) from the Upper Jurassic, are more similar to P. conulum but do not possess conical processes. P. mixtispinosum (Klement) differs from P. conulum by possessing two kinds of processes (i) approximately 50 processes of moderate length and (ii) numerous fine hairs covering the shell surface. OCCURRENCE. P. conulum has only been observed at certain horizons in the Upper Cenomanian of Fetcham Mill, Compton Bay and Escalles in samples FM 690, CB 19, E 165, E 159 and E 153. Genus CORONIFERA Cookson & Eisenack emend. EMENDED DIAGNOSIS. Shell subspherical to ovoidal, bearing numerous simple or bifurcating processes. Processes solid or hollow, closed distally, and joined proxi- mally by low crests or membranes. Apical process distinctive, simple or branched. Antapical process large, tubular, often terminating with denticulate margin. Archae- opyle apical. TYPE SPECIES. Coronifera oceanica Cookson & Eisenack 1958. Lower Cretaceous (Albian) ; Australia. REMARKS. The diagnosis has been emended to include the presence of an apical archaeopyle and low crests joining the processes, and the positions of the two distinctive processes. Coronifera differs from Diphyes Cookson (1965) in that the processes are never tubular and open distally, and by the presence of a reticulum joining the basal portion of the processes. However, both genera possess a large tubular antapical process and an apical archaeopyle, and are probably closely related. 162 CENOMANIAN NON -C ALC AREOU S MICROPL ANKTON, i Coronifera oceanica Cookson & Eisenack (PI. 8, figs. 8, ii) 1958 Coronifera oceanica Cookson & Eisenack : 45, pi. 12, figs. 5, 6. 1958 Coronifera oceanica Cookson & Eisenack; Eisenack : 407, pi. 25, fig. i. 1964 Coronifera oceanica Cookson & Eisenack; Cookson & Hughes : 56, pi. 9, figs. 8, 9. 1967 Coronifera oceanica Cookson & Eisenack; Clarke & Verdier: 77, pi. 17, fig. 7. DESCRIPTION . The shell is subspherical to ovoidal, thin-walled, and bears numerous processes of length between one-quarter and one-third of the shell diameter. There is, occasionally, a slight apical prominence beneath the apical process. Fine fibres radiate from the bases of the processes over the shell surface reminiscent of those on the shell surface of Exochosphaeridium striolatum (Deflandre). The pro- cesses are weak, fairly flexuous and commonly joined to each other by a network of low crests or fine membranes. The latter may be proximal or may extend along the entire length of the processes. Distally the processes are closed and may be simple, bifurcate or trifurcate. A large tubular process is present at the antapex; it is open distally and terminates with a denticulate margin. At the apex, when attached, there is a process which is only slightly larger than the typical processes but is usually branched and, therefore, distinctive. The large archaeoplye, developed in the majority of specimens, has an angular margin and forms opposite the antapical process. A number of specimens of C. oceanica were observed in the Albian sample from Fetcham Mill (FM 886). They resemble the specimens of Eisenack (1958) from the Aptian of Germany in that the processes are fewer, more solid and are usually simple. DIMENSIONS. Range of observed specimens: diameter of central body 31 (40) 54 /x, maximum length of processes 11 (15-3) 22 /z. Number of specimens measured, 15. REMARKS. C. oceanica has previously been recorded from the Upper Aptian of Germany (Eisenack, 1958), Albian of Australia (Cookson & Eisenack, 1958) and basal Cenomanian of England (Cookson & Hughes, 1964). The surface reticulation was not described in the Australian type material but appears to be present on the photographed specimens. It was first commented on by Cookson & Hughes. OCCURRENCE. In addition to the German and Australian records, C. oceanica is infrequent to common at all horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles, and is also common in the Albian sample (FM 886) from Fetcham Mill. It was not present in the Turonian sample (FM 520). One specimen was observed in the North American material, in sample Sas 1084 (Albian) from Saskatchewan. Thus the known stratigraphic range is from Upper Aptian to Cenomanian. Cyst-Family EXOCHOSPHAERIDIACEAE Sarjeant & Downie 1966 Genus EXOCHOSPHAERIDIUM Davey, Downie, Sarjeant & Williams 1966 REMARKS. Exochosphaeridium differs from Trichodinium Eisenack & Cookson (1960) in that the latter possesses a well developed cingulum. CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 163 Exochosphaeridium phragmites Davey, Downie, Sarjeant & Williams (PI. 7> %. 5) 1966 Exochosphaeridium phragmites Davey, Downie, Sarjeant & Williams: 165, pi. 2, figs. 8-10. DIMENSIONS. Range of observed specimens: maximum diameter of central body 41 (51-9) 67 IJL, minimum diameter of central body 32 (46-7) 57 //,, maximum length of processes 10 (18-5) 40 p,. Number of specimens measured, 18. REMARKS. Superficially E. phragmites resembles E. striolatum (Deflandre) which, however, has a definitely striated periphragm. Trichodinium paucispinum Eisenack & Cookson (1960) is also similar but has fewer processes and a well developed cingulum. OCCURRENCE. E. phragmites is rare at most horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles, and is recorded from the Albian sample, FM 886. It has not been recorded in the North American material. Exochosphaeridium pseudohystrichodinium (Deflandre) (PI. ii, figs. 4, 5) 1937 Hystrichosphaeridium pseudohystrichodinium Deflandre : 73, pi. 15, figs. 3, 4. 1966 ? Exochosphaeridium pseudohystrichodinium (Deflandre) ; Davey, Downie, Sarjeant & Williams : 166. 1967 Baltisphaeridium pseudohystrichodinium (Deflandre); Clarke & Verdier: 75, pi. 15, fig. 7. (see also for earlier references). EMENDED DIAGNOSIS. Shell spherical to ovoidal; shell wall thick, with pitted surface. Processes numerous, slightly fibrous, occasionally bifurcating medially, slender, broadening slightly proximally ; distally truncated or terminated with small bifurcation. Cingular processes rarely aligned. Apical process sometimes branched and slightly larger than normal. Archaeopyle precingular, formed by loss of one or two plate areas. HOLOTYPE. Slide AH. 55, Laboratoire de Micropaleontologie, Ecole Practique des Hautes Etudes, Paris. (Figured by Deflandre 1937, pi. 15, fig. 3). Upper Cretaceous flint from the Paris Basin. DIMENSIONS. Range of type material: shell length 49 to 54/11, shell width 38 to 45 /i, overall length 80-90 ju,. Range of observed specimens: diameter of central body 35 (43-8) 54 p, maximum length of processes 13 (16-8) 21 /u,. Number of specimens measured, 12. DESCRIPTION. The shell wall is moderately thick (c. 1-5 ^), pitted (never striated), and bears a large number of broad-based processes. All the processes appear to be basically the same except for the apical process which is typically branched and is usually larger than normal. The author was permitted, through the courtesy of Professor Deflandre, to make a detailed examination of the holotype and paratype. In the holotype the cingular processes were arranged in a definite circular manner around the shell. Such an 164 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i alignment was not observed with certainty in any of the Cenomanian specimens studied. The archaeopyle, present in the paratype, is precingular and usually formed by the loss of two plate areas. REMARKS. The Cenomanian specimens differ from the holotype only in that aligned circular processes were not observed. This may be due to unfavourable preservation or orientation of the specimens. The diagnosis of E. pseudohystricho- dinium has been emended to include a description of the apical process and the archaeopyle. The overall form of E. pseudohystrichodinium resembles species included in the " hirsutum " group, but the typical fibrous shell periphragm is absent. The forma- tion of the archaeopyle by the loss of two precingular plates has also been observed in E. striolatum var. truncatum nov., indicating a relationship. OCCURRENCE. This species is rather restricted, being common in samples FM 690, E 165 and E 159 and rare in samples FM 750, E 207 and FM 520 (Turonian). It appears, therefore, to be most common in the Upper Cenomanian. It was not recorded in the samples from North America. The recorded stratigraphic range is from Cenomanian to Eocene (Pastiels 1948). Exochosphaeridium striolatum (Deflandre) comb. nov. 1937 Hystrichosphaeridium striolatum Deflandre : 72, pi. 15, figs, i, 2. DIAGNOSIS. Shell subspherical to ovoidal. Processes numerous, variable, fibrous, often bifurcate medially and sometimes terminating with small fork. Pro- cesses distally may be pointed or blunted. Fibres pass down length of processes onto shell surface and there join with similar fibres from adjacent processes. Apical process and precingular archaeopyle may be present. HOLOTYPE. Slide AH. 89, Laboratoire de Micropaleontologie, Ecole Practique des Hautes Etudes, Paris. (Figured by Deflandre 1937, pi. 15, fig. i). Upper Cretaceous flint from the Paris Basin. REMARKS. Through the courtesy of Professor Deflandre, I was able to examine the holotype and paratype of E. striolatum. On the holotype one process appeared to be unusually thick and may have been apical in position. Neither holotype nor paratype was observed to possess an archaeopyle but this may have been due to the fact that the lower surfaces of the specimens were extremely dark and could not be studied. Exochosphaeridium striolatum (Deflandre) var. truncatum nov. (PI. 7, figs. 1-3) DERIVATION OF NAME. Latin, truncatus, shorten by cutting off with reference to the truncated extremities of the processes. DIAGNOSIS. A variety of E. striolatum possessing subspherical to ovoidal shell; shell wall fibrous or lightly pitted. Processes numerous, typically fibrous and blunted CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 165 distally, slender or subtriangular, rarely branched. Apical process large, often foliate. Precingular archaeopyle, commonly present, formed by loss of two plate areas, rarely one. HOLOTYPE. B.M. (N.H.) .51982 (i). Lower Chalk, Bureau de Recherches Geo- logiques et Minieres Borehole, Escalles, Pas de Calais, at 159 metres depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Holotype : diameter of central body 66 by 67 /z, length of processes 17-22^. Range: diameter of central body 34 (56-1) 81 ^, maximum length of processes 6 (17-8) 27 p. Number of specimens measured, 24. DESCRIPTION. The shell possesses a moderately thick wall (c. i /*) but it is quite often distorted, especially when an archaeopyle is developed. The shell surface (peri- phragm) is typically fibrous; the fibres pass down the length of the processes onto the shell surface and, there, join up with similar fibres from adjacent processes. Some specimens are less conspicuously fibrous, the fibres being apparent near the bases of the processes and only extending a little way onto the shell surface. The remainder of the shell surface in these forms is lightly pitted. The processes may be fairly slender to subtriangular and are occasionally joined proximally. A small number of processes are subdivided medially. The processes are typically trun- cated distally but may be slightly bulbous. Process alignment was not observed. The apical process is larger than the other processes and often foliate; the endo- phragm occasionally forms a small apical bulge beneath it. The archaeopyle is typically formed by the removal of two precingular plates, as is apparent by its shape. Rarely only one plate is lost. Detached opercula consisting of two precingular plates have been located (pi. 7, fig. 3). REMARKS. E. striolatum var. truncatum nov. is a member of the " hirsutum " group, a group which is characterized by having fibres radiating from the bases of the processes over the shell surface. The truncate nature of the processes, although obvious, has not previously been remarked on. Exochosphaeridium spinosum var. deflandrei (Lejeuner-Carpentier 1941) is similar, the processes sometimes being bulbous distally, but they are very slender and appear never to be truncated. Two other forms, E. striolatum (Deflandre) (illustrated by Lejeune-Carpentier 1941 as Hystrichosphaeridium hirsutum text-figs. 1-4) and E. (Hystrichosphaeridium) cf. hirsutum (Cookson & Eisenack 1958), are also comparable, differing in that their processes are not truncated. Baltisphaeridium bifidum Clarke & Verdier (1967) is similar but possesses fewer and more slender processes which, however, are bifid distally. The above forms all belong to the " hirsutum " group and are rather similar, differing only in the detail form of the processes and perhaps in archaeopyle formation, which is noted here for the first time. Although the precingular archaeopyle is typically, but not always, formed by the loss of two plate areas, this is not considered reason enough for the erection of a new genus. This species may indicate a trend towards the genus Lingulodinium Wall (1967) where four or five precingular plate areas are lost. OCCURRENCE. E. striolatum var. truncatum is infrequent to fairly common at all horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. 166 CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i At Fetcham Mill it has also been recorded from the Albian (sample FM 886) and Turonian (sample FM 520). A single specimen was obtained from the Upper Wood- bine of Texas, but the species was absent from the Saskatchewan samples. OTHER SPECIES The following species and variety are here included in Exochosphaeridium on the basis of similarity in structure : Exochosphaeridium spinosum (White 1842) comb. nov. 1842 Xanthidium spinosum White: Microsc. J., 11, 35-40, pi. 4 fig. 6. Exochosphaeridium spinosum var deflandrei (Lejeune-Carpentier 1941) comb. nov. 1941 Hystrichos- phaeridium spinosum var. defiandrei Lejeune-Carpentier: Annls. Soc. geol. Belg., 63 (bull, 3), 684, figs. 6, 7. Cyst-Family AREOLIGERAGEAE Evitt emend. Sarjeant & Downie 1966 Genus CYCLONEPHELIUM Deflandre & Cookson emend. Williams & Downie 1966 REMARKS. Four species are described from the Cenomanian, C. distinctum Cook- son & Eisenack being the only one fairly common throughout this stage. C. mem- braniphorum Cookson & Eisenack, C. vannophorum sp. nov. and C. paucispinum sp. nov. are restricted in occurrence. One species C. eisenacki sp. nov., is described from the Albian of Saskatchewan. In the Cenomanian, the genus is most abundant at Escalles, rarest at Saskatchewan and absent from Texas, possibly indicating a preference for an open water environment. Cyclonephelium distinctum Deflandre & Cookson (PI. n, figs. 6-8, 10 ; Figs. 16 C, D, F) 1955 Cyclonephelium distinctum Deflandre & Cookson : 285, pi. 2, fig. 14. 