oe oe a Pe es Sx ore eS Peete eee — me Sorat ee a a ee 4 th a le aaa lic as « \) TE la het Bulletin of the British Museum (Natural History) Upper Cretaceous ammonites from the Calabar region, south-east Nigeria P. M. P. Zaborski ie a : i 7. Geology series Vol 39 No1 27 June 1985 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four scientific seriesy Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and an Historical series. Pape inthe Bulletin are primarily the results of research carried out on the unique and ever-growing collections of the Museum, both by the scientific staff of the Museum and by specialists from elsewhere who make use of the Museum’s resources. Many of the papers are works of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself, available separately, and individually priced. Volumes contain about 300 pages and several volumes may appear within a calendar year. Subscriptions may be placed for one or more of the series on either an Annual or Per Volume basis. Prices vary according to the contents of the individual parts. Orders and enquiries should be sent to: Publications Sales, British Museum (Natural History), Cromwell Road, London SW7 5BD, England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Geol.) © Trustees of the British Museum (Natural History), 1985 The Geology Series is edited in the Museum’s Department of Palaeontology Keeper of Palaeontology: DrH. W. Ball Editor of the Bulletin: Dr M. K. Howarth Assistant Editor: Mr D.L. F. Sealy ISBN 0 565 07006 1 ISSN 0007-1471 Geology series British Museum (Natural History) Vol 39 No 1 pp 1-72 Cromwell Road London SW7 5BD Issued 27 June 1985 Upper Cretaceous ammonites from region, south-east Nigeria 491450 S35 ive : a] Dee eae P. M. P. Zaborski C8GINNL 2 Z Contents SORT OMIDS [SMES at ets saccisscle Bis. wots ssiciauciewra se ECE RSLNS Sa caleisladieslaetapiealne QuleRlaceeideemace saab sis NTTOGCICHIOM ."s sdadetoodacbadsese coscasecboc sheen accede scROn rece Sree eee RRO TORTS eer Te Mma. (ames SLIM AULC KE SCHIP ULOMS) aa. serer tien toad eth racusiileaas seiscwe dasorease yocteGarith eames tees kooe dl sce en BamilvaGaudny ceratidae Spath...ce-ntaas.asrnoe-ae02/4- seuinetecen totes asaceasecoaseineashassdhseses GremHSPARAPALATYCEVGS SUIMUZU .5..¢.¢...<-..1a;cateass poe son decabawedsahiede cage GEN Anagaudryceras involvulum (Stoliczka) GENUS GALOBTGATs { GHOSSOWNTCE: sgoncsdeososdsdeeeboenedepacpodesecconsacneconernotertar sa: Gardnyceras beantalyense:@ollignont eeeacceasssaaeeseee eee ase cea ee eee Gaudryceras varicostatum van Hoepen Heeatiraiypa MUTT AC INTC KS Sees cataiieiir-snsle feejfesisen slnaid onied Seisecuatie ves Gate meeateRcmt ae tn oeaeee Solerrmiky torial heres (Grill Ze kesacsosontendersa os concresececostopcerecsecssocrascsacodbapconos. Gems TUTE ATTEN + le eyrusfoeserac 97 Pseudaspidoceras footeanum (Stoliczka) ............. 000. cece eee eee 98 Bamilvavascoceradac DOUVINE sca ceic cs ce eis cs Mave os © cescls es been oe 100 Genusaigascoceras Choltat .c aatiics cee dace teense e ses tek eet neces ges 100 Vascoceras (Paravascoceras) harttii (Hyatt).........00... cece eee ee eee 100 RSTISTETN GES a 6. oid.0.0.6 c.oless cig. EMRE OAR CIRC OF ARETE Ro Ine ae eae ae ee 101 LRG ue oils duce dita cake Bel cle Oaceetic tic ROG oil ONE Rene Val ct are ne Oe 104 Bull. Br. Mus. nat. Hist. (Geol.) 39 (2): 73-105 Issued 29 August 1985 1 BRITISH ML 74 M. K. HOWARTH Synopsis About one hundred ammonites collected from 24 localities near Novo Redondo indicate various dates in the Cenomanian and Lower Turonian. The Lower Cenomanian ammonites include large examples of Sharpeiceras, especially S. florencae, and also many specimens of Desmoceras latidorsatum lemoinei, and species of Ostlingoceras and Mariella. Middle Cenomanian ammonites include large specimens of Euomphaloceras (E.) cunningtoni and Forbesiceras obtectum. One locality has a mixture of the Upper Cenomanian ammonites Calycoceras naviculare and Euomphaloceras (Kanabiceras) septemseriatum, and the Lower Turonian ammonites Pseudaspidoceras footeanum, Vascoceras (Paravascoceras) harttii and Proplacenticeras stantoni. These are the first records of ammonites in Angola referable to the middle part of the Lower Turonian. Introduction The rich ammonite faunas from the Cretaceous rocks of the coastal regions of Angola have been described by many authors. All the stages from the Albian to the top of the Cretaceous contain ammonites, and an earlier review of the faunas by Howarth (1965: 340-343, 400-405), was followed by an up-to-date review of the biostratigraphy of the Albian to Turonian stages by Cooper (19785). Novo Redondo is about 270 km south of Luanda, and lies in the southern half of the Cuanza sedimentary basin. Lower and Middle Cenomanian ammonites from the sur- rounding district were described by Cooper (1973), and included notable examples of the genera Turrilites, Euomphaloceras and Forbesiceras. The new collection of about 100 ammonites described here includes more examples of these three genera, together with a fauna of large specimens of the Lower Cenomanian genus Sharpeiceras, of which only one Angolan specimen has been described and figured before (Haas 1942). Also included is a small number of ammonites from a locality a short distance south of Novo Redondo that contains species of both Upper Cenomanian and Lower Turonian age. The latter species are slightly younger than the basal Turonian ammonites of Salinas, revised by Cooper (1978a), and older than the mid- Turonian ammonites of Mogamedes that were described by Howarth (1968); they fill an age gap that has not hitherto been represented by ammonites in Angola. The collecting localities, which are shown in Fig. 1, all lie within sheet 184, Novo Redondo, of the 1:100000 geological survey maps of Angola (Lapao 1972). In fact the localities lie between 10 km south and 25 km NNE of Novo Redondo, and can be readily plotted on the detailed geological map of sheet 184; Fig. 1 outlines the geological relationships in the area. Pre- Cambrian basement rocks to the east are overlain by Cretaceous rocks in a coastal strip, here 25-30 km wide, which start with non-marine Aptian sediments, overlain by marine Albian, Cenomanian, and some representatives of other stages, up to the top of the Cretaceous. The total thickness of the Cretaceous is 700-1000 m. More details of the stratigraphy within each stage, and the ammonites from which the age had been determined, were given by Lapao (1972: 10-32), but no details are available of the stratigraphy at each ammonite locality. They appear to represent single dates, except for the mixture of Upper Cenomanian and Lower Turonian ammonites at locality 1007. North of Novo Redondo the Lower Cenomanian is shown separately on Lapao’s map (1972), but the area it occupies does not entirely accord with the ages given by the ammonites at some localities. Localities 67, 72 and 76 near the northern edge of the map contain the Lower Cenomanian ammonite Sharpeiceras, but they occur within the area shown on the map as Middle/Upper Cenomanian. A few km south of Novo Redondo the facies becomes somewhat different and the Cenomanian is not divided. No divisions within the Cenomanian are shown on the sketch map of Fig. 1: instead, the Lower or Middle Cenomanian dates suggested by the ammonites are shown by lettering on appropriate areas of the map. It can be seen that all the localities north of Novo Redondo are Lower Cenomanian, while those south of Novo Redondo are mainly Middle Cenomanian in age. CENOMANIAN & TURONIAN AMMONITES FROM ANGOLA REDONDO |® LC e301 275. MC , 263 \ LT e430 ole 1007 © 953 Recent, Pleistocene Senonian Turonian, Cenomanian Albian (top) Lower Turonian Upper Middle Cenomanian Lower Collecting localities 75 Fig. 1 Map of the collecting localities and an outline of the geology in the area around Novo Dis Redondo, Angola. 76 M. K. HOWARTH Localities, faunas and ages Lower Cenomanian Localities 67, 23-5 km NNE of Novo Redondo; 72, 20 km NNE of Novo Redondo; 100, 15 km NNE of Novo Redondo; 188, 6 km NE of Novo Redondo; and 526, 11-5 km NE of Novo Redondo Sharpeiceras florencae Spath Locality 76, 19-5 km north of Novo Redondo Phylloceras (Hypophylloceras) seresitense Pervinquiére Sharpeiceras florencae Spath Puzosia sp. indet. Localities 86, 16-5 km NNE of Novo Redondo; and 118, 11-5 km NE of Novo Redondo Desmoceras (D.) latidorsatum lemoinei Collignon Puzosia sp. indet. Locality 88, 20 km NNE of Novo Redondo Desmoceras (D.) latidorsatum lemoinei Collignon Sharpeiceras florencae Spath Puzosia sp. indet. Locality 90, 18 km NNE of Novo Redondo Sharpeiceras mexicanum (Bose) Localities 99, 15-5 km NNE of Novo Redondo; and 102, 16-5 km NNE of Novo Redondo Sharpeiceras florencae Spath S. mexicanum (Bose) Locality 301, 2-5 km SSE of Novo Redondo Ostlingoceras (O.) cf. rorayense (Collignon) Locality 350, 8-5 km NNE of Novo Redondo Phylloceras (Hypophylloceras) seresitense Pervinquiére Mariella (M.) oehlerti (Pervinquieére) Desmoceras (D.) latidorsatum lemoinei Collignon Locality 458, 8-5 km NNE of Novo Redondo Phylloceras (Hypophylloceras) seresitense Pervinquiére Desmoceras (D.) latidorsatum lemoinei Collignon Gaudryceras sp. indet. Most of the genera and species found at these localities are accurate indicators of a Lower Cenomanian age. Sharpeiceras is probably confined to the Lower Cenomanian, and S. florencae occurs in the Carcitanensis Zone in England, the basal zone of the Cenomanian in the scheme adopted by Kennedy & Hancock (1978: V.14). The subspecies Desmoceras latidorsatum lemoinei is characteristic of the Lower Cenomanian, as are also the two Turrilitidae species Ostlingoceras (O.) rorayense and Mariella (M.) oehlerti. Therefore a Lower Cenomanian age is applicable to all 15 localities listed above. The Puzosia sp. indet. recorded at localities 76, 86 and 88, and the Gaudryceras sp. indet. at locality 458, are based on poorly-preserved fragmentary specimens that are not specifically identifiable; they are not described in the systematic section of the paper. Localities of probable Lower Cenomanian age Locality 89, 18-5 km north of Novo Redondo Gaudryceras sp. indet. CENOMANIAN & TURONIAN AMMONITES FROM ANGOLA 77 Localities 162, 9-5 km NNE of Novo Redondo; and 319, 8-5 km NNE of Novo Redondo Phylloceras (Hypophylloceras) seresitense Pervinquiére These three localities can be dated as Cenomanian, and probably Lower Cenomanian, by their proximity to other localities that are definitely dated as Lower Cenomanian (localities 89 and 162 are near to 76 and 350 respectively), and both the ammonites are associated with Lower Cenomanian species at other localities. Middle Cenomanian Locality 263, 4-5 km SE of Novo Redondo Euomphaloceras (E.) cunningtoni (Sharpe) var. meridionale (Stoliczka) Locality 275, 5 km SE of Novo Redondo Forbesiceras obtectum (Sharpe) E. (E.) cunningtoni (Sharpe) var. meridionale (Stoliczka) Locality 419, 8 km south of Novo Redondo Turrilites (T.) acutus Passy Forbesiceras obtectum (Sharpe) Locality 430, 5 km SE of Novo Redondo E. (E.) cunningtoni (Sharpe) var. meridionale (Stoliczka) Locality 953, 6 km SE of Novo Redondo Turrilites (T.) acutus Passy Forbesiceras obtectum (Sharpe) E. (E.) cunningtoni (Sharpe) var. meridionale (Stoliczka) Only three species of ammonites occur at these five localities, but the presence of all three together at locality 953 suggests that all five localities are of approximately the same age. The most accurate indicator of that age is probably Turrilites (T.) acutus, which occurs most commonly in the Acutus Zone, the middle one of the three zones of the Middle Cenomanian in north-west Europe (Kennedy 1971: 102; Kennedy & Hancock 1978: V.15). The genus Euomphaloceras first appeared in the Middle Cenomanian, and EF. cunningtoni occurs in the Acutus Zone, though it is more abundant in the underlying Costatus Zone (Kennedy 1971: 90, 102). Forbesiceras obtectum also occurs in the Acutus Zone, and well as at lower levels in the Middle and the top of the Lower Cenomanian (Kennedy 1971: 48, 102). If a single age can be assigned to these five localities, it is most probably Turrilites acutus Zone, Middle Cenomanian. The underlying Turrilites costatus Zone is a less likely possibility. Upper Cenomanian and Lower Turonian Locality 1007, 5-5 km south of Novo Redondo Proplacenticeras stantoni (Hyatt) var. bolli (Hyatt) Calycoceras (C.) naviculare (Mantell) Euomphaloceras (Kanabiceras) septemseriatum (Cragin) Pseudaspidoceras footeanum (Stoliczka) Vascoceras (Paravascoceras) harttii (Hyatt) Both Upper Cenomanian and Lower Turonian strata are present at this locality. Calycoceras naviculare occurs in the Naviculare Zone and in the overlying Geslinianum Zone of the Upper Cenomanian, but neither it nor the genus Calycoceras go higher than the top of the Cenomanian (Wright & Kennedy 1981: 33). Euomphaloceras septemseriatum is characteristic of the Gesli- nianum Zone and is probably confined to that horizon. These two species indicate the presence of at least the Geslinianum Zone of the Upper Cenomanian, but there is no evidence for the uppermost zone of the Cenomanian, i.e. the Juddii Zone, in the scheme adopted by Wright & Kennedy (1981: 8). Of the other three species, Pseudaspidoceras footeanum is the clearest 78 M. K. HOWARTH indicator of a Lower Turonian age, because the genus does not occur below the Turonian, and P. footeanum itself occurs in the Lower Turonian of Israel (Freund & Raab 1969: 14, 73), England (Wright & Kennedy 1981: 82) and southern India. The genus Vascoceras first appears in the Geslinianum Zone, Upper Cenomanian, but it is more characteristic of the Turonian, and V. (Paravascoceras) harttii is similar to species that occur near the top of the Lower Turonian in Portugal (see p. 101). Species of Proplacenticeras first occur in the Upper Cenomanian, Geslinianum Zone of Europe (Wright & Kennedy 1981: 123), but P. stantoni var. bolli is nearly identical to specimens that are found in the Lower Turonian Britton Formation in Texas (Moreman 1942: 192, Young & Powell 1978: XXV.18). So the evidence of the ammonites at locality 1007 suggests that beds of both Upper Cenomanian, Geslinianum Zone, and Lower Turonian age are present. Systematic descriptions All the specimens are in the collections of the British Museum (Natural History) and catalogue numbers have the prefix letter C. Measurements are given in the order diameter (D), whorl height (Wh), whorl breadth (Wb), diameter of the umbilicus (U), and figures in brackets are proportions of the diameter. Suborder PHYLLOCERATINA Arkell, 1950 Family PHYLLOCERATIDAE Zittel, 1884 Genus PHYLLOCERAS Suess, 1865 Subgenus HYPOPHYLLOCERAS Salfeld, 1924 Type species. Phylloceras onoense Stanton, 1895. Phylloceras (Hypophylloceras) seresitense Pervinquicre Fig. 2 1907 Phylloceras velledae var. seresitense Pervinquiére: 52. 1910 Phylloceras velledae var. seresitense Pervinquiére; Pervinquieére: 9; text-fig. 2; pl. 1, figs 1-3. 1977 Phylloceras (Hypophylloceras) seresitense Pervinquiére; Kennedy & Klinger: 363-365; pl. 4, figs 6, 7; pl. 6, fig. 4; pl. 7, fig. 4; pl. 9. 1978 Phylloceras (Hyporbulites) seresitense seresitense Pervinquiére; Collignon: 2; pl. 1, fig. 1. 1979 Phylloceras (Hypophylloceras) seresitense Pervinquiére; Cooper & Kennedy: 177-181; figs 1, 2, 31 (q.v. for full synonymy). MATERIAL. Six specimens, C.80994-5 (loc. 76), C.81010 (loc. 162), C.81020 (loc. 319), C.81034 (loc. 350) and C.81046 (loc. 458). MEASUREMENTS. D Wh Wb Wb: Wh C.80995 136 81-6 (0-60) 39-3 (0-29) 0-48 C.80994 — 70:3 37-5 0-53 REMARKS. The best specimen is the half ammonite, 136 mm in diameter, figured in Fig. 2. C.80994 is a quarter-whorl fragment of similar size, and the remaining four are much smaller fragments. They are referred to P. (H.) seresitense on the basis of their general whorl propor- tions, the slightly flexuous striate ornament, and the tetraphylloid saddle endings of the suture lines. The umbilicus appears to be nearly or completely closed, and low, slightly flexuous constrictions can be seen on the figured specimen. This species has been fully discussed by Kennedy & Klinger (1977) and Cooper & Kennedy (1979), and in the latter paper specimens from the Upper Albian of Angola were described and figured. A small example from the Upper Albian was also figured by Collignon (1978: pl. 1, fig. 1). The specimens now described extend the range of the species in Angola up into at least the Lower Cenomanian. Many European examples were described in three papers by Wiedmann (1962a, 1962b, 1964), who divided the species into three subspecies according to the degree of CENOMANIAN & TURONIAN AMMONITES FROM ANGOLA 79 compression of the whorls and the size of the umbilicus. Cooper & Kennedy (1979: 180; fig. 2) found considerable variation in whorl compression, and did not accept that the subspecies divisions were valid. This view seems to be supported by the present few specimens which have whorl breadth/height ratios of 0-48 and 0-53 at about 120-135 mm diameter, and agree with Cooper & Kennedy’s (1979: 181; fig. 2) graph of whorl dimensions extrapolated to slightly larger diameters. OccurRENCE. This species ranges from the Upper Aptian to the middle of the Cenomanian. Suborder ANCYLOCERATINA Wiedmann, 1960 Superfamily TURRILITACEAE Gill, 1871 Family TURRILITIDAE Gill, 1871 Genus OSTLINGOCERAS Hyatt, 1900 Type SPECIES. Turrilites puzosianus d’Orbigny, 1842. Ostlingoceras (Ostlingoceras) cf. rorayense (Collignon) Fig. 5 1964 Turrilites rorayensis Collignon: 49; pl. 330, fig. 1479. 1975 Ostlingoceras (O.) rorayense (Collignon); Forster: 186; pl. 6, figs 7, 8. 1978 Ostlingoceras (O.) rorayense (Collignon); Klinger & Kennedy: 11; pl. 1, figs M, N; pl. 5, fig. G; pl. 8, fig. E. MATERIAL. One specimen, C.81019, from loc. 301. Remarks. This is a fragment, only 13 mm high, of three small whorls that are sinistrally coiled, have a small apical angle, and almost flat sides to the whorl. The ornament consists of weak ribs just below the upper whorl suture which end in a row of tubercles about the middle of the flank, then a narrow smooth band down to the second row of tubercles just above the lower whorl suture. A third row of tubercles is not visible, but could be hidden by the overlap of the whorls. The close coiling, low apical angle, flat sides, weak ribs and small tubercles suggest that this small fragment belongs to the subgenus Ostlingoceras, and comparison is made with Collignon’s species which has been revised by Klinger & Kennedy (1978) on the basis of more than 150 specimens from the Lower Cenomanian of Zululand. OccurRENCE. The Lower Cenomanian of Madagascar, Mozambique, Zululand and Angola. Genus MARIELLA Nowak, 1916 Type species. Turrilites bergeri Brongniart, 1822. Mariella (Mariella) oehlerti (Pervinquicre) Figs 4, 6 1907 Turrilites gresslyi Pictet & Campiche; Boule, Lemoine & Thévenin: 57; pl. 13, fig. 2. 1910 Turrilites oehlerti Pervinquiére: 53; pl. 5, figs 14-17. 1929 Turrilites gresslyi Pictet & Campiche; Collignon: 41; pl. 6, fig. 15. 1929 Turrilites oehlerti Pervinquiére; Collignon: 41; pl. 6, figs 16, 17. 1964 Turrilites oehlerti Pervinquiére; Collignon: 15; pl. 320, figs 1398, 1399. 1978 Mariella(M.) oehlerti oehlerti (Pervinquiére) Klinger & Kennedy: 31; pl. 3, fig. E; pl. 4, fig. E; pl. 6, figs H-N; pl. 7, fig. G; pl. 8, figs G, H. MatTERIAL. Three specimens, C.81035-7, from loc. 350. DescriPTIon. The collection consists of one fragment of half a large whorl of about 55 mm diameter, and two small fragments showing parts of two and three whorls respectively from about 6 to 25 mm diameter. All are coiled sinistrally. The whorl of the larger specimen has a 80 M. K. HOWARTH CENOMANIAN & TURONIAN AMMONITES FROM ANGOLA 81 flattened outer side and is angled at the upper whorl suture but more rounded at the lower one. The smaller specimens have more rounded whorls. The main ornament consists of four rows of tubercles, the upper row just above the middle of the whorl, the next row just below the middle of the whorl, and the two lowest rows much closer to each other and near the lower whorl suture. The lowest row is usually hidden by the overlap of the next larger whorl. There is an equal number of tubercles in each row, aligned obliquely in the direction of the spire. The tubercles in the lowest row are smaller than the others. The tubercle density is about 15 per half whorl in the larger fragment and about 12 per half whorl in the smaller ones. Bands between the tubercles and up to the upper whorl suture are smooth, and ribs are only weakly developed on the bottom of the whorl in the zone of overlap. No suture-lines are visible. Remarks. These specimens are not sufficiently well preserved to show whether they belong to the nominate subspecies or to the subspecies sulcata of Klinger & Kennedy (1978: 33; pl. 3, fig. D; pl. 8, fig. D). The latter differs from the nominate subspecies in having a spiral groove between the second and third rows of tubercles. Such a groove is not present in these Angolan examples, although there appears to be a groove between the top and second rows in the largest specimen which might be a preservational feature. All the morphological features that can be seen, including the density of tubercles, match those of Pervinquiére’s species, which was revised by Klinger & Kennedy (1978) on the basis of several hundred specimens from Zululand. OccurRENCE. The Lower Cenomanian of north Africa, Angola, Zululand and Madagascar. Genus TURRILITES Lamarck, 1801 Tyre species. Turrilites costatus Lamarck, 1801. Turrilites (Turrilites) acutus Passy Fig. 3 1832 Turrilites acutus Passy: 7; pl. 16, figs 3, 4. 1973 Turrilites (Turrilites) acutus Passy; Cooper: 47; figs 2D, 3A—C, 8D, 13B. 1978 Turrilites (T.) acutus Passy; Klinger & Kennedy: 7; pl. 2, figs A-E. MATERIAL. Three specimens, C.81040-1 from loc. 419, and C.81067 from loc. 953. Description. Each of the three large specimens consists of parts of two sinistrally coiled whorls up to 50-60 mm diameter. They are preserved mainly as internal moulds, though no suture-lines can be seen. Three rows of prominent tubercles form the main ornament, an upper row of large, pointed tubercles just above the middle of the whorl, a row of smaller pointed tubercles on the bottom half of the whorl, and a row of small tubercles that are just exposed above the lower whorl suture. There are approximately equal numbers of tubercles in each row. Those in the middle and upper rows are elongated obliquely to form weak ribs between the rows and stronger ribs that reach nearly up to the upper whorl suture. Remarks. Several closely similar specimens from the Middle Cenomanian at nearby localities south of Novo Redondo were described and figured by Cooper (1973: 47), and the species has been described by Klinger & Kennedy (1978: 7) on the basis of a large number of specimens Fig. 2 Phylloceras (Hypophylloceras) seresitense Pervinquiére from Lower Cenomanian at loc. 76. C.80995, x0-65. Fig. 3. Turrilites (T.) acutus Passy from Middle Cenomanian at loc. 419; Fig. 3b is the top of the spire. C.81040, x1. Figs 4,6 Mariella (M.) oehlerti (Pervinquiére) from Lower Cenomanian at loc. 350; Figs 4b and 6b are the bases of the spires. Fig. 4, C.81037, x1. Fig. 6, C.81035, x1. Fig. 5 Ostlingoceras (O.) cf. rorayense (Collignon) from Lower Cenomanian at loc. 301. C.81019, x2. Fig. 7 Proplacenticeras stantoni (Hyatt) var. bolli (Hyatt) from Lower Turonian at loc. 1007. C.81058, x0-67. 82 M. K. HOWARTH from the Middle Cenomanian of Zululand. A full synonymy can be found in the latter paper. The three specimens described here agree in all characters, and show that at these large sizes the tubercles in the lowest row are exposed just above the next larger whorl. OccurRENCE. Middle and low Upper Cenomanian of Europe, Africa, Madagascar and North America. Suborder AMMONITINA Hyatt, 1889 Superfamily DESMOCERATACEAE Zittel, 1895 Family DESMOCERATIDAE Zittel, 1895 Genus DESMOCERAS Zittel, 1884 TYPE SPECIES. Ammonites latidorsatus Michelin, 1836. Desmoceras (Desmoceras) latidorsatum lemoinei Collignon Figs 8-13 1928 Desmoceras (Latidorsella) latidorsata (Michelin); Collignon: 157; pl. 2, figs 4, 5. 1928 Desmoceras (Latidorsella) lemoinei Collignon: 158; pl. 2, fig. 6. 1942 Desmoceras kossmati Matsumoto: 26; fig. 12. 1954 Desmoceras kossmati Matsumoto; Matsumoto: 249; pl. 1, figs 1-6; pl. 6, fig. 6. 21958 Desmoceras (Pseudouhligella) elgini Young: 292; pl. 39, figs 4-20, 24-5, 30-1. 1959 Desmoceras kossmati Matsumoto; Matsumoto: 7; pl. 2, fig. 2. 1984 Desmoceras (Desmoceras) latidorsatum (Michelin); Wright & Kennedy: 61; pl. 3, figs 3, 5, 7-8, 13. MATERIAL. Twenty-two specimens, C.80982-6 (loc. 86), C.80996-7 (loc. 88), C.81008 (loc. 118), C.81021-33 (loc. 350) and C.81045 (loc. 458). MEASUREMENTS. D Wh Wb N Wh:Wb C.80982 55:5 25-2 (0-45) 25-4 (0-46) 14-0 (0-25) 0-99 C.80983 34-5 17-5 (0-51) 17-2 (0-50) 7-3 (0-21) 1-02 C.81023 24-8 12-0 (0-48) 10-6 (0-43) 5-1 (0-21) 1-13 C.81024 34-0 15-4 (0-45) 17-4 (0-51) 8-5 (0-25) 0-88 C.81025 26-0 11-9 (0-46) 14-5 (0-55) 6-0 (0-23) 0-82 DeEscriPTION. The 22 specimens range in size from 66 mm diameter down to 13 mm. Body chambers are preserved in nearly half of them, and are of all sizes (therefore many are immature specimens), while septa are present in one specimen up to a diameter of 44 mm. None have a complete adult mouth border preserved. The whorl shape is subcircular, with flattish sides, an evenly rounded venter, and rounded umbilical edges. The whorl proportions vary from slightly depressed to slightly compressed (whorl height/breadth ratios from 0-88 to 1-13). The shell surface is smooth and unornamented, apart from curved or biconcave constrictions which are irregularly developed in some specimens from sizes at least as small as 13 mm diameter. These constrictions are projected well forwards on the venter, where they are usually accompanied by a raised collar immediately behind. REMARKS. Collections of Desmoceras latidorsatum (Michelin) show much variation in whorl proportions and shape and strength of constrictions, as was described by Wiedmann & Dieni (1968: 131) and Cooper & Kennedy (1979: 239). The nominal species is of Middle Albian age, Figs 8-13. Desmoceras (D.) latidorsatum lemoinei Collignon from Lower Cenomanian. Figs 8, 12, 13 from loc. 86; Figs 9-11 from loc. 350. All x1. Fig. 8, C.80982; Fig. 9, C.81025; Fig. 10, C.81023; Fig. 11, C.81024; Fig. 12, C.80983; Fig. 13, C.80981. Fig. 14 Forbesiceras obtectum (Sharpe) from Middle Cenomanian at loc. 419. C.81078, x0-6. ae CENOMANIAN & TURONIAN AMMONITES FROM ANGOLA 83 84 M. K. HOWARTH and the overall morphological change in collections from younger horizons is considered sufficient to give subspecific names to them. Those in the uppermost Albian of Angola are more depressed and have been called Desmoceras latidorsatum perinflatum Cooper & Kennedy (1979: 237), although the correct name for this subspecies appears to be D. latidorsatum inflatum Breistroffer (1933) or D. latidorsatum reynesianum Haas (1952) according to their synonymy. In the Lower Cenomanian species that are now described, the whorls have become slightly more compressed again, so that all the measurable ones fall below the scatter of points of the Upper Albian specimens on Cooper & Kennedy’s graph (1979: 240, fig. 38, upper). They also differ from the Middle Albian subspecies in having weaker and fewer constrictions. The slight morphological shift in these features is sufficient for them to be referred to a different chronological subspecies, for which /emoinei Collignon appears to be the earliest available name. Collignon’s material from the Lower Cenomanian of Madagascar consists of about 25 small pyritized specimens of 10-20 mm diameter. He referred some of them to Desmoceras latidorsatum, but for the slightly more compressed ones the new species D. lemoinei was proposed. All are more compressed than the Upper Albian subspecies, and they have weakly- developed constrictions. The holotype of D. lemoinei has the following approximate dimen- sions: diameter 11 mm, whorl height 6-1 mm, whori breadth 5-6 mm; whorl height/breadth ratio 1-09. Similar, slightly compressed, specimens are found rarely in the Lower Cenomanian condensed facies of south and south-west England (Wright & Kennedy 1984). The same, slightly compressed, subspecies occurs in the Lower Cenomanian of Japan, where it is said to be common, and was described as D. kossmati Matsumoto (1942, 1954). Specimens are larger, attaining 50-60 mm diameter, and the whorl height/breadth ratio again falls below that of the Upper Albian subspecies. The same name was used by Matsumoto (1959) for examples from the Lower Cenomanian of California. Similar specimens occur in the Lower Cenomanian of Texas and have been described as D. (Pseudouhligella) elgini Young (1958); these have even more compressed whorls, with height/breadth ratios in the range 1-15 to 1-55 at 15-30 mm diameter, but they overlap substantially with Lower Cenomanian collections from other areas and it is likely that e/gini is also a synonym of D. latidorsatum lemoinet. Superfamily HOPLITACEAE H. Douvillé, 1890 Family PLACENTICERATIDAE Hyatt, 1900 Genus PROPLACENTICERAS Spath, 1926 Type SpEcIES. Placenticeras fritschi Grossouvre, 1894. Proplacenticeras stantoni (Hyatt) var. bolli (Hyatt) Fig. 7 1876 Placenticeras intercalare Meek: 471 (pars). 1903 Placenticeras stantoni var. bolli Hyatt: 214; pl. 40, figs 3-7; pl. 41, figs 1-7; pl. 42, figs 1, 2; pl. 43, figs 1, 2. 2? 1930 Placenticeras merenskyi Haughton: 363; pl. 11, figs 1-3. 1942 Proplacenticeras stantoni var. bolli (Hyatt) Moreman: 219. 1965 Proplacenticeras memoriaschloenbachi (Laube & Bruder) var. ambiloensis Collignon: 14; pl. 381, fig. 1646. 1965 Proplacenticeras stantoni (Hyatt) var. bolli (Hyatt); Collignon: pl. 383, fig. 1651. MATERIAL. One specimen, C.81058, from loc. 1007. Description. The specimen is wholly an internal mould, and consists of one-third of a whorl of body chamber, starting at the final septum of the phragmocone at a whorl height of 67-5 mm, and ending at an irregularly broken aperture at a whorl height of about 80 mm. The latter corresponds to an overall diameter of about 175 mm, and the whorl breadth at that size is about 49 mm. Only part of the small umbilicus is preserved, because there is a break about halfway down the umbilical wall. The umbilical edge is rounded and merges into the whorl side. The CENOMANIAN & TURONIAN AMMONITES FROM ANGOLA 85 greatest whorl breadth is near the umbilical edge, and from there the nearly flat whorl sides converge towards the venter, which is narrow and tabulate. The ventrolateral edges are well marked but rounded. The surface of this internal mould is smooth except for very faint traces of sinuous growth lines, one of which is accentuated almost to the strength of a constriction, especially on the outer half of the whorl and across the venter. The venter is smooth except for a very faint trace of a central ridge. The impressed dorsum shows that the venter of the next inner whorl was slightly concave between sharply-angled and slightly raised ventrolateral edges on which were a series of low undulating clavi that alternated from side to side. The last septum has most of the saddles preserved, but the deeper (posterior) parts of the lobes are missing. The first lateral saddle is divided into three by two adventitious lobes, and it slopes backwards to the first lateral lobe which is the deepest part of the suture-line. The second lateral saddle and the two or three auxiliary saddles slope forwards again towards the umbilicus. RemarkS. The Angolan specimen is only part of a large body chamber, but it is identified on the basis of its being an exact match with the final portion of a very fine complete specimen (BM(NH) C.53926) from the Britton Formation, Eagle Ford Group, near Britton, Texas. The latter specimen was obtained from W. S. Adkins and was labelled by him. Comparison of the two shows that the apparent umbilical seam of the Angolan specimen is in fact a curved break halfway down the umbilical wall, and the real umbilicus was somewhat smaller. Umbilical tubercles are not seen, probably because they faded just before the body chamber, as in the Texas specimen where they are well developed on the earlier whorls. The impression of the next inner whorl shows alternating clavi bounding a slightly concave venter, features that are typical of Proplacenticeras. The correct name for this species depends on further taxonomic work to investigate the relationships between three names currently in use for specimens from the Britton Formation in Texas. These are Proplacenticeras cumminsi (Cragin 1893), P. stantoni var. bolli (Hyatt 1903) and P. pseudoplacenta var. occidentale (Hyatt 1903), as described by Moreman (1942: 219). The oldest name (P. cumminsi) depends on the interpretation of an unfigured holotype, while the two Hyatt names still await the designation of type specimens. In view of the large amount of variation found within Campanian species of Placenticeras by Wolleben (1967), it is possible that only one species of Proplacenticeras was present in the Lower Turonian Britton Formation in Texas. Until such a revision is done, the name attached to the Angolan specimen is that of the best figured of Hyatt’s forms. Another similar specimen, though perhaps with a somewhat more complicated suture-line, has been described as Proplacenticeras cf. pseudoplacenta (Hyatt) by Matsumoto & Miller (1958: 355; pl. 45, fig. 2), and is part of a mid-Turonian (Woolgari Zone) assemblage of ammonites from Texas. A single specimen from Namibia that was made the holotype of Placenticeras merenskyi by Haughton (1930: 363; pl. 11, figs 1-3) is very close in all visible features to the inner whorls of the Texas specimens, and the specific name is probably a synonym. Its Upper Cenomanian age is deduced from an associated fauna of Exogyra. Other examples of Proplacenticeras occur in Madagascar, and at least some of those figured by Collignon (1965: pl. 381, fig. 1646; pl. 383, fig. 1651) are probably the same as the Texas species. Different species of Proplacenticeras are abundant in the Coniacian of south-east Africa, but Turonian rocks are not known in that area (Kennedy & Klinger 1975: 278). Species of Proplacenticeras in the Upper Cenomanian of Europe are P. memoriaschloenbachi (Laube & Bruder) and P. orbignyanum (Geinitz), both being more compressed than P. stantoni var. bolli. OccurRrENCE. Lower Turonian of Angola, Namibia, Madagascar and Texas. Superfamily ACANTHOCERATACEAE Grossouvre, 1894 Family FORBESICERATIDAE Wright, 1952 Genus FORBESICERAS Kossmat, 1897 TYPE SPECIES. Ammonites largilliertianus d’Orbigny, 1841. 86 M. K. HOWARTH Forbesiceras obtectum (Sharpe) Fig. 14 1853. Ammonites obtectus Sharpe: 20; pl. 7, fig. 4. 1971 Forbesiceras obtectum (Sharpe); Kennedy: 47; pl. 9, fig. 3; pl. 16, fig. 3; pl. 46, fig. 3 (see for synonymy). 1973 Forbesiceras obtectum (Sharpe); Cooper: 48; fig 5, 6A, 6B. 1984 Forbesiceras obtectum (Sharpe); Wright & Kennedy: 94; pl. 12, fig. 4; pl. 14, figs 1, 2; pl. 15, fig. 4. MATERIAL. Eight specimens, C.81077 from loc. 275, C.81078-83 from loc. 419, and C.81084 from loc. 953. REMARKS. This species has been described by Kennedy (1971), who gave a full synonymy, and by Cooper (1973), who had a considerable number of large specimens from localities near Novo Redondo. The present collection consists of eight large, worn internal moulds, all wholly septate, and with no portion of the original shell or ornament preserved. The most complete specimen has much of one whorl preserved up to a maximum size of 205 mm diameter, while all the others are large fragments of a quarter of a whorl or less, the largest having a whorl height of 150 mm and breadth of 54 mm, which corresponds to a diameter of about 225 mm. The largest specimen recorded by Cooper was 305 mm in diameter. OccurrENCE. Middle Cenomanian of England, Angola and many other areas of Asia and Africa. Well dated occurrences are Middle Cenomanian, but the species may also occur in the Lower Cenomanian (see Wright & Kennedy 1984: 95). Family ACANTHOCERATIDAE Grossouvre, 1894 Subfamily MANTELLICERATINAE Hyatt, 1903 Genus SHARPEICERAS Hyatt, 1903 TYPE SPECIES. Ammonites laticlavius Sharpe, 1855. Discussion. This genus and its type species have been discussed by Matsumoto, Muramoto & Takahashi (1969: 258) and Kennedy (1971: 64). Characteristic features are the high, flat-sided whorls, long ribs, and umbilical, mid-lateral and lower and upper ventrolateral tubercles. A second lateral tubercle may appear above the mid-lateral position. The ventrolateral tubercles enlarge into horns in the adults of some species, and the ribs become widely spaced. Such massive-whorled, horned adults are usually between 200 mm and 400 mm in diameter. Sharpeiceras appears to have been derived at the base of the Lower Cenomanian from early Mantelliceras stock, by the development of a mid-lateral tubercle at small diameters which was retained throughout growth. Sharpeiceras is an uncommon genus in most areas, but it is of widespread distribution and is restricted to the lower and middle zones of the Lower Ceno- manian, and so is an accurate age indicator (Kennedy & Hancock 1977: 132, 1978: V.14). Sharpeiceras florencae Spath Figs 15-19 1904 Acanthoceras laticlavium (Sharpe); Douvillé: 239; pl. 31, fig. 3. 1925 Sharpeiceras florencae Spath: 198, pl. 37. 2.1931 Acanthoceras (Mantelliceras) falloti Collignon: 41; pl. 4, figs 9-12. 1933 Sharpeiceras florencae Spath; Collignon: 67; pl. 6, fig. 5. 1942 Sharpeiceras goliath Haas: 7; figs 5-7. 2.1956 Sharpeiceras occidentale Benavides-Caceres: 465; pl. 54, figs 5, 6. 1959 Sharpeiceras florencae Spath; Matsumoto: 69, 70; fig. 28. 1962 Tlahualitoceras tlahualitoense Kellum & Mintz: 276; pl. 6, fig. 1; pl. 7, figs 1, 2; pl. 8, fig. 1. 1964 Sharpeiceras schlueteri Hyatt, Collignon: 102; pl. 353, fig. 1564. ? 1964 Sharpeiceras vohipalense Collignon: 104; pl. 354, fig. 1565. 1971 Sharpeiceras florencae Spath; Kennedy: 67; pl. 25, fig. 2. 1982 Sharpeiceras occidentale Benavides-Caceres; Renz: 68; pl. 21, fig. 1. CENOMANIAN & TURONIAN AMMONITES FROM ANGOLA 87 MATERIAL. 13 specimens: C.80975 from loc. 67, C.80977 from loc. 72, C.80979-80 from loc. 76, C.80990-1 and C.81005 from loc. 88, C.81000 from loc. 99, C.81003-4 from loc. 100, C.81006 from loc. 102, C.81016 from loc. 188, and C.81052 from loc. 526. MEASUREMENTS. D Wh Wb U Wh/Wb C.81006 192-5 77-7 (0-40) 58-3 (0:30) 56 (0-29) 1-33 C.81005 a 79 63:5 — 1-24 C.80979 — 89 74 — 1-20 C.81003 — 56 44 — 1-27 Description. The most complete specimen (C.81006, Fig. 17) is 208 mm in diameter at the aperture. Owing to poor preservation septa cannot be seen, though it appears to have between a quarter and half a whorl of body chamber. The ribs and tubercles are of moderate strength up to about 135 mm diameter, then they become much stronger and more widely spaced. From the same size the lower ventrolateral tubercle becomes much enlarged into a horn, to such an extent oo mi he ae ¥ ye! By; logy series Vol39No3 31 October 1985 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique and ever-growing collections of the Museum, both by the scientific staff of the Museum and by specialists from elsewhere who make use of the Museum’s resources. Many of the papers are works of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself, available separately, and individually priced. Volumes contain about 300 pages and several volumes may appear within a calendar year. Subscriptions may be placed for one or more of the series on either an Annual or Per Volume basis. Prices vary according to the contents of the individual parts. Orders and enquiries should be sent to: Publications Sales, British Museum (Natural History), Cromwell Road, London SW7 SBD, England. W orld List abbreviation: Bull. Br. Mus. nat. Hist. (Geol.) © Trustees of the British Museum (Natural History), 1985 The Geology Series is edited in the Museum’s Department of Palaeontology Keeper of Palaeontology: Dr H. W. Ball Editor of the Bulletin: Dr M. K. Howarth Assistant Editor: Mr D. L. F. Sealy ISBN 0 565 07008 8 ISSN 0007-1471 Geology series British Museum (Natural History) Vol 39 No 3 pp 107-189 Cromwell Road London SW7 S5BD Issued 31 October 1985 4\ShH MUSE a PUBLICATIONS UN ISSUED 3 10CT 1985 The systematics and palaeogeography || , P. E. S. Whalley _ Department of Entomology, British Museum (Natural History), Cromwell SW7 5BD Contents SSVI PSUS Merersreclalefeicicrsioreistaisiaelelsyetesasai=)-icicietale e si’ss\ejs}eieieralale/s a/c is’ s\sieie/oieva.e ele lecorainsietoveisioteroreisiarea 107 ITA GR CMA UG EL © MM ornreteratctoceletetetetcroteietetoiatsierslaieieTa%e'sic/efe/s sle:etsis{sve(s1e sles [avero(a eielsleleloleve\siersicveisieretureisiciere.c «/ecc1e 108 Geolosyapresenyationrand! C6CMMIGQUES) . * Figs 17-18 Regiata scutra sp. nov. (Orthoptera). Fig. 17, holotype, In.64027, forewing, 21 mm long. Fig. 18, paratype, In.64007, forewing. 132 P. E. S. WHALLEY Fig. 19 Regiata scutra sp. nov. (Orthoptera). Paratype, In.59146, forewing. b, enlargement of anal area and wing base. Note break in wing in front of anal area (arrowed) caused when the curved wing was flattened. Wing membrane heavily spotted. OTHER MATERIAL. [n.59392. Discussion. This species is similar to the grylloid Karataogryllus gryllotalpiformis Sharov, from the Jurassic of Asia, but differs in the shape of the median vein and the very sinuous subcostal vein. R. scutra was probably similar to Recent gryllids where the wings are folded over the back. When these were flattened during fossilization they split at the base of the wing, seen in Fig. 19b (arrow) as a fine break between the anal area and the rest of the wing. R. scutra also had the apex of the wings rolled together or even curled up as shown by In.64007 (Fig. 18). When the surface of the rock is wet the fossil appears to be covered in granules which, when examined dry, are seen to be tiny pits. Although these pits are denser in the anal area they occur over most of the forewing. The pitting shows up clearly in the subcostal area where they are arranged regularly along the subcostal veins. The anterior median and postcubital veins on some specimens are thickened; these thickened veins were almost certainly slightly raised above the wing and may have formed part of the stridulatory mechanism, but there is no sign of the mirror which, in Recent forms, is bounded by CuP and a branch of CuA. Vein 1A, which is also part of the stridulatory mechanism in Recent species, is, however, prominent in the fossil. R. scutra can be distinguished from Hagla gracilis Giebel, which is common in some Lower Lias LOWER JURASSIC INSECTS OF DORSET 133 same bedding plane as an ammonite. deposits in Britain, by a number of characters. The displacement of the postmedian and anterior cubital veins towards the posterior margin occurs in both Hagla and Regiata but they differ considerably in the form of the subcostal veins and the arrangement of the anal veins. R. scutra has a less complex, though equally broad, costal area similar to H. gracilis, but the median and cubital veins are nearly parallel towards the apex, suggesting a rolled or curved apex to the wing. The split apparent in most of the fossils between the anal area plus Cu + M and the rest of the wing is consistent with the flattening of the wing in life, which could have been held in the typical gryllid-like (box-like) fashion over the body. Until further studies are made of the Haglidae s. lat. the classification of Regiata as a haglid must remain provisional (see p. 130). Genus PROTOHAGLA Zeuner, 1962 TYPE SPECIES. Protohagla langi Zeuner, by original designation. Lower Lias, U.K. Protohagla langi Zeuner, 1962 Fig. 22 1962 Protohagla langi Zeuner: 165; pl. 26, figs 1, 2. No new material of this species has been found since Zeuner originally described it. P. langi is a large insect and had a wingspan of about 150 mm. Zeuner (1962: 166) said that in this specimen (In.59018) the fore and hind wings had been nipped off simultaneously at their bases. He speculated that this had happened when the insect was caught, with the wings closed, by a predator over the land. However, examination shows that the underlying wing is another forewing, so Zeuner’s hypothesis is untenable. Family GRYLLIDAE Latreille, 1802 Genus PROTOGRYLLUS Handlirsch, 1906 TYPE SPECIES. Protogryllus dobbertinensis (Geinitz 1880), by subsequent designation of Zeuner (1939: 188). Lias, Europe. DiaGnosis. Gryllids with distinct harp in male forewing but without mirror (Handlirsch 1906: 424). Fig. 21 Orichalcum ornatum gen. et sp. nov. (Orthoptera). Holotype, In.53983, forewing, 26mm long. Note the patterning on the wing. Fig. 22 Protohagla langi Zeuner (Orthoptera). Holotype, In.59018, forewing, 61 mm long. Fig. 23 Protogryllus magnus Zeuner (Orthoptera). In.51016, forewing, 26mm long. a, part. b, counterpart. * 134 LOWER JURASSIC INSECTS OF DORSET 135 Protogryllus magnus Zeuner, 1937 Fig. 23a, b 1937 Protogryllus magnus Zeuner: 155. 1939 Protogryllus (Archaegryllodes) magnus Zeuner; Zeuner: 192. Zeuner based the description of P. magnus on two incomplete male specimens, the holotype from the Upper Lias of Dumbleton, Gloucestershire and a second specimen from the Lower Lias of Binton, Worcestershire. Only slight differences have been found between these and the Dorset specimens. The curvature of the median vein, where it curves forward towards the radial, is more gentle in the Dorset specimens than in the holotype. Otherwise the size and rest of the venation are similar. Zeuner’s figure of 22mm wing length is based on the holotype and is probably an underestimate; even so, the Dorset specimens are larger and with the difference in localities, may represent a distinct species. Zeuner (1937) placed P. magnus in the subgenus Archaegryllodes Haughton (1924) on the basis of the median and radial veins touching in the forewing. He likened it to a Triassic species, A. stormbergensis Haughton from South Africa. In 1939 Zeuner suggested that Pro- togryllus grandis Zeuner might be the female of P. magnus, but no further evidence for this hypothesis has been elicited from the Dorset specimens. There is more of the very specialized venation preserved in the Dorset specimens than in the holotype of P. magnus; the stridulatory area of the male forewing can be seen in Fig. 23a. MATERIAL. In.44003, In.51016 (Fig. 23), In.51022, In.59373. Dimensions: 26:7 x 7-4mm. Genus MICROMACULA nov. TYPE SPECIES. Micromacula gracilis sp. nov. Lower Lias, U.K. D1AGnosis. Gryllids with narrow wings, dense maculations. Anal area more extensive than subcostal area. Distinct fan present between Cu and M in centre of wing. NAME. ‘Small spot’. Micromacula gracilis sp. nov. Figs 24-25 DIAGNOsIs. As genus. NAME. ‘Slender’. DESCRIPTION. Forewing with broad base, narrowing apically. Subcostal area reticulate; sub- costal vein reaching costal margin well before middle of wing. Costal margin with distinct hump on basal part. Radial vein running roughly parallel to median vein along wing. Cu sharply angled along edge of anal area with CuP close and three anal veins. Whole wing covered with fine spots. Cubital with fan-shaped arrangement of branches to medial veins. Anal area larger than subcostal area, strongly reticulate and covered with spots. Ho orype. In.49230 (Fig. 24); Birchi nodule, Black Ven, Charmouth, Dorset; part and counter- part. Jackson colln. Dimensions: 11 x 2:6mm, forewing; 12-13 mm, hindwing. PARATYPES. In.49594, In.51041, In.53913, In.53934, In.53969, In.59126 (Fig. 25). OTHER MATERIAL. [n.53926, In.53973, In.64031. Discussion. The holotype of M. gracilis is from the Turneri Zone whereas most of the para- types are from the younger Obtusum Zone. This is the only species certainly known to occur in both the Turneri Zone and Obtusum Zone, but for most species zonal age data is unavailable (see p. 110). Specimen 1n.59126 (Fig. 25) has both forewings and part of the hindwings preserved. The hindwing is longer than the forewing and is membranous, but is heavily folded and no details of the venation can be seen. The forewings are strongly sclerotized and similar to those of many 136 P. E. S. WHALLEY In.59126, pair of forewings, length 11 mm. Recent species. The wing was evidently curved round the body in life, as in modern gryllids, because the flattened fossil shows cracks in the position of original folds. M. gracilis was probably similar in appearance to the Recent species Amsurgus lateralis Chopard (Gryllidae, Trigonidiinae), a cricket from the Pacific area. The forewings of the Recent species are broadly patterned giving a reticulate appearance, and the surface is covered with small spots; these are larger on the veins, and from them hairs arise. Little is known of the biology of the Recent species, which is considered to be typically tree-dwelling, rather than ground-living. Suborder CAELIFERA Superfamily LOCUSTOPSIDEA Family LOCUSTOPSIDAE Handlirsch, 1906 Genus LOCUSTOPSIS Handlirsch, 1906 TYPE SPECIES. Locustopsis elegans Handlirsch, by subsequent designation of Cockerell, 1915. (Handlirsch 1906: 421; Zeuner 1942: 8; Sharov 1971: 90-97). Trias-Upper Lias, Europe. LOWER JURASSIC INSECTS OF DORSET 137 D1aGnosis. CuA divides into three branches. Sc extends almost to apex of wing. R and M run separately to base of wing. M with three, rarely two, branches. Locustopsis spectabilis Zeuner, 1942 Figs 26-28 1942 Locustopsis spectabilis Zeuner: 8; fig. 1. TYPE LOCALITY. Strensham, Worcestershire; Upper Trias. DISCUSSION. This species was described by Zeuner on the basis of a single, slightly folded, forewing from Strensham and he mentioned a possible second specimen from Warwickshire. There are a number of specimens from Dorset which I attribute to this species, although they differ slightly from the holotype in the relative proportions of the wing; this could be due to the slight folding in the holotype. The Dorset specimens show a weakly-developed precostal vein while the actual costal vein is well developed with lateral branches. The subcostal vein reaches the wing tip while the radial, subcostal, median and cubital veins are separated almost to the base of the wing. The median vein divides into three branches; in some specimens the third branch comes off the postmedial while in others it comes off the antemedial. Most of the specimens from Dorset, and the holotype from Worcestershire, have a three-branched cubital vein, but two Dorset specimens (In.49591, In.51043 (Fig. 28)) have an extra branch, making a four-branched cubital. Recent Orthoptera have considerable intraspecific variation in wing venation (Ragge 1955) but, in Recent species, a four-branched cubital is unusual. The rest of the venation of these two specimens is similar to the other Dorset specimens and they are con- sidered to be merely intraspecific variants. MATERIAL EXAMINED. In.49203, [n.49208, In.49565 (Fig. 26), 1n.49586, In.49591, In.49593 (Fig. 27), In.49600, 1n.49614, In.50992, In.51001, In.51035, In.51043 (Fig. 28), (? In.53900), In.53908, In.53914, In.53917, (? In.53984), In.59121, In.59131, In.59357, In.59358A, In.59367, In.64039. DIMENSIONS. 22—29 mm. Order PHASMATODEA, stick-insects Superfamily XIPHOPTERIDEA Family AEROPHASMATIDAE Martynov, 1928 This family was based originally on specimens from the Jurassic of Karatau. Sharov (1968) commented on traces of dense short hairs on these specimens, which are also clearly visible on the new species from Dorset. Genus DURNOVARIA nov. TyPE SPECIES. Durnovaria parallela sp. nov. Lower Lias, U.K. DiaGnosis. Elongate wing with roughly parallel veins. Costal vein simple, unbranched; costal area densely hairy. Median three-branched with MP and Cu having a common stem. Name. Latin name of Dorchester. Durnovaria parallela sp. nov. Figs 29, 33-34 DraGnosis. As genus. Name. ‘Parallel’. DESCRIPTION. Precostal, costal and much of subcostal areas very hairy, rest of wing strongly pigmented and probably hairy. Sc short, reaching to costal margin approximately level with first fork of radial vein. Four radial veins (plus radial sector) almost parallel along wing. Median and radial veins clearly fused at base of wing. Median (strictly M + CuA) divides very early into MA, which again forms three branches, and MP/Cu branch. The latter divides into 138 P. E. S. WHALLEY Figs 26-28 Locustopsis spectabilis Zeuner (Orthoptera). Fig. 26, In.49565, forewing, 24mm long. a, part. b, counterpart. Fig. 27, In.49593, forewing. Fig. 28, In.51043, forewing. LOWER JURASSIC INSECTS OF DORSET 139 Fig. 29 Durnovaria parallela gen. et sp. nov. (Orthoptera). Holotype, In.59171, counterpart. See Fig. 33. Figs 30-32 Orthoptera, saltatoria-type legs. Fig. 30, In.59127, tibia and tarsi, total length 24mm. Stout spines on tibia. a, part. b, counterpart. Fig. 31, In.59391, femur 10-9 mm, tibia 10mm. Note ridge of spines along both. Fig. 32, In.53919, femur 9:2 mm. 140 P. E. S. WHALLEY 1A Fig. 33. Durnovaria parallela gen. et sp. nov. (Orthoptera). Diagram of forewing venation. A = anal area, C = costal vein, Cu = cubital veins, M = median veins, R = radial veins, Sc = subcostal vein. two, a postmedial and anterior cubital. The posterior branch of M + Cu forms CuA. CuP comes off near the base of M from the common stem of M + CuA and runs unbranched to the wing margin. The first anal has a short apical branch, the second is nearly as long but unbranched, while the third is short. There are many small reticulations in the area between 2A and the wing margin. Hototyree. In.59171 (Figs 29, 34); Flatstones, Stonebarrow, Charmouth, Dorset; part and counterpart. Jackson colln. Dimensions: 34 x 7:7mm; estimated wing length 36mm. DIscussIONn. This species is separated from the Chresmodidae (sensu Sharov, 1968) by the branching of the postmedian and anterior median veins. It differs from Aerophasma Martynov in having a weakly-developed cubital vein and in the division of the anterior median vein into two branches. The long parallel-vein appearance is typical of the Xiphopteroidea. The very hairy nature of the wing, particularly clear in the anterior basal part of the wing, is character- istic of the Aerophasmatidae, a family currently known from the Jurassic of the U.S.S.R. The other families in the Xiphopteroidea are Triassic (Xiphopteridae, Aeroplanidae), Jurassic (Necrophasmatidae) or Cretaceous (Cretophasmatidae). All are regarded as related to, though not necessarily direct ancestors of, modern stick-insects (Phasmidae). we a : Pues Fig. 34 Durnovaria parallela gen. et sp. nov. (Orthoptera). Holotype, In.59171, part, forewing, 34mm long. See Fig. 29. Order HEMIPTERA, plant bugs Suborder HOMOPTERA Six homopterous insects occur in the Jackson collection. Three of these are considered to belong to a species in the Cicadidea, one is unidentifiable but in the same superfamily, one is placed in the Cicadellidea and the last is in the Superfamily Fulgoridea. Currently there are six species of Mesozoic Homoptera known from the U.K.; these are three species of Margaroptilon (Handlirsch 1906) and one of Homopterites (Handlirsch 1906); both genera are in the Cicadel- lidae, one in Tettigarctidae and one in Cicadidae (Whalley 1983). LOWER JURASSIC INSECTS OF DORSET 141 Superfamily CICADIDEA Family TETTIGARCTIDAE Bekker-Migdisova, 1949 There are only two living species of this family which is common in the fossil record from the Triassic onwards. Species of several genera are known from the Mesozoic while the sister- group, Cicadidae, are rare in the Mesozoic (Whalley 1983). The fossils from Dorset, which have similar venation to Shuraboprosbole Bekker-Migdisova, are very much smaller. The vein which represents the base of the median, linking it and the anterior cubital to the postcubital as a cross-vein, cannot be seen in the Dorset specimens and was probably not present. Genus PARAPROSBOLE nov. TYPE SPECIES. Paraprosbole rotruda sp. nov. Lower Lias, Dorset, U.K. D1AGNOosIs. Base of M + CuA thickened and curved. Cross-vein from M + CuA missing. Name. Like Prosbole, but distinct. Paraprosbole rotruda sp. nov. Figs 35—36 DIAGNosIs. As genus. Name. A genus of Recent moths (Heinrich 1956: 225). DescriPTION. Base of M + CuA curved and thickened. R clearly divided as far as nodal line. CuA and anal veins unbranched. Nodal line double, pigmented. Hotorypee. [n.59374 (Fig. 35); Flatstones, Stonebarrow, Charmouth, Dorset; part and counter- part. Jackson colln. Dimensions: Base of wing to nodal line 12-13mm. Nodal line, from anterior to posterior wing margins, 8-9 mm. Total wing length 20-22 mm. PARATYPES. [n.48162 (Fig. 36), 1n.64395. Discussion. None of the wings is complete but comparison of the base of the wing (holotype) and the width of the wing at the nodal line with all the specimens suggests they are all conspecific, even though one is represented only by the apical part of the wing. They differ from the Triassic Cicadoprosbole in having fewer short branches on the radial vein near the apex of the wing. In specimen In.48162 (Fig. 36) the ambient vein is clear, with traces of the membrane distal to this visible. Tettigarctidae, gen. et sp. indet. The single specimen is incomplete and cannot be identified further than to family. It has a clearly preserved nodal line and parts of a few apical veins can be seen. There are some strongly marked sclerotized areas towards the centre of the front margin of the wing but the whole wing is much smaller than in Paraprosbole. MATERIAL. In.51011 (not figured). Dimensions: Apical vein c. 4-5-Smm. Nodal line c. 4-5 mm. Basal part of wing to nodal line c. 7-5 mm. Superfamily CICADELLIDEA, gen. et sp. indet. Fig. 37 The single specimen does not have the nodal line as clearly preserved as in the Tettigarctidae but has a strongly curved R + M + Cu. Some of the apical veins are preserved. MATERIAL. In.49225. Dimension: Wing length c. 8 mm. ‘Suo] wu g ‘xordde BuIA ‘SUIOA [e}IGQNno + [eIpoul + [eIpel Jo aseq = KX ‘S7ZHPU] (e1o}dowoyY) yopur ‘ds yo ‘usd ‘vapyapeoD LE “sly ‘Zuo wu g ‘xoidde ‘our jepou = N-N ‘7918p Ul ‘edAjesed ‘og “B14 ‘aedigjun0s ‘q y1ed ‘e ‘BuO; WUIg ynoge sI YOIyM ‘oul; [epou Jo uorIsod syieU N-N ‘pLe6S Ul ‘odAjopoy ‘cg “31.4 (e19}doul0#) ‘Aou “ds yo ‘uad wpnajou ajoqsoidvavg 9€-SE SBIM a. cg to ee ee eee JE : Pb ae x ‘ : z * 3 A ete J! LOWER JURASSIC INSECTS OF DORSET 143 Superfamily FULGORIDEA Family CIXIIDAE Spinola, 1839 Genus MESOCIXIELLA Martynov, 1937 TYPE SPECIES. Mesocixiella asiatica Martynovy, by original designation. Upper Triassic and Lower Jurassic, U.S.S.R. Mesocixiella (?) fennahi sp. nov. Fig. 38 D1AGNosIs. Fulgoroid with short branches to R,. Front margin of tegmina straight. Name. After Dr R. G. Fennah, specialist in Fulgoroids. DESCRIPTION. Head regularly curved, narrow. Rostrum reaching to metathorax. Sc vein long, R probably unbranched near apex. Two wings overlap and some of the veins visible in the figure are from the lower wing. M forks with several more apical forks. Eyes just visible, trace of antenna below eye. Ho.oryPe. 1n.53942; Flatstones, Stonebarrow, Charmouth, Dorset. Jackson colln. Dimensions: Head to tip of wing, 5-9 mm; Wing approx. 4mm. Head to tip of abdomen, 3-7 mm. Discussion. I am grateful to Dr R. G. Fennah for comments on this specimen in which he indicated the general cixiid-like appearance, although pointing out that some of the diagnostic features cannot be seen. Cixiids are an extant group with a long fossil history from the Permian (Rohdendorf 1962). M. (?) fennahi is only provisionally placed in the genus Mesocixiella. Suborder HETEROPTERA The Heteroptera are currently the subject of a joint research project by the present author with W. R. Dolling and Y. I. Popov. The paper (in press) will contain a detailed taxonomic account of the new species and a discussion of their affinities. The basic information from this study will be summarized here for use in the evaluation of the fauna. Thirty specimens of Heteroptera from the Dorset Lias occur in the Jackson collection, and one additional specimen from the same locality is preserved in the City of Bristol Museum and Art Gallery. The results of the study showed that, while there were aquatic Heteroptera which could be placed in Recent families, all the terrestrial species belonged to extinct families. This confirms similar results with the Heteroptera from the Upper Lias of Germany found by Popov & Wootton (1977), where they were unable to place any of the terrestrial species in extant families. In the forthcoming revision of the Heteroptera of the Dorset Lias, one species is placed in Archegocimicidae, eight in the Pachymeridiidae, two in the Corixidae, and four (including the specimen from the Bristol Museum) in the Belostomatidae. Of the remaining 15 specimens, 14 are placed in the Infraorder Cimicomorpha and will be put in a new family, while the remain- ing specimen is in the Infraorder Leptopodomorpha and will also be placed in a new family. The proposed names for the various taxa which will appear in the joint paper are not given here to avoid nomenclatural complications, but the registration numbers will allow correlation between the general discussion here and the subsequent descriptions of these species in the joint paper. Infraorder CIMICOMORPHA Family ARCHEGOCIMICIDAE Handlirsch, 1906 Archegocimicidae, gen. et sp. indet. In.51007 is part of a wing which is provisionally referred to this family. Family nov. 1 - This family is erected to include the series of 14 specimens, all with clear cuneal and medial fractures and with the head, thorax and abdomen covered with short curved bristles. 144 P. E. S. WHALLEY < gsi Bog te bee 39 ber % LOWER JURASSIC INSECTS OF DORSET 145 Gen. et sp. nov. 1A Figs 39-40 D1aGnosis. Head elongate, slightly longer than pronotum. Eyes small, hairy. Legs slender, rather short with femora scarcely reaching beyond body margin. Cuneal and medial fracture present. Pronotum strongly transverse, its lateral margins explanate and embracing the head. MATERIAL. Four specimens from Black Ven, nine specimens from Stonebarrow, one specimen from Charmouth. In.49216, In.49243, In.49567, In.49576, In.49589, In.51048, In.53909, 1n.53986, 1n.53995, In.59104, In.59128 (Fig. 40), In.59137 (Fig. 39), In.59147, In.59358B. Dimensions: 8-8-10 mm body length. Jackson colln. Discussion. This is a very distinctive species of Heteroptera with a characteristically shaped head and prothorax, and conspicuous curved setae. Judged on the number of specimens in relation to the total number of Heteroptera in the collection, this species was common in the area from which the Dorset Lias insects were derived. Infraorder LEPTOPODOMORPHA Family nov. 2 This family is erected for a single species and genus from the Dorset Lias. Gen. et sp. noy. 2A Fig. 41 DIAGNOsIs. Pronotum with narrow anterior collar and transverse sulcus dividing it into smaller anterior and larger posterior lobes. Forewing with cuneal and medial fracture; membrane clearly differentiated from corium. Forewings greatly widened in costal area. Pronotum, scutel- lum, clavus and corium finely punctate. MATERIAL. Black Ven, In.59152. Dimensions: 5-4 x 3:-5mm wing. Jackson colln. Discussion. Although the head is missing from the single specimen, the rest of the body is sufficiently well preserved to show characters of the new family. Infraorder PENTATOMOMORPHA Family PACHYMERIDIIDAE Handlirsch, 1906 A family with a number of species and genera described from the Mesozoic. Gen. et sp. nov. 3A DIAGNOsIs. Forewing narrow, as in Pachymeridium Geinitz, but with three branches of the radial vein arising close together near the apex of the medial fracture. Clavus and corium punctate, with cross-markings in ‘herring-bone’ pattern. MATERIAL. Four specimens from Black Ven, three specimens from Stonebarrow. In.49566, In.51039, In.53901, In.53941, In.59105, In.59151, In.64014. Dimensions: 8-9 mm wing. Jackson colln. Fig. 38 Mesocixiella (?) fennahi sp. nov. (Homoptera). Holotype, In.53942. R = rostrum, E = eye. Total length of insect including wings 5-9 mm. Figs 39-40 Gen. et sp. nov. 1A (Heteroptera). Fig. 39, In.59137. Length overall 9mm. Fig. 40, In.59128, forewing. Co = corium, M = membrane. Fig. 41 Gen. et sp. nov. 2A (Heteroptera). In.59152. Length of body 5-4mm. Fig. 42 Gen. et sp. nov. 8A (Heteroptera). Bristol Museum specimen no. 111/1958. Arrow indicates metathoracic gland. Length of insect 32 mm. 146 P. E. S. WHALLEY When the Dorset pachymerids in the Jackson collection were examined, several possible new species were separated; these are listed below but although distinct from sp. nov. 3A, they are not sufficiently well preserved for description. All are from the Flatstones, Stonebarrow, Dorset. Pachymeridiidae species 4A. Material: In.51028. Dimensions: 7-5 mm wing. Pachymeridiidae species 5A. Material: In.51005. Dimensions: 6-7 mm wing. Pachymeridiidae species 6A. Material: In.50999. Dimensions: 6:3 mm wing. Infraorder NEPOMORPHA Family CORIXIDAE Leach, 1815 Two specimens of aquatic corixid bugs are ascribed to this family but they are too poorly preserved to warrant a formal description and name. MATERIAL. 2 specimens from Black Ven, Charmouth; In.59140, In.59166. Dimensions: 9- 10mm x 5mm body. Jackson colln. Family BELOSTOMATIDAE Leach, 1815 This family of aquatic bugs is represented by three specimens in the Jackson collection and one well-preserved specimen from the same Dorset locality in the City of Bristol Museum and Art Gallery. The Dorset specimens are the earliest known examples of the family, which was evidently well differentiated into a discrete group by the Lower Lias. Most of the specimens show the movable breathing siphon characteristic of this aquatic family. Genus MESONEPA Handlirsch, 1906 TYPE SPECIES. Mesonepa primordialis Handlirsch (1906: 637), by subsequent designation of Popov (1971: 116). Jurassic, western Europe. Mesonepa species 7A DIAGNOsIs. Smaller than other species of Mesonepa. MATERIAL. Two specimens from Stonebarrow, Dorset; In.51014, In.59383. Dimensions: 13 x 6mm. DISCUSSION. This is one of the smallest species known in the family but has many of the features typical of Recent species. The scutellum is broad and is a unique feature, suggesting that the earliest belostomatids may have resembled the naucorids. Gen. et sp. nov. 8A Fig. 42 DiAGNosIs. Belostomatids with anterior tarsi and tibia not fused. MATERIAL. One specimen from Charmouth, Dorset; in City of Bristol Museum and Art Gallery. Dimensions: 32 x 18-5 mm. Discussion. This species, which looks very similar at first glance to Recent species, differs in the structure of the forelegs. In Recent species the forelegs are modified into grasping organs and the tarsi and tibia are fused together. With this they seize their prey. In the new species the tarsi are free and not fused to the tibia and presumably a slightly different technique for grasping the prey was used. The preservation of the metathoracic gland (Fig. 42, arrowed) shows clearly, but the fact that some of the components were red as in modern species is even more remarkable. A preliminary examination of the contents of this gland has indicated a high sulphur level for the red material and may well indicate the presence of organo-sulphur compounds (see p. 112). LOWER JURASSIC INSECTS OF DORSET 147 Order RAPHIDIOPTERA, snakeflies Adults and larvae of this order are predatory, feeding on other insects. The Raphidioptera have a long fossil history but the Permian and Upper Carboniferous species which have been ascribed to the order are probably not correctly assigned (Carpenter 1967). There are only about 160 living species in the world, although Carpenter (1967) considered that they were a more diverse group in the Mesozoic. Most of the fossils are known only from the wings (e.g. Carpenter 1936, Martynova 1947, 1961, Macleod 1970) while the characteristically shaped body with elongate pronotum is rarely preserved. Although the exact shape of the prothorax is not clear in the first species described below, it is certainly somewhat elongated. Family MESORAPHIIDIDAE Martynov, 1925a D1AGNosIs. Subcostal vein ends around the middle of the costal margin. Costal area narrower than in Recent Raphidioptera. Rs originated well before centre of wing. Lower and Upper Jurassic. Genus METARAPHIDIA nov. TYPE SPECIES. Metaraphidia confusa sp. nov. Lower Lias, U.K. DIAGNOsIS. Forewing with two large radial cells between R and Rs. Short pterostigmal cross- veins. Hindwing broadly similar, subcostal vein slightly shorter than in forewing. Name. Varied from Raphidia. Figs 43-44 Diagram of wing venations in Raphidioptera. A = anal veins, Cu = cubital veins, M = median veins, MA = anterior median vein, R = radial veins, Rs = radial sector, Sc = subcostal vein. Fig. 43, Priscaenigma obtusa gen. et sp. nov. Fig. 44, Metaraphidia confusa gen. et sp. nov. 148 P. E. S. WHALLEY Discussion. The presence of pterostigmal cross-veins differentiates this genus from others in the Mesoraphidiidae. Pterostigmal cross-veins are also found in Recent Raphidiidae but the very elongate radial cells are characteristic of Metaraphidia. There are also more marginal branches (8-9) to the radial veins than in other genera of Mesoraphidiidae. Metaraphidia confusa sp. nov. Figs 44-45 DIAGNOsIS. As genus. Discal cell small, less than half length of either radial cell. NAME. ‘Mixed up’. DESCRIPTION. Forewing with hairy costal margin, base of wing obscured. Rs branches off R towards base of wing. Base of M and origins of MA and MP obscured. MP short, probably fused for part of length with Cu. Posterior median cells all broadly similar in shape. Fore and hind wings overlying one another. No trace of pterostigma. Body and head partially obscured. Prothorax longer than width of head. Abdomen slender, tip enlarged. Ho.otype. GSM 117552 (Fig. 45); Charmouth, Dorset; in British Geological Survey collection. Dimensions: Forewing 14:5 x 4-8mm. Discussion. This snakefly was evidently washed up against an ammonite and buried with it. Also trapped against the same ammonite was another insect; all three animals were fossilized together. The second insect, probably dipterous but poorly preserved, was fossilized partly under the snakefly (Fig. 45, A-D). Most of the ammonite (A) has been cut away to show the underlying insects, of which the dipterous wing (B) is just faintly visible next to the larger insect. The snakefly head (C) and prothorax (D) are just visible on the specimen but three of the four wings are tightly overlapped and individual details are difficult to discern. Fortunately one forewing is displaced slightly clear of the others. Family BAISSOPTERIDAE Martynova, 1961 D1aGnosis. Raphidians with a long or short Sc. Cells in radial and median area of wing more numerous than in Mesoraphidiidae. Lower and Upper Jurassic. Genus PRISCAENIGMA nov. TYPE SPECIES. Priscaenigma obtusa sp. nov. Lower Lias, U.K. DiaGnosis. Raphidians with very long basal radial cell and 3—4 more radial cells to apex of wing. The long basal cell separates this genus from Baissoptera Martynova. Name. ‘Ancient riddle’. Priscaenigma obtusa sp. nov. Figs 43, 46 DIAGNOosIs. As genus. Name. ‘Blunt’. DESCRIPTION. Forewing, subcostal vein simple, reaching to near wing apex. One basal subcostal cross-vein present (but area distorted and others may be present), pterostigma obscure. M arising off R. Intramedian cell broader apically. CuA roughly parallel to hind margin with 1-2 apical branches. Several rather elongate cells in radial and median areas. Ho.otype. [n.53898 (Fig. 46); Black Ven, Charmouth, Dorset; part and counterpart. Jackson colln. Dimensions: 12:6mm long. Discussion. This specimen has been fossilized with its wings together, and as a result it is difficult in one or two places to distinguish the underlying wing veins from those of the top LOWER JURASSIC INSECTS OF DORSET 149 SAR: 75. Se: collection no. GSM 117552. Length of forewing 14-5mm. The anterior part of the insect body was covered by an ammonite which has been partly removed; there is also a small dipterous specimen below this ammonite. A = ammonite, B = dipterous wing, C = head of Metaraphidia, D = protothorax of Metaraphidia, E = abdomen of Metaraphidia. Fig. 46 Priscaenigma obtusa gen. et sp. nov. (Raphidioptera). Holotype, [n.53898, forewing 12-6mm long. wing. In particular there is a short, indistinct, vein between the subcosta and the costal margin about half-way along the wing. This is almost certainly part of the underlying wing (? a hindwing) and not the subcostal vein of the fore (top) wing. The actual subcosta of the forewing appears to extend almost to the apex of the wing. Initially this specimen was considered to be a chrysopid (Neuroptera), with some general similarities to the aberrant Recent genus Kimochrysa Tjeder, 1966 (Chrysopidae); in particular 150 P. E. S. WHALLEY the long radial cell and the shape of the intramedian cell are similar. However, the origin of the radial sector is atypical of the Chrysopidae and the marginal veins in the fossil are unbranched. The cells in the median/apical area of the wing are more numerous than in the Meso- raphidiidae and are more typical of the Baissopteridae, although the described species of Baissoptera Martynova all have shorter subcostal veins. The radial cell in Priscaenigma is longer than in Baissoptera and the new species may well merit a new family of its own. Order MECOPTERA, scorpion flies The publication of their bibliography of the Mecoptera by Schlee & Schlee (1976) and of Willmann’s (1978) catalogue of fossil Mecoptera has brought together the very scattered liter- ature of this order. Following Hennig (1981) and other workers, the two main suborders of the Mecoptera, Eumecoptera and Protomecoptera, are accepted here. The Protomecoptera, which are exclu- sively fossil, have many subcostal veinlets and a modified cubital vein. The vast majority of Recent and fossil species are placed in the Eumecoptera, in which the number of costal veinlets is reduced and the cubital vein is unbranched. Hennig interpreted the cubital vein as a compos- ite one with M4. From the Lower Permian into the early Jurassic, Mecoptera fossils are common. They are known from the Lower Cretaceous and into the Tertiary (Willmann 1978) but are far less abundant in the Recent fauna. In the late Triassic and throughout the Jurassic fossil Mecoptera are numerous both as species and individuals, their abundance at this period contrasting sharply with the small size of the Recent fauna, with only about 300 extant species. This is the reverse of the position with, for example, the Diptera or Lepidoptera, which are rare as fossils in the Jurassic but abundant in the Recent fauna. Recent Mecoptera are mostly carnivorous in adult and larval stages, feeding on insects. Mecoptera do not feed on the wing like dragonflies but hunt for their prey while it rests or feed on dead insects. There is no reason to suppose, from a study of the few preserved mouthparts of fossil Mecoptera, that their feeding habits were any different from those of their living relatives. z, Pee ret Nata o ETO - abe Bet ‘ 2 Fie Se “ Fig. 47 Orthophlebia capillata sp. nov. (Mecoptera). Holotype, In.53924, part. A = anterior of thorax, P = posterior of thorax, F = forewing, H = hindwing, T = thorax. Arrow indicates chisel marks, not abdomen. See Figs 48, 51. LOWER JURASSIC INSECTS OF DORSET 151 The classification of fossil Mecoptera is based mainly on small venational differences. These differences are used at both the specific and generic level but unfortunately we do not know how much intraspecific variation there is in these characters. Some of the finer points, such as the minor branching of veins near the wing margin, may well prove to be intraspecific. There are 15 specimens in the Jackson collection although several are represented by only small fragments. Three species are recognized. Suborder EUMECOPTERA Family ORTHOPHLEBIIDAE Handlirsch, 1906 D1AGNosIs. Forewing with long subcostal vein reaching the pterostigma. Radial vein diverging from Sc with five or seven branches. Usually five medial veins. This family is known from the Triassic to the Jurassic in Europe, Asia and Australia. Fig. 48 Orthophlebia capillata sp. nov. (Mecoptera). Holotype, In.53924, counterpart. a, thorax, 75mm long. A = anterior margin of thorax, H = hindwing base. b, posterior part of thorax to show setae (arrowed). See Figs 47, 51. SY P. E. S. WHALLEY Genus ORTHOPHALEBIA Westwood, 1845 TyPE species. Orthophlebia communis Westwood (in Brodie 1845: 102), by subsequent desig- nation of Tillyard, 1933. Upper Triassic and Jurassic of Europe and Asia. DIAGNOsIS. Orthophlebiids with six or seven branches to Rs. Orthophlebia capillata sp. nov. Figs 47-49, 51 DiaGnosis. Large orthophlebid with curved Cu, in forewing joining Cu,. Hindwing with Cu, and M joined. Veins of hindwing with Y-shaped 1A. NAME. ‘Hairy’. DESCRIPTION. Forewing, Sc reaches pterostigma; R, branches in pterostigma; Rs gives rise to seven final branches on wing margin. Anterior median vein has short branch near wing margin (M, + M,) and a single M,. M, and M, arise from a short postmedial vein. Cubital vein divides near base into curved anterior and straighter posterior branches. Three anal veins. Few cross-veins preserved. Whole wing covered with microtrichia with larger hairs along veins. Hindwing with humeral vein; seven-branched Rs. M with four branches and Cu, arising from near base of M. One cross-vein from CuP to 1A forms part of the distinctive Y-vein formed by 2A. Thorax partially preserved, with clear indication of setae (Fig. 48b), particularly eight post-dorsal setae on metanotum. Prothorax shorter than broad. Part of hind leg, femur 5mm long, covered with short hairs. RS Za 3A 1 Fig. 49 Orthophlebia capillata sp. nov. (Mecoptera). Diagram of wing venations; F—forewing, H— hindwing. A = anal veins, Cu = cubital veins, M = median veins, R = radial veins. LOWER JURASSIC INSECTS OF DORSET 153 Hototyee. In.53924 (Figs 47-49, 51); Flatstones, Stonebarrow, Dorset; part and counter- part. Jackson colln. Dimensions: Forewing length 24mm, width 63mm. Hindwing, 21:5 x 6mm. PARATYPES. In.49579, In.49601, In.53899, In.59380. DIscussION. The new species from Dorset, with a wingspan of 50mm, is larger than any other species of Orthophlebia except the huge O. grandis Martynov 1927, from Turkestan, which has a wingspan of 70mm. Even O. gigantea Tillyard, from the Trias of Strensham (U.K.), had a wingspan of only 38mm and most Liassic species were much smaller. All the specimens of O. capillata show the very hairy nature of the wings, with hairs on the membrane as well as along the veins. In.53899 shows the microtrichia particularly well when the rock surface is wet; some larger hairs are preserved on the membrane while even larger setal bases are clearly visible along the anal veins: presumably there were originally long setae on these veins. The costal margin is covered with short stout hairs, while the pterostigma is covered by a dense patch of microtrichia. The wing membrane as preserved in some specimens looks as though it may have been slightly thickened and perhaps translucent in life. The smaller O. anglicus described from the Lower Lias of Gloucestershire lacks the Y-shaped second anal vein present in the hindwing of O. capillata. Genus PROTORTHOPALEBIA Tillyard, 1933 TYPE SPECIES. Protorthophlebia latipennis Tillyard (1933: 28), by original designation. Upper Trias to Upper Jurassic, Europe and Asia. D1AGnosis. Orthophlebiids with a wing length of less than 10mm. Rs five-branched. Protorthophlebia latipennis Tillyard, 1933 Figs 50, 52-53 1933 Protorthophlebia latipennis Tillyard: 29, fig. 6. DiaGnosis. As genus. Three anteradial veins; setal bases on anal veins large and prominent. DESCRIPTION. Sc and R, end at pterostigma. R, forked near wing margin. Two-branched MA, three-branched MP. Three anal veins; CuA curved joining CuP near base. Long hairs on Sc, R, and M at the base, large setal bases on anal veins. Whole membrane covered with short, fine hairs with short, stout hairs along costal margin. SS ere — Fig. 50 Protorthophlebia latipennis Tillyard (Mecoptera). Diagram of wing venation. A = anal vein, M = median vein, Cu = cubital vein, R = radial veins, Sc = subcostal vein. 154 P. E. S. WHALLEY Ns 6 oe 9 PSE ‘3% Se OME IEE es ” ee eee MOF ee tha “Hie | * - an ae a 3 * a a ‘ Ge eee nee nn omen ern Me in Ace een ats Moon So mmee acoso dosooceune enc 472 Synopsis The lithostratigraphy of deposits laid down during the Bartonian in the Hampshire Basin is described. The new formal terms Barton Clay Formation and Becton Sand Formation are defined. The Elmore Formation is reduced to member rank within the Barton Clay Formation. A pattern of cyclic sedimenta- tion within the Barton Clay Formation, Becton Sand Formation and part of the Boscombe Sands is recognized. Fifty-three mammalian taxa of species group rank, represented mainly by teeth but also by some important cranial and postcranial elements, are described from the Creechbarrow Limestone and Barton Clay Formations. Of these, 12 are new: Gesneropithex figularis, Nannopithex quaylei, Microchoerus wardi, M. creechbarrowensis, Europolemur collinsonae, Plesiarctomys curranti, Sciuroides rissonei, Suevosciurus authodon, Heterohyus morinionensis, Vulpavoides cooperi, Plagiolophus .curtisi and Haplobunodon venatorum. Plagiolophus curtisi is divided into two subspecies, the nominate one and P. curtisi creechensis. Treposciurus helveticus is here raised from subspecific rank and a new subspecies, T. h. preecei. is added. Some species formerly placed in Anchomomys (order Primates) are included in Gesneropithex (order Lipotyphla). The contents of the family Pseudosciuridae and of the genera Cebochoerus and Acotherulum are revised. Genus records from Creechbarrow are the earliest for Treposciurus and Suevosciurus and the latest for Nannopithex, Europolemur and Vulpavoides. Of the named taxa, 11 of the genera and all except three of the species are new records for the English Palaeogene. Biostratigraphic correlation within the Bartonian of Europe, especially that by means of mammals, is discussed. Two mammalian concurrent range biozones, spanning the late Lutetian and Bartonian, are defined. Correlation from the non-marine to marine facies is attempted by integrating the mammal data with those from other organisms. Tentative conclusions are drawn concerning depositional environments; mammalian palaeoecology is discussed for the Creechbarrow locality. Some aspects of palaeogeography are considered for the Euro- pean Bartonian. Introduction Land mammals have long been known from Ludian deposits of the Hampshire Basin (e.g. Wood 1846) and have been the subject of sume recent studies (e.g. Bosma 1974, Cray 1973, Insole 1972). They are much scarcer in the underlying marine Barton Clay and the discovery of BARTONIAN MAMMALS OF HAMPSHIRE BASIN 193 several specimens by amateur geologists at Barton Cliff was a matter of interest (Hooker 1972). However, the small number of specimens and the unlikelihood of any more coming to light seemed to preclude further studies. An independent investigation of the equivocally dated non-marine Creechbarrow Limestone, launched in 1975, led to the discovery of a rich mamma- lian fauna of Bartonian age (Hooker 1977b). The small number of mainland European Barton- ian mammal sites makes this fauna important on a continental scale. The pioneers of Creechbarrow geology, Hudleston (1902a, 1902b, 1903) and Keeping (1910, 1912), made large excavations, but without undertaking any sieving of the sediment. In recent decades sieving has come to the fore as a field technique (e.g. Kiihne 1969, McKenna 1962, Insole 1972), and the rich fauna described here was retrieved solely by this means. It was decided to undertake a detailed taxonomic treatment of all the available English Bartonian mammals, which are known only from the Hampshire Basin. Some basic litho- stratigraphic classification was found to be necessary for much of the sequence of strata, because of the mixed nature of earlier stratigraphic studies and the recent indiscriminate application of formal lithostratigraphic rank terms to non-lithostratigraphic or undefined units. The presence of land mammals in the marine as well as the fresh-water facies of the Hamp- shire Basin suggested the possibility of biostratigraphic correlation across this otherwise diffi- cult environmental boundary. A possible reconciliation of some of the discrepancies between current mammalian and marine invertebrate zonal and dating schemes in Europe was thus envisaged. A brief interpretation of the environments and mammalian palaeoecology and palaeobiogeography was also considered a useful corollary and to make the study as com- prehensive as possible. Materials and methods Collecting Collecting by the author and his colleagues was restricted to the Creechbarrow locality. Material from other localities was mainly the result of collecting by enthusiastic and competent amateur geologists. The first attempt at excavation of the Creechbarrow Limestone at Creechbarrow was carried out on a Tertiary Research Group field meeting from 19-20 July 1975. Trench excavation was begun on the east side of the hill just below the summit, but had to be abandoned because of the amount of hillwash and talus, probably from old excavations. A second excavation (Hole 1) was opened higher and immediately east of the summit trigonometrical point. In this hole, below the soil layers, marl and rotted limestone were found and 27kg were removed. This sample yielded the first remains of mammals since Keeping’s work (1912). Hole 1 was widened and deepened on 2 August 1975 to become Hole 2 (subsequently considered together with Hole 1) and 282 kg of matrix were removed. Hole 3 was dug on 6-7 December 1975 and 840 kg were removed. More intensive excavation work funded by the British Museum (Natural History) (BM(NH)) was carried out from 30 May—4 June 1976 and resulted in the removal of 4190 kg of matrix from Holes 4 and 5. A further BM(NH) excavation from 28 July-12 August 1978 resulted in Holes 6 and 7, about 5700 kg being removed from Hole 6. Matrix from Holes 1-5 was removed entire for laboratory processing. Most of that from Hole 6 was initially wet-sieved in a nearby flooded quarry, using an Alcon water pump and a stack of sieves ranging in mesh size from 10mm to 0-Smm. The concentrate was then trans- ported back to the BM(NH). A map of the holes is shown in Text-fig. 4, p. 209. The present report covers material recovered by complete processing of the marl matrix from Holes 1-3 and part of Hole 4 (bags 1—36); and initial coarse processing for larger, rarer species from the rest of Hole 4 and Holes 5—6. Matrix processed in the laboratory was first allowed to dry, weighed and allowed to break down in hot water before being sieved through 0-5 mm mesh size. The residue was dried and graded above and below 4mm mesh. The dried fraction >4mm was then sorted for fossils. The 0-5—4mm fraction was treated with dilute acetic acid to remove the limestone fragments. The insoluble residue was washed of acid salts, dried and 194 J. J. HOOKER graded above and below 1 mm mesh for greater ease of sorting. In a trial sample of the lower size fraction, nothing but a few bone fragments was found. Although all of this size fraction was retained for further concentration in the future, only that above 1 mm was sorted further for vertebrates and any other fossils that had survived the acid treatment. This fraction was found to be rich in mammals, at a concentration of about one tooth per kilogram of unprocessed matrix (dry weight). The inorganic insoluble residue in the acid-treated samples is mainly angular to subangular quartz sand. All the material collected is in the Department of Palaeontology, British Museum (Natural History) and the distribution of numbered mammalian specimens per excavation hole is record- ed below. Holes 1/2: M35390-444, M35446-599, M36499, M36790-829, M36860. Hole 3: M35600—6259, M36380—447, M36497-8, M36500. Holes 4-5: M35445, M37090-S71. Hole 6: M37690-719. Specimens already in museum collections are listed below. The BM(NH) cast number follows in brackets those in other museums, where relevant. Barton: M11090, M12346, M26176, M26238, M26552-3, M26649, M29090-1, M34864-S. Elmore: M36491, GM.978110—1 (M36493), GM.978110—2 (M36492), GM.978110—3 (M41977). Hengistbury: IGS. GSM88617 (M31996). Creechbarrow: SMC. C9968 and C9969 (M33501). Curation of small teeth The small teeth were all mounted on stainless steel entomological pins, inserted in corks, which were in turn inserted into flat-bottomed glass tubes. This allowed the register number to be written on an object (the cork) actually attached to each specimen, as well as on the glass tube, and should minimize long-term loss or muddling of specimens through study. Measurements The larger specimens were measured using an 8-inch Mitutoyo vernier caliper. Smaller speci- mens in the BM(NH) were measured using a calibrated micrometer eyepiece in a Nikon SMZ-10 binocular microscope; similarly calibrated micrometer eyepieces on a variety of bin- ocular microscopes were used in other institutions. The main tooth measurements are: Maximum mesiodistal dimension (I); Maximum buccolingual dimension (w); Maximum buccolingual dimension of trigonid of lower molariform teeth (w,); Maximum buccolingual dimension of talonid of lower molariform teeth (w,). Thus in the case of incisors, e.g. of rodents, the mesiodistal dimension is in fact in a medio- lateral direction and the buccolingual dimension in an anteroposterior direction. Where prob- lems arise and/or exceptional parameters are used, this is discussed in the text. The standard deviation (s) and coefficient of variation (v) are calculated according to the methods of Simpson et al. (1960). Photographic work Pl. 35, fig. 1c (p. 434) was taken on a Leitz Ortholux II. Most of the other light macrographs were taken using a Zeiss Tessovar or a Leitz Aristophot. In nearly every case, the specimens were first coated with ammonium chloride to overcome the confusing effects of shine on tooth enamel and of preservational colour. The scanning electron micrographs (SEM) were taken using an ISI Stereoscan 60A. The specimens for scanning electron microscopy were cleaned with acetone, mounted on aluminium stubs using Durofix and photographed at 2kv without coating. Absence of coating and use of Durofix as a mounting medium meant that the speci- mens could be easily returned to their normal storage without the tedious task of removing the coating. Teeth are figured as from the left side, i.e. figures of right teeth are shown reversed in nearly every case. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 195 Taxonomic procedure For each taxon in the systematic section (pp. 212-415), except where identification is only at a high taxonomic level, a diagnosis (new or emended) or a reference to the most recent diagnosis is given. Synonymy lists for genera are not intended to be comprehensive but to give a range of the names that have been applied. Those for species refer mainly to the British material under discussion. Historical accounts of wider synonymy are given where relevant. Procedure and terminology follows Matthews (1973). Where information from the English material has allowed clarification or changes in supra- specific groupings, either a stratophenetic (e.g. Suevosciurus) or a cladistic (e.g. Microchoerinae) approach has been employed. The choice was dependent on the nature of the available data (see Fortey & Jefferies 1982). The importance of size differences in cheek teeth of mammals for distinction at species level is recognized (see Gingerich 1974, 1976a). High coefficients of variation (i.e. c.8 or more) for mainly length measurements of preultimate molars with or without polymodality of the normal curve is taken to indicate more than one species. Whilst it is always impossible with fossils to be completely certain, if large numbers of specimens produce a normal curve with a low coefficient of variation and continuous range of morphological variation, it is highly probable that one is dealing with a biospecies. If the species concept is a difficult one, the problems of relating assemblages (especially ones in close stratigraphic succession) to named species, from which they differ only slightly, or of deciding on erection of subspecies, are almost insoluble within our present system of taxonomy. A general principle has been followed that a subspecific name may be used if the difference (either morphological or size) is noticeable but less than is found between closely related species from one locality, belonging to the same or a related genus. In cases of doubt, the specific prefix ‘aff’ has been used, ‘cf.’ being retained for uncertainty of identification on account of incom- pleteness. Abbreviations Institutions and Collections. The following abbreviations are used throughout, appended to mammal specimen numbers, except that Department of Palaeontology, BM(NH), specimen numbers are not given the BM(NH) prefix: they may be recognized either by the prefix ‘M’ or absence of a prefix for mammalian specimens, and the prefixes GG, In, P, R and V for other biota. BGS British Geological Survey, Keyworth and London (formerly Institute of Geological Sciences, London). BM(NH) British Museum (Natural History), London. CGH Crochard, Girardot, Herman Collection, Brussels, Belgium. FSL Faculté des Sciences, Université Claude Bernard, Villeurbanne, Lyon, France. GH Geiseltalmuseum, Martin-Luther-Universitat, Halle-Wittenberg, D.D.R. GIU Instituut voor Aardwetenschappen (formerly Geologisch Instituut), Rijksuniversiteit Utrecht, Netherlands. GM Gosport Museum, Hampshire. IRSNB Institut Royal des Sciences Naturelles, Brussels, Belgium. LGM Musée Géologique de Lausanne, Switzerland (LM =Lausanne Museum, Stehlin’s numbering). MNHN Institut de Paleontologie, Muséum National d’Histoire Naturelle, Paris, France. NMB Naturhistorisches Museum, Basel, Switzerland. SMC Sedgwick Museum, Cambridge. UM Laboratoire de Paléontologie des Vertébrés, Université de Montpellier, France. UMMP University of Michigan Museum of Paleontology, Ann Arbor, U.S.A. Teeth I = incisor DI = deciduous incisor C = canine P = premolar M = molar DC = deciduous canine DP = deciduous premolar J. J. HOOKER 196 Cie \ BARTONIAN MAMMALS OF HAMPSHIRE BASIN 197 Other L = left R = right ICZN = International Commission for Zoological Nomenclature. ISSC = International Subcommission on Stratigraphic Classification. Localities and stratigraphy The Bartonian Stage in the concept of continental European palaeomammalogists has com- prised the Auversian and Marinesian Substages and it is this period of time which is covered here. The recent restriction of the Bartonian by Curry (1981) is discussed below under correla- tion, p. 415. Four mammaliferous localities attributed to the Bartonian (one Auversian, three Marinesian, see correlation section) are now known in the Hampshire Basin of southern England. Text-fig. 1 shows the outcrop of Bartonian strata in the London and Hampshire Basins, i.e. upper part of Boscombe Sands; Barton Clay Formation (formally defined here for the first time); Becton Sand Formation (formally defined here for the first time); “Upper Bagshot Beds’; and Creechbarrow Limestone Formation. Hooker & Ward (1980: tab. 1) show the whole London and Hampshire Basins sequence of vertebrate-bearing Palaeogene strata. Elmore, Hampshire This locality is a foreshore exposure just SE of Lee-on-the-Solent, Hampshire (National Grid Reference SU 563001 to SZ 565996). Between here and Stubbington, 3km to the NW (SU 545018) there is a series of intermittent tidally-controlled exposures of Bracklesham Beds (Selsey and Huntingbridge divisions of Curry et al., 1977). They dip at c.2° SW and contain a fully marine invertebrate and fish fauna (see Kemp et al. 1979). The strata of the Huntingbridge division at Elmore have been described in detail by Kemp et al., (1979) and named the Elmore Formation. Its lithology is described as ‘predominantly interbedded silty clays, clayey silts and sandy silts, usually extensively bioturbated, passing in some areas to a monotonous sequence of clayey silts ... . Disseminated fine glauconite is usually present, but medium and coarse glauco- nite only occurs in the basal 5m; occasional pebbles have been seen at the basal contact’. It is thus easily distinguished from the underlying ‘shelly and nummulitic sandy clays and silty sands of the Selsey division’ from which it is separated by a disconformity. Its top is defined by the overlying Nummulites prestwichianus bed, which is a biostratigraphic unit, but which is some- times also lithologically distinctive. Apart from this marker horizon, the Elmore Formation was not clearly distinguished from the overlying Barton Clay by Kemp et al. (1979). It is here included as a member within the Barton Clay Formation. (See discussion below under Barton Cliff locality.) Text-figure 1 Outcrop map of English Bartonian strata. 1 = pre-Bartonian strata. 2 = the edge of the Tertiary outcrops, with the dots outside. 3 = Boscombe Sands. 4 = Barton Clay Formation. 5 = Becton Sand Formation. 6 = Boscombe Sands, Barton Clay Formation and Becton Sand Formation undifferentiated. 7 = Creechbarrow Limestone Formation and unnamed sands below. 8 = Upper Bagshot Beds. 9 = younger Tertiary overlying Bartonian strata. The Upper Bagshot Beds outcrop in the London Basin, the remaining Bartonian and post-Bartonian strata in the Hampshire Basin. The northern edge of the mainland outcrop of the Barton Clay Formation is shown broken. Inclusion herein of the Elmore Formation (Huntingbridge division of the Brackle- sham Group) within the Barton Clay Formation means that the basal boundary in this area has not yet been mapped. It is thus shown tentatively, based on information from Studley Wood, Hunt- ingbridge, Marchwood, Dibden, Fawley and Elmore. Small solid triangles indicate towns. Abbreviations: AB=Alum Bay; Bag= Bagshot; Bar = Barton; C= Creechbarrow; CB = Christchurch Bay; E=Elmore; He = Hengistbury; Hi= Highcliffe; Mi = Milford; Mu = Mudeford; NM = New Milton; TB = Totland Bay; WB = Whitecliff Bay. Sources of data: Reid (1898, 1902a, b) and White (1915, 1921a) and related one inch geological maps; Arkell (1947), Clarke (1981), Gardner et al. (1888) and Hooker (1975). 198 J. J. HOOKER BARTONIAN MAMMALS OF HAMPSHIRE BASIN 199 There are only four mammalian specimens in total. They were surface-collected by Messrs D. J. Kemp and W. J. Quayle and their provenance in the lower part of the Elmore Member (all that is exposed at Elmore) is based on preservation and proximity to the outcrop, coupled with the knowledge that beach material moves little. In support of this last point, the abundant beach-collected Elmore shark fauna retains its individuality from that of Long Bank, Lee (Selsey division), less than 1 km to the NW (see Kemp et al. 1979: 95). Hengistbury, Dorset This locality is a sea cliff about 1-5km long and a maximum of 35m high, on a promontory known as Hengistbury Head (SZ 1790). It comprises, from below upwards: Boscombe Sands, Barton Clay Formation (formerly known as Hengistbury Beds) and unnamed sands (incorrectly called Highcliff Sands by Gardner (1879) but herein tentatively referred to the Becton Sand Formation (q.v., p. 205). Historical reviews as well as detailed sections and faunal lists were given by Hooker (1975) and Curry (1977). The only mammalian specimen from this locality is, according to preservation, from the ‘Hengistbury Beds’ (Hooker 1977a). The ‘Hengistbury Beds’ are sandy clayey silts to clayey silty sands with pebble beds, richly glauconitic in the lower part and with siderite nodules in the upper part, with a shallow marine invertebrate and fish fauna. They are here formally included in the Barton Clay Formation (see below under Barton Cliff locality) following Prestwich’s (1849) identification, but within which they probably require separate member status. Final evidence that the ‘Hengistbury Beds’ are lateral equivalents of and do not dip below the Barton Clay has been provided by the Christchurch borehole (Freshney & Edwards 1983). It shows Barton Clay and the subjacent Boscombe Sands to rest directly on a thick sequence of non-glauconitic sands with interbedded clays and lignites. Barton Cliff, Hampshire Introduction. The sea cliffs of Christchurch Bay from Christchurch Harbour in the west to Milford-on-Sea in the east are up to 30m high and divided into four stretches by river valleys, the steep-sided ones being locally known as bunnies. The second stretch from the west is delimited westwards by Chewton Bunny (marking also the Hampshire—Dorset border) and eastwards by Becton Bunny and is known as Barton Cliff. Barton-on-Sea is situated about midway along its length. The strata in Christchurch Bay dip fairly regularly and very gently (c.2°) NE. In upward stratigraphic succession are exposed: the Boscome Sands, Barton Clay Formation, Becton Sand Formation (both defined here), Lower Headon Beds and basal unit of the Middle Headon Beds. The brackish to freshwater Lower Headon Beds in this section contain a well-known early Ludian mammal fauna at Beacon and Hordle Cliffs. At least three levels in the Barton Clay, at 35m, 40m and 48 m below the base of the Headon Beds, contain a few important marine and terrestrial mammals (PI. 1; Text-fig. 2; Halstead & Middleton 1972; Hooker 1972). Lithostratigraphy of the ‘Barton Beds’ — Historical background (see Text-fig. 3). As recorded by Curry (1958a), the term ‘Barton Clays’ was coined by Prestwich (1847: 355) for the ‘clays of Barton’. Prestwich (1846) and previous authors (see Prestwich 1846: 224-226) had considered these clays to be part of the London Clay. Although the type locality of the Barton Clay is Plate 1 Sections in Barton Clay Formation, Barton Cliff, Hampshire. For scale, see bed thicknesses in Text-fig. 2, p. 201. Fig. 1 Wave-washed lowest cliff terrace between Chewton Bunny and Barton-on-Sea. Beds A3 to D are shown. Bed C is delimited by nodule bands (n) and is also marked by a central, prominent, light-coloured non-glauconitic clay band. Talus (t) moves on a slide plane (s) over the lowest part of bed D. Fig. 2 Upper cliff terraces between Chewton Bunny and Barton-on-Sea. Beds E to F are shown, capped by Pleistocene plateau gravel (g). Nodule bands (n) separate beds E and F and also upper and lower parts of bed F. 200 J. J. HOOKER Barton Cliff, most of Prestwich’s attention to this stratigraphic unit was focused on Alum and Whitecliff Bays on the Isle of Wight, as these are the localities at which he had drawn up his detailed sections in 1846. At Alum Bay, Prestwich’s Barton Clay is easy to define because it is sandwiched between two sand units and there is a prominent pebble bed at the base. The upper sands, termed Headon Hill Sand by Prestwich (1846: 243) are succeeded everywhere by what has long been known as the Fluvio-marine series (Prestwich 1846: 243). In Christchurch Bay, however, a bed of clay divides the upper sands into two parts. As Curry (1958a: 14) noted, it is not clear whether the top of the Barton Clay at the type locality is at the top of this clay bed or lower down at the top of the main mass of clay. Moreover, it does not appear that at the time Prestwich was very well aquainted with the Barton section as he did not specify this particular point. In a later work, however (Prestwich 1857: 108), it is evident that the junction was intended to be the lower of the two alternatives. Prestwich (1857: 108) used the term ‘Barton series’, comprising the Barton Clay and Headon Hill Sand. Fisher (1862: 86-91) raised the base of the Barton Series in Alum Bay from the pebble bed of the original definition to the Nummulites prestwichianus bed, about 14m above, relegating the strata below to the Bracklesham Beds. Keeping (1887: 71) recognized the N. prestwichianus bed in Whitecliff Bay. Gardner et al. (1888) divided the Barton Beds (=Series) into Lower, Middle and Upper on the basis of faunal changes through the sequence. The Geological Survey in mapping used the terms Barton Clay and Barton Sand, the latter to replace Headon Hill Sand (Reid 1898: 30). Burton (1929) further divided the Barton Beds into thirteen lettered beds with an additional fourteenth (L) which referred to the Lignite Bed which Gardner et al. (1888: 596, fig. 4) considered to commence the Headon Beds. Burton based his beds or ‘horizons’ on faunal content but they often coincided with distinct lithologies and their limits were defined using lithological markers such as nodule bands (e.g. see Pl. 1). Curry (1958b: 12) returned to Prestwich’s concept of the base of the Barton Beds in Christchurch Bay, as this was ‘the natural break in the succession’. No major changes were then made until Stinton (1975), in order to comply with the Stratigraphical Code of the Geological Society of London (Harland et al. 1972, which differs little from Hedberg, 1972, 1976) created the term Barton Formation, incorporating the Barton Beds of Prestwich plus the Lower Headon Beds. Lithostratigraphy of the ‘Barton Beds’ — Interpretation. Prestwich’s (1847) first attempt to define the Barton Clay and Headon Hill Sand involved lithostratigraphic considerations. He ignored fauna except to show that the Barton Clay was not the same age as the London Clay. The subsequent altering of the lower limit of the Barton Clay (Fisher 1862, Keeping 1887) was based essentially on a fossil occurrence: that of N. prestwichianus, although Fisher also referred to lithological differences. Likewise the threefold division of Gardner et al. (1888) was based on biostratigraphy. Burton’s system, whilst being claimed as faunal in the title of his 1933 paper, was really a composite of biostratigraphy and lithostratigraphy. Stinton’s (1975) Barton For- mation was stated to follow Curry (1958a). However, Curry (1958a: 14) in the section dealing with Barton Beds upheld the limits and divisions of Gardner et al. and Burton. Nevertheless, he used (Curry 1958a: 5) the term ‘Bartonian Stage’ (chronostratigraphy) to comprise the time of deposition of both the Barton Beds and the Lower Headon Beds (different lithostratigraphic entities), and this may be what Stinton meant. In fact, as Stinton described it, the Barton Formation coincides exactly with the Bartonian cycle of Stamp (1921: 153). Sedimentary cycles or cyclothems are stated by Hedberg (1972: 305; 1976: 36) to be informal terms for the purposes of international stratigraphy and therefore not to form the basis of lithostratigraphic terminology. The Barton Formation has therefore no lithostratigraphic foundations, nor more- over have its included members, which coincide with the biostratigraphically-based Lower, Middle and Upper divisions of the Barton Beds of Gardner et al., plus the lower division of the Headon Beds. See also comments by Daley et al. (1979) and Stinton & Curry (1979). The original limits of the Barton Clay and Headon Hill Sand (Prestwich 1846, 1847) corre- spond closely to a basic lithostratigraphic breakdown of the sequence concerned. This fact is emphasized by Curry’s (1958a: 14) statement: ‘In areas where the Barton Beds are unfossiliferous, a lithological division into Barton Clays below (corresponding roughly to the BARTONIAN MAMMALS OF HAMPSHIRE BASIN 201 z 7) 2 c E 3 En eee Petey. 5 oes 2 Tei aicecne Key to lithological symbols used in © uw £ 9 Is Text-figs 2, 4, 5, 67 & 68: ov c 9g 7 4,9, . a esas es ~ a SI ase =| 2 Ke} (=e - a s_| 8 a eS a a 4 5 6) eee) SS ag 65 Fae eres ee it 8 9 uw ve ve a B 10 wal LS D = 3 6 \ —— 1 AANA 0 $ 13(Z°V] 14 1624 | z o 6 2 | a 5 5 3 Fre fore Sees: 16 17 BE ( | ian } \ Location of mammals: Microchiroptera sp. indet. 1 Basilosaurus sp. MIDDLE Palaeotherium aff? muehlbergi Plagiolophus curtisi (?) Zygorhiza wanklyni Zygorhiza wanklyni Basilosaurus sp. Cebochoerus minor Scale in metres: BARTON CLAY FORMATION (main mass) 5 Text-figure 2 Stratigraphical column of Barton Clay and Becton Sand Formations in Christchurch Bay. Occurrences of mammals are indicated; those on left are located with precision; those on right are located approximately according to preservation and locality data. Key to symbols: 1, silty clay; 2, sandy silty clay; 3, clayey silty sand; 4, sand; 5, calcareous sand; 6, sandy marl; 7, sandy limestone; 8, limestone; 9, siderite; 10, ‘claystone’ (?phosphatic); 11, shells abundant; 12, pebbles/cobbles; 13, burrowed horizon; 14, lignitic, often laminated clay; 15, gypsum; 16, major glauconite occurrence; 17, minor glauconite occurrence. J. J. HOOKER 202 ‘snupiydiMysadd sanijnuun yj] = ‘d'yy ‘ske[D pure spues UNO 431W = "O'S'H ‘Spurs YpoysIy = SH ‘uonewO Arenjsq = “Wy sq ‘weysopyovig = 1g ‘(adpom 1oddn) uonewioy Ae[D uoVIeg = “OG :SUONRIADIQGY ‘SUONRWIO. puvs u0}d9g puL AvI_D UOJIEg 9} 0} poragjoI soy AVG YOINYD}SIIYO,. Ul BILI}S JO SUOT]ROYIsse|O BSUNSeIIUOD ¢ JINSY-}xa], [peg ‘d “nN | Kel - g ©) beau uoyeg 4 Lz uojieg $2) H a , uoljews04 Jaquieyy 3}1104B1H JO spag $O4BIH speg uoyeg 19M07 uoyieg JaM07 JOMO7 uojVeg SIPPIN auley uojieg Jaquey usien speg u0jeg AIPPIN uoljews04 Oc Peg eweyo 7 ° ca 3 » ee ° =| uoljewso4 pues Jaqueyy u0}oeg Auung u0}90qg uooeeg uoyeg spog uojyeg saddn pues uojo0g ulasay ulasey Sl6L fe6l ‘6c6L sé6sl sesl 18 32 cg9sl ZS8L LS8L Zvst pasodoid pesodoid UO}UNS uoyng piey saupsey yausi4 YOIMIS2dq YBN UYoIMIS2dd SajoAo “yesysouyy BARTONIAN MAMMALS OF HAMPSHIRE BASIN 203 Lower and Middle Bartons) and Barton Sand(s) (Upper Bartons) above has been used’. White (1921a: 98) considered the term Headon Hill Sands to have been superseded by Barton Sand; in fact its abandonment avoids potential confusion with the Headon Beds. In order to conform with the correct application of the procedure for international strati- graphy, it is here proposed that Barton Clay Formation and Becton Sand Formation be used for Prestwich’s Barton Clays and Headon Hill Sand respectively, and that the use of the term ‘Barton Formation’ with its included members (Stinton 1975) be discontinued. This proposition both follows the new system and maintains as far as possible original definitions, providing mappable units which are a prime requisite of geological formations. Formal Designation of the Barton Clay Formation i. TERM: Barton Clay Formation. ii. STATUS: Formal lithostratigraphic unit. il. STRATOTYPE LOCALITY: Cliff sections in Christchurch Bay from Friar’s Cliff, Mudeford in the west to just east of Barton-on-Sea in the east. iv. GRID REFERENCE: SZ 194927 to SZ 242927. v. HYPOSTRATOTYPES: Cliff sections in Alum Bay (SZ 305854 to SZ 304855) and Whitecliff Bay (SZ 641862 to SZ 642863). Both show sequences more chronologically extended in Barton Clay lithology than does the stratotype. vi. LITHOLOGY. Grey to brown silty, usually shelly (but sometimes decalcified), sometimes medium to very fine sandy, clay, with subordinate clayey sandy silts. Glauconite occurrence varies from abundant large unrolled grains to absent. The glauconitic beds contain irregular patches of very coarse subangular sand. Subsidiary beds of very fine sand also occur. There are several layers of calcareous, phosphatic and sideritic nodules. The sediments are strongly bioturbated except for the sand lithology, which is laminated and shows lenticular bedding. At the western (Hengistbury) end of the depositional basin, the sediments contain more sand and silt and less clay than at the type and most other localities. The clay mineralogy is dominantly an illite/kaolinite suite, although mainly illitic/smectitic in the Elmore facies (see Curry et al. 1968; Gilkes 1968, 1978; Blondeau & Pomerol 1969). Two sharp junctions occur within the formation between cyclothems (described below). Rolled flint pebbles occasionally occur through the main mass, becoming common in the extreme west, and rolled flint cobbles and/or pebbles tend to form a basal bed. vii. Biota. An abundant marine invertebrate fauna, especially molluscs, and a notable shark tooth and teleost otolith fauna. Marine and land-derived tetrapods, as well as the fruits, seeds, cones and wood of land plants also occur. The most up-to-date comprehensive list is by Burton (1933). Subsequently, numerous specialist papers have appeared which include studies of organisms from the Barton Clay; these include foraminifera (Curry 1937, Murray & Wright 1974), brachiopods (Elliott 1954), molluscs (numerous papers by Wrigley; see list in Cox 1954), ostracods (Keen 1978: 439), malacostracan crustaceans (Quayle & Collins 1981, Quayle 1982), asteroids and ophiuroids (Rasmussen 1972), fish otoliths (Stinton 1975-80), sharks (Ward 1980), turtles (Moody 1980), birds (Walker 1980), marine mammals (Halstead & Middleton 1972), calcareous nannoplankton (Martini 1971, Aubry 1983), dinocysts and acritarchs (Bujak et al. 1980), palynomorphs (Gruas-Cavagnetto 1970), and macroplants (Chandler 1960, 1978). Some of the bivalve molluscs (e.g. Psammotaea) can be found in vertical life position in the sediment, whereas many other molluscs show evidence of having been buried as long dead shells colonized inside or out, or both, by bryozoans, oysters, serpulid worms and the gastropod Capulus, etc. vili. THICKNESS. Approximately 40m at Barton Cliff. It is c. 74m at Alum Bay and c. 60m at Whitecliff Bay. The latter is calculated at 90° to the strike; note that Gardner et al.’s (1888: 604) measurement of 368 ft 1 in (c. 112m) for the sequence of the N. prestwichianus bed to the top of the Barton Sand must have been taken directly from the cliff section which is oblique to the strike. They noted that ‘the section is not quite at right angles to the outcrop, and a diagonal direction may somewhat exaggerate the thickness’ but apparently made insufficient allowance 204 J. J. HOOKER for this. The true measurement for this sequence is 97 m (c. 318 ft). The thickness at Alum Bay is probably exaggerated by tectonic disturbance (see Curry 1942: 99). ix. RELATIONSHIPS. a. Position of the basal junction. The base bed of very glauconitic clayey sand to silt rests sharply, usually with a basal pebble bed, on different horizons of the light- coloured Boscombe Sands or non-marine lignitic clays in the west, and on more kaolinitic shelly sandy clays and clayey sands of the Selsey division (Bracklesham Group) (see Curry et al. 1968: fig. 4) in central and eastern parts of the area of deposition. b. Evidence. Kemp et al. (1979) erected the Elmore Formation for the strata between the shelly sandy clays of the Bracklesham Beds (Selsey division) and the N. prestwichianus bed of the Barton Clay in Whitecliff Bay, but excluded the clays between the pebble bed and the N. prestwichianus bed in Alum Bay as being a different facies. Two main problems are posed here by the Elmore as a formation. Firstly, it was described from a detailed section at Elmore, but as only the lower part was exposed at this locality, Whitecliff Bay was chosen as the type locality. Unfortunately no detailed section has ever been published of this part of the sequence in Whitecliff Bay, as it is normally vegetated in the cliff and obscured by beach material on the foreshore (see Prestwich 1846, Fisher 1862, Gardner et al. 1888). Recently the author had the opportunity of examining a freshly scoured exposure at the cliff foot, which showed an almost complete sequence of green, poorly glauconitic clays and silts, for about 10m below blue clays with Nummulites rectus. The occurrence of N. prest- wichianus in a richly glauconitic sand layer in the cliff, however, could not be traced to the cliff foot and may be cut out by a fault, parallel with the strike, which is visible at about the appropriate horizon (c. 6m below N. rectus according to Curry et al. 1972). Nevertheless there appeared to be no change in lithology upwards from typical Elmore facies until the N. rectus clays were reached. Secondly, the Elmore Formation includes beds of richly glauconitic sand and sandy clay with coarse quartz grit at its base (e.g. Studley Wood and Huntingbridge). These are very similar to beds at Alum Bay and Hengistbury which were expressly excluded from the formation by Kemp et al. (1979), but included in Prestwich’s definition of Barton Clay. Thus the Elmore facies appears no more distinct than any of the other various minor facies which occur within the Barton Clay; and it is more appropriate to include it within the Barton Clay Formation as a member than treat it as a formation in its own right. This procedure, moreover, does not contravene historical precedent. Fisher (1862: 80) noted that ‘the character of the matrix at Hunting Bridge approaches more nearly to some of the Barton deposits than to any of the Bracklesham strata’. It was only the fossils that caused him to state ‘the species are so decidedly of a Bracklesham type, that I have no hesitation in classing the deposit as a part of that series’. It is easy to separate Fisher’s biostratigraphy from his lithostratigraphy. Prestwich (1847) too included the upper part of the Huntingbridge division in Alum Bay in his original concept of the Barton Clay (see Fisher 1862: 87-91, and Text-figs 67—68 herein, for correlation). It seems therefore that the basal junction of the Barton Clay is successively younger when one traces it from Whitecliff Bay to Alum Bay and from there to High Cliff and Hengistbury (see Text-fig. 68 for its base at the different localities). c. Position of the top junction. Where bed J is absent, the top is the transitional boundary from predominantly clay to predominantly sand within the middle cyclothem (see below); i.e. the middle of Burton’s bed H, below the Chama Bed (sensu stricto). Where bed J is present, the top is within the upper cyclothem, the transitional boundary from predominantly clay to predominantly sand near the top of bed J. To the west (Hengistbury) the Barton Clay Forma- tion (lower part) seems to be beginning to pass laterally into a sandy, pebbly unit, but is then truncated by modern erosion (Hooker 1975). d. Evidence. In Whitecliff Bay, the upper part of the Barton Clay was called ‘Chama Bed’ by Gardner et al. (1888), in allusion to one of its shelly fossils and equivalent position in the lithological sequence to that at Barton. They noted, however, that it appeared more clayey than at the type locality. In fact, it consists of dark grey clays, becoming sandy upwards and with small tabular or irregularly ovoid siderite nodules. Its lithology thus resembles Barton bed J. It BARTONIAN MAMMALS OF HAMPSHIRE BASIN 205 rests with a sharp junction on blue-grey sandy clays, which are correlated on dinocyst evidence (Bujak et al. 1980) with the sequence upper bed H-—lower bed I at Barton. The best time correlations here are of the Whitecliff Bay ‘Chama Bed’ with Barton bed J and of the imme- diately underlying sandy clays in Whitecliff Bay with Barton bed I. This suggests that the sand lithology of bed I has wedged out eastwards or northeastwards. It is thus logical to include bed J at Barton in the Barton Clay Formation, despite Prestwich’s (1857) decision, in view of the above evidence. x. DISTRIBUTION: Limited by the area of outcrop in the Hampshire Basin. Main sections recorded: Barton Cliff, Alum Bay, Whitecliff Bay (Gardner et al. 1888), Hengistbury (Curry 1977; Hooker 1975); Studley Wood (Stinton 1970); Huntingbridge and Elmore (Kemp et al. 1979). Former sections: Afton, Gunville (Curry 1942), Hinton Admiral (Gardner et al. 1888), Bransgore, Poulner, Totton (Burton 1933), Fawley (Curry et al. 1968), Marchwood. Also numerous wells and boreholes (Whitaker 1910, 1917; Freshney 1978). xi. AGE: Bartonian (Auversian plus Marinesian) to earliest Ludian?, late middle/late Eocene. xii. SYNONYMy: Approximately equal to the Barton Clay of Prestwich (1847). Comprises Burton’s (1929, 1933) beds Al to H (lower half) and J (= Beacon Bunny Beds of Prestwich, 1857) plus the beds between the base of Al and the pebble bed, all in Christchurch Bay. Includes Hengistbury (Head) Beds of Gardner (1879) as western unit and Elmore Formation and Huntingbridge bed(s)/division as eastern unit within the lower part of the formation. Equals the lower part of the Barton Beds/series of authors. Includes the Highcliff Sand and Clay of Wright (1851) (= A3; not the Highcliff Sand of Gardner, 1879). Formal Designation of the Becton Sand Formation i. TERM: Becton Sand Formation. Derived from Becton Bunny which divides Barton Cliff from Beacon Cliff and on either side of which occur the strata defined. The riame is also chosen as the Formation nearly coincides with Stinton’s (1975: 7) informal Becton member. Although the term Barton Sand is well known (see views of Lawson 1979), the concept of a formation distinct from the Barton Clay is considered to be best achieved by avoiding name repetition. The concept of the Barton Beds as a whole is non-lithostratigraphic and requires other definitions. ii. STATUS: Formal lithostratigraphic unit. ili. STRATOTYPE LOCALITY: Cliff sections in Christchurch Bay from west of Sea Road Gap in Barton Cliff to the west, eastwards to Long Mead End (Taddiford Gap) at the eastern extrem- ity of Beacon Cliff. iv. GRID REFERENCE: SZ 229931 to SZ 262922. v. LirHOLoGy. Fine sands, largely decalcified, clayey and silty at the base. Glauconite is absent or sparse and fine-grained. Mica is frequent especially towards the top. The sediments are strongly bioturbated and, especially in the lower sand unit, display abundant pellet-lined burrows of Ophiomorpha. Rolled flint pebbles occur at the base of a clay unit in the upper part in the east of the basin (Whitecliff Bay). One sharp junction is sometimes recognizable within the formation in the west, between the middle and upper cyclothems, as a slightly indurated ferruginous band (Alum Bay). Division into two wedges occurs through interdigitation with bed J of the Barton Clay Formation. At Barton Cliff the latter is seen to pass laterally westwards into sand of the same facies as bed K. The transition takes place over a distance of about 1 km and the interlocking tongues of sand and clay are complicated by prominent dewatering structures. vi. BloTA. This is generally sparse but essentially rather similar to that of the Barton Clay Formation and many of the references given there will suffice here. The biota is, however, more restricted, especially in the upper part which is less marine. No mammals are known. vii. THICKNESS. 25m at Barton Cliff; 27-4m at Alum Bay (according to Curry & Edwards, 1972, but subject to the same tectonic disturbance as the Barton Clay Formation at this locality); and 55m in Whitecliff Bay (see comment on previously published thickness under definition of Barton Clay Formation, pp. 203-204). 