1963 Circulodinium deflandrei Alberti, Baltes: 587, pi. 6, figs. 9-11. 1967 Cyclonephelium distinctum Deflandre & Cookson; Clarke & Verdier: 22, pi. i, figs. 6, 7. (see also for earlier references) . DESCRIPTION. This species is extremely variable, as pointed out by Cookson & Eisenack (1962), and a member of most Cenomanian assemblages studied. The shell, which is always somewhat flattened, may be subcircular to ovoidal in outline and has a smooth or lightly granular shell wall. The regular outline is sometimes broken by an apical protuberance and more rarely by two reduced, antapical horns. The latter are of unequal size and when present the antapical region is slightly con- cave (Fig. i6F). The bald areas, typical of this genus, may occupy almost all of the ventral and dorsal surfaces of the shell leaving only the peripheral regions to bear processes, or may be practically invisible beneath the encroaching processes. Usually, however, these bald areas are circular to ovoidal in shape and occupy approximately thirty per cent of each side of the shell. The processes are usually abundant, extremely variable in form and typically under one-third of the shell width in length. They are solid and usually widen distally and proximally, may be fine or taeniate, and are often distally bifurcate. The bases of the processes are CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 167 occasionally thickened and rarely a line of thickening on the shell surface joins one process with a neighbouring one. Equally rarely, and only when the processes are broad, they may be joined distally. A large apical archaeopyle is always present. This is angular, possesses a pro- nounced zigzag margin and on the ventral surface usually a relatively deep sulcal notch. Although the archaeopyle is always discernible, the operculum is sometimes still attached. It then behaved as a lid, returning to its original position after the escape of the encysted organism. DIMENSIONS. Range of observed specimens: height of shell (operculum absent) 41 (54-5) 82 ju, width 48 (61-8) 81 \L, maximum length of the processes 4 (10-7) 21 /z. Number of specimens measured, 50. REMARKS. C. distinctum appears to be extremely variable, and specimens showing extremes in variation sometimes appear to be transitional to other species. OCCURRENCE. Found throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles, and rare to common at all horizons. Two samples from Sas- katchewan, samples Sas 890 and Sas 805, also contain C. distinctum. The species had a wide geographical distribution throughout most of the Cretaceous and is of little value in detailed stratigraphy. Cyclonephelium membraniphorum Cookson & Eisenack (PL ii, fig. 9) 1958 Cyclonephelium compactum Deflandre & Cookson : 48, pi. 12, fig. 8 only. 19626 Cyclonephelium membraniphorum Cookson & Eisenack : 495, pi. 6, figs. 8-14. 1964 Cyclonephelium membraniphorum Cookson & Eisenack; Cookson & Hughes : 44, pi. 10, figs. 5, 6. 1967 Cyclonephelium membraniphomm Cookson & Eisenack; Clarke & Verdier: 23, pi. 2, figs, i, 2. DESCRIPTION. The shell is always dorso-ventrally flattened, both the apex and antapex being circular to subcircular in outline. An apical archaeopyle is always developed, with a zigzag margin and sulcal notch. The bald areas, typical of this genus, may be relatively large or quite small and are surrounded by high membranes. The latter are braced by stout supports which arise from lines of thickening on the shell surface. These lines of thickened periphragm may be curved and may form semi-circular rings. When the latter are present, the membranes are in the form of wide tubular projections. The periphragm of the shell wall and the membranes is granular or pitted. The membranes may sometimes be fenestrate. In some specimens attributed to C. membraniphorum the bald areas are practically non-existent. The membranes in these forms are particularly well developed and commonly are in the form of wide tubes or funnels. Distally they have a sub- circular to polygonal outline and may possess a thickened outer rim bearing irregular pieces of periphragm (PI. n, fig. 9). DIMENSIONS. Range of observed specimens: length of shell (operculum missing) 37 (50-6) 79 ft, width 41 (54-1) 82 /u, maximum height of membranes 6 (13-3) 26^. Number of specimens measured, 16. GEOL. 17, 3 12 168 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i REMARKS. Many of the specimens observed appear identical with the Albian to Cenomanian forms from Australia, except that the former are considerably smaller, the largest of them only falling within the size range of the type material. It is not thought practical to separate the European forms solely on the size distinction and they have, therefore, been placed in the Australian species. An interesting variation is the apparent absence of the bald areas, although the overall shape and the apical archaeopyle typical of Cyclonephelium are present. OCCURRENCE. In addition to the Australian records, C. membraniphorum is rare and spasmodic throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles, though absent elsewhere. This species has also been recorded from the Turonian and Lower Senonian by Clarke & Verdier (1967). Cyclonephelium vannophorum sp. nov. (PL 9, fig. 3: PI. ii, figs, n, 12; Fig. i6E) DERIVATION OF NAME. Latin, vannophorum, fan-bearer with reference to the fan-shaped processes. DIAGNOSIS. Shell subcircular in outline, with slight apical prominence and one or two reduced antapical horns. Shell wall coarsely granular with areas devoid of processes of moderate size. Processes numerous, short, solid, of irregular shape and often confluent distally. Apical archaeopyle with zigzag margin. HOLOTYPE. B.M. (N.H.) slide .51986 (i). Lower Chalk, Compton Bay, Isle of Wight at 15 feet 6 inches above the base of the Cenomanian. Upper Creatceous (Cenomanian) . DIMENSIONS. Holotype: length of shell (including operculum) 65 /*, width 62 /u, length of processes up to 8//,. Range: length of shell (without operculum) 47 (57-8) 70 /*, width 56 (65-5) 78 p, maximum length of processes 3 (5-3) 8/u,. Number of specimens measured, 11. DESCRIPTION. When two antapical horns are present the portion of the shell between them is concave and from this region a broad furrow passes towards the apex. The furrow or sulcus decreases in width and depth in this direction and disappears just posterior to the archaeopyle margin. The processes vary greatly in size from mere enlarged granules (0-5 ju,) to 8 p, in length. In the larger processes the stem is quite narrow, the distal third widening rapidly (PI. 9, fig. 3) and is some- times bifurcate. The processes are often joined to form a short line on the shell surface. Rarely the cingulum is just discernible by a concentration of small processes along its borders. REMARKS. C. vannophorum sp. nov. is most closely comparable to ?C. attadalicum Cookson & Eisenack (19626) from the Aptian /Albian of Australia. The processes are similar in form but the shell of ?C. attadalicum is more polygonal and the cingulum is always well defined. OCCURRENCE . C . vannophorum has only been recorded from the Lower Cenomanian sample CB 3 (Compton Bay) and it there comprises about 2% of the micro- CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 169 E F FIG. 16. Cyclonephelium eisenacki sp. nov., A. Complete specimen (x 700). B. Specimen illustrating apical archaeopyle ( x 700). Cyclonephelium distinctum Deflandre & Cookson, C. D. and F. Illustrating variation in the shape of the shell of this species (processes re- moved) (X 700). Cyclonephelium vannophorumsp. nov., E. Specimen with well developed apical archaeopyle (processes removed) (X 700). iyo CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i plankton content. The unusual distribution and the similarity to a Lower Cretaceous species indicate that C. vannophorum is possibly a derived form. Cyclonephelium paucispinum sp. nov. (PI. 9, figs, i, 2) DERIVATION OF NAME. Latin, paucus, few; spina, thorn with reference to the scarcity of the processes. DIAGNOSIS. Shell subpolygonal in outline with well defined antapical horn. Shell wall lightly to coarsely granular, bearing small number of irregularly distri- buted processes confined to peripheral region of shell. Processes of moderate size, solid and widening distally. Apical archaeopyle with a zigzag margin and sulcal notch. HOLOTYPE. B.M. (N.H.) .51981 (2). Lower Chalk, Bureau de Recherches Geo- logiques et Minieres Borehole, Escalles, Pas de Calais at 165 metres depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Holotype : length of shell (operculum missing) 54 //,, width 81 p, length of processes 1-5-12 /x. Range: length of shell (operculum missing) 40 (72-4) 92 /z, width 54 (87-0) 112 /A, maximum length of processes 5 (10-2) 19 p. Number of specimens measured, 12. DESCRIPTION. The shell is subpolygonal in outline, the cingular region being the widest portion of the shell. Posteriorly there is a pointed antapical horn. The few processes present are of moderate size and if close together tend to anastomose both distally and proximally. REMARKS. The large size, the paucity of processes and the subpolygonal shape of the shell distinguish C. paucispinum sp. nov. from all previously described species. The processes most closely resemble those of C. distinctum but are fewer in number. OCCURRENCE. C. paucispinum is rare in samples E 165, CB 3, and common in sample CB n. The distribution is, therefore, rather erratic in the Lower, Middle and Upper Cenomanian. Cyclonephelium eisenacki sp. nov. (PI. 8, figs. 3, 4; PL 9, fig. 4; Figs. I 7 A, B) 1960 Aptea cf. polymorpha Eisenack & Cookson : 9, PI. 3, figs. 2-4. DERIVATION OF NAME. The species is named after Professor A. Eisenack. DIAGNOSIS. Shell subtriangular, flattened, with convex sides. Apical horn well developed; antapical horns, if present, very reduced. Greater part of shell surface ornamented by complex network of low crests and short, capitate processes. In centre of both dorsal and ventral sides there is a circular area devoid of ornamenta- tion. Apical archaeopyle always present. CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i 171 HOLOTYPE. B.M. (N.H.) V. 51980(2). Upper Lower Colorado (Fish Scale Zone?), International Yarbo Borehole No. 17, Saskatchewan at 1023 feet depth. Lower Cretaceous (Albian). DIMENSIONS. Holotype: shell length 70 /z, width 66 p, height of crests 1-5 p. Range: shell length 70-92/11 (4 complete specimens measured); shell length (oper- culum missing) 55 (63-1) 71 /z, width 66 (72-1) 84 /z, height of crests 2 (4-2) 7 /z. Number of specimens measured, 12. DESCRIPTION. The shell surface is ornamented by a very characteristic network of low crests or lamellae which are often broken, so forming isolated processes. A poorly-marked cingulum, outlined by the crests, is occasionally discernible. The apical operculum is often in position. REMARKS. The illustrations of Aptea cf. polymorpha indicate that these speci- mens belong in the genus Cydonephelium and may well be conspecific with the Canadian specimens here described. This species differs from Aptea polymorpha Eisenack (1958) by the absence of the numerous fine processes and the outer mem- branous structure which they support. Pseudoceratium turneri Cookson & Eiesnack (1958), from the Aptian /Albian of Australia, possesses a similar but stronger orna- mentation. Also the apical and antapical horns are usually much better developed, although the specimen of P. turneri illustrated (loc. cit. pi. 5, fig. 5), approaches a number of the C. eisenacki examples. OCCURRENCE. C. eisenacki sp. nov. is fairly common in two samples from Saskatchewan, samples Sas 1084 and 1023. It has not been recorded elsewhere in the material examined. Cyst-Family ADNATOSPHAERIDIAGEAE Sarjeant & Downie 1966 Genus ADNATOSPHAERIDIUM Williams & Downie 1966 Adnatosphaeridium chonetum (Cookson & Eisenack) comb. nov. (PL 10, figS. II, 12.) 19626 ICannosphaeropsis choneta Cookson & Eisenack : 493, pi. 4, figs. 8-10. DESCRIPTION. The shell is spherical to subspherical and bears a number of more or less tubular processes which possess bands of thickening for support. The pro- cesses are usually joined to their neighbour either along their entire length, or only distally by membranes. However, isolated tubular processes are occasionally visible. The impression obtained from most specimens is of a complex membranous network perpendicular to the shell surface. A well developed apical archae- opyle is typically present. DIMENSIONS. Range of observed specimens: shell diameter 27 (34-1) 46^, length of processes 6 (10-4) 24 /z. Number of specimens measured, 15. REMARKS. The Cenomanian specimens examined strongly resemble, but are slightly smaller than, the type material from the Cenomanian of Australia. This 172 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i species, tentatively placed in Cannosphaeropsis by Cookson & Eisenack, is here transferred to Adnatosphaeridium because of the presence of interconnecting processes and an apical archaeopyle. OCCURRENCE. Apart from the Australian occurrence, it is infrequent to common in the Middle and Upper Cenomanian of Fetcham Mill and Escalles, not occurring below samples FM 790 and E 195. Only one specimen was recorded from Compton Bay, in sample CB 19. Cyst-Family HYSTRICHOSPAERAGEAE O. Wetzel emend. Sarjeant & Downie 1966 Genus HYSTRICHOSPHAERA O. Wetzel emend. Davey & Williams 1966 Hystrichosphaera ramosa var. ramosa (Ehrenberg) (PI. 10, figS. I, 2, 5) 1838 Xanthidium ramosum Ehr. : pi. i, figs, i, 2, 5. 19660 Hystrichosphaera ramosa (Ehr.) var. ramosa Davey & Williams : 33, pi. i, figs, i, 6; pi. 3, fig. i; text-fig. 8. (See also for earlier references). 1967 Hystrichosphaera furcata (Ehr.) Clarke & Verdier: 48, pi. 8, figs. 12, 13. DIMENSIONS. Range of observed specimens: diameter of central body 29 (40-7) 56 fi, maximum length of processes 13 (19-8) 27 jii. Number of specimens measured, 32. OCCURRENCE. This variety has a known stratigraphic range from the Oxfordian (Jurassic) to the Ypresian (Eocene). It is a common variety throughout the Ceno- manian of Fetcham Mill, Compton Bay and Escalles, and is rare to infrequent in the material from Saskatchewan and Texas. Hystrichosphaera ramosa (Ehrenberg) var. gracilis Davey & Williams 1955 Hystrichosphaera ramosa (Ehr.) Deflandre & Cookson : 263, pi. 5, fig. 8. 1963 Hystrichosphaera ramosa (Ehr.) Gorka : 48, pi. 6, figs. 6, 7. 1966^ Hystrichosphaera ramosa (Ehr.) var. gracilis Davey & Williams : 34, pi. i, fig. 5; pi. 5, fig. 6. DIMENSIONS. Range of observed specimens: diameter of central body 28-35 //,, maximum length of processes 16-22 /u,. Number of specimens measured, 5. OCCURRENCE. The known stratigraphic range of this variety is from the Ceno- manian (England) to the Miocene (Australia). It is rare to infrequent at most hori- zons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles, but has not been recorded in the North American material. CENOMANIAN NON-CALCAREOUS MICROPLAN KTON, i 173 Hystrichosphaera ramosa (Ehrenberg) var. multibrevis Davey & Williams (PL 10, figs. 3, 4) I 955 Hystrichosphaera furcata (Ehr.) Valensi : 586, pi. 4, fig. 4; pi. 5, fig. 12. 1958 Hystrichosphaera furcata (Ehr.) Eisenack : 406, pi. 25, figs. 4-8. 19660 Hystrichosphaera ramosa (Ehr.) var. multibrevis Davey & Williams : 35, pi. i, fig. 4; pi. 4, fig. 6; text-fig. 9. DIMENSIONS. Range of observed specimens : diameter of central body 31 (39-7) 51 n, maximum length of processes 10 (14-3) 21 p. Number of specimens measured, 13. OCCURRENCE. H. ramosa var. multibrevis has a stratigraphic range from the Lower Cretaceous (Hauterivian) to the Eocene (Ypresian). It is rare to infrequent in all samples examined from Fetcham Mill, Compton Bay and Escalles. It has also been recorded from the Lower Cretaceous (Albian) of Saskatchewan, in samples Sas 967, 1023 and 1084. Hystrichosphaera ramosa (Ehrenberg) var. reticulata Davey & Williams 19660 Hystrichosphaera ramosa (Ehr.) var. reticulata. Davey & Williams : 38, pi. i, fig. 2. DIMENSIONS. Range of observed specimens: diameter of central body 33 (45-2) 59 p, maximum length of processes 13 (15-3) 17 //,. Number of specimens measured, 13- OCCURRENCE. This variety is rare to infrequent at most horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It has also been recorded in two Albian samples from Saskatchewan, samples Sas 967 and 1084. Hystrichosphaera cingulata var. cingulata (O. Wetzel) J 933 Cymatiosphaera cingulata O. Wetzel : 28, pi. 4, fig. 10. 19660 Hystrichosphaera cingulata (O. Wetzel) Davey & Williams : 38, pi. i, fig. 9. (See also for earlier references). 1967 Hystrichosphaera cingulata var. cingulata (O. Wetzel) Clarke & Verdier: 45, pi. 8, figs. 9, 10. DIMENSIONS. Range of observed specimens: diameter of central body 26 (36-8) 48 ju, maximum height of crests 5 (7-0) 13 /z. Number of specimens measured, 16. OCCURRENCE. H. cingulata is infrequent at all horizons throughout the Ceno- manian of Fetcham Mill, Compton Bay and Escalles. It has not been observed in the North American material. The recorded stratigraphic range is from Albian to Middle Miocene. Hystrichosphaera cingulata (O. Wetzel) var. reticulata Davey & Williams 19660 Hystrichosphaera cingulata var. reticulata Davey & Williams : 39, pi. i, fig. 10; pi. 2, fig- 4- 1967 Hystrichosphaera cingulata var. perforata Clarke & Verdier : 46, pi. 9, figs. 2-4, text-fig. 19. DIMENSIONS. Range of observed specimens: diameter of central body 33 (42-8) 59 ft, maximum height of crests 7 (10-2) 17 /u.. Number of specimens measured, 14. 174 CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i OCCURRENCE. H. cingulata var. reticulata is rare to infrequent throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It appears to be slightly more common in the Middle and Upper Cenomanian at these localities. It has not been observed in the North American samples. Clarke & Verdier (1967) also record this variety from the Turonian and Senonian. Hystrichosphaera crassimurata Davey & Williams 19663 Hystrichosphaera crassimurata Davey & Williams : 39, pi. i, fig. u. 1967 Hystrichosphaera cingulata var. polygonalis Clarke & Verdier: 47, pi. 8 figs. 7, 8, text-fig. 20. DIMENSIONS. Range of observed specimens: diameter of central body 36-46 /z, height of crests, up to 14 jit. Number of specimens measured, 4. OCCURRENCE. H. crassimurata is very rare to rare in the Middle and Upper Cenomanian of Fetcham Mill, Compton Bay and Escalles, the lowest samples in which it is found being FM 790, E 195 and CB 9. It has also been recorded from the Upper Woodbine Formation of Texas but is absent from the Saskatchewan material. Hystrichosphaera crassipellis Deflandre & Cookson 1954 Hystrichosphaera crassipellis Deflandre & Cookson: text-fig. 5. ig66a Hystrichosphaera crassipellis Deflandre & Cookson; Davey & Williams : 40, pi. i, figs. 7, 8. (See also for earlier references). 1967 Hystrichosphaera crassipellis Deflandre & Cookson; Clarke & Verdier: 48, pi. 8, fig. n; pi. 9, fig. i. DIMENSIONS. Range of observed specimens: diameter of central body 34 (47-9) 68 p,, maximum length of processes 10 (17-8) 29 p. Number of specimens measured, 19. OCCURRENCE. H. crassipellis is rare at most horizons throughout the Cenomanian of Fetcham Mill, Compton Bay and Escalles; it has not been recorded in the North American material. The recorded stratigraphic range is from Cenomanian to Middle Miocene (Gerlach 1961). Genus ACHOMOSPHAERA Evitt 1963 Achomosphaera ramulifera (Deflandre) (PL 10, fig. 7) 1935 Hystrichosphaera cf. ramosa (Ehr.) Deflandre : pi. 5, fig. n. 1966^ Achomosphaera ramulifera (Deflandre) Davey & Williams : 49, pi. 2, fig. 3. (See also for earlier references) . 1967 Achomosphaera ramulifera (Deflandre); Clarke & Verdier: 40, pi. 8, fig. i. DIMENSIONS. Range of observed specimens: diameter of central body 36 (44-7) 61 /A, maximum length of processes 16 (25-8) 36 p. Number of specimens measured, 14. CENOMANIAN NON-CALCAREOUS MICROPL AN KTON, i 175 OCCURRENCE. This species has a known stratigraphic range from the Cenomanian to the Eocene (Pastiels 1948). It is a very rare to infrequent species at most horizons in the Cenomanian from Fetcham Mill, Compton Bay and Escalles. It is absent from the North American samples. Achomosphaera sagena Davey & Williams ig66a Achomosphaera sagena Davey & Williams : 31, pi. 2, figs, i, 2. 1967 Achomosphaera reticulata Clarke & Verdier: 41, pi. 8, figs. 2, 3, text-fig. 16. DIMENSIONS. Range of type material: diameter of central body 35 (48-4) 59^, maximum length of processes 17 (20-8) 28 ^. Number of specimens measured, 12. OCCURRENCE. A. sagena is a very rare to rare species occurring spasmodically in the Cenomanian samples from Fetcham Mill, Compton Bay and Escalles. Clarke & Verdier (1967) have also recorded it (as A . reticulata) from the Turonian and Senonian. Genus Hystrichodinium Deflandre emend. Clarke & Verdier 1967 Hystrichodinium voigti (Alberti) (PI. 10, figs. 6, 10) 1961 Heliodinium voigti Alberti : 33, pi. 8, figs. 1-5. 1966^ Heliodinium voigti Alberti; Sarjeant : 142, pi. 16, fig. 2; text-fig. 36. 1967 Hystrichodinium voigti (Alberti) Clarke & Verdier: 38. DESCRIPTION. The shell is ovoidal to subpolygonal. The periphragm is smooth or slightly granular, and forms sutural crests and ribbon-like processes. The former are variably developed, may occasionally be absent, but when present give rise to long, flexuous processes along their length. When the crests are absent the pro- cesses arise directly from the shell surface. The processes are not confined to the plate boundaries and may be distributed at random over the shell surface. The processes are thin-walled, typically simple and occasionally terminate with bifurcate or trifurcate extremities. The cingulum (4-5^ wide) is always delimited by low crests and is strongly helicoid. A precingular archaeopyle, formed by the loss of plate 3", is typically present. One detached operculum has been located and bears five processes. DIMENSIONS. Range of observed specimens: shell length 40 (49-6) 62 /*, width 40 (48-4) 58 ^, length of processes 27 (35-6) 48 //,. Number of specimens measured, 11. OCCURRENCE. H. voigti is rare to fairly common and found spasmodically through- out the Cenomanian of Fetcham Mill, Compton Bay and Escalles. It was originally described by Alberti (1961) from the Lower Barremian to ? Lower Aptian of Germany. The specimens described by Sarjeant (19660) are from the Cenomanian of Fetcham Mill. Hystrichodinium dasys sp. nov. (PI. 10, figs. 8, 9) DERIVATION OF NAME. Latin, dasys, hairy or shaggy with reference to the abundant hair-like processes. 176 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i DIAGNOSIS. Shell subspherical, thin-walled, smooth or slightly granular. Tabula- tion very faintly marked ; cingulum, and more rarely, precingular and postcingular plate boundaries visible. Processes numerous, short, fine and flexuous, aligned along plate boundaries, also intratabular. Archaeopyle not observed. HOLOTYPE. B.M. (N.H.) .51982 (3). Lower Chalk, Bureau de Recherches Geo- logiques et Minieres Borehole, Escalles, Pas de Calais at 159 metres depth. Upper Cretaceous (Cenomanian). DIMENSIONS. Holotype : diameter of central body 58 by 61 p, length of processes 10-14 p. Range: diameter of central body 42 (53-0) 70^, maximum length of processes 7 (10-6) 14 p. Number of specimens measured, 13. DESCRIPTION. The shell, being thin-walled, is always found deformed. Only two specimens possess a faint tabulation. This is best seen in the holotype which has a clear but lightly defined cingulum and two or three precingular and post- cingular plate boundaries visible. The numerous processes are hair-like, approxi- mately 0-2 11 wide for most of their length, probably hollow and terminate with a point. The most characteristic feature of the processes is their extreme flexibility. REMARKS. The characteristic processes and the presence of a tabulation differ- entiate H. dasys sp. nov. very easily from all previously described microplankton species. OCCURRENCE. This species is rare to common in all samples from Escalles except the lower three (E 201, E 207, and E 213). One specimen was recorded from Comp- ton Bay (sample CB 7) and two from Fetcham Mill (sample FM 520, Turonian). The species was not recorded from the North American material. The distribution of H. dasys possibly indicates an environmental difference between Escalles and Fetcham Mill/Compton Bay, perhaps in the depth of water or distance from land. IV. REFERENCES ALBERT:, G. 1959. Zur Kenntnis der Gattung Deflandrea Eisenack (Dinoflag.) in der Kreide und im Alttertiar Nord und Mitteldeutschlands. Mitt. geol. Stlnst. Hamb., Hamburg 28 : 93-105, pis. 8, 9. 1961. Zur Kenntnis mesozoischer und alttertiarer Dinoflagellaten und Hystrichos- phaerideen von Nord- und Mitteldeutschland sowie einigen anderen europaischen Gebieten. Palaeontographica, Cassel, Stuttgart, 116, A : 1-58, pis. 1-12. BALTES, N. 1963. Dinoflagellate si Hystrichosphaeride cretacice din Platforma moezica. Petrol Gaze, Bucuresti, 12 : 581-597, pis. 1-8. 1965. Observatii asupra microflorei cretacice inferiare din zona R. Bacaz. Petrol Gaze, Bucuresti, 16 : 3-17, pis. 3, 4. BARROIS, C. 1876. Recherches sur le terrain Cretac6 Superieur de 1'Angleterre et de 1'Irlande. Mem. Soc. geol. N., Lille, 3, 189-205. BRISTOW, H. W. 1889. The Geology of the Isle of Wight. Mem. geol. Surv. U.K., London. BROWN, C. A. 1960. Palynological Techniques, i-vi, 1-188 Baton Rouge, Louisiana. BROWN, R. W. 1956. Composition of Scientific Words. 882 pp. Reese Press, Baltimore, Md. CHATTON, E. 1952. Dinoflagelles, in Traite de Zoologie, I, 309-390. Masson, Paris. CLARKE, R. F. A. & VERDIER, J. P. 1967. An investigation of microplankton assemblages from the Chalk of the Isle of Wight, England. Verh. K. ned. Akad. Wet., Amsterdam, 24, 3 : 1-96, pis. 1-17. CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i 177 CLARKE, R. F. A., DAVEY, R. J., SARJEANT, W. A. S. & VERDIER, J. P. 1968. A Note on the Nomenclature of some Upper Cretaceous Dinoflagellate Taxa (in press) . COOKSON, I. C. 1956. Additional microplankton from Australian late Mesozoic and Tertiary sediments. Aust. J. mar. Freshwat. Res. Melbourne, 7, i : 183-191, pis. i, 2. 1965. Cretaceous and Tertiary Microplankton from South-Eastern Australia. Proc. R. Soc. Viet., Melbourne, 78, i : 85-93, pis. 9-11. COOKSON, I. C. & EISENACK, A. 1958. Microplankton from Australian and New Guinea Upper Mesozoic sediments. Proc. R. Soc. Viet., Melbourne, 70, i : 19-79, pis. 1-12. igdoa. Microplankton from Australian Cretaceous sediments. Micropaleontology , New York, 6, i : 1-18, pis. 1-3. - 19606. Upper Mesozoic microplankton from Australia and New Guinea. Palaeontology, London, 2, 2 : 243-261, pis. 37-39. - 1961. Upper Cretaceous microplankton from the Belfast No. 4 bore, South- Western Australia. Proc. R. Soc. Viet., Melbourne, 74, i : 69-76, pis. n, 12. - 19620. Some Cretaceous and Tertiary microfossils from Western Australia. Proc. R. Soc. Viet., Melbourne, 75, 2 : 269-273, pi. 37. 19626. Additional microplankton from Australian Cretaceous sediments. Micropalaeon- tology, New York, 8, 4 : 485-507, pis. 1-7. COOKSON, I. C. & HUGHES, N. F. 1964. Microplankton from the Cambridge Greensand (mid-Cretaceous). Palaeontology, London, 7, i : 37-59, pis. 5-11. COOKSON, I. C. & MANUM, S. 1964. On Deflandrea victoriensis n. sp., D. tripartita Cookson & Eisenack, and related species. Proc. R. Soc. Viet., Melbourne, 77, 2 : 521-524, pis. 76. DAVEY, R. J., DOWNIE, C., SARJEANT, W. A. S. & WILLIAMS, G. L. 1966. Fossil Dino- flagellate Cysts attributable to Baltisphaeridium, in Studies of Mesozoic and Cainozoic Dinoflagellate Cysts. Bull. Br. Mus. nat. Hist., London. Suppl. 3 : 157-175. DAVEY, R. J. & WILLIAMS, G. L. 196612. The Genera Hystrichosphaera and Achomosphaera, in Studies of Mesozoic and Cainozoic Dinoflagellate Cysts. Bull. Br. Mus. nat. Hist., London. Suppl. 3 : 28-52. 19666. The genus Hystrichosphaeridium and its allies, in Studies of Mesozoic and Cainozoic Dinoflagellate Cysts. Bull. Br. Mus. nat. Hist., London. Suppl. 3 : 53-105. DEFLANDRE, G. 1934. Sur les microfossiles d'origine planctonique conserves a 1'etat de matiere organique dans les silex de la craie. C.r. hebd. Seanc. Acad. Sci., Paris, 199 : 966-968. 1935. Considerations biologiques sur les micro-organismes d'origine planctonique conserves dans les silex de la craie. Bull. biol. Fr. Belg., Paris, 69 : 213-244, pis. 5-9. I936a. Les Flagelles fossiles. Aper9u biologique et paleontologique. R61e geologique. Actual, scient. ind., Paris, 335, 98 pp., 135 figs. - 19366. Microfossiles des silex cretaces, i. Generalites. Flagelles. Annls. Paleont., Paris, 25 : 151-191, pis. 11-20. 1937- Microfossiles des silex cretaces n. Flagell6s incertae sedis. Hystrichosphae- ridees. Sarcodines. Organismes divers. Annls. Paleont., Paris, 26 : 51-103, pis. 8-18. 1943. Sur quelques nouveaux Dinoflagelles des silex cretaces. Bull. Soc. geol. Fr., Paris, 5, 13 : 499-5 9, pl- 17- - 1947. Sur une novelle Hystrichosphere des silex cretaces et sur les affiniti6s du genre Cannosphaeropsis O. Wetzel. C.r. hebd. Seanc. Acad. Sci., Paris, 224 : 1574-1576, figs. 1-5. i952a. Protistes. Generalites. In Piveteau, J. (ed.) : Traite de paleontologie, 1, Paris: Masson. 782 pp. (pp. 116-124, n g s - 80-113; pp. 322-326, figs. 1-35 relevant). 19526. Dinoflagelles fossiles. In Grasse, P.P. (ed.) : Traite de Zoologie, vol. i Paris: Masson. 1071 pp. 830 figs. (pp. 391-406, figs. 300-339 relevant). DEFLANDRE, G. & COOKSON, I. C. 1954. Sur le microplankton fossile conserve dans diverses roches sedimentaires australiennes s'etageant du Cretace inferieur au Miocene superieur. C.r. hebd. Seanc. Acad. Sci., Paris, 239, 19 : 1235-1238, figs. 1-17. - 1955. Fossil microplankton from Australian late Mesozoic and Tertiary sediments. Aust. J. mar. Freshwat. Res., Melbourne, 6, 2 : 242-313, pis. 1-9. 178 CENOMANIAN NON-CALCAREOUS MICROPL ANKTON, i DEFLANDRE, G. & COURTEVILLE, M. 1939. Note preliminaire sur les microfossiles des silex cr6taces du Cambresis. Bull. Soc. fr. Microsc, Paris, 8 : 95-106, pis. 1-3. DESTOMBES, J. P. 1961. Deformations subies par les assises Cretacees dans la partie Septen- trionale du Boulonnais. Bull. Servs. Carte geol. Fr., Paris, 57, 261 : 1-5. DOWNIE, C., EVITT, W. R. & SARJEANT, W. A. S. 1963. Dinoflagellates, hystrichospheres and the classification of the acritarchs. Standford Stud. Geol., Palo Alto, 7, 3 : 1-16. DOWNIE, C. & SARJEANT, W. A. S. 1963. On the interpretation of some Hystrichosphere genera. Palaeontology, London, 6, i : 83-96. 1964. Bibliography and Index of Fossil Dinoflagellates and Acritarchs. Mem. geol. Soc. Am., Washington, 94. - 1966. The Morphology, Terminology and Classification of Fossil Dinoflagellate Cysts, in Studies of Mesozoic and Cainozoic Dinoflagellate Cysts, Bull. Br. Mus. nat. Hist., London. Suppl. 3 : 10-17. EHRENBERG, C. G. 1838. Uber das Massenverhaltnis der jetzt labenden Kieselinfusorien und iiber ein neues Infusorien-Conglomerat als Polierschiefer von Jastraba in Ungarn. Abh. Akad. Wiss. Berlin (1836), 1 : 109-135, pi. i. 1843. Uber einige Jura-Infusorien Arten des Corallrags bei Krakau. Mber. Akad. Wiss. Berlin, 61-63. 1854. Mikrogeologie. Leipzig. 486 pp, 41 pis. EISENACK, A. 1958. Mikroplankton aus dem norddeutschen Apt nebst einigen Bemerkungen uber fossile Dinoflagellaten. Neues Jb. Geol. Paldont., Abh., Stuttgart, 106, 3 : 383-422, pis. 21-27. - 1961. Einige Erorterungen uber fossile Dinoflagellaten nebst Ubersichtiiber die zur Zeit bekannten Gattungen. Neues Jb. Geol. Paldont., Abh., Stuttgart, 112, 3 : 281-324, pis. 33-37- EISENACK, A. & COOKSON, I. C. 1960. Microplankton from Australian Lower Cretaceous sediments. Proc. R. Soc. Viet., Melbourne, 72, i : i-u, pis. 1-3. EVITT, W. R. 1963. A discussion and proposals concerning fossil Dinoflagellates, Hystrichos- pheres and Acritarchs. Proc. natn. Acad. Sci. U.S.A. Washington, 49 : 158-164, 298-302, figs. 1-4. EVITT, W. R. & DAVIDSON, S. E. 1964. Dinoflagellate Studies, i. Dinoflagellate Cysts and Thecae. Stanford Stud. Geol., Palo Alto, 10, i : 1-12, pi. i. FIRTION, F. 1952. Le Cenomanien inf&rieur du Nouvion-en-Thierache : examen micropaleon- tologique. Annls. Soc. geol. N., Lille, 72 : 150-164, pis. 8-10. GERLACH, E. 1961. Mikrofossilien aus dem Oligozan und Miozan Nordwestdeutschlands, unter besonderer Berucksichtigung der Hystrichosphaerideen und Dinoflagellaten. Neues Jb. Geol. Paldont., Abh., Stuttgart, 112, 2 : 143-228, pis. 25-29. GIGNOUX, M. 1955 Stratigraphic Geology W. H. Freeman & Co. GOCHT, H. 1957. Microplankton aus dem nordwestdeutschen Neokom i . Paldont. Z., Berlin, 31, 3-4 : 163-185, pis. 18-20. - 1959. Microplankton aus dem nordwestdeutschen Neokom n. Paldont. Z., Berlin, 33, 1-2 : 50-89, pis. 3-8. G6RKA, H. 1963. Coccolithophorides, Dinoflagelles, Hystrichosphaerides et microfossiles incertae sedis du Cretac superieur de Pologne. A eta palaeont. polon., Warszawa, 8, i : 3-90, pi. i -i i. 1965. Les Microfossiles du Jurassique Superieur de Magnuszew (Pologne). A eta. palaeont. polon., Warszawa, 10, 3 : 291-334, pis. 1-5. GRAY, D. A. 1965. The Fetcham Mill (Surrey) Borehole. Bull. geol. Surv. Gr. Br., London, 23 : 65-116. HANCOCK, J. M. 1959. Les Ammonites du Cenomanian de la Sarthe, in Colloque sur le Cretace Superieur Francais. Comptes Rend. Soc. Sav.-Dijon 1959, 249-252. JEFFORDS, R. M. & JONES, D. H. 1959. Preparation of slides for Spores and other Micro- fossils. /. Paleont., Tulsa, 33, 2 : 344-347. CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i 179 JUKES-BROWNE, A. J. 1903. The Cretaceous Rocks of Britain. 2, The Lower and Middle Chalk of England. Mem. geol. Surv. U.K., London. KAYE, P. 1964. Observations on the Speeton Clay (Lower Cretaceous) . Geol. Mag., London, 101 : 34 -356- KLEMENT, K. W. 1960. Dinoflagellaten und Hystrichosphaerideen aus dem Unteren und Mittleren Malm Sudwestdeutschlands. Palaeontographica, Cassel, Stuttgart, 114, A : 1-104, pis. i-io. LEJEUNE-CARPENTIER, M. 1939. L'etude microscopique des silex (7 i6me Note). Un nouveau Peridinien cretacique, Gonyaulax wetzeli. Annls. Soc. geol. Belg., Liege, 62, 10-11: 6525- 529, figs. 1,2. 1941. L'etude microscopique des silex (9 i6me Note). Sur Hystrichosphaeridium hirsutum (Ehrenberg) et quelques formes voisines. Annls. Soc. geol. Belg., Li6ge, 63, 3 : 67192, figs. 1-9. 1943. L'etude microscopique des silex (n ieme Note). Une Hystrichosphaeridee a classer parmi les Peridiniens. Annls. Soc. geol. Belg., Liege, 67, i : 622-28, figs. 1-6. 1946. L'etude microscopique des silex (12 ieme Note). Especes nouvelles ou douteuses de Gonyaulax. Annls. Soc. geol. Belg., Li6ge 69, 4 : 6187-197, figs. 1-5. MALIAVKINA, V. S., SAMOILOVITCH, R. S., VOITZIEL, Z. A., KLUMKO, S. A., ROVUNA, L. V., IVAN- OVA, I. A., MARKOVA, L. G. & MTCHEDLISHVILI, N. D. 1961. Pollen and spores of Western Siberia; Jurassic-Paleocene; A symposium (in Russian). Trudy vses. neft. Nauchno-Issled. geol. raev Inst., Lenningrad, Moskva, 177 : 1-657, P^ 8 - 1-84. MANUM, S. & COOKSON, I. C. 1964. Cretaceous Microplankton in a sample from Graham Island, Arctic Canada, collected during the Second " Fram " Expedition (1898-1902). Skr. norske Vidensk-Akad., Oslo, 17 : 1-36, pis. 1-7. NEALE, J. W. & SARJEANT, W. A. S. 1962. Microplankton from the Speeton Clay of Yorkshire. Geol. Mag., London, 99, 5 : 439-458, pis. 19, 20. NOREM, W. L. 1953. Separation of Spores and Pollen from Siliceous Rocks. /. Paleont., Tulsa, 27, 6 : 881-883. NORRIS, G. & SARJEANT, W. A. S. 1965. A Descriptive Index of Genera of Fossil Dinophyceae and Acritarcha. Bull. geol. Surv. N .Z. Paleont., Wellington, 40. PASTIELS, A. 1948. Contributions a 1'etude des microfossiles de 1'Eocene beige. Mem. Mus. r. Hist. nat. Belg., Bruxelles, 109 : 1-77, pis. 1-6. PEAKE, N. B. & HANCOCK, J. M. 1961. The Upper Cretaceous Rocks of Norfolk, in The Geology of Norfolk. Trans. Norfolk Norwich Nat. Soc., Norwich, pp. 293-339. POCOCK, S. A. J. 1962. Microfloral analysis and age determination of strata at the Jurassic- Cretaceous boundary in the Western Canada plains. Palaeontographica, Cassel, Stuttgart, 111, B : 1-95, pis. 1-15. Palynology of the Jurassic Sediments of Western Canada (to be published). ROSE, C. B. 1835-36. A Sketch of the Geology of West Norfolk. Phil. Mag., London, 3, 7 : 171-182, 274-279, 370-376 and 3, 8 : 28-42. ROSSIGNOL, M. 1964. Hystrichosph6res du Quaternaire en Mediterranee orientale, dans les Sediments Pleistocenes et les boues Marines Actuelles. Revue Micropaleont, Paris, 2 : 83-99, pis. 1-3. SARJEANT, W. A. S. 1959 Organic-Shelled Microplankton of the Callovian and Oxfordian. Doctoral Thesis, University of Sheffield. - 1962. Microplankton from the Ampthill Clay of Melton South Yorkshire. Palaeontology, London, 5, 3: 478-497, pis. 69-70. 1965^. The Xanthidia. Endeavour, London, 24, 91 : 33-39, figs. 1-19. - 19656. Microplankton from the Callovian (S. Calloviense Zone) of Normandy. Revue Micropaleont., Paris, 3 : 175-184, pi. i. - 1966^. Dinoflagellate Cysts with a Gonyaulax-type tabulation, in Studies of Mesozoic and Cainozoic Dinoflagellate Cysts. Bull. Br. Mus. nat. Hist., London. Suppl. 3 : 107-156. i8o CENOMANIAN NON-CALCAREOUS MICROPLANKTON, i SARJEANT, W. A. S. 19666. Further Dinoflagellate Cysts from the Speeton Clay (Lower Cretaceous), in Studies of Mesozoic and Cainozoic Dinoflagellate Cysts. Bull. Br. Mus. nat. Hist., London. Suppl. 3 : 199-214. SARJEANT, W. A. S. & DOWNIE, C. 1966. The Classification of Dinoflagellate Cysts above Generic Level. Grana palynol., Stockholm, 6, 3 : 503-527. SMITH, W. E. 1957. The Cenomanian Limestone of the Beer District, South Devon. Proc. geol. Ass., London, 68 : 115-135. - 1961. The Cenomanian Limestone and contiguous Deposits west of Beer. Proc. geol. Ass., London, 72 : 91-134. 1965. The Cenomanian Limestone of Seaton. Proc. geol. Ass., London, 76 : 121-136. STANLEY, E. A. 1965. Upper Cretaceous and Palaeocene Plant Microfossils and Palaeocene Dinoflagellates and Hystrichosphaerids from Northwestern South Dakota. Bull. Am. Paleont., Ithaca, 49 : 179-355. STEIN, F. VON. 1883. Der Organismus der Infusions-thieve, 3. Abth., 2 Halfte: Die Natur- geschichte der arthrodelen Flagellaten. Leipzig: Engelmann, 30 pp., 25 pi. (pp. 18-19 rele- vant) . TASCH, P. K., McCLURE, K. & OFTEDAHL, O. 1964. Biostratigraphy and taxonomy of a hystrichosphere dinoflagellate assemblage from the Cretaceous of Kansas. Micro- paleontology, New York, 10, 2 : 189-206, pis. 1-3. VALENSI, L. 1955. Etude micropaleontologique des silex du Magdalenien de Saint- Amand (Cher). Bull. Soc. prehist.fr., Paris, 52, 9-10 : 584-596, pis. 1-5. VOZZHENNIKOVA, T. F. 1965. Vvedenye v izuchenye iskopayemyx Peridineyevyx vodoroslei. Dokl. Akad. Nauk. SSSR (Izdatelstvo " Nauka " Moskva), Moskva, Leningrad, 156 pp., figs. 1-50. WALL, D. 1967. Fossil Microplankton in Deep Sea Cores from the Caribbean Sea. Palaeon- tology, London, 10, i : 95-123, pis. 14-16. WETZEL, O. 1933. Die in organischer Substanz erhaltenen Mikrofossilien des Baltischen Kreide-Feuersteins. Palaeontographica, Cassel, Stuttgart, 77 : 141-188, figs, i-io, 78 : i-no, pis. 1-7. WETZEL, W. 1952. Beitrag zu Kenntnis des dan-zeitlichen Meeresplanktons. Geol. Jber., Berlin (for 1950), 66 : 391-419, pi. A. WHITAKER, W. 1859. Catalogue of the Rock Specimens in the Museum of Practical Geology, London. Ed. 2, p. 296. WHITE, H. H. 1842. On fossil Xanthidia. Microsc. /., London, 11 : 35-40, pi. 4. WHITE, H. J. OSBORNE. 1921. A Short Account of the Geology of the Isle of Wight. Mem. geol, Surv. U.K., London. WILLIAMS, G. L. 1965. Organic-Walled Microfossils aid Oil Search. Oil Gas J., Tulsa, pp. 108-112. WILLIAMS, G. L. & DOWNIE, C. 19660. The genus Hystrichokolpoma, in Studies of Mesozoic and Cainozoic Dinoflagellate Cysts. Bull. Br. Mits. nat. Hist., London. Suppl. 3 : 176-181. 19666. Further Dinoflagellate Cysts from the London Clay, in Studies of Mesozoic and Cainozoic Dinoflagellate Cysts. Bull. Br. Mus. nat. Hist., London. Suppl. 