206 J. J. HOOKER vill. RELATIONSHIPS. Rests with transitional, partially interdigitating, contact on Barton Clay Formation. The top is overlain sharply by the Lower Headon Beds. At the type locality the overlying bed is a thin green sandy clay, uppermost bed K of Burton (1933). This is overlain by a thick shelly lignitic clay (described as bed L of the Barton Beds by Burton (1929) and as the lignite bed of the Lower Headon Beds (‘Lower Freshwater Formation’) by Wright (1851)). The lignitic clay bed was treated as two different beds by Stinton (1975: 7). Rootlets descend from this horizon into the beds below. ix. DISTRIBUTION: Limited by the outcrop in the Hampshire Basin. Main sections: Barton and Beacon Cliffs (Burton 1929, 1933), Alum Bay, Whitecliff Bay (Gardner et al. 1888), Warwick Slade (Stinton 1970), Totland (Fowler et al. 1973). Former sections: Poulner, Fawley (Burton 1933), Brockenhurst (White 19216). Also numerous wells and boreholes (Whitaker 1910, 1917; Freshney 1978). x. AGE: Late Bartonian (? Marinesian) to early Ludian?; late middle/late Eocene. xi. SYNONYMY: Equivalent almost exactly to Headon Hill Sand (Prestwich 1846), Barton Sand (Reid 1898) and Upper Bagshot Sand of Tawney (1882). Includes Barton Beds/series in part and comprises Burton’s (1929, 1933) beds H (upper) (= Chama Bed s.s.), I and K, except uppermost green clay. Also includes the lower part of the Becton Bunny Beds (Gardner et al. 1888) and the Long Mead End Sands (Tawney 1882). Sands above the ‘Hengistbury Beds’. These at Hengistbury may tentatively be included in the Becton Sand Formation as a third, lowest wedge (see Text-fig. 68), but it cannot be confirmed that it was once continuous with known parts of the formation. Discussion of B.G.S. concept. The current concept of the Barton Formation, as developed by the B.G.S. in the course of mapping the Southampton sheet, is as follows. It comprises Barton Sand, Barton Clay, Huntingbridge division (including Elmore Formation) and sandy clays with Nummulites variolarius near the top of the Selsey division (Bracklesham Group). They divide the Barton Formation into a Becton Sand Member (= Barton Sand) and a Barton Clay Member (= the rest) (E. Freshney, personal communication 1981). Their Barton Formation is, however, mainly a cyclothem and so does not constitute a formation in Hedberg’s (1976) sense. Moreover, inclusion of the Selsey division clays causes problems. Sandy clays are interspersed with sandy silts and silty sands in the Selsey division at several levels and localities, not just at the top. Those at the top resemble those lower down in being shelly and having the same clay mineralogy (Curry et al. 1968). On the basis of detailed lithology, the latter would have to be included also, and distinctions from the Bracklesham Group would break down. The characteristics of the Barton Clay Formation as a glauconitic or non-glauconitic silty clay with only subsidiary glauconitic clayey sands are important, as are those of the Bracklesham Group as glauconitic sands and sandy clays with intercalated lami- nated non-glauconitic clays. Curry et al. (1968) distinguished three facies in the Fawley Trans- mission Tunnel and the marked change from the second to the third facies was between the Selsey and Huntingbridge divisions (see Kemp et al. 1979). This would thus seem the most appropriate position to place the Bracklesham Group—Barton Clay Formation boundary. Cyclothems in Christchurch Bay (Text-fig. 2). Two sharp breaks with erosive bases occur in the sequence of the Barton Clay and Becton Sand Formations. They separate three somewhat similar sequences of sediments. The sharp breaks occur at the junctions of Burton’s beds A3/B and I/J. The complete idealized sequence of facies is as follows: glauconitic clayey sand/silt (a); glauconitic sandy silty clay (b); glauconitic silty clay (c); non-glauconitic silty clay (d); non- glauconitic sandy silty clay (e); non-glauconitic clayey silty sand (f); and non-glauconitic sand (g). The middle sequence is complete, the lower sequence is almost complete whilst the upper only consists of the last three facies. All three coarsen upwards and therefore in part form similar cyclothems (cycles). The gradual changes within each cycle make the boundaries between Burton’s beds (except those coinciding with cycle boundaries) difficult to define, espe- cially when nodule bands or oyster beds are impersistent (e.g. in a restricted section). Certain BARTONIAN MAMMALS OF HAMPSHIRE BASIN 207 detailed differences are noticeable between the three cycles. In the lower, the change from one lithology to the next occurs initially by intercalation of a thin bed of the new type. This is followed by further intercalations which become steadily thicker until they exceed and finally exclude those of the old type. In fact in the lower cycle the non-glauconitic sand facies occurs only as intercalations with non-glauconitic silty clay (the non-glauconitic sandy clay facies being cut out); the former never excludes the latter. The middle cycle undergoes only gradual lithological changes, as does the upper. The middle cycle also has a sequence of four nodule bands. The nodules in the first and second bands (B/C and C/D junctions) are subspherical to subovoid; those of the third (E/F junction are subovoid/lenticular; and those of the fourth (within F) are lenticular. Moreover, the Stone Band (G) could be regarded as the final expres- sion of this shape trend, as it is a semi-continuous tabular layer. A slight interruption to the otherwise steady sequence of the middle cycle is the glauconitic sandy interval in the middle of D. The junction of the Barton Clay and Becton Sand Formations is difficult to recognize. A lithological boundary has previously been taken above the Chama Bed (H) and below the overlying decalcified sands (I). However, when weathered and leached (as at the top of Barton Cliff) the upper part of H appears more sandy and light-coloured and is difficult to distinguish from overlying bed I. To overcome this problem, the formational boundary has been placed at the point where sand first predominates in the sediment. This is roughly in the middle of bed H (base of Chama Bed s.s.) for the lower sand wedge (I) and the top of bed J for the upper sand wedge (K). The sharp junctions between the lower and middle and between middle and upper cycles have been modified by penecontemporaneous burrowing organisms, so that in places the junctions appear indistinct at close range. These burrows occur throughout the sequence, but are more easily seen at sharp lithological junctions. The burrows often have lined walls (Ophiomorpha), which may show glauconite or limonite enrichment. The large burrow systems from the base of B penetrate well into A3, disturbing the sediment, and often contain fossils from B which have thus been introduced secondarily into A3. To further complicate matters, in the lower part of B, some rolled shells occur with patchy black staining which is a preservation diagnostic of A3. It appears that they were derived into B following a phase of erosion which removed the top of A3. The regressive phase of the upper cycle is continued through the overlying Lower Headon Beds, where the overall facies is non-marine (see Plint (1984) for stratigraphic and sedimento- logical coverage). Creechbarrow, Dorset Of the four localities, this one has yielded by far the largest number of mammalian specimens. On Creechbarrow Hill (SY 921824) over 80m of clays, sands and conglomerates overlie the ‘Dorset Pipe Clay series’ and are capped by the Creechbarrow Limestone Formation (Hooker 1977b). The mammal fauna is from the latter. The deposits above the pipe clays occur as an outlier and are only known to outcrop on Creechbarrow Hill. There are no natural exposures, so knowledge of the succession is limited to mines in the pipe clays, old pits in the pebbly sands (Hudleston 1902a, 1903; Arkell 1947), temporary excavations in the Creechbarrow Limestone and a borehole (Pl. 2; Text-figs 4-6). Arkell (1947: 233-241) reviewed the history of investigations at the site to that date, but it was not until 30 years later that the limestone was shown to be Bartonian rather than Ypresian (Lower Bagshot’) or Ludian (Bembridge Limestone) in age (Hooker 1977b). Hooker (1977b) and Preece (1980a) also summarized the history of the site. The succession near the summit. Hudleston’s (1902a, 1903) descriptions of the succession as exposed in small, now overgrown, pits are detailed, although unfortunately he saw no need to publish a map of these (Hudleston 1902b: 169, footnote). The lithic descriptions here are limited to the exposures made between 1975 and 1978 in the region of the limestone for the purpose of collecting its biota (see Text-fig. 5). J. J. HOOKER 208 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 209 SY 9125 N 4 Lithology pees | Thickness (cm )|_Lithostratigraphy | ete D 1-40 l 20 - 40 superficial deposits 0 - 30 0 - 30 Creechbarrow 0 - 160 Limestone [ Formation 9 194.16m 0.D. N o 100 unnamed oO 1 metre es 30 sot sands 4 4 4 4 16 | Text-figure 4 Map of recent Creechbarrow exca- 16 vations. Spot height of trigonometrical point at 20 summit of the hill is shown. Excavation holes 1 a and to 7 are numbered. The position of the hand- cored borehole in hole 5 is indicated. Iso- 31 pachytes of the Creechbarrow Limestone Formation are given. clays Text-figure 5 Stratigraphical column of strata exposed in the recent Creechbarrow excava- tions. For key to lithological symbols of solid deposits, see Text-fig. 2, p. 201. For description of superficial deposits, see p. 211. The Creechbarrow Limestone Formation. This comprises bed 12 (see Text-fig. 5), a buff marl containing variable sized limestone clasts. The limestone (sensu stricto): a cream coloured, mottled with buff, soft to hard, massive micritic limestone with drusy sparite, scattered angular to subangular quartz grains up to c. 1mm diameter, oncoliths often containing gastropods (Hudleston’s ‘horseshoe concretions’), occasional slug plates, moulds of land snails and vertebrate remains. Hard limestone in the recent excavations was mainly encountered in the basal rubbly soil layer (A). None of the Plate2 Creechbarrow, Dorset. Fig. 1 General view from the south-west. The outcrop of the unnamed sands, clays and gravels is marked by the dark cover of bracken (Pteridium aquilinum (L.)). The outcrop of the Creechbarrow Limestone Formation at the summit is marked by the light cover of grass. Fig. 2 Section in north-east corner of Hole 6. For explanation of bed numbers see Text-fig. 5 (above) and text. Scale is in centimetre units. 210 J. J. HOOKER limestone appeared to be truly in situ, although Hudleston (1902a: fig. 2) claimed to have found it so in his ‘summit pit’. The buff marl: a buff calcareous silty clay with abundant angular to subangular quartz grains up to 1 mm diameter and containing limestone clasts (see above). It also contains oncoliths often containing gastropods, snail opercula, slug plates, vertebrate remains and derived silicified Cretaceous bryozoans. Interpretation of bed 12. The size and abundance of limestone clasts within the buff marl increases towards the centre of the summit of the hill, until it finally consists of a mass of large limestone boulders with the marl simply filling the narrow gaps in between. This suggests that the marl may be the result of decomposition in situ of softer parts of the limestone. A piece of soft limestone was broken down in dilute acetic acid and yielded a very poorly sorted sand with minor silt and clay components and vertebrate debris. The greater proportion of clay which occurs in the marl suggests that the above interpretation of the origin of the marl is too simple. It is perhaps more likely that there was originally a series of passage beds from manganiferous clay (bed 11) through marl to the limestone, which has become greatly modified by solution. Today, the marl and limestone lithologies do not form discrete beds. Sequences of marl through to limestone occur in several parts of the “Fluvio-marine series’ in the Isle of Wight (see Insole 1972) and some may be comparable. That solution must almost certainly have occurred is indicated by the abundance in the marl of oncolith-enclosed gastropods, Bembridgia opercula, slug plates and vertebrate remains, especially mammals, all of which have been encountered more sparsely in the soft limestone. Keeping (1912: 131) noted that ‘there is everywhere evidence of great disturbance of the strata, whether we refer this chiefly to large movements of faulting and overthrust, or the more superficial action of landslips, soil creep etc. The result has been a kneading up together of various deposits, so as to produce in many parts a mass much resembling some boulder clays’. In the recent excavations this statement is supported not only by the jumbled nature of the various limestone blocks embedded in the buff marl but also by a detached lobe of the manganiferous clay enclosed within the buff marl in the SW corner of Hole 5. Evidence for bedding is notably lacking. It is quite likely that the disturbance and solution features could have resulted from permafrost and solifluction activity during the Pleistocene. Beds I-11. These all consist of manganiferous, very poorly sorted clay/silt/sand in which the grains are subangular to subrounded and up to 1 mm in diameter (see Text-fig. 5). Bed 11: pale brown sandy silty clay with occasional angular fragments of flint. This formed the floor of the excavations, in which it was exposed in places. It also occurred in the top metre of the shallow hand-cored borehole put down from the floor of Hole 5. Beds 1-10 were exposed only in this shallow borehole. Bed 10: pale brown clayey silty sand. Bed 9: pale brown very sandy silty clay. Bed 8: pale grey silty clay with low sand content. Bed 7: pale brown very sandy silty clay. Bed 6: pale grey very silty clay with low sand content. Bed 5: pale grey sandy silty clay. Bed 4: pale brown slightly clayey silty sand, the clay content reducing considerably in middle of bed. Bed 3: pale brown very clayey silty sand. Bed 2: very pale grey calcareous silty sand. Bed 1: whitish buff very calcareous silty sand with small ovoid and tubular calcareous concretionary bodies (? oncoliths) 2-5 mm in diameter. The base of the hand-cored borehole was at 187-8m OD. After an unrecorded interval of 3-9 m, at 183-9m OD is the top of an English China Clays (E.C.C.) borehole which was put down in 1973 at SY 92138 82367 and penetrated 72-2 m of sediment, the base being at 111-7m OD. The recovery was very incomplete but the lithology was recorded as mainly hard sand or loamy sand, with frequent layers of flints which finally caused the drilling rods to jam and twist off before the pipe clays were reached. At about 150ft down (138m OD) a different lithology was BARTONIAN MAMMALS OF HAMPSHIRE BASIN 211 recorded, described as ‘almost white silty clay w. brown yellow mottle and veining throughout’. This might represent the brick clay, exposed in Mr Bond’s brickyard; but this stratum was recorded by Arkell (1947) as occurring from 20—-80ft below the limestone. There are many inconsistencies and gaps in our knowledge concerning the strata below the limestone and detailed logging of these would be a worthwhile future project. I am very grateful to English China Clays for allowing me to consult and reproduce data from their borehole log of Creech- barrow. The Superficial Deposits. Bed A is very variable in thickness and probably resulted from weathering of the limestone, perhaps once naturally exposed at the surface. Hudleston (1902a: 250) surmised that a limestone needle was once exposed at the summit of the hill and later artificially levelled to produce limestone rubble (= bed A herein). If so, then the fine sandy soil of bed B was the natural soil formed subsequent to the building of a ‘lodge’ of which only the foundations now remain. This homogeneous horizon is devoid of human debris and wedges out rapidly westwards and summitwards in Hole 6. Hudleston (1902a: 253) later contradicted his own surmise when he suggested rapid disintegration of the limestone ‘when exposed to the atmosphere’ as the only explanation for absence of fragments on the flanks of the hill and its lack of use as a building stone. It seems more likely, therefore, that bed B was the natural soil formed before human occupation. Bed C was found to contain fragments of limestone, pottery, glass and roofing ‘slate’ (composed of Purbeck limestone) and a human phalangeal bone, as well as numerous black (? burnt) patches. This could have formed during occupation and/ demolition of the ‘lodge’. Bed D, the modern topsoil, has presumably developed since demo- lition, which, according to Hudleston (19025), took place in the mid 18th Century. BN) i \a\} AO \192.90 a La borethole| | y 192.81 Text-figure 6 Three-dimensional diagrams of stratigraphy in Creechbarrow holes 1~—7. For lithologi- cal explanation see Text-figs 2, 5 and text. Figures are spot heights O.D. in metres of corners of holes measured from trigonometrical point. The floors of the excavations are shown blank. De J. J. HOOKER Fragments of soft red limestone, perhaps representing Keeping’s (1910: 437) ‘Cherry Marl’ (a workmen’s term, given informal bed status by Cooper, 1976: 9), were found in bed C. Evidence of probable burning in bed C, the omission of any mention of the bed in Keeping’s (1912: 130-131) or Hudleston’s (1902a, b, 1903) detailed sections and its absence in situ in the sections exposed during the recent excavations all suggest that this rock has been discoloured by fire: a test, which involved burning a soft piece of Creechbarrow Limestone, produced a comparable red colour. Structure. Hudleston (1902a: 249) calculated a dip of 10°-12° N for the Creechbarrow Lime- stone. Bloomfield (1913) showed similar dips for the underlying beds, but in this case appar- ently on the eastern side dipping westwards. Hudleston showed the limestone to continue some way down the northern slope of the hill and Arkell’s (1947: 233) map reflects this. In the recent excavations, the base of the limestone appeared to rise at the northern end, the outcrop edge swinging sharply to the west, thus suggesting an east—west striking synclinal structure for the beds at the summit (Text-figs 4, 6). No excavations were made on the northern slope during the recent visits, only flint gravel being seen there in a long shallow trench which may be the remains of one of Hudleston’s pits. It cannot be confirmed that limestone on the north face is in situ; but equally it cannot be affirmed that the dip of the beds in the recent excavations could not be due to recent earth movements (e.g. soil creep, solifluction, slumping due to mining of the pipe clay, etc.). Palaeogene sedimentology and structure of the whole area of the western Hampshire Basin (Wareham Basin) is the subject of a recent study by Plint (1981, unpublished Ph.D. thesis, Oxford), some aspects of which have been published (Plint 1982, 1983). Systematic palaeontology The taxa described are listed below with locality data (Table 1). The taxonomic order is essentially founded on Simpson (1945) but incorporating various subsequent modifications. Some of these are discussed under the relevant groups. Others concern the placing of Cetacea after Condylarthra from which they are considered to be derived (e.g. see Gingerich & Russell 1981); and the basic twofold division of Eutheria into ‘insectivore-primate-rodent’ and ‘carnivore-ungulate’ groups as advocated by Lillegraven (1969). Dental nomenclature diagrams for various key groups explain the terms used in this work (Text-figs 8, 31, 43 and 54). For mammalian ranges after the Bartonian, Ludian is used in preference to Priabonian and Stampian in preference to Rupelian, because of difficulties of correlation within each pair, in contrast to the relative ease of referring relevant European mammal faunas to the Ludian and Stampian. See Cavelier (1979) for information on correlating Ludian and Stampian with Pria- bonian and Rupelian. See also the notes on taxonomic procedure, p. 195. Table 1 The English Bartonian mammals and their localities: E— Elmore, H — Hengistbury, C — Creechbarrow, B-— Barton, beds C—-F. All are Class MAMMALIA Linnaeus, Subclass THERIA Parker & Haswell. Page EHCB Infraclass METATHBERIA Huxley, s..0:.0.cetecaeurnuaseencenee see eeen ence saceeeatinemseucaet 215 Superorder MARSUPIALIA Illiger, Order MARSUPICARNIVORA Ride .......... 215 Superfamily, DIDETPHOIDEA Grayirea-sceecee eco eee cee eee 215 Family, DIDEEREIDAEIGray vesaccecsrscecche ote tee ra onilae eo seleisetete eters 215 Subfamily DIDELPHINAE Gray, Tribe DIDELPHINI Gray .............. 215 Genus Amphiperatherium Filhol ..................00ecceeeeene cent cece eee eees 216 1. Amphiperatherium aff. goethei Crochet .............2.2--.0ee eee eeee eee 216 == = 2. Amphiperatherium fontense Crochet ..............0eeeeceeee eee eeeeee ees 220 ——xX-— BARTONIAN MAMMALS OF HAMPSHIRE BASIN Table 1 (continued) Page TT AGASS PEO MNES AY Grill ed erarecare setae oe ewe ee et Ea eae SEE 222 Oxceme ROME UMAERTAGROMED Sc ccc6 os cceccccmacn cosa ree eee omens 222 Pearman hy JEANAIOIL SSSI DAN BM Cte) o\vstapechadsdéoonencuoudodesocesccadoecadccaceodas 222 3, 2 IPaMtiO ESTE, oN GE SO WNC cococongsonosanonoscaonannassapoosponec 222 One? IL OINA2 silAW sbi <2 ip eeecsnnonneedsronmesrtds acne cad ccadesc dec atootceoeenober 223 SubordemuRINACEOMORPEAISabant-eessreceeeecstee saree er ererenceen eter eres: 223 Feronlhy AMP SUUL SM NOMI DYANE ISGP ocooopscccaacoss0dsuecenovosnnonooceoaanoner 223 (GENUS CEsagrO DHEA? ANAS? sm coosspeadocansoososnneanhoosbaoponppponoanaee 224 4. COMA MNDEHCIGHIS SUONS eccconsc0cdae0900004 snosoondonduRNDbOSOde 234 RanmilyNIiCemlmiEE RIDA Simpsonucessse-reeeeesceceeee nerer eter rece etre 239 Subfamily NYCTITHERIINAE Simpson ..................00..cceeceeeeeeeees 239 Genus SG. Geua Crave mace setepasece secrets teres PRA ee a oo ame 240 SS CHACUGIS PIMOS tear Grsc orate eines cickee oreo MAR nee 241 OrdenCHIROPMERABlumenbach a2... ccpsmeeie ciceciensoeeeoeiiat ects ae see eeeeecie erie 241 Suborder MICROCHIROPTERA Dobson ...............0000ceee eee e cece cece eeeeeees 241 I Reaiaot Mi/i0GV Gls tie See Ante oar eet Senre tir eines See io Goines Geist pete Donaan aopeinc acres 241 OF (ENE oy Waals WR Sag ommee neers tan eseoraed semen os GadocosedodeasaaHcnon 241 U5 SISO TES Oy 000 ee So nlersoepoeee Fave e Son ae tho ue arode ere deideaogs ataabandee 243 So Hedy GUS oh CAG Be sommmanneadeonvadsnedeeousodeson nos saboncshoosmnoaconno 245 OME PENMCES PING etn seen aac accor enienien saree aarcae ea eo nares 245 OrdemeRMMAMES sbinnacuserrasc cen aace. asec me cniaciaesoeern on cne cee reece cece 245 SOT etal RO SHMUMNOMIT Be teas yssccacscetet sido ovens eacsersocyesetcnenereusonererovolsvares chs. ate favas labs aleee taba 245 infraonrdem ARSE ORMES Gregory Gee ccriec: ceases cence sceeriet enc meceeiaece 245 HamilyiOMONMYVIDAE dnrowessant so-seeececureseracece ee aera corer teers 245 Subfamily MICROCHOERINAE Lydekker .................... 00. cece e eee 245 GenuseNannopineastenliny errr secece ete rn ae cee hearer er ear erreEre 249 OMNGnnopithexrquaylerspynOverneceeseee eee eee nee eect 249 il; INCERO DUNDES) ll sedsangudtaocdes teaonaeodsgouemep coos dededosoncdoennse 251 GenuswEscudolonisistenlinwensancsad-oece cease noe teck reine eeeaceamernrne 252 IDES Psendolonisich crusajont Wouls;Se SUGTe ee eneeeeeeeee een eee cere nee: 256 GEMS VITCHOGROCTUSIWVIOOG ore cssreict- execs cre stasa' are cicuaeeceietegante erarapaerovorerajeteione 8 ero a7 Dil, 13), AVINGROE HOG TIS WGI SO WONG: Sanacocaacsecoccondcooooedcandonsscacenoaas 259 14. Microchoerus creechbarrowensis Sp. NOV. .........+00seeee eevee eee e ees 261 iniaorder 2B MIURIF ORIMES Gregory eaneseece secre see eeeeeseeeeeee ae: 267 Rannily ADAP TDA Drouessar ts . Its curved cusp indicates it is not a lower tooth. pad—paraconid pasd—parastylid pocd—postcristid pocld—postcingulid pomecd—postmetacristid pred—protocristid precld—precingulid protd—protoconid prsd—protostylid Tald—talonid basin TaldN—talonid notch Trd—trigonid basin TrdN—trigonid notch BARTONIAN MAMMALS OF HAMPSHIRE BASIN 223 Text-figure 9 A-—B, Pantolestidae indet. from Creechbarrow. Occlusal and buccal views respectively of left P?/? (M36390). C-D, Crypto- pithecus major (Lydekker) from the Lower Headon Beds, Hordle Cliff. Occlusal and buccal views respectively of left P?/* (M34756). All x 8. D P?/3: An upper tooth from Hordle Cliff (M34756; Text-fig. 9C-D) compares well in morphol- ogy with those of Cryptopithecus sideroolithicus Schlosser 1890, identified by Heissig (1977: 216, figs 6-8) as P* and P*. It seems likely that the latter are really P? and P*, as two teeth from Hordle Cliff, with morphology typical of that of a pantolestid P*, are present in the BM(NH) collections. These are broader with a distinct protocone and are considered to belong to Cryptopithecus major (Lydekker 1887) Heissig 1977. Two teeth from Creechbarrow closely resemble M34756 but are slightly smaller. The better preserved (M36390) has a weaker para- style which is situated higher on the crown and the preparacrista extends further apically. The distal and distolingual regions are missing in both teeth. Order LIPOTYPHLA Haeckel 1866 (rank emend. Gregory 1910) Insectivore classification has had a chequered history (see Butler 1972). Certain fossil and modern groups that at one time or another have been included peripherally in the Insectivora are generally now given ordinal status of their own (e.g. Apatotheria, p. 327, and Proteutheria, p. 222). For the rest Butler (1972) advocated the use of a restricted insectivore order, using Haeckel’s (1866) name Lipotyphla, for the extant erinaceids, talpids, soricids, solenodontids, tenrecids and chrysochlorids and several extinct families. Novacek (1976) preferred to use the order Insectivora in a restricted sense for the Lipotyphla. The name Lipotyphla has the advan- tage of being narrower in its range of meaning than Insectivora. Butler (1972: 255) listed six important characters common to this group: 1, absence of caecum; 2, reduction of jugal; 3, expansion of maxilla in orbital wall, displacing palatine; 4, mobile proboscis moved by a series of muscles which influence the form of the skull; 5, reduced pubic symphysis; 6, haemochorial placenta. Characters 2-5 are the most likely to be useful in the fossil record and only 2-4 have been recognized in the early Tertiary. Moreover, skulls are rare and further extrapolation is by means of the teeth. Suborder ERINACEOMORPHA Saban 1954 (sensu Sigé 1976) Family AMPHILEMURIDAE Heller 1935 TYPE GENUS. Amphilemur Heller 1935 (including Alsaticopithecus Hurzeler 1947b). The syn- onymy of Russell et al. (1975) is upheld against Koenigswald & Storch’s (1983) claimed distinc- tions between Amphilemur and Alsaticopithecus, which I cannot substantiate. INCLUDED GENERA. Gesneropithex Hiirzeler 1946 (including Anchomomys Stehlin 1916 in part) and Pholidocercus Koenigswald & Storch 1983. 224 J. J. HOOKER RANGE. Lutetian to Ludian, Europe. EMENDED DIAGNOSIS (modified from Koenigswald & Storch 1983). Dental formula 3443. I+-P3 of approximately equal size, although C! may be rather larger and P! rather smaller. Thus relatively little differentiation of the antemolar dentition, although sharp size and com- plexity change from P3 to P4. Lower teeth mesial to P, single-rooted and bladelike, and becoming progressively procumbent and overlapping forwards. P+ premolariform. P* essen- tially without metacone; P, with unicuspid talonid or with very small entoconid and distinct cristid bisecting the distal protoconid slope and extending for the length of the talonid basin. Molars relatively bunodont with rather inflated main cusps. M+ tend to be slightly longer than M$. Upper molars with buccal angles blunt, a mesostyle at least on M?, narrow stylar shelves and well developed paraconule and metaconule with weak cristae. M' * quadrate, essentially as long as broad, with postmetacrista short and weak; paracone and metacone low and buccal and better separated from each other than metacone and paracone of successive teeth; hypocone medium to large, conical; erinaceid crest may be present. Lower molars with mesiodistally compressed trigonid, low transverse paracristid without paraconid which does not reach lingual margin; broadly rounded talonid basin; little height distinction between trigonid and talonid; talonid wider than trigonid at least on M,_,; hypoconid higher than entoconid; cristid obliqua contacts distal trogonid wall below and buccal to the base of the protoconid—metaconid notch; entoconulid present; wear moderately transverse (c. 30°—40° to crown base) and normally penetrates no lower than buccal cingulum. M,_, hypoconulid just lingual to midline, separated from postcristid by distal grooves. Protoconid lower than metaconid on M,_,, subequal on M,. Mental foramina occur below P, , below P,/M, and sometimes below P,. Dermal ossification of tail known in one genus. DISCUSSION. Koenigswald & Storch (1983) have recently provided compelling dental and post- cranial evidence, from four skeletons of Pholidocercus from the Lutetian of Messel, W. Germany, which places this family in the Lipotyphla rather than the Primates, where it has frequently been classified in the past. I (Hooker 1982) independently came to the same conclu- sion, on the basis of cranial and dental remains of Gesneropithex from the Bartonian and Ludian of the Hampshire and Paris Basins. Koenigswald & Storch (1983) included in the family the poorly known type species of Gesneropithex but excluded other species that are referred to it herein. Inclusion of the latter, which considerably expand our knowledge of the genus, requires certain changes to the family diagnosis: e.g. size does not always decrease continuously from M+ to M3, one species showing much intraspecific size variation in M3. Those charac- ters in the diagnosis which are italicized are unique to the Amphilemuridae and distinguish them from the other erinaceomorph families Adapisoricidae, Erinaceidae and Nyctitheriidae. Other characters distinguish them from only one or two of these families, but until the relation- ships within the Erinaceomorpha are better known the diagnosis cannot be made more precise. Comparison with the diagnosis of Russell et al. (1975) of the Adapisoricidae shows relatively few differences from that of the Amphilemuridae. In fact separation of the two leaves the Adapisoricidae based only on primitive characters. A possible solution to this problem may be a more detailed breakdown of the Adapisoricidae into smaller family units. This has already begun (Bown & Schankler 1982; Gingerich 1983, at subfamily level; Novack et al. 1985). Genus GESNEROPITHEX Hiirzeler 1946 TYPE SPECIES. G. peyeri Hiirzeler 1946. Ludian fissure filling; Gosgen Pumpstation, Canton Solothurn, Switzerland. INCLUDED SPECIES. G. latidens (Teilhard 1921) comb. nov.; G. grisollensis (Louis & Sudre 1975) comb. nov.; G. figularis sp. nov.; and G. sp. indet. (from Robiac; Sudre 1969a,b). RANGE. Auversian to Ludian. England, France and Switzerland. EMENDED DIAGNOsIS. Upper incisors, canine and P'/* with prominent distal cusp. Lower anterior teeth strongly procumbent. C, without talonid cusp. P? relatively narrow. P? with BARTONIAN MAMMALS OF HAMPSHIRE BASIN 225 small protocone and parastyle. P* approximately as long as broad with prominent parastyle. P, with subequal metaconid and protostylid both much smaller than protoconid; paraconid median. M'” essentially square in outline, hypocone as far from protocone as metacone is from paracone. Lower molars with paracristid far from lingual edge and without enamel wrinkling. DIFFERENTIAL DIAGNOSIS. Amphilemur and Pholidocercus have the upper teeth in front of P? with weaker distal cusp; anterior premolars with lower length : width ratio; stronger upper molar mesostyles, M’ * more trapezoidal with smaller hypocones closer to protocones; and P, without protostylid. Amphilemur has P, with a larger metaconid and more lingual paracristid. Pholidocercus has lower teeth in front of P, less procumbent; C, with talonid cusp; P? without paracone or parastyle; P* relatively shorter with smaller parastyle; P, with shorter talonid; and lower molars with paracristids extending further lingually and more concave talonid basins with enamel wrinkling. Note that Gesneropithex and Amphilemur are known mainly from teeth and it is not known whether either were like Pholidocercus in the postcranial skeleton or presence of dermal ossifi- cation. TAXONOMIC HISTORY. The treatment of species herein referred to Gesneropithex by authors is given below in chronological order. Koenigswald & Storch (1983) have comprehensively dealt with other members of the family in a similar way. Teilhard (1921) described Anchomomys latidens, but the generic assignment was provisional. The adapid Anchomomys was then considered a tarsiid. Hiurzeler (1946, 1947a) described Gesneropithex peyeri as a primate in Prosimii, incertae sedis. Simons (1962, 1963) and Crusafont-Pairo (1967) placed Gesneropithex in the Adapidae. Russell et al. (1967: 40) placed Gesneropithex provisionally in the Adapidae. Sudre (1969a) figured an M! and DP* from Robiac as Dichobune sp., an artiodactyl. Sudre (1969b) figured a mandibular fragment with M,_,; from Robiac as Gesneropithex sp. Szalay (1971b) considered Gesneropithex an adapisoricid. Szalay (1974) considered “Anchomomys’ latidens neither a close relative of Anchomomys, nor an adapid, nor a primate, but believed it to be an erinaceid insectivoran. Russell et al. (1975) synonymized Alsaticopithecus with Amphilemur, placing it in the family Amphilemuridae as Primates incertae sedis. They also selected the mandibular fragment (NMB GP128) as lectotype of Gesneropithex peyeri which they considered was not an adapi- soricid, but did not suggest what it could be. The paralectotype M? (NMB GP127) they reidentified as Amphilemur sp. Louis & Sudre (1975) described a new species which they placed in “Anchomomys’ (A, grisollensis) in ? Adapidae. They figured a probable DP* as a DP® and a DP* of Micro- choerus or Necrolemur (cf. Schmidt-Kittler 1977a: fig. 10 for Microchoerus DP*) as a DP* of ‘A. grisollensis. They considered the new species related to ‘A.’ latidens, neither being true Anchomomys and together probably with A. pygmaeus (Rutimeyer 1890) (now a Periconodon, see Gingerich, 1977a) belonging to a primate group of as yet uncertain affinities. Hooker (1977b) recorded Gesneropithex from Creechbarrow. Kay & Cartmill (1977: 36) showed that the interorbital index of ‘the erinaceoid “A.” latidens’ plotted well within the insectivores especially the tupaiids, but also alongside the micro- syopid primate Palaechthon. Sudre (1978b) reidentified the M' and DP*%, previously (1969a) identified as Dichobune sp., as Mouillacitherium aff. elegans, another dichobunid artiodactyl. Hooker & Insole (1980) reidentified the Creechbarrow Gesneropithex as Amphilemur sp. 1. Koenigswald & Storch (1983) included Gesneropithex peyeri in the Amphilemuridae but excluded ‘Anchomomys’ grisollensis. RECENT CONTRIBUTIONS. Two recent contrasting classifications of Gesneropithex are those of Szalay (1971b) and Russell et al. (1975), who also discuss Amphilemur in the same context. New 226 J. J. HOOKER material from Creechbarrow described here as Gesneropithex figularis and from Hordle Cliff identified here as Gesneropithex aff. grisollensis (Louis & Sudre 1975) has an important bearing on their contrasting views. In the light of this material, the views of Szalay and Russell et al. will be considered in detail. Szalay’s (1971b) reasons for removing Gesneropithex from the Primates and reclassifying it in the ‘Adapisorex-like erinaceotans’ were: 1. Shallower mandible than any early Tertiary primate. 2. P, unique if primate — narrow crown, mesial position of paraconid and large talonid unlike any adapid, anaptomorphid, microchoerid or early cercopithecid. Total complex of characters in P, compatible with Adapisorex gaudryi. 3. Distal border of talonid of lower molars barely rounded, and postcristid between prominent entoconid and hypoconid nearly straight when viewed in horizontal plane. 4. Tiny, centrally placed (molar) hypoconulid is deep in groove between entoconid and hypo- conid. 5. (Molar) trigonid relatively erect, unlike the slightly procumbent talonid (=trigonid?) of adapids, microchoerids or anaptomorphids. 6. Upper molar lacks parastyle and metastyle and the nature of the cuspidate hypocone indi- cates that the genus most likely does not share homologous advanced similarities with upper molars of early Tertiary primates. 7. M”’ is similar in position and conformation of hypocone to such erinaceotans as A. gaudryi, Macrocranion tupaiodon and Galerix exilis. Russell et al. (1975: 171-172) admitted that G. peyeri P,-M, showed many adapisoricid characters. They dismissed Szalay’s (1971b) arguments for transfer of Gesneropithex to the erinaceoids by equating them with those that they used to return Amphilemur to the Primates. They (1975: 172) repeated some of the distinctions from the Adapisoricidae that they had already made for Amphilemur (1975: 170-171), but did not suggest relationships with this genus. They stated (1975: 171) that some specimens of the primates Pelycodus, Hemiacodon or Omomys have a mandible shallower than in Amphilemur (Gesneropithex has a mandible of similar depth). They suggested homologies for the four and a half alveoli mesial to P, which Szalay had ignored, and considered that the most mesial, also being the largest, had housed the canine and was thus unlike an adapisoricid. Koenigswald & Storch (1983: 473-474) reinterpret- ed these alveoli as single-rooted P,—I,. Szalay’s arguments were mainly concerned with removing Gesneropithex from the Primates, whereas the arguments of Russell et al. served only to remove it from the Adapisoricidae, without suggesting what it could be. The latters’ reasons why Amphilemur was not an adapiso- ricid are of direct relevance and are listed below: 1. Either I, or lower canine missing makes the dentition atypical for the Adapisoricidae which have ——. metaconid. 3. Metaconid is high (compared to protoconid) on M,. 4. Hypoconid instead of entoconid is highest of the talonid cusps. 5. P, and molars swollen and with broad talonid. NEW DISCOVERIES. The new material provides an answer to two points raised above: the dental formula (point 1 of Russell et al. above), and the jaw depth (Szalay’s point 1 above). Dental formula. Two mandibular rami identified as G. aff. grisollensis from the Hordle Cliff Mammal Bed are complete to the very front. One shows seven alveoli in front of P,; the other shows six alveoli in front of P; which itself has fused mesial and distal roots occupying a single alveolus (see Text-fig. 1OE and G). The six teeth preserved in front of P, in A. eocaenicus are all single-rooted, and isolated teeth of almost identical morphology have also been recovered at Creechbarrow and from the Hordle Cliff Mammal Bed. Thus the most anterior alveolus in the BARTONIAN MAMMALS OF HAMPSHIRE BASIN DOs, Text-figure 10 A—D, F, Gesneropithex figularis sp. nov. from Creechbarrow. A, D, lateral and medial views respectively of right posterior mandibular fragment with M, (M35407). B, C, F, lateral, medial and anteromedial views respectively of right anterior mandibular fragment (M37406). E, Gesneropi- thex aff. grisollensis (Louis & Sudre) from the Hordle Cliff Mammal Bed; lateral view of right mandibular fragment with P;-M, (M50199). Alveoli are labelled; S = symphysis. G—H, lateral and medial views respectively of almost complete left mandibular ramus of Gesneropithex aff. grisollensis from the Hordle Cliff Mammal Bed (M50198). It shows P,—-M, and alveoli for I1,-P, and M,. I-J, lateral and medial views respectively of left mandibular ramus of Recent hedgehog, Erinaceus europaeus Linné. A-E, G-J x 2;F x 6:6. 228 J. J. HOOKER Hordle Cliff jaws must have contained an I,. This conforms with Koenigswald & Storch’s (1983) interpretation of Amphilemur eocaenicus as having possessed an I, and also the undoubted presence of this tooth in Pholidocercus. There is thus no support from the dental formula for reference of Gesneropithex to the Primates. The only primate with three incisors (Purgatorius) already shows some heterodonty here, unlike amphilemurids (Kielan-Jaworowska et al. 1979: 250). Several teeth from Creechbarrow closely resemble the tooth identified as I, (by comparison with the anteriormost tooth preserved in the jaw of A. eocaenicus) but differ in being broader and having a less oblique shape. These are herein tentatively identified as the elusive I,. The Creechbarrow edentulous anterior mandibular fragment can be matched up with the Hordle Cliff jaws and the alveoli identified as I,—-P,. Erosion at the front of the jaw means that only the deep recesses of the alveoli still remain (Text-fig. 10B, C and F). Jaw depth. One of the two mandibular rami from Hordle Cliff mentioned above (M50198) is essentially complete (Text-fig. 10G—H). It was collected broken during sieving and skilfully mended by the collector Mr R. Gardner. P,-M, are present but quite heavily worn. The horizontal ramus is shallow like A. eocaenicus and G. peyeri and the ascending ramus shows a striking resemblance to that of the modern hedgehog Erinaceus europaeus L. (see Text-fig. 10I-J). When tracings of the posterior horizontal and ascending rami of a range of early Tertiary and modern prosimian primates are superimposed on that of G. aff. grisollensis, with the bases of the tooth rows aligned, the primate ascending rami are not bent upwards and lie posterior and ventral to that of Gesneropithex (e.g. Text-fig. 11E). (Szalay & Delson’s (1979: 50, fig. 17E—-F) figure of a reconstructed Palenochtha jaw closely resembles an erinaceomorph in the characters mentioned above; however, study of the material of this genus in the UMMP shows it to be like other plesiadapiform primates and unlike erinaceomorphs.) The same is true for those prosimians with shallow as with deep jaws. It is probable that the rotation in primates was necessary to move the ascending rami backwards away from the enlarging eyes. The ascending ramus of most modern lipotyphlans and of known adapisoricids is like that of Gesneropithex in orientation. It is evident that the difference of opinion on jaw depth between Szalay (1971b) and Russell et al. (1975) is partly a question of definition. Although not necessar- ily shallower below the tooth row than in a primate, the upward bending of the ramus more posteriorly in Gesneropithex gives it a shallower appearance in this region, especially when the coronoid process is mainly broken away. Thus both dental formula and jaw shape in Gesneropithex are dissimilar from primates and similar to adapisoricids. This supports Szalay’s (1971b) conclusions and removes the first objec- tion of Russell et al. (1975) to the inclusion of Amphilemur in the Adapisoricidae. OTHER EVIDENCE. Cranium. An examination of the unique holotype cranium of G. latidens (MNHN Qu11012) from the Phosphorites du Quercy (exact horizon and locality unknown) provides further relevant characters (see Text-fig. 11A—C). The following are typical of lipotyphlans: 1. Extensive preorbital depression extending posteriorly right to orbit which has sharp laterally projecting anterior edge for accommodation and attachment of nasolabial muscles (see Butler 1956). 2. Lacrymal foramen inside orbit (as in Macrocranion, see Maier, 1979: 43). 3. When the complete mandibular ramus of G. aff. grisollensis from Hordle Cliff is occluded with the cranium of G. latidens (the two are very closely related if not conspecific) the distance between the anterior edges of the orbit and the ascending ramus leaves room for only a small eye. In addition, the infraorbital foramen is large. This excludes most primates except for the earliest genus known from cranial remains — Palaechthon (see Kay & Cartmill 1977), where the infraorbital foramen is also large. There are, however, no other special similarities between Gesneropithex and Palaechthon. Some striking similarities can, however, be observed between Gesneropithex and the modern galericine erinaceids Hylomys, Neotetracus and Podogymnura. In these genera and in tupaiids, however, the position of the orbit and infraorbital foramen vary considerably (see Table 3). BARTONIAN MAMMALS OF HAMPSHIRE BASIN 229 Table 3 Variation in position of orbit and infraorbital foramen in Gesneropithex latidens and other species. Position of Position of Species infraorbital anterior foramen orbital edge Gesneropithex latidens P* middle M? paracone Macrocranion tupaiodon P? distal edge M? paracone Palaechthon nacimienti P? middle M! paracone Hylomys suillus P* parastyle M!’ paracone Neotetracus sinensis P* parastyle M! paracone Podogymnura truei M! middle M? paracone Echinosorex gymnurus P*/M! junction M! distal half Tupaia glis P? distal edge M! mesial edge Anathana elliotti P? distal edge P* parastyle Urogale cylindrura P? middle M? middle Dendrogale melanura P? middle M! middle Ptilocercus lowi P* middle P* middle Hylomys, Neotetracus and Podogymnura have a depression for the nasolabial muscles but only in Hylomys and Neotetracus has this a dorsal limiting ridge as in Gesneropithex. The orbit, as reconstructed in Text-fig. 11B, is more like Podogymnura than the other genera in position, size and apparent weakness of the dorsal edge. The irregular grooves on the frontal region of Gesneropithex are very similar in scale and pattern to Neotetracus and Hylomys; in contrast they are more anastomosing in Podogymnura. The unique cranium of G. latidens is considerably distorted. It has been shortened, with the anterior region of the parietals having been sheared to form an imbricate structure. The right frontal has also been pushed anteroventrally to slide beneath the dorsal edge of the right maxilla, reducing the angle between the two, reducing the overall height of the cranium in this region and further shortening the whole cranium. In the reconstruction of the skull of Gesnero- pithex, an attempt has been made to reverse these processes. It is probable that the structural displacements which have taken place diagenetically were at sutures. The only ones visible on the specimen are those delimiting part of the lacrymal and maxilla (Text-fig. 11C). There is no suggestion of a jugal having a facial or orbital extension. The absence of this extension is a diagnostic lipotyphlan character but although the jugal has a facial extension in the tupatids (Order Scandentia), the sutures are often completely fused and invisible. Maier (1979: 43) stated that in Macrocranion ‘a slender process of the zygomatic (= jugal) forms the lower border of the orbit nearly reaching the lacrimal region; this is a feature more typical of the menotyphlan insectivores than of lipotyphla’; but the suture is not obvious from his figure (1979: 44, fig. 4). The orbit of G. latidens, now mainly cleaned of matrix by Dr D. E. Russell, shows the presence of a lacrymal foramen and below it the much larger opening of the infraorbital canal; the arrangement is most like Podogymnura (see Text-fig. 11C-D) among modern galericines. See also Butler (1956: fig. 7) for orbital bones of Echinosorex and contrasting insectivores. The back of the cranium has a prominent occipital crest, unlike Hylomys, Neotetracus or Podogymnura but more like the larger Echinosorex (Text-fig. 11A—B). Teeth. The most relevant features of the teeth in the Amphilemuridae are compared with the early Tertiary prosimians and the Adapisoricidae in Table 4. They further support Szalay’s (1971b) views on the erinaceoid relationships of Gesneropithex and indicate that the Amphi- lemuridae should not be included in the Order Primates. The characters 2—5 of Russell et al. (1975), listed on p. 226 as reasons why Amphilemur is not an adapisoricid, will now be considered. The P, and lower molars seem to be no more swollen and to have talenids no broader than those of some adapisoricids. The reduction of the lower molar paracristid (paraconid absent) is associated with increase in size of the upper molar J. J. HOOKER 230 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 231 Table 4 Features of the teeth of Amphilemuridae compared with early Tertiary prosimians and Adapisoricidae. Early Tertiary Character prosimians Amphilemuridae Adapisoricidae Tendency to develop Ze e Nannopithex fold y e ne Tendency to develop 3 = ‘erinaceid’ crest y YES outa expanded reduced reduced Presence of . molar paraconid yeS no Someumics Position of molar ee Baceal buccal on M, , protoconid re metaconid mesiobuccal on M, heterodont, homodont, procumbent, canine or incisor homodont, blade-like where known Anterior lower enlarged, other procumbent, (e.g. Macrocranion, teeth teeth upright or, blade-like see Tobien 1962: if procumbent, Litolestes, see Schwartz broad & Krishtalka 1976) hypocone and more transverse shear between the teeth. The increase in height of the lower molar hypoconid is associated with the greater separation of the upper molar paracone and metacone. The lower molar metaconid is the same height as the entoconid as in adapisoricids; therefore the protoconid must have reduced in size. This is likely to be related to the reduction of the upper molar metastyle causing the metacone of one tooth to lie nearer to the paracone of the following tooth. These modifications are typical of mammals whose teeth have attained the quadritubercular state, as for instance has happened independently in ungulates and primates. Most of the other characters of Amphilemur and Gesneropithex link them strongly with the Adapisoricidae. Those that do not, instead of pointing to relationships with primates, suggest modifications for a less insectivorous, more herbivorous diet, as has already been suggested for Macrocranion, a closely related adapisoricid (see Maier 1979). A possible ancestry of the Amphilemuridae could be sought among the pre-Middle Eocene Adapisoricidae. Neomatronella from the European early Eocene is a potential candidate. It shows certain characters in common with the Amphilemuridae like a similarly reduced M, talonid; lower molar paracristid not reaching lingual edge; M, _, hypoconulid distinctly Text-figure 11 A, dorsal view of holotype cranium of Gesneropithex latidens (Teilhard) (MNHN Qu11012) from the Phosphorites du Quercy; x3-3. Abbreviations: M = matrix; H = hole; F = frontal; iof = infraorbital foramen; Max = maxilla; Pat = parietal; tfe = edge of temporal fossa; L = lacrymal; O = orbitosphenoid. Oblique hatching indicates broken surface of bone and arrow the position and orientation of midline of posterior region of cranium. B, reconstruction of Gesneropithex skull, based on holotype cranium of G. latidens (MNHN Qu11012) and left (reversed) mandibular ramus (M50198) from Hordle Cliff, with addition of teeth of G. figularis sp. nov. from Creechbarrow; x 2. C—D, posterolateral views of right orbital regions of crania of Gesneropithex latidens (holotype) and Recent Podogymnura truei Mearns (BMNH ZDS53.660) respectively; x 3-3. L.F. = lacrymal foramen; arrow indicates course of infraorbital canal. E, outlines of mandibular rami of Gesneropithex aff. grisollensis (Louis & Sudre) (unbreken line) x2, and of Pseudoloris parvulus (Filhol) (broken line) about x 4, superimposed with the cheek tooth rows aligned. G. aff. grisollensis is based on a specimen from Hordle Cliff (M50198), and P. parvulus both on a specimen from Hordle Cliff (M50200) and on MNHN unnumbered from the Phosphorites du Quercy. J. J. HOOKER N N ‘(OE TLEW) (Passeacs) © 14311 ‘Z “(PZ9S EW) (Pesseacs) "dC I43tI “A-M (SOPLEW) (passaaod) ¢ AC WS ‘A— (COTPSEW) (Posseacs) SW 1YBII “L “(ES9SEIN) (P2819A2) “IW IYBII ‘S (C89SEW) 'W Yl “A (T89SEW) (Pesseres) °_ 14811 ‘© (SL 9SEW) €d Ul ‘d (ZOPSEW) “d Yl ‘O (OLOSEW) 'd Yel ‘N (899SEW) (Pesseaos) suTURD JOMOT IYBII ‘WW “(999SEW) (Passedos) ©] IYSsII “Ty (EQ9SEW) (passancg) 7] 1431 “ (6ETLEW) &'LYPl ‘£ (6S9SEW) M36227 RI? 1-95 1-10 M36226 IE 1-80 1-10 M36225 RI?? 2:20 1-10 M37404 RI?? (1:90) (1:10) M37142 RI? 1:80 1-05 M36793 E12? 2:00 1:20 M37403 IL(Ce 1:75 1-15 M37141 Ley 1-75 0:85 M36442 LIP 1-60 1:00 M35652 RDP*? DSS = M35651 EDRs? (2:40) ~ M37405 RDP*? 2:50 1-80 M36417 ER 3-05 = M37133 LIP 3-25, = M35653 Lie 3:30 3-70 M36383 ILI - - M35417 LM! 3-85 (3:75) M35654 RM! 3:90 3:90 M35419 LM! = 4-00 M37134 LM’ 3:80 = M35420* RM? 3:80 (4-40) M35418 LM? 3:70 4-25 M35655 LM? 3-90 4-50 M35657 RM? = - M37135 LM? 4-15 4-85 M37136 RM? = (4:30) M35658 LM? 2:95 SoD) M35656 LM? Sell5 PIS) M35659 LM? 3:50 4-05 M35660 LI,? = 125 M37138 RI,? 2:30 - M35661 LI ,? = ie3s M37139 LI,? 2:80 1-20 M35663 RI, 2:30 0-90 M35662 LI, DNS 0-90 M35664 RI,? = 1-15 M35665 LI, (2:40) 1-00 M35667 RI, = (0:95) M35666 RI, 2:60 0-80 M35668 RC, BESS 1:00 M35670 LP, (2:50) 1:10 M37120 LP, 2-60 1-30 M35672 RP, DDS 1-10 M35671 LP, 2:30 1-05 M35673 RP, 2:85 1-00 M37121 RP, 2:20 1-00 M35674 RP, = ~ M35669 LP, 2:20 1:10 M35675 RP, 2:65 1-30 M35676 RP, 2:40 1-25 Table 5 continued 236 J. J. HOOKER Table 5 Continued No. Tooth l Ww, W> M37123 LP, 2:65 1-65 M37122 LP, (2:90) 1-65 M35403 RP, 2:50 1-45 M35404 RP, 2:65 1-55 M35402 LIP 2-20 1-30 M35677 IL) 2:35 1-35 M35678 LP, 2:50 1:50 M35684 RDP, = 1-85 - M37124 LDP, = 1-65 = M35681 RP, 3-60 (2-15) M35680 LP, 3-40 (1-85) M35679 LP, (3:30) 2:05 M35682 RP, (3-70) 2:10 M35405 RP, - 2-05 M35406 RP, 3-60 2:05 M37125 LM, (3-75) (2°85) 3-00 M35685 LM, 4-00 2:80 (2-95) M35407 RM, in jaw 4-55 3:10 3-40 M35683 RM, 3-85 2-90 (3-10) M35409 RM, 4-40 355 3:45 M37126 RM, 4:05 3:20 3:50 M37129 RM, = = = M37128 LM, (4-00) 3-00 2:70 M37691 RM, - = DSS) M37690 LM, (3-75) 2:65 2-40 M35688 LM, 3-85 SpilS) YS M35689 RM, = = 2:60 M35412 RM; - 2:50 2:30 M35410 RM, 3i/5 2:70 DSS M35411 RM, = = 2:90 M37130 RM, 4-15 3:20 2:85 1° is intermediate in morphology between I* and the upper canine. It has small ?parastyle, the mesial margin being less ridged, and a smaller ?metastyle. The distribution of facetting is similar to that of ?I? but less extensive. The upper canine differs from that of A. leemanni in having small ?parastyle and ?metastyle. P'/? is smaller and lower-crowned than I? and has prominent parastyle and metastyle. G. latidens P? and A. leemanni ?P! in contrast have no parastyle. There appears to have been a single root pinched in the middle. All the more mesial upper teeth are single-rooted with no sign of pinching, whilst all the more distal upper teeth have three roots. P? has the protocone broken off but this cusp must have been much smaller than on P*. Length of this tooth, however, is much closer to that of P* than it is in G. latidens or G. aff. grisollensis. The morphology is, however, quite similar. Teeth differing from P? only in being smaller and having thinner enamel, a lower paracone, very small metacone on the postparacrista, larger parastyle and metastyle and very small protocone are identified as DP?. P*: on M36383, the rather buccally orientated postprotocrista joins the distal cingulum, as on G. grisollensis; on M35653 it stops short and the distal cingulum extends more lingually as in G. latidens. There is no sign of a metacone, but a small cuspule occurs in M35653 and M37133 immediately mesial to the metastyle. A similar condition occurs in a P* of G. grisol- lensis from Grisolles where there also occurs a second more mesial cuspule which probably represents a metacone. BARTONIAN MAMMALS OF HAMPSHIRE BASIN DB The morphology of the upper molars is essentially the same as in G. grisollensis. There is individual variation in the development of the pre- and post-paraconule cristae: one or other may be missing. There is also some overlap in the morphology of M? and M®. For instance the M7’, M35655, has a distal interstitial facet, but has a relatively lingually tapering outline with smaller hypocone than usual and a large mesostyle like M*. Conversely, the M*, M35656, has no distal interstitial facet and is the correct size for M* (smaller than M7), but has only a slightly lingually tapering outline and a large hypocone like M?. These variations can most simply be explained by differing positions of the field of development of various characters with respect to the tooth (see Butler, 1939, 1967). 21, differs from I, in being larger, the crown base being relatively broader, the mesial and distal edges flaring occlusally and the angle of the occlusal with the mesial edge being slightly less acute. The mesial interstitial facet is also longer and extends further down the crown, which on this side passes basally straight into the root without changing orientation. From the preserved !, alveoli on the two Hordle Cliff jaws, the tooth was just as procumbent as I, (cf. Heller 1935: pl. 1, figs la, b) but judged from the mesial interstitial facet must have been orientated more transversely in the jaw. M37139 is unworn and shows a faintly scalloped edge but is not deeply dissected like the lower incisors of the adapisoricid Litolestes (see Schwartz & Krishtalka, 1976). From overail size and the size and extent of the bevelled worn edges of M35660 and M35661, I' must have been at least as large as ?I* and have occluded along the whole length of I,. Whether or not there was an upper mesial diastema is not certain. I, is much the same as in A. eocaenicus. The mesial and distal crown edges are parallel and the mesial one flares slightly mesially away from the root. The mesial interstitial facet is restricted to the occlusal end of the mesial crown margin and is parallel to the mesial edge of the root. I, is blade-like, very procumbent, and mesially evenly tapering; the protoconid shows lin- gually as a rib but projects little if at all occlusally. The mesial interstitial facet is oblique immediately below the mesial end of the crown. The lower canine is blade-like and procumbent like I, but the mesial taper is less marked and the protoconid projects occlusally. Midway along the distal crest of the protoconid is a concave, buccally bevelled facet caused by contact with the upper canine. The mesial interstitial facet is in the same position as on I. P, is less blade-like, has a rounded distolingual shelf, a small talonid cusp and a shorter mesial overhang than in the canine. The mesial interstitial facet is more horizontal and is further under the mesial overhang. M35673 has a more delicate cuspate aspect than the other specimens and might be a deciduous tooth. P, is less procumbent than P,, has very weak development of metaconid and protostylid (often situated low down on the distal slope of the protoconid) and the talonid consists of a transverse ridge forming the distal crown margin and rising buccally to form a small ?hypoconid. P, is like P, but relatively shorter, more upright, less procumbent, and with a distinct mesially projecting paraconid. It thus contrasts with A. eocaenicus in which P, is almost identical to P,. An alternative is that some P3s have been misidentified as P,s. From alveolar evidence, this and all the more mesial teeth were single-rooted, those more distal double-rooted. P, is much the same as in G. grisollensis but the preprotocristid appears to have a more buccal sweep and the paraconid may be bicuspid. Two trigonid fragments are identified as DP, on the basis of low crown height, thin enamel and mesially projected paraconid region. Protoconid and metaconid are subequal in height. M35684 is unworn and shows a papillate paraconid. Of the lower preultimate molars, one can be confidently assigned to M,, as the tip of its broken root was found in situ in an otherwise edentulous mandibular fragment (M35407). The tooth has been restored to its original position in the jaw. There is less difference between its mesial and distal width measurements than some isolated teeth which also have a longer trigonid. The latter are here identified as M, . Overall morphology is almost identical to that of G. grisollensis M,s and Ms. 238 J. J. HOOKER > B WE) 2700. ech ord z : J 5 5 9 3.0 1 Text-figure 14 A, scatter diagrams of length (1) against width (w) in M'? and P* of species of Gesneropithex. A = G. latidens (Teilhard); Q = G. grisollensis (Louis & Sudre) from Grisolles; O = G. aff. grisollensis from the Hordle Cliff Mammal Bed; V = G. aff. grisollensis from Hordle Cliff (Rodent Bed?); © = G. figularis sp. nov. from Creechbarrow; << = G. peyeri Hurzeler from Gosgen Pumpstation (G) and Eclépens B (E); 3 = G. sp. indet. from Robiac. Symbols filled in on left are M's, those filled in on right are M’s. B, scatter diagram of length against width in I,,—P, of Gesneropithex figularis from Creechbarrow (crosses and outline symbols) and C,—P, of Amphilemur eocaenicus Heller (syntype GH.7416) from Geiseltal (solid symbols linked by line). x =1,.; + =I); 2=1,;0 =C,;0 =P,;0 =P); V = P,;; A = P,. Measurements in millimetres. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 239 M, shows considerable variation in size and hypoconulid development. The hypoconulid may be nearly as large as the entoconid, situated slightly lingual to the midline and distinctly separated from either entoconid or hypoconid (e.g. M37130) or it may be smaller, lingual in position and completely fused to the entoconid (e.g. M35410). M35688 is intermediate in morphology. Louis & Sudre (1975) described M, of G. grisollensis from three teeth whose morphology is like M37130. RELATIONSHIPS. There appear to have been at least four lineages of Gesneropithex in total, spanning Auversian to mid-Ludian (viz. G. peyeri, G. grisollensis—latidens, G. figularis and the G. sp. from Robiac). They were probably derived from a species of Amphiiemur of the early to middle Lutetian, but curiously no member of the Amphilemuridae is known from the rich late Lutetian to Auversian Egerkingen fissure deposits. Louis (1976: 49), however, recorded some teeth from the Auversian locality of Arcis-le-Ponsart, which he considered were close to G. grisollensis. Additional records of G. peyeri, as used in comparisons in this paper (e.g. Text-fig. 14A) are based on an unworn M? (NMB Mt1227) and a DP* (NMB Mt1465) from Eclépens B. It is also probable that the M! and DP* figured by Sudre (1969a, 1978b) as dichobunid artiodactyls belong to the same species as his G. sp. from Robiac (Sudre 1969b). The hypocone being larger than the metaconule on both would be unusual for Mouillacitherium, as he himself noted (1978b), and also for artiodactyls generally. The outline and cusp pattern is typical of Gesnero- pithex and probably represents a new species. An M, (CGH RC29) from Robiac is virtually unworn and shows some special features. G. figularis appears closely related to G. grisollensis, which in turn is almost indistinguishable from G. latidens in many features. Comparisons are also hampered because the holotype of G. latidens is unique and from an unknown stratigraphical level. There is some variation in development of the ‘erinaceid’ crest in G. grisollensis despite Louis & Sudre’s (1975) statement. On G. latidens there is overall less cresting of the cusps, longer mesial upper molar cingula and more continuous ectocingulum, but it is not clear whether this is merely individual variation. Specimens of Gesneropithex from the Mammal Bed and ?Rodent Bed, Hordle Cliff, are difficult to classify, some looking like G. grisollensis, others more like G. latidens. None have the P* distal cingulum bypassing the incipient hypocone and passing lingually, or such a weak distal hypocone crest as in G. latidens; in P* size the Mammal Bed specimens are more like G. grisollensis, but the P, metaconid appears to be smaller. Species of Gesneropithex have a patchy stratigraphical and geographical distribution so it is not surprising to find a new species at Creechbarrow. It seems likely that G. figularis became extinct before the Ludian, being replaced in southern England by the smaller G. aff. grisollensis. Family NYCTITHERIIDAE Simpson 1928 The modern concept of this family was first proposed by Robinson (1968). European members were treated in detail by Sigé (1976) who divided the family into two subfamilies: Nyctitheriinae Simpson 1928 and Amphidozotheriinae Sigé 1976. Earlier history of the family was briefly given by Cray (1973: 36) and need not concern us here. Subfamily NYCTITHERIINAE Simpson 1928 TYPE GENUS. Nyctitherium Marsh 1872. INCLUDED GENERA. Leptacodon Matthew & Granger 1921; Saturninia Stehlin 1940; Scraeva Cray 1973; and Pontifactor West 1974. RANGE. Middle Palaeocene to late Eocene, North America; early Eocene to early Oligocene, Europe. DiaGnosis. See Robinson (1968). 240 J. J. HOOKER Genus SCRAEVA Cray 1973 [ = Arvaldus Cray 1973] TYPE SPECIES. Scraeva hatherwoodensis Cray 1973. Microchoerus Bed, upper Middle or lower Upper Headon Beds; Headon Hill, Isle of Wight. INCLUDED SPECIES. S. woodi Cray 1973. RANGE. Bartonian (Creechbarrow Limestone) to Ludian (Bembridge Limestone), southern England. DIAGNOsIS. See Cray (1973: 37) and Sigé (1976: 11, 78). eo. t A Text-figure 15 Scraeva sp. indet., from Creechbarrow. All x 20. A-B, right P* fragment (reversed) (M35647). A is buccal and B occlusal view. C, occlusal view of left M'/? (M35648). D, occluso-buccal view of left lower molar trigonid (M35649). E-G, right M,. talonid (reversed) (M35424). E is lingual view, F buccal and G occlusal view. H—J, right P, trigonid (reversed) (M35650). H is lingual view, I buccal and J occlusal view. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 241 Scraeva sp. indet. (Text-fig. 15) v. 1977b Nyctitheriidae indet.; Hooker: 141. v 1980 Scraeva sp.; Hooker & Insole: 37. MATERIAL. Right P* buccal fragment (M35647); two left M'/? fragments (M35422, M35648); right M'/? fragment (M37143); right P, trigonid (M35650); left M,/. talonid (M35423); right M,,2 talonid (M35424); left lower molar trigonid (M35649). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION. The material representing this taxon is poor, but both M,,, talonid fragments show the presence of a mesoconid, the only character by which this genus can be distinguished from Saturninia (see Text-fig. 1SE). The size of the specimens appears intermediate between that of S. hatherwoodensis and S. woodi. No upper teeth of Scraeva have been published, although two upper molars were figured in a thesis by Insole (1972: pl. 6, figs 1-2). The right P* fragment from Creechbarrow lacks the protocone. The connate paracone and metacone are tall cusps, the latter shorter and narrower than the former. The parastyle is prominent and the tooth was very waisted between the paracone/metacone region and the (now missing) protocone region. These last two characters are present in Sigé’s (1976: 41, fig. 46; 42, fig. 48) figures of P*s of Saturninia hartenbergeri Sigé 1976. In Saturninia the P*s especially are intraspecifically variable for some features (see Sigé 1976: 22, fig. 12), but it is noteworthy that the paracone and metacone are lower and further apart in Scraeva woodi and Saturninia gracilis than in M35647. The best preserved of the three M‘'/*s (M35648) is lacking the buccal region along with the parastyle, paracone, metacone and metastyle. The postparaconule crista is weak but the preme- taconule crista is strong as in S. woodi. The hypocone is prominent and the postflexus is deep as in S. woodi. The precingulum is weak like S. woodi, not as strong as is usual in S. hartenbergeri. The development of the premetaconule crista is like not only S. woodi but also Saturninia grisollensis Sigé 1976 and S. hartenbergeri. The P, trigonid, although slightly worn, appears overall higher than in S. hatherwoodensis or S. woodi. This would fit with the relatively high paracone and metacone of P*. The metaconid appears relatively higher than in S. hatherwoodensis but this feature seems variable in S. woodi. COMMENTS ON THE STATUS OF Scraeva. It is accepted that Scraeva and Saturninia are very closely related (see Sige 1976: 78-79), but on the basis of the meagre Creechbarrow material their dichotomy must have been before the Bartonian. Oddly, however, Sigé’s (1976: 57, fig. 73) figure of M,,, of Saturninia grandis Sigé 1976 shows a mesoconid. Sigé did not mention the character in his description. As S. grandis is the largest species of Saturninia, approximately equal in size to Scraeva woodi, it is worth considering the possibility that the presence of a molar mesoconid is purely a feature of large size, not of generic relationships. This argument is weakened by the fact that another large species of Saturninia, S. beata, apparently has no mesoconid. It is also possible of course that Saturninia grandis should be considered a Scraeva. Conclusions of this kind cannot be made on the basis of the poor material from Creechbarrow but must await fuller study of material from the Fluvio-marine series in comparison with Saturninia species. Order CHIROPTERA Blumenbach 1779 Suborder MICROCHIROPTERA Dobson 1875 Microchiroptera gen. et sp. indet. 1 (Text-fig. 16) v. 1972 ?0momyoidea indet.; Hooker: 180-181. vp. 1/980 Chiroptera indet.; Hooker & Insole: 38. MatTeRIAL. Left lower molar trigonid fragment (M35710); left M,, talonid fragment (M35712); right M, talonid fragment (M29090). 242 J. J. HOOKER Text-figure 16 Microchiroptera gen. et sp. indet. 1. AE, left lower molar trigonid fragment (M35710) and left M,,. talonid fragment (M35712), almost certainly same specimen, from Creechbarrow. Views: A, buccal; B, occlusal; C, lingual; D, mesial; E, distal. F—I, right M, talonid fragment (reversed) (M29090) from the Barton Clay (Bed F?), Barton. Views: F, buccal; G, occlusal; H, lingual; I, distal. All x 16-5. HORIZONS AND LOCALITIES. M29090 is from the Barton Clay (probably around Bed F), Barton Cliff at Barton-on-Sea; the others are from the Creechbarrow Limestone Formation, Creech- barrow. DESCRIPTION. M35710 and M35712 were found in the same excavation hole. Although they are of identical preservation, size and wear their actual physical fit is poor. Nonetheless this is probably due to abrasion during weathering of the limestone and they almost certainly belong BARTONIAN MAMMALS OF HAMPSHIRE BASIN 243 to the same specimen, a left M,,.. Its length is estimated 2:00 mm, the trigonid width is 1-22 mm and talonid width 1:25mm. It is a robust, fairly low-crowned tooth. It is somewhat worn on the lingual sides of the protoconid and hypoconid tips and the paracristid. There are extensive buccal phase facets on the mesiobuccal walls of the protoconid, precingulid and hypoconid and smaller ones on the distobuccal edges of the protocristid, postmetacristid, postcristid and hypoconulid. The tip of the hypoconid is chipped on the buccal side. There is a continuous buccal cingulum termin- ating with the precingulid and postcingulid. This parallels the basal crown margin which undulates round the trigonid, reaching a low point at the mesiobuccal corner. There is essen- tially no ectoflexid. There are abraded patches on the mesial sides of the paraconid and the lingual end of the precingulid, but it is not certain whether or not they represent interstitial facets. The trigonid is long but the paracristid curves lingually to join the paraconid. The metaconid is a prominent cusp, very slightly higher than the paraconid but much lower than the proto- conid. It is strongly salient mesial of the deeply notched protocristid. The protoconid has a prominent lingual rib. The talonid is much lower than the trigonid, and narrower at the cusp tips although the same width basally. The hypoconulid is only slightly nearer the entoconid than the hypoconid, thus not being truly nyctalodont (see Menu & Sigé 1971). An important feature is the very low crestiform entoconid, which is little more than an occlusally convex bulge of the entocristid and lower than the hypoconulid. The cristid obliqua is essentially longitudinal, slightly concave buccally and joins the protoconid just buccal of the trigonid notch. M29090 is smaller than M35712 (0:93mm wide) and has no distal interstitial facet. It is slightly narrower and the hypoconulid protrudes further distally. The entoconid is worn but was of similar morphology to M35712, although probably even lower. In other respects the two talonids are almost identical. The curious character of the entoconid in both suggests that they may belong to the same taxon. Discussion. The hipposiderid Palaeophyllophora Revilliod 1917b is similar in having a rela- tively low semi-nyctalodont talonid with often weak entoconid (see Sige 1978: pl. 1, fig. 2). However, the entocristid is nevertheless high, the whole tooth is higher-crowned, the talonid is shorter relative to the trigonid, the paracristid is obliquely straight and the cusps are less prominent. The archaeonycterine palaeochiropterygid Archaeonycteris Revilliod 1917a has a lower molar trigonid rather similar to the Creechbarrow specimen (see Russell & Sigé 1970: pl. 6, figs 1c, 3). It also has a similarly-placed hypoconulid and can have a fairly weak entoconid (see Russell et al. 1973: text-fig. 61). Other European Palaeogene bats show little in common with this taxon from Creechbarrow and Barton. The low tubercular nature of the teeth and the near median hypoconulid are primitive characters and suggest affinities with the superfamily Palaeochiropterygoidea Revilliod 1917a. The special entoconid character perhaps suggests closer affinities with Archae- onycteris than with any other genus, but certain other specializations are probably only con- vergent with Palaeophyllophora. Should it become better known, it may be found to be a new genus. Microchiroptera gen. et sp. indet. 2 (Text-fig. 17A—D) vp. 1980 Chiroptera indet.; Hooker & Insole: 38. MATERIAL. Left lower molar trigonid (M35711). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION. The fragment is 1:17mm wide. The trigonid is buccally acute with a straight, slightly oblique paracristid. The buccal cingulum is complete to the precingulid. The metaconid H J Text-figure 17 Microchiroptera from Creechbarrow. A—D, left lower molar trigonid fragment (M35711) of Gen. et sp. indet. 2. Views: A, buccal; B, occlusal; C, lingual; D, mesial. E—G, left upper canine (M36790) of Gen. et sp. indet. 3. Views: E, buccal; F, occlusal; G, lingual. H—J, left lower canine (M36421) of Gen. et sp. indet. 4. Views: H, buccal; I, occlusal; J, lingual. All x 16-5. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 245 is slightly salient mesially. The protoconid is concave lingually. The paraconid is heavily worn and would probably have been slightly lower than the metaconid. The edges of the paracristid and protocristid and the tips of the protoconid and metaconid are slightly worn. The tooth as a whole is slightly corroded. Discussion. The fragment is more gracile, more crestiform and has higher cusps than M35710, which also differs in many of the characters described. It must therefore represent a different taxon from Microchiroptera gen. et sp. indet. 1. Unfortunately the talonid is missing and the morphology of the trigonid could be encountered in a very wide variety of microchiropteran taxa. Microchiroptera gen. et sp. indet. 3 (Text-fig. 17E—G) MATERIAL. Left upper canine (M36790). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION. Much of the main cusp is broken away and parts of the lingual cingulum are abraded. In occlusal view the tooth is subovoid in outline. The strong shelf-like lingual cin- gulum continues round the mesial edge where it migrates further up the crown. Distobuccally there is a slight ridge-like basal swelling not developed into a true cingulum. The lingual side of the main cusp is concave, the buccal side convex. Length is 1:77 mm, width 1:53 mm. Discussion. What is preserved of this tooth is very like the hipposiderid Pseudorhinolophus Schlosser 1887, especially in the lingual concavity. A slight difference is that the lingual cin- gulum is slightly shallower at its distal end. Size is similar to that of P. weithoferi Revilliod 1917b. If the similarities indicate true affinities with this genus, M36790 cannot belong to Microchiroptera gen. et sp. indet. 1. Although Microchiroptera gen. et sp. indet. 2, amongst many other taxa, could fit in the genus Pseudorhinolophus, it is a little small for this upper canine, M36790, to be associated with it. Microchiroptera gen. et sp. indet. 4 (Text-fig. 17H—J) MATERIAL. Left lower canine (M36421). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION. Most of the main cusp is broken away leaving only the distal and lingual areas of the base of the crown together with the root. There is a strong cingulum continuous round the region preserved. The latter is slightly cuspate distally and there are two worn papillae lin- gually. The main cusp has a prominent lingual rib and a distolingual crest. Discussion. The mesiolingual angle is more acute than it is in Pseudorhinolophus and the lingual cingulum is stronger, so it is unlikely that this tooth belongs to the same taxon as Microchiroptera gen. et sp. indet. 3, but on grounds of size it could belong to either of the other two. It is too incomplete for further comment. Order PRIMATES Linnaeus 1758 Suborder PROSIMII Illiger 1811 Infraorder TARSITFORMES Gregory 1915 Family OMOMYIDAE Trouessart 1879 (sensu Szalay 1976) Subfamily MICROCHOERINAE Lydekker 1887 (sensu Simons 1961) (including Necrolemurinae Simpson 1940 and Pseudolorisinae Simpson 1940) TYPE GENUS. Microchoerus Wood 1844. INCLUDED GENERA. Pseudoloris Stehlin 1916, Nannopithex Stehlin 1916, Necrolemur Filhol 1873 and ?Pivetonia Crusafont-Pairo 1967. 246 J. J. HOOKER RANGE. Late Ypresian to late Ludian, Europe. INTRODUCTION. I follow Szalay (1976; see especially pp. 168-170 for historical review) in allo- cating this subfamily to the Omomyidae, rather than Russell et al. (1967), who placed it in the Tarsiidae. This is done on grounds of greater morphological similarity. Classically separated from the North American Omomyidae essentially on geographical grounds, the Micro- choerinae appear not to have been diagnosed since they have acquired the present generic content, nor since being placed in the present family. It is almost certainly a natural group, as Simons (1961) thought. Szalay (1976: 413) considered that the only features of the Micro- choerinae which might separate them from the other omomyids was tibiofibular fusion and extreme inflation of the mastoid, and that the enlarged first lower incisor was a primitive Table 6 Coefficients of variation for length and width of teeth of Recent Tarsius. T. bancanus T.b.b. T.b.b. T.b.b. + T.b.n. borneanus + T. syrichta + T.b. natunensis + T. syrichta Tooth N Vv N Vv N Vv N Vv I’ ] 8 4-45 10 8-41 9 4-63 11 8-04 Ww 8 9-54 10 10-78 9 9-59 11 10-50 I? 1 7 4-35 8 4-65 8 5-63 9 5:57 Ww 7 4:47 8 10-02 8 4:17 9 9-37 C} l 8 4-33 10 528)7/ 9 5-49 11 5-87 Ww 8 4-94 10 9:97 9 4-83 11 9-46 jp l q/ 4-73 10 7-44 8 7:14 11 8-06 Ww 7 3-00 10 11-02 8 3:97 11 10-49 jp | 8 7-94 10 7-69 9 7-81 11 TESS Ww 8 3-46 10 3-15 9 3-24 11 3-00 Pe | 8 5-63 10 5-19 9 6:17 11 5-65 Ww 8 5-43 10 5:29 9 5:36 11 5:28 M’ l 8 3-68 10 3-50 9 3-69 11 3-54 Ww 8 2:39 10 3-53 9 2°58 11 3-61 M? l 7 2-94 9 2:95 8 3-06 10 3-07 Ww 7 2:29 9 2°86 8 DajsD 10 2-70 M? l 8 4-50 10 4-08 9 4-21 11 3-87 Ww 8 5:87 10 S03) 9 9759 11 5-12 I, l 7 3-92 8 4-11 8 4-11 9 4-17 Ww 7 7-97 8 7-67 8 8-62 9 8-44 Cy | 7 5-30 8 5-67 8 6-22 9 6:26 Ww 7 7-16 8 7:19 8 6°87 9 6-87 P, l 7 6-69 9 6-52 8 7-64 10 7-15 Ww 7 3-61 9 9-37 8 5-63 10 9-33 P, | 7] 5:47 8 5:97 8 5:49 9 5-83 Ww 7 5-63 8 6:14 8 5:21 9 5375) P, l 8 8-08 9 7-80 9 9-11 10 8-70 Ww 8 6:63 9 6:20 9 7-02 10 6:63 M, l 8 5-60 10 4-99 9 5-30 11 4-82 Ww 8 3-14 10 3-75 9 4:97 11 4-93 M, l 7 6-09 9 5-43 8 5-78 10 5-30 Ww 7 3-82 9 3-51 8 5-34 10 4-80 M; l 8 3-74 10 3-60 9 4:27 11 4-11 Ww 8 4:26 9 4-05 9 5-40 10 5:10 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 247 feature of the subfamily. Szalay & Delson (1979: 258) considered the conformation of the enlarged upper and lower first incisors as the primary diagnostic feature. Slight separation of the first pair of upper incisors was considered to differentiate them from the Tarsiidae and other omomyids where the anterior dentition was known. Most of these features are omitted from the diagnosis below as they are known for so few of the genera in the whole family that their value cannot be reasonably assessed. DENTAL FORMULA. This was interpreted by Simons (1961) as 547-743. Small perforations in the lower jaw anterior to the enlarged lower tooth were thought by him to be the alveoli of a very small incisor. These were reinterpreted by Gingerich (1975a, 1976b) and Szalay (1975, 1976) as anterior mental foramina. Gingerich (1976b: 91) reinterpreted the lower formula as 1133 by direct comparison with Tarsius. Szalay (1976) in contrast interpreted it as 2 1 2 3 by direct comparison with Omomys or Tetonius (presumably with respect to their mode of tooth reduction rather than their actual dental formula). If one considers details of occlusion in Microchoerus (see Simons 1961: text-fig. 2) it is more reasonable that P? should occlude with an equal-sized P, rather than with the lower canine as it would have to according to Szalay’s interpretation. The small tooth immediately distal to the enlarged lower incisor could then be either the lower canine or I,. As the larger of the two alternatives in the upper jaw is the canine, it is almost certain that this is the one which would be retained. Thus Gingerich’s interpretation of the dental formula is followed here. Consequently it seems that reduction and loss for face shortening has involved teeth different in the Microchoerinae from the Omomy- inae or Anaptomorphinae. EMENDED DIAGNOSIS. Dental formula: 74-34. I+ enlarged; upper canine and P? subequal, not greatly reduced, approximately same length as P*; P, , subequal, slightly smaller than P,; I? and lower canine greatly reduced. ORIGINS. It is unlikely that the subfamily is derived from Teilhardina as Hiirzeler (1948b) thought, because Nannopithex occurs by the late Ypresian (see Hooker & Insole 1980: 38) with already too great modification of the anterior dentition (i.e. enlargement of I+). As yet fragmen- tary omomyid remains from the English early Eocene (Hooker 1980, Hooker & Insole 1980) might provide alternative candidates. RECENT COMPARATIVE MATERIAL. Because of small sample sizes of this subfamily in the Creech- barrow assemblages, and because ranges of variation in some published measurements of Microchoerus species are large, I sought an independent control. In searching the literature for tooth measurements of the Recent tarsiid Tarsius, I found that Swindler (1976: 68, 210-211, tabs 43-46) gave statistics for length and width of C}-M3. Unfortunately Swindler pooled the different species to increase the number of measurements. I have calculated the coefficients of variation of these measurements and the majority are indeed higher than normally found in a single species (see Gingerich 1974). The numbers of specimens available to Swindler were: T. spectrum (1 male, 5 females), T. bancanus (1 male, 1 female), T. syrichta (7 males, 9 females). I therefore measured the length and width dimensions of all the teeth of Recent Tarsius bancanus and T. syrichta available in the BM(NH) Zoology Department (a maximum of 12 specimens). The coefficients of variation are given in Table 6, and plots of certain teeth in scatter diagrams in Text-fig. 18. Gingerich (1974: 896) demonstrated that low coefficients of variation (v) were characteristic for cheek teeth of a range of mammalian taxa, mainly primate (from various sources). This did not include Tarsius. Tarsius, as far as the relatively small sample of mainly one species can show, has generally low coefficients of variation for all the teeth. Intraspecific variation is high only in I' width, P? length, I, width, lower canine width and P, length. M, and M, length is more variable than in the corresponding upper teeth and may result partly in variation in size and position of the paraconid and also in its wear. P, length variation may be artificially high because it is difficult to measure, being oblique in side view. The largest sample was of T. bancanus borneanus from Borneo. Numbers of other species or subspecies were too low to be compared for v, other than by adding to the data for T. bancanus 248 J. J. HOOKER Text-figure 18 Scatter diagrams of length (1) against width (w) in upper canine, P*, M' and M? of species of Recent Tarsius; x = T. syrichta (Linné) from the Philippines; @ = T. bancanus borneanus Elliott from Borneo; [] = T. bancanus natunensis Chasen (type) from the S. Natuna Islands. Individuals of known sex are indicated. An enlarged solid circle indicates two identical measure- ments. Measurements in millimetres. borneanus to see whether this was then increased or decreased. Either or both happened with the more anterior teeth, but the molars were much more constant (see Table 6). For application of these data to the Creechbarrow microchoerines, see under the relevant species. Niemitz (1977) made a biometrical study of Recent Tarsius and reduced the recognizable species to three, and subspecies to two of T. spectrum, possibly two of T. bancanus and one of T. syrichta. He did not measure the teeth but, comprehensively done, this would seem a worthwhile project. Although the populations of Tarsius on the various Far Eastern islands BARTONIAN MAMMALS OF HAMPSHIRE BASIN 249 cannot have been separated for long geologically, it is evident that isolation has produced slightly different dimensions for some of the teeth, mainly the unicuspid anterior ones. NOTE ON MEASUREMENTS. Because of the problem in deciding on the mesiodistal and bucco- lingual parameters in I+, the length and width dimensions given for these teeth alone (including Recent Tarsius) are anteroposterior and mediolateral respectively. Genus NANNOPITHEX Stehlin 1916 TYPE SPECIES. Nannopithex pollicaris Stehlin 1916. Late Lutetian fissure fillings; Egerkingen, Grey Marl Facies (= Huppersand), Switzerland [= Necrolemur filholi Chantre & Gaillard 1897, Auversian fissure filling; Lissieu, France (fide Simons 1961: 66-67) ]. INCLUDED SPECIES. N. raabi (Heller 1930) Simons 1961, N. quaylei sp. nov. and N. sp. 1. Szalay (1971a: 2) synonymized Pivetonia Crusafont-Pairo 1967 with Pseudoloris. From Crusafont- Pairo’s figures, the M, , paraconids and long trigonids are more suggestive of Nannopithex, although size is close to Pseudoloris parvulus. Whatever the correct genus, Pivetonia isabenae Crusafont-Pairo 1967 is a valid species. RANGE. Late Ypresian to Bartonian (Marinesian), England, France, East Germany and Switzer- land. EMENDED DIAGNOSIS. Tooth enamel wrinkling may be present; mesiolingual crest from para- cone on I?-P* absent; P* * paracone—protocone crest present: P* metastylar wing short; upper molar Nannopithex fold present; M‘~? outline subtriangular to subquadrate, broader than long; M! 2 hypocone usually much smaller than protocone; M' * paracone single and meta- conule small and usually single; M’* mesostyle absent; I, narrow with distolingual cingulum strong; P, , relatively moderately elongated with high distal cingula; M, trigonid length > talonid; M,_, protocristid straight and complete; M,_, entoconid lower than metaconid; M,_, hypoconulid very small and single to absent; M, hypoconulid lobe moderately broad, short to long, unicuspid to bicuspid; paraconid present on M,_,, usually fusing with metaconid on M,; angular process may not be expanded; coronoid process may be high. Nannopithex quaylei sp. nov. (PI. 4, figs 9-14; Pl. 5, figs 1, 5; Pl. 6, fig. 1; Text-figs 19-21) Vv 1977b ?Pronycticebus sp.; Hooker: 141. v 1980 Pseudoloris sp. 1; Hooker & Insole: 39. Name. After Mr W. J. Quayle, for help with field work. Hovorype. Left M?, M37145. PI. 4, fig. 10. PARATYPES (11). Left I1 (M37154); right I’ (M35434); left P* (M35759); two right P*s, one almost complete (M35760), the other a buccal fragment (M37144); left M* lacking metaconal region (M35721); left I, (M35736); left P,; (M35748); left M,/. talonid fragment (M35720); right M,,2 talonid fragment (M35425); left M3 talonid fragment (M37147). DOUBTFUL MATERIAL. Four upper molar fragments (M35426, M35723, M36440, M37146); left M,; trigonid fragment (M37152). HORIZON AND LOCALITY. DiaGnosis. Large Nannopithex, M? length 2-75 mm; tooth enamel surface smooth; P* parastyle distinct; M, with indistinct entoconid. DIFFERENTIAL DIAGNOSIS. N. raabi, N. filholi and N. sp. 1 (p. 251) are smaller, have a distinct M, entoconid, the enamel wrinkled in varying degrees and an indistinct P* parastyle; N. raabi and N. filholi appear slightly higher-crowned; N. filholi sometimes has a larger M'? hypocone, broader bicuspid M; talonid and higher M,_, entoconid. 250 J. J. HOOKER BARTONIAN MAMMALS OF HAMPSHIRE BASIN 251 DESCRIPTION. Fragmentary representatives of several different important tooth types can be referred to this species. M3s are not much larger than those attributed to N. sp. 1 (q.v.). The characters of the left M, talonid used in the diagnosis might be due partly to individual variation. More material is obviously needed (see Text-figs 19-21 for evidence of size separation from N. sp. 1). Even though most of the teeth are slightly worn, this is not thought to have been the cause of lack of enamel wrinkling. The teeth of N. sp. 1 in the same states of wear still have visible wrinkling. The M,,, talonid fragments appear longer for their width than in the other Nannopi- thex species but trigonids are unknown. Most of the other details of the teeth are typical of other Nannopithex species and evident from the figures (PI. 4, figs 9-14, Pl. 5, figs 1, 5 and PI. 6, fig. 1). ConcLusions. For the first time two contemporaneous species of Nannopithex at a single locality are demonstrated, as well as an upward extension of the range of the genus. Nannopithex sp. | (Pl. 4, figs 1-8; PI. 5, fig. 4; Text-figs 19-21) v. 1977b Nannopithex filholi (Chantre & Gaillard); Hooker: 141. v. 1980 Nannopithex filholi (Chantre & Gaillard); Hooker & Insole: 38. MATERIAL. Left I’ (M35728); right I’ (M35740); left I? (M35741); right P? buccal half (M37151); right P?/* protoconal fragment (M37161); left M* broken buccally (M35724); right M?” paraconal fragment (M35725); left M? protoconal fragment (M35722); left P, (M37149); two right M,s (M35715-—6); right M,,, talonid fragment (M35717); two left M3s (M35427, M37153); right M, talonid fragment (M35727). HORIZON AND LOCALITY. DESCRIPTION. Much of the material from Creechbarrow is fragmentary, which makes accurate measurement difficult. The typically least variable teeth, the first and second molars, are espe- cially incomplete whereas two M33; are nearly complete. Most of the specimens compare quite well with N. filholi (including N. pollicaris) but are larger. The enamel wrinkling is variable. The M? (M35722) is quite large and closer to N. quaylei in size than to any other described Nannopithex M?s; it is tentatively distinguished on the enamel wrinkling. Two of the three M,s (M35427, M37153) are similarly large but have wrinkled enamel in the talonid basin and a prominent entoconid. A tentative conclusion from this evidence is that N. sp. 1 has proportionately larger M3s than N. filholi. An alternative explanation is that these teeth belong to N. quaylei. However, M35727 is rather smaller than the other two and more material is needed to be certain of this difference. The M,s (M35715-6) show, better than N. quaylei, the Pseudoloris-like elongation of the talonid (see Table 9). In this character and in slightly lower crown height they differ from N. raabi and N. filholi. A fairly close relationship to Pseudoloris may be indicated (see Text-fig. 24). Text-figs 20-21 demonstrate the existence of size variation greater than expected for a single species of Nannopithex at Creechbarrow, but the fragmentary nature of some crucial tooth types renders N. sp. 1 unnameable in the present state of knowledge. Plate 4 Scanning electron micrographs of occlusal views of teeth of Nannopithex from Creechbarrow, x 15. Figs 1-8 Nannopithex sp. 1. Fig. 1, left I (M35741). Fig. 2, buccal half of right P? (reversed) (M37151). Fig. 3, lingual half of right P*/* (reversed) (M37161). Fig. 4, lingual half of left M? (M35724). Fig. 5, lingual half of left M? (M35722). Fig. 6, left P, (M37149). Fig. 7, right M, (reversed) (M35715). Fig. 8, left M3 (M35427). See above. Figs 9-14 Nannopithex quaylei sp. nov. Fig. 9, right P* (reversed) (M35760). Fig. 10, holotype left M? (M37145). Fig. 11, left M? lacking metaconal region (M35721). Fig. 12, left P,; (M35748). Fig. 13, right M,,, talonid fragment (reversed) (M35425). Fig. 14, left M, talonid fragment (M37147). See p. 249. Wp) J. J. HOOKER Text-figure 19 Scatter diagrams of length (I) against width (w) in P*, M*, M, and M, of species of Nannopithex from various European localities: © = Geiseltal (N. raabi (Heller)); 1] = Bouxwiller (N. raabi); @ = Egerkingen y and Grey Marl Facies (= Huppersand) (N. filholi (Chantre & Gaillard)); A = Lissieu (N. filholi); < = Creechbarrow. Data for Egerkingen and Lissieu mainly from Hiirzeler (1948). The Creechbarrow plots represent N. quaylei sp. nov. for the upper teeth and N.sp. 1 for the lowers. Measurements in millimetres. Genus PSEUDOLORIS Stehlin 1916 TYPE SPECIES. Necrolenmur parvulus Filhol 1890b, from the Phosphorites du Quercy. INCLUDED SPECIES. P. reguanti Crusafont-Pairo 1967 and P. crusafonti Louis & Sudre 1975. RANGE. Bartonian (Marinesian) to Ludian, England, France, Switzerland, West Germany and Spain. EMENDED DIAGNOSIS. Tooth enamel wrinkling absent; mesiolingual crest from paracone on I'-P* absent; P** paracone—protocone crest absent; P* metastylar wing long; upper molar Nannopithex fold absent; M‘? outline subtriangular, broader than long; M! 7 hypocone BARTONIAN MAMMALS OF HAMPSHIRE BASIN 253 25 30 35 40 45 50 55 60 65 10 25 -30 35 40 A5 50 .55 60 65 10 25 30 35 40 45 50 55 60 65 70 2 Bevea Y coe nit 0 05 A0 15 -20 Text-figure 20 Histograms of log. length x width of: A, I’ of Nannopithex; B, I' of Recent Tarsius; C, I, of Nannopithex; D,1, of Recent Tarsius. In A and C, hollow squares = N. quaylei sp. nov. and N. sp. 1 from Creechbarrow; hatched squares = N. raabi (Heller) from Geiseltal. In B and D, hollow squares = T. bancanus Horsfield; hatched squares = T. syrichta (Linneé). Measurements in milli- metres. much smaller than protocone; M!? paraconule and metaconule small and single; M! ? meso- style absent; I, narrow with distolingual cingulum weak; P, , relatively very elongated with low distal cingula; M, trigonid length equal to talonid; M,_, protocristid straight and com- plete; M,_, entoconid same height as metaconid; M,_, hypoconulid very small and single to absent; M,; hypoconulid lobe narrow, short, unicuspid; lower molar paracristid without para- conid on M, _; and not joined to metaconid by crest; angular process not expanded; coronoid process low. o .20 25 .30 -35 Cc Text-figure 21 Histograms of: A, log. talonid width of M, of Nannopithex quaylei sp. nov. and N. sp. 1 from Creechbarrow; B, log. maximum width of M, of Recent Tarsius, hollow = T. bancanus Horsfield, hatched = T. syrichta (Linné); C, log. talonid width of M,, of N. quaylei and N. sp. 1 from Creechbarrow; D, log. talonid width of Recent Tarsius, hollow = T. bancanus, hatched = T. syrichta, dotted = M,, undotted = M, . Measurements in millimetres. Plate 5 Scanning electron micrographs of teeth of Nannopithex and Pseudoloris from Creechbarrow, <5), Figs 1,5 Nannopithex quaylei sp. nov. Fig. 1, buccal view of right P* (reversed) (M35760). Fig. Sa—c, right I’ (reversed) (M35434); a, lateral (distobuccal), b, posterior (distolingual) and c, mesial views. See p. 249. Figs 2-3 Pseudoloris cf. crusafonti Louis & Sudre. Fig. 2, occlusal view of right M? (reversed) (M37148). Fig. 3, occlusal view of right M, (reversed) (M35726), broken distally. See p. 256. Fig. 4a—c Nannopithex sp. 1. Left 1' (M35728); views as Fig. Sa—c. See p. 251. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 255 Plate 6 Scanning electron micrographs of lower incisors of Nannopithex and Microchoerus from Creechbarrow, x 15. Views as PI. 5, fig. Sa—c. Fig. la—c Nannopithex quaylei sp. nov. Left I, (M35736). See p. 249 Figs 2-3 Microchoerus wardi sp. nov. Fig. 2a—c, left I, (M35747). Fig. 3a—c, left DI, ? (M35719). See p. 259. 256 J. J. HOOKER Pseudoloris cf. crusafonti Louis & Sudre 1975 (Pl. 5, figs 2-3; Text-fig. 22) v. 1977b Pseudoloris cf. crusafonti Louis & Sudre; Hooker: 141. v. 1980 Pseudoloris crusafonti Louis & Sudre; Hooker & Insole: 39. HOLOTYPE OF SPECIES. M? (UM GRI.443). Calcaire de St Ouen (Bartonian); Grisolles, Aisne, France. DIAGNOSIS OF SPECIES. See Louis & Sudre (1975). MATERIAL. Right M? (M37148) and right M, broken distally (M35726). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION AND DISCUSSION. P. crusafonti was considered by Louis & Sudre (1975) to be intermediate in size between P. parvulus and P. reguanti. Identification of the Creechbarrow material is complicated by Schmidt-Kittler’s (1977a: 190-192, text-fig. 11) recent finds in the Ludian fissure filling of Weissenburg 8, southern Germany. On the basis of size, Schmidt- Kittler considered that two species were represented within the P. crusafonti assemblage from Grisolles, the lower molars overlapping with P. parvulus, the upper teeth being larger. Text-fig. 22 uses Schmidt-Kittler’s (1977a: text-fig. 11) and Louis & Sudre’s (1975: 824, tab. 7) data and adds new material from Creechbarrow and the Hordle Cliff Mammal Bed (the latter mainly in R. Gardner’s private collection). M37148 is slightly larger than either of the Grisolles M’s or that from Weissenburg 8. It resembles P. crusafonti in morphology except for the absence of pre- and post-metaconule cristae. At first sight this seems a relatively insignificant difference, but in all the upper molars of P. parvulus available from the Hordle Cliff Mammal Bed, presence of such cristae is constant. Oddly, M35726 plots close to P. reguanti for size, the Weissenburg 8 tooth being larger and the P. crusafonti M, being smaller. From the pattern encountered in Recent Tarsius (p. 247) and Eocene Microchoerus (p. 260), M,_> are more variable than M!? ® in size. Thus it need not be doubted that the type P. crusafonti assemblage is homogeneous. More material is needed from all localities for numerical and morphological variation to be evaluated, before definite identification of non-type material can be made. Text-figure 22 Scatter diagrams of length (1) against width (w) in M, and M? of species of Pseudoloris from various European localities: @ = Hordle Cliff Mammal Bed; © = Creechbarrow; © = Weissenburg 8; (] = type assemblage of P. crusafonti Louis & Sudre from Grisolles; A = Perriere; A = Le Bretou; @ = Malpérié; @ = holotype of P. reguanti Crusafont-Pairo from San Cugat de Gavadons; V = Eclépens [B]; WV = Nordshausen bei Kassel. Data from Weissenburg 8, San Cugat, Eclépens and Nordshausen from Schmidt-Kittler (1977a: text-fig. 11); those from Grisolles, Perriére, Le Bretou and Malpérié from Louis & Sudre (1975: tab. 7). Measurements in millimetres. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 257 Genus MICROCHOERUS Wood 1844 TYPE SPECIES. Microchoerus erinaceus Wood 1844. Lower Headon Beds (early Ludian), Hordle Cliff, England. INCLUDED SPECIES. M. edwardsi (Filhol 1880a); M. ornatus Stehlin 1916 (doubtfully distinct from M. edwardsi); M. wardi sp. nov.; M. creechbarrowensis sp. nov.; M. sp. from Grisolles. RANGE. Bartonian (Marinesian) to late Ludian, England, France, West Germany, Switzerland and Spain. EMENDED DIAGNOSIS. Tooth enamel wrinkling present; mesiolingual crest from paracone on I'_P* present; P* * paracone—protocone crest may be present; P* metastylar wing short; upper molar Nannopithex fold present; M'~? outline subquadrate, slightly broader than long to equidimensional; M’ 7 hypocone nearly as large as protocone; M’ paraconule and metacon- ule large and metaconule double; M' * mesostyle present; I, broad with distolingual cingulum strong; P, 3 relatively moderately short with high distal cingula; M, trigonid length greater than talonid; M,_, protocristid broken and displaced; M,_, entoconid lower than metaconid; M, _, hypoconulid small to large, single to double; M, hypoconulid lobe broad, moderately long to long, bicuspid; lower molar paracristid with paraconid on M,, without paraconid and fusing with metaconid on M,_, ; angular process expanded; coronoid process low. DIFFERENTIAL DIAGNOSIS. Necrolemur has no M!~? mesostyle, small or absent M, 5 hypo- conulid, smaller M'~? paraconule and metaconule; and M,_, protocristid is straight and complete to slightly broken and displaced and M3 are smaller with M, hypoconulid lobe usually shorter and narrower. EVOLUTIONARY GRADE IN Microchoerus. Louis & Sudre (1975: 819-821) described the first recorded as well as the oldest known association of Microchoerus and Necrolemur at the same locality. They listed features which for them distinguished the two genera. They stated that a mesostyle is only sometimes present on Microchoerus upper molars, in contrast to most authors (e.g. Cooper 1910, Stehlin 1916, Hill 1955, Cray 1973, Schmid 1979) who have used it as a constant diagnostic feature. Study of the sample of Microchoerus from Grisolles in the UM has shown that, although variable in size and often weak, the mesostyle is constantly present even in this early (Marinesian) assemblage. Godinot (1985) has, however, recently documented a Ludian Microchoerus lineage in which the mesostyle is sometimes absent. Another character mentioned by Louis & Sudre was the greater lingual spread of the molar paracone and metacone in Necrolemur. Although difficult to assess, this seems to result from the smaller size of the paraconule and metaconule (the latter usually doubled) in Necrolemur. An examination of the occlusal relations of the upper and lower molars of Necrolemur and Microchoerus has shown that this, along with the remaining characters listed by Louis & Sudre, constitute important grade features which, however, could not entirely separate the two genera unless they occurred in association at the same locality. These grade changes are outlined below; in each case the equivalent changes in the upper and lower teeth are placed alongside. For simplicity they are divided into two regions of the tooth: mesial and distal. A. Mesial: M?!? M, > 1. Small paraconule, single paracone Protocristid straight and unbroken. 2. Large paraconule, paracone single or Protocristid broken in middle, lingual incipiently double joined by crest. half more distal than buccal half and may be confluent with cristid obliqua. 3. Large paraconule, double paracone. Protocristid broken in middle, lingual half more distal than buccal half and may be confluent with cristid obliqua; mesoconid development on cristid obliqua. 258 J. J. HOOKER B. Distal: M2 M 1-2 1. Small metaconule, single or doubled. Postcristid high, complete and even, without hypoconulid. 2. Large buccal metaconule and small Postcristid lower, bent in middle just lingual metaconule. buccal to small hypoconulid. 3. Both metaconules large. Postcristid broken by larger doubled hypoconulids. Necrolemur zitteli is essentially at stage 1; N. antiquus may reach stage 2 (see also Godinot 1985); Microchoerus erinaceus is at stage 2 with some individuals tending towards stage 3; M. edwardsi has most individuals at stage 3; M. ornatus is at stage 3; M. wardi (p. 259) is at stage 2, with some tendency towards stage 3; M. creechbarrowensis (p. 261) is at stage 2; and M. sp. from Grisolles is between stages 1 and 2. These characters indicate slight overlap in grade between late (Ludian) Necrolemur (N. antiquus) and early (Marinesian) Microchoerus (M. sp. from Grisolles). Only the presence of a mesostyle in Microchoerus then totally distinguishes it from Necrolemur. In those from Grisol- les, it is a very weak structure. The evolution of Microchoerus from an animal with Necrolemur- Table 7 Length (l) and maximum width (w) measurements of Microchoerus species from Creechbarrow. Measurements in millimetres. Microchoerus wardi Microchoerus creechbarrowensis Tooth No. ] w No. ] . Ww Cc M37159 2:00 1-80 M35428 (2-70) 1-85 M37160 2:15 1:60 M35730 2:50 1:75 M35742 2:30 1-60 Pp? M35745 2:20 2:40 p+ M37155 2:60 3-20 M! M35436 2:90 3-40 M35731 3-25 (4-10) M35746 2:70 3-25 M35732 3:30 4-10 M37162 2:65 3-25 M2 M35437 2:60 3-45 M35733 3-00 3-90 M? M37163 2:05 2°75 M35734 2:55 3-45 M37156 2:50 3-00 I, M35747 1:70 1:65 DI,? M35719 1-30 1-05 PP M35439 2:25 1-55 1 M35432 2:50 2:00 M37157 2°85 2-50 M35431 2:95 2:10 M37158 2:85 2:25 M, M35749 2:95 2:20 M37165 3-05 2:40 M, M37166 2:95 2:30 M35737 3-25 2:60 M35750 2:90 2:05 Mi)2 M37167 = 2:15 M35739 — (2:45) M,; M35752 - 1:90 M35433 - 2:55 M35751 ~ 1-95 M35753 = 1-85 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 259 like M' ? is likely to have involved an increase in size accompanied by incipient development of a mesostyle and a molar grade change from 1 towards 2. THE SPECIES OF Microchoerus. There has long been disagreement on how best to divide the genus at species level or lower, and on what criteria. M. erinaceus and M. edwardsi have been distinguished by some on size (e.g. Depéret 1917, Insole 1972) and others on degree of com- plication of tooth enamel by wrinkling and duplication of cusps (e.g. Stehlin 1916, Hill 1955, Cray 1973, Schmidt-Kittler 1971b, Louis & Sudre 1975, Schmidt-Kittler 1977a, Schmid 1979). In the latter case, M. edwardsi has often been reduced to the rank of a subspecies of M. erinaceus, because of overlap in the variable characters which define it for these authors. Insole (1972) considered that two distinct size groups occurred together through much of the Fluvio- marine series (Ludian) of the Isle of Wight. On the basis of measurements of the type speci- mens, he attributed the larger one to M. erinaceus, the smaller to M. edwardsi. He considered M. ornatus to represent merely a variant of M. edwardsi in which enamel complication was extreme. At most horizons the sample sizes were low for at least one of the two species represented; but whereas the smaller form normally showed a duplicated molar paracone, the larger normally showed a single paracone (Insole 1972, text-figs 43-44, tab. 19). This then suggested that Schmidt-Kittler’s (1971b) evolutionary series of M. erinaceus erina- ceus to M. erinaceus edwardsi was untenable, and that various small specimens from continen- tal Europe attributed to M. erinaceus, or aff. or cf. to this species, have probably been misidentified. Insole’s scatter diagrams also suggest that the same two size types of Micro- choerus may have existed in continental Europe during the Ludian, but the sample sizes are very small. Further support, however, comes from rather large ranges of size and possible bimodality in some of Schmid’s histograms (1979: 309, fig. 5), for instance “Euzet/Perriére’. In this example, of course, the variation could alternatively have been increased by lumping two assemblages of slightly different ages. The coefficient of variation of Schmidt-Kittler’s (1971b) M, length measurements from his M. erinaceus edwardsi assemblage from Ehrenstein 1A is greater than 10. Perhaps the smaller measurements belong to Necrolemur or a similar-sized undescribed species of Microchoerus. The sample of Microchoerus teeth from the English Bartonian (Creechbarrow) is small, but two sizes appear to be represented. One is the size of M. edwardsi, the other is smaller. The M! ? plot of Microchoerus from Grisolles, according to Louis & Sudre’s (1975) measurements, is intermediate between the two from Creechbarrow (Text-fig. 23A). It is considered that, despite the small number of specimens, two species are represented at Creechbarrow, because both of the distance of size separation, and of the total size range spanned by the two (coefficient of variation of M' is greater than 10). They both also differ from M. erinaceus, M. edwardsi and the undescribed species from Grisolles. Microchoerus wardi sp. nov. (Pl. 6, figs 2-3; Pl. 7; Text-fig. 23; Tables 7-8) vp. 1977b Microchoerus sp.; Hooker: 141. v. 1980 Microchoerus sp. 1; Hooker & Insole: 38. Name. After Mr D. J. and Mrs A. Ward, for help with field work. Hotorype. Right M!, M37162. Pl. 7, fig. 4. PaRATYPES. Two left upper canines (M37159-60); right upper canine (M35742); right P? (M35745); right P* protoconal fragment (M35743); two complete right M's and one fragment (M35436, M35746, M35435); right M? (M35437); left M? (M37163); left I, with tip broken (M35747); left DI, ? (M35719); right P, (M35439); right P, (M35432); left M, (M35749); right M, (M37165); left M, (M37166); right M, (M35750); right M,,. talonid fragment (M37167); left M,,. buccal fragment (M37168); three right M,s, one lacking much of the talonid (M35752), the other two lacking the trigonid (M35751, M35753). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. J. J. HOOKER 260 ‘[ENPIAIPUL 9UO Jo Y}99} UIOf souTT ‘soI}OUNT] [TUN UL s}USWIOINSvIJ] (PO88T) [OU[LA Wor Ad1aNd 10J pur (916) Ul[YoIg WoT, SAYdOINI}U_ IOJ ‘(¢L6T) 2IPNS 2 sSInOT Wo safostiH oj (7 6]) a[osuy wos [tH UopRoH] 10) (€L6]) AID WO YD IpsoH 40} eyeq EW = [oquids pros ‘LY = joquiAs oulpjnO “g wWreideIp ul AoIINd) Wo (JOYII4) Ispapmpa ‘W Jo odAjojoy pue y WeiseIp Ul soydso1oIjUq WO UI[YI}g snipUsO “PF sdAjO[OH{ = VY smaller and relatively shorter than M*; P* much shorter than broad with weak parastyle; P, with small unicuspid talonid and metaconid present to absent; M, > protocristid and postcristid nearly transverse. J. J. HOOKER 268 ar ae =r(l am art AF nF am ar, ar é +e 7 ANNO] ysry ssad01d plouoloD Ih oF a +6 art oa + ate anes =i = é +% _ sok ou uorsuedxo ssaooid iejnsuy Op = be = = = = = = + + —% — — poautof jou poutof pluoovjoul — prjstiovied fw 6E ae = = aL = = = = - poleyur pojeyur jou aqo] plnuosodAy peoig g¢ ap 3P + + + + cts =f pr a Sate = pidsnoiq pidsnorun qo] pynuosodcy *W LE ~ + S3u0l}s yeom UOOINsuOd pryynuosodAy fe OF z = = = au = = = + + — - ~ peionjsuos. —- payorjsuooun aqoy pynuosoddy fW s¢ + + + - - 4p fe Sf peoiq MOLIBU aqo] pynuosoddy We PE + + + + $0 Sf ff ES ee Suo] joys aqo] pyynuosoddy WEE + + at ar + oF ar + + + +4 + - juosqe juasoid aqheu pluoovied *w 7¢ + + + + —-/+ -/+ = = = peoe|dsip 1ysrens pusioojoid “TW OTE A ffte = = = = = — — uay401q uayoiqun pusiojsod “"';we OE + + - + /+ —/+ = = — — s{Pprur ur yuaq yystens Pusioysod “ "WW 67 5 at = = = = = = = = = — g[qnop Ajyjensn a[3urs pynuosodky “yw gz 3F ar ar + ar - = = - yue}suoo a[qQuleA pynuosodsy “ "We LZ as = = = + + —% = - pruoovj}ow = pruoorjoul > WYyBIoy pruosojua “ ' YW 97 at st + + + + + + + + —, —/+ - ou sok pluooevied “We sz = = = = + + —% — - ou sok Ppluoovied '! PZ - = = sp AR a = = pruoje} = pruojer< yisua] pruosin "We EZ A aiff = = fot = = _ = — — — - - sok ou PIuUOoOsoU IvJOW IOMOT ZZ ar ar +% arth ar a + ate arih oar + ar = [| skemye Z SOWIT}SWIOS sjoo1 fg [Z = = 1h = = = = = 1h Se = = = I< i 2) suoniodoid m/[ * “gq 07 a =, = = - _ ~ rh P= = = MOT ysry winjnsuro yeisip © “q 61 J ae aa) +4 +e + + +4 43 EE +% +4 + }e013 WSIS uoHonpal aulue. IOMOT 8] i _ +i + = —% _ yeom suos uIN[NsUID JenBurjoysip '] LT a tug dt + + + +6 -6 - ii — proiq MOLIBU yipia "TOT Ju ue ae a ate + + + = ae — —/+ _ suol}s yeom auosodéy -,W SI ae 4 + ab = 3u0I}s yeom yyZuers a[A\sosay pT + + + + ar = = = - — — - — sok ou g[Aisosoul -_,W EI aT ee ft = =f. = /+ = = = — _ a[qnop [suis auoovied , WW ZI + + - + fe afte fap GP SSR 21qnop 9[3uls gjnuosejeut -_ WW TT ae ab 4 + =e = /)e = _ _ - = -- _ adie] [jeuls ojnuoovjour pue ajnuoovied - We Ol + + + + + - + — = yysua] = yisue] < PIM |W 6 ae ae ae + 4 + + + /+ - ayeipenbqns = sejn3urisjqns oulyino -_,N 8 + + — _ - ou sok Ploy xaysidouuvyy rejou teddy, 1 = = = _ = = = = pes, nls = = Suoms yeom qjAjseied .q 9 2s = = = = = = = + + = = ~ suo] joys SUIM IeJA]seJOU ,q ¢ + + + 4p + + + + try ar aA +6 + ye913 1YsIIs uononpal 7] + ae a si + a 4b + ate = — — - ~ sok ou jSso19 auoovIed [ensuljoissy = ¢ + + + = = — - _ poonpel poonpal jou EN 7 + + + - - + + + SoS = a 4 sak ou SulpULM joweug | = = =< = = z z z Sos 2 2 z poourape aatytutid Jg}0vIeYD SF 8. eS EG ee ae Be kG ge See en gay oo ae eS Se a es 2 = ey a 5 = g ie iss) — na = 2 A = BARTONIAN MAMMALS OF HAMPSHIRE BASIN 269 3 : t 3 ~ 6 pa a i 7 2 6 S = = 3 3 3 8 3 2 Se 28 | ee o ES a > = cS ov = coe 5 0 i) 3 > 3 a Sey aN = = 8 z 9 = 2 Ry Q a = i & = ry ws ® a = 2 oO a N = o oO 3 oO . & ) a a C) o a o % a 2 2 2 2 2 = 2 = = pant 3 2 2 SS #8 10,29,31 XS a, 64 We 5,717,19,20,24,26,35,39 Text-figure 24 Cladogram of species of the omomyid subfamily Microchoerinae. This is probably one of the most parsimonious possibilities but has not been subjected to computer analysis. Some characters are unknown for a few species only (e.g. Ps. crusafonti, Na. filholi, Na. quaylei), whilst others are generally poorly known (e.g. 40, 41). For instance, characters 40 and 41 are known in most species of Necrolemur and Microchoerus but it is not known whether they were incipiently developed in Na. filholi like characters 8, 11, 34 and 37. Na. filholi appears to be very variable morphologically (see Hurzeler 1948b), some individuals differing little from Na. raabi, while others show marked trends towards Necrolemur and Microchoerus. At terminal or near terminal branchings, often only size (the larger indicated by *) or a parallelism (+) are available as advanced character states. The large number of characters separating Pseudoloris from the rest might indicate a more distant relationship for this genus or alternatively a lack of intermediates. The correct alternative might become apparent from a detailed study of all three omomyid subfamilies and a better knowledge of Pivetonia. Abbreviations: M.= Microchoerus; Na. = Nannopithex; Ne. = Necrolemur; Pi. = Pivetonia; and Ps. = Pseudoloris. Character numbers are identified in Table 9, opposite. DIFFERENTIAL DIAGNOSIS. Protoadapis differs in having M'~ with small metaconules, P* with large parastyle and P, with constant metaconid. Cercamonius has P, with metaconid and P, greatly reduced. Mahgarita has M' ? metaconule, P$ greatly reduced, P* with reduced protocone. Pronycticebus and Donrussellia have large P* parastyle, transversely elongated M!? with metaconules and P, with metaconid and bicuspid talonid. Periconodon has M' ~~ metaconule and pericone and weak, distally placed P, metaconid. Agerinia has strong P, metaconid and P; = P, in height. Anchomomys is smaller, has M‘~* postflexus and oblique protocristid and postcristid. Microadapis has M! metaconule, P,, metaconid. Adapis, Cryptadapis and Leptadapis have weak M'~ postprotocrista and nearly molariform lise Discussion. Weigelt (1933: 123) described the type species in detail but gave no formal generic diagnosis. Simons (1962: 12-14) revised the type species, placing it in Protoadapis, whose upper Table 9 Characters and character states of species in the Microchoerinae. Characters are estimated as primitive (—) or advanced (+) based on outgroup comparison with the rest of the Omomyidae. ‘? indicates lack of information on a character. Na. = Nannopithex; Ps. = Pseudoloris; Ne. = Necrolemur; M. = Microchoerus. 270 J. J. HOOKER Stampian p M. M. Ludian Ps. parvulus edwardsi [-— i] Ps. 7 Ne. i Mari crusafonti zitteli ' 1 jarinesian lM. wardi Ll M. | ¢ see / a spl —= ate sis y ~ Be M.sp.(Grisolles) 4 N 7 1 FZ Auversian . SS y 7 Ne. cf. = ae - er zitteli ~ ~ ~~ eS = / “Ef Na. fitholi Pi. SF = isabenae ~ ei ee & = Lutetian y SS S Na. raabi “ Ypresian Text-figure 25 Phylogenetic model for Microchoerinae, derived from the cladogram in Text-fig. 24 (p. 269) and from stratigraphical occurrences. The numbers refer to the grades described for Microchoerus (p. 257) and also as they affect Necrolemur. Their limits are arbitrary as they exhibit gradual changes, and in some cases do not coincide with recognized species boundaries. In the case of M. sp. from Grisolles this is because the assemblage exhibits a grade intermediate between 1| and 2. In the case of N. antiquus Filhol, the change from N. zitteli Schlosser to N. antiquus reflects other changes; moreover, known specimens of N. antiquus are mainly from the old collections of the Quercy Phosphorites, involving mixed assemblages in which both grades 1 and 2 occur. For abbreviations see Text-fig. 24. dentition was then unknown. Russell et al. (1967: 38) used absence of an M!? (‘M,,’ in Russell et al. must be an error) metaconule as a specific character of ‘P. klatti. Wilson & Szalay (1976) resurrected Europolemur for the type species. Simons (1962: 12), Russell et al. (1967: 38) and Szalay & Delson (1979: 125) all quoted an M,_» hypoconulid as a distinctive feature of E. klatti. The two available specimens with lower teeth lack this cusp, although there is a buccally oblique ridge running down the distal postcris- tid wall. Its absence has been omitted from the generic diagnosis herein because E. collinsonae does have a M,_, hypoconulid. The variable occurrence of a P, metaconid in E. klatti (present in Simons’ (1962: pl. 3, fig. F), absent in Szalay & Delson’s (1979: fig. 57B, D-E) figures) is considered to represent the begin- ning of an important trend in simplifying P,, carried a stage further in E. collinsonae below. With the description of the new species below, Europolemur is no longer monotypic, Gin- gerich’s (1977a: 61) main reason for reuniting E. klatti with Protoadapis, and it is easier to understand the importance of its generic characters. Europolemur collinsonae sp. nov. (PI. 9, figs 3—4; Pl. 10; Table 10) v. 1977b ?Caenopithecus sp.; Hooker: 141. v. 1980 ?Protoadapis sp.; Hooker & Insole: 31, 38. Name. After Dr M. E. Collinson for her help with field work and other aspects of this project. Hotorype. Left M*, M37169. PI. 9, fig. 3. PaRATYPES. Left M? (M35440), right M? (M37170), left P, (M37171), two right M,s with incomplete talonids (M35755, M37173), right M, (M35756), right M,,/. talonid fragment (M37172), two left Ms, one a trigonid fragment (M35757, M37693). BARTONIAN MAMMALS OF HAMPSHIRE BASIN 271 Plate 9 Light macrographs of upper molars of Europolemur, x 8. Figs 3a and 4a are buccal views, the rest occlusal. Fig. 1 Europolemur klatti Weigelt from Bouxwiller, France. Cast of right M? (reversed) (NMB Bchs648; holotype of synonym, Alsatia dunaifi Tattersall & Schwartz 1984). Fig. 2. Europolemur aff. klatti Weigelt from Egerkingen y, Switzerland. Cast of right M? (reversed) (NMB unnumbered; BM(NH) cast no. M42083). Figs 3,4 Europolemur collinsonae sp. nov. from Creechbarrow. Fig. 3a, b, holotype left M7 (M37169). Fig. 4a, b, right M? (reversed) (M37170). See opposite. J. J. HOOKER Di. ‘OLZ d 208 (LSLSEW) SW YI O-ep ‘SLY (9SLSEW) (Passarod) “IWR ‘O-VE “BI (CSLSEI) (Passoaos) pruoyey ayopduroour ym "YA Ys ‘O-eZ “BLY (TLILEW) "d Yel ‘9-e] “BLY “SMarA jensuly] ‘9 pue yesnjooo “q ‘Teoonq “eg x ‘MOIIeqYde01D Woy ‘Aou ‘ds apuosuljjo2 anwajodoung JO 4199} Y99Y9 Jomo] JO sydeiso1OVU IYSIT OT Bed BARTONIAN MAMMALS OF HAMPSHIRE BASIN 273 HoRIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. D1aGnosis. Large Europolemur (M? length 4-9 mm); M? with hypocone nearly as tall as proto- cone, centrocrista buccally flexed, postprotocrista distally arched and metacone as far from paracone as is protocone; M* with small hypocone; P, with postprotocristid essentially straight, ending just short of distolingual corner, and without paraconid or metaconid; M, _, with entoconid distal to and nearly as tall as hypoconid and with hypoconulid present. DIFFERENTIAL DIAGNOSIS. E. klatti is slightly smaller, M* with smaller hypocone, straight centro- crista, only slightly distally arched postprotocrista and metacone closer to paracone than is protocone; M® lacks hypocone; P, with metaconid or with postprotocristid kinked at mid- point in position of absent metaconid and with very small paraconid; M, , with lower ento- conid, which is transversely opposite hypoconid, and without hypoconulid. DESCRIPTION. Because of the small number of specimens, documentation of variation is restricted to M>, M, and M; (two specimens of each). Of the M?s, M37170 (PI. 9, fig. 4) is broader lingually than M35440, which is triangular in shape, and its hypocone appears larger, although M35440 is more worn. The crown base of the hypocone is drawn out into a point lingually on M35440 over a transversely elongated and mesiodistally short lingual root. M, can be distinguished from M, by having a longer trigonid which is angled instead of bevelled mesiolingually (Pl. 10, figs 2-3). There is no paraconid. On both M;s the buccal cingulum is stronger than on the M,, probably an individual difference. The only complete M, (PI. 10, fig. 4) is considerably worn but shows the presence of an entoconid smaller than the hypoconid. The tooth as a whole is narrower than the M, and is only very slightly larger than a topotype EF. klatti M, (GH XXII/7-1961) (see Szalay & Delson 1979: 124, fig. 57B, D-E). PuyLoGeny. A right M* (NMB unnumbered, cast in BM(NH) no. M42083) of Europolemur from Egerkingen y is intermediate in morphology between E. klatti from Geiseltal and Bouxwil- ler and E. collinsonae from Creechbarrow. It is the same size as E. klatti but has a hypocone nearly as large as in E. collinsonae; its postprotocrista is slightly more distally arched than in E. klatti; the buccal flexing of its centrocrista is intermediate between the two species; and in the remaining characters it is similar to E. klatti (see Pl. 9, figs 1-3). It is thus best referred to as E. aff. klatti, but if more material should be found it may be shown to belong to a new species. It is also intermediate in time between E. klatti and E. collinsonae and it is\likely that the three represent an evolving lineage. Some of the trends are similar to those which produced the Table 10 Length (Il) and mesial (w,) and distal (w,) width measurements of Europolemur collinsonae from Creechbarrow. Two width measurements are only given for lower molariform teeth. Measure- ments in millimetres. No. Tooth | Ww, W> M37160 LM? 49 6:2 M35440 LM? 4:2 (6-7) M37170 RM? 4-1 6-1 M37171 LP, 4-2 DS M35755 RM, = (2:9) - M37173 RM, = (2:8) = M35756 RM, 5-1 35 3-7 M37172 RM, - - 3-9 M35757 LM, 5-6 3-0 PUSS) M37693 LM, = 2°8 = 274 J. J. HOOKER Lutetian Caenopithecus, with which E. collinsonae was first confused (Hooker 1977b). In parti- cular, the buccal flexing of the centrocrista, unlike that of Caenopithecus, did not go as far as development of a mesostyle, nor did M3 expand, nor lower molar metastylids form. Instead the hypocone grew larger and the reduction of P, was different in aspect (Pl. 10, fig. 1), the metaconid being lost instead of being shifted lingually, and the protoconid remaining high. Both Europolemur and Caenopithecus probably share a common ancestry in Protoadapis (see Gingerich 1977a: 77, fig. 8). Genus LEPTADAPIS Gervais 1876 [incl. Paradapis Tattersall & Schwartz 1984 and ?Arisella Crusafont-Pairo 1967] TYPE SPECIES. Adapis magnus Filhol 1874. Phosphorites du Quercy, Raynal, France. INCLUDED SPECIES. L. ruetimeyeri Stehlin 1912, L. priscus (Stehlin 1916) Szalay & Delson 1979, L. stintoni (Gingerich 1977a) Schwartz & Tattersall 1985, L? capellae (Crusafont-Pairo 1967) ?Szalay & Delson 1979. RANGE. Bartonian—Stampian, England; ?Bartonian—late Ludian, France; late Lutetian— Auversian, Switzerland; ?Lutetian, Spain; Ludian, W. Germany. EMENDED DIAGNOSIS. Medium to large adapine; dental formula 4+44, canines large and pointed, uppers with apically converging ridges and grooves, lowers slightly procumbent; inci- sors and P+ reduced (P+ half the linear dimensions of P$); P{ semimolariform; upper molar postprotocrista weak and without metaconule; M, , hypoconulid median and postcristid slightly oblique; M, normally without entoconid; facial region of cranium nearly as high as long; mandibular ramus of approximately equal depth front to back. DIFFERENTIAL DIAGNOSIS. Microadapis differs in having large hypocone and paraconule and metaconule on M! and large entoconid on M,. Cryptadapis has slightly narrower lower cheek teeth and upper molars with slightly larger hypocone. Adapis has smaller canines, the uppers ungrooved and buccolingually compressed, the lowers with a mesially sloping crested tip; incisors and P+ not reduced (P+ nearly as large as P3); M,_» hypoconulid close to entoconid and postcristid strongly oblique; M, often with small sharp entoconid; facial region of cranium much lower than long; mandibular ramus deepening posteriorly. All other adapines (where relevant parts are known) have strong upper molar postprotocrista and non-molariform P#. Discussion. Leptadapis has been used much less frequently as a genus for the species L. magnus than has Adapis, probably because it is evidently closely related and the number of species involved is small. Sometimes Leptadapis has been used as a subgenus of Adapis (e.g. Stehlin 1912; Gingerich 1975b). More recently Szalay (1974: 131-132) has supported separation of the two as genera, with a brief functional analysis of the slightly different anterior dentitions. Gingerich (1977a: 71-72, figs 2, 5) described Adapis stintoni as both a morphological and stratigraphical intermediate between A. magnus and A. parisiensis, thus providing reasons for synonymizing Leptadapis with Adapis. Gingerich (1977b, 1981) reiterated his view of the A. magnus — A. stintoni— A. parisiensis lineage but also noted (1977b: 171) a personal communi- cation from Sudre that two sizes of Adapis occurred in the French locality of Pont d’Assou (Tarn), which if confirmed would thus contradict his evolutionary scheme. An examination of the holotype (M32135) has shown that ‘A.’ stintoni resembles referred L. magnus specimens in that its mandible is uniformly shallow from front to back, and in its large lower canine and small P, (judged from alveoli), almost transverse lower molar postcristids and mesial P, metaconid. It is thus here referred to Leptadapis. It should not be synonymized with A. parisiensis as Szalay & Delson (1979: 139) have done. (Note that while this paper was in press, Schwartz & Tattersall (1985: 92) also recombined ‘A.’ stintoni with Leptadapis.) The realization that ‘A.’ stintoni is in reality a Leptadapis led me to check further lines of evidence bearing on the contrasting classifications and phylogenies. Two sizes of Leptadapis BARTONIAN MAMMALS OF HAMPSHIRE BASIN 275 5 6 7 8 | Text-figure 26 Scatter diagram of length (1) against width (w) of upper molars of species of Leptadapis from various European localities: <= Creechbarrow; A = Lower Headon Beds, Hordle Cliff; Y = Lignite Bed, Headon Hill Limestone, Headon Hill, the smaller plot belonging to the holotype of L. stintoni (Gingerich); x = Lacey’s Farm Quarry; + = Lower Hamstead Beds, Bouldnor Cliff; LJ = Ehrenstein 1A; © = holotype of L. magnus (Filhol) from the Quercy Phosphorites; O =a syntype of L. ruetimeyeri Stehlin (NMB Ef417) from Egerkingen «. Large rectangles span length and width measurements obtained by Gingerich (1977b) from Euzet. Symbols solid on left = M', on right = M? and completely solid = M?; outline symbols = M'/?. Data for Lacey’s Farm Quarry from Insole (1972), for Ehrenstein 1A from Schmidt-Kittler (1971b) and for the Phosphorites from a cast. Measurements in millimetres. Lines join teeth of one individual. appear to occur together at three further localities: 1, Ehrenstein 1A (see Text-fig. 26 and Schmidt-Kittler’s (1971b) upper molar measurements); 2, the lignite bed in the Headon Hill Limestone (the type locality of L. stintoni), where a large M* (M20204) has been found (see Text-fig. 26); and 3, the Lower Headon Beds, where L. magnus occurs at Hordle Cliff and a much smaller incomplete M'/* from HH2 at Headon Hill. Insole (1972: 204205, pl. 7, fig. 2) recorded as M!’?, small enough to belong to L. stintoni, from the Osborne Beds of Lacey’s Farm Quarry, Isle of Wight, as A. magnus (see Text-fig. 26). An even smaller M‘/* has been found in the post-Grande Coupure Lower Hamstead Beds of Bouldnor Cliff (R.L.E. Ford private collection). Acquisition of a lower molar metastylid (Text-fig. 43) appears to have stratigraphical impor- tance (see below). Its variable occurrence in Leptadapis specimens from the Quercy Phosphor- ites housed in the BM(NH) seemed to have the potential for providing further evidence. Text-fig. 27B shows a scatter diagram of lower molars from this collection separated on 276 J. J. HOOKER B 6 7 8 9 10 1 Text-figure 27 A, scatter diagram of length (l) against width (w) of lower molars of species of Leptadapis from various European localities: i= Creechbarrow; A = Lower Headon Beds, Hordle Cliff; V = holotype of L. stintoni (Gingerich) from the Lignite Bed, Headon Hill Limestone, Headon Hill; (1) = Ehrenstein 1A; © =a syntype of L. ruetimeyeri Stehlin (NMB Ef401) from Egerkingen «. Large rectangles span length and width measurements obtained by Gingerich (1977b) from Euzet. Symbols solid on left = M,, on right = M, and completely solid = M;; outline symbols = M,,,. Data for Ehrenstein 1A from Schmidt-Kittler (19715). B, Scatter diagram of length (1) against width (w) of lower molars of species of Leptadapis from the Quercy Phosphorites in the BM(NH): (M37185); left M?* (M37186) (possibly the same individual as M37185); left M, (M37187); and a fragment of left M,,. (M35762). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. EMENDED DIAGNOSIS. Medium-sized Plesiarctomys, length of M, = 5:-38mm. Upper and lower molars with moderately spaced buccal and lingual cusp tips, with shallow intervening basins with a few coarse lophules and no enamel wrinkling. Height of talonid basin above crown base in M, about half the length of the tooth. M,_, hypolophulid nearly continuous. Lower molar sinusid narrow and ectolophid complete. M' ? endoloph with very shallow to absent lingual groove separating protocone from hypocone. Anterior margin of pterygoid fossa of mandible vertical. Upper incisors nearly as long (mesiodistally) as wide (buccolingually), with one or two faint sulci. REASONS FOR EMENDATIONS. Characters here removed from Wood’s (1970) diagnosis are those which I consider are subject to individual variation and had originally been included because no more than one specimen of any cheek tooth type was then known. 1. The apparent absence of two hypoconulids from the Creechbarrow M, and the variation of this character in P. gervaisii (Wood 1970: 253, figs 4A—C) make it unsuitable as a specific character. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 289 2. The presence of only three roots (one mesial, two distal) instead of four in the Creechbar- row M, and some variation in their number in P. spectabilis (Wood 1970: 265) requires emendation of this character. For P. gervaisii, Wood (1970: 249) stated ‘lower molars with three or, rarely, four roots apiece’, but (1970: 254) ‘All of the lower molars preserved have three roots’. In view of the other close similarities of the Creechbarrow M, to the holotype of P. hurzeleri, it seems best to consider these as conspecific and the root number as slightly variable. 3. Variation in size of M' hypocone in P. gervaisii (Wood 1970: 251, figs 2A, C, D) suggests that a similar degree of difference between the Creechbarrow and Robiac upper preultimate molars is not of specific value. 