3 : 215-235. WRIGHT, C. W. 1963. Geology of the Yorkshire Coast. Geol. Assoc. Guide, London, No. 34. PLATE i Gonyaulacysta exilicristata sp. nov. Fetcham Mill Borehole (depth, 730 feet). FIG. i. Dorsal surface of holotype. Slide PF. 3987 (i). x 500. FIG. 2. Ventral surface. Slide PF. 3987. X 500. Carpodinium obliquicostatum Cookson & Hughes FIG. 3. Lower Chalk, Fetcham Mill Borehole (depth, 750 feet). Lateral view. Slide PF. 3988. x 500. FIG. 4. Lower Chalk, Escalles Borehole (depth, 159 metres) . Dorsal surface with archaeopyle. Slide E 195/3. X I2 5 Histocysta pa I la sp. nov. Fetcham Mill Borehole. FIG 5. Holotype; view of archaeopyle and attached operculum. Slide PF. 3052. x 500. FIG. 6. Paratype; Antapical view. Slide PF. 3991 (i). X 500. Gonyaulacysta delicata sp. nov. FIG. 7. Ventral surface of holotype. V. 51979 (i). X 500. FIG. 8. Paratype; V. 5 1979(2). X 500. Gonyaulacysta sp. A. Figured specimen. Lower Chalk, Fetcham Mill Borehole. FIG. 9. Ventral surface. Slide PF. 3987 (2). x 500 FIG 10. Medial view. Slide PF. 3987 (2). x 500. Butt. Br. Mus. nat. Hist. (Geol.) 17, 3 * PLATE i m 4 GEOL. 17, 3. PLATE 2 Cribroperidiniutn intricatum sp. nov. FIG. i. Ventral surface of holotype. V. 51980 (i). x 500. FIG. 2. Lower Colorado, Saskatchewan (depth, 1,023 feet). Dorsal surface with operculum in situ. .51980. x 500. FIG. 3. Lower Colorado, Saskatchewan. Detached operculum. Slide Sas 1023/3. X 500. Microdinium setosum Sarjeant. FIG. 4. Lower Chalk, Fetcham Mill Borehole (depth, 840 feet). Lateral view. Slide PF. 3036. x 640. Microdinium variospinum sp. nov. FIG. 5. Lower Chalk, Escalles Borehole (depth, 195 metres). Antapical sutural spines visible. Slide E 195/3. X 640. FIG. 6. Lateral view of holotype. V. 51981 (i). x 640. IMicrodinium crinitutn sp. nov. FIG. 7. Lower Chalk, Fetcham Mill Borehole (depth, 770 feet). Slide FM 770/15. x 640. FIG. 8. Dorsal surface of holotype Slide PF. 3990 (i). x 640. Microdinium distinctum sp. nov. FIG. 9. Ventral surface of holotype. Slide PF. 3989 (i). X 640. FIG. 10. Dorsal surface of holotype. X 640. FIG. ii. Lower Chalk, Fetcham Mill Borehole (depth, 790 feet). Ventral surface. Slide FM 790/16. x 640. Bull. Br. Mus. nat. Hist. (Geol.) 17, 3 PLATE 2 f 8 v ' - V W 10 11 t V PLATE 3 Ellipsodinium rugulosum Clarke & Verdier FIG. i. Lateral view of specimen showing operculum partially detached. Slide PF. 3988. X 1250. Fromea amphora Cookson & Eisenack Lower Chalk, Fetcham Mill Borehole. FIG. 2. Slide FM 770/6 (depth, 770 feet). X 500. FIG. 3. Specimen with operculum attached (depth, 650 feet). Slide PF. 3041. x 500. Microdinium veligerum (Deflandre). FIG. 4. Lower Chalk, Fetcham Mill Borehole (depth, 650 feet). Antapical view to show crestal cavities. Slide FM 650/5. x 1250. Apteodinium granulatum Eisenack. FIG. 5. Lower Chalk, Fetcham Mill Borehole (depth, 690 feet). Ventral surface. Slide FM 690/12. x 640. FIG. 6. Lower Chalk, Escalles Borehole (depth, 165 metres). Lateral view. V. 51981. X 640. Cassiculosphaeridia reticulata sp. nov. FIG. 7. Lower Chalk, Fetcham Mill Borehole (depth, 730 feet) . Operculum partially detached. Slide FM 730/9. X 500. Chytroeisphaeridia euteiches sp. nov. FIG. 8. Holotype showing archaeopyle. V. 51982 (2). x 500. FIG. 9. Lower Chalk, Escalles Borehole (depth, 159 metres). Archaeopyle with detached operculum. V. 51982. x 500. Epelidosphaeridia spinosa (Cookson & Hughes) Lower Chalk, Fetcham Mill Borehole. FIG. 10. Dorsal surface. Slide PF. 3992, (depth, 770 feet), x 500. FIG. ii. Ventral surface with sulcus. Slide PF. 3992 (depth, 770 feet). X 500. FIG. 12. Complete specimen. Slide PF. 3048 (depth, 770 feet), x 500. Bull. Br. Mus. nat. Hist. (Geol.) 17, 3 PLATE 3 m f * i A \ I 8 x- -JW - " ': 10 11 12 PLATE 4 Hystrichosphaeridium deanei Davey & Williams FIG. i. Lower Chalk, Compton Bay. Complete specimen (137 feet above base of Chalk). Slide CB 19/2. x 500. Hystrichosphaeridium difficile Manum & Cookson Upper Colorado, Saskatchewan. FIG. 2. Complete specimen, archaeopyle partially detached. .51983 (depth, 805 feet). X 500. FIG. 6. Detached operculum bearing 4 processes. Slide Sas 805/3 (depth, 805 feet). X 500. FIG. 7. Detached operculum, lateral view. Slide Sas 805/3 (depth, 805 feet). X 500. X 500. Cassiculosphaeridia reticulata sp. nov. FIG. 3. Holotype; view of archaeopyle. V. 51981 (4). x 500. Microdinium veligerum (Deflandre). FIG. 4. Lower Chalk, Fetcham Mill Borehole (depth, 750 feet). Dorsal surface. Slide PF. 3988. x 640. Microdinium cf . ornatum Cookson & Eisenack FIG. 5. Lower Chalk, Escalles Borehole (depth, 189 metres). Dorsal surface. Slide E 189/4. X 640. Hystrichosphaeridium radiculatum Davey & Williams FIG. 8. Lower Chalk, Escalles Borehole (depth, 159 metres). Lateral view showing archaeo- pyle. V. 51982. x 500. Hystrichosphaeridium mantelli Davey & Williams. FIG. 9. Lower Chalk, Escalles Borehole (depth, 153 metres). Lateral view to show archaeo- pyle. Slide E 153/3. X 500. Polysphaeridium laminaspinosum Davey & Williams Lower Chalk, Fetcham Mill Borehole. FIG. 10. Apical view showing archaeopyle. Slide PF. 3035 (depth, 840 feet), x 500. FIG. ii. Antapical view. Slide PF. 3035 (depth, 840 feet), x 500. Bull. Br. Mus. not. Hist. (Geol.) 17, 3 PLATE 4 8 10 11 PLATE 5 Oligosphaeridium anthophorurn (Cookson & Eisenack) Lower Colorado, Saskatchewan. FIG. i. Enlargement to show process extremities. Slide Sas 1023/3 (depth, 1023 feet). X 975- FIG. 2. Detached operculum. Slide Sas 1023/1 (depth, 1023 feet). X 500. FIG. 3. V. 51980 (boring depth, 1023 feet), x 500. Oligosphaeridium prolixispinosum Davey & Williams FIG. 4. Lower Chalk, Compton Bay (116 feet above the base of the Chalk). Specimen showing " bald " cingular region. Slide CB I7/C. X 500. Hystrichosphaeridium tubiferum (Ehrenberg) Lower Chalk, Fetcham Mill Borehole. FIG. 5. Detached operculum. Slide FM 690/12 (depth, 690 feet). X 500. FIG. 8. Slide PF. 3987 (depth, 730 feet), x 500. Oligosphaeridium complex (White) FIG. 6. Lower Chalk, Escalles Borehole (depth, 159 metres). Detached operculum composed of 4 plates. V. 51982. x 500. FIG. 7. Lower Chalk, Fetcham Mill Borehole (depth, 750 feet). Specimen illustrating the 6 precingular processes. Slide FM 750/13. x 500. Hystrichosphaeridium bowerbanki Davey & Williams FIG. 9. Lower Chalk, Compton Bay (116 feet above the base of the Chalk). Lateral view, archaeopyle to the north. Slide CB I7/C. x 500. Bull. Br. Mus. nat. Hist. (Geol.) 17, 3 PLATE 5 . '/ 5 GEOL. 17, 3. PLATE 6 Oligosphaeridiutn reniforme (Tasch) FIG. i. Lower Colorado, Saskatchewan (depth, 1023 feet). Slide Sas 1023/1. x 500. Tanyosphaeridiutn variecalamum Davey & Williams FIG 2 Holotype. Slide PF 3035 (2) x 500 FIG 5 Lower Chalk, Compton Bay (28 feet above base of Chalk). Precingular, cingular and postcingular processes aligned into three series. Slide CB 5/C x 500 Litosphaeridium siphoniphorum (Cookson & Eisenack) Lower Chalk, Fetcham Mill Borehole. FIG. 3. Lateral view showing precingular and postcingular processes, and antapical process. Slide PF. 3987 (depth, 730 feet), x 500. FIG. 4. Detached operculum. Slide FM 690/14 (depth, 690 feet). X 975. Callaiosphaeridium asymmetricum (Deflandre & Courteville) . FIG. 6. Upper Greensand, Fetcham Mill Borehole (depth, 886 feet). Lateral view illustrating epitractal archaeopyle. Slide FM 886/2. x 500. Cleistosphaeridiutn polypes (Cookson & Eisenack) FIG. 7. Upper Woodbine Formation. Enlargement to show process extremities. Slide T5/3- X 975- FIG. 8. Lower Chalk, Compton Bay (151 feet above base of Chalk). Specimen possessing apical archaeopyle. Slide CB 2i/C. x 500. Cleistosphaeridiutn polypes var. clavulum nov. FIG. 9. Type. Slide PF. 3995 (i) x 500. FIG. 10. Lower Chalk, Fetcham Mill Borehole (depth, 840 feet). Enlargement to show process extremities. Slide PF. 3035. x 975. fCleistosphaeridium aciculare sp. nov. FIG. ii. Upper Colorado, Saskatchewan (depth, 835 feet). Specimen with numerous fine processes. V. 51988. x 500 (phase contrast). FIG. 12. Holotype. V. 51979 (3). X 500. Bull. Br. Mus. nat. Hist. (Geol.) 17, 3 PLATE 6 8 12 , *t 10 PLATE 7 Exochosphaeridium striolatutn var. truncatum nov. FIG. i. Lower Chalk, Fetcham Mill Borehole (depth, 750 feet). Apical view showing two partially detached precingular plates to the north. Slide PF. 3988. x 500. FIG. 2. Type; precingular archaeopyle clearly shown to be formed by the removal of two plates. V. 51982 (i). x 500. FIG. 3. Lower Chalk, Escalles Borehole (depth, 165 metres). Detached operculum composed of two precingular plates. .51981. x 500. fCleistospaeridium flexuosum Davey et al. FIG. 4. Lower Chalk, Fetcham Mill Borehole (depth, 690 feet). Slide PF. 3030. x 500 (phase contrast). Exochosphaeridium phragmites Davey et al. FIG. 5. Lower Chalk, Escalles Borehole (depth, 165 metres). Apical process with precingular archaeopyle to the north-west. V. 51981. x 500. Cleistosphaeridium huguonioti var. pertusutn nov. FIG. 6. Lower Chalk, Fetcham Mill Borehole (depth, 690 feet). Enlargement showing vacuolated processes with small distal bifurcations. Slide FM 690/14. x 975. FIG. 7. Type. Slide PF. 3040 (2). x 500. FIG. 9. Lower Chalk, Escalles Borehole (depth, 159 metres). Specimen possessing apical archaeopyle. Slide E 159/4. X 500. Cleistosphaeridium heteracanthum (Deflandre & Cookson) FIG. 8. Lower Chalk, Compton Bay (137 feet above base of Chalk). Slide CB 19/2. x 500. Cleistosphaeridium huguonioti (Valensi) FIG. 10. Lower Chalk, Fetcham Mill Borehole (depth, 730 feet). Detached operculum. Slide FM 730/19. x 975. ? Cleistosphaeridium parvum sp. nov. FIG. ii. Holotype. V. 51981 (3). x 975. FIG. 12. Lower Chalk, Escalles Borehole (depth, 159 metres). Cingular region devoid of processes clearly visible. Slide E 159/1. X 975. Bull. Br. Mus. nat. Hist. (Geol.) 17, 3 PLATE 7 / k * ' * 7 PLATE 8 Cleistosphaeridiutn armatum (Deflandre) FIG. i. Lower Chalk, Escalles Borehole (depth, 189 metres). View of apical archaeopyle. Slide E 189/4. x 975- FIG. 2. Lower Chalk, Compton Bay (116 feet above base of Chalk). Enlargement to show process extremities. Slide CB iy/C. x 975. FIG. 12. Lower Chalk, Compton Bay (i 16 feet above base of Chalk). Slide CB I7/C. x 500. Cyclonephelium eisenacki sp. nov. FIG. 3. Lower Colorado, Saskatchewan (depth, 1,023 feet). Specimen with archaeopyle developed. .51980. X 500. FIG. 4. Holotype. V. 51980 (2). x 500. Prolixosphaeridium conulurn sp. nov. FIG. 5. Holotype; complete specimen. V. 51981 (5). x 500. FIG. 6. Lower Chalk, Compton Bay (59 feet above base of Chalk). Apical archaeopyle to the north. Slide CB 9/2. X 500. Cleistosphaeridium multifurcatum (Deflandre) FIG. 7. Lower Chalk, Escalles Borehole (depth, 207 metres). View of apical archaeopyle. Slide E 207/5. X 500. FIG. 10. Lower Chalk, Compton Bay (15 feet above base of Chalk). Complete specimen. V. 51986. x 500. Coronifera oceanica Cookson & Eisenack FIG. 8. Lower Chalk, Fetcham Mill Borehole (depth, 730 feet). Specimen possessing an apical archaeopyle and an antapical process. Slide FM 730/13. x 500. FIG. ii. Lower Chalk, Escalles Borehole (depth, 189 metres). Complete specimen. Slide E 189/4. x 500. Surculosphaeridiutn longifurcatum (Firtion) FIG. 9. Lower Chalk, Fetcham Mill Borehole (depth, 730 feet). Lateral view to show apical archaeopyle, deeply furcate precingular processes and completely subdivided cingular processes. Slide PF. 3987. X 500. Bull. Br. Mus. nat. Hist. (Geol.) 17, 3 PLATE 8 ** 10 12 PLATE 9 Cyclonephelium paucispinum sp. nov. FIG. i. Holotype. V. 51981 (2). x 500. FIG. 2. Lower Chalk, Compton Bay (15 feet above base of Chalk). Complete specimen with operculum partially detached. V. 51986. x 500. Cyclonephelium vannophorum sp. nov. FIG. 3. Enlargement of holotype to show apical archaeopyle, and the shape of the processes. V. 51986 (i). X 975- Cyclonephelium eisenacki sp. nov. FIG. 4. Lower Colorado, Saskatchewan (depth, 1,023 feet). Complete specimen. Slide Sas 1023/3. X 500. Hystrichokolpoma ferox (Deflandre) Lower Chalk, Fetcham Mill Borehole (depth, 840 feet). Slide FM 840/11. FIG. 5. Lateral view (bottom of specimen by transparency) showing precingular, cingular and postcingular processes ; large antapical process to the south and fine sulcal processes to the west, x 500. FIG. 6. Medial section, x 500. FIG. 7. Lateral view (top of specimen) showing precingular, cingular and postcingular processes, x 500. Bull. Br. Mus. nat. Hist. (Geol.) 17, 3 PLATE 9 3 v PLATE 10 Hystrichosphaera ramosa var. ramosa (Ehr.) Lower Chalk, Fetcham Mill Borehole (depth, 750 feet). FIG. i. Ventral surface. Slide PF. 3988. x 500. FIG. 2. Dorsal surface with precingular archaeopyle. Slide PF. 3988. x 500. FIG. 5. Detached operculum (boring depth, 770 feet). Slide FM 770/15. x 500. Hystrichosphaera ramosa var. multibrevis Davey & Williams Lower Chalk, Fetcham Mill Borehole. FIG. 3. Slide PF. 3988 (depth, 750 feet), x 500. FIG. 4. Slide PF. 3988 (depth, 750 feet), x 500. Hystrichodiniutn voigti (Alberti) Lower Chalk, Escalles Borehole. FIG. 6. Archaeopyle to the north-east. V. 51982 (depth, 159 metres). FIG. 10. Detached operculum bearing 4 processes. Slide E 165/1 (depth, 165 metres). X 500. Achomosphaera ramulifera (Deflandre) FIG. 7. Lower Chalk, Escalles Borehole (depth, 159 metres). Specimen showing precingular archaeophyle and apical process. V. 51982. x 500. Hystrichodinium dasys sp. nov. FIG. 8. Holotype illustrating cingulum. V. 51982 (3). x 500. FIG. 9. Middle Chalk, Fetcham Mill Borehole (depth, 520 feet). Specimen with unusually stout processes proximally. Slide FM 520/7. x 500. Adnatosphaeridium chonetum (Cookson & Eisenack) FIG. ii. Lower Chalk, Escalles Borehole (depth, 165 metres). Complete specimen, V. 51981. X 500. FIG. 12. Lower Chalk, Fetcham Mill Borehole (depth, 730 feet). Apical archaeopyle present. Slide PF. 3987. x 500. Bull. Br. Mus. nat. Hist. (Geol.) 17, 3 PLATE 10 8 10 -- 12 PLATE ii Trichodinium castaneum (Deflandre). Lower Chalk, Escalles Borehole. FIG. i. Lateral view showing precingular archaeopyle and cingulum, V. 51989 (depth, 195 metres), x 500. FIG. 2. Dorsal view showing small apical horn. Slide E 159/2 (depth, 159 metres), x 500. FIG. 3. Complete specimen possessing a cingulum and possibly a sulcus, Slide E 189/4 (depth, 189 metres), x 500. Exochosphaeridium pseudohystrichodinium (Deflandre) Lower Chalk, Escalles Borehole. FIG. 4. Complete specimen with operculum partially detached, V. 51982 (depth, 159 metres). X 500. FIG. 5. Complete specimen illustrating pitted surface, V. 51981 (depth, 165 metres), x 500. Cyclonephelium distinctum Deflandre & Cookson FIG. 6. Lower Chalk, Escalles Borehole (depth, 159 metres). Specimen with unusually long processes, V. 51982. x 500. FIG. 7. Lower Chalk, Escalles Borehole (depth, 159 metres). Complete specimen; archaeo- pyle in the act of developing. V. 51989. x 500. FIG. 8. Lower Chalk, Compton Bay (15 feet above base of Chalk). Typical specimen with archaeopyle developed. V. 51986. X 500. FIG. 10. Lower Chalk, Escalles Borehole (depth, 165 metres). Detached operculum. V. 51981. x 500. Cyclonephelium membraniphorum Cookson & Eisenack FIG. 9. Lower Chalk, Escalles Borehole (depth, 153 metres). Apical archaeopyle well illustrated. Slide E 153/3. X 500. Cyclonephelium vannophorum sp. nov. FIG. ii. Holotype with operculum partially detached. V. 51986 (i). x 500. FIG. 12. Lower Chalk, Compton Bay (15 feet above base of Chalk). Specimen with apical archaeopyle developed. V. 51986. x 500. Bull. Br. Mus. nat. Hist. (Geol.) 17, 3 PLATE ii \ \ w? 10 PRINTED IN GREAT BRITAIN BY ADLARD & SON LIMITED BARTHOLOMEW PRESS, DORKING A REDESCRIPTION OF W. CARRUTHERS' TYPE GRAPTOLITES ISLES STRACHAN BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 17 No. 4 LONDON: 1969 A REDESCRIPTION OF W. CARRUTHERS' TYPE GRAPTOLITES BY ISLES STRACHAN Pp. 181-206; 5 Plates; 8 Text-figures BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY) GEOLOGY Vol. 17 No. 4 LONDON : 1969 THE BULLETIN OF THE BRITISH MUSEUM (NATURAL HISTORY), instituted in 1949, is issued in five series corresponding to the Departments of the Museum, and an Historical series. Parts will appear at irregular intervals as they become ready. Volumes will contain about three or four hundred pages, and will not necessarily be completed within one calendar year. In 1965 a separate supplementary series of longer papers was instituted, numbered serially for each Department. This paper is Vol. 17, No. 4 of the Geological (Palaeontological) series. The abbreviated titles of periodicals cited followthose of the World List of Scientific Periodicals. World List abbreviation: Bull. Br. Mus. nat. Hist. (Geol.). Trustees of the British Museum (Natural History) 1969 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) Issued 17 January, 1969 Price i 45. A REDESCRIPTION OF W. CARRUTHERS' TYPE GRAPTOLITES By ISLES STRACHAN CONTENTS Page I. INTRODUCTION ......... 183 II. DESCRIPTIONS ......... 185 Leptograptus capillaris . . . . . . . 185 Pleurograptus linearis . . . . . . . 186 Dicellograptus elegans . . . . . . . 187 Dicellograptus moffatensis . . . . . . . 189 Dicranograptus clingani . . . . . . . 190 Climacograptus minimus ....... 191 Climacograptus minutus . . . . . . . 193 Cvyptograptus tricornis . . . . . . . 194 Monograpius clingani . . . . . . . 195 Monograptus intermedius . . . . . . . 196 Diver sograptus? capillaris . . . . . . . 198 Rastrites maximus ........ 200 Cyvtograptus murchisoni ....... 202 " Dendrograptus " lentus . . . . . . . 204 III. REFERENCES .......... 204 SYNOPSIS The fourteen species of graptolites described and figured by William Carruthers, mainly in his paper of 1868, are redescribed from the type material where possible. Two neotypes are proposed. Some of the species appear to be unsatisfactory by modern standards, but it is hoped that refiguring of the types will assist future work. I. INTRODUCTION IN their monograph of British Graptolites (1901-18) Elles & Wood described and figured most of the previously known species as well as numerous new forms. In many cases, however, they did not use the original type material either because it was not available or because later collections had provided other, sometimes better, material. In the fifty years which have elapsed since then, discrimination of grapto- lite species has become keener, particularly by workers overseas who have frequently, in their subdividing, formally designated type specimens without consideration of the availability or value of the early material. It seems appropriate that the species described in the nineteenth century should be re-examined so that these workers should be able to recognize the old species accurately with the aid of modern descriptions and illustrations. The work could be tackled systematically by genus and species but this involves consideration of generic and specific limits in deciding GEOL. IJ, 4 15 184 W. CARRUTHERS' TYPE GRAPTOLITES the scope of a paper. It is easier to treat the species for redescription by discussing one author at a time and I hope to undertake further papers of a similar scope to this. William Carruthers (1830-1922) was born at Moffat, Dumfriesshire, and early developed an interest in the local geology, particularly in the graptolites for which the area had then recently become famous through the researches of Harkness and Salter. His own publications on graptolites are comparatively few in number but it is clear that by 1867 he was regarded as an expert on the group, since he contributed the article on graptolites to the fourth edition of Murchison's " Siluria ". In the following year he described more fully the new species erected in that memoir. He strongly disagreed with Nicholson on the interpretation of a number of graptolite features and they carried on an acrimonious correspondence in the pages of the Geological Magazine during 1867. At the same time he was also describing fossil plants, to which he later devoted all his research. In 1870-72 he was assisting the young amateur John Hopkinson with his descriptions of new species of graptolites from the Moffat Shales but they both appeared to realize that the publication in 1876 of Lap worth's paper on " Scottish Monograptidae " marked a new and more critical approach to graptolite systematics and stratigraphy and neither of them published further papers on the subject. It is, however, clear from Lapworth's later papers that he maintained a friendly association with Hopkinson and Carruthers and the latter's manuscript notes on graptolites are in the Lapworth Library, Birmingham University. These notes contain no indication of how or when Lap- worth acquired them but there are no references to papers after 1872 and it is possible that Carruthers gave them to Lapworth before taking up more administrative duties at the British Museum (Natural History) in 1880 and confining his research to palaeobotany. The new species described by Carruthers are found in two main papers, in 1858 and 1868. The first of these was reprinted the following year in a more accessible journal and, as already mentioned, the 1868 paper described properly the species which were only illustrated in the appendix to " Siluria ". Carruthers had some harsh remarks to make about Nicholson's drawings of graptolites and, on the whole, his own are generally sufficiently accurate to allow recognition of the type specimens. Most of these were catalogued at once in the British Museum (Natural History) collections where they can be traced. Almost all his species can be recognized fairly easily. The type of Rastrites maximus is the only one which has not been found and a neotype is proposed for that species. It is a pleasure to acknowledge the ready co-operation which I have had from those at the British Museum (Natural History) who have had charge of the graptolite collections and to thank their photographers who have provided most of the illus- trations for the plates. I must also express my thanks to the Geological Survey & Museum and to the Royal Scottish Museum for the loan of specimens over a period which has been much longer than I had at first intended. Specimens with Q numbers are in the British Museum (Natural History); with GSM, in the Institute of Geological Sciences, London; with BU, in the Geology Department, University of Birmingham; and with SM, in the Sedgwick Museum, Cambridge. W. CARRUTHERS' TYPE GRAPTOLITES 185 II. DESCRIPTIONS Leptograptus capillaris (Carruthers) PL i 1868 Cladograpsus capillaris Carruthers: 130, pi. 5, figs. 7, ya. 1876 Leptograptus capillaris (Carr.) Lapworth : 9, pi. 3, fig. 72. 1903 Leptograptus capillaris (Carr.); Elles and Wood : 112, pi. XV, figs. 4a-d. ?I954 Leptograptus capillaris (Carr.); Sherrard 195, pi. n, fig. 8. Original description. " Extremely slender polypary, with remote branches and very minute hydrothecae; about twenty-four in an inch. It is not so abundant as C. linearis, and is easily distinguished by its capillary appearance. It is probably the same species as that figured and described by Emmons in his American Geology, vol. i, p. 109, pi. i, fig. 7, under the name of Nemagrapsus capillaris. Loc. Moffat ". Comments on usage. There appears to have been little confusion over this species. Although he quoted Emmons' species, Carruthers clearly intended his own form to rank as a new species and the American form is now referred to Thamnograptus. The original description refers to " remote branches " which are not shown by the type specimens or original figure but one illustration of Cladograpsus linearis of the previous year (Carruthers 1867, pi. 2, fig. I7a) shows the characteristic curved stipes with secondary branches. The original of this has not however been recognized. TYPE MATERIAL. The syntypes, British Museum (Natural History) Q-30, were refigured by Elles & Wood. As lectotype I select the large rhabdosome on the upper right of Carruthers' original figure (see PI. i, fig. 2). The precise locality and horizon are not known but Elles & Wood give it as Hartfell Spa?, Hartfell Shales. DIAGNOSIS. Very slender Leptograptus with markedly curved stipes, maximum breadth 0-5 mm. thecae eight to ten per cm. REVISED DESCRIPTION. The species normally occurs crowded on the slabs so that details of the stipes are not readily seen. The sicula is inconspicuous but the stipes appear to grow initially horizontally or even somewhat downward before curving gracefully upwards and inwards, forming loops on occasion. The stipes widen from an initial breadth of 0-2 mm. to a maximum of 0-5 mm. over a distance of at least 4 or 5 cm. and appear to be twisted so that the thecae can appear on either side of the curve. The sicula is 2 mm. long and about 0-2 mm. wide but details of the origin of the stipes are not known. The thecae appear to be simple leptograptid tubes but the preservation of the distal parts of the stipes is such as to make measurements of their length and overlap impossible. GEOLOGICAL HORIZON. Lapworth (1878) recorded the species only from the zone of Pleurograptus linearis but Elles & Wood (1903) also recorded it from the zone of Dicranograptus clingani. The type slab shows no associates, nor do most of the specimens in the Lapworth Collection. Lithologically, however, the slabs agree with the upper part of the Lower Hartfell Shales. DISTRIBUTION. This species is known from only a few localities in the South of Scotland and is not recorded from beds of a similar age in South Wales. Ruedemann i86 W. CARRUTHERS' TYPE GRAPTOLITES and Decker recorded it from the Viola Limestone in Oklahoma but gave no figures (Ruedemann 1947). Sherrard (1954) recorded it from New South Wales but her figure shows stipes which appear to be too broad and not curved enough. They may be one of the forms of Leptograptus flaccidus which occurs about the same hori- zon, e.g. L.f. arcuatus Elles & Wood. Thomas (1960) illustrated a form which is clearly not L. capillaris as the stipes are too straight, and this then casts doubt on the range of the species as given by Thomas, which goes up to the Ashgill Series. Pleurograptus linearis (Carruthers) PL 2 ; Figs. la-c 1858 Cladograpsus linearis Carruthers : 467, fig. i. 1859 Cladograpsus linearis Carruthers : 24, fig. i. non 1867 Cladograpsus linearis Carruthers; Carruthers : 369, pi. 2, fig. 17. ?i8&7a Cladograpsus linearis Carruthers; Carruthers : 540, fig. 8. 1867 Pleurograpsus linearis (Carruthers) Nicholson : 257, pi. n, figs. 1-5. 1868 Cladograpsus linearis Carruthers ; Carruthers : 129. 1876 Pleurograptus linearis (Carruthers) ; Lapworth, pi. 3, fig. 69. 1904 Pleurograptus linearis (Carruthers); Elles & Wood : 119, pi. 16, fig. 7, pi. 17, fig. i. ORIGINAL DESCRIPTION. " From a short and very slender base the zoophyte divides into two stems, each supporting the cells on their upper sides. Branches are given off at irregular intervals from these principal stems. The length of the polypidom is very great; one specimen I have been able to trace for nearly three feet . . . The polypidom at its origin, near to the slender base, is very narrow, being little more than a fine line; as it increases in length it increases in breadth, until it is fully two-fifths of a line broad. The cells are very remote from each other, and are, at first sight, from the slight indentation they make in the stem, scarcely perceptible, giving the Graptolite the appearance as if it were a clear line. The mouth of the cell is straight and at right angles to the axis ; it makes an inden- tation equal to about one-sixth of the breadth of the polypidom. The number of cells in an inch is about eighteen. Type locality Hartfell." COMMENTS ON USAGE. Carruthers' later figures (? 1867, i867a) are different from his original but there has been no difficulty in the recognition of his species. Carruthers accepted Nicholson's specimen of Pleurograpsus linearis although he was very critical of the detail of the figures. Nicholson's specimen was refigured by Elles & Wood who also figured a variety simplex in which the branches are much more widely spaced. TYPE MATERIAL. Carruthers' original specimen is in the British Museum (Natural History), (^.848, and the counterpart is amongst material presented to the Royal Scottish Museum, Edinburgh, by Carruthers in 1858 (RSM 1858 . 10 .4). Nicholson's specimen, Q . 27, was said by Elles & Wood to be the type specimen but it is clearly not. The lower of the two large specimens on the type slab has four lateral branches and is recognizable as the specimen figured in 1858 (see PL 2, figs, i, 2). REVISED DESCRIPTION. Rhabdosome consisting of two or three main stipes arising from the sicula with secondary branching on one or both sides. Stipes up W. CARRUTHERS' TYPE GRAPTOLITES 187 to I mm. broad; thecae simple leptograptid, about eight per cm., with apertural excavations occupying about one-third of the rhabdosome breadth. The sicula is not seen on any of the figured material and is probably associated in older rhabdo- somes with one of the branches, a feature also seen in species of Leptograptus where a third branch may be developed. Young specimens on the type slab, however, suggest that the sicula is about 2 mm. long. The stipes are initially very slender but show some secondary thickening in older specimens where the thecae on the main stipes are difficult to distinguish. There is considerable variation in the branching, Nicholson's specimen (PL 2, fig. 3) being much stiffer than Carruthers' or the other specimen figured by Elles & Wood in which the tertiary branched stipes are aligned in flowing curves. This suggests that the variation in attitude of the stipes is simply a depositional effect. fl ^ FIG. i. Pleurograptus linearis Carr. a. Enlargement of proximal part of main stipe showing 2 lateral branches and possible central branch from the sicula. Counterpart of type slab, R.S.M. 1858.10.4. The 2 parallel distal parts of branches are the left hand pair from the holotype (PI. 2, fig. i). b, c. Young specimens showing sicula, type slab, Q.8^8. All X5- DISTRIBUTION. P. linearis appears to be confined to the one horizon in the south of Scotland where, however, it is quite common if all the fragments showing this type of branching really belong to it. Ruedemann (1908; 1947) figured a single specimen from the Utica Shale of New York State which shows similar branching at rather more distant intervals, approximating in that respect to var. simplex which is also recorded by Thomas (1960) from Australia. Dicellograptus elegans (Carruthers) PI. 3, fig. i. ; Figs 2a, b. Didymograpsus elegans Carruthers : 369, pi. 2, fig. i6a. Didymograpsus elegans Carruthers; Carruthers : 129, pi. 5, figs. 8a, ?d. Dicellograpsus elegans (Carruthers) Hopkinson : 24, pi. i, fig. 3. Dicellograptus elegans (Carruthers) ; Lapworth, pi. 4, fig. 87. Dicellograptus elegans (Carruthers) Dicellograptus elegans (Carruthers) 1867 1868 1871 1876 1877 1904 ?I947 1954 ORIGINAL DESCRIPTION. " Branches of the polypary divaricating at various angles, and with a slight curve within a short distance of the proximal origin of the Dicellograptus elegans (Carruthers) Dicellograptus elegans (Carruthers) ; Sherrard, pi. 10, fig. 6. Lapworth : 141, pi. 7, fig. 8. Elles & Wood : 159, pi. 23, figs. 2a-e. Ruedemann : 380, pi. 63, fig. i. i88 W. CARRUTHERS' TYPE GRAPTOLITES polypary. The hydrothecae are rounded at the apex, and free throughout a con- siderable portion of their length and the intervening spaces are rounded at the base ; about twenty-two cells in an inch. The initial process is obvious in young specimens, but I have not been able to detect it in old individuals ; the outer apex of the angle ornamented with three short strong spines. Loc. Moffat ". COMMENTS ON USAGE. Since Elles & Wood refigured the type specimen there has been no difficulty with this species. The originals of Carruthers' figures 8b and 8c (BM(NH) Q-54) were recognized by Hopkinson (1871) as being a distinct species (D. morrisii Hopkinson) and not merely young stages of elegans as Carruthers thought. TYPE MATERIAL. The holotype, Q.850 (PL 3, fig. i), is from the Hartfell Shales, Dobbs Linn. The species appears to be commonest in the P. linearis Zone but the associates on the type slab (Dicellograptus pumilus and Climacograptus spp.) give no direct confirmation of horizon. DIAGNOSIS. Dicellograptus with markedly introverted thecae and with distinct sigmoid curvature of stipes near proximal end, so that the stipes curve first upwards, then outwards, then upwards again. FIG. 2. Dicellograptus spp. a. D. elegans Carr. Detail of thecal aperture in proximal part of stipe. Q.850. x 30. b. D. elegans Carr. Distal thecae. (3.850. X5- c. D. moffatensis Carr. Distal thecae. (3.843. X5- REVISED DESCRIPTION. The proximal double curvature is quite distinctive and the stipes have an almost uniform width of i mm. throughout their length. The first two thecae have prominent apertural spines which, with the virgella, give the three-spined proximal end noted by Carruthers. The thecae number eight to ten per cm. and are of the strongly introverted type with marked ventral curvature (group IV of Elles & Wood, which however are not introtorted, see Bulman 1944 : 37). GEOLOGICAL HORIZON. Lapworth (1878) and Elles & Wood (1904) recorded the species only from the P. linearis zone, but in the summary range chart at the end of the Monograph, Elles & Wood also recorded it as common in the underlying zone of D. clingani. Elles (1925) also recorded it from both zones so its precise range must await a revision of the Hartfell Shales. Foreign records are surprisingly scanty but it seems to occur in Australia at the same horizon. W. CARRUTHERS' TYPE GRAPTOLITES 189 Dicellograptus moffatensis (Carruthers) PL 3, figs. 5, 6; Fig. 2c 1858 Didymograpsus Moffatensis Carruthers : 469, fig. 3. 1859 Didymograpsus Moffatensis Carruthers; Carruthers 1868 Didymograpsus Moffatensis Carruthers; Carruthers 26, fig. 3. 129. 25, pi. i, fig. 4. 1871 Dicellograpsus Moffatensis (Carruthers) Hopkinson 1875 Dicellograptus moffatensis (Carruthers); Hopkinson & Lapworth : 654, pi. 34, fig. 5a. 1877 Dicellograptus Moffatensis (Carruthers); Lapworth : 141, pi. 7, fig. 9. 1904 Dicellograptus moffatensis (Carruthers); Elles & Wood : 157, pi. 23, figs. la-f. ORIGINAL DESCRIPTION. ' The base terminates in three distinct spinous pro- cesses. The zoophyte bifurcates from the base. The general appearance is like the figure; or occasionally the lines form an acute angle for about a quarter of an inch, then suddenly expand in slight curves, almost at right angles, for a short dis- tance, when they again recur to their original direction. The branches are united for about a quarter of a line by a slight web, which in some specimens is terminated in a fine process of short length, taking the direction of a line bisecting the angle. The cells are arranged in the outer margins; they are very remote, and penetrate the polypidom to scarcely one-fourth of its breadth; they form slight openings on the margin of the polypidom, first entering at a right angle, and then suddenly turning downwards. These openings are lengthened ovate pouches, answering exactly in shape and size to the cell-serratures of the margin. The number of cells in an inch is about twenty. The breadth of the polypidom is about two-thirds of a line. Type locality Hartf ell ." COMMENTS ON USAGE. In 1868, Carruthers noted D. divaricatus (Hall) and D. anceps (Nicholson) as synonyms of his own species but this is merely a reflection of the confused state of graptolite systematics at the time. Elles & Wood refigured Carruthers' type but the species does not seem to have been widely recognized. TYPE MATERIAL. This is one of the few species of which Elles & Wood made any discussion of type specimens. They considered that Carruthers' specimen was not a satisfactory type and that a specimen from the Lapworth Collection should be taken as the type. Carruthers' original figure is admittedly poor (PL 3, fig. 5) but the type slab shows three specimens which agree fairly well with it and show the essential characters of the species. The abrupt widening of the stipes on Lap worth's specimen, which Elles & Wood wanted as a character of the species, appears to be the result of slight shearing. It is impossible to decide which of the three specimens was the original of Carruthers' figure so the specimen (Q . 843) figured by Elles & Wood is here selected as lectotype (see PL 3, fig. 6). It is from the Hartf ell Shales of Hartf eU and is associa- ted on the slab with Orthograptus cf . whitfieldi. This probably indicates a low horizon in the Hartf ell Shales. Lapworth (1878) records it no higher than the zone of Climacograptus wilsoni. REVISED DESCRIPTION. Stipes sub-parallel initially, then diverging making an angle of about 45, sometimes later converging. Stipes widening from an initial breadth of 0-4 mm. to a maximum of about 1-5 mm. Thecae eight to ten per cm., igo W. CARRUTHERS' TYPE GRAPTOLITES with markedly curved ventral walls and introverted apertures. The proximal end usually shows a membrane in the axil of the stipes, obscuring the sicula. The virgella and first thecal spines variably developed. The thecae are poorly preserved but distally appear to be very similar to those of D. elegans (Fig. 2c). It is unfortunate that both Carruthers and Hopkinson included D. divaricatus (Hall) in their synonymy of this species as it makes it difficult to establish what they regarded as the diagnostic features of the species. The general shape of the rhab- dosome appears to be the most characteristic feature, particularly the narrow axil and later divergence. In this respect the specimen figured by Elles & Wood in their Monograph (pi. 23, fig. ic) is not at all typical but its only associates are the pair of specimens figured on the same plate (pi. 23 fig. ib) which have the typical shape. As the rhabdosomes were flexible in life to some extent, this poses the prob- lem of how much reliance should be placed on rhabdosome shape as a specific character in Dicellograptus and other forms with long slender stipes. D. moffatensis can be readily separated from other British Dicellograptus by size and shape of rhab- dosome, D. morrisii Hopkinson being the most similar but with a more open axil. D. moffatensis var. alabamensis Ruedemann 1908 is not related at all and is clearly a Dicranograptus, close to D. brevicaulis Elles & Wood 1904. DISTRIBUTION. Elles & Wood recorded the species widely throughout the British Isles but it seems doubtful now if it occurs in the Lake District. The specimen figured by Hopkinson & Lapworth from Llanvirn (SM . Ai740o) is a poorly-preserved, bent dichograptid (O. M. B. Bulman, personal communication). The specimen from Abereiddy Bay refigured by Elles & Wood may well come from the locality referred to the " Dicranograptus Shales " (Cox, 1915 : 304) and not from the D. murchisoni Shales for which the area is best known. If this is so, the species ranges from PLlandeilo Series to Caradoc Series (wilsoni Zone). It has been recorded from Australia but Thomas (1960) does not list it in his range chart and so presumably does not consider it to be present. Linnarsson recorded it from Scania but Hadding (1913) transferred this form to his new species D. vagus. Dicranograptus clingani Carruthers PI. 3, figs. 2-4; Fig. 3a 1868 Dicranograptus Clingani Carruthers : 132, pi. 5, figs. 6a-c. 1870 Dicranograptus Clingani Carruthers; Hopkinson : 358, pi. 16, figs. 4a-c. 1876 Dicranograptus Clingani Carruthers; Lapworth, pi. 3, fig. 76. 1877 Dicranograptus Clingani Carruthers; Lapworth : 141, pi. 6, fig. 43. 1904 Dicranograptus Clingani Carruthers; Elles & Wood : 165, pi. 24, figs. ia-1. 1915 Dicranograptus Clingani Carruthers; Hadding : 22, pi. 3, figs. 1-8. ORIGINAL DESCRIPTION. " Polypary with a short diprionidian portion, the proxi- mal end furnished with three very delicate spines; hydrothecae forming a slight serration along the margin; twenty-one cells in the inch. Loc. Moffat." COMMENTS ON USAGE. Elles & Wood put this species in a group by itself on the basis of the thecal characters approximately straight ventral walls and horizontal W. CARRUTHERS' TYPE GRAPTOLITES 191 apertures. These characters are clearly seen in Hopkinson's figures and serve to differentiate the species from other forms with a short biserial portion. TYPE MATERIAL. Elles & Wood refigured as " type specimen " the original of Carruthers' fig. 6a and this can be taken as a designation of a lectotype. The specimen, Q-55 (PI. 3, figs. 3, 4), is from the Hartfell Shales, Hartfell Spa. Carruthers' fig. 6b, Q.842 (PL 3, fig. 2), is also from this locality. REVISED DESCRIPTION. Dicranograptus with short biserial portion consisting of three or four pairs of thecae, and short, straight uniserial stipes diverging at about 40. The biserial portion has a uniform breadth of about i mm. and the branches are similarly uniform at about 0-8 mm. The virgella is usually prominent as a short spine and the first two thecae may have sub-apertural spines. The ventral walls of the later thecae are straight and the apertures are horizontal in excavations which occupy about one-third of the breadth. The uniserial stipes appear to be rarely more than 2 cm. long although Hadding figures a specimen with stipes nearly 4 cm. long. The distal thecae number eight to ten per cm. but the stipes are usually twisted so that the thecae are in scalariform view (Fig. 3a) and it is difficult to decide the precise thecal shape. The apertural excavations are clearly marked by lists and it is probable that the straight ventral walls of the proximal thecae continue in the distal ones. FIG. 3. a. Dicranograptus clingani Carr. Proximal end of paratype. (3.842. X5- The right hand side of the specimen is poorly preserved, b. Climacograptus minutus Carr. Lectotype. Q8o. X5- c. Cryptograptus tricornis Carr. Proximal end of lectotype showing basal spines. X5- DISTRIBUTION. The species is common in the Lower Hartfell Shales of the Moffat area and is also found at Conway. Elles & Wood recorded it from equivalent beds in Ireland. It is found in Scandinavia but has not been recorded from North America. Thomas (1960) gives records but omits it from his range chart, so casting doubt on the records. The type specimen has no associates but Elles & Wood recorded it only from its own zonal association. Climacograptus minimus (Carruthers) PI. 4, fig. 3; Fig. 4c 1868 Diplograpsus minimus Carruthers : 74, 130, pi. 5, figs. iaa, b. ?igo6 Climacograptus minimus (Carruthers) Elles & Wood : 191, pi. 27, figs. xa-g. 192 W. CARRUTHERS' TYPE GRAPTOLITES ORIGINAL DESCRIPTION. " This agrees with D. pristis in general appearance, and in the form and arrangement of the cells, except that the whole polypary and all its parts are so very small. Had I met with only a few specimens, I would have con- sidered it as merely an accidental variety, but I have seen so many, all agreeing in size, that I cannot doubt that it is a good species, especially as young specimens of D. pristis early attain their full breadth, and the increase of the polypary is by addition to its distal end, and not to the size of the already formed hydrothecae, just as in the living Sertulariadae. About thirty-eight cells to one inch. Loc. Moffat." FIG. 4. Figures to show Carruthers' clear distinction between his two small species of diplograptid. a, Copy of original MS drawing of " Diplograptus minimus " which has appended note "38 to inch ". b, Similar copy of Climacograptus minutus which has notes " 32 to 40 to an inch, nearly opposite ". COMMENTS ON USAGE. In his 1868 paper, Carruthers clearly distinguished two small biserial species and Fig. 4, taken from his notebook, illustrates his ideas of the differences, one clearly climacograptid, the other what he called diplograptid (now orthograptid) . His type slab however shows a large number of small, poorly- preserved rhabdosomes which appear to be almost all climacograptid and Elles & Wood certainly regarded his species in this light. In his description of the species (1868 : 130) he called it Diplograpsus minutus although elsewhere, in the explanation of the plate published in the first part of the paper (1868 : 74), he called it D. mini- mus, agreeing with his MS notes, and usage of this prior name avoids the homonym which results from the transference of the species to Climacograptus. Elles & Wood did not re-figure Carruthers' material, and their specimens, from the Hartfell Shales (clingani and linearis Zones), reach a breadth of 2 mm. Carruthers' specimens on the type slab are only about I mm. wide and are associated with some slender uniserial stipes, which look very like Monograptus sp. (atavus or acinaces type). It thus seems probable that Elles & Wood's species is not the same as Carruthers' but it is clear from the foreign references (e.g. Ross & Berry 1963, Obyt & Sobolevskaya 1964) that it is Elles & Wood's species which is nowadays W. CARRUTHERS 1 TYPE GRAPTOLITES 193 recognized under this name. Carruthers' original notes are of little use since the only measurements he gives are the thecal numbers per inch. A sketch of " Diplogr. minimus " clearly shows thecae of an orthograptid type in accordance with his comparison of it with D. pristis but, as mentioned above, the common form on the type slab appears to have climacograptid thecae. The species must remain for the moment in an unsatisfactory state since a proper appraisal of Elles & Wood's form must await critical re-examination of the Upper Ordovician climacograptids. TYPE MATERIAL. Q.82, a slab crowded with poorly-preserved specimens. REVISED DESCRIPTION. The rhabdosome is about i mm. wide and up to 10 mm. long. Thecal details are obscure. HORIZON. As noted above, the type slab shows slender uniserial stipes very similar to monograptids of the vesiculosus and cyphus Zones, in contrast with the description by Elles & Wood of the species from the Hartfell Shales (dingani and linearis Zones) . Climacograptus minutus Carruthers PL 4, fig. i, ; Figs 3b, 4b 1868 Climacograptus minutus Carruthers : 132, pi. 5, fig. loa. ?i9o6 Climacograptus minutus Carruthers; Elles & Wood : 211, pi. 27, figs. i2a-c. ORIGINAL DESCRIPTION. ' This is a very minute but well-marked species, never attaining a greater size then represented on the Plate. There are at the rate of from thirty-two to forty cells in the space of an inch. Loc. Moffat." COMMENTS ON USAGE. This species does not seem to have been widely recorded, probably because of its unsatisfactory nature. Carruthers' original specimens are of widely differing widths and Elles & Wood based their account of the species on other material from the British Museum (Natural History). Packham (1962) revised some of the British Silurian diplograptids but did not discuss this species. His C. tangshanensis linearis appears to be close to Carruthers' fig. lob (1868). TYPE MATERIAL. Carruthers' original specimens can be recognized and appear to be of two different forms. His fig. loa (Q.8o) is here selected as lectotype (PL 4, fig. i) . The original of fig. lob (Q . 1372) is a narrower form and fits well with Climacograptus scalaris miserabilis Elles & Wood. Elles & Wood's figured material (Q . 849) is all somewhat distorted and it is clear that they did not use Carruthers' original specimens for their description, probably through some confusion with " Diplogr aptus minutus ". REVISED DESCRIPTION. Rhabdosome 7 mm. long. 1-2 mm. broad; thecae thirteen per cm., with large excavations occupying about one quarter of the breadth of the rhabdosome and about equal to the length of the free ventral wall. The apertures are nearly opposite each other (as recorded in Carruthers' notes) and not alternate as stated by Elles & Wood. The proximal end of the type specimen is poorly preserved but there appears to be a stout virgula which is prolonged beyond the distal end of the rhabdosome. The type specimen is completely flattened and it is not possible to make out any details of a median septum. 194 W. CARRUTHERS' TYPE GRAPTOLITES The uniform breadth of the rhabdosome and the large opposite excavations com- bine to make this form quite distinct from other British climacograptids. Un- fortunately the range of variation is not known and since the type specimen has no associates, its precise horizon is also doubtful. It remains therefore an unsatis- factory species. Cryptograptus tricornis (Carruthers) PL 4, figs. 4-6; Fig. 30 1858 Diplograpsus tricornis Carruthers : 468, fig. 2. 1859 Diplograpsus tricornis Carruthers; Carruthers 1867 Diplograpsus tricornis Carruthers; Carruthers 1868 Diplograpsus tricornis Carruthers; Carruthers 25, fig- 2. 290, pi. i, figs. loa-d. 131, pi. 5, fig. ua, b. 1880 Cryptograptus tricornis (Carruthers) Lapworth : 171, pi. 5, figs. 27a-e. 1908 Cryptograptus tricornis (Carruthers); Elles & Wood : 296, pi. 32, figs. I2a-d. 1908 Cryptograptus tricornis (Carruthers) ; Ruedemann : 443, pi. 28, figs. 1-4. 1913 Cryptograptus tricornis (Carruthers); Hadding : 40, pi. 2, figs. 13-14. 1915 Cryptograptus tricornis (Carruthers); Hadding : 325, pi. 6, fig. 15. 1934 Cryptograptus tricornis (Carruthers); Hsu : 87, pi. 6, figs. I3a-m. 1937 Cryptograptus tricornis (Carruthers) ; Bulman : 5, t-fig. 8. 1945 Cryptograptus tricornis (Carruthers); Bulman : 29, pi. 2, figs. r-8. 1960 Cryptograptus tricornis (Carruthers); Thomas, pi. 6, fig. 69. ORIGINAL DESCRIPTION. " This species can be readily distinguished by the three spines which adorn its base, and which are almost always preserved. The central spine is a continuation of the line of the axis ; it is shorter than the lateral ones . . . The polypidom is more slender than in D. foliaceus, which in general outline it some- what resembles. The axis is slender, and produced beyond the other parts of the fossil. The cell-walls are well marked, extending upwards from the axis to the boundary of the fossil. Each cell forms a rhomb whose outer border is slightly indented, giving the boundary of the fossil a faintly serrated aspect. When the fossil is preserved so as to show the serratures, the spines are so compressed that the central one is almost or altogether lost. When the spines are well preserved and in the position described, no traces of the individual cells are discernible ; the boundary of the fossil is an unbroken line ". In 1868, Carruthers added " When I described this species I had not detected the mouths of the cells in those specimens in which they should have been shown on the upper surface. In more perfectly preserved specimens since obtained these have been beautifully shown ". COMMENTS ON USAGE. This species has been widely recognized as it is easily identified from the original description and figures. Lapworth (1880) discussed the varying appearance of the thecae at some length and introduced the genus Crypto- graptus for this and allied species. Hadding (1915) showed by comparison with Glossograptus that the two stipes were in lateral contact (the monopleural arrange- ment of Jaanusson 1960) and removed the genus from the Diplograptidae. The structure of the proximal end was not, however, elucidated until 1938 when Bulman described isolated specimens from the Balclatchie Limestone and the two different W. CARRUTHERS' TYPE GRAPTOLITES 195 aspects of the basal spines noted by Carruthers was explained. A number of varieties have been described, differing mainly in the breadth of the rhabdosome. DIAGNOSIS. Rhabdosome biserial, monopleural, 2-4 cm. long widest at base when preserved in lateral view, maximum width 1-8 mm. but typically narrower. Thecae ten to twelve per cm. Basal spines conspicuous, but short. TYPE MATERIAL. BM(NH) (^.1299, presented to the museum in 1860, is almost certainly Carruthers' original slab. It is crowded with specimens up to 35 mm. long and 1-6 mm. broad but it is not possible to recognize the original figured speci- mens. Accordingly, one of the better specimens has been selected as lectotype (PI. 4, fig. 4). The association on the slab includes abundant Corynoides calicularis Nich. Similar pieces in the Lap worth Collection (Birmingham University) are labelled " The Cornice, Hartfell " and, although this is not marked on Lapworths' published map of Hartfell, it would appear from the text that the horizon is lowest Hartfell, zone of Climacograptus wilsoni. DESCRIPTION. No detailed description is needed since that in the Monograph is satisfactory and has been recently supplemented by Bulman's detailed account of isolated specimens. There seems to be some variation in the thecal count, speci- mens from Girvan having consistently higher (twelve to sixteen per cm.) counts. Elles & Wood described the variety schaeferi which Lapworth had figured earlier and claimed that it was " somewhat wider than the typical form ". Examination of the material in Lapworth's collection does not confirm this difference but there seems to be a difference in thecal shape and the basal spines are not conspicuous. DISTRIBUTION. This species has been recorded from all continents and some of the more recent records are given in the synonymy. The stratigraphic range was given by Elles & Wood as Arenig Series (extensus Zone) to Caradoc Series (clingani Zone), a longer range than any other Ordovician species, and it is possible that critical examination of a large number of specimens from the lower horizons might show them to be distinct from the typical form from the early Caradoc. Sherrard (1954) records it from the zone of P. linearis but no illustration is given to confirm this. Monograptus clingani (Carruthers) PL 5, figs. 1-5 1867 Graptolithus Clingani Carruthers : 369, pi. 2, fig. 8. 1868 Graptolithus Clingani Carruthers; Carruthers : 127, pi. 5, figs, iga, b. 1876 Monograptus Clingani (Carruthers); Lapworth, pi. i, fig. 24. 18760 Monograptus Clingani (Carruthers); Lapworth : 501, pi. 20, figs. sa-c. non 1897 Monograptus Clingani (Carruthers); Perner, pi. n, figs. 15-17. 1913 Monograptus Clingani (Carruthers); Elles & Wood : 463, pi. 46, figs. na-f. 1951 Monograptus clingani (Carruthers); Bulman : 322, t-fig. 5. 1956 Monograptus clingani (Carruthers); Bondarenko & Keller : 91, t-fig. 2. ORIGINAL DESCRIPTIONS. 1867. "... a beautiful small species, which at first I referred to G. millepeda, M'Coy, but that species is certainly the proximal end of G. Becki, and this differs from it in having a very broad common base, from which the hydrothecae rise ." 1868. " Polypary, small and arcuate, with a broad common I 9 6 W. CARRUTHERS' TYPE GRAPTOLITES canal, and slender somewhat recurved hydrothecae. This beautiful little graptolite I long supposed to be only the proximal portion of some other species, but the large number I have met with, all equally perfect, none larger than fig. iga, and many showing the prolongation of the axis beyond the distal end, together with the great breadth of the common canal (forming two-thirds of the breadth of the whole polypary), unlike the early portion or proximal fragment of any graptolite with which I am acquainted, have induced me to consider it a good species ..." COMMENTS ON USAGE. Carruthers' descriptions gave no real details of the species. From the syntypes it appears that two different species may be confused but as both are young rhabdosomes it is not possible to be sure. The characters of the proximal thecae are not determinable from the lectotype, chosen by Pfibyl (1948), which is completely flattened but they appear to be of the priodon type with well marked hooks to the apertures (PI. 5, fig. 5). The distal thecae are similar. Lapworth recorded a " Clingani Band ", some 6 inches thick, in his Dobbs Linn sections and numerous specimens in his collection bear this label. In a recent resurvey of the section, Toghill (1965) recognized the unit again but recorded M. dingani from a wider horizon. TYPE MATERIAL. The specimen (Q.87) figured by Carruthers as fig. iga was referred to as " type specimen " by Elles & Wood in the explanation of their plate and the second specimen called " co-type " (Q.8-4), leaving them perhaps of equal status. Pfibyl (1948) however clearly stated that Carruthers' fig. iga is the lecto- type which is unfortunate as this specimen has a broader common canal than in most of the other specimens referred to the species. This may be only a preserva- tional feature as specimens in the Lapworth collection from DufTkinnel, the type locality, are all of the common type in which the hooks occupy at least half of the breadth of the rhabdosome. REVISED DESCRIPTION. Rhabdosome dorsally curved at the proximal end, dis- tally becoming more or less straight; widening from an initial breadth of 0-6 mm. to a maximum of about 1-5 mm. within the first 6 or 7 mm. The thecae are of priodon type with about one-third of the length involved in the hook which occupies half of the breadth of the stipe. The shape of the thecae varies a good deal with the type of preservation but there appears to be no overlap of the thecae (Bulman 1951, fig. 5), the broad prothecal portion occupying the whole of the breadth of the rhabdosome. DISTRIBUTION. Lapworth recorded the species in abundance