4. Other modifications are to make the diagnosis compare feature for feature with P. cur- ranti. DIFFERENTIAL DIAGNOSIS. P. savagei is smaller and has an incomplete M,, ectolophid. P. spectabilis and P. hartenbergeri are smaller and higher-crowned; their unworn buccal and lingual cusp tips are closer together; their enamel is coarsely wrinkled, often in a radiating pattern; a lingual groove separating M’~* protocone and hypocone is present; they have one mesial and one distal lower molar root; their incisors lack sulci; and P. hartenbergeri incisors are also relatively shorter. P. gervaisii is slightly larger and has two mesial and one distal lower molar roots. P. curranti is smaller and lower-crowned; its unworn buccal and lingual cusp tips are further apart; its enamel is finely wrinkled; its lower molar sinusid is broader. DescriPTION. P*: This tooth is badly corroded but is important as being the first upper premolar to be attributed to this species (PI. 14, fig. 1). This is based on size and the steeper slope of the lingual wall compared to P. spectabilis. From Wood’s (1970) figures, the outline of P. spectabilis P* varies considerably and M35761 does not differ significantly except that the parastyle is more prominent, thus increasing the length/width ratio. A small hypocone is weakly differentiated from the protocone on the postprotocingulum. The paraconule is buccal to the midline and the paracone and metacone appear small and joined by a strong buccal cingulum with no sign of a mesostyle or mesoloph. However, bad preservation makes clear recognition of cusps and crests difficult. The tooth is less square and less molarized than LM 2936 from Mormont described under P. curranti (p. 287). M2: Almost square in outline (PI. 14, fig. 2) in contrast to the triangular right M‘/? (FSL 4916) from Robiac (Wood 1970: 256, fig. 6A), but very similar to the right M'/? from Mormont (LGM 40463: LM 2935) (figured by Major, 1873: pl. 3, fig. 7 as Sciuroides sp.). In both the Creechbarrow and Mormont specimens, the hypocone is larger than in FSL 4916 and there is a strong mesoloph stemming from a weak mesostyle. Unfortunately both teeth are considerably worn, obscuring most of the detail. M?: The suggestion that M37185 and M37186 belong to the same individual is based on near identity of wear and morphology of the M°s and matching interstitial facets of right M? and M?, supported by similar wear and preservation (PI. 14, fig. 2). All three teeth came from the same excavation hole. The outline of the two M*s from Creechbarrow tapers more rapidly than in the right M? (FSL 4917) from Robiac, which is trapezoidal, having a lingually more salient hypocone. There is some evidence of a mesoloph, lacking on FSL 4917. That specimen has a gap between the mesostyle and metacone, whereas the Creechbarrow M?s have a contin- uous buccal cingulum. Other potential features have been removed by wear. M, : M37187 is identified as an M, rather than an M, because the outline has a slight mesial taper (PI. 14, fig. 3). It otherwise differs from the M, of the holotype (Wood 1970: 255, fig. 5A) in having a wider sinusid; no obvious differentiation of the posterolophid into two hypo- conulids; the hypolophulid joining the mesoconid instead of the hypoconid; a stronger ecto- lophid; and a single mesial root pinched in the middle. More specimens of lower molars are required before the exact degree of variation in root number can be assessed. The left M,,. (M35762) is broken and corroded but shows a strong ectolophid and the hypolophulid joining the mesoconid as in M37187. 290 J. J. HOOKER Discussion. P. hurzeleri appears to be very closely related to P. gervaisii from which it differs mainly in slightly smaller size. It could stand as a good morphological ancestor for P. gervaisii, whose type horizon is higher (late Ludian). That P. gervaisii may have already evolved by the earliest Ludian is suggested by the presence at Eclépens of a P, (NMB Mt.1468), mentioned by Wood (1970). This locality contains fissures of both Marinesian (Eclépens-Gare and Eclépens A) and earliest Ludian (Eclépens B) age (see correlation section, pp. 422-429). The P. hurzeleri M!/? (LGM 40463: LM 2935) from Mormont unfortunately lacks more detailed locality data. ?Manitshinae, gen. et sp. indet. (Pl. 13, figs 9-13; Text-fig. 29) v. 1980 Plesiarctomys sp. 3; Hooker & Insole: 39. MATERIAL. Right P* (M35452); broken right M‘/? (M35767); left M? (M35536); right DP, (M35768): right P, (M37189). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION. The specific association of these teeth is by no means certain. They are smaller than any known species of Plesiarctomys except perhaps the unnamed upper premolar from the Lutetian of Montllobar, Spain, described by Wood (1970: 271). Apart from size, the upper teeth share prominent isolated or near isolated paraconules and metaconules and the lack of the typically inflated robust appearance of Plesiarctomys teeth. Their morphology is approached only slightly by the low-crowned, basined teeth of Plesiarctomys curranti. The anteroloph, endoloph and posteroloph form a single three quarters encircling loph on which protocone and hypocone can hardly be individually recognized. P*: This tooth lacks the buccal wall and is triangular in outline (Pl. 13, fig. 9). Large paraconule and metaconule arise from narrow preprotocrista and postprotocrista respectively. The metaconule is more isolated than the paraconule. The anteroloph broadens buccally towards a parastyle which is mainly broken away. A fissure separates paracone and metacone. The posteroloph bulges slightly distolingually but there is no sign of a hypocone. M!'/*: A large buccal part of this tooth is missing (PI. 13, fig. 10). It is considered to be a right because the more isolated of the two intermediate conules by comparison with P* appears to be the metaconule; and also because the more prominent of the two lingual cusps is probably the protocone. M32: This well-preserved tooth, unlike the previous two, is scarcely worn (PI. 13, fig. 11). Its outline is a right-angled triangle with the distolingual edge as the hypotenuse and the mesial edge the shortest side. The metacone is only slightly smaller than the paracone and the meta- conule is larger than the paraconule. There is no hypocone and the prominent posteroloph bulges distally and meets the metacone buccally. There is a prominent mesostyle terminating a strong mesoloph which extends one third the distance across the trigon basin. Distal to the main mesostyle is a fissure and then a small accessory mesostyle joined to the metacone. The enamel is finely wrinkled and two slightly coarser folds occur buccal and lingual of the para- conule. Protocristae are lacking except for a slight development between the metaconule and protocone; and the paraconule and metaconule would have remained distinct even after heavy wear. The anteroloph is weak and forms only a narrow shelf in front of the paracone. DP,: This tooth is corroded and enamel is missing from its buccal wall, but it shows finely wrinkled enamel in the talonid basin (PI. 13, fig. 12). It is identified as a DP, by analogy with the high length/width ratio of this tooth compared to P, in Plesiarctomys. There appears to be a nearly complete hypolophulid joining the hypoconid as can occur in Plesiarctomys (see Wood 1970: fig. 12K). The large mesial cusp probably combines both protoconid and metaconid and there is no cingulum mesial to this. P,: This tooth is also corroded and distobuccally broken (PI. 13, fig. 13). It has a distinct protoconid which is slightly smaller than the metaconid, as in Plesiarctomys curranti. There is a deep talonid notch, no hypolophulid and the talonid basin enamel is finely wrinkled. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 291 Discussion. This small assemblage probably belongs to an undescribed genus. Despite some similarities, it does not fit the diagnosis of Plesiarctomys. There are some striking similarities to the ‘genre indétermine B’ from the early Eocene of Avenay and Condé-en-Brie, France, of Michaux (1968: pl. 10, figs 7-9). The large isolated intermediate conules, strong encircling lingual loph with indistinct cusps, and finely wrinkled, deeply basined trigon are features in common, suggesting that the two taxa may be congeneric. That they are distinct at species level is indicated by the absence of a hypocone on the Avenay upper preultimate molar. Michaux (1968) did not place his ‘genre indéterminé B’ in any paramyid subfamily. The suggestion that the present form might be a manitshine is based on its vague similarity to Plesiarctomys. A rodent lower right incisor from Creechbarrow (M37190) (Text-fig. 30F—I) fits none of the other taxa described here, unless it is a juvenile Ailuravus. It is the correct size for the present form, and has some unusual features. It is an almost symmetricaily compressed D-shape in cross section and has a buccolingually elongated pulp cavity, not triangular or dagger-shaped as in Plesiarctomys. There is a single distobuccal carina. The enamel is as thin as in the incisors of Ailuravus. The mesial enamel band extends one fifth of the width (buccolingual) of the tooth. Distally the enamel extends more than half way round the tooth. From figures in Wood (1962) this degree of enamel extension occurs in Ischyrotomus compressidens (Peterson 1919), Reithro- paramys debequensis Wood 1962, Leptotomus sciuroides (Scott & Osborn 1890), L. mytonensis Wood 1962, L. bridgerensis Wood 1962, L. parvus Wood 1959, L. huerfanensis Wood 1962, L. costilloi Wood 1962, L. grandis Wood 1962, L. leptodus (Cope 1883), Paramys delicatior Leidy 1873 and P. delicatus Leidy 1873 amongst the North American paramyids and Ailuravus spp. amongst the European ones. It does not occur, however, in Paramys francesi, the species that Michaux (1968: 175) considered closest to his ‘genre indéterminé B’. This could be a primitive character of no importance in isolation, or M37190 may not belong to the same species as the other teeth assigned to the present form. Subfamily AILURAVINAE Michaux 1968 DIAGNOsIS. See Wood, 1976a: 122. Genus AILURAVUS Ritimeyer 1891 [incl. Palaeomarmota Haupt 1921, Megachiromyoides Weigelt 1933, Aeluravus and Maurimontia Stehlin & Schaub 1951] TYPE SPECIES. Ailuravus picteti Ritimeyer 1891. Late Lutetian fissure fillings, Egerkingen, Canton Solothurn, Switzerland. INCLUDED SPECIES. A. michauxi Hartenberger 1975, A. macrurus Weitzel 1949, A. stehlinschaubi Wood 1976a. RANGE. Late Ypresian to Bartonian; Britain, France, East and West Germany and Switzerland. EMENDED DIAGNOSIS. See Wood, 1976a: 123-124. REMARKS. Following Wood (1976a: 145) it is considered that the two species described but not named by Michaux (1968: 159-161) fall within the variation of a single species. They were later named by Hartenberger (1975: 780-781) but his diagnoses do not allow them to be distin- guished and no measurements were given. Therefore, A. remensis Hartenberger 1975 (p. 781) is herein synonymized with A. michauxi Hartenberger 1975 (p. 780) on page priority. The four species listed above are not easy to distinguish because of the small number of specimens of each species and the large amount of individual variation. Even though 4A. macrurus is known from complete skeletons with fur preserved, the teeth are usually in occlusion and cannot easily be seen. Creechbarrow is only the second known locality for A. stehlinschaubi and considerably increases the number of teeth known. DP?, P*, DP,, I’ and I, were previously unknown. 292 J. J. HOOKER BARTONIAN MAMMALS OF HAMPSHIRE BASIN 293 Ailuravus stehlinschaubi Wood 1976a (Pl. 14, figs 4-13; Text-fig. 30J-K; Table 12) v. 1869 Hyracotherium Owen?; ?Pictet & Humbert: pl. 25, fig. 5. vy. 1951 Maurimontia picteti Stehlin & Schaub: 20, fig. 18; 206, fig. 310. vy. 1962 Maurimontia picteti Stehlin & Schaub; Wood: 239, fig. 88G—H. v* 1976a Ailuravus stehlinschaubi Wood: 141-145, fig. 6. y. 1977b Ailuravus stehlinschaubi Wood; Hooker: 141-142. y. 1980 Ailuravus stehlinschaubi Wood; Hooker & Insole: 39. Hotorype. Right maxilla with DP*, P*-M? (LGM 39559: LM 2906). Bartonian fissure filling, Eclépens-Gare, Canton Vaud, Switzerland. PaRATyPEs. Right mandibular ramus with chipped P,-M, (LGM 39561: LM 2910), and left M! (NMB Mt1767). Occurrence as holotype. MATERIAL. DP? (M35453); P? (M35769); DP* (M35771); 3 P*s (M35770, M37191-2); 6 M!/2s (M35454-6, M35772, M37193, M37697); 2 DP4s (M35774-5); M,). (M35457); 8 trigonid and talonid fragments of P4/M,/. (M35458—-9, M35776-8, M35780, M37194, M37698); M3 trigonid (M35779) and talonid (M35773) fragments; 12 undetermined fragments of cheek teeth (M35460—3, M35784, M37195); I’ (M35698); and 6 I,s (M35781-3, M37196-8). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. RANGE. Recorded only from the Bartonian of Eclepens-Gare, Switzerland; and Creechbarrow, England. EMENDED DIAGNOSIS. (Slightly modified from Wood, 1976a: 142). Large species of Ailuravus, length of M‘ = 4-50-5-14mm (see also measurements in Table 12 and Wood, 1976a: 127); cusps (especially conules) of upper cheek teeth tend to be elongate mesiodistally; valleys between cusps deep; enamel wrinkling moderate to dense; single paraconule; metaconule developing a mesio-distal elongation or doubling; protostyle weak to missing; mesostyle very prominent and buccally extended, especially on P*; M!'~ with hypocone consistently only slightly smaller than protocone; M? without hypocone; mesoconid of lower teeth very large, triangular, and cut off from buccal margin of tooth; posterolophid of molars almost non- existent except for the hypertrophied hypoconulid; incisors almost circular in cross section and small (I, mesiodistal dimension 40-52% of M'/? length); mental foramen more than the length of P, in front of P,. DIFFERENTIAL DIAGNOSIS. A. michauxi is smaller with cone-shaped intermediate conules on upper cheek teeth; valleys shallow with much less enamel wrinkling; paraconule may be double; protostyle prominent; mesostyle and hypocone smaller; molar posterolophid more continuous; [incisors and jaw unknown]. A. macrurus has cone-shaped intermediate conules; valleys shallow with much less enamel wrinkling; paraconule double; protostyle prominent; mesostyle smaller; hypocone may be smaller; molar posterolophid more continuous; incisors more buccally tapered and slightly larger; mental foramen more posterior. A. picteti has less elongate intermediate conules; valleys may be shallower and may have less enamel wrinkling; paraconule double; protostyle may be prominent; mesostyle smaller; hypo- Plate 14 Light macrographs of occlusal views of cheek teeth of paramyids from Creechbarrow, x 8. Figs 1-3 Plesiarctomys hurzeleri Wood. Fig. 1, left P* (M35761). Fig. 2, right M7? * (reversed) (M37185). Fig. 3, left M, (M37187). See p. 288. Figs 4-13 Ailuravus stehlinschaubi Wood. Fig. 4, right DP? (reversed) (M35453). Fig. 5, left DP* (M35771). Fig. 6, left P? (M35769). Fig. 7, left P* (M37191). Fig. 8, right M'/? (reversed) (M37193). Fig. 9, left DP, (M35775). Fig. 10, left P,? trigonid fragment (M35778). Fig. 11, right M,,2 (reversed) (M35457). Fig. 12, left M, trigonid fragment (M35779). Fig. 13, left M, talonid fragment (M35773). See above. 294 J. J. HOOKER Table 12 Length (1) and mesial (w,) and distal (w,) width measurements of Ailuravus stehlinschaubi from Creechbar- row. Only one width measurement is given for incisors and upper cheek teeth. Measurements in millimetres. No. Tooth l Wi W> M35453 DP? 2:30 1-85 M35769 p2 2:70 2:39 M35770 Ip (5-20) - M37191 p* (5-70) 6:30 M37192 Pe 5:35 - M35454 M!? 4-50 ~ M35772 M'? 4-75 55 M37193 M'? 4-95 5:60 M37697 M?? 4-75 5:25 M35774 DP, (4-50) 3-10 (3-60) M35775 DP, 4-20 (2:60) (3-25) M35778 Pry ~ 3-45 = M35780 1nd - 3-45 - M37698 P,? ~ — 375 M35457 M,/2 4-90 3-80 4-35 M35458 Mi; - 3-45 - M37194 M,) - 3-65 = M35459 M,)2 - = 4-30 M35776 M,;2 - - 4-15 M35779 M, - 4-00 - M35773 M, — ~ p55 M35698 lis 2-60 2-40 M35781 I, 2:05 2-60 M35782 I, 2-15 (2:90) M35783 I, 2:05 (2:90) M37196 I, 2:00 2:35 M37197 I, 2-30 2:85 M37198 I, 2:35 2-95 cone may be smaller; mesoconid usually joined to hypoconid by buccal spur as well as by ectolophid; incisors slightly larger (inferred from alveoli); mental foramen more posterior. DESCRIPTION. Wood (1976a) has described this species in detail, so only new features or varia- tions will be dealt with here. Most of the tooth measurements of the Creechbarrow material are slightly less than their counterparts in the type series (see Table 12 and Wood, 1976a: 126-127, tabs 1—2). Wrinkling of the enamel is very variable in extent and detail. It may be relatively weak (PI. 14, fig. 7) or very strong (PI. 14, fig. 8). It usually has a narrow core of dentine. Because of the variation, there is overlap with A. picteti, but here the degree of wrinkling is usually less. DP? and P® (PI. 14, figs 4, 6): Wood (1976a: 143-144) suggested that these teeth were probably absent in the holotype, referring to the absence of alveoli in Stehlin & Schaub’s (1951) fig. 18, drawn before jaw breakage caused by DP* removal. However, Pictet & Humbert’s (1869: pl. 25, fig. 5b) buccal view shows the maxilla to have been truncated immediately anterior to DP*. The question can never be settled for the holotype, but M35453 and M35769 show the morphologies of DP? and P? respectively and indicate that both these teeth could occur in A. stehlinschaubi. M35453 is the smaller and has much thinner enamel than M35769 or any of the P*s or molars, which is typical for a milk tooth. It has a prominent parastyle, distally placed ?paracone with two distal crests, a minute ?protocone at the distolingual corner and a ?preparacrista with a small cuspule at about its midpoint. M35769 is narrower distally, has a central paracone with a buccally flexed postparacrista which joins a distal cingulum bearing a BARTONIAN MAMMALS OF HAMPSHIRE BASIN 295 small lingual cuspule; a curved crest runs down the lingual side of the paracone; there is neither parastyle nor protocone. DP* (PI. 14, fig. 5): This tooth is broken mesiobuccally but is essentially similar to that of the holotype DP*. P+: M37191 is lacking most of the parastyle but from the spacing of its main cusps seems to be relatively broader than the holotype P*, has slightly more wrinkled enamel and a small hypocone (PI. 14, fig. 7). M35770 has the protocone broken away and is much more densely wrinkled than either M37191 or the holotype. M37192 is heavily, naturally worn and shows a similar degree of wrinkling to M37191. M"/?: M! and M? of the holotype can be distinguished on length/width ratio. Among the isolated Creechbarrow upper preultimate molars, the shorter broader ones are tentatively identified as M' and the single relatively longer one as M2. There is high variation in enamel wrinkling, buccal cingulum strength and mesostyle size. There are also differences in the distal elongation or doubling of the metaconule: in M35772 and M37697 there is a second meta- conule developed distal to the first; in M35454 and M35455, the first metaconule has a strong distal crest extending to the distal cingulum; and in M37193, the whole area distal to the first metaconule is covered with dense enamel wrinkling (PI. 14, fig. 8). DP, (Pl. 14, fig. 9): Two submolariform teeth (M35774 and M35775) have thinner enamel than any of the molars and so are identified as DP,, a tooth not previously recorded for the genus Ailuravus. They show variation in degree of enamel wrinkling, as do the molars. The principal differences from the holotype P, are presence of a smaller, rounder, less triangular mesoconid, large isolated lingually situated hypoconulid, an extensive cingulum-bordered sinusid and lower crown height. Lower Permanent Cheek Teeth: Only the M,,., M35475, is complete (Pl. 14, fig. 11). All the others are either trigonid or talonid fragments. The M, talonid (M35773) shows, better than the poorly preserved paratype, the crescentically crested hypoconid and the almost total isolation of this cusp, the hypoconulid, entoconid and mesoconid (PI. 14, fig. 13). The trigonid (M35779) is identified as M, because of the distal position of the protoconid relative to the metaconid and the apparent depth of the ectoflexid (PI. 14, fig. 12). The remaining talonid fragments add little to our existing knowledge of the species. Of the trigonids, those with the more convergent cusps are tentatively identified as P, (e.g. Pl. 14, fig. 10) from comparison with the admittedly imperfect paratype premolar and molars. Incisors: The finding of seven Ailuravus incisors at Creechbarrow is in marked contrast to the normal rarity of these teeth (see Wood, 1976a: 125). The cross-sectional shape of the lowers is similar to that of A. macrurus figured by Wood (1976a: fig. SG), but rounder (Text-fig. 30K). The single upper is also much rounder than Wood’s (1976a: fig. 5D) figured I’ of A. macrurus (see Text-fig. 30J). Discussion. Although not recorded from other localities, it is possible that the record of Ailuravinae gen. indet., which is an incisor (J.-L. Hartenberger, personal communication 1978), from Robiac (Sudre 1969a: tab. opposite p. 142) might be shown to belong to this species if cheek teeth can be found. Wood (1976a: 145) considered A. stehlinschaubi to have descended from Lutetian A. picteti. The Ailuravus sp. indet. from the Auversian Sables a Batillaria bouei (Sables d’Auvers) of Arcis-le-Ponsart, Marne, France, may provide an intermediate between the two (Louis 1976: 49). Thus A. stehlinschaubi is a good age marker for the European Bartonian, if not more strictly the Marinesian. The genus appears to have become extinct at the end of the Bartonian. Infraorder uncertain Superfamily THERIDOMYOIDEA Alston 1876 This superfamily has been classifed in various ways infraordinally in the Rodentia (e.g. Schaub 1958, Romer 1966, Simpson 1945), but it is probably derived from subfamily Paramyinae of the 296 J. J. HOOKER protolophule | metalophule |! metalophule i} sinusid endoloph ectolophid Text-figure 31 Dental nomenclature of pseudosciurid cheek teeth. A, left upper molar. Buccal edge towards top of page, lingual towards bottom. B, left lower molar. Buccal edge towards bottom of page, lingual towards top. Abbreviations: A. antph—anteroloph B. antd—anteroconid hyp—hypocone antphd—anterolophid hypphle—hypolophule antphld—anterolophulid hypst—hypostyle ectstd—ectostylid mes—mesocone entd—entoconid mesph—mesoloph (partial) hypd—hypoconid mest—mesostyle hypld—hypoconulid met—metacone hypphld—hypolophulid metle 1—metaconule 1 mesd—mesoconid metle 2—metaconule 2 mesphd—mesolophid (partial) par—paracone mestd—mesostylid parle—paraconule metd—metaconid past—parastyle metphld—metalophulid postph—posteroloph postphd—posterolophid prost—protostyle (= anterocone) protd—protoconid prot—protocone Paramyidae (see Hartenberger 1969: 56). No evidence is forthcoming on these questions from the English Bartonian material. Suffice to say that theridomyoids are hystricomorphous and sciurognathous and have cheek teeth which range from low-crowned to moderately high- crowned but always retain the roots. The complexities of convergence in hystricomorphy are discussed by Wood & Patterson (1970). Their dental formula is +9+43 but sometimes DP# is retained and not replaced by P#. They have usually been divided into two families: the Pseudosciuridae Zittel 1893 for the lower-crowned and Theridomyidae Alston 1876 for the higher-crowned types. However, Har- tenberger (1971) produced a model of parallel evolution of higher-crowned forms (which would have been classified as theridomyids) from new (Estellomys) and hypothetical lower-crowned forms (which would have been classified as pseudosciurids or as transitions between the two families). He assimilated the two families under the name Theridomyidae which he divided into six subfamilies, all considered to be monophyletic. These were later increased to seven (Hartenberger 1973) and comprised: Pseudosciurinae Zittel 1893, Oltinomyinae Hartenberger 1971, Sciuroidinae Hartenberger 1971, Columbomyinae Thaler 1966, Theridomyinae Alston 1876, Issiodoromyinae Lavocat 1951 and Remyinae Hartenberger 1973, with the genus Quer- cimys Thaler 1966 of uncertain family status. Tarnomys was described later (Hartenberger & Schmidt-Kittler 1976), but although considered related to Pseudosciurus, it was classified at no lower level than superfamily Theridomyoidea. The Pseudosciurinae and Sciuroidinae contained all the genera which had been included in the Pseudosciuridae before 1973. Bosma (1974) accepted the traditional separation of the Pseudosciuridae and Theridomyidae and considered that Hartenberger’s transition forms should be placed arbitrarily in one or the other family. She BARTONIAN MAMMALS OF HAMPSHIRE BASIN 297 did not recognize the subfamily division of the Pseudosciuridae (presumably meaning the Pseudosciurinae and Sciuroidinae) but gave no further explanation. This division into two subfamilies is untenable for the following reasons. Hartenberger (1973) described a new species of Suevosciurus (S. romani) in the Pseudosciurinae which in fact is identical with Sciuroides siderolithicus, type species of the type genus of the Sciuroidinae. Therefore the type species of Sciuroides was included in both the Pseudosciurinae and Sci- uroidinae. Apart from S. siderolithicus, which was not treated systematically, all the other Sciuroides species which he put in the Sciuroidinae are, according to Bosma (1974), referrable to Paradelomys; see details below. Only one of Hartenberger’s diagnostic features for the two subfamilies allows comparison: the size of the anterior palatal foramina, and here there is much overlap. From Hartenberger’s (1973) text-fig. 2, only one lineage in his Theridomyidae is shown to have evolved from truly brachyodont to hypsodont and thus bridged his ‘Pseudosciuridae’/ ‘Theridomyidae’ grade boundary. This is the Estellomys—Oltinomys lineage. The principle of evolution of high-crowned from low-crowned forms is not in doubt here. It is a common and well documented trend among fossil mammals. Even if this happened in parallel, however, no forms originally placed in the Pseudosciuridae have necessarily given rise to any originally placed in the Theridomyidae. Problems in classification always occur when the stem of two suprageneric groups such as these is approached. It is not necessarily solved, however, by lumping the two groups. In post-Bartonian time there are two well-defined groups which differ considerably in crown height and must have occupied very different ecological niches. It seems desirable to recognize these as different families and seek more reliable charac- ters to define each. Three combined features which characterize all the higher-crowned forms plus Estellomys are: 1, a mesoloph which tends to join the hypocone; 2, absence of a metalop- hule I; and 3, an oblique ectolophid without mesoconid. Conversely the Pseudosciuridae have lesser or greater tendencies towards: 1, mesoloph not joined to hypocone; 2, presence of metalophule I independent of the mesoloph; and 3, presence of mesoconid. The characters outlined above support the retention of these two families, but pose problems for Paradelomys andTarnomys which are somewhat intermediate. All the species except typical P. crusafonti have a metalophule I, although in the late forms (P. spelaeus and T. quercinus) it sometimes partially fuses with the mesoloph (see Schmidt-Kittler 1971a: fig. 27). However, the earliest assemblage known of P. crusafonti (that of Grisolles) has a metalophule I and a short mesoloph. It is evident that the long, but weak, mesoloph which joins the hypocone in assemblages from Robiac times onward is formed by fusion of the short buccal mesoloph to the lingual half of metalophule I, with abortion of the distobuccal half of metalophule I. All the theridomyids may also have undergone a similar modification and the genera here attributed to the Pseudosciuridae may be only those that retained it as a primitive feature. Nevertheless, unlike P. crusafonti, typical theridomyids all show a strong hypocone-joining mesoloph. This is more likely to have been derived from a morphology like that occurring in Lutetian Protadel- omys alsaticus Hartenberger 1969. Some individuals of this species (Hartenberger 1969: pl. 2, figs 4, 7) have a long, strong mesoloph joining the metaconule, which suggests a trend towards the Theridomyidae as construed here. Paradelomys as defined by its type species alone, P. crusafonti, had developed its deep sinus by at least the Bartonian, but still had an oblique ectolophid and no anterolophulid or mesoconid. Three other species that have been referred to Paradelomys (P. depereti, P. quercyi and P. spelaeus) have a metalophule I, non-oblique ectolophid, weak mesoconid, antesinusid and anterolophulid. They might usefully be placed in a new genus. None of the brachyodont theridomyoids that developed metalophule I are known to have given rise to hypsodont taxa, but all that instead developed a strong mesoloph—hypocone link did give rise to hypsodont taxa. This tends to justify the retention of both families as probably monophyletic natural adaptive groups. If it should be found, however, that the characters used to differentiate them were independently evolved, then a modifed system of Hartenberger’s subfamilies might prove to be a suitable solution, or different characters again might be used to differentiate the two families. 298 J. J. HOOKER Family PPEUDOSCIURIDAE Zittel 1893 TYPE GENUS. Pseudosciurus Hensel 1856. INCLUDED GENERA. Protadelomys Hartenberger 1968, Sciuroides Major 1873, Treposciurus Schmidt-Kittler 1970, Suevosciurus Dehm 1937, Paradelomys Thaler 1966 and Tarnomys Har- tenberger & Schmidt-Kittler 1976. RANGE. Lutetian to Stampian; England, Belgium, France, Spain, Switzerland and West Germany. EMENDED DIAGNOSIS. Cheek teeth brachyodont; uppers with metalophule I and/or with rugose to reticulate enamel and only short mesoloph; lowers with mesoconid and/or with bent ecto- lophid. TAXONOMIC TREATMENT. Pseudosciurids are represented at Creechbarrow by single species of Sciuroides, Treposciurus and Suevosciurus. They are abundant as individuals and the Suevo- sciurus comprises more than 20% of the total mammal fauna. Because of the varied generic and specific combinations which have been proposed within this family, the maximum number of available characters has been used in a stratigraphical context to minimize the confusing effects of evolutionary reversals and parallelisms. Often these characters are not present in all individ- uals, infraspecific variation being high. Their percentage occurrences are tabulated in order, to avoid as much as possible use of imprecise words like ‘usually’ and ‘most’, which however still need to be used in the generic diagnoses. Equally important as a taxonomic method in the face of such large morphological variation is the estimation of coefficients of size variation. In any one assemblage low coefficients of variation for tooth size give confident species limits (see Gingerich 1974). DENTAL NOMENCLATURE (Text-fig. 31). Bosma (1974: 16-17) stated that she was following Wood & Wilson (1936) for her nomenclature of the dental pattern. She did not in fact closely follow these authors, possibly because their terminology was designed for only cricetid, heteromyid and sciuravid, not theridomyoid, teeth. Moreover, in Bosma’s (1974) text-fig. 2, of a labelled pseudosciurid right upper molar, some of the cusp names have been transposed as if the tooth were a left. These errors are repeated in Bosma & Insole (1976: 2, text-fig. 1). Hartenberger (1973: 6) stated that he was directly following Wood & Wilson (1936) and Schaub (1958) (the latter presumably only for the higher crowned types), but his system differs somewhat from that of Bosma and more closely resembles that of Wood & Wilson for a cricetid type of tooth. I essentially follow Wood & Wilson’s and Hartenberger’s schemes with the following exceptions: 1, I use paraconule instead of protoconule to maintain uniformity of nomenclature for all mammals with this cusp. The choice is because the paraconule has the same spatial relationship to the paracone, not to the protocone, as the metaconule has to the metacone. 2, I tentatively recognize a mesocone. 3, Bosma (1974) named the loph joining the metacone to the hypocone the metaloph, and the intermediate conule on it the metaconule. The more mesial parallel loph she called metalophule I, from Wood & Wilson (1936). Hartenberger used Wood & Wilson’s term metalophule II for Bosma’s metaloph but retaining her meaning of metaconule. In post- Lutetian pseudosciurids, there are often two distal lophules on upper molariform teeth, which can be homologized or at least compared for identity of position with metalophules I and II of the cricetid type (Wood & Wilson 1936: 389, fig. 1). Both or either may bear or be reduced to a conule. Of these only the more distal one has been named — the metaconule. In Lutetian pseudosciurids and the presumed ancestral paramyids with tribosphenic teeth, there is a single cusp occasionally on a single crest between metacone and protocone. It can easily be homolo- gized with the metaconule. Individual variation in this region in species of Protadelomys as illustrated by Hartenberger (1969) shows incipient development of two lophules and two conules, all derived from or developed around the original metaconule. Therefore metaconule 1 herein is the previously unnamed cusp on or replacing metalophule I, whilst metaconule 2 is the metaconule of Bosma and Hartenberger. 4, I tentatively recognize a hypostyle as the swelling on the posteroloph. 5, the enamel bridge which sometimes links metalophule II to the BARTONIAN MAMMALS OF HAMPSHIRE BASIN 299 posteroloph I name the hypolophule. 6, The crest, which sometimes trends distobuccally from the protocone, I consider can be referred to an incipient protolophule II. 7, I recognize the anterolophulid as used by Mein & Freudenthal (1971 : 3) for cricetid teeth. DEVELOPMENT OF METALOPHULE I. There is variation in the position of the lingual attachment of metalophule I in pseudosciurids. The condition in upper first and second molars is as follows. In Sciuroides it ranges between the lingual half of metalophule II and half way between the middle of the endoloph and the hypocone. In Treposciurus, it ranges between the middle of the endoloph and half way between this point and the hypocone. In Suevosciurus (vestigially), Tarnomys and Paradelomys it is predominantly close to or immediately distal to the middle of the endoloph. In DP*, the lingual attachment in all these genera has the same range as in the M! * of Sciuroides. Pseudosciurus has no metalophule I. Late members of the respective genera tend to have a lingual attachment intermediate between the hypocone and the middle of the endoloph, whereas in early members it is at opposite extremes away from this position. Thus there appears to be a time-linked trend towards the intermediate position as the ‘optimum’; and a slightly different mode of origin of an otherwise homologous structure. It appears therefore that, in M‘! * of Sciuroides and in DP* of all the genera, the metaloph begins to split from the buccal end. It is not until the split has progressed right to the endoloph that the ‘optimal’ position is reached. Some support for this hypothesis is provided by Protadel- omys lugdunensis Hartenberger 1969, from the Auversian of Lissieu and shown by Hartenberger (1969) to be closely related to Sciuroides (under the name of Suevosciurus). P. lugdunensis has an undivided metaconule but two crests link it to the metacone and only one to the hypocone (Hartenberger 1969: pl. 4, figs 2-3). Metalophule I in M!~? of the other genera appears to have developed as a crest along the line of the postprotocrista, as the metaconule split in response to the expansion of the disto- lingual portion of the tooth. That this may have developed fairly rapidly from a morphology closely resembling that of Protadelomys cartieri from the late Lutetian of Egerkingen y (see Hartenberger 1969: pl. 1, fig. 1) is suggested by an unusual individual within the morphologi- cally very variable assemblage of Treposciurus helveticus preecei from Creechbarrow. This M!/? (Pl. 18, fig. 8) has no metalophules but a single massive metaconule in a position which would be intersected by a straight line drawn between the protocone and metacone. It is thus slightly more mesial than the metaconule in P. lugdunensis. Other variants show different grades of development of the metalophules, mainly I, which nearly always either joins, or is very close to, the middle of the endoloph, where a mesocone is often present. Thus the most primitive genus in the Pseudosciuridae (Protadelomys) lacks metalophule I and could be regarded as providing potential ancestors for both theridomyoid families. As discussed above, important characters are shared between P. alsaticus and typical Theridomy- idae and between P. lugdunensis and typical Pseudosciuridae. More specimens from Lutetian— Auversian sites may eventually result in the splitting up of the genus Protadelomys, but it is here provisionally retained undivided within the Pseudosciuridae. NOTE ON ENAMEL WRINKLING. A character of frequent occurrence in the Pseudosciuridae is wrinkling of the enamel. Even at their strongest the wrinkles are not dentine-cored and have often been removed by natural tooth wear. Their pattern may vary, from isolated granules, to a series of ridges, to reticulation. They have been mapped in detail for Pseudosciurus by Schmidt- Kittler (1971a). The wrinkling may be coarse to fine, low or high, extensive or restricted to only some areas of the tooth. Complicated as they may seem these various features are directly related to the intensity of the wrinkling. Slight wrinkling (grade 1 of the character analysis tables 14-15, 17-18, 20-21) produces low, isolated granules; more intense wrinkling (grade 2) produces a system of ridges; and the maximum observable intensity (grade 3) produces a high reticulate pattern which fills the tooth valleys, leaving deep circular cavities between. Fineness and coarseness vary independently of these grades and together with wear make objective comparisons difficult, but they are nevertheless attempted for the three pseudosciurids from Creechbarrow and their relatives. 300 J. J. HOOKER IDENTIFICATION OF ISOLATED PREULTIMATE MOLARS. Isolated first and second molars are difficult to distinguish from one another, but those uppers with near square outline are regarded as M! and those with distally tapering outline as M?; mesially tapering lowers are identified as M, and those which are rectangular as M,. These identifications are only tentative and are attempted for the purposes of the percentage character analysis tables only, where expression of characters in each tooth morphology is important. In the tables and graphs of length and width measurements they are not separated, but nevertheless still produce sufficiently low coefficients of variation. Genus SCIUROIDES Major 1873 TYPE SPECIES. Sciurus siderolithicus Pictet & Humbert 1869. Marinesian fissure filling, Eclépens- Gare, Canton Vaud, Switzerland. INCLUDED SPECIES. S. russelli (Hartenberger & Louis 1976), S. ehrensteinensis Schmidt-Kittler 1971a and S. rissonei sp. nov. RANGE. Marinesian; England, France, Switzerland. Ludian; England, Switzerland and ?France. EMENDED DIAGNOSIS. Upper cheek teeth with: high endoloph; strong, dentine-cored, complete to incomplete metalophule I; strong paraconule; presence of a metaconule 2; DP* with straight mesiolingual margin and unicuspid parastyle. Lower cheek teeth with: high, complete, straight oblique to bent ectolophid joining hypolophulid separately from hypoconid; weak mesoconid. Lower molars with: small to large, often median anteroconid on M,_,; anterolophulid usually present; anterolophid weak; metalophulid low and incomplete. DP, with: crown height lower than that of molars; metalophulid absent. M3 hypolophulid high and complete. Anterior palatine foramina project just posterior to maxillary-premaxillary suture (known only for S. siderolithicus). PREVIOUS DIAGNOSES. Major (1873) gave no formal diagnosis and Schlosser (1884: 76-77) was the first to give an adequate description. Major’s concept of the genus included species now included in other pseudosciurid genera (see Dehm 1937; Schmidt-Kittler 1971a). Schmidt- Kittler (1971a: 29) gave a brief history of treatment of the genus and a comprehensive diag- nosis. Hartenberger (1973: 21) did not discuss Schmidt-Kittler’s diagnosis or concept of Sciuroides, but gave his own short diagnosis: ‘Créte longitudinale joignant protoconide et hypoconide assez développée. Présence presque constante aux molaires supérieures d'un metalophule I. Thus his own new species, Suevosciurus (Treposciurus) romani, fits this diagnosis and is in fact inseparable from the type species of Sciuroides. Hartenberger included in Sciuroides (apart from the type species) ‘S.’ quercyi, ‘S. intermedius’ and ‘S. ruetimeyeri. He incorporated Sciuroides and Paradelomys into his subfamily Sciuroidinae. One of his characters of this subfamily was medium to large anterior palatine foramina. ‘S.’ quercyi and ‘S. intermedius’ (= ‘Paradelomys’ depereti, see Bosma 1974: 54) fit the medium category; those of ‘S.’ ruetimeyeri are unknown; and those of the type species of Sciuroides are short and thus prevent its inclusion in the type subfamily. In contrast, Schmidt-Kittler’s (1971a) concept of Sciuroides produces no anomalies and Sue- vosciurus russelli Hartenberger & Louis 1976 can readily be accommodated in it. Slight emen- dations and additions are made above and involve the following points. 1. Schmidt-Kittler (1971a: 29 and table 4) distinguished Sciuroides from the other pseudosciurid genera by the absence of enamel wrinkling. This was based on the few available specimens of S. siderolithicus and the unique M?” of S. ehrensteinensis. Bosma (1974: 32-33) described further material of the latter from the Headon Beds of the Isle of Wight and noted that the enamel was slightly rugose. In some individuals of all four species there may be minor unnamed crests branching from the major lophs as well as numerous irregularities of the enamel but they tend to be coarser than those which occur in Treposciurus or Suevosciurus. Enamel wrinkling is thus not of generic significance. 2, The weakness of the anterolophid and metalophulid is an important character not previously mentioned. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 301 EXCLUDED SPECIES. 1. Adelomys (Sciuroides) fontensis Thaler 1966, from the early Ludian of Fons 1. Schmidt-Kittler (1971a: 35) thought it was probably synonymous with S. siderolithicus. Hartenberger (1973: 29) doubted that it was even a pseudosciurid. It has the interrupted metalophulid of Sciuroides and is about the size of S. aff. siderolithicus from Weidenstetten. Without more material its status is problematical. 2. Suevosciurus (Treposciurus) romani Hartenberger 1973, from the late Bartonian of Robiac. Bosma (1974: 33) put it in Sciuroides. Comparison of casts of Hartenberger’s figured specimens with the type series of S. siderolithicus and referred material from the Quercy Phosphorites in the BM(NH) has shown no tangible differences. UM RBN900 (P*—M! according to Hartenber- ger 1973: pl. 2, fig. 1) is reidentified as DP*-M! because of the large parastyle and distinct paraconule of the DP*. The species is here synonymized with S. siderolithicus and the Robiac specimens are considered good confirmatory evidence of the Bartonian age of Eclépens-Gare. 3. Sciuroides intermedius Schlosser 1884. Schmidt-Kittler (1971a) placed this species in his genus Treposciurus. The material which Hartenberger (1973: 22) referred to S. intermedius is Adelomys depereti Stehlin & Schaub 1951 (see Bosma 1974: 54). 4. and 5. Adelomys depereti Stehlin & Schaub 1951 and Sciuroides quercyi Schlosser 1884. Bosma (1974: 55) placed both in Paradelomys. They cannot be Sciuroides because they have a deep mesially-directed sinus and its closure buccally (when present) can be shown to be com- posed of the lingual end of metalophule I fused with incipient protolophule II. This closure is thus not homologous with the strong shallowly-indented endoloph of Sciuroides. 6. Sciurus ruetimeyeri Pictet & Humbert 1869. This species was referred by Hartenberger (1973: 21) to Sciuroides. Schmidt-Kittler (1971a: 72-73) thought it should go in a genus of its own while retaining it as ‘“Sciuroides’ rutimeyeri (sic). As interpreted by the better preserved of the two figured syntypes (Pictet & Humbert 1869: pl. 14, fig. 7) (LGM 40464: LM 2930), it is a species of Paradelomys very similar to and possibly conspecific with P. crusafonti. The upper molar from Eclépens B identified by Schmidt-Kittler (1971a: text-fig. 28) as ‘Sciuroides’ riitimey- eri fits better with Paradelomys depereti. Sciuroides rissonei sp. nov. (Pl. 15; Text-figs 32-33; Tables 13—15) v. 1977b Sciuroides cf. russelli (Hartenberger & Louis); Hooker: 141. v. 1980 Sciuroides aff. russelli (Hartenberger & Louis); Hooker & Insole: 39. Name. After Mr A. Rissoné for help with field work. Ho.orype. Right M,,. (?M,), M37386. PI. 15, fig. 10. PARATYPES. (72): 7 DP* (M35532, M36101—4, M37366-7), 3 P* (M35539, M36099, M37365), 20 M' (M35533-5, M36038, M36105-13, M37368-73, M37699), 8 M? (M35537, M36054, M36115—-7, M37374-6), 9 DP, (M36119-23, M37379—82), 3 P, (M36118, M37377-8), 19 M,/2 (M35540—5, M36124-33, M37383-5), 3 M3 (M36134—-5, M37700). DOUBTFULLY REFERRED MATERIAL. 20 upper incisors (M35579-81, M36136—41, M36175, M37387-95, M37701); lower incisor (M3614?). TYPE HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. D1aGnosis. Medium-sized species of Sciuroides, mean length of M'/? 2-68 mm. Lingual cusps of upper molars less than twice the height of the buccal cusps; height of buccal cusps of lower molars less than tooth width. Less than 25% of M'/*s have metalophule I joining endoloph. 70% of M,).s lack enamel wrinkling. More than 85% of lower premolars and molars have oblique ectolophid which joins hypolophulid between a quarter and a third of the distance from the hypoconid. DIFFERENTIAL DIAGNOSIS. S. russelli has stongly wrinkled molar enamel. Its lower premolar and molar ectolophids are bent in the middle, the mesial halves being oblique, the distal halves J. J. HOOKER 302 ‘(6609EW ) (PastaAod) TOE 20g (PETOEIN) ©W Yl TI Blt (SZI9EW) (2° W) “TW Yl TT Bt (98ELEW) (Passeea) (2 WW) WA 14s adAjofoy ‘OT “314 (ST I9EW) "d Yl ‘6 ‘SIA (Z8ELEW) (Possorcs) "Gq 1811 ‘g “BI (LOT9EW) (Pass9Aed) (0 WN) 7/,W 181 “ZL “BIA (L9OELEW) (PasseAcd) -dC 1811 ‘9 “B14 (ZESSEW) pd Yl ‘S “Bt (PLELEW) WC Wl ‘b Bt “(TTT 9€ WW) (P2st2A04) (0 7) 2, 1481 “€ “S14 “(9OT9EW) (2, ZW) 2, Al ‘7 BT pd WS ‘] “SI ¢-Z] x ‘MOIIEQYII9ID WO “AOU “ds 1auOssi4 SaplouniIg JO Y99} Yooyd Jo sydeIZoIIW UOII[a BSuluUvIG CT ALI BARTONIAN MAMMALS OF HAMPSHIRE BASIN 303 Table 13 Statistics of length and width measurements of cheek teeth of Sciuroides rissonei from the Creechbarrow Limestone Formation, Creechbarrow. (N = number of specimens; OR = observed range; X = mean; s = standard deviation; v = coefficient of variation (in brackets where N is too few for v to be significant). See p. 194. SS Longin —— > ———$___—_ Width ——___"_""— Tooth N OR X Ss Vv N OR X Ss V DP* 2:45-2:75 2:577 0:1203 4-67 2:20-2:75 2-414 0-1725 7-15 p+ 2:50-3:20 2:850 0:4949 (17-36) 2:60-3:00 2-767 0-2082 (7:52) M?/2 1 2:50-2:95 2-682 0:1030 3:84 1 2:65-3:10 2-806 0-1116 3:98 2:40-2:90 2-631 0-1438 5:47 DAS S252 SOL075 4-26 2:80-2:95 2-867 0-0764 (2:67) 2:55-2:90 2:775 0-1096 3-95 3:20-3:25 3-225 0-0354 (1-10) 2:30-2:90 2:531 0-1811 7-16 1:70-2:20 1-856 0-1568 8-45 2:00-2:40 2-267 0-2309 (10-19) 2:25-2:50 2:366 0-0831 $79) 2:40-2:55 2-475 0-1061 (4:29) _ ND W OOOO YN) WwW +] ~) a) - NDWwWNONND == (with the mesoconid) parallel with the buccal border of the tooth. The ectolophid joins the hypolophulid slightly nearer the hypoconid. Note that Hartenberger & Louis (1976) recognized this species as congeneric with ‘Suevo- sciurus romani but did not realize that both belong in the genus Sciuroides. Their diagnosis reflects this: ‘“Suevosciurus légérement plus petit que S. mutabilis. D4 et P4 inférieures et supérieures peu molarisées’. In fact S. russelli is slightly larger than the type assemblage of Treposciurus mutabilis, according to Schmidt-Kittler’s (1971a: text-fig. 13) measurements. S. ehrensteinensis is known from a small number of teeth from the English Headon Beds and the German fissure filling Ehrenstein 1A. It is slightly larger than S. rissonei or S. russelli, has higher lingual upper cheek tooth cusps and higher buccal and lingual lower cheek tooth cusps. Its metalophule I usually joins the endoloph (fide Bosma 1974: 33), and its ectolophids are bent. S. siderolithicus is slightly smaller and has a bent ectolophid which joins the hypolophulid nearer the hypoconid. DESCRIPTION. Close relationship between this species and S. russelli is reflected in the initial records from Creechbarrow (see synonymy list, p. 301). It is considered, however, that identifi- cation of S. rissonei as S. russelli on this basis would confuse, not clarify. The enamel wrinkling is an important specific character within the genus Treposciurus (see p. 308) and is thus also considered important in Sciuroides. The obliquity of the ectolophid is a notable primitive character within the family. Tables 14-15 show percentage variation of characters for upper and lower cheek teeth. In addition, the paraconule is always large and the endoloph always complete on upper cheek teeth. Not all the incidences of characters are random. There is an 80% correlation between either one of metalophules I and II being broken and the other being complete in M"”. Fifteen specimens are involved. In one specimen (M37699) both I and II are almost broken. Two specimens (M36112 and M36038) have both I and II broken, but in addition have extra unnamed lophules. In M35535 metalophule II looks complete but is separated from the hypo- cone and joined to the lingual end of the posteroloph by a hypolophule. Pl. 15, figs 2-3 show metalophule I broken and metalophule II complete, while in fig. 7 the condition is reversed. PI. 15, figs 5 and 6 show variations in shape and size of the parastyle in DP*. Pl. 15, figs 10-11 show grade 1 enamel wrinkling, while fig. 12 shows grade 2 wrinkling. Unlike S. ehrensteinensis from the Isle of Wight (Bosma 1974: pl. 4, fig. 2) no mesostylids are present on any of the Creechbarrow teeth and only a faint ectostylid on one M3. S. russelli Ms sometimes have a crest joining the hypolophulid to the posterolophid; in S. rissonei there is one M, and one DP, where a projection distally from the hypolophulid fails to join the posterolophid. The inter- ruption of the metalophulid is a good generic feature. There are two individual M/s, however, where this crest is complete, but in both the anterolophulid is missing. Scoring Characters + (N) of respective teeth units DP* p+ M! Metalophule I joins lingually: 1 29 0 25 metalophule II lingual to 71 100 50 metaconule 2 (1), hypocone (2), 3 0 0 25 distal endoloph (3) (7) (3) (8) (10) (0) Metalophule I broken/unbroken B 86 66:7 57 (7) (3) (7) (10) (0) U 14 33-3 43 Metalophule II broken/unbroken B 57 33-3 50 (7) (3) (6) (10) (6) U 43 66-7 50 Enamel wrinkling (0-3) 0 100 0 100 (7) (3) (7) (10) (6) 1 0 0 0 2 0 33-3 0 3 0 66:7 0 Mesostyle size (0-4) 0 0 0 0 (7) (2) (8) (10) (6) 1 0 0 0 D, 0 0 13 3 86 100 25 4 14 0 62 Mesostyle saliency: prominent (3), 0 >// 50 0 slight (2), non (1), ectoflexus (0) 1 43 50 40 (7) (2) (5) (9) (7) 2 0 0 60 3 0 0 0 Mesostyle mesiodistally elongated 0 85-7 0 75 (7) (2) (4) (8) (5) 1 14-3 100 25 Mesoloph length (0-3) 0 0 0 0 (7) (3) (7) (9) (1) 1 28-6 100 0 2 57-1 0 100 3 14-3 0 0 Protolophule I broken/unbroken B 16-7 66-7 0 (6) (3) (8) (10) (7) 0) 83-3 33:3 100 Metaconule 2 presence/absence 0 0 50 0 (7) (2) (8) (9) (0) 1 100 50 100 Hypolophule absent (0), partial 0 85-7 0 85-7 (1), complete (2) 1 14-3 50 14-3 (7) (2) (7) (9) (0) 2 0 50 0 Posteroloph broken/unbroken lingually B 42-9 0 28-6 (7) (3) (7) (9) (0) U 57-1 100 71-4 Mesocone presence/absence 0 85:7 100 50 (7) (3) (8) (10) (7) 1 14-3 0 50 Sinus depth: shallow (1) to 1 0 0 0 deep (4) D 100 33-3 Bie (7) (3) (8) (10) (8) 3 0 66-7 25 4 0 0 37-5 Protostyle presence/absence 0 100 50 14-3 (6) (2) (7) (10) (8) 1 0 50 85-7 Hypostyle presence/absence 0 28-6 100 0 (7) (3) (8) (10) (0) 1 71-4 0 100 M2? BARTONIAN MAMMALS OF HAMPSHIRE BASIN 305 Tables 14-15 Percentage character analysis of upper (Table 14) and lower (Table 15) cheek teeth of Sciuroides rissonei from Creechbarrow. ‘Scoring units’ give states for characters described in the left hand column. The numbers given under the tooth-headed columns on the right are percentages and refer to the number of teeth showing that particular character state. The lowest row of figures in the left hand column are the respective numbers of each tooth represented. See also p. 303. Characters + (N) of representative teeth Distance along hypolophulid from hypoconid of junction with ectolophid (8) (3) (7) (10) (2) Orientation of ectolophid: straight oblique/bent parallel (7) (3) (7) (19) (2) Anteroconid size (1—3) (0) (0) (7) (10) (2) Anterolophulid absent (0), weak (1), strong (2) (0) (0) (7) (10) (2) Mesoconid length as % depth of sinusid (6) (3) (7) (10) (2) Mesolophid absent (0), + complete (1), 4 complete (2) (8) (3) (7) (10) (2) Enamel wrinkling (0-2) (8) (3) (7) (10) (2) Mesostylid presence/absence (8) (3) (7) (10) (2) Ectostylid presence/absence (8) (3) (7) (10) (2) Hypoconulid presence/absence (8) (3) (7) (10) (2) Distal crest to hypoconulid presence/absence (8) (3) (7) (10) (2) Scoring units DP, Pe M, M, M,; Length —————__ —$>—_—__ Width ——_____—__ Tooth OR 5 : 5 N OR Z § v DP* 3 2:10-2:27, 2418 00854). (6.92), Sin 1k872>-10N 1-99 0 10g tmaammeaeter p4 2 1-77-2-05' 1-01 04980 (037) 2 188-203 1-96 Of06m Gm M2 18 1-95-2-25 2-10 0-0834 3-98 19 1:98-2:28 2-15 0:0946 4.39 M2 6 1:85-2:13 205 00972 475 #5 183-221 206 0-146.N IAM DP, eo a a iL dee bs = 2 P, 4 193-210 201 00850 (423) 4 1:58-1:89 1-72 01365 em M,. 14 2:05-2:40 2-18 0-0996 457 12 1:73-2:10 1-93 01093 5-66 M 7 233-271 248° 01337 540 7 181-200 1-91 0.0629 seon BARTONIAN MAMMALS OF HAMPSHIRE BASIN 309 TYPES AND RANGE. Holotype (LGM 40209) and paratypes from the Sidérolithique, Eclepens (B?), Canton Vaud, Switzerland; teeth in the NMB (referred herein) also from Eclépens B. Schmidt-Kittler (1971a: 53-57) only recorded his material as being from Eclépens, but all rodent material from the new fissures was collected from Eclépens B (personal communication, D. Rigassi 1980). EMENDED DIAGNOSIS. Mean length of M'/? 2:33 mm. Range 2:08-2:64 mm. Treposciurus helveticus preecei subsp. nov. (PI. 16; Pl. 17, figs 1-6; Text-figs 34-35; Tables 16-18) v. 1977b Treposciurus sp.; Hooker: 141. v. 1980 Treposciurus sp.; Hooker & Insole: 39. Name. After Dr R. Preece for help with field work. Types. As the definition of this subspecies is based on a size range which overlaps with the other subspecies, and as its morphological variation is very high, the designation of a holotype is inappropriate. Therefore contrary to Recommendation 73A of the International Code of Zoological Nomenclature (1985: 149), I give all members of this type series equal status in the form of 75 syntypes: 6 DP* (M36034—7, M36100, M37331), 4 P* (M36032-3, M37330, M37337), 25 M'/? (M35509-10, M35512-4, M36039, M36041—53, M36067, M37332-6), 7 M? (M35515-8, M36055, M37338-9), 1 DP, (M36058), 6 P, (M35538, M36056-7, M37340-2), 19 M,/, (M35519-29, M36059—60, M36062—3, M37343-6), 7 M, (M35530—1, M36064—-6, M37347-8). DOUBTFULLY REFERRED MATERIAL. 53 upper incisors (M35564—73, M36068—96, M36170—4, M37349-57): 14 lower incisors (M35574-8, M36097—8, M37358—64). TYPE HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. D1aGnosis. Mean length of M’/? 2:10 mm. Range 1:95-2:25 mm. COMMENT ON TAXONOMY. The Treposciurus from Creechbarrow is very similar in morphology and range of morphological variation to the type and referred material of T. mutabilis helveticus from Eclépens B. The only tangible difference is in the size, it being smaller but with overlap- ping range. As there are morphological differences between T. m. helveticus and T. m. mutabilis, especially now that the former is known from more material from the type locality, it is proposed here that the subspecies T. m. helveticus be raised to the rank of species. Material from the Headon and Osborne Beds, referred by Bosma (1974) and Bosma & Insole (1976) to T. m. helveticus, is morphologically very similar to T. m. mutabilis. It differs only in being slightly smaller, but there is much overlap in measurements. Therefore it is proposed to refer this material to T. mutabilis with no division into subspecies. The overlap in size between the Creechbarrow Treposciurus and T. helveticus Schmidt-Kittler 197la, new rank, from Eclépens B is much less and it is proposed to recognize them as conspecific but belonging to different subspecies. DESCRIPTION. Of the three English Bartonian pseudosciurids, T. helveticus preecei is the most variable in morphology, especially in the upper molars. Since two particular morphologies, which characterize respectively T. mutabilis and T. intermedius, are present at Creechbarrow, it could be doubted that only a single species were represented there. However, the lack of any clear-cut morphological differences on which subdivisions could be made, the unimodal size distribution and its low coefficient of variation (Table 16) especially in the first and second molars, and the presence of similar morphological range in the Eclépens B assemblage of T. helveticus, serve to remove doubt. In Tables 17-18, morphological variation is given as in Tables 20-21 for Suevosciurus autho- don (pp. 319-321). The major variations in M'/” are illustrated in PI. 16, figs 2-3, 7, 9-12. PI. 16, fig. 11 shows a morphology almost identical to that of T. mutabilis; the only difference is that the enamel in the centre of the tooth is not quite so strongly or deeply reticulate as is usual for T. mutabilis. Fig. 12 shows a morphology almost identical to that of T. intermedius; the 310 J. J. HOOKER replacement of the lingual end of metalophule II by a hypolophule joining the posteroloph is a variant feature which can often occur in any species of Treposciurus or Suevosciurus. The former tooth is the largest in the assemblage and the latter one of the smallest, thus corre- sponding to the respective size differences of the Ludian species they resemble. Apart from these extremes there are specimens that show a mainly incomplete metalophule I which usually joins the endoloph just distal to the midpoint, unlike T. intermedius where it joins the endoloph a a Aa A e a A ° fe} A w ° ° @ » e ao 90 0 2 ° fo) w © A fo) fe} @ T.h. preecei ul/2 =o 0 2 ° T.h. helveticus ul/2 =A ° mio =a 3 m2 =A T. h. preecei M A =2 @=2 4 Text-figure 34 Scatter diagrams of length (1) against width (w) in DP*, P*, DP, and P, of Treposciurus helveticus preecei subsp. nov. from Creechbarrow, and in M!” of Treposciurus helveticus helveticus Schmidt-Kittler from Eclépens B and M'/? and M3 of T. helveticus preecei from Creechbarrow. Measurements in millimetres. Lines join teeth of one individual. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 311 a a a a a A A a a ° (0) a ° 4 Aa MP o 4 ° Aa of e e es A ° e e feay e ° 0 T. h. preecei My/2=0 M3 =e T. h. helveticus My/9= 4 A 4 2 ; Text-figure 35 Scatter diagrams of length (1) against width (w) in M,,. and M, of Treposciurus helveticus helveticus Schmidt-Kittler from Eclépens B and T. helveticus preecei subsp. nov. from Creechbarrow. Measurements in millimetres. Lines join teeth of one individual. nearer the hypocone. Some specimens have smooth enamel and no metalophule I at all and there are complex intermediates between most of these types. Fig. 7 shows a doubled meta- conule 2, whilst Fig. 9 shows a single, huge, undivided metaconule such as occurs in the late Lutetian Protadelomys cartieri from Egerkingen y, Switzerland. One very curious M’/? (M35514) is relatively elongated, has no metalophule I, a very shallow sinus and a strong lingual cingulum completely spanning the sinus and rising distally to about a third the crown height of the hypocone. There is less variation in the lower molars. No M,,.s show features typical of T. mutabilis and the lower molars of T. intermedius have no special distinguishing characters. M35524 is peculiar in having a double mesoconid. An M,; (M37348) deserves special mention (PI. 17, fig. 6); unlike the typical shape shown in PI. 17, fig. 4, its outline is a parallelogram, the acute angles being the mesiolingual and distobuccal corners. PHYLOGENETIC RELATIONSHIPS. From the polymorphic pattern described above it is possible to formulate a phylogenetic model for the three Treposciurus species. This involves a cladogenesis with T. helveticus as the common ancestor of T. mutabilis and T. intermedius. The two Ludian species would have developed by selection of different morphs already present in the Bartonian species. The two daughter species are thus more specialized than the ancestral species. This speciation event seems to have occurred at about the Bartonian/Ludian boundary, at a time when there was a widespread marine regression with mammalian migrations and extinctions within Europe (see Garimond et al. 1975), thus increasing selection pressures. It is not certain whether the differences between the two subspecies of T. helveticus reflect geographical or stratigraphical separation. There is some biostratigraphical evidence that Eclépens B is younger than Creechbarrow (see correlation section, pp. 425—427). Alternatively, if the subspecies were geographical, it is possible that T. intermedius arose in the region of southern England and T. mutabilis in the region of Switzerland. 32 J. J. HOOKER Scoring Characters + (N) of respective teeth units DP* p+ M! M? M? Metalophule I joins lingually: 0 0 75 37-5 63-6 metalophule II lingual to 1 20 0 0 0 metaconule 2 (1), hypocone (2), D 80 25 12:5 9-1 distal endoloph (3), central endoloph (4); 3 0 0 0 18-2 is absent (0) 4 0 0 50-0 9-1 (5) (4) (8) (11) (0) Metalophule I broken/unbroken B 60 100 100 90:9 (5) (4) (8) (11) (0) U 40 0 0 9-1 Metalophule II broken/unbroken B 0 25 50 18-2 WS (5) (4) (8) (11) (4) U 100 75 50 81-8 DS Enamel wrinkling (0-3) 0 83-3 0 0 12-5 0 (6) (3) (7) (8) (6) 1 0 0 85-7 50-0 0 2 16-7 33-3 14-3 37S) 0 3 0 66-7 0 0) 100 Mesostyle size (0—4) 0 0 0 0 0 80 (5) (3) (7) (11) (5) 1 0 0 0 0 0 D 0 66:7 0 21 0 3 80 33-3 100 90-9 0 4 20 0 0 0 20 Mesostyle saliency: 0 De 0 0 0 0 slight (2), non (1), ectoflexus (0) 1 50 66:7 85:7 100 40 (4) (3) (7) (10) (5) 2 25 33-3 14-3 0 60 Mesostyle mesiodistally elongated 0 100 100 87-5 100 0 (5) (3) (8) (11) (5) 1 0 0 265) 0 100 Mesoloph length (0-3) 0 20 33-3 25 DiD, 100 (5) (3) (8) (1) (5) 1 40 66-7 12:5 18-2 0 2 20 0 50 45-5 0 3 20 0 12:5 9-1 0 Protolophule I broken/unbroken B 0 66-7 75 37:5 20 (5) (3) (8) (8) (5) U 100 333 25 62:5 80 Metaconule 2 presence/absence 0 33-3 0 U2es) 80 (6) (3) (8) (10) (0) 1 66-7 100 87-5 20 Hypolophule absent (0), partial 0 100 100 87:5 90 (1), complete (2) 1 0 0 0 0 (6) (3) (8) (10) (0) 2 0 0 12:5 10 Posteroloph broken/unbroken lingually B 50 0 87:5 44-4 (4) (4) (8) (9) (0) U 50 100 12:5 55-6 Mesocone presence/absence 0 100 100 62:5 80 SyFil (6) (4) (8) (10) (7) 1 0 0 SPS 20 42-9 Sinus depth: shallow (1) to 1 0 0 12-5 0 0 deep (4) 2 20 75 Biles 20 14-3 (5) (4) (8) (10) (7) 3 80 25 Bie5 50 57-1 4 0 0 12:5 30 28-6 Protostyle presence/absence 0 20 66:7 62:5 36-4 50 (5) (3) (8) (11) (4) 1 80 33-3 BIS 63-6 50 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 313 Scoring Characters + (N) of respective teeth units DP* p+ M! M? M? Hypostyle presence/absence 0 25 100 DS 60 (4) (4) (8) (10) (0) 1 75 0) 75 40 Paraconule small (1), large (2) 1 0 33-3 0 0 28-6 (6) (3) (8) (11) (7) 2 100 66-7 100 100 71-4 Endoloph broken (1), complete (2) 1 0 25 50 12-5 0 (6) (4) (8) (9) (5) D, 100 75 50 87-5 100 DP* parastyle bicuspid 0 66:7 (3) 1 33:3 Tables 17-18 Percentage character analysis of upper (opposite and above) and lower (below) cheek teeth of Treposciurus helveticus preecei from Creechbarrow. For explanation see Table 14, p. 305. Scoring Characters + (N) of respective teeth units DPA Py M, M, M, Distance along hypolophulid <} 100 DS 45-4 50 83-3 from hypoconid of junction rt 0 75 36:4 50 16:7 with ectolophid 4 0 0 18-2 0 0 (1) (4) (11) (6) (6) Anteroconid size (1—3) 1 40 33-3 50 (0) (0) (10) (6) (6) 2 40 * 66:7 50 3 20 0 0 Anterolophulid absent (0), weak 0) 87:5 83-3 80 (1), strong (2) 1 05) 16:7 0 (0) (0) (8) (6) (5) 2) 0 0 20 Mesoconid length as % depth of 40 100 0 0 0 0 sinusid 50 0 0 10 0 0 (1) (4) (10) (6) (6) 60 0 0 0 33-3 0 70 0 25 30 33-3 16-7 80 0 IS 40 16-7 16-7 90 0 0 10 16-7 66:6 100 0 0 10 0 0 Mesolophid absent (0), + complete (1) 0 100 25 45:5 66:7 50 (1) (4) (11) (6) (6) 1 0 75 54-5 333}°3} 50 Enamel wrinkling (0-2) 0 100 0) 75-0 80 0 (1) (4) (8) (5) (6) 1 0 25 12:5 20 50 2 0 75 12:5 0 50 Mesostylid presence/absence 0 100 100 70 60 83-3 (1) (4) (10) (5) (6) 1 0 0 30 40 16-7 Ectostylid presence/absence 0 100 100 100 83-3 100 (1) (4) (8) (6) (6) 1 0 0 0 16-7 0 Hypoconulid presence/absence 0 0 50 70 60 66:7 (1) (4) (10) (5) (6) 1 100 50 30 40 33-3 Distal crest to hypolophulid 0 100 100 90-9 100 100 (1) (4) (11) (6) (6) 1 0 0 9-1 0 0 314 J. J. HOOKER BARTONIAN MAMMALS OF HAMPSHIRE BASIN 315 Genus SUEVOSCIURUS Dehm 1937 [incl. Suevosciurus (Microsuevosciurus) Hartenberger, 1973] Type SPECIES. Sciuroides fraasi Major 1873. Ludian fissure filling; Orlinger Tal, West Germany. INCLUDED SPECIES. S. ehingensis Dehm 1937, S. minimus (Major 1873), S. palustris (Misonne 1957) and S. authodon sp. nov. Note that Bosma (1974: 53) has removed Hartenberger’s (1973: pl. 1, figs 6-13) ‘Suevosciurus aff. minimus’ from Fons 4 from Suevosciurus and compared it with Treposciurus. RANGE. Marinesian—Ludian, England; Ludian—Stampian, W. Germany; Ludian, France and Switzerland; early Stampian, Belgium; Oligocene, Majorca. EMENDED DIAGNOSIS. Upper cheek teeth with: endoloph high, usually complete; M! ? meta- lophule I usually missing, occasionally present, not dentine-cored, developed as ridge lingually only or as metaconule 1, in which case usually attached to endoloph just distal to midpoint; paraconule strong to weak on DP* and M' °; metaconule 2 usually absent; DP* with medium to large, often bicuspid parastyle and usually concave mesiolingual margin. Lower cheek teeth with: ectolophid complete, but low, to incomplete, bent parallel, joining hypolophulid at or close to hypoconid; mesoconid nearly always strong. Lower molars with: M,_, anteroconid usually large, usually buccal to midline: -anterolophulid nearly always absent; anterolophid strong; metalophulid high and complete. M, hypolophulid incomplete or low and complete. DP¥ with: crown height equal to that of molars; metalophulid present. Anterior palatine foramina project just posterior to maxillary—premaxillary suture (known for S. fraasi and S. ehingensis). PREVIOUS DIAGNOSES. The most recent was by Hartenberger (1973: 11), but this reflects his inclusion in the genus of species referable to Treposciurus and Sciuroides (see discussions under these genera). The diagnosis herein is closer to that of Schmidt-Kittler (1971a: 39) but involves the reidentification of his P¢ as DP and the characters of subsequently described material from the Isle of Wight (Bosma 1974) and Creechbarrow (herein). Suevosciurus authodon sp. nov. (Pl. 17, figs 7-12; Pl. 18; Text-figs 36-39; Tables 19-21) v. 1977b Suevosciurus sp.; Hooker: 141. v. 1980 Suevosciurus sp. 1; Hooker & Insole: 39. Name. Noun in apposition, from Greek adf@s, afresh, and ddovs, tooth, in allusion to the tooth replacement in this species. Hotorype. Right DP*, M35469. PI. 18, fig. 5. PARATYPES. (336): 27 DP* (M35470-1, M35787, M35797-813, M36447, M37205-10); 23 P* (M35464-8, M35486, M35785-6, M35788-96, M37199-204); 102 M'/? (M35472-7, M35814-75, M3721 1-35); 25 M?® (M35488-90, M35876-88, M37236—44): 2 DP, (M35897, M37247); 13 Py (M35491—2, M35889—-96, M35898, M37245-6); 102 M,,. (M35493-504, M35899-960, M37248— 75); 42 M; (M35505-8, M35961—87, M37276—86). DOUBTFULLY REFERRED MATERIAL. 44 upper incisors (M35549-53, M35988—36008, M36161-2, M37287-302); 67 lower incisors (M35554—63, M36009-31, M36163—9, M37303-29). Plate 16 Scanning electron micrographs of upper cheek teeth of Treposciurus helveticus preecei subsp. nov. from Creechbarrow, x 18. All are syntypes. Fig. 1, left P* (M36032). Fig. 2, left M'/? (M'?) (M36045). Fig. 3, left M‘/? (M2?) (M36043). Fig. 4, left M? (M35516). Fig. 5, left DP* (M36100). Fig. 6, right DP* (reversed) (M36037). Fig. 7, right M'/* (M!2) (reversed) (M35513). Fig. 8, right M? (reversed) (M35518). Fig. 9, right M!/? (M2?) (reversed) (M35512). Fig. 10, right M'/? (M2) (reversed) (M37334). Fig. 11, right M'‘/? (M72) (reversed) (M36051). Fig. 12, left M‘/* (M*?) (M36039). See p. 309. J. J. HOOKER 316 “S1E d 205 (LO6SEI) (4 'W) “WW Yl ‘ZI 3IA (L68SEW) (pasiaos) "Gq ISI “TT SI (196SEW) SW Wel Ol Sta (OZ6SEW) (possaaod) (7) “W148 ‘6 “BL “(ST6SEW) (6'W) “TW Yl ‘8 “Std “(l68SEW) "d Yl “L “BIH “ST Xx “AOU “ds uopoyiny snaniosoaang ZI-L ssi ‘60¢ “d 928 (SPELEW) (passeros) © ISI ‘9 “BLY (SSO9EW) "dC YI ‘S “BIA (S9O9EW) (Poss9od) © Al YSU “p “SI (SPELEW) (Pass9roy) (4 WW) © WAL YB ‘E “BLA c/T IAT WUBI ‘7 “BI (ZPELEW) (Passoaod) %q 1YSt ‘| “SI gy] x ‘sadAjuds “Aou ‘dsqns tadaaud snoijaajay sniniosoda4], 9—]| Ssly “MOIIVQYOIIID WO sniniasoaang pure sn.inidsoda. J, JO YII9} Y99YO J9MO] JO SydvVIsOINW UOI}DaTa BuIUULIS = LT BRI (OPELEW) (Passere.) (2 ') BARTONIAN MAMMALS OF HAMPSHIRE BASIN 31 7/ TYPE HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DiaGnosis. Mean length of M'? 1:-49mm; range 1:350-1:625mm. P# present. DP* mean length/width ratio 1-08. 70% M's have size 2 metaconule 1. DESCRIPTION. One of the most important discoveries in the Creechbarrow material is of upper and lower fourth permanent premolars. Teeth previously identified as such can now be conclu- sively identified as upper and lower fourth deciduous premolars. The possibility of DP4 reten- tion in Suevosciurus was first suggested by Bosma (1974: 40). She reasoned that no distinction could be made between deciduous and permanent premolars despite the abundant material available and that ‘P*’ had a strong parastyle typical of DP* in other pseudosciurid genera. The supposed characters of ‘P* can no longer be used as a constant diagnostic feature for Suevo- sciurus and even though retention of DP¥ characterizes the best-known of the various species, it cannot be used in the generic diagnosis with the inclusion of S. authodon. Nevertheless the increase in crown height of DP can be used and this probably foreshadowed later loss of P#. There is much variation in practically all the known morphological characters of Suevo- sciurus and size has been used by previous workers almost exclusively to distinguish between species. The low coefficients of variation for the large tooth assemblage from Creechbarrow indicate the presence of a single species (Table 19). An attempt is made here to provide numerical scales for features of variable size (e.g. mesostyle), or just to indicate presence or absence, and to show percentage occurrence within the Creechbarrow assemblage (see Tables 20-21). It is hoped that this will encourage other workers to do the same for their material, to allow species-level distinctions on at least percentage characters. Some are suggested under Phylogenetic Relationships (p. 321). Table 19 Statistics of length and width measurements of cheek teeth of Suevosciurus authodon from Creechbarrow. (Notation as in Table 13, p. 303). OR X s Vv N OR Tooth bell n < DPE* 17. =: 1-250-1:550 =1-43 =: 0-086 6:00 7p) > WMeilsUSieowsy —le3¥h (08) 7:38 Re DSi ICO25) sec 49 0:05 5-06 22. =1:350-1:550 1:45 0-066 4:58 M?? 84 = 1350-1625 1:49 0-055 3-71 86 =: 11375-1675 =—-:1:57 ~—- 0-068 4:36 M? 24 =1-250-1:500 1:38 0-059 4-25 Jy esisjesy7s) te ({O)5)3) 3-73 DP, 2 1-350-1-400 1:38 0-035 (2-57) 2097 5— SOS O06 (LO 0) Pr 121-350-1525 1:44 0-048 3:33 10 =1-050-1-250 1:18 0-070 5-89 M,,2 88 1-425-1-700 1:57 0-066 4-18 96 1-250-1:500 1:35 0-056 4-17 M, 42 = 1-550-1:850 1:70 0-082 4-83 40 1:150-1:500 1:34 0-063 4-72 The specimens figured in Plate 18 show some of the range of variation in M'’’s relevant to an understanding of the character developments detailed in Table 20. Metalophule I may be a complete ridge (fig. 8), a lingually restricted ridge (figs 3, 9), reduced to just metaconule | (fig. 2) or missing (fig. 7). Size 1 is shown in figs 2—3, size 2 in fig. 9 and size 3 in fig. 12. There may be a minor ridge of enamel joining metaconule 1 to either the endoloph, hypocone, metalophule II or a combination of the three. Fig. 9 shows links to both the endoloph and metalophule II. Metaconule 2 is sometimes present (figs 8, 11) but usually absent (figs 2-3, 7, 9-10, 12). Enamel wrinkling may be fine low beading (fig. 2) or finely (fig. 7) to coarsely (fig. 9) reticulate, and may show degrees of obscuration by wear (figs 8, 12). The sinus may be shallow (fig. 9), of medium depth (fig. 2) or deep (fig. 12); it is nearly always complete. The mesostyle has been given five arbitrary size numbers: 0 (fig. 10), 1 (fig. 8), 2 (fig. 2), 3 (figs 11-12), 4 (figs 3, 7, 9). Its buccal saliency is sometimes but not always related to its size; e.g. there may be an ectoflexus (fig. 10) or the mesostyle may be non-salient (grade 1: fig. 8) or salient in two size grades (2: fig. 3; 3: fig. 9). The mesostyle may be a circular object (fig. 2) or it may be elongated into a partial J. J. HOOKER 318 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 319 Table 20 Percentage character analysis of upper cheek teeth of Suevosciurus authodon from Creech- barrow. For explanation see Table 14. Scoring Characters + (N) of respective teeth units DP* pt M! M? M? Metalophule I shape: metaconule 1 1 only (1), ridge (2) 2 30 11-1 16:7 (10) (0) (18) (30) (0) Metalophule I size 0 23-1 100 53) 3-2 (13) (15) (19) (31) (0) 1 23-1 0 31-6 19-4 2 46-1 0 42:1 54-8 3 WET 0 21-0 22-6 Metalophule I joins endoloph (1), 0 10 39-0 13-3 hypocone (2), metalophule II (3), 1 0 50-0 40-0 endoloph and hypocone (4), hypocone 2 60 5:5 13:3 and metalophule II (5), endoloph and 3 30 0 16:8 metalophule II (6), all three (7), none (0) 4 0 5:5 0 (10) (0) (18) (30) (0) 5 0 0 3-3 6 0 0 10-0 7 0 0 3:3 Metalophule II broken/unbroken B 5 25 0 12-7 83:3 (20) (20) (23) (55) (12) U 95 75 100 87:3 16:7 Enamel wrinkling (0-3) 0 100 0 10 0 0 (5) (16) (10) (22) (11) 1 0 0 60 63-6 0 D 0 50 30 36-4 45:5 3 0 50 0 0 54-5 Mesostyle size (0—4) 0 0 0 0 1:9 12:5 (18) (21) (22) (52) (16) 1 11-1 0 9-1 1:9 0 2 44-4 4:8 9-1 32:8 0 3 38-9 47-6 59-1 36:5 25-0 4 5-6 47-6 BAT] 26:9 62:5 Mesostyle saliency: prominent (3), 0 26:3 30 4:8 2:0 0 slight (2), non (1), ectoflexus (0) 1 68-4 60 57:1 43:1 40 (19) (20) (21) (51) (20) 2 5)23} 10 33-3 35-3 45 3 0 0 48 19-6 15 Mesostyle mesiodistally elongated 0 100 66:7 100 86:5 0) (19) (21) (22) (52) (21) 1 0 333)°3) 0 13-5 100 Mesoloph length (0-2) 0 5:6 66:7 Apes} 44-2 85-7 (18) (21) (22) (52) (21) 1 61-1 23-8 54-5 46-2 14-3 2 33-3 9-5 18-2 9-6 0 Table 20 (continued) Plate 18 Scanning electron micrographs of upper cheek teeth of Suevosciurus authodon sp. nov. from Creechbarrow, x 25. Fig. 1, right P* (reversed) (M35795). Fig. 2, left M'/? (M!?) (M35821). Fig. 3, left M!'/? (M2?) (M35481). Fig. 4, right M? (reversed) (M35489). Fig. 5, holotype right DP* (reversed) (M35469). Fig. 6, left DP* (M36447). Fig. 7, right M'/* (M2?) (reversed) (M35861). Fig. 8, right M'/? (M2?) (reversed) (M35855). Fig. 9, left M'/* (M*?) (M35833). Fig. 10, left M*/? (M2?) (M35838). Fig. 11, right M'/? (reversed) (M35860). Fig. 12, left M'/? (M2?) (M35844). See p. 315. 320 J. J. HOOKER Table 20 continued Scoring Characters + (N) of respective teeth units DP* p+ M! M2 M? Protolophule I broken/unbroken B 9-5 50 4-4 3-6 MBPS) (21) (16) (23) (55) (17) U 90-5 50 95-6 96-4 76:5 Metaconule 2 absence/presence 0 66:7 65 78-3 62:3 (21) (20) (23) (53) (0) 1 33-3 315) 21:7 Bie Hypolophule absent (0), partial 0 85:7 79:0 95:2 70-4 (1), complete (2) 1 14:3 10:5 4:8 3-7 (21) (19) (21) (54) (0) 2 0 10:5 0 25-9 Posteroloph broken/unbroken lingually B 33-3 36:8 13-0 1:8 (21) (19) (23) (55) (0) U 66-7 63-2 87-0 98-2 Mesocone absence/presence 0 100 100 94-1 91-9 100 (21) (18) (18) (40) (14) 1 0 0 5-9 8-1 0 Sinus depth: shallow (1) to 1 0) 0 0 0 4:8 deep (4) D) Sp) 20 4-4 18-2 14-3 (19) (20) (23) (55) (21) 3 68:4 75 Swen 65-5 Sal 4 26:3 5 43-4 16:3 23-8 Protostyle absence/presence 0 S28) 100 26:3 59:5 75 (17) (18) (19) (37) (16) 1 47-1 0 1 40-5 25 Hypostyle absence/presence 0 80-9 73-7 33-3 47-2 (21) (19) (18) (36) (0) 1 I@eil 26:3 66:7 52:8 Paraconule absent (0), small (1), 0) 0 5 0 0 0 large (2) 1 0 5 4-4 535) 36:8 (21) (20) (23) (55) (19) 2 100 90 95-6 94-5 63-2 Endoloph broken (1), complete (2) 1 0 21-1 4-4 3-6 0 (20) (19) (23) (55) (15) D 100 78-9 95-6 96-4 100 DP* parastyle bicuspid 0 53-3 (15) 1 46:7 DP* mesiolingual margin concave 0 60 (20) 1 40 mesoloph (grade 1: fig. 7; or 2: fig. 12); alternatively it may be mesiodistally elongated (figs 3, 9). Whereas in advanced species of Suevosciurus the M!? paraconule is often reduced, this is so in only four Creechbarrow specimens (e.g. fig. 9). Metalophule II may be broken buccal and/or lingual to the position of metaconule 2 (whether present or absent) and there may be a hypolophule (fig. 7). Other minor cusps which may be present are protostyle (= anterocone), mesocone and hypostyle (fig. 12). When the usual method of distinguishing M! from M? (see under Taxonomic Treatment of the Pseudosciuridae, p. 298) was used for this species, it was found that M?s outnumbered M's two to one. One explanation is that the M* type morphology commonly occurred in M'. On M?, although the paraconule is often weaker than on M! ”, it may also be accompanied by a crest which joins it to the anteroloph, as on one M? of Sciuroides rissonei (M361 15). Because of the distal tapering of the outline of this tooth, a metacone is present, as shown in PI. 18, fig. 4, in only three out of twenty specimens. The M,,,s are less variable than the upper molars. The ectolophid/hypolophulid connection BARTONIAN MAMMALS OF HAMPSHIRE BASIN 321 varies from at or very close to the hypoconid (PI. 17, fig. 8) to about a quarter the distance along the hypolophulid from the hypoconid (fig. 9). The anteroconid is always present and often large; fig. 8 shows a size | anteroconid, fig. 9 a size 2 and fig. 12 a size 4. Figs 8-9 show the typical 60—70% filling of the sinusid by the mesoconid. Enamel wrinkling resembles that on M! *. Most Suevosciurus sp. from the Mammal Bed of Hordle Cliff have the buccal end of the mesoconid fused to the hypoconid. Schmidt-Kittler (1971a: 44) noted this occasionally also in S. fraasi from southern Germany. Proportionately fewer specimens from Creechbarrow than from Hordle have this character. PHYLOGENETIC RELATIONSHIPS WITH OTHER SPECIES OF Suevosciurus. S. authodon appears primi- tive in its retention of P¥; relatively short DP*; normally large paraconule; normal presence of Table 21 Percentage character analysis of lower cheek teeth of Suevosciurus authodon from Creech- barrow. For explanation see Table 14. Scoring Characters + (N) of respective teeth units DP, Pi M, M, M,; Distance along hypolophulid <} 50 55:6 78 47 65:6 from hypoconid of junction 4 50 44-4 22 53 31-3 with ectolophid no link 0 0 0 0 spil (2) (9) (32) (34) (32) Anteroconid size (1—4) 1 3-1 22-9 74:3 (0) (0) (32) (35) (35) 2 21-9 31-4 DT] 3 59-4 37:1 0 4 15-6 8-6 0 Anterolophulid absent (0), weak (1) 0 80-7 86:7 40-7 (0) (0) (31) (30) (27) 1 19-3 13-3 59-3 Mesoconid length as % depth of 50 0 11-1 22:6 27:8 6:1 sinusid 60 50 22:2 B25 19-4 12:1 (2) (9) (31) (36) (33) 70 0 33-4 35-5 27-8 33-3 80 50 22:2 6-4 13-9 27:3 90 0) iVileil 0 2:8 15-1 100 0 0 0 8-3 6:1 Mesoconid fused buccally with 0) 100 100 67:7 78:8 78-8 hypoconid 1 0 0 B23 21-2 21:2 (2) (9) (31) (36) (33) Enamel wrinkling (0-3) 0 100 0 8-7 22:2 0 (2) (6) (23) (18) (21) 1 0 0 73-9 50-0 14-2 2 0 50 17-4 27:8 42.9 3 0 50 0 0 42-9 Mesostylid absence/presence 0 100 100 I) 100 70 (2) (9) (33) (38) (26) 1 0 0 6-1 0 30 Ectostylid absence/presence 0 100 100 96:9 83:8 100 (2) (9) (32) (37) (33) 1 0 0 BA 16:2 0 Hypoconulid absence/presence 10) 100 44-4 SES 71-9 86-7 (2) (9) (33) (32) (30) 1 0 55:6 48-5 28-1 13-3 Distal crest to hypolophulid 0 100 778 97-1 85-7 96:1 absence/presence 1 0 22:2 2:9 14:3 3-9 (2) (9) (34) (35) (26) 322 J. J. HOOKER incomplete metalophule I with large metaconule 1; and normally complete M, hypolophulid. In all other assemblages of Suevosciurus species where cheek teeth mesial to the molars are known, only DP¥ (never P$) have been found; this tooth tends to have a higher length/width ratio than in S. authodon (except in the holotype of S. ehingensis), although there is some overlap. In assemblages from the early Ludian of the Isle of Wight a short incomplete meta- lophule I is occasionally present (Bosma 1974: 40); it also occurs in the Bavarian assemblages (Schmidt-Kittler 1971a: text-fig. 17). In assemblages of Suevosciurus other than from Creech- barrow, the M‘/? metaconule 2 is normally missing. It is not mentioned by Bosma (1974) as occurring in the Isle of Wight assemblages and only occasionally as a vestigial feature by Schmidt-Kittler (1971a: 40, text-fig. 17a). Attempts to compare the Creechbarrow Suevosciurus with existing material from the liter- ature are hampered by the lack of numerically delineated information on morphological varia- tion and the belief that the various species only differ in size. oes: wr nouu dl ~ Ln) =Z=OO0OO00o 1.4 5 ! Text-figure 36 Scatter diagrams of length (1) against width (w) in DP*, P*, M‘/? and M® of Suevosciurus authodon sp. nov. from Creechbarrow. Measurements in millimetres. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 323 c O Oo © © 6 ® @@soms ul ©®62506O 6 oOo Oo ® © 2 ° (0) (0) ie) Oo ©) © ¢ ° OR=s2 Yo © ©. C& © ° Ons oe} O ° O=4 Oro M12 Text-figure 37 Scatter diagrams of length (1) against width (w) in DP,, P,, M,,. and M, of Suevosciurus authodon sp. nov. from Creechbarrow. Measurements in millimetres. Schmidt-Kittler (1971a) identified two principal lineages of Suevosciurus which differed in size where found together at any one locality. The sequence of fissure-fill localities he dated by correlating other elements of their mammal faunas with those known from stratified sequences elsewhere in Europe. The smaller lineage he referred to S. fraasi, the larger to S. ehingensis. Each lineage fluctuated in size somewhat but there was a net size increase with time. Some very small teeth from Ehrenstein 1A and Weissenburg 6 he removed from his two lineages and identified as S. minimus on the basis of size. Bosma (1974) also demonstrated a slight size increase in a sequence of three superposed horizons in the Isle of Wight Ludian. She noted the problem of identification posed by those of Schmidt-Kittler’s assemblages with only one species of Suevosciurus, as there is overlap if all assemblages are considered together. To avoid identify- ing them by horizon determined in turn by the associated fauna, she suggested a classification based on arbitrary size groupings. The result was the identification of parts of the S. ehingensis 324 J. J. HOOKER lineage as S. fraasi and parts of the S. fraasi lineage as S. palustris. In addition, she calculated that there was an undescribed species intermediate in size between S. fraasi and S. ehingensis. On the same basis, material from the Headon Beds was identified as S. palustris and that from the Osborne Beds (Bosma & Insole 1976) as S. fraasi. Two different displays of the published data plus some new data are presented here in an attempt to test the credibility of the lineages and also to search for ways of identifying species other than by purely single parameters or by stratigraphy. Firstly (Text-fig. 38) histograms of the log of length x width of M'/* have been plotted in stratigraphical order. The record is not as dense as it is for part of the Wyoming early Eocene (see Gingerich 1976a) and more than one phylogenetic interpretation is possible. After the Grande Coupure most of the Bavarian local- ities contain a larger and a smaller species which show only a slight size increase with time. Of these the smaller one can readily be identified with S. fraasi on size. The larger, however, is considerably smaller than the type assemblage of S. ehingensis from the later (Antoingt Zone) Ehingen 1 locality. Where two species are present in Bavarian fissure fillings prior to the Grande Coupure, there is more size variation between localities and consequent overlap in total. Moreover, the smaller species is in most cases the more abundant, except in the last locality prior to the Grande Coupure (Weissenburg 2) where the larger has a wide observed range into which the one smaller specimen (doubtfully attributed by Schmidt-Kittler to S. fraasi) could possibly be incorporated. If the two post-Grand Coupure lineages could be shown to be derived from this single variable species, then the smaller one could no longer be identified as S. fraasi, whose type locality is slightly older. On the other hand, the Weissenburg 2 histogram is bimodal and the length measurements produce a coefficient of variation of 9-11, twice that of the unimodal assemblages with at least twenty specimens. It appears thus that two species are represented in the Weissenburg 2 assemblage, but more than just the smallest specimen should be referred to S. fraasi (see Schmidt-Kittler 1971a: 109). Overlap of measure- ments, however, means that it is impossible to identify all specimens. If more specimens from earlier pre-Grande Coupure Bavarian sites were known, they would probably overlap like those of Weissenburg 2. Nevertheless, the presence of the same two lineages from Weissenburg 8 until Ronheim | is supported by the length x width histograms. Turning now to the Isle of Wight assemblages, those from the Headon Beds were identified by Bosma (1974) as S. palustris. She nevertheless noted that on the single upper molar of this species in the GIU from the type locality, the mesostyle was hardly developed. The holotype and three topotype M!’’s in the IRSNB, now cleaned of wax, show a similar poor development of the mesostyle; so does GIU HB801. Schmidt-Kittler (1971a: 40) mentioned the presence of well-developed mesostyles in his assemblages and Bosma (1974: 40) noted that although vari- able the mesostyle was usually relatively high in her assemblages. Out of 74 M'/*s from Creechbarrow, only 3 have mesostyles as poorly developed as those of S. palustris. Specimens from the Osborne Beds of Lacey’s Farm Quarry, Isle of Wight, were identified by Bosma & Insole (1976) as S. fraasi on the basis of size. They are intermediate between the larger and smaller species present at Weissenburg 8, a locality probably only slightly younger than Lacey’s Farm Quarry. They overlap with the smaller species and probably would also overlap with the larger species if more specimens were known. The Lacey’s Farm Quarry assemblage appears a good candidate for potential common ancestry of both the larger and smaller lineages. Alterna- tively, S. minimus from Ehrenstein 1A could have given rise to the smaller species and the Lacey’s Farm Quarry assemblage to the larger species. Unfortunately little is known of S. minimus from any locality so that assessment of its position in the phylogeny is not yet possible. Either of the two above alternatives appear more plausible in the light of the Isle of Wight material, however, than Schmidt-Kittler’s (1971a: 119) idea that S. minimus was the common ancestor of both lineages. The placing of S. minimus in an ancestral position probably arose because he thought the holotype came from the Bartonian fissure of Eclepens-Gare. However, a label with the specimen in the LGM in Stehlin’s writing gives ‘Entreroches’, which is a later Ludian fissure. That this is the correct provenance is supported by the record of ‘une trés petite mandibule inférieure portant quatre molaires et une incisive’ belonging to a ‘Rongeur’ by Harpe & Gaudin (1854: 127). Ehingen1 Bernloch1 Burgmagerbein 2. Ronheim1 Schelklingen 1 Herrlingen 1 Ehingen 12 Hoogbutsel_--~ ; Ehrenstein 1B Weissenburg 2 Bernloch1A Mohren 6 Entreroches__ — Weissenbu rg 8 Ehrenstein 1A__ Lacey’s Farm Headon Hill 3-4 Totland Ba Headon Hill 2 Hordle Cliff MB Creechbarrow BARTONIAN MAMMALS OF HAMPSHIRE BASIN S.fraasi ——>— eel Ey els Switzerland 0.1 0.2 0.3 325 S. ehingensis ~ THIL oa Ptr. S| C1 S. Germany = = S. England alec Sbs_JO 10 ——S. authodon NE5 0 (Recs pe ee ee la ene a ee Cla malas es one 04 0.5 0.6 07 0.8 0.9 1.0 Log I xw Text-figure 38 Histograms of log. length x width in stratigraphic order of M‘? of species of Suevosciurus from localities in England, Switzerland, southern Germany and Belgium, spanning Bartonian to Stampian. For measurements and references to localities see Cray (1973), Bosma (1974), Bosma & Insole (1976), Insole (1972) and herein for England, Stehlin (1903) for Switzerland, Schmidt-Kittler (1971a) for southern Germany and Misonne (1957) for Belgium. The histogram of S. palustris (Misonne) is shown solid to indicate that it is morphologically separable from its contemporary S. fraasi Major from Ehrenstein 1B. The single plot of S. minimus (Major) from Entreroches is estimated from the unique M,, by squaring its length. Names of species arrowed indicate type assemblages or age of type assemblage (S. fraasi). 326 J. J. HOOKER The second combination of parameters is shown in Text-fig. 39. It plots length x width against length/width of DP*. The lowest length/width ratio is shared by the Creechbarrow species and the holotype of S. ehingensis, which can be distinguished easily on absolute size. The larger species from Ehrenstein 1B, Herrlingen 1 and Schelklingen 1 have higher ratios. The smaller species from Bernloch 1A, Ehrenstein 1B, Ehingen 12, Herrlingen 1, Schelklingen 1 and Ronheim | in total spans the ratio of the larger species but is less variable (possibly an artifact of lower numbers) in each locality; in the smaller species the assemblages with the smaller teeth have a higher length/width ratio than those with the larger teeth. Specimens from Lacey’s Farm Quarry fit easily within the range of S. fraasi. However, as no pre-Grande Coupure DP*s belonging to the larger Bavarian species are known, no concomi- tant contrast can be made. The teeth from the Headon Beds are similar in size to the smallest in the smaller Bavarian lineage from Ehrenstein 1B. Both were identified by Bosma as S. palustris. 1.0 A of 3 A w Text-figure 39 Scatter diagram of length x width against length/width in DP* of species of Suevo- sciurus. Key: x = Creechbarrow; + = HH2 and TB; V = HH3-4; W = Lacey’s Farm Quarry; 4~ = S. fraasi Major from Bernloch 1A; A = S. fraasi from Ehrenstein 1B; A = the larger lineage from Ehrenstein 1B; © =S. fraasi from Ehingen 12; [1] = S. fraasi from Herrlingen 1; = the larger lineage from Herrlingen 1; © = S. fraasi from Ronheim 1 and Schelklingen 1; @ = the larger lineage from Schelklingen 1; and << = S. ehingensis Dehm from Ehingen 1. For data sources see Text-fig. 38, p. 325. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 327 The Headon Beds teeth can be distinguished by their length/width ratios, which are interme- diate between those from Ehrenstein 1B and those from Creechbarrow, slightly overlapping both. Their stratigraphical trend is towards higher length/width ratio. In the light of the few Hordle Cliff specimens and the histograms shown here, Bosma’s (1974) claimed size increase through the Headon Beds does not appear to be a significant trend. On the basis of mesostyle size, neither the Headon Beds Suevosciurus nor the smallest members of the smaller Bavarian lineage deserves to be identified as S. palustris. Can Suevosciurus authodon be considered the common ancestor for all later species of the genus? It is the oldest and most primitive known. No assemblages with two species are known until Ehrenstein 1A. Without a better Bartonian record, it is not possible to judge whether or not S. authodon was actually the common ancestral species for Suevosciurus, but it nevertheless represents a good morphological ancestor. The absence of any authenticated records of Suevo- sciurus from non-British Bartonian sites suggests that it may have originated in the area of southern England and later spread to the rest of Europe. RELATIONSHIP WITH Treposciurus. Although Suevosciurus authodon has many of the characters of later Suevosciurus species, these are often less developed or occur in a smaller percentage of individuals in the assemblage. Moreover, its ability to replace DP4 with P¢ and the characters of this latter tooth closely resemble Treposciurus helveticus preecei. Close relationship is indi- cated and it is likely that they had a very recent common ancestry within the Bartonian. RECOMMENDATIONS ON APPLICATION OF CURRENT SPECIFIC NAMES. 1. S. ehingensis should be restricted to material from Ehingen 1, Bernloch 1B and Burgmagerbein 2, comprising very large teeth and a DP* with a length/width ratio of about 1. 2. S. palustris should be restricted to the small teeth from Hoogbutsel with upper molar mesostyles minute or missing. 3. S. fraasi should be used for the smaller of the two main Bavarian lineages, characterized by having DP* with a very prominent parastyle. 4. S. minimus should be used for the holotype M,_; from Extreroches, about which more information is needed before its status can be assessed, and the isolated molars from Ehrenstein 1A and Weissenburg 6. This leaves, as two definable unnamed species, members of the main larger Bavarian lineage (referred by Schmidt-Kittler 1971a to S. ehingensis) and the small teeth from the Headon Beds (referred by Bosma 1974 to S. palustris). More problematical are the teeth from Lacey’s Farm Quarry. More material from critical levels and in more geographical areas would improve our knowledge of the interrelationships within this complex genus, whose biostratigraphical poten- tial is high. Order APATOTHERIA Scott & Jepsen 1936 I follow Sigé (1975) and Russell et al. (1979) in recognizing this group as a distinct order. Its position herein adjacent to the Carnivora and distant from the Proteutheria reflects its relation- ships with the palaeoryctids as advocated by Szalay (1968), themselves showing specializations towards creodonts, carnivores and condylarths (see Van Valen, 1966). It comprises only one family unless Aethomylos Novacek 1976 is an apatothere (see Novacek, 1976: 40-44, figs 16-19). Family APATEMYIDAE Matthew 1909 TYPE GENUS. Apatemys Marsh 1872. INCLUDED GENERA. Jepsenella Simpson 1940; Labidolemur Matthew & Granger 1921; Stehlin- ella Matthew 1929c; Sinclairella Jepsen 1934; Eochiromys Teilhard 1927; Heterohyus Gervais 1852. RANGE. Middle Palaeocene—Oligocene, North America; late Palaeocene—late Eocene, Europe. DiaGnosis. See McKenna (1963: 13). = é 7 é é = ae = “WW episeq *W puryeq snwie1 SUIpUusoSY BZ = é = é é . 1F = dap AO TTEYS ajqipueus jo yidoq = LZ - é — b é - + — opIM MOJICU IOSIOUI IOMOT 97 - 4 — é é — - +/— sof ou Ajjeoong SurIsjnq pryjnuosodAy Fw sz , TUO|e} a . a © e a * vy 2s oS Zurpusixs aqo| Angee QW PZ + — — é é = - — Fal 0 (0) yoys yysua] dsno pynuosoddhy Ww €Z + = 1 é d = = aH = ysry MO] ones ysug] “W:*W 7 _ _ — é + + — + anbijgo OSIOASUBI} pysiojsod PW. IZ + Ma - + Suo1}s yeom YISUIS OVWIP We OT ar = = é = = ofr os juosoid yuosqe quil] pystiovied [eIsow J. 6] + = = é + = ar iar SuOoI}s yeom pluoovied WOT bar ar AF é = = ar ap SuoI}s yeom pluosojus FL oF ar SF é = = + = y0Ys suo] Yj8ug] pruosin FOOT — = = é = af - = suo] y10ys yjsua] pruoyey FWOS i — ~ i + ~ _ _ so ou yojou pruosojua @ WSOP] é — — é — - + - SoA ou pynuosoddy jensuy “WeSC — — — + + + - fi 3Uu01}S yerom eystioodcyjsod FWeSFZI + — - é BuOol}s yeom yysuasjs vysti90}01djsod FSO | - — ~ + + = _ i ou so B}SII00}01djsod JFSOOT . 4 = a in 5 aS ; eae ae auosodAy ¢ Sululof winynZu1s0j01djsod FES 6 + + = — — — — fi 93 IP] [jews suosodAy Fg — — — — + — é suo]s yeom yysues eystiorjounjsod - WL - — - - — _ + é juasqe juosoid Rysiiovjourjsod - JW 9 = = ap = ar + db é doop Mo][eys snxayowe -W OS — — + + ii + é é ou sok juares Ayjeisout a[Ayseied -We of — é 4 fi éi yeom suo]s yisus}s OWIP WN € + + = + é — é é ou sox gfAjsered juorfes AyJeoong 2 Z + é + é + é — é suols yeom yiysuens aAjseied |W Sel = = Sel x 2a eo] = poaourape oatqtuntid JooeIvyD S = a De < =. 2 < s Suse os = i ek: E Re & E : g Sap ud an a “4, AQ PaIVOIPUT SI IOJOBVIYO B UO UOPUIIOJUI JO YOR] ‘SIV[OW JMO] SA[OAUI CZ 0} E] SoJORIeYO ‘sipjour Jaddn aAjoAut Z] 0} J siaj}OvIeYD “aepikulayedy ay} JO 3so1 Oy} YIM UOSTIeduIOD dnoi3jno uO paseg (+) psouURApPe JO (—) dATITUNIId se pa}eUT}sa oIe SIDJOVILYD 1941930} pasapisuod aie Kay} ‘satoads 19y}0 ay} 10J UMOUYUN AT]eN}ZITA S19}9vILYO ‘IfOaATe UO paseq *g pue *q JO SOZIS 9AI}R[91 BY} }d90xa SIo}OvIBYO [je ut jeonuapt svodde sijioni6 “YY pue iapjafnay “FY sy ‘sndyousajazy Jo satoads uvisoid,-jsod yuaJayIp ay} ut (azis ydaoxa) s1ajovIeYS JO UOTINGINSIG ZZ AGEL 328 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 329 Genus HETEROHYUS Gervais 1852 [incl. Necrosorex Filhol 1890c, Amphichiromys and Heterochiromys Stehlin 1916, Heterohyus (Chardinyus) and H. (Gardonyus) Sigé 1975] TYPE SPECIES. H. armatus Gervais 1852. Middle Lutetian, Bouxwiller, France. INCLUDED SPECIES. H. europaeus (Stehlin 1916) Teilhard 1922; H. gracilis (Stehlin 1916) Teilhard 1922; H. heufelderi Heller 1930 (doubtfully distinct from H. gracilis, but see discussion in Russell et al., 1979: 226); H. sudrei Sige 1975; H. quercyi (Filhol 1890c) Teilhard 1922; H. nanus Teilhard 1922; H. morinionensis sp. nov.; H. sp. from Malpérié and La Bouffie (see Sige 1975: 665); H. sp. from Hordle Cliff (see Cray 1973: 63-64); H. sp. from Mutigny; H. spp. 1, 2, 3 from Avenay; H. spp. [1, 2] from Grauves and Cuis (see Russell et al., 1979: 227-231). RANGE. Ypresian—Ludian, France; late Lutetian, Switzerland; early Lutetian, D.D.R.; Marinesian—early Ludian, England. DiaGnosis. See Sigé (1975: 656) for the most recent formal diagnosis; but see Russell et al. (1979: 222-226) for comprehensive discussion of generic characters. Discussion. Species of this genus, like the others in the Apatemyidae, are rare faunal elements and are known mainly from scattered, isolated teeth. A few jaws are known but the only specimen with upper and lower teeth associated (referred specimen of H. quercyi) is lost (fide Sigé 1975: 658). There are thus often problems in associating both different tooth types in a single row and uppers and lowers of the same species. Occlusal studies have not been published. Table 22 shows the distribution of characters in the different post-Ypresian species (named and unnamed). Various attempts have been made in the past to divide the genus Heterohyus into different genera or subgenera (see synonymy list above, Sigé 1975 and Russell et al. 1979). The table shows that most of the available characters have a rather random distribution with few congruencies. The position is further complicated by the incompleteness of different taxa and the consequent poor overlap of parts known. Small numbers of specimens restrict the knowledge of intraspecific variation. However, some such variation can be seen in Table 22 in characters 22, 24, and 25. Neither a phenetic comparison, using all the characters, nor a cladistic one using only advanced characters (giving shared character states as percentages of the shared characters in both cases) give reliable results (see Table 23). The only exceptions are the very high dissimi- Table 23. Number of character states in common between the different species of Heterohyus. Numbers are percentages of the total characters shared. Above the oblique line only advanced character states are used; below it all the character states are used. Characters are from Table 22. 0 az) as) = 2 2&8 8 8 g awe & 7 = = Sos 8 = e 8 o Bt SaaS ee Sa ici? asin cr 3 v Sa a a = = > Se) acjee gs se = se) se) = H. armatus 25 20 17 36 H. europaeus 1 14 12 20 H. gracilis and H. heufelderi H. sudrei H. sp., Malpérie H. nanus H. morinionensis H. quercyi 330 J. J. HOOKER larities of H. gracilis/heufelderi from H. europaeus and H. quercyi and of H. morinionensis from H. gracilis/heufelderi and H. sudrei. The only high similarities are H. quercyi with H. armatus and the H. sp. from Malpérié with H. gracilis/heufelderi, both artificially increased by the reduced number of shared anatomical parts known: H. armatus is only known from lower teeth, H. sp. from Malpérié only from one M?. A further complication is that the upper dentition attributed to H. quercyi may not belong to this species. Future collecting in Europe may well allow more characters to be added for different species, thus helping to clarify relationships. Heterohyus cf. sudrei Sigé 1975 (Pl. 19, fig. 7; Text-fig. 40E; Tables 22, 23) Hototyre of H. sudrei. Left M*? (UM RBNS5170). Lower Calcaire de Fons (Marinesian), Robiac-Nord, France. RANGE of H. sudrei. Marinesian to early Ludian, France. DiaGnosis of H. sudrei. See Sige (1975: 657). MATERIAL. Left M,,, talonid fragment (M35707). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION. This specimen cannot be included in the best represented species from Creech- barrow (H. morinionensis) as it is slightly smaller; it has an entoconid notch, a slightly lower entoconid, and the buccal hypoconid wall is more vertical, meeting the postcristid at an acute angle when viewed distally. This compares well with H. sudrei (see Text-fig. 40E-F). In H. morinionensis and H. nanus the angle is 90° or more (Text-fig. 40A—D). The rather longitudinal orientation of the cristid obliqua and its apparent buccal position of attachment to the trigonid suggests that the tooth is more likely to be an M, than an M,j. It has a large distal interstitial facet. Wear on the crown is slight. Just lingual of the midpoint of the postcristid on the mesial side is the faint swelling of a hypoconulid. There is a well-marked valley between this and the entoconid. Because of lingual breakage no width measurement is possible. However, it is slightly smaller than the M,s of H. sudrei figured by Sigé (1975: pl. 2, figs 1—2). Heterohyus aff. nanus Teilhard 1922 (Pl. 19, fig. 6; Text-fig. 40C; Table 24) v. 1980 Heterohyus sp. 2; Hooker & Insole: 40. Ho.otyre of H. nanus. Left mandibular ramus with incisor, P,, M, ; (MNHN Qu8732). Phosphorites, Mémerlein, France. RANGE of H. nanus. Marinesian to Ludian, France. D1AGnosis of H. nanus. See Sigé (1975: 659). MATERIAL. Left M, (M35709). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION AND DISCUSSION. The tooth is unworn but broken distolingually so that the entoconid is now missing. The angle that the postcristid makes with the buccal hypoconid wall, however, is obtuse, suggesting that the entoconid was developed as in H. nanus (i.e. larger than in H. sudrei). There is a small paraconid and the trigonid is open mesially, the mesial limb of the paracristid being absent, its place being taken by a mesial cingulum. Both trigonid and talonid are relatively short and the whole tooth is thus not much longer than wide. All these characters suggest strong affinities with H. nanus. It is, however, smaller (Table 24), being closer in size to M, of H. sp. from Malpérié and La Bouffie (Sigé 1975: 665; pl. 1, figs 14-15). There appears to be a slight size increase with time for M', M* and M, from the BARTONIAN MAMMALS OF HAMPSHIRE BASIN 331 Table 24 Length (1) and trigonid (w,) and talonid (w,) width measurements of teeth of Heterohyus from Creechbarrow. Two width measurements are only given for lower molariform teeth. Measurements in millimetres. No. Tooth l Ww, W> H. morinionensis: M35702 M! ~ 2:65 M35703 M? 2°85 _ M35400 M? 2-35 (3-50) M35705 M, 2:40 2:00 M35706 M,; - = H. aff. nanus: M35709 M, 1:70 1:50 — Robiac assemblage to those of Malpérie and Perriere (Sigé 1975: text-fig. 3); the Creechbarrow tooth might represent an earlier stage in this trend, but this can only be corroborated by the finding of more teeth and tooth types. Heterohyus morinionensis sp. nov. (Pl. 19, figs 1-5; Text-fig. 40D; Tables 22-24) v. 1977b Heterohyus (Chardinyus) sp.; Hooker: 141. v. 1980 Heterohyus sp. 1; Hooker & Insole: 40. Name. Morinio, a Roman station in the present borough of Wareham, near Creechbarrow. Hotorype. Left M?, M35400. PI. 19, fig. 3. ParatyPes. Left M' (M35702); right M? (M35703); right M? (M37119); lingual fragment of left M? (M35704); right M, (M35705); left M, talonid fragment (M35706). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DiaGnosis. Medium-small species of Heterohyus (M? length 2-85mm; see Table 24). M? with parastyle mesially but not buccally salient, and ectoflexus shallow. M' * with weak postmeta- crista. Upper molars with large, well differentiated hypocone with weak posthypocrista and weak postprotocrista. M, without hypoconulid. M, _, without entoconid notch. Lower molars with short trigonid and talonid; strong entoconid; weak paraconid; no mesial paracristid limb; and transverse postcristid. M;/M, length ratio low. M, with short hypoconulid on short hypoconulid lobe, bulging buccally. DESCRIPTION. M!: The postprotocingulum is weak and the large, lingually salient hypocone is separated from it by a notch (PI. 19, fig. 1). The parastyle and metastyle are broken away but the ectoflexus appears to have been at best very shallow. There is very weak postflexus. The distal cingulum peters out lingually but a continuation down the distal side of the hypocone takes the form of faint papillae. M?: This tooth is broken lingually and has a deep postflexus (PI. 19, fig. 2). The metacone is formed lingually into a strong ridge confluent with the postprotocrista. There is a moderate ectoflexus. M2: On the holotype the parastyle is broken at its tip but when complete would have been very prominent buccally (Pl. 19, fig. 3). The postprotocrista appears very weak but may have been largely removed by wear. The two other fragmentary specimens show a constancy for the distinctness of the hypocone from the protocone, although the hypocone is slightly smaller in M35704 than in the other two specimens. M37119 is less worn than the holotype but shows the postprotocrista to be absent, a sharp even valley linking the trigon and talon basins. M,: The paraconid is broken away and the edge of the talonid notch, metaconid and entoconid slightly chipped (PI. 19, fig. 4). Otherwise the tooth is well preserved, moderately 332 J. J. HOOKER BARTONIAN MAMMALS OF HAMPSHIRE BASIN 333 worn with well-defined wear facets. The entoconid is distally positioned, confluent with the postcristid and there is no hypoconulid. A weak lingual rib on the protoconid half crosses the trigonid basin. There is a weak mesial cingulum but no mesial paracristid arm, so that the trigonid basin is open mesially. M,: This talonid fragment is truncated through the talonid notch and just behind the tip of the hypoconid (PI. 19, fig. 5). Its most distinctive feature is the short buccal hypoconulid lobe. The hypoconulid appears to project buccally slightly beyond the hypoconid. The hypoconulid is like that in H. quercyi in one respect, that it is slightly constricted, unlike H. nanus where it merges imperceptibly in occlusal view with the rest of the talonid. Shape of the talonid appears to be variable (see discussion below). Discussion. When all characters are considered (including size), H. morinionensis is most similar to H. quercyi. This is, however, reliant on the correct attribution of the lost cranium and mandible figured by Teilhard (1922: text-fig. 39; pl. 4, figs 16-18). One must therefore first consider the likelihood of the specific association of this specimen with the holotype mandible with M,. In all the sufficiently known species of Heterohyus except H. nanus and H. quercyi there is a correlation between the upper molar postprotocrista and the lower molar paraconid. If the former is present the latter is weak; if the former is absent, the latter is strong. In H. nanus, lower molars with weak paraconids are ‘associated’ with upper molars without postproto- cristae. In H. quercyi, M, with a strong paraconid is apparently associated with upper molars with postprotocristae. Sigé (1975: pl. 1, figs 6-7) figured an M' and M? from Aubrelong 2 which he considered close in size and morphology to H. quercyi, but which he placed with some doubt in H. sudrei. His first identification as H. quercyi? is tentatively accepted here. These teeth are very similar to those of H. morinionensis, differing only in being slightly larger, having slightly stronger post- protocristae and shallower ectoflexus in M?. H. morinionensis upper molars occur at Creechbarrow with three different types of lower molars. Two of these have already been identified as H. cf. sudrei and H. aff. nanus, both being smaller than the third type, which are of appropriate size for the H. morinionensis upper molars. Furthermore, the M, of the third type (M35705) has a deeply grooved oblique buccal phase facet on the distal hypoconid wall (see Text-fig. 40D), which is considerably stronger and more oblique than the same facet on otherwise similar lower molars of H. nanus. It is most likely to have been caused by an M? metacone strengthened by the buccal end of a postprotocrista. Although the trigonid of M, is unknown in H. morinionensis, judged from a mandibular ramus of H. nanus (Sigé 1975: pl. 2, fig. 8a—c) the paraconid alters little in strength from M, to M;. The M, paraconid in H. morinionensis is thus likely to have been considerably smaller than that of the holotype of H. quercyi. The quality of Teilhard’s (1922) retouched photographs of the referred cranium and mandible of H. quercyi is poor and the exact details of the teeth are confusing, the different scale photographs being partly contradictory. The lower teeth are mainly obscured, but M, (his pl. 4, fig. 17) does appear to have a strong paraconid, thus supporting his identification. It is possible Plate 19 Teeth of Heterohyus and Lophiotherium from Creechbarrow. Figs 1-5 Heterohyus morinionensis sp. nov. Scanning electron micrographs of occlusal views of molars, x 16. Fig. 1, left M' (M35702). Fig. 2, right M? (reversed) (M35703). Fig. 3, holotype left M* (M35400). Fig. 4, right M, (reversed) (M3570S). Fig. 5, left M, talonid fragment (M35706). See p. 331. Fig.6 Heterohyus aff. nanus Teilhard. Scanning electron micrograph of occlusal view of left M,, x 16 (M35709). See p. 330. Fig. 7 Heterohyus cf. sudrei Sigé. Scanning electron micrograph of occlusal view of left M,,, talonid fragment, x 16 (M35707). See p. 330. Figs 8-9 Lophiotherium siderolithicum (Pictet). Light macrographs of lower teeth, x 3. Fig. 8a-c, associated left P,, M>_3; a, buccal, b, occlusal and c, lingual views (M37705). Fig. 9, lingual view of right lower canine (reversed) (M36828). See p. 347. 334 J. J. HOOKER that the strengthening of the lower molar paraconid plus paracristid is a trend, independent of that occurring earlier in H. armatus and H. europaeus, and associated with similar strength- ening of the upper molar postprotocrista and with a slight modification of the occlusal relation- ships of the upper and lower molars. H. morinionensis also appears to differ from H. quercyi in having a buccally salient M? parastyle. The possibility cannot be excluded that this was broken in Teilhard’s figured speci- men. Two other characters, depth of M? ectoflexus and shape of the M, hypoconulid, may be intraspecifically variable. The first is supported by two M/?s of an undescribed species from Hordle Cliff, where the difference in depth is at least as great as between the Creechbarrow and Aubrelong 2 specimens. The second is supported by two specimens of H. armatus from Boux- willer in the NMB (Bchs490 and Bchs306): the former is like the holotype (see Stehlin 1916: text-fig. 346b) where the hypoconulid is buccally situated and projecting and the lobe is short; the latter has a much longer median hypoconulid lobe. Given the above conclusions, H. morinionensis is closely related to H. quercyi and is a good candidate both morphologically and stratigraphically to be its ancestor. H. morinionensis has no known specialized characters which would prevent such a possibility. Heterohyus spp. indet. (Text-fig. 40G—W) MATERIAL. Six incisor fragments (M35699—700, M35708, M37116-8). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION. These specimens show size differences, and some confusing morphological differ- ences at least some of which are ontogenetic. M35708 is much smaller than any of the others and almost certainly belongs to H. aff. nanus as represented by the M,. It is rolled and the distal enamel band is little wider than the mesial one. The former was probably once wider, being worn away parallel to the buccal edge by post-mortem abrasion (Text-fig. 40U—W). All the other incisor fragments are essentially the same size and presumably belong to the commoner H. morinionensis or to H. cf. sudrei, or both. The morphology of Heterohyus incisors has not been covered in much detail by authors and it is possible that they are not distinctive at species level. H. sudrei has no incisors attributed to it, whereas they are known for most of the other named species. The Creechbarrow incisors vary in width of the distal enamel band and cross-sectional shape. M35700 is the largest fragment. It is truncated at both ends but can be oritntated by the apical narrowing of the pulp cavity. The enamel is slightly rugose, the distal enamel band is wide and the mesial enamel band concave. Apically there is a sharp mesiobuccal angle, but passing basally this angle becomes slightly rounded and begins to migrate lingually, thus making the buccal edge more evenly rounded and narrowing the mesial concavity. At this point, the mesial enamel has a noticeably pitted surface (Text-fig. 40K—N). M35699 is a small fragment which appears to correspond in position to the apical end of M35700. In M37117, the buccal edge, although rather eroded, appears to have been rounded without a sharp mesiobuccal edge; the mesial enamel band is flat; and the distal enamel band is about two-thirds the width of that of M35700. In incisors of H. nanus and H. quercyi, at least, the distal enamel band narrows in a basal direction. It thus seems reasonable to suggest that M37117 represents a more basal region than M35700 (Text-fig. 40O—P). M37116 is rather eroded but has a complete layer of cement, is a thin oval in cross section and tapers basally. It is thus considered to be a root fragment. Basally the pulp cavity is open but is reduced apically to a small lumen and must thus represent an old individual (Text-fig. 40Q-T). M37118 is a crown fragment which differs significantly from the others (Text-fig. 40G—J). It preserves the naturally worn tip and its mesial enamel band is convex, not concave, for most of Text-figure 40 Heterohyus teeth. A-F, distal views of lower preultimate molars, x8. A, right M, (reversed) (UM MPR4) of H. nanus Teilhard from Malpérie; B, right M, (reversed) (UM RBN6004) of H. nanus from Robiac; C, left M, (M35709) of H. aff. nanus from Creechbarrow; D, right M, (reversed) (M35705) of H. morinionensis sp. noy. from Creechbarrow (f = facet caused by M7? metacone); E, left M,,. (M35707) of H. cf. sudrei from Creechbarrow; F, right M, (reversed) (P. Louis collection ROB330) of H. sudrei Sigé from Robiac. (A, B and F drawn from casts). G—W, incisors of H. spp. indet. from Creechbarrow, x4. G—J, ?upper right incisor crown fragment (reversed) (M37118); K—N, adapical crown fragment of lower left incisor (M35700); O-P, adbasal crown fragment of lower incisor (M37117); Q—T, root fragment of right lower incisor (reversed) (M37116); U-W, crown fragment of lower incisor (M35708). G, K, O, Q and U are distal views; H, L, P, R and V are mesial views. I, J, M, N, S, T and W are transverse sections, shown as if of left teeth and viewed basally, so that distal is to the left and mesial to the right; I, M and S are from apical ends, J, N and T from basal ends; enamel is shown as solid black and cement is delimited by a double line. 336 J. J. HOOKER its length. It could represent a different taxon. However, the occlusobuccal angle of the worn tip is much less acute than for lower incisors (see Teilhard 1922: text-figs 40b and 41; Stehlin 1916: text-figs 340, 349; Heller 1930: pl. 5, fig. 4a—c; and Sigé 1975: text-fig. 1b—c), as it is in rodents. Also the tooth is convex mesially as in the upper incisor (‘I'?’) of Apatemys figured by Russell et al. (1979: pl. 1, fig. 10). These facts support it being an upper incisor (I'), in contrast to all the other fragments which are lowers (I,). Upper incisors of only two specimens of Heterohyus have been described and figured. One is that of the lost skull of H. quercyi; the other is an isolated I' cusp tip fragment from Mutigny (Russell et al. 1979: 227; pl 3, fig. 8). The latter was provisionally attributed on the basis of differences from those belonging to Apatemys from the same locality. Whereas the Mutigny I'? has a complete enamel crown and is bicuspid like Apatemys, that of H. quercyi is, according to Teilhard, ‘sciuroide, émaillée sur sa face externe’. Under these circumstances, it seems unlikely that the I’ of H. quercyi was bicuspid at least during most of its life. M37118 at its basal end is beginning to develop a sharp mesiobuccal edge. Like the lower incisors the enamel is slightly rugose. The worn tip is obliquely bevelled, wear being more intense mesially than distally. If the different types of lower incisor fragment indeed represent successive segments, the minimum length for this tooth measured along the curvature would be 30mm. A more realistic estimate, based on more gradual narrowing of the distal enamel band, would be 40mm. From the size of the pulp cavity in M35700 compared to M37116, it is likely that the root remained open until a fairly advanced wear stage. Order CARNIVORA Bowdich 1821 Superfamily MIACOIDEA Cope 1880 (rank emend. Simpson 1931) Family MIACIDAE Cope 1880 Subfamily MIACINAE Cope 1880 (rank emend. Trouessart 1885) TYPE GENUS. Miacis Cope 1872. INCLUDED GENERA. Uintacyon Leidy 1872; Tapocyon Stock 1934; Vulpavus Marsh 1871; Vass- acyon Matthew 1909; Oodectes Wortman 1901; Paroodectes Springhorn 1980; Palaearctonyx Matthew 1909; Pleurocyon Peterson 1919. RANGE. Eocene, North America and Europe; Upper Oligocene, Asia? DIAGNosIs. See Matthew (1909: 345). COMMENT. Miacines are generally rare in European Eocene faunas. They have been described mainly by Filhol (1876a), Teilhard (1915), Guth (1964), Beaumont (1966), Quinet (1968), Rich (1971) and Springhorn (1980, 1982). Flynn & Galiano (1982) have shown the Miacoidea and Miacidae to be paraphyletic, but have not resolved the interrelationships of the now caniform “Miacinae’, although they have added other genera. The old classification is thus retained as a conservative measure. Genus MIACIS Cope 1873 ?Miacis sp. indet. (Text-fig. 41A—O) v. 1977b Miacidae indet.; Hooker: 141. vp. 1980 Mluacidae indet.; Hooker & Insole: 41. MATERIAL. Right P* (M36206); right and left P*s broken mesially (M36207-8); right P, (M36202); three left M, trigonid fragments (M36210—1, M37568); right M, (M37569). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 337 ay Text-figure 41 Miacinae teeth. A-O, ?Miacis sp. indet. A—-C, left M, trigonid fragment (M37568); D-F, right M, (reversed) (M37569); G-I, right P, (reversed) (M36202); J-L, right P? (reversed) (M36206); M—O, right P* (reversed) (M36207). P—U, Miacinae indet. P-R, Left P?? (M36204); S—U, right P,? (reversed) (M36200). A, D, G, J, M, R and S are occlusal views. B, E, H, K, N, P and U are buccal views. C, F, I, L, O, Q and T are lingual views. All are from Creechbarrow; x 6:6. 338 J. J. HOOKER DESCRIPTION. The association of these few broken isolated teeth is arbitrary. The size ratios of the different tooth types fit Miacis, but depend on this association. The form is very similar in size to Miacis exilis (Filhol 1876b) Teilhard 1915 from the Phosphorites du Quercy. P?: This tooth is 2°7 mm long and 1-2 mm wide. It fits Guth’s (1964: 360) description of that of M. exilis in the complete cingulum and its development distally into a metastyle (Text-fig. 41J—L). The preparacrista curves lingually to join the lingual cingulum halfway between the paracone and the mesial tooth margin. A faint parastyle is represented by a slight cingular swelling. P*: M36208 only preserves the metastylar wing. M36207 only lacks the protocone and parastyle, if there was one (Text-fig. 41M-—O). The cingulum is strong and continuous and the enamel has faint wrinkles which are orientated vertically. M36207 appears to have a longer metastylar wing than do the miacine P*s from the Sparnacian of Dormaal figured by Quinet (1968: pl. 6, figs S—9) and also M. exilis (Guth 1964: pl. 15, top left fig.). It is similar to a miacine P* from Mormont (LGM 40937) but here the buccal cingulum is interrupted at the paracone. P,: This tooth is 1-8 mm wide. It is broken in the parastylid region (Text-fig. 41G—I). The cingulum appears to have been complete buccally but incomplete lingually. The small hypo- conid is low and distal on the distal protoconid crest and salient only lingually; it is, however, larger than on any of the M. exilis P4s figured by Teilhard (1915: 112, text-fig. 2) or by Guth (1964: pl. 15, top right fig.). M,: Of the three trigonid fragments, two are too corroded or broken to be worthy of description. M37568, however, is better preserved but has the metaconid broken (Text-fig. 41A-C). It is 2-7 mm wide. The buccal cingulum extends for only a short distance between the paraconid and protoconid. The paraconid has a bluntly keeled mesial edge. The distal edge of the trigonid is nearly vertical and the metaconid was probably as high or higher than the paraconid. In occlusal view the mesiolingual corner is acute, resembling in this way Guth’s (1964: pl. 16) top right rather than top central figure of M. exilis. This feature thus appears to be of only individual significance. M,: This tooth is broken and corroded basally, especially round the talonid and in the trigonid and talonid basins (Text-fig. 41D-—F). It nevertheless shows much lower trigonid cusps than the M,, the paraconid and metaconid being subequal. A prominent cingular spur occurs on the mesiobuccal corner. Of Guth’s (1964: pl. 16) occlusal views of M. exilis lower molars, in cusp pattern M37569 most closely resembles the top left figure; the other two have a more oblique arrangement. It is estimated to have been 3-2 mm long and 2:4 mm wide. DISCUSSION. Too little is known of this form to be more precise about its affinities, but it testifies to the presence of at least one stoat-sized carnivore in the Creechbarrow fauna. ?Miacinae, gen. et sp. indet. (Text-fig. 41P—U) MATERIAL. Left P?? (M36204); right P? (M36200). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. Discussion. These rather undiagnostic teeth indicate the probable presence of a second miacine in the Creechbarrow fauna, a little more than half the size of ?Miacis sp. indet. The left P?? is estimated at 1-1 mm wide. The right P,? is 0-8 mm wide. Order CONDYLARTHRA Cope 1881 Family PAROXYCLAENIDAE Weitzel 1933 TYPE GENUS. Paroxyclaenus Teilhard 1922. INCLUDED GENERA. Kopidodon Weitzel 1933; Vulpavoides Matthes 1952 (including Russellites Van Valen 1965); Pugiodens Matthes 1952 (doubtfully distinct from Vulpavoides); Spaniella Crusafont-Pairo & Russell 1967; ?Kochictis Kretzoi 1943. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 339 RANGE. Sparnacian to Bartonian, possibly to Ludian or even to Chattian (Kochictis), Europe. The Asian record based on Dulcidon Van Valen 1965 has been removed by Tobien (1969: 30-31). DIAGNosIs. See Van Valen (1965: 389). Discussion. This small, rare, endemic European family numbers less than twenty published specimens. Fortunately three of the genera are known from skulls with nearly complete denti- tions and in addition there are several skeletons from Messel. The most recent detailed studies are by Tobien (1969), Rich (1971) and Koenigswald (1983). The small number of specimens means that the degree of intraspecific variation is virtually unknown in most cases, with consequent disagreement among authors on intrageneric variation also (see below). Van Valen’s (1965, 1967) rank of family is followed here. Tobien (1969) ranked it as a subfamily of the Arctocyonidae. Ordinal affinity follows Russell & McKenna (1962) and Tobien (1969), rather than Van Valen (1965) who placed them with a query in the Insectivora. Generic content essentially follows Tobien (1969) except that Russellites is synonymized with Vulpavoides. Genus VULPAVOIDES Matthes 1952 [incl. Russellites Van Valen 1965, ? = Pugiodens Matthes 1952] TYPE SPECIES. V. germanica Matthes 1952. Early Lutetian, Geiseltal, D.D.R. (? = Pugiodens mirus Matthes 1952; see Van Valen 1965, Tobien 1969). INCLUDED SPECIES. V. simplicidens (Van Valen 1965) comb. nov.; V. cooperi sp. nov. RANGE. Early Lutetian, D.D.R.; middle Lutetian, France; late Lutetian, Switzerland; and Mari- nesian, England. EMENDED DIAGNOSIS (modified from Van Valen, 1965: 392). Upper molars relatively short (length from 58-75% of width) with: cusps moderately high (M’ paracone height from crown base from c. 90-100% of tooth length); paracone and metacone relatively close together; postmetaconule crista weak to absent; paraconule weak to absent; metaconule weak; postpro- tocrista confluent with metaconule; strong centrocrista not strongly indented buccally or lin- gually; preparaconule crista absent; ectoflexus weak to moderate. P? * with: metacone present; and lingual half much shorter than buccal half. Note that the premolar characters are known only for the type species. Discussion. Van Valen (1965: 390) synonymized Vulpavoides germanica (based on a crushed cranium with upper dentition) with Pugiodens mirus (based on a mandibular ramus with lower dentition) on similarity of occlusal relationships. The choice of the latter as the senior synonym was probably to avoid the suggestion of unjustified miacid relationships implied by the alterna- tive Vulpavoides (i.e. like Vulpavus). It was unfortunate, however, as the mandible was already lost (Van Valen 1965: 391) and could not be located when I visited Halle in 1979. Tobien (1969: 28, 35) gave reasons for a certain amount of doubt regarding the synonymy of the two genera and preferred to use Vulpavoides as the senior synonym. If the synonymy is accepted, then Van Valen should be followed as ‘first reviser’ according to the ICZN rules (1985: 53). Vulpavoides is used here because of doubt over synonymy and the present restriction of comparisons to the upper teeth. Van Valen (1965: 392) considered his new genus Russellites to be the one most closely related to Pugiodens (his sense), yet distinct. He later (1967: 259) synonymized it without comment with Pugiodens. Tobien (1969: 29) considered it distinct and to share a number of characters with Kopidodon not shared with Vulpavoides: i.e. P* not transversely elongated; M? outline identi- cal; M? distobuccal corner not rounded and mesiobuccal corner projecting; M? buccal edge perpendicular to transverse axis; and less reduction of M?. However, the P* character based on the roots does not necessarily indicate a similarity with the very characteristic massive crown without metacone in Kopidodon. Also, the orientation of the buccal edge to the transverse axis shows some individual variation. In the holotype M' of 340 J. J. HOOKER V. germanica it is perpendicular. In the M! figured by Heller (1930: pl. 1, fig. 3a—b), it is oblique, the mesiobuccal angle being obtuse. In the holotype M’ of ‘R.’ simplicidens, the buccal edge is perpendicular, but in NMB Bchs656 (M') from Bouxwiller, it is oblique, the distobuccal angle being obtuse. Furthermore the M? outlines in Kopidodon and Russellites are not exactly identi- cal. The remaining characters can be more than matched by the following which are shared between Russellites and Vulpavoides but not with Kopidodon: upper molars with postmetacrista low, outline short and broad with paracone and metacone consequently closer together, strong centrocrista not strongly indented buccally or lingually and preparaconule crista absent. Vulpavoides cooperi sp. nov. (Text-fig. 42A—D, I-K) v. 1977b Russellites sp.; Hooker: 141. v. 1980 Pugiodens sp.; Hooker & Insole: 42. Name. After Mr J. Cooper, for help with field work. Hotorype. Right M', M37570. Text-fig. 42A—D. DOUBTFULLY REFERRED SPECIMEN. Left DP*? fragment (M35646). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. D1AGNosis. Small Vulpavoides (M' 3-0 mm long x 5:2 mm wide); M! relatively short and broad (length 58% of width); with no paraconule; paracone and metacone close together; high preprotocrista joining paracone lingually; crestiform metaconule forming buccal end of post- protocrista; no postmetaconule crista; sloping height of slightly worn paracone equal to the tooth length. Ectocingulum and ectoflexus weak. DIFFERENTIAL DIAGNOSIS. V. germanica Matthes is larger with stronger upper molar ecto- cingulum and deeper ectoflexus. V. germanica and V. simplicidens (Van Valen) have relatively longer M! with lower, more widely spaced paracone and metacone, cuspate paraconule and metaconule present; postmetaconule crista low; and low preprotocrista reaching only to paraconule. DESCRIPTION AND DISCUSSION. M37570 is considered to be an M’ as it has an oblique buccal edge with an obtuse mesiobuccal angle. M*s and M?s have acute mesiobuccal angles. There are no sharply delimited wear facets but dentine is exposed on the protocone, meta- conule, paracone and metacone and also in the trigon basin and extensively over the stylar shelf, the latter merging with that on the paracone and metacone. A buccal phase facet on the distal side of the metacone suggests an angle of shear with the M, trigonid of about 45°. Similar smooth enamel surfaces but with no definable edges occur on the mesial side ‘of the metacone and on both sides of the paracone. The holotype is compared in Text-fig. 42 with the referred M' of V. simplicidens from Bouxwiller. The weak ectocingulum is similar in both and contrasts with the stronger one in V. germanica. The obliquity of the protocone is also like V. simplicidens, but M' of V. germanica is too worn to compare for this character. One of the most striking features of M' of V. cooperi is the partial zalambdodonty. This is not so much the result of independent drawing together of the paracone and metacone, as rather an accommodation response of these cusps to extreme differential shortening of the tooth as a whole. Nevertheless, the mesocylix (the occlusal cavity within an upper molariform tooth) is very shallow buccally and must have occluded with an M, with a very reduced talonid. In the Eutheria, partial to complete zalambdodonty is common in palaeoryctoids, creodonts, various lipotyphlan families (see Van Valen 1966, Butler 1972), anagalids and poss- ibly dinocerates (Matthew 1928: 970). In the Condylarthra it appears otherwise to occur in the Mesonychidae. No milk teeth have hitherto been recorded for paroxyclaenids, but M35646 may be a DP* as it has relatively thinner enamel than M37570 but a somewhat similar cusp pattern (Text-fig. 42I-K). Only the lingual half is preserved, but the protocone has the same oblique orientation BARTONIAN MAMMALS OF HAMPSHIRE BASIN 341 DY ON, Text-figure 42 Vulpavoides teeth. A-D, holotype right M' (reversed) (M37570) of V. cooperi sp. nov. from Creechbarrow. E-H, left M'! (NMB Bchs656) of V. simplicidens (Van Valen) from Bouxwiller, drawn from cast. I-K, left DP*? fragment (M35646) of V. cooperi? from Creechbarrow. A, E and I are occlusal views. B, F and J are mesial views. C, G and K are distal views. D and H are buccal views. All x 6-6. and long lingual slope. It differs in the more sharply acute lingual outline, and presence of a paraconule which is buccal to the metaconule, both being isolated by clefts from the protocone. There is the faintest suggestion of a postparaconule crista. Although only the holotype is definitely referrable to this taxon, its distinctness and the rarity of all members of the family Paroxyclaenidae are considered sufficient justification for descrip- tion of a new species on an isolated M’. 342 J. J. HOOKER Little can be said of the relationships to the other species of the genus, but V. simplicidens of middle to late Lutetian age is a potential ancestor for V. cooperi. The minimum changes required would be molar shortening, accompanied by conule reduction and paracone and metacone height increase. V. simplicidens has no known specialized characters which would prevent such a relationship. Order CETACKEA Brisson 1762 Suborder ARCHAEOCETI Flower 1883 Family DORUDONTIDAE Miller 1923 Genus ZYGORHIZA True 1908 Zygorhiza wanklyni (Seeley 1876) Kellogg 1936 * 1876 Zeuglodon Wanklyni Seeley: 428-432. v. 1907 Zeuglodon Wanklyni Seeley; Andrews: 124-127, fig. 1. 2? 1933 Zeuglodon wanklyni Seeley; Burton: 161. v. 1936 Zygorhiza? wanklyni (Seeley) ?Kellogg: 174-175. v. 1972 Zygorhiza wanklyni (Seeley); Halstead & Middleton: 186-187, text-fig. 1. v. 1980 Zygorhiza wanklyni (Seeley); Hooker & Insole: 41. The history of occurrence of this whale from the Barton Clay of Barton has been fully reviewed by Halstead & Middleton (1972). In summary, the holotype skull is lost and referred material is restricted to a few isolated vertebrae. It is worthy of note that a series of about two dozen associated vertebrae is in the collections of the Bournemouth Natural Science Society. They are currently being conserved by Paul Clasby of Lymington and will be the subject of a joint project when the conservation work is finished. Detailed stratigraphical information is not recorded with any of these specimens, but the abundance of glauconite grains on all except M12346 (figured Halstead & Middleton, 1972: text-fig. 1) strongly suggests the range of beds B—D as the provenance. M12346 may be from beds E or F as its matrix consists of non-glauconitic clay. Family BASILOSAURIDAE Cope 1867 Genus BASILOSAURUS Harlan 1834 Basilosaurus sp. indet. v. 1972 Basilosaurus sp. indet.; Halstead & Middleton: 187-189, text-figs 2-3. v. 1980 Basilosaurus sp.; Hooker & Insole: 41. No new specimens have come to light from the Barton Clay of Barton since Halstead & Middleton’s (1972) record, but I call attention to a minor inaccuracy on p. 187 of their paper. Like most of the Zygorhiza specimens, the Basilosaurus vertebra, M26552, collected by Ken Hobby, has abundant adherent glauconite grains indicating a provenance from beds B—D, not bed E as stated. A personal communication from Mr Hobby shortly after he made the find in 1966 supports this provenance, although the specimen was not actually found in situ but on the slopes of Barton Cliff east of Chewton Bunny. The absence of glauconite grains from the other vertebra (M26553), collected by Paddy Blackwell, suggests a provenance of beds E-—F (cf. D or E according to Halstead & Middleton, 1972: 189). Order PERISSODACTYLA Owen 1848) Superfamily EQUOIDEA Hay 1902 Family PALAEOTHERIIDAE Bonaparte 1850 (sensu Remy 1967) TYPE GENUS. Palaeotherium Cuvier 1804. INCLUDED GENERA. Pachynolophus Pomel 1847a, Anchilophus Gervais 1852, Propachynolophus Lemoine 1891, Propalaeotherium Gervais 1849, Lophiotherium Gervais 1859, Paraplagiolophus BARTONIAN MAMMALS OF HAMPSHIRE BASIN 343 ectph AN prphd mephd hyphd ~ JL JL J Df Lr met! hyp 1st 2nd 3rd A B crescent crescent crescent metph Text-figure 43 Dental nomenclature of palaeotheriid cheek teeth. A, left upper molar. B, distal half of right M, and complete right M,. Abbreviations: A. cr—rhinocerotoid crista B. entd—entoconid dhc—distal hypoconal crest hyd—hypoconid ectph—ectoloph hyld—hypoconulid hyp—hypocone hyphd—hypolophid hys—hypostyle mephd—metalophid mes—mesostyle metd—metaconid met—metacone metsd—metastylid metl—metaconule msd—mesoconid metph—metaloph paphd—paralophid mets—metastyle pasd—parastylid par—paracone prphd—protolophid pas—parastyle prtd—protoconid prt—protocone prtl—paraconule prtph—protoloph Depéret 1917, Plagiolophus Pomel 1847b, Leptolophus Remy 1965 and Pseudopalaeotherium Franzen 1972. RANGE. Late Ypresian to Stampian of Europe. Genus PROPALAEOTHERIUM Gervais 1849 TYPE SPECIES. Palaeotherium isselanum Blainville 1864. Lutetian, Issel, Aude, France. INCLUDED SPECIES. P. helveticum Savage, Russell & Louis 1965, P. argentonicum Gervais 1859, P. rollinati Stehlin 1905a, P. hassiacum Haupt 1925, P. messelense (Haupt 1925) Savage, Russell & Louis 1965, P. parvulum (Laurillard 1849) Depéret 1901, P. sinense Zdansky 1930 and ?P. hengyangense Young 1944. RANGE. Early Lutetian—early Ludian?, Europe; late Eocene, China. D1aGnosis. See Savage et al. (1965: 57). Propalaeotherium aff. parvulum (Laurillard 1849) Depéret 1901 (Pls 20-21; Text-fig. 44A; Table 25) v. 1977b Propalaeotherium cf. parvulum (Laurillard 1849); Hooker: 141. v. 1979 Propalaeotherium cf. parvulum (Laurillard 1849); Kemp et al.: 102. v. 1980 Propalaeotherium cf. parvulum (Laurillard 1849); Hooker & Insole: 42. Ho.otype of P. parvulum. This is an M? from the Lutetian of Argenton, Creuse, France (Blainville 1839-64: genus Lophiodon, pl. 3), whose ‘present location [is] unknown’, according 344 J. J. HOOKER to Savage et al. (1965: 66). Stehlin (1905a: 401) was also unable to locate this specimen but described and figured (1905a: text-fig. 25) three topotypes in Bordeaux Museum and the Ecole des Mines, Paris (the latter now in the MNHN). Another topotype right M? similar to but more worn than the holotype (as figured by de Blainville) exists in the MNHN (unnumbered; cast in BM(NH) no. M42170). MATERIAL. Right P?* and left P*, almost certainly associated (M36498), two P*s (M37464, M37466), P* (M37465), P?/* (M37706), five M?s (M35596—7, M37468-9, M37707), DP*? in two non-fitting halves (M37467), P, (M37538), P4 (M36176), five M,/.s (M35598—9, M37470-2), M3 (GM 978110-1; cast in BM(NH) no. M36493) and DP? (M37708). HORIZONS AND LOCALITIES. The M;, is from the Elmore Member, Barton Clay Formation, Elmore. The rest of the material is from the Creechbarrow Limestone Formation, Creechbarrow. DIAGNOsIS (quoted from Savage et al., 1965: 66). “Smaller animals referable to Propalaeo- therium, comparable in size to very large individuals of Hyracotherium or of Propachynolophus maldani (estimated skull lengths: 140 to 160 mm). Upper molars with strong and somewhat bulbous mesostyles. Protoloph and metaloph with strong posterior flexure at labial extremity giving direct connection with paracone and metacone respectively. Upper premolars non- molariform, and with no mesostyles. P, may have incipient entoconid; otherwise lower pre- molars non-molariform’. This appears to be the most recent diagnosis of the species. However, Stehlin’s (1905a: pl. 9) figures show that the flexure of the protoloph and metaloph is by no means a constant feature, and that the upper distal premolars can show a moderate stage of molarization with small hypocones and mesostyles. In view of the uncertainties regarding the limits and definition of P. parvulum, no attempt will be made here to alter the diagnosis of Savage et al. (1965). AUTHORS’ CONCEPT OF THE SPECIES. The name P. parvulum has tended to be used for all small Lutetian to Bartonian Propalaeotherium, not referable to P. messelense. Stehlin (1905a) fre- quently referred to the great variation in size and especially morphology in assemblages from Egerkingen, Chamblon and Mormont, some of which he referred to as varieties but without naming them. He (1905a: 430) was unable, however, to sort out these varieties into any groupings according to the different Egerkingen fissures or facies. A study of Stehlin’s (1905a: text-fig. 26; pl. 9) figures shows that there are quite striking differences in length/width proportions of the upper molars and distal premolars. There is also some tendency towards molarization in the case of the relatively longer P*s (e.g. Stehlin, 1905a: pl. 9, figs 67, 70 in contrast with figs 1, 2, 45, 57). These differences become most marked when the material from Mormont and Creechbarrow is considered. Unfortunately these two localities have yielded few specimens. The scatter diagrams in Text-fig. 44A are presented to show the extent of the differences in length/width proportions of P7>-M? from the different Egerkingen fissures and facies, and from Mormont (mainly from Stehlin’s (1905a) measurements) and Creechbarrow (Table 25). Unfor- tunately there were very few measureable lower molars available to Stehlin and, whereas the majority of upper molars are from Egerkingen «, most upper premolars are from the aberrant facies. Nevertheless certain patterns seem to emerge. Upper molars of the Egerkingen y fissure and grey marl and aberrant facies tend to be slightly larger than those from « and f fissures. Those from Mormont and Creechbarrow both plot bimodally for size and the larger specimens are relatively shorter and broader than the smaller specimens. Similarly, in the case of P*, single specimens from « and f plot smaller than those from the grey marl and aberrant facies. The Mormont and Creechbarrow P*s and P*s are again bimodal but more on size and much less on proportions than are the upper molars; however, their total range is no greater than that of the aberrant facies P*s. Two isolated M,/.s from Creechbarrow are very similar in absolute size and proportions to those of Stehlin’s (1905b: text-fig. 44) lower jaw (NMB Ec3), with P, 4, M,_,, referred to as ‘Unbestimmbare Mandibularmaterialien kleiner Palaeohippiden von Eger- kingen’, but (1905b: 542) thought most likely to belong to ‘einer der vielen Varietaten von BARTONIAN MAMMALS OF HAMPSHIRE BASIN 345 om! ZA we r [] m2 N| (10 m3 Lo Eclépens-Gare Robiac Creechbarrow ee ese Eclépens-Gare Creechbarrow Robiac : Hy 1.75 1.80 1.85 1.90 1.95 2.00 log lI. x w. Text-figure 44 A, scatter diagrams of length (I) against width (w) of P* * and upper molars of Propalaeotherium parvulum (Laurillard) and P. aff. parvulum from various European localities. L] = Egerkingen grey marl facies; © = Egerkingen aberrant facies; V = Egerkingen y; A = Egerkingen ~; © = Egerkingen f#; 17 = Eclépens-Gare; {* = Creechbarrow; = Gentilly (Upper Calcaire Grossier). Of the upper premolars, solid symbols = P?, outline symbols = P+. Of the upper molars, symbols solid on left = M', on right = M? and completely solid = M?; outline symbols = M'/*. Measurements in millimetres. Lines join teeth of one individual. B, histograms of log. length x width of upper molars of Lophiotherium siderolithicum (Pictet) and the lectotype of L. robiacense Depéret, from Eclépens-Gare, Creechbarrow and Robiac. Original measurements in millimetres. Propalaeotherium parvulum’. The lower molars of two other Egerkingen specimens (Stehlin 1905a: text-fig. 26; pl. 9, fig. 35) are relatively much narrower. The greatest differences in size or length/width proportions of this material occur in the P?+s and M’s, classically some of the intraspecifically most variable teeth (Gingerich 1974). Thus no convincing separation into two (or more) species can be made on the basis of these scatter diagrams. Until a more detailed synthesis is done of all the material (especially that from Egerkingen), it is best to accept only one species as occurring at any one time. Whether or not 346 J. J. HOOKER Table 25 Length (1) and maximum width (w) measurements of Propalaeotherium aff. par- vulum from Creechbarrow and Elmore. Measurements in millimetres. No. Tooth l Ww M36498 P? 84 10-0 M36498 p 90 10-5 M37464 P 79 100 M37466 ps 700) “a8 M37465 ps all wae 684 M37707 Me 100 IS? M35596 Me 106 (146) M37468 Me 10-4 (142) M37538 P, le wis! M36176 P, 90 64 M35598 Vem Cee M37470 Mi, 94. 0 GM 978110-1 M, 139 67 M37708 DP? 75 oT this should be considered conspecific with P. parvulum, based on material from the type locality, cannot be definitely decided here, hence the ‘aff.’ prefix to the English Bartonian specimens. DESCRIPTION. Comparisons of the Creechbarrow material with the topotypes is restricted to upper molars. The Creechbarrow upper molars differ from the topotypes in having a deep fissure in the protoloph between the paraconule and protocone. This fissure is very shallow in the virtually unworn topotype M"”? figured by Stehlin (1905a: 401, text-fig. 25). It is evident from other upper molars from Argenton that the paraconule became fused to the protocone at an early wear stage. Upper molars from Eclépens-Gare (and Mormont undifferentiated, almost certainly also from Eclépens-Gare) are very similar to those from Creechbarrow in the depth of their protoloph fissure, as is the composite upper dentition from Lissieu figured by Depéret (1901: pl. 4, figs 2-3). Upper molars from Egerkingen are in general intermediate in the depth of this fissure but vary somewhat between the two extremes. Those from the younger fissures « and f appear similar in this feature to those from the older ones. All the Creechbarrow M?s are large and relatively short and broad, unlike the small and relatively more elongated M? from Mormont (Stehlin 1905a: pl. 9, fig. 68). Their mesostyles are also very strong. The P*s and P*s can be divided into two morphological types: Type A Type B Outline: Triangular Subquadrate Width/length ratio: High Low Hypocone: Absent Present In addition, type B teeth are smaller than type A teeth but not enough are available to test usefully the coefficient of variation. M37465 is fairly worn, closely resembles LM 46 from Mormont (Stehlin 1905a: pl. 9, fig. 67) but the hypocone appears larger; the wear makes judgement of cusp size difficult, but the remains of the lingual valley between the protocone and hypocone is nearly central on the lingual margin. M37466 is less worn, narrower lingually and with closer protocone and hypocone than M37465; unlike the latter there is also a metostyle which is distal to the midpoint of the ectoloph. Both M37465 and M37466 are type B teeth (PI. 20, figs 1-2); the rest are type A (e.g. Pl. 20, figs 3-4). The P, has no paraconid (PI. 21, fig. 2). The metaconid is represented only by a slight lingual swelling halfway along the postprotocristid. There is a small hypoconid. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 347 The P, is of roughly rectangular outline with protoconid and hypoconid in line mesiodistally (Pl. 21, fig. 3). The paralophid is weak, directed mesiolingually and dies out well buccal of the lingual margin. The entoconid is absent. In these features it is like NMB Ec3 from the Eger- kingen grey marl or aberrant facies (Stehlin 1905b: 541, text-fig. 44). NMB Ecl from Eger- kingen a (Stehlin 1905Sa: 424, text-fig. 26) has semimolariform P,, with a fairly large entoconid; strong convex paralophid continuous to the lingual margin; and talonid broader than trigonid. The M, 2s are also more like those of Ec3 than Ecl in being relatively short and broad and in having the buccal crescents of trigonid and talonid more acute and the paralophid weaker (PI. 21, fig. 4). The M, is heavily worn. Its length/ width proportions are intermediate between those of Ecl and Ec3. It also has an expanded hypoconulid lobe like Ec1. The DP,,? is similar in general cusp arrangement to the non-molariform DP, (M35585) of Lophiotherium siderolithicum (see p. 352), but is longer and narrower; the talonid is lower, the entoconid being entirely missing; and the mesial slope of the metaconid is shallower, the base meeting the mesial end of the straight paralophid (PI. 21, fig. 1). Genus LOPHIOTHERIUM Gervais 1859 TYPE SPECIES. Lophiotherium cervulum Gervais 1859. Ludian, Euzet, Gard, France. INCLUDED SPECIES. L. pygmaeum (Depéret 1901) Stehlin 1905b and L. siderolithicum (Pictet 1857) Depéret 1901 (?syn. L. robiacense Depéret 1917). Note that L. magnum Matthes 1977, L. geiseltalensis Matthes 1977 and L. voigti Matthes 1977 all appear to be synonymous with existing species of Propalaeotherium, but represent milk dentitions (see also Franzen 1980b: 91). RANGE. Late Lutetian to Marinesian, Switzerland; Marinesian, England; Marinesian to Ludian, France; Ludian, Spain, Majorca. D1AGnosis. See Savage et al. (1965: 72). Lophiotherium siderolithicum (Pictet 1857) Depéret 1901 (Pl. 19, figs 8-9; Pls 22-23; Text-fig. 44B; Tables 26—27) v. 1912 Dictulumus leporinus (Owen) (sic); Keeping: 130 (error for Dichobunus leporinus; in reality Dicho- bune leporina Cuvier). v. 1977b Lophiotherium robiacense Depéret; Hooker: 141. v. 1980 Lophiotherium robiacense Depéret; Hooker & Insole: 42. TYPES AND HISTORICAL BACKGROUND. Pictet (1857: 53-57; pl. 4, figs 1-3) described and figured a right maxilla with P?-M? (the last tooth broken lingually) and figured (1857: pl. 4, fig. 4) an isolated worn upper right molar as representing his then new species Hyracotherium siderolithi- cum; he designated no types, so his two specimens were syntypes. Stehlin (1905b: 456) recog- nized the maxilla as the type and noted that it was lost; he did not mention the isolated molar. Following Stehlin’s decision, therefore, the maxilla is recognized here as lectotype. Because it is such an advanced individual and virtually inseparable from typical maxillae in the type assemblage of L. cervulum from Euzet, Stehlin (1905b) referred it to this species. He also recognized, however, that the assemblage from Eclépens-Gare as a whole, although variable, was more primitive than typical L. cervulum yet more advanced than Lutetian L. pygmaeum. He therefore designated three phylogenetic stages A, C and E for L. pygmaeum, L. ‘cervulum’ from Mormont (Eclépens-Gare) and L. cervulum from Euzet respectively. He thought that the mixing of the Mormont faunas (see Stehlin, 1903: 10-13) was the cause of such high morphological variation in the premolars, the type of L. siderolithicum thus being from a younger fauna and synonymous with L. cervulum. Depéret (1917) named the varieties atavum, transiens and progressum for individuals in the Euzet L. cervulum assemblage which resembled respectively Stehlin’s stages A, C and E. He then gave the name L. robiacense to two specimens from Robiac, which he considered to represent the constant state of stage C. Within Le Mormont, apart from a few teeth from St Loup, L. siderolithicum in fact is known J. J. HOOKER 348 A]UIe1199 SOL *36P9E WN) jesnjooo ‘q pue [eoong ‘ BPX < ‘eve ‘daag (LOLLEW) (Pessoaes) WW 14st “q “eg “BI (¢ “SIA SB [eNprArpur sures rd YQ “bh “BIA (S6P9EW) (Pastoral) ed ISH “q “VE “BLY (SOPLEW) 4d Yl “9 “PZ “BIA (99PLEW) ed Yl GP “SIA “‘SMOIA MOLIVQYIIIID WO (pieypliney]) wnjnasnd ye wnisayjoavjodoig JO Yy}99} yaoyd Jaddn jo sydeisoioew JYysIT YZ eld 349 BARTONIAN MAMMALS OF HAMPSHIRE BASIN ‘9 pue [esnjos0o ‘q ‘Jeong (86SSEIN) “TW YPl ‘Bp “BLY (OLIOEW) "d Yel O-PE “BIA (SESLEW) (Passaaad) © ‘ ‘ere doag Bp X ‘MOLILQYOIDID WO (pseypline Tq) wnjnaind ye wnisayjoavjodo.ig JO Y}99} Y29Y9 Jamo] Jo sydesso1OeW WYSIT d ysis ‘0-87 “BLY (80LLEW) 6°dC Yo O-e] “BLY “smara [eNsury IZ 938 Id 350 J. J. HOOKER only from Eclépens, associated with an almost exclusively Marinesian mammal fauna (see Hartenberger, 1973: 68, for list), and de la Harpe (1869: 459) noted that out of five or six fissures in the quarry at Gare d’Eclépens, almost all of the animal fossils came from a single fissure. The degree of size variation is low (see Text-fig. 44B), supporting the idea that the old collections of Eclépens L. siderolithicum constitute essentially a single assemblage. Thus the high morphological variation recorded by Stehlin for the premolars of L. siderolithicum is considered to be an accurate representation of the species. The position is, however, unfor- tunately complicated in that other fissures were exploited after the time of de la Harpe and the specimens were not differentiated in the LGM collections (M. Weidmann, personal communi- cation 1983). The species name does not appear to have been used in major zoological literature since Depeéret (1917). Savage et al. (1965) did not even include it in a synonymy list, considering that the Mormont material probably belonged to L. robiacense. They nevertheless doubted the constancy of the premolars of L. robiacense, the species being represented at most by two specimens. Sudre (1969a: 103-105, 118) pointed out that both specimens came from a very high level in the Robiac section which Roman (1904) considered to be lowest Ludian. As such, they are likely to be little older than the specimen from Fons 1 figured by Remy (1967) as L. aff. cervulum. Remy (1967: 29-30) decided that species of Lophiotherium could only be determined on tooth morphology by statistical procedures, individual specimens being useless. He then showed that L. pygmaeum, L. robiacense and L. cervulum could be distinguished by P’ */M!-° length ratio with no overlap. However, he was only able to measure one specimen of each of L. pygmaeum and L. robiacense; and the former is composite (Stehlin 1905b: 452, text-fig. 31). It is considered here that the best procedure is to recognize L. siderolithicum as a statistically well-represented species, the type assemblage being essentially homogeneous, and a senior synonym of L. robiacense (if the two are identical). MATERIAL. (54): Associated left P*-M? (M36177); two P's (M37410-1); P? (M36497); two P?s (M37412, M37703); two P*s (M37413-4); twelve M!/2s (SMC 9969, M35584, M36499, Table 26 Length (1) and maximum width (w) measurements of Lophiotherium siderolithicum from Creechbarrow. Measurements in millimetres. No. Tooth ] Ww No. Tooth l Ww M37410 P} 5-4 4-4 M37430 1 77 5:7 M36497 p? 4-1 = M37705 P, 7:3 5:9 M37703 Pp? = 8-0 M37705 M, 78 6:2 M37413 p+ Wed 9:7 M37431 M,/2 - 5°5 M36177 p* 7-0 9-7 M37432 M,/2 7:8 5:9 M36177 M! 7-1 9-9 M37433 M,)2 8-2 5:5 M36177 M2 7:2 10-7 M37434 M,)2 7:6 6-1 M35584 M?/2 75 - M37436 M,/2 8-0 5:4 M37415 M?/2 7-4 9-8 M35587 Mi). (7:3) 5:5 M37416 M?/2 8-1 10-6 M37705 M, 11-0 6:3 M37417 M?/2 7:8 10-9 M37438 M, 10-5 5-6 M37418 M?/2 7-6 9:7 M37440 M, — 6-1 M37419 M?/2 8-7 11-9 M37441 M, (11-0) 5:7 M37421 M?/2 8-5 10:8 M37442 M, (11-0) 5:8 C9969 M?/2 7-6 10-3 M36180 M, 10-5 5:5 M36178 M? 7-0 10-3 M35585 DP, 6:6 4-1 M37424 M? — 10-6+ M35590 DP, — 4:5 M37426 DP? 7:4 7:2 M37447 DP, 7-4 5-0 M37427 DP* 76 9-1 M37449 DP, — 5:6 M37428 DP* 6:5 7-6 M35586 DP, 7-6 5:0 M35588 P, 7-0 5:3 M37451 DP, 6:9 5:3 M37429 PA il 5-1 M37452 DP, TET oF BARTONIAN MAMMALS OF HAMPSHIRE BASIN 351 M37415-23); three Ms (M36178, M37424-5); upper molar fragment (M36179); DP? (M37426); three DP*s (M37427-8, M37704); associated left P,, M, 3, (M37705); five P,s (M35588, M35594, M37429-31); eight M,/.s (M35587, M35589, M37432-7); eight M3s (M36180, M37438—44); two lower molar fragments (M35591, M37445); two DP,s (M35585, M37446); six DP3s (M35586, M35590, M37447-S0); four DP4s (M35592-3, M37451—2); lower canine (M36828). DOUBTFULLY REFERRED MATERIAL. (23): Two upper canines (M37453—4); twenty one incisors (M35595, M36228—-35, M36237-8, M36792, M37455-63). HoRIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. EMENDED DIAGNOSIS (modified from Savage et al. (1965: 79-80) for L. robiacense). Molariform upper teeth with very strong and discrete paraconules, weak to strong, indistinct to discrete metaconules; P* hypocone missing to as developed as protocone, usually moderately to well developed; P* hypocone missing or of height equal to protocone, poorly to well separated from it, usually moderately separated; P* hypocone missing or of height equal to protocone, poorly to moderately separated from it. P, entoconid moderately to strongly developed; P; entoconid weak to moderate; P, metaconid weak to strong. DESCRIPTION. In tooth size the Creechbarrow assemblage is slightly larger than the Eclépens- Gare assemblage (see Text-fig. 44B) but there is much overlap. In morphology the two are very similar (Table 27), the latter having been described in detail by Stehlin (1905b). In response to Remy’s (1967) comments (see p. 350) I have attempted to divide into cate- gories the grades of molariformity of the upper (those with the greatest number of easily recognizeable character states) premolars and score numbers of individual teeth for both assemblages. The different premolars are mainly isolated and often difficult to distinguish from one another. It seems possible to distinguish P*s from P*s because the latter have smaller, more mesial metaconules and are smaller overall. Moreover, they molarize differently: in P* the hypocone ‘grows’ gradually on the postprotocingulum, some distance behind the protocone, moving further distally in advanced stages; in P? the hypocone, from its moment of appear- ance, is the same height as the protocone and progressively buds off distally from it. P*s molarize like P*s and are difficult to separate except on smaller size and narrower proportions. P? is a small narrow wedge-shaped tooth with poorly differentiated protocone and hypocone. In Table 27, the number of specimens from both Eclépens-Gare and Creechbarrow and the lectotype of L. robiacense from Robiac are shown for each category of the four upper premolars. Although the peaks of the Creechbarrow character state distributions do not always correspond to those of Eclépens-Gare, they are not so skewed to one or other end of the scale as they are in L. cervulum or L. pygmaeum. The isolated upper deciduous premolars (PI. 22, figs 56) are identified as DP* and DP* and resemble those figured by Stehlin (1905b: 459, text-fig. 36) from Eclépens-Gare. The lower cheek teeth from Creechbarrow also closely resemble those of the Eclépens-Gare assemblage, being only slightly larger (see Table 26 for measurements). Recognizing the differ- ent isolated lower premolars is difficult. Moreover, the fully molariform P,4s are very like M,s or M,s. Three molariform teeth from the left side of the jaw, from Hole 6 at Creechbarrow (Text-fig. 4, p. 209), have identical preservation and very similar wear states (PI. 19, fig. 8). One is M;; an M, can be associated with it by matching up interstitial facets; the third does not fit against the M, but is almost certainly from the same individual. Its hypolophid is weaker than that of M, and its entoconid, although well developed, is slightly lower than the metaconid. It is thus identified as P, and is as fully molariform as that of L. cervulum figured by Savage et al. (1965: 74, text-fig. 35c—d). Of three isolated teeth, identified as P,, one is fully molariform like the Hole 6 one; another is worn but would have had the entoconid little more than half the height of the metaconid (PI. 23, fig. 4); and a third is intermediate between the two. Lower deciduous premolars are also difficult to separate from one another when isolated. Those here identified as DP; and DP, (e.g. Pl. 23, figs 2-3) are very similar to those figured by Stehlin (1905b: 469, text-fig. 39). An unworn non-molariform tooth with thin enamel and a 352 J. J. HOOKER Table 27 Numbers of specimens in each character state of the four upper premolars, for assemblages of Lophiotherium at Eclépens-Gare and Creechbarrow. At right, condition in the lectotype of L. robiacense from Robiac. Eclépens- Creech- Tooth Category Gare barrow Robiac ip 1. Hypocone undifferentiated on 3 0 postprotocingulum 2. Hypocone very small on 3 0 postprotocingulum 3. Hypocone smaller than protocone, 8 0 1 postprotocingulum broken 4. Hypocone = protocone, closer than 3 2 paraconule to metaconule 5. Hypocone = protocone, same distance 6 1 apart as paraconule and metaconule P32 1. Hypocone undifferentiated on 3) 0 1 postprotocingulum 2. Hypocone very close to protocone 7 1 3. Hypocone closer to protocone than 9 1 paraconule to metaconule 4. Hypocone-protocone distance 4 0 = paraconule—metaconule distance p2 il, 5 0 1 2 Same numbered categories as for P* 4 1 3}. 5 0 Pe 1. Metacone absent 2 0 2. Metacone small 4 0 3. Metacone subequal with paracone 0 1 1 papillate buccal cingulum between protoconid and hypoconid (M35585) is tentatively identified as DP,. Its metaconid and metastylid are scarcely separated even at the tip, the entoconid is very small and mesial and distal parts are of approximately equal width (PI. 23, fig. 1). As DP, it thus resembles L. pygmaeum more than it does L. siderolithicum as figured by Stehlin (1905b: 469, text-figs 38-39), and contrasts strongly with a more typical Creechbarrow DP, (M37446). A palaeotherioid curved lower canine with a distal facet (M36828) (Pl. 19, fig. 9) resembles those of L. cervulum figured by Depéret (1917: pl. 15). It is of appropriate size to belong to L. siderolithicum. While attempting to identify miscellaneous canines from Creechbarrow, I noticed a strong resemblance between M36828 and the isolated canines described from Eger- kingen and Mormont as Choeromorus jurensis Stehlin 1908 and C. helveticus Pictet & Humbert 1869 (particularly Pictet & Humbert 1869: pl. 25, fig. 11a). Because of their resemblance to miniature pig canines, these authors assumed that they were congeneric with the contemporary genus Cebochoerus, which was based on cheek teeth, and at that time placed in the family Suidae (now Cebochoeridae, p. 389). An anomaly was caused by the knowledge of certain tooth rows with typical Cebochoerus cheek teeth associated with non-Choeromorus-like canines. Stehlin ‘resolved’ the problem by assuming that the genera Cebochoerus and Choeromorus could only be distinguished on the canines as they had identical cheek teeth. Sudre (1978): 52-54) discussed the problem in more detail and considered two other possible solutions: either that the canines of Cebochoerus were sexually dimorphic or that the canines attributed to Choeromorus belonged to another kind of animal. They have never been found in situ in a jaw with a Cebochoerus cheek dentition, so Sudre’s second solution seems the more likely. In fact a close resemblance exists between Stehlin’s (1908: 708, text-figs 101-102) two crucial specimens and the lower canines of Anchilophus cf. dumasii figured by Remy (1967: 23, text-fig. 13; pl. 5, fig. 1). Anchilophus occurs in both the Egerkingen and Mormont faunas identified on cheek BARTONIAN MAMMALS OF HAMPSHIRE BASIN 353 teeth and generic identity of the canines with these seems very likely. The strong pig-like curvature to these teeth was evidently to overcome accommodation problems in the long, shallow mandibular symphyses of both Lophiotherium and Anchilophus. Two upper palaeotherioid canines of appropriate size and proportions are tentatively attrib- uted to L. siderolithicum, as are various incisors which resemble those in Depéret’s figures of L. cervulum (1917: pl. 15, fig. 9). CONCLUSIONS. The presence of L. siderolithicum at Creechbarrow is good evidence of later Bartonian age. Its slightly larger size compared to the type assemblage might indicate either geographical differences or a temporal trend towards the even larger specimens from Robiac. There is no independent evidence either way, but the Lophiotherium lineage appears to have undergone no significant or directional change in size throughout its known stratigraphical range. Genus PLAGIOLOPHUS Pomel 1847b [ = Paloplotherium Owen 1848a] TyPE SPECIES. Palaeotherium minus Cuvier 1804. Premiére Masse du Gypse, Montmartre, Paris, France. INCLUDED SPECIES. P. cartieri Stehlin 1904b, P. lugdunensis (Depéret & Carriére 1901), P. cartail- haci Stehlin 1904a, P. annectens (Owen 1848a), P. fraasi Meyer 1852, P. javalii (Filhol 1877b) (doubtfully distinct from P. fraasi) and P. curtisi sp. nov. RANGE. Late Lutetian—Stampian, Europe. EMENDED DIAGNOSIS (modified from Viret (1958: 359) and largely complementing Franzen’s (1968: 16) diagnosis of Palaeotherium). Medium-small to medium-—large Palaeotheriidae, cranium length 150-300 mm. DP? replaced by P! or both absent. Cheek teeth with or without cement; with high percentage of pericanalicular dentine (Remy 1976); and semihypsodont. Upper molars with buccal cusp height from 70-100% of tooth width and from 1-3—1-6 times the lingual cusp height. Upper molariform cheek teeth with prominent paraconules and high oblique metalophs. P* with or without metacone and without hypocone; P* with or without elongate metaloph and without hypocone; P; with or without hypoconid and without ento- conid; P, with or without small entoconid and with high hypoconid. Successive molars increase in size distally. Lower cheek teeth with weak interrupted buccal cingulum; and lingual valley of first crescent deep. M,_, with hypoconulid. Twinned metaconid/metastylid on molari- form lower cheek teeth. Postcanine diastema medium short to medium long (lowers from 29-36% of cheek tooth row; uppers from 15-26% of cheek tooth row). Narial incision varying in depth from P' to M’, the lower border consisting of premaxillae plus maxillae. Astragalus trochlear ridges at angle of 75° to navicular facet. Plagiolophus curtisi sp. nov. For type material, see under nominate subspecies. For synonymy, see under each subspecies. DiaGnosis. 1, small to medium-sized Plagiolophus (M? length 13-0-16-5mm); 2, tooth cement absent; 3, P' present; 4, P? metacone present; 5, P* metaloph short, not reaching distal cingulum; 6, P* + metaloph nearly longitudinally orientated; 7, M!? distal hypoconal crest strong and longitudinally orientated; 8, upper cheek teeth with high distal cingulum; 9, upper molar paracone and metacone buccally salient; 10, upper molar metaloph obliquity medium; 11, molar crown height medium (height of buccal cusps = maximum width of tooth); 12, upper molar buccal/lingual height ratio low; 13, upper molar protoloph fissure between paraconule and protocone moderately deep; 14, upper molar ‘rhinocerotoid’ crista present; 15, M3 moder- ately distally expanded; 16, canines large and parallel orientated; 17, P; hypoconid present; 18, P, entoconid absent; 19, lower molar buccal/lingual height ratio 1:1; 20, molar enlargement gradient low. Muzzle broad anteriorly and mandibular symphysis narrow; palatal concavity J. J. HOOKER 354 ‘Lee d 99g (LLI9EW) zN--d YI] “6 “SIA (SLI9EW) (passaaad) .AL 1Y8t1 ‘*g “B14 (6966 OWS) 7/,W lL “SIA (LZPLEW) dC Yl “9 “89 “BLY (97PLEW) ¢d Cl Ol ‘S “BIN (ET PLEW) ed YQ “ep “BIA (COLLEW) (Pessoal) -q WB “gq ‘VE “BI (LOPIEW) zd WAI “7 Bla (OIPLEW) (Pessoaod) dq WYStI ‘gq ‘ey “B14 (pexyjnsun pue q) [esn[o90 pure (e) [eooNg ak SMOIA “p X ‘MOIIVQYIIDID WOT (19)91g) WNIIYINOAapIS WNi4ayjO1ydoT JO 4190} Yooyo Jaddn jo sydeisoINVW YSIT ZZ *LId 355 BARTONIAN MAMMALS OF HAMPSHIRE BASIN “Lye ‘d 29g (SEpLEW) *W Yay ‘9-eg “BLY (OSPLEW) *d Yl O-¥p “SLA (ZSPLEW) (Pesseaes) ’Gq WB ‘O-KE “BI (LppLEW) £dC Wal O-87 BLA (S8SSEW) “dC Yel O-P] “BIA SMOIA [ensuly ‘9 pur [esnooo ‘q ‘eoong ‘Bp x ‘MOIIEQYOIIID WO (19391 gq) WNIIYINOJapIS WNLaYyIO1YdOT JO 499} Y2oYd IOMO] Jo sydeiso1OeVW WSIT ¢€79Id 356 J. J. HOOKER 24 : 22 * * * S 205 o vv = 181 a 8 Ix (e) 164 voy 14—- 40 60 80 100 120 140 ~—«: 1€o |x w of lower canine Text-figure 45 Scatter diagram of length x width of lower canine against mean width of M? of Plagiolophus and Paraplagiolophus. Measurements in millimetres. | = Plagiolophus annectens (Owen) from Hordle Cliff; W = P. minor (Cuvier) from La Débruge; > = P. fraasi Meyer from the Phosphorites du Quercy; V = P. sp. from Egerkingen y; « = P. curtisi curtisi sp. & subsp. nov. from Barton Cliff; @ = P. curtisi creechensis subsp. nov. from Creechbarrow; a | (EE fie ee me ames eee a | A 20 25 I. 10 ( er ae w. D> yan Aa (O) T a Ut ) a F i T a T T T T Tr i; T T aes =a | 10 15 20 B Text-figure 46 Scatter diagram of length (1) against width (w) of (A) upper and (B) lower cheek teeth of Plagiolophus curtisi curtisi sp. & subsp. nov. from Barton Cliff (solid symbols), and P. curtisi creechensis subsp. nov. from Creechbarrow (outline symbols). Measurements in millimetres. O = P3; A = P3; 0 = P$; © = Mt; V = M3; << = M3; [> = M, 2. Lines join teeth of one individual. P* is similar to P? but the paracone and metacone seem better separated. The metaloph has similar mesial crests to P* and bends at its distal end to fuse with the protocone. The resultant hook-shaped structure (‘crochet’ of Stehlin, 1904a: 461) resembles that of P* of P. cartailhaci (Stehlin 1904a: pl. 12, fig. 1) and was the main reason for earlier misidentification of M26176 as P. cartailhaci (Hooker 1972). The ‘crochet’ also occurs in some individuals of P. annectens (Depéret 1917: pl. 9, fig. 2; Remy 1967: text-fig. 4). The cingulum is interrupted lingually. Upper molars: In M26176, the protocone and paraconule are fused by wear on M' but not on the less worn M?-*. On the even less worn M26238, the paraconule is connected to the protocone. The paraconule protrudes less distally from the protoloph than in higher-crowned species like P. annectens. The buccal cingulum in M26176 is very weak (where this region is not eroded away), but it is not certain whether or not this constitutes individual variation; M26238 is also eroded here. On the M?s of both individuals (PI. 24, figs 1b-c, 2a—b), the subterminal metastyle is absent from the basal half of the buccal wall, but a detached, more distal basal swelling could also be construed as part of the metastyle. In M26176, the hypostylar or distal cingular region is relatively expanded in length and height. It is formed from two curved crests, from the distal BARTONIAN MAMMALS OF HAMPSHIRE BASIN 359 Plate 24 Light macrographs and x-ray of upper teeth and cranial fragments of Plagiolophus, x 1. Figs 1-2 Plagiolophus curtisi curtisi sp. & subsp. nov. from Barton. Fig. 1a—f, holotype (M26176): a, left dorsolateral view of palatal fragment; b, buccal view of left P>-M?; c, ventral view of palatal fragment and occlusal view of P?—M?; d, dorsal view of palatal fragment; e, dorsal view of left frontal fragment; f, ventral view of left frontal fragment. Fig. 2a, b, buccal (a) and occlusal (b) views of right M? (M26238). See p. 356. Fig. 3. Plagiolophus fraasi Meyer from Caylux: side view x-ray of right maxillary-premaxillary fragment with canine (M1673). J. J. HOOKER 360 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 361 ends of the ectoloph and hypocone, which meet in the middle of the distal tooth margin; the ectolophal part projects distally more than the hypoconal part. M26238 is slightly less expand- ed distally than M26176, especially from the ectoloph, the distal hypoconal crest is lower and the two crests meet at a V-shaped distal concavity. Potential variation in this subspecies can only be observed on these two M?s, and appears to be restricted to height of connection of paraconule to protocone (probably affecting the other molars as well) and degree of distal expansion. The curious subterminal metastyle is likely to be at least of common occurrence, if not a constant character. Too few specimens are known to test constancy of size but both M?s are very close and the other teeth of M26176 are all larger than any of their equivalents (mainly different individuals) from Creechbarrow. Lower premolars: The only remnant of P;, the distal root, is as broad as on the Creechbarrow P,s where a hypoconid is present (PI. 25, fig. 1d). The P; of P. annectens, which lacks a hypoconid, has a small peg-like distal root. Moreover, the presence of a P,; hypoconid in M26176 can be extrapolated from the presence of a P*? metacone. P, is as far as can be observed typical for the genus. Lower molars: Right M, is missing the mesial half of its lingual margin and is buccally eroded; both M,s are better preserved and similar to M,; right M, is complete. The metaconid and entoconid (and hypoconulid in M;) have broad bases as seen lingually and tend to reduce the depth of the valleys enclosed by the first and second (and third on M;) crescents (PI. 25, fig. 1c, e). The effect is to give the teeth a robust appearance. Some specimens of other species of the genus have slight development of this feature, but its occurrence also in the Creechbarrow lower molars suggests that it might represent more than just individual variation in this case. Skull: What remain of the side walls of the maxillae diverge dorsolaterally from the diastematal ridges more than in other species of the genus, presumably to house the large parallel upper canine roots. Only the left postcanine diastema is complete posteriorly and neither is complete anteriorly. The maximum measurable length is 14-4mm, but they are estimated to have been at least 23-0 mm long, thus similar in relative length to those of P. annectens. The internal nares extend anteriorly to between the level of M? and M?. The infraorbital foramen is above the junction of P* and P*, in contrast to P. annectens and P. minor where it is above the middle of P*, and to P. cartailhaci and P. fraasi where it is above the anterior half of M!'. It migrates posteriorly in ontogeny, but moves little (less than the length of P*) after the permanent premolars have erupted, a stage passed by the specimens considered here. Immediately behind the infraorbital foramen there is part of a preorbital fossa (PI. 24, fig. 1b). These may be of two kinds in the other equoid family, the Equidae (Gregory 1920): an upper (lacrymal’) and a lower (‘malar’), which sometimes merge to form one. Its preserved antero- ventral edge is sharp and extends posteriorly for the length of P* in the ‘malar’ region, but its dorsal extent is unknown. Gregory (1920: 282) stated that neither Palaeotherium nor Paloplo- therium (= Plagiolophus) shows pronounced preorbital fossae, but, in addition to M26176, a maxilla of P. fraasi (M14734) from the Bembridge Marls of Bembridge shows part of a ‘malar’ fossa (Text-fig. 47B). Shallow preorbital depressions in P. annectens (Owen 1848a: pl. 3, fig. 1) and another P. fraasi specimen (M1733) from the Phosphorites du Quercy reach dorsally to the nasals, covering both ‘malar’ and ‘lacrymal’ regions, but like modern Equus lack sharp rims Plate 25 Light macrographs of lower teeth, jaw fragments and radius of Plagiolophus, x 1. Fig. la—-g Plagiolophus curtisi curtisi sp. & subsp. nov. from Barton, holotype (M26176): left mandibular ramus with P,, M,_, in occlusal (d) and lingual (e) views; right mandibular ramus with M,_; in buccal (a), occlusal (b) and lingual (c) views; mandibular symphysis in right lateral (f) and dorsal (g) views. See p. 356. Fig. 2a—c_ ?Plagiolophus sp. indet. Proximal half of right radius from Barton (M34864). Views are anterior (a), proximal (b) and posterior (c). See p. 364. 362 J. J. HOOKER (Gregory 1920: pl. 18). The apparent individual nature of this fossa in Plagiolophus contrasts with the hipparionine equids, where it is considered a generic character (Woodburne & Bernor 1980). The cranial roof is known only from a fragment interpreted as part of the left frontal (Pl. 24, figs le-f; Text-fig. 49B). Its dorsal surface was flat, the supraorbital rim thin and the postorbital process thin and horizontal, in contrast to other species where it is gently domed with the supraorbital rims thick and the postorbital processes strongly downturned (e.g. Text-fig. 49A). The leading edge of the orbit appears to have extended forward to a point above the junction of M! and M?, further forward than in other species: M? midpoint in P. annectens and P. fraasi and M? ® junction in P. cartailhaci. As for the infraorbital foramen, the orbital position is ontogenetically variable, but the specimens concerned have similarly worn teeth. The orbital floor and adjacent areas are incomplete but show the course of foramina and other structures (Text-fig. 47A). The orbital opening of the maxillary foramen and probably also the anterior postpalatine canal are as in Pachynolophus lavocati (Remy 1972: 60-62, text-fig. 9), P. livin- ierensis and Hyracotherium leporinum (Savage et al. 1965: 44, 46-47, text-fig. 20), P. annectens and P. fraasi. The squamosal is perforated by a temporal canal, opening posteriorly at the postglenoid foramen and anteriorly at the temporal foramen (Text-fig. 48). The glenoid region differs from that of P. fraasi as follows: the main postglenoid process is slightly more prominent but less posteriorly expanded; and the smaller posteromedial accessory process is shorter but more sharply differentiated from the main process. Individual variation in these features is unknown. A appc distal root Text-figure 47 A, Plagiolophus curtisi curtisi sp. & subsp. nov., holotype M26176 from Barton Cliff. Dorsal view of maxillary fragment (incorporating part of the palatine) drawn as from the right side, consisting of the left maxillary fragment from P* forwards (reversed) and right maxillary fragment from M!' backwards. Abbreviations: appc = anterior postpalatine canal; mms = maxillary sinus (main part), tms = maxillary sinus (turbinate part); vtc = ventral turbinate crest; for other abbrevia- tions see Text-fig. 51, p. 366. x 1. B, Plagiolophus fraasi Meyer, M14734 from the Bembridge Marls of Bembridge. Lateral view of right maxillary fragment (reversed). x 1. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 363 apgp mpgp Text-figure 48 Plagiolophus, glenoid region. A—C, holotype of Plagiolophus curtisi curtisi sp. & subsp. nov. (M26176) from Barton Cliff. D, P. fraasi Meyer (M1733) from the Phosphorites du Quercy. A, lateral view of right glenoid region; B, posterior view of right glenoid region; C, ventral view shown as right glenoid region, but composite of right and left; D, ventral view of left glenoid region (reversed). Abbreviations: Al = alisphenoid; apgp = accessory postglenoid process; gf = glenoid fossa; mpgp=main postglenoid process; Pat =parietal; pgf=postglenoid foramen: Sq = squamosal; tf = temporal foramen. All x 1. The mandibular symphysis is massive for housing the canine roots. Its ventral edge is angled anterodorsally below a point just behind the emergence of the canines. In contrast the small symphyses of other species have an even, anterior tapering, shape. The symphysis in M26176 terminates an unknown distance in front of the premolar row; in P. annectens it is just in front of P,. The maximum length measurable on the lower diastemata of M26176 is 23-0mm, the same as that estimated for the complete upper diastema. As there was apparently no tooth immediately following the breakage of the lower diastema, P, was probably absent, P, begin- ning the row behind the main mental foramen. In dorsal view the symphysis is only slightly concave laterally and the diastematal ridges are almost straight. In other species it is laterally more deeply concave and so are the ridges. Like the palate, the interdiastematal symphysial concavity is shallower in P. curtisi than in other species of Plagiolophus. The anteromedial edge of the ascending ramus forms an acutely angled ridge because the masseteric ridge lies posterolateral to it. This contrasts with other species where both ridges normally lie side by side and are weaker and more rounded. The curvature of the posterior edge of the coronoid process indicates a more gentle slope towards the condyle than in other species. The same character in a specimen of Paraplagiolophus codiciensis (MNHN, labelled ‘M. Guérin Cat. 20’) is associated with a more posteriorly expanded angular region and a lower, less recurved coronoid process. The condyle is similar to that of P. annectens but is thicker and 364 J. J. HOOKER tfe pop tfe Ye Text-figure 49 Plagiolophus, posterior cranial views. A, syntype of P. annectens (Owen) (29729) from Hordle Cliff. B, holotype of P. curtisi curtisi sp. & subsp. nov. (M26176) from Barton Cliff. Abbreviations: f = fossa; lf = lacrymal foramina; lp = lacrymal process; pop = postorbital process; tfe = edge of temporal fossa. x 1. has a sharp anterolateral ridge. Posteroventrally the beginning of the angular edge is orientated more like P. codiciensis than a species of Plagiolophus (Text-fig. 50). Tentatively referred radius: The proximal articulation measures 14-0mm anteroposteriorly and 23-3mm mediolaterally. Both are probably slight underestimates as the bone is somewhat eroded anteriorly and laterally. The maximum length of the incomplete shaft is 56-5 mm (PI. 25, fig. 2). It is very similar in size and morphology to radii of P. annectens from Hordle Cliff and La Débruge and in morphology to those of P. minor from La Débruge. All species of Palaeo- therium examined in comparison had a medially bowed radius, resulting in the proximal articulation being slightly oblique, compared to the nearly straight P. annectens radius. Palaeo- therium radii also tend to be shorter and broader but this feature cannot be observed on M34864. Plagiolophus curtisi creechensis subsp. nov. (Pl. 26, fig. 1; Text-figs 45, 46, 52C—D, 53; Table 28) v. 1977b Plagiolophus aff. sp. nov.; Hooker: 141. v. 1980 Plagiolophus sp. 1; Hooker & Insole: 42. Name. From the tithing of Creech in the parish of Church Knowle, overlooked by Creech- barrow. Hotorype. Associated upper and lower dentitions (M36181): left C', P?-M?; right P?-M?; left I,?, C,, P,-M,; right C,, P,-M,. Numerous indeterminate fragments, probably mainly from the jaws, are likely also to be associated. PARATYPES. (18): Upper canine (M37473); two P?s (M37474-5); two or three P*s (M37476-7, 2M37478); P* (M37479): four fragmentary upper molars, mainly if not all M‘/*s (M37480-3); two P,s (M36182, M37484); P; (M37485); two M,/.s (M37486-7); M3 (M37709); lower molar trigonid fragment (M37488). DOUBTFULLY REFERRED MATERIAL. Two incisors (M37489—90); left petrosal (M37491). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. D1aGnosis. 1, small (M? length 13-0mm); 2, P?-* protocones conical; 3, small metaconule on upper molar metaloph; 4, no accessory subterminal metastyle. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 365 DESCRIPTION. Canines: It has been shown above that the canines of M26176 are not splayed like those of other species of Plagiolophus but have essentially parallel orientation and the uppers are more laterally spaced than the lowers. Because of ectental cheek tooth motion in Plagiolophus, the striations produced on the transverse-vertical facets of the canines in the ‘splayed’ species run longitudinally on the uppers and transversely on the lowers (Text-fig. Text-figure 50 Ascending mandibular rami (shown as right) of Plagiolophus and Paraplagiolophus, x 0-75. A-B, Paraplagiolophus codiciensis (Gaudry) (MNHN unnumbered) from Jumencourt (Upper Calcaire Grossier); posterior and medial views respectively of left ascending ramus (reversed). C—D, Plagiolophus curtisi curtisi sp. & subsp. nov., holotype (M26176) from Barton Cliff; medial and posterior (condyle only) views respectively of fragmentary right ascending ramus. E-F, Plagiolophus annectens (Owen) (28229, Bravard Cat. no. G148) from La Debruge; posterior and medial views respectively of left ascending ramus (reversed). Abbreviation: df = dental foramen. 366 J. J. HOOKER Text-figure 51 Plagiolophus curtisi curtisi sp. & subsp. nov., holotype (M26176) from Barton Cliff. Partial reconstruction of skull, x 0-5. A, left lateral view of left and right composite. B, ventral view of cranium. C, dorsal view of mandible. Abbreviations: appf = anterior postpalatine foramen; iof = infraorbital foramen; mf= mental foramen; Pal = palatine; pof = preorbital fossa; pop = postorbital process; pppf = posterior postpalatine foramen. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 367 52A-B). Observation of striations on P. curtisi creechensis should confirm whether or not the canine angle was as in P. curtisi curtisi or as in the other species. Unfortunately no striations are preserved, although parts of the facets remain on some of the incomplete canines. The apical half of the left lower canine of the holotype preserves almost all the facet, including part of its basal margin towards the lingual side. Here the facet margin slopes buccobasally, leaving a lingual ridge unworn at the same crown level. If wear had been completely transverse, this ridge would almost certainly have sustained wear (Text-fig. 52C—D). It is thus tentatively concluded that wear was oblique and that the canines in the lower jaw at least had approx- imately parallel orientation as in P. curtisi curtisi. The lower canines are almost as large as those of M26176, whilst an upper (M37473) fits reasonably well in the remains of the upper alveoli of M26176. Upper premolars: P' is known to have existed from mesial interstitial facets on the P’s. Complete P?s from two individuals (M36181, M37474) show the metacone and mesostyle to be absent, but the paraconule to be a distinct cusp. On one the metaconule is joined to the distal cingulum, on the other it is isolated. The only complete P®* is the left one of the holotype. It has equal-sized paracone and metacone lying close together with the paracone bulging lingually more than the metacone, as in M26176. The large metaconule is linked by a narrow crest (missing on the right holotype P*) to the distal cingulum. There is no mesostyle. There are two individuals to show variation in P*. At similar wear stages, M36181 shows the metaconule joined to the protocone in a complete ‘crochet’ structure as in M26176, whereas in Text-figure 52 Occlusal views of canines in Plagiolophus showing orientations at beginning of buccal phase. A-B, P. fraasi Meyer from the Phosphorites du Quercy; C—D, P. curtisi creechensis subsp. nov. from Creechbarrow. A, Right upper canine (M1673); B, left lower canine (M1750); C, eft upper canine (reversed) (M37473); D, left lower canine. Wear facets of A and B show striations indicating direction of motion; oblique line delimiting lower edge of facet in D indicates suggested maximum transverse angle of mastication. Oblique hatching indicates broken surfaces. x 5. 368 J. J. HOOKER BARTONIAN MAMMALS OF HAMPSHIRE BASIN 369 M37479 the distal end is bent lingually at right angles, separated from the protocone, and a small bulge crosses the valley almost meeting the distal cingulum. The mesial end of the metaloph is simple in both specimens. M36181 is the only individual complete enough to show that, although a mesostyle is absent, a very faint ridge is present at the base and top of the ectoloph between the paracone and metacone. The cingulum is interrupted lingually on P?~+, but the contrast with P. curtisi curtisi is likely to be of an individual nature. Upper molars: Only the holotype shows distinguishing characters. The buccal cingulum is interrupted opposite the metacone on M?, otherwise it is complete although not very strong. Degree of separation of the paraconule from the protocone is very similar to M26176. The degree of distal expansion of M? is intermediate between that described above for M26176 and M26238, but similar to both in style. The subterminal metastyle is absent from this tooth in M36181, but the more distal basal swelling is present. Lower premolars: Available P,s have no mesial interstitial facet, so it is concluded that P, was absent. In three P,s there is slight variation in the size of the metaconid. It ranges from insignificant and style-like to three-quarters the height of the protoconid. There is a low, almost longitudinal metalophid which has a slight lingual bend at its distal end, the only recognizeable homologue of the hypoconid. The buccal cingulum varies from weak and interrupted to completely missing. The hypoconid is smaller on P; (M36181) than on P, and the whole talonid is much lower and narrower than the trigonid. The small size of the talonid corresponds to the proximity of the paracone and metacone of P? with which it occludes. Because of disto-occlusal flaring, the hypolophid of M37485 has been almost entirely removed by heavy wear. On the holotype left P, the talonid is slightly wider than, and nearly as high as, the trigonid. The latter is less open than on P;, being more like that of the molars but with a weaker paralophid. Lower molars: The shallowing of the valleys already mentioned under P. curtisi curtisi is hardly visible on M36181, but it is present in a curious form on an M, (M37709) which has the hypoconulid and part of the hypolophid broken away. Lying in the valley enclosed by the second crescent there is a transverse rib which merges lingually with a mesial extension of the base of the entoconid. Apart from slightly lower crown height, the lower molars are otherwise similar to those of P. annectens. Tentatively referred petrosal: The identification of this isolated bone with P. curtisi creechensis should be considered as no more than very tentative. The only perissodactyl genus in the Creechbarrow fauna the right size is Plagiolophus, but as so few published accounts of the ear region of European perissodactyls exist, the taxonomic relevance of morphological differences cannot easily be assessed. M37491 has been compared with an as yet undescribed petrosal in situ in a cranium of P. fraasi (M1733) from the Phosphorites du Quercy, two isolated petrosals from the Hordle Cliff Plate 26 Light macrographs of cheek teeth and x-rays of symphyses of Plagiolophus, ~ 1. Fig. la-f Plagiolophus curtisi creechensis sp. & subsp. nov., holotype (M36181) from Creechbarrow. a, buccal view of left P7>-M?; b, occlusal view of left P7-M?; c, occlusal view of right P?-M?; d, buccal view of left P,-M;; e, occlusal view of left P,-M;; f, lingual view of left P,-M;; note that hypoconulid lobe is broken on M3. See p. 364. Figs 2-3 Plagiolophus annectens (Owen) from Hordle Cliff. X-rays of mandibular symphyses. Fig. 2a, b, younger individual (BM(NH) 29730) in dorsal (a) and lateral (b) views. Fig. 3, older individual (BM(NH) 29705) in dorsal view. Fig. 4a, b Plagiolophus curtisi curtisi sp. & subsp. nov., holotype from Barton Cliff (M26176). X-ray of mandibular symphysis in (a) dorsal and (b) lateral views. See p. 356. 370 J. J. HOOKER Mammal Bed (R. Gardner private collection) and figures and descriptions of Pachynolophus livinierensis Savage, Russell & Louis 1965 (Savage et al. 1965: 48, 51-52, text-fig. 22; pl. 1) and Pachynolophus lavocati Remy 1972 (Remy 1972: 66-69, text-fig. 11; pl. 4). (Note that Savage et al. 1965: pl. 1 shows a right petrosal, not a left as stated). The specimen is worn and broken but its significant differences from the other petrosals mentioned above will be described. Terminology used here follows Savage et al. (1965), MacIntyre (1972) and Remy (1972). Most of the edges of the bone are broken away except for the posterior half of the lateral edge. The main position of the foramina and other structures are indicated in Text-fig. 53. In ventral view, the basic pattern of the foramina is the same in all the material mentioned but, in contrast to the others, in M37491 there is a facial canal bridge. The Fallopian hiatus is thus more anterior and occurs in an area of marginal breakage. The rim of the fenestra ovalis is also raised and, anteromedially of these two areas, there is left a distinct oval fossa in the area of the sulcus arteriosus promontorii. On the dorsal surface, the two depressions of the internal auditory meatus have an anterolateral/posteromedial spatial relationship as in M1733. The Fallopian aqueduct is over- hung by an anterior lip and is orientated anterodorsally. The utricular fossette occupies all the posterolateral half of the anterolateral depression. The cochlear fossette is large and the saccu- lar fossette posterolateral to it. The foramen of Morgagni is very small and sharp-rimmed and almost vertically perforates the posterolateral wall of the posteromedial depression. Lateral to the internal auditory meatus is the floccular (subarcuate) fossa, which in M1733 and the Hordle Cliff specimens is absent. Discussion of P. curtisi. Characters 6, 7 and 8 of the diagnosis of P. curtisi are shared with P. annectens, P. minor and P. fraasi of the Ludian. Characters 3, 4, 5, 12, 17, 18, 19 and 20 are shared with an undescribed species from the late Lutetian of Egerkingen y. Characters 10, 11 and 15 are intermediate between the two types; so are the proximity of the P* paracone and fen A Text-figure 53. Eroded left petrosal (M37491) from Creechbarrow, possibly belonging to Plagiolophus curtisi creechensis subsp. nov. A, ventral view; B, dorsal view. Abbreviations: cf = cochlear fossette; f = unnamed fossa; Fa = Fallopian aqueduct; Fac = Fallopian aqueduct (concealed by facial canal bridge); fen o = fenestra ovalis; fen r = fenestra rotunda; Fh = Fallopian hiatus; fir = floccular recess (subarcuate fossa); fM=foramen of Morgagni; fmm =fossa muscularis minor; IAM = internal auditory meatus; Pr = promontorium; sf = saccular fossette; uf = utricular foss- ette; VII = facial canal (probable course of VIIth (facial) nerve). Oblique hatching indicates broken surfaces. x 3-3. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 37 metacone and small size of the P, hypoconid. Characters 9 and 16 are unique to P. curtisi. The characters of the earlier species are considered primitive and those shared by the later species advanced. The intermediate characters fit well with the known Bartonian age of the nominate subspecies, whilst its unique characters suggest that it did not give rise to any other known species. The presence of P* metacone and P; hypoconid in this and the undescribed Egerkingen + species has relevance to an understanding of the affinities of Plagiolophus. It supports the derivation of this genus from a low-crowned palaeothere such as Propalaeotherium (referred to Equidae by Savage et al., 1965) by increase in crown height. The non-molariform P3 in advanced Plagiolophus can then be regarded as a secondary simplification, rather hem as evidence of derivation from a condylarth stock independent of Hyracotherium' (see Butler, 1952a, 1952b; Remy, 1967, 1976). Detailed discussion of this question will be covered elsewhere. When first studying the Creechbarrow and Barton Plagiolophus material, three alternative classifications appeared possible: 1, simply to consider all the material as one taxonomic unit at species level, on the basis of the unique characters listed above; 2, to separate the two assemblages as different species on the basis of a few minor characters; or 3, to describe them as one species but giving each assemblage a different subspecific name. The mixing of the assemblages resulting from solution 1 was abandoned as confusing. The small size of the assemblages does not give a clear enough idea of the constancy or variation of the dis- tinguishing characters, so solution 2 was abandoned. The apparently constant size difference between the two assemblages is only slightly less than that which occurs between two closely related sympatric species (P. annectens and P. minor from La Débruge), so solution 3 was used. Because both subspecies occur in the same general area (Creechbarrow and Barton are only 32km apart) and because they have not been found in continental European Bartonian sites (where nevertheless other species of Plagiolophus are known), it is logical to consider them as probably stratigraphical subspecies or stages in a single lineage. Of the characters which distinguish the two subspecies, 2 and 3 can be considered primitive in P. curtisi creechensis and advanced in P. curtisi curtisi, as they are shared with P. sp. from Egerkingen y and P. annectens respectively. Character 4 may also be an advanced character for P. curtisi curtisi, but it occurs in Auversian P. cartieri and not in any of the other species, older or younger. The evidence as a whole suggests a slightly earlier age for the Creechbarrow Limestone than for Barton Clay beds D/E. Genus PALAEOTHERIUM Cuvier 1804 TYPE SPECIES. P. magnum Cuvier 1804. Premiére Masse du Gypse, Montmartre, Paris, France. INCLUDED SPECIES. P. eocaenum Gervais 1875, P. siderolithicum (Pictet & Humbert 1869) Franzen 1968, P. lautricense Stehlin 1904a, P. ruetimeyeri Stehlin 1904b, P. pomeli Franzen 1968, P. castrense Noulet 1863, P. medium Cuvier 1804, P. crassum Cuvier 1805, P. renevieri Stehlin 1904b, P. muehlbergi Stehlin 1904b, P. curtum Cuvier 1812, P. duvalii Pomel 1853, P. crusafonti Casanovas-Cladellas 1975. RANGE. Late Lutetian—late Ludian, SUM es See Franzen (1968) for details of horizons and localities of the species. DIAGNOsIS. See Franzen (1968: 16). Palaeotherium aff. ?muehlbergi Stehlin 1904b (Pl. 27, fig. 1) vp. 1980 Palaeotherium sp.; Hooker & Insole: 43. LecrotyPe of P. muehlbergi. Left P?~* (Palaontologisches Institut und Museum Ziirich no. A/V 242). Ludian fissure filling, Obergosgen, Switzerland. RANGE of P. muehlbergi. Ludian; England, France, Switzerland and West Germany. D1aGnosis of P. muehlbergi. See Franzen (1968: 102, 104). J. J. HOOKER NX lo) BARTONIAN MAMMALS OF HAMPSHIRE BASIN 373 MATERIAL. Left mandibular fragment with distal fragment of M, and complete M; (M34865). HORIZON AND LOCALITY. Barton Clay Formation (not in situ), on high terrace of Barton Cliff below Naish Estate, near Barton-on-Sea, Hampshire; probably about SZ 223931. The jaw cavity is filled with brown concretionary claystone full of moulds of molluscs, mainly Turri- tella; there is no glauconite. The preservation thus suggests Bed F, which was exposed here at the time of collection in 1976 by Mr I.C. Daniels of Southbourne, Dorset. DESCRIPTION. The distal width of the remains of M, is 16-4mm. The M, is 35-9mm long; the trigonid width is 15-9 mm, the talonid width is 13-9 mm and the hypoconulid width 10-6 mm. The M, is moderately worn and exact crown height is unknown, but comparison with specimens at similar wear stages indicates that it is fairly low crowned for Palaeotherium. The crescents of M; are sharply convex, not blunt and rounded as in P. magnum. The hypoconulid lobe is moderately bent and the distal limb is orientated essentially lingually. The extent of wear on the hypoconulid and hypoconid suggests that when unworn the former was not much lower than the latter. The lingual cingulum is mainly strong but fades between the entoconid and hypoconulid. The buccal cingulum is weak and interrupted at the protoconid, hypoconid and hypoconulid. COMPARISONS AND DISCUSSION. Eight described species can approach the dimensions of the Barton specimen: P. pomeli (Marinesian), P. ruetimeyeri, P. castrense (late Lutetian— Marinesian), P. medium, P. crassum, P. curtum, P. renevieri and P. muehlbergi (Ludian). P. pomeli differs in having higher-crowned teeth and an M1-3 enlargement. P. medium and P. crassum have relatively longer and narrower M,s with a more continuous buccal cingulum. Moreover, the earlier subspecies, which are closer in age to the Barton specimen, are more different in size than the later ones. These species and P. curtum tend to have rounder, less angled crescents. P. renevieri is very similar to P. crassum. The remaining three species are very similar to M34865. Both P. ruetimeyeri and P. castrense have strong continuous buccal cingula, but on P. muehlbergi these can be weak and interrupted. P. muehlbergi also has characteristically angled crescents but the hypoconulid has a stronger and more mesially directed distal hook. Of the Marinesian species in the above list, the M, from Robiac, attributed to P. aff. ruetimeyeri ruetimeyeri by Franzen (1968: pl. 4, fig. 8), is closest in size, being only very slightly smaller. This too, though, has a continuous buccal cingulum. Franzen (1968: 109) considered the possibility that P. muehlbergi praecursum Franzen 1968 (the earliest subspecies from Hordle Cliff) could have evolved from P. ruetimeyeri, but that the typical and best known assemblage from Egerkingen « was very much older and more primi- tive than P. muehlbergi. He noted that the Robiac specimens did not provide enough informa- tion to support this derivation. The Barton specimen demonstrates the presence in the Marinesian of a P. muehlbergi-like species, but more complete material would be necessary to test out the above evolutionary hypothesis and to provide a more accurate identification. Palaeotherium sp. indet. (Pl. 27, fig. 2) vy. 1979 cf. Palaeotherium sp.; Kemp et al.: 102. vp. 1980 Palaeotherium sp.; Hooker & Insole: 43. MATERIAL. Lateral fragment of distal articulation of left humerus (M36491). Plate 27 Light macrographs of Palaeotherium and Lophiodon, x 1. Fig. la-c Palaeotherium aff. ?muehlbergi Stehlin (M34865) from Barton Cliff. Left mandibular fragment with M, and distal root of M, . a, buccal, b, occlusal and c, lingual views. See p. 371. Fig. 2a, b Palaeotherium sp. indet. (M36491) from Elmore. Mediodistal fragment of right humerus. a, anterior and b, distal views. See p. 373. Fig. 3a, b_ Lophiodon cf. cuvieri Filhol (GM.978110—3) from Elmore. Cast (M41977) of left M2, not sprayed with ammonium chloride. a, occlusal and b, buccal views. See p. 374. 374 J. J. HOOKER HoRIZON AND LOCALITY. Elmore Member, Barton Clay Formation (not in situ), foreshore at Elmore, near Lee-on-Solent, Hampshire. DESCRIPTION. No useful measurements can be taken of this fragment. It compares very well in size and morphology with the relevant part of a complete humerus of P. magnum stehlini Depéret 1917 (BM(NH) 29744a) from Hordle Cliff. The characters available, however, do not appear to be species characteristic. The only Palaeotherium species of the appropriate size known from the Auversian is P. castrense. Specific identification, however, cannot be made, although it seems unlikely that it was conspecific with the specimen from Barton. One must hope that this locality will eventually yield some dental remains of the genus. Suborder ANCYLOPODA Cope 1889 (rank emend. Radinsky 1964) (sensu Hooker 1984) Family LOPHIODONTIDAE Gill 1872 TYPE GENUS. Lophiodon Cuvier 1822. REMARKS. The classification by Radinsky (1964: 3-4, 7-9) and Savage et al. (1966: 31-35) of Lophiaspis Depéret 1910 and Paleomoropus Radinsky 1964 in the family Eomoropidae Matthew 1929a is followed here, rather than that by Fischer (1964: 62-64, 67, fig. 21; 1977a: 917-918), who included them in the Lophiodontidae. RANGE. Same as Lophiodon (q.Vv.). Genus LOPHIODON Cuvier 1822 TYPE SPECIES. Palaeotherium tapiroides Cuvier 1812. Middle Lutetian; Bouxwiller, Alsace, France. INCLUDED SPECIES. L. remensis Lemoine 1878, L. tapirotherium Desmarest 1822, L.? buchsowilla- num Desmarest 1822, L. parisiense Gervais 1852, L. cuvieri Filhol 1888a, L. thomasi Déperet 1906, L.? isselense Fischer 1829, L. rhinocerodes Rutimeyer 1862, L. lautricense Noulet 1851, L. sardus Bosco 1902, L.? leptorhynchum Filhol 1888a. For additional doubtful species see Fischer (1964: 67). For differing views of species, see Fischer (1964, 1977a, 1977b), Jaeger (1971) and Sudre (1971). Discussion. Dedieu (1977, June) placed L. buchsowillanum (type species), L. isselense, L. leptor- hynchum and L. compactus Astre 1960 (the last raised to species level) in his new genus Paralophiodon, which he referred to a new monotypic subfamily Paralophiodontinae of the family Isectolophidae. Fischer (1977a, July) placed L. buchsowillanum in his new monotypic lophiodontid genus Rhinocerolophiodon (thus a junior objective synonym of Paralophiodon) and referred L. leptorhynchum to Lophiaspis. Until more detailed evidence is published, a conserva- tive approach is adopted here. RANGE. Late Ypresian to Bartonian (Marinesian), Europe and possibly south-west Asia. DIAGNOsIS. See Fischer (1977b: 1129). Lophiodon cf. cuvieri Filhol 1888a (PI. 27, fig. 3) v. 1980 Lophiodon cf. cuvieri Filhol; Hooker & Insole: 43. SynTypes of L. cuvieri. Cranium and left mandibular ramus. Upper Calcaire Grossier (late Lutetian); Jouy, Marne, France (Filhol 1888a: 144-154; pls 14-15; pl. 16, figs 3-5; pl. 17, fig. 5). Neither Schroeder (1916) nor Fischer (1964) were able to find the specimens although they were stated to be in the Sorbonne, Paris by Filhol (1888a: 144) and Stehlin (1903: 116). Lophiodon cuvieri Watelet 1864 is a nomen nudum, as this author gave no figures or any information which can be construed as a description or even mention of a single distinguishing feature. The species must therefore take its authorship and date from Filhol (1888a). BARTONIAN MAMMALS OF HAMPSHIRE BASIN 37/5) RANGE of L. cuvieri. Early or middle Lutetian, D.D.R.; ?Lutetian, B.R.D. (see Tobien 1961) (L. cf. cuvieri from the Bartonian Gehlbergschichten of Helmstedt is thought to be reworked, see Russell et al., 1982: 13); middle to late Lutetian, ?7Auversian, France; late Lutetian to ?Auver- sian, Switzerland; ?Auversian, Britain. EMENDED DIAGNOSIS of L. cuvieri (translated and slightly modified from Fischer 1964; additions are italicized). Medium-sized species. Maxillary teeth large and rounded (M! length 31mm). Preultimate upper molars slightly broader than long. Cingula reduced to a minimum. Upper molars with completely tapiroid metacone (the convexity on the outer wall is strong); the parastyle appears rounded and closely fitted to the paracone. Upper premolars with or without small, poorly differentiated hypocone. The base of the lingual side of the incisors mostly covered with very strong enamel wrinkles and papillae. MATERIAL. Left M2, not in situ, but almost certainly from the Elmore Member, Barton Clay Formation, foreshore at Elmore, near Lee-on-Solent, Hampshire. The specimen is in Gosport Museum (GM 978110-3) and a cast (M41977) in the BM(NH). Mr DJ. Kemp of Gosport Museum collected the specimen and presented the cast to the BM(NH). DESCRIPTION. The tooth is moderately worn so that dentine is continuously exposed from parastyle to protocone and from ectoloph to hypocone. The mesial edge of the tooth is mainly missing as far back as the long axis of the protoloph, except for the buccal region of the parastyle and a small part at the mid-point of the mesial edge below the crown base. Estimated maximum length perpendicular to the mesial edge is 38-5mm. Maximum width parallel to the mesial edge is 41:7mm. Measurements of M? from Filhol’s (1888a) pl. 16, figs 3-4 of the syntype cranium of L. cuvieri are 36-4mm long and 42-0mm wide, so size and length/width ratio are very similar to this species. The size, shape and proximity of the parastyle to the paracone is also similar to Filhol’s figures of L. cuvieri. Only the buccal cingulum 1s slightly more extensive: it is weakly developed between the parastyle and paracone and is present buccal to the metacone, continuing distally to meet the distal cingulum. Although M? in Lophiodon is less distinctive of the ‘tapiroid’ versus ‘rhinocerotoid’ condition than are the upper preultimate molars, the Elmore M? can be distinguished from that of most other similar-sized species as follows. In L. isselense it is relatively longer and narrower and the parastyle is separated from the paracone; in L. tapiroides, it is relatively longer and narrower and the parastyle, although no more detached than in L. cuvieri, is mesiodistally compressed. In L. parisiense it is very similar morphologically, although slightly but significantly smaller. L. thomasi from the French Bartonian is probably the species most similar to L. cuvieri, being distinguished only by the presence of a distinct P* hypocone. The unique holotype consists of an upper incisor, part of P??, complete P*, and M!? only, so comparison with the Elmore M? is not possible. Whereas specific identity with either L. cuvieri or L. thomasi is equally possible, the determination as L. cf. cuvieri is considered the more practical alternative. Discussion. When Depeéret (1906) first described L. thomasi from the Calcaire de Ducy of Sergy, Aisne, France (see Thomas 1906 for locality details), he distinguished it from L. parisiense by its P* hypocone, slightly less tapiroid M’ ? and slightly larger size. He did not compare it with L. cuvieri, from which it differs only in its distinct P* hypocone. Fischer (1964: 67, text-fig. 21) considered that L. thomasi was derived from L. parisiense, but L. cuvieri seems a better candi- date for ancestry both morphologically and stratigraphically. Stehlin (1903: 128-146) attributed various cheek teeth from Egerkingen to L. cuvieri or ‘L. cuvieri?. Amongst these he described (1903: 135-136) several P*s (as ‘P”’) and P*s (as ‘P"’), mainly from fissure «, as “L. cuvieri? and as having incipient hypocones. They appear thus to form good morphological intermediates between L. cuvieri and L. thomasi, which, if phylogen- etically intermediate, ought to be of Auversian age. Similar incipient hypocone development on P’*s and a P® attributed to L. cuvieri from the French Auversian locality of Arcis-le-Ponsart (Louis 1976: 50) would seem to confirm both the postulated ancestor—descendant relationship and an Auversian age for Egerkingen «. On the other hand, Sudre (1971: 92) noted beginnings 376 J. J. HOOKER of hypocone formation on a P? and P® of L. cuvieri from the late Lutetian (similar horizon to the types) of Pargny. So it would appear that large assemblages are necessary to account for individual variation (cf. Sudre, 1971, for L. lautricense). Upper premolars are eagerly awaited from the sparsely mammaliferous Elmore site. Lophiodon cf. lautricense Noulet 1851 v. 1977a Lophiodon cf. lautricense Noulet; Hooker: 91—94. v. 1980 Lophiodon cf. lautricense Noulet; Hooker & Insole: 43. Ho.otyPe of L. lautricense. Mandible with broken teeth. Sables du Castrais; Braconnac, near Lautrec, Tarn, France (Muséum d Histoire Naturelle de Toulouse, unnumbered). RANGE of L. lautricense. Marinesian; France, Switzerland, West Germany. DiaGnosis of L. lautricense. No formal diagnosis is available, but there are comprehensive accounts by Sudre (1971) and Stehlin (1903). MATERIAL. Right symphysial fragment with roots of P, , and root and alveolus of P, (BGS GSM88617; cast in BM(NH), M31996). For details see Hooker (1977a). Order ARTIODACTYLA Owen 1848b Suborder PALAEODONTA Matthew 1929b Superfamily DICHOBUNOIDEA Gill 1872 (rank emend. Viret 1961) Family DICHOBUNIDAE Gill 1872 (sensu Sudre 1978b) TYPE GENUS. Dichobune Cuvier 1822. INCLUDED GENERA. Protodichobune Lemoine 1891; Aumelasia Sudre 1980; Buxobune Sudre 1978b; Meniscodon Ritimeyer 1888; Mouillacitherium Filhol 1882; Hyperdichobune Stehlin 1910b; Metriotherium Filhol 1882; Synaphodus Pomel 1848; Messelobunodon Franzen 1980a; ?Lantianius Chow 1964; ?Chorlakkia Gingerich, Russell, Sigogneau-Russell & Hartenberger 1979. RANGE. Late Ypresian—Chattian, Europe; middle to late Eocene, Asia? DIAGNOSIS. See Sudre (1978b: 18). Genus HYPERDICHOBUNE Stehlin 19106 TYPE SPECIES. Dichobune spinifera Stehlin 1906. Ludian, Entreroches, Canton Vaud, Switzer- land. INCLUDED SPECIES. H. nobilis (Stehlin 1906) Stehlin 1910b; H. spectabilis Stehlin 1910b; H. hammeli Sudre 1978b; ?H. langi (Rutimeyer 1891) Sudre 1972; H. sp. 1 and ?H. sp. 2 herein. RANGE. M. Lutetian—Auversian, France; late Lutetian—late Ludian, Switzerland; Marinesian, England. DiaGnosis. See Sudre (1978b: 35). Hyperdichobune sp. | (Pl. 28; Table 29) v. 1977b ?Hyperdichobune spectabilis Stehlin; Hooker: 141. v. 1980 Hyperdichobune sp.; Hooker & Insole: 43. MATERIAL. P* (M36800); three M‘/*s (M36196, M37492, M37710); two M?s (M37493-4); DP* (M37459); M,,, talonid fragment (M37553); M, (M36815); lower molar trigonid fragment ? (M36195); and DP, talonid fragment (M37496). cc popac premec ypid Text-figure 54 Dental nomenclature of artiodactyl cheek teeth, modified considerably from Coombs & Coombs (1977: 585, fig. 1). A, left upper molar; buccal edge towards top of page, lingual towards bottom. B, left lower molar, and C, left lower fourth deciduous premolar; buccal edges towards bottom of page, lingual edges towards top. Abbreviations: A. cale—central accessory conule B & C. bpacd—buccal paracristid cc—centrocrista cdo—cristid obliqua eclm—ectocingulum eccld—ectocingulid hyp—hypocone ectstd—ectostylid mest—mesostyle met—metacone metle—metaconule mets—metastyle par—paracone parle—paraconule past—parastyle poclm—postcingulum pomec—postmetacrista pometlc—postmetaconule crista popac—postparacrista poparlc—postparaconule crista poprc—postprotocrista poprclm—postprotocingulum preclm—precingulum premec—premetacrista premetlc—premetaconule crista prepac—preparacrista preparlc—preparaconule crista preprc—preprotocrista prot—protocone prphle—protolophule 377 entd—entoconid entld—entoconulid entstd—entostylid hypd—hypoconid hypld—hypoconulid lpacd—lingual paracristid metd—metaconid metstd—metastylid pacd—paracristid pad—paraconid pasd—parastylid pocd—postcristid pocld—postcingulid poentcd—postentocristid pomecd—postmetacristid prcd—protocristid precld—precingulid prentcd—pre-entocristid protd—protoconid pspad—pseudoparaconid (of Hershkovitz, 1971). 378 J. J. HOOKER Table 29 Length (1) and trigonid (w,) and talonid (w,) width measurements of teeth of Hyper- dichobune sp. 1 from Creechbarrow. Two width measurements are only given for lower molariform teeth. Measurements in millimetres. No. Tooth l Ww W> M36800 RP* (3-7) 3-5 M37710 LM‘? (4-1) - M36196 RM‘? ~ (4-5) 4-8 M37493 ~=LM? 45 (5-0) M37494. =RM? (4-8) 5-4 M37492 UM (4-5) 4-4 M37459 RDP* (3-8) (4-0) M37553 RM, - - 3:3 M36815 RM, - 3-0 2-6 M37496 ~=6 LDP, - ~ 2:3 HORIZON AND LOCALITY. Creechbarrow Limestone Formation. Creechbarrow. DESCRIPTION. P*: The parastyle, base of paracone and preprotocrista are broken away. There is a large, distally-placed metacone two-thirds the size of the paracone. The distal cingulum is weak and swollen somewhat in the middle. An incomplete buccal cingulum is preserved round the metacone (Pl. 28, fig. 1). H. spectabilis from Entreroches (NMB Mt723) has a smaller metacone closer to the paracone; continuous buccal cingulum; stronger, more extensive distal cingulum with larger bulge forming incipient metaconule; and larger protocone with proto- lophule joining paracone. M'?: M36196 is heavily worn and very weathered (Pl. 28, fig. 3). It has a large, distally skewed mesostyle; a strong buccal cingulum only between the mesostyle and metastyle; and apparently no lingual cingulum. It evidently had a fairly large hypocone which is completely fused by wear to the metaconule and protocone. The outline is somewhat oblique, the lingual cusps being slightly more mesial than the buccal cusps. Of M37710, only the lingual half remains (PI. 28, fig. 6). The paraconule is small, close to the protocone, does not bulge mesially and there is a distinct postparaconule crista. The metaconule is nearly as large as the protocone and the hypocone is much smaller. What remains of the outline is similar to M36196. M37492 lacks the metaconal region, is smaller than M36196 or M37710 and may be an M! (PI. 28, fig. 2). It is heavily worn but its development of lingual cusps is very similar to that of M37710. The mesial margin is straight and the parastyle is prominent. H. spectabilis M‘~* (Mt723) are similar in position of mesostyle, prominence of postparaconule crista and outline; but the hypocone is larger and there is a distinct mesial bulge to the paraconule. M?: M37493 (PI. 28, fig. 7) is a very corroded tooth and thus difficult to compare character for character with M37494 (PI. 28, fig. 4), whose parastylar area is broken. The latter, however, is slightly larger, seems to have a more rounded outline, stronger cingula and more mesially bulging paraconule. It is thus more like H. spectabilis than is M37493, whose paraconule is more like that of the M‘'/*s, M37710 and M37492. Both M?s have postparaconule cristae and no hypocone, but are relatively shorter teeth than the two known M?s of H. spectabilis (Mt723 and Mt396, the latter figured by Stehlin, 19106: 1100, text-fig. 225). DP*: Like M37492, M37459 has lost the metaconal region; it is quite worn, has thinner enamel and is lower-crowned than the molars; and it has a strong parastyle and postpara- conule crista (Pl. 28, fig. 5). It appears to have been narrower mesially than distally and there is a strong cingulum round the more buccal protocone. It has more the proportions of the DP* of Mouillacitherium cf. elegans from Mormont (Stehlin 1906: pl. 12, fig. 30) than that of H. spinifera (ibid.: pl. 12, fig. 40). BARTONIAN MAMMALS OF HAMPSHIRE BASIN 379 M,,2: This is very similar to the holotype M,_, of H. spectabilis, but the hypoconulid is a larger, distally-projecting cusp and there is no crest joining the cristid obliqua to the entoconid (Pl. 28, fig. 10). The wear state of the two specimens is very similar. M,: The hypoconulid lobe is broken away and the tooth is very heavily worn (PI. 28, fig. 11). The characters that remain visible are similar to H. spectabilis but like the M,,, there is no crest joining the cristid obliqua to the entoconid. M,/2/3: This doubtfully attributed trigonid fragment has a distinct paraconid slightly smaller than the metaconid but close to it (PI. 28, fig. 9). Whether H. spectabilis had the same character is difficult to tell as the teeth are much more worn. However, the holotype M,; has two deep areas of dentine separated by a shallower dentine isthmus; this suggests that intervening enamel had, at only a slightly earlier wear stage, separated a paraconid from a metaconid. | DP,: This talonid fragment is like the M,,, but smaller, lower-crowned and with thinner enamel, its hypoconulid is more vertically orientated and its outline begins to taper mesially (PI. 28, fig. 8). Discussion. Stehlin (1910b) could see no difference except size between H. spinifera and H. spectabilis, both from Entreroches. The only tooth types common to both, however, are P*-M! and P,—-M.,. H. spinifera seems nevertheless to have an M’ with paraconule not mesially salient and weaker cingula; a P* with a weaker incipient metaconule and better separated paracone and metacone and no protolophule; and a lower molar without a crest joining the cristid obliqua to the entoconid. From this evaluation and the above description, the Creechbarrow material appears to incorporate characters of both H. spinifera and H. spectabilis. This could mean that two species each related to those from Entreroches are represented, or that the morphological characters distinguishing the meagre sample of H. spinifera and H. spectabilis reflect intraspecific variation in a monospecific Creechbarrow assemblage, if not also in the Entreroches assemblages. More specimens are required to test these ideas. The Creechbarrow teeth do not belong to either H. hammeli, H. nobilis or ?H. langi. The first has a narrower P* with small protocone and M'” apparently with no postparaconule crista. The type series, however, appears to be composite: NMB Bchs430 with its large hypocone and overall arrangements of cusps (Sudre 1978b: text-pl. 1, fig. 5) fits the lipotyphlan Amphilemur leemanni (whose type locality, Bouxwiller, is the same as that of H. hammeli); also the absence of a P, hypoconid on the holotype makes it unlikely that the P* would have had subequal paracone and metacone like the paratype UM BUX66166 (Sudre 19785: text-pl. 1, fig. 6). Stehlin (1910b: 1101) distinguished H. nobilis from H. spinifera mainly on the former’s less molariform referred P* (Stehlin 1906: pl. 12, fig. 25), in which it also differs from H. sp. 1. °H. langi is a very different species which may not even belong to this genus (see below under 2H. sp. 2). The M’” from Lissieu, thought at first by Sudre (1972: 127, text-fig. SC) to be H. nobilis and later (Sudre 1978b: 36) to be specifically indeterminate, is similar to the Creechbarrow molars, but from the figure appears to lack a postparaconule crista. ?Hyperdichobune sp. 2 (PI. 29, fig. 6) v. 1977b Rhagatherium sp.; Hooker: 141. v. 1980 Rhagatherium sp.; Hooker & Insole: 44. MATERIAL. Left M‘/? (M37546), right M‘/?/? lingual half (M36187); and doubtfully a right M‘/? metaconal fragment (M37547). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION. M37546 is triangular in outline, slightly broader than long, with the mesiobuccal corner a right angle. Outline and cusp pattern are very similar to Mixtotherium but with very small hypocone and a paraconule well separated from the protocone by a fissure. The centro- crista is buccally flexed and there is a large mesostyle. The paracone and metacone are buccally 380 J. J. HOOKER BARTONIAN MAMMALS OF HAMPSHIRE BASIN 381 salient and there is no buccal cingulum. The cusps generally appear gracile and are separated by broad deep valleys. The premetaconule crista is short, curving back to join the metacone mesiolingually. The tooth is 5-6mm long and 6-1 mm wide. In M36187, the hypocone is little more than a papilla at the distobuccal end of the post- cingulum. M37547 shows the metastyle joined to the mesostyle by a buccal cingulum. Discussion. The affinities are somewhat enigmatic. Strong similarities can be seen with ?H yper- dichobune langi from Egerkingen on the one hand and with Mixtotherium on the other. Deci- sions are hampered by only one tooth type being known for ?Hyperdichobune sp. 2. It could be doubted that ?H. langi is a Hyperdichobune for the following reasons. Firstly, on M? the near triangular outline tapers distally, the mesiobuccal corner being acute; second, the upper molar paraconule is large and well separated from the protocone; and third, on the upper molars the very weak postparaconule crista is directed towards the mesostyle, not towards the middle of the paracone. No premolars were recorded for “Dichobune’ langi by Stehlin (1906: 623-625), nor did he (19106) refer it with the other three species to his new genus Hyperdichobune. Sudre (1972: 121-126, text-figs 3-4) referred upper and lower molars from Lissieu to the species and included it in Hyperdichobune on the basis of an isolated P*. The P* was later considered (Sudre 1978b: 36) to be wrongly identified and to require replacing by another one that he had referred to H. nobilis (Sudre 1972: 127, text-fig. 5B). The latter tooth is like P* of Mixtotherium in its more sharply triangular outline, buccal cingulum round the metacone, large, mesially salient para- style and small paraconule and metaconule. If it really belongs to ?H. langi, it suggests a close relationship with Mixtotherium; alternatively it may belong to Mixtotherium infans. 2H. langi M? differs from that of ?H. sp. 2 from Creechbarrow mainly in being smaller, relatively shorter and having a larger hypocone: Mixtotherium differs from ?H. sp. 2 in having a strong buccal cingulum, less buccally salient paracone and metacone, the paraconule fused to the protocone and no hypocone. The intermediate state of ?H. sp. 2 suggests affinities between 2H. langi and Mixtotherium and possible derivation of the latter from the Dichobunidae. The absence of either premolars associated with molars or of any known basicranial in ?H. langi means that we do not know whether any of the other distinctive features of Mixtotherium, such as absence of mastoid exposure, had already evolved in these dichobunids. When better known a new generic name may be found necessary for ?H. angi and ?H. sp. 2. Family MIXTOTHERIDAE Pearson 1927 TYPE AND ONLY KNOWN GENUS. Mixtotherium Filhol 1880b. See also comments under ?H yper- dichobune sp. 2, above. Genus MIXTOTHERIUM Filhol 1880b (Complicated synonymy dealt with by Stehlin, 1908: 799-828). TYPE SPECIES. M. cuspidatum Filhol 1880b. Phosphorites du Quercy; France (figured Filhol, 1882: pl. 9, figs 1-7). INCLUDED SPECIES. M. gresslyi Rutimeyer 1891 (including M. priscum Stehlin 1908); M. infans Stehlin 1910b; M. leenhardti Stehlin 1908; M. depressum (Filhol 1884) Stehlin 1908; M. quercyi (Filhol 1888) Stehlin 1908. Plate 28 Light macrographs of cheek teeth of Hyperdichobune sp. 1 from Creechbarrow, x 9. Occlusal views, except Fig. 1a. Fig. la, b, right P* (reversed) (M36800). a, buccal and b, occlusal views. Fig. 2, left M‘/? (M37492). Fig. 3, right M’/? (reversed) (M36196). Fig. 4, right M? (reversed) (M37494). Fig. 5, right DP* (reversed) (M37459). Fig. 6, lingual half of left M'/? (M37710). Fig. 7, left M* (M37493). Fig. 8, talonid fragment of left DP, (M37496). Fig. 9, trigonid fragment of right M,,,,, (reversed) (M36195). Fig. 10, talonid fragment of right M, . (reversed) (M37553). Fig. 11, right M, with broken hypoconulid lobe (reversed) (M36815). See p. 376. 382 J. J. HOOKER RANGE. Late Lutetian—Ludian; England, France and Switzerland. DIAGNOsIS. See Viret (1961: 902). DIscuSSION. This genus is very local in occurrence, being almost entirely restricted to the fissure deposits of Switzerland and France, although it is not uncommon where it occurs. It was recently discovered in the Upper Calcaire Grossier of the Paris Basin (Ginsburg et al. 1977) and the Creechbarrow records are the first for England. It has been found in the recent excavations in Le Quercy at the early Ludian sites of La Bouffie and Perriére (Crochet et al. 1981). Once considered related to anoplotheres or anthracotheres, the genus has sometimes been included in the Cebochoeridae (e.g. Sudre 1972; Simpson 1945) because of superficially similar skull shape and lack of mastoid exposure; its possible relationships with the dichobunids are discussed herein under Hyperdichobune above. Of the described species, only M. cuspidatum, M. gresslyi and M. infans are well character- ized. M. quercyi and M. depressum appear very close to typical M. cuspidatum, whilst M. leenhardti and M. priscum are difficult to separate from M. gresslyi. M. cf. priscum and M. cf. gresslyi from Mormont and Quercy were tentatively separated by Stehlin (1908) from the typical assemblages of these species on stratigraphical, not morphological grounds. It is prob- a) 4 5 6 7] 8 ! Text-figure 55 Scatter diagram of length (1) against width (w) in P*-M®? of Mixtotherium. D = M. ‘priscum’ Stehlin from Egerkingen Huppersand; V = M. ‘priscum’ from Egerkingen y; A= M. gresslyi Rutimeyer and M. infans Stehlin from Egerkingen «; <] = M. ‘priscum’ from Lissieu; © = M. ‘cf. priscum’ from the Phosphorites du Quercy; © = M. aff. gresslyi from Creechbarrow; 47 = M. ‘cf. gresslyi’ from Mormont. * indicates specimens where the parastyle is broken. H = holotype of M. priscum; L = lectotype of M. gresslyi; S = syntype of M. infans. Symbols solid above = P*; solid on left = M!; solid on right = M?; completely solid = M?; and outline = M'. Measurements in millimetres. Lines join teeth of one individual. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 383 able that the slightly deeper horizontal mandibular ramus of M. leenhardti (Stehlin 1908: text-fig. 129) could result from old age of the individual, as the teeth appear quite worn, compared for instance to his text-fig. 131 of M. cf. priscum. Gaining an accurate idea of the range of intraspecific variation in this genus is difficult in view of the small number of speci- mens and the poor locality and stratigraphical data for the old Quercy material. The larger collections from Egerkingen, including those from Egerkingen y made after Stehlin’s pub- lication of Mixtotherium, offer better potential. The Creechbarrow material, although not including osteological parts, consists of enough teeth to demonstrate variation for a number of features. A revision of the complete genus is highly desirable. From the scatter diagram (Text-fig. 55), the Creechbarrow material makes a plot interme- diate between M. gresslyi and M. priscum according to Stehlin’s concepts. However, Stehlin (1908: 822) also thought that M. priscum might in the light of further material be divided into two or three species. He was struck by its high morphological variation. He could not convinc- ingly separate it morphologically from M. gresslyi except by referring the very small individual from Egerkingen « (NMB Eg159; Stehlin 1910b: pl. 17, fig. 21), with more quadrate M' and elongated P*, to the latter species. In Text-fig. 55, Eg159 plots closer to the figured syntype M? of M. infans than to the type M!' of M. gresslyi. Sudre (1972: 133-134, fig. 9) described and figured M. gresslyi from Lissieu, but in size his material (isolated upper molars) is close to Eg159 and M. infans. His figures suggest that in morphology it differs too; its high selenodonty, deeply incised parastyles and in one case (1972: fig. 9c) absence of paraconule more closely resemble the xiphodont Haplomeryx. Text-fig. 55 shows no clear size difference between M. gresslyi and M. priscum if one excludes Eg159. M. priscum from Egerkingen « can easily be accommodated within M. gresslyi, whose type is from the same fissure. The assemblages of M. priscum from Egerkingen Huppersand (grey marl facies) and Egerkingen y appear slightly larger, but more specimens would almost certainly produce a normal range overlapping exten- sively with M. gresslyi. The differences are no greater than would be expected from assemblages of one species from different localities or slightly different stratigraphical horizons. It is pro- posed here that the species M. priscum be synonymized with M. gresslyi, whilst recognizing that a subspecific difference might be demonstrated if a systematic study of the Egerkingen material were undertaken. Table 30 compares coefficients of variation of various combinations of the different assemblages. The Creechbarrow assemblage (here referred to as M. aff. gresslyi) is intermediate in size between the Egerkingen y/Huppersand assemblages (‘M. priscum’) and the Egerkingen « assemblage (type M. gresslyi). It is possibly closer to the latter. Mixtotherium aff. gresslyi Rutimeyer 1891 (Pl. 29, figs 1-4; Pl. 30; Text-fig. 55; Tables 30-33) v. 1977b Mixtotherium priscum Stehlin; Hooker: 141. v. 1980 Mixtotherium priscum Stehlin; Hooker & Insole: 44. Lectotype of M. gresslyi (following Sudre 1972: 133, as ‘type’). Right maxillary fragment with P?-M!. Auversian, Egerkingen «; Canton Solothurn, Switzerland (figured Riitimeyer 1891: pl. 6, fig. 1 and Stehlin 1908: 817, text-fig. 129). Table 30 Coefficients of variation (v) for length of M’/? in Mixtotherium assemblages. N = number of specimens. Assemblages N Vv M. gresslyi and M. ‘priscum’ from Egerkingen « 4 35/3) Ditto and M. ‘cf. gressly? from Mormont 3) 6:16 Ditto and M. aff. gresslyi from Creechbarrow 11 4-68 Ditto and M. ‘priscum’ from Egerkingen y and Huppersand 14 7:34 M. ‘priscum’ from Egerkingen y and Huppersand and M. aff. gresslyi from Creechbarrow M. aff. gresslyi from Creechbarrow 384 J. J. HOOKER MATERIAL. Two P?s, broken mesially (M37505, M37559); two P*s (M37506—7); five M!/2s, one fragmentary (M36184, M37509-10, M37514, M37512); seven M?s, three fragmentary (M36185-— 6, M37508, M37515, M37511, M37513, M37516); three fragmentary upper molars (M36818, M37517-8); two P,,, talonid fragments (M36438, M37560); two P3s lacking mesial half (M36797, M37558); two P,s lacking talonids (M37519—20); seven M,,, talonid fragments (M36193, M36796, M36809-10, M37522-4); two M, talonid fragments (M36408, M36500); nine lower molar trigonid fragments (M36188—-9, M36198, M36827, M37521, M37525-7, M37711); two DP,? fragments (M36416, M36808); DP? talonid fragment (M36194); two DP, trigonid fragments (M36411, M37550). HorIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. RANGE OF THE SPECIES (including aff. and cf.). Late Lutetian, France and Switzerland; Auversian, Switzerland; Marinesian, England and Switzerland; and ?Marinesian and Ludian, France. EMENDED DIAGNOSIS OF SPECIES. Medium-sized species (mean length of M' = 65mm); upper molars broader than long (length c. 80% of width); depth of adult mandibular ramus beneath the distal end of M, 2:5 times length of M,. DIFFERENTIAL DIAGNOSIS. M. cuspidatum is larger; has upper molars with approximately equal length and width; and deeper mandible. M. infans is smaller. Table 32 Trigonid (w,) and talonid (w,) Table 31 Length (1) and width (w) measure- width measurements of lower teeth of Mix- ments of upper teeth of Mixtotherium aff. totherium aff. gresslyi from Creechbarrow. It gresslyi from Creechbarrow. Measurements was not possible to measure length in any of in millimetres. *indicates parastyle missing. these teeth. Measurements in millimetres. No. Tooth ] Ww No. Tooth WwW, W> M37505 RP? - 3-9 M36438 EPS, = 3-0 M37559 RP? E 4-0 M36797 LP, ~ 3-5 M37506 LP* 6:7 6-2 M37558 LP, = 3-5 M37507 RP* 6:4 59 M37560 RP, = 2-7 M36184 LM"? 6-0 69 M37519 Bi, 3-4 o M37509 LM!/2* 61 79 M37520 RP, 3-4 = M37510 LM1/2* 5-8 75 M36796 LM,» a 45 M37514 RM!” 6:0 - M37522 LM, a 4-2 M36185 LM? 6:3 713 M6193 RM,» S 4] M37508 LM?* 5-7 7 M36809 RM, ,, F 3-6 M37515 LM?* 6:2 8-0 M37523 RM, 5 = 3-8 M36186 RM? 6:3 7:8 M37524 RM, = 4-7 a a > a eS et M36500 LM, = 4:5 M36189 EM jap 4-0 B M37521 LM Gay 4-2 = M37525 EMaee 4-0 2 M37526 EM at ais 4-5 ze M36188 RM, 3 4-1 : M37527 RM, 35 4-2 = M37711 RM, 3-9 _ M36808 LDP,? 2-0 - M36416 RDP,? 1:9 - M36194 LDP,? _ 27 M37550 RDP, 3-0 = BARTONIAN MAMMALS OF HAMPSHIRE BASIN 385 DESCRIPTION. Comparison is made with a maxilla from the Quercy Phosphorites (M33522), casts of a maxilla (NMB Eh409) and mandibular ramus (Eh424) from Egerkingen y and original M? (M12017), mandibular ramus (M12015) and M, (M12016) from Egerkingen undif- ferentiated (although the ochreous sandy bolus matrix suggests fissure y also); as well as with Stehlin’s figures (1908, 19105). P?: Both teeth are broken mesially through the paracone. Their metacone is larger and better separated from the paracone than on Eh409 and M33522. In fact the metacone is nearly as large as the paracone. The protocone is also larger in M37559 than on these two and M37505. P*: Both have strongly buccally flexed centrocristae and strong mesostyles — more so than M33522. There is a strong buccal cingulum on M37506 behind the mesostyle, which broadens the stylar shelf and pushes the metacone further from the buccal margin than is the case in M37507. In the latter the buccal cingulum is restricted to the mesial side of the metastyle. The metaconule is larger in M37507 and has a premetaconule crista joining it to the metacone at a mesiolingual position. The parastyle is more mesially salient in M37506 than M37507. Both have overall smaller parastyles than M33522. Upper molars: See Table 33 for the distribution of the main variable characters. There is a partial positive correlation between length of the premetaconule crista and the proximity of the paraconule to the protocone; otherwise the characters appear randomly distributed. There is nothing to suggest that more than one species is represented. M? can generally be distinguished from M' or M? (when absence of a recognizable distal interstitial facet is unreliable), by being relatively longer and having a more lingually displaced metacone. P, 3: Two specimens from Egerkingen y show P; to be very variable in this assemblage. One (M12015) resembles the P, of another (Eh424) in the small size of the hypoconid, absence of a metaconid and shallow lingual concavity of the protoconid—paraconid—parastylid crest. The P; of Eh424, on the other hand, has a small distolingual metaconid joined to the protoconid by an oblique protocristid; the larger hypoconid is slightly displaced buccally; and the protoconid— paraconid—parastylid crest is more deeply concave lingually. On this basis, the distal fragment (M36438) from Creechbarrow could be either a P, or a simple P;. Size favours it being a P,. On this basis, it could also be argued that the equal-sized M37560 (also a distal fragment), with a small metaconid close mesiolingually to the hypoconid, might be a complex P, rather than a P,. M36797 and M37558 are convincing Ps. They are better preserved but still incomplete. mesially. Both have equal-sized metaconids which lie equidistantly distolingual to the proto- conid and mesiolingual to the hypoconid, but, unlike Eh424, displaced from the lingual tooth edge and with distolingual oblique crests. The hypoconid on M37558 is smaller than on M36797 and the same size as the metaconid. The cristid obliqua lives up to its name on M36797 (as on Eh424) but is near longitudinal on M37558. P,: Both are trigonid fragments which show a buccal paraconid opposite a large lingual pseudoparaconid and a buccally concave paracristid. The specializations here are thus remi- niscent of those of an artiodactyl DP,. That these are not DPys is corroborated by enamel thickness and the presence at Creechbarrow of two DP,s referable to this species. One of the latter is complete enough to show a better divided mesial lobe with the pseudoparaconid larger than the paraconid. Stehlin (1908: 826-7; 1910b: pl. 20, figs 4, 12-13, 47) had some problems distinguishing between the two tooth types. Identification of Ef209 as a DP, was based on narrower outline and pseudoparaconid higher than paraconid. Whether the morphology of the Creechbarrow P,s occurs in M. gresslyi from Egerkingen depends on whether Stehlin’s (1910b: pl. 20, fig. 47) enigmatic tooth (NMB Ef238) is a P, or DP,. However, Ef238 has a pseudo- paraconid larger than the paraconid, not the other way round, and it could be argued that the Creechbarrow specimens are constant for this character. There is a slight difference in size of the parastylid between the two Creechbarrow P,s, but otherwise they are remarkably similar. Lower molars: These are all fragmentary, mainly having been broken into trigonid and talonid parts. The parastylid varies somewhat in strength (PI. 30, figs 7-8); in two specimens it J. J. HOOKER 386 “6LE d 296 8 x “(9PSLEW) 2), W WOT 7 ds aunqoyoipiadd Hy, q ‘eg “SI 68¢ d 208 6 x ‘(ET89EW) ded YOT Jopur ds wniwayioixiw, q “Bg “BI "Ege d 208 (S8I9EW) -W Yl ‘bp 3t4 ‘(PSTOEW) 2, W Ul 4 ‘VE “BIA (9OSLEW) pd Wl 9 “PZ “BIA (SOSLEW) (P2SdOA2) ed WUBI ‘T “BI “¢-¢ x “JOKOUIINY 1 ]ssaub ye wniayIoIxIW p—| S81y ‘(poxtjnsun JO q) [esnfooo pur (e) [ed9Ng s1¥ SMITA “MOIIBQYIIIID woy ‘z ‘ds aunqoysip.iada yi, pur joput ‘ds wnisayjoix1py{, “JaXawny 1M ]ssaub “ye wnisayj0ix1py JO 41901 yaoyd Joddn jo sydeviso1oew jYSIT 67 Pld 387 BARTONIAN MAMMALS OF HAMPSHIRE BASIN “ege ‘daag (QOSOEW) 1UsUITEY pruojer “Py Ye] ‘2-LOT “BLA (ZZSLEW) owsey pruojey “WW Ye] “6 ‘Sly (OTSLEW) Josey pruosiy © TW yal *g By (SZSLEW) JuoWsey prucsin ©/2/" yy yoy‘, “SLA (6ISLEW) *d Yl ‘2-k9 “BLY (SSSLEW) Apetsow uayosg *g 9] ‘¢ “BIL (SEP9E) TUOWBeY pruoye) {dq Yel ‘bp BLq (OSSLEW) (passaaal) UWIBeIy prluosin *qq 14311 ‘¢ “SI (P61SEW) 'dC Wl ‘7 BIA (SO89EW) “dC Yel 1 BA (9) fensury pure (pexyjnsun 10 (e) eo0ng a1” SMAIA “S-¢ X ‘MOIIEQYIIDID Woy JaAoWINY M]ssaub “Ye wini1ay10IX1JAI JO 43991 Y29YO 1aMO] Jo sydeiso010ew JYSIT O¢ RId q) [esnjo50 388 J. J. HOOKER Table 33 Distribution of intraspecifically variable characters in upper molars of Mixtotherium aff. gresslyi from Creechbarrow. extent of distance of premetaconule paraconule crista as from protocone fraction of as fraction of distance from _ distance from extent of extent of extent of mesostyle metaconule to protocone to lingual post- buccal No. Tooth loop mesostyle paracone cingulum cingulum cingulum M36184 M!'? — open >t 2 postproto- —_ just to absent cingulum metaconule parastyle— mesostyle M36190 M!'/? — open = = = = - M37509 M!'? — widely 4 4? postproto- 4 round - open cingulum metaconule M37510 M!'? narrowly 4 2 postproto- —_ just to continuous open cingulum metaconule M37512 M!'? - - 4? round all round - protocone metaconule M37514 M1? — closed c.4 - ~ - between parastyle and adjacent stylar cuspule M36185 M? widely 4 2 postproto- 5 round absent open cingulum metaconule parastyle— mesostyle M36186 M? widely <} 4 round + round absent open protocone metaconule parastyle— mesostyle M37508 M? closed 4 5 - - continuous M37511 M? open? >F - - just to - metaconule M37513. M?? open <4 - - - broken in front of mesostyle « M37515 M? widely 5 4 postproto- - broken at least open cingulum behind mesostyle M37517 -M?/2/3 - +? - postproto- 4 round ~ cingulum metaconule M37518 M?/2/3 - - - postproto- —_ just to - cingulum metaconule is nearly as weak as in the M. gresslyi specimens from Egerkingen y. A typical character of Mixtotherium M,_,s is the presence of a distal cingulum which loops round the distolingual corner of the tooth from the hypoconid to run up the distal wall of the entoconid. On some Creechbarrow specimens it is joined to the postcristid by a median to near lingual crestiform hypoconulid; on all the specimens the lingual limb of the postcristid is present although it may not reach the tip of the entoconid (sometimes missing on M. cuspidatum); on one specimen the distal cingulum does not extend buccally from the hypoconulid. DP.?: Only mesial fragments are preserved. They are similar to Stehlin’s figured DP, from Mormont (see p. 385) but are narrower and the metaconid appears to have been absent. M36808 has a parastylid which is slightly smaller than the paraconid and is situated mesio- BARTONIAN MAMMALS OF HAMPSHIRE BASIN 389 lingually to it. An extra small stylar cusp occurs mesial to the parastylid and makes the mesial outline taper to a point. The buccal cingulum is papillate and the lingual cingulum is weak and interrupted. M36416 is broken mesially as well as distally but seems to be beginning to broaden just before the mesial truncation. A very small cusp at the level of the non-existent lingual cingulum and distolingual to the large protoconid may represent the metaconid. A steep longitudinal protocristid bypasses it and is truncated by the distal breakage (cf. Stehlin 1910b: pl. 20, fig. 24). DP,;: M36194 is lower-crowned and has thinner enamel than the P 3s. It has a similar arrangement of cusps but the postcristid reaches the lingual margin at a small entoconid with a sharp distal crest; the cristid obliqua is only slightly oblique; and what is preserved of the confluent distal and buccal cingula is strong and papillate. Compared with the DP, that Stehlin figured from Mormont (1910b: 1116, text-fig. 231) the metaconid appears to extend further lingually and the cingula appear stronger, but the figure is not clear on the latter point. DP,,: M37550 is a trigonid fragment which differs from Stehlin’s (1910b: pl. 20, fig. 12) figure in its more transverse relationship between the protoconid and metaconid and is slightly narrower in outline. M36411 is more fragmentary but shows the protoconid and metaconid to have a more oblique relationship than Stehlin’s specimen. Variation appears to be high in this tooth. Discussion. It is difficult to judge the significance of the more obvious differences between the Creechbarrow assemblage and the various Egerkingen assemblages. The relative constancy of the position of the P; metaconid, the large P? metacone and the P, paraconid lobe, the last two perhaps being linked by occlusal relationships, suggest specializations in the Creechbarrow assemblage. Some evidence for variation in these characters in M. gresslyi and M. cuspidatum, however, demands a cautious approach. A statistical study of the genus could be enlightening. ?Mixtotherium sp. indet. (PI. 29, fig. 5) MATERIAL. Left DP?? (M368 13). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION AND DISCUSSION. Length is 3:6mm. The outline is wedge-shaped, the enamel thin, and the tooth is fairly worn and broken distolingually where a protocone was probably once present. The paracone is by far the largest cusp. The parastyle is large and buccally as well as mesially salient. The metacone is very small and only lingually salient, a feature increased by wear from the hypoconid of a ?7DP,. The distal third of the buccal edge bulges and is bent buccally. A weak mesiolingual cingulum is raised high on the flank of the paracone. This tooth differs from the DP? of Mixtotherium cuspidatum figured by Stehlin (1908: 803, text-fig. 120) in being much smaller and less molariform. The latter has a mesostyle, a large metacone and a crest joining the metacone to the protocone. Evidence for possible attribution of the Creechbarrow specimen to Mixtotherium comes from a referred DP? of M. ‘priscum’ (= M. gresslyi, see p. 383) from Egerkingen y (NMB Eh414). This tooth is associated in a maxillary fragment with a DP* which is likewise far less molariform than that of M. cus- pidatum. The Egerkingen y DP? is very similar to M36813. It differs in being nearly twice the linear size; in having a straighter buccal edge with a buccal cingulum bordering the metacone; and in having a lower mesiolingual cingulum. The size of M36813 would be appropriate to that of M. infans but no DP?s have been attributed to this species. Specific identification based on this single tooth would be unreliable. Family CEBOCHOERIDAE Lydekker 1883 TYPE GENUS. Cebochoerus Gervais 1852. INCLUDED GENERA. Acotherulum Gervais 1850 (?including Moiachoerus Golpe-Posse 1972, fide Sudre 1978b: 82) and Gen. nov. (for Choeromorus Gervais 1852, sensu Jaeger 1971). 390 J. J. HOOKER RANGE. Middle Lutetian to early Stampian, Europe. D1aAGnosis. See Sudre (1978b: 49). INTRODUCTION. This small family has long been the subject of discussion over generic content and nomenclatural priorities (e.g. Stehlin 1908, Depéret 1917, Pearson 1927, Sudre 1978b). The main causes of these problems are the rather low anatomical diversity (taxonomic differences often being minor and easily confused with intraspecific variation) and the poor quality or loss of type specimens. A preoccupation with increase in size as the main criterion in the recognition of lineages has also impeded the unravelling of the morphological details of the group. The problem of the genus Choeromorus (sensu Stehlin 1908), whose only difference from Cebochoerus was supposed to be the presence of enlarged pig-like lower canines and extrapo- lated small first premolars, was essentially overcome by Sudre (1978b: 52-54); he considered the affinity of the canines, which were isolated, with the Cebochoeridae to be unproven and that the problem therefore never really existed. The probability that these teeth belong to palaeotheres is discussed herein above under Lophiotherium (p. 352). The problem caused by synonymy of the type species of Cebochoerus and Acotherulum was discussed by Sudre (1978b: 49-51). He proposed the changing of the type species of Cebo- choerus to C. lacustris (following the apparent sentiments of Gervais (1859: 198)) in order to stabilize the Cebochoeridae type genus and to maintain its ‘common usage’ distinction from Acotherulum. No formal proposal has yet been made to the ICZN regarding this change of type species. Sudre’s separation of the two genera is followed here, but a different content of species for each is shown to be necessary. It is worthy of note that the holotype of C. anceps is no longer missing (as recorded by Depéret 1917: 102 and Sudre 1978b: 50) but is in the collections of the University of Lyon (FSL 6796). Genus CEBOCHOERUS Gervais 1852 (sensu Gervais 1859) [incl. Cebochoerus (Gervachoerus) Sudre 1978b in part] TYPE SPECIES. C. anceps Gervais 1852. Ludian; La Débruge, Vaucluse, France. This is pending formal application to the ICZN for change of type species to C. lacustris Gervais 1856 (from the Ludian limestone of Souvignargues, Gard, France), the holotype of which was reported lost by Sudre (1978b: 74). INCLUDED SPECIES. C. lacustris Gervais 1856; C. minor Gervais 1876 (including C. helveticus (Pictet & Humbert 1869) Sudre 1978b); C. ruetimeyeri Stehlin 1908; C. robiacensis Depéret 1917; C. fontensis Sudre 1978b; and C. aff. fontensis. RANGE. Late Lutetian—late Ludian; England, France, Switzerland and Spain. EMENDED DIAGNOsIS. P+ large and caniniform; C, incisiform. Upper cheek teeth with strong mesial and distal cingula and crests. Upper molars with: large paraconule; lingual crown height > buccal crown height; fused lingual roots; postprotocrista restricted to distal wall of proto- cone; protolophule which may or may not join paracone. P* protocone lingual to paracone. P? protocone lingual or distolingual to paracone; P? outline may or may not narrow mesially. P* protocone present or absent. Lower molars with angles of protoconid and hypoconid crests acute. Supraorbital foramina above level of M* or M?. Adult horizontal mandibular ramus about twice as deep below M, as below P,. SIGNIFICANCE OF THE SUBGENUS Gervachoerus. Sudre (1978b: 56-57) erected his new subgenus with C. minor as its type species. The characters which he used, however, were based on a referred cranium and mandible from Euzet (figured by Depéret 1917: pl. 17, figs 4-6). Contrary to his diagnosis of Gervachoerus, the holotype maxilla of C. minor (MNHN Qu60b) has upper molars with central accessory conules, and a topotype maxilla (MNHN Qu71) has P® roots indicating a longer, narrower tooth than P* and which is broader distally than mesially. Material from Lamandine (the type locality) is not complete enough to comment on the other characters listed for Gervachoerus, but the only species known to have the mandible shallow BARTONIAN MAMMALS OF HAMPSHIRE BASIN 391 Text-figure 56 Disto-occlusal views of upper preultimate molars of Cebochoerus and Acotherulum, about x 4. A, left M* of C. fontensis Sudre (UM F4-214) from Fons 4; B, left M? of C. aff. fontensis (UL 6145) from Euzet; C, right M? (reversed) of C. robiacensis Depéret (M7528) from Caylux; D, right M! (reversed) of A. quercyi (Stehlin) (MNHN Qu7075) from Bach; E, left M! of C. ‘helveticus’ (Pictet & Humbert) (LGM LMS808) from Mormont; F, left M! of C. minor Gervais (MNHN Qu11283) from Lamandine. A, B, D, E and F are drawn from casts. from front to back is ‘G. campichii, here shown to belong to Acotherulum (see p. 399). The presence on the upper molars from Lamandine of a protolophule joining the paracone and almost equal buccal and lingual crown height indicates close relationships with C. lacustris and C. helveticus (which Sudre referred to the nominate subgenus) rather than to the Euzet skull or to other species Sudre placed in Gervachoerus, such as G. fontensis (see Text-fig. 56). It is concluded that: 1, Gervachoerus, based as it must be on C. minor, is not sufficiently distinct from the nominate subgenus, based on C. lacustris, to warrant subgeneric separation; 2, Sudre’s concept of Gervachoerus is almost entirely a phylogenetic one which includes species attributable on characters to both Cebochoerus and Acotherulum; and 3, the Euzet material cannot belong to C. minor and requires a new specific name, if it is not conspecific with C. fontensis (it is referred to here as C. aff. fontensis). Cebochoerus minor Gervais 1876 (Pl. 31, fig. 1; Text-figs 57, 58B, 59, 60A) vy. 1972 Cebochoerus sp.; Hooker: 180-181. v? 1977b Creodonta or Carnivora indet.; Hooker: 141. v. 1978b Cebochoerus helveticus (Pictet & Humbert)?; Sudre: 72, 74. vy. 1980 Cebochoerus sp. 1; Hooker & Insole: 44. Hototyre. Left maxilla with M! * (MNHN Qu60b). Early Ludian limestone; Lamandine, Quercy, France (see PI. 31, fig. 2). This specimen was to have been figured by Gervais (1876: pl. XI, figs 7-8), but his plate XI appears never to have been published. RANGE OF SPECIES. Marinesian, England; early Ludian, France and Spain; probably Mari- nesian, Switzerland and France. 392 J. J. HOOKER MATERIAL. Slightly crushed cranium in three pieces, lacking braincase, and with crowns of left P3_M? and right P*-M!; roots of left and right C’-P' and left P?; and alveoli of right 1272 (M26649). HorIZON AND LOCALITY. Bed C (0:6 m (= 2 ft) below top nodule band), Barton Clay Formation, Barton Cliff, east of Chewton Bunny, Hampshire. In wavewashed lower part of cliff at about SZ 226930. Collected by Mr & Mrs P. Clasby. DOUBTFULLY REFERRED MATERIAL. Five mainly poorly-preserved anterior permanent and/or deciduous lower premolars (M36199, M36403, M36801, M37551, M37565) from the Creech- barrow Limestone Formation, Creechbarrow. EMENDED DIAGNOSIS. Medium-sized Cebochoerus (mean length of M' = 7-5mm; see Text-fig. 57). Upper molars with: protolophule joining paracone, and lingual crown height = buccal crown height. P* width/length ratio 1-4. P*® with protocone distolingual to paracone; and outline narrowing mesially. [Crown of P* unknown]. Supraorbital foramina above level of M?. Frontals not extended posteriorly from postorbital processes. DESCRIPTION OF CRANIUM. The specimen is very fragile and after being cleaned of matrix was filled with carbowax to give it support. Slight crushing has bent the ventral part of the left maxilla, posterior to the infraorbital foramen, plus the jugal, the ventral half of the lacrymal and the palate in a medial direction, so that in ventral view the teeth anterior to left P*? appear to splay laterally. The left nasal and frontal have detached from the maxilla at the intervening sutures and slid beneath it slightly. The result in dorsal view is that the midline suture has come to lie nearer to the left than the right maxilla. Incisors. The right premaxilla has a complete circular alveolus for I°. It is about 20° from the vertical and not separated from the canine by a diastema. Parts of the I* alveolus are present, —— Vida LEG J Text-figure 57 Scatter diagram of length (1) against width (w) in upper molars of the Cebochoerus minor/helveticus/lacustris complex. © = C. lacustris Gervais from Souvignargues (H = holotype) and Mémerlein; [] = C. ‘helveticus’ (Pictet & Humbert) from Mormont; A = C. minor Gervais from Lamandine: V = C. minor from Sosis; [> = C. minor from Le Bretou; © = C. minor from Fons 6; 3= C. minor from Barton. Symbols solid on left = M!, solid on right = M?; completely solid = M?: outline = M'/2. Measurements in millimetres. Lines join teeth of one individual. a =He - - z w BARTONIAN MAMMALS OF HAMPSHIRE BASIN 393 anteromedial to I° and orientated at about 45° to the vertical. There is no evidence for an I’ but there would be room for it between I? and the midline. In the Mémerlein cranium of C. lacustris, the premaxilla appears blunter anteriorly than that of M26649 with the incisor alveoli more transversely orientated. There may, however, be some distortion in the Mémerlein speci- men, hidden beneath a crust of matrix. No other known cebochoerid crania have this region preserved. Canines. These are two-rooted and only slightly smaller than the right canine of C. lacustris (proportional with the cheek teeth). The upper canine of the Euzet C. aff. fontensis cranium appears both proportionally and absolutely smaller. Premolars. Like the canines, the P's are two-rooted and absolutely but not proportionally slightly smaller than those of C. lacustris, but are both absolutely and proportionally larger than those of C. aff. fontensis. They are separated from both canines and P’s by short dia- stemata as in these other two species. The two roots of P? have been damaged by the axis of diagenetic bending of the cranium. P? is long, narrow and wedge-shaped with a small distolingual protocone, which makes the tooth wider distally then mesially. The roots on a topotype of C. minor (MNHN Qu71; PI. 31, fig. 3) suggest a similar crown morphology. Both are certainly very different from C. aff. fontensis whose P? has a large lingual protocone. C. lacustris from Mémerlein and C. ‘helveticus’ (Stehlin 1908: pl. 14, fig. 17) have a P? protocone larger than that of the Barton specimen, but like the latter and unlike C. aff. fontensis it is distolingual in position. In the Barton specimen there is a small metacone midway along the postparacrista, as in C. lacustris and C. ‘helveticus’ and unlike C. aff. fontensis. The P*s are transverse and triangular, with pre- and post-cingula which do not meet round the protocone. There is a distal accessory conule, more distinct on the right than on the left tooth. C. ‘helveticus’ (NMB Mt2) has a very similar P*. On the Mémerlein C. lacustris, the pre- and post-cingula are stronger, the former bypassing the postprotocrista and fusing with the ‘paracingulum’. C. aff. fontensis has a less transverse P* with short weak precingulum and no postcingulum and the protocone with its crests is higher. Molars. These resemble C. minor, C. lacustris and C. ‘helveticus’ in the presence of a protolo- phule joining the paracone and in the premetaconule crista tending to form a prominent central accessory conule and to join the metacone. Buccal cingula are restricted to between the para- cone and metacone and near the parastyle and metastyle. There is a very small mesostyle on M? and a larger one on M? associated with buccal flexing of the centrocrista. These cingular features show much variation in C. minor from Lamandine, three specimens including the holotype having a strong cingulum round the metacone, one having a complete buccal cin- gulum and two others being like the Barton specimen. The fact that the Memerlein C. lacustris upper molars have the buccal cingulum continuous round the metacone is thus not considered a taxonomic difference either from the Barton specimen or from specimens of C. minor from the type locality. Similar variation occurs in C. ‘helveticus’ but no specimens appear to exist with a complete buccal cingulum. Osteology. The cranium is smaller than, but otherwise very similar to, that of C. lacustris from Meémerlein, which has been extensively described and figured by Pearson (1927). Only those features which are new or differ from this species are described here. Recognition of sutures and orbital foramina is made difficult by surface erosion and crushing. Text-figs 58-59 attempt to show these and other features and reconstruct the cranial shape. I have relied heavily on the well-preserved cranium of A. saturninum (MNHN Qu16366, holotype of Leptacotherulum cadurcense) to guide interpretation of sutures and foramina (Text-fig. 58A; see also Russell 1964: 263, text-fig. 57). The left frontal preserves twinned supraorbital foramina, probably an individual character (cf. Sisson & Grossman 1953: 137, for Bos taurus L.). Their position above M? is like C. lacustris (see Pearson 1927). In the crania of all other species of Cebochoerus and Acotherulum, where this feature is known, they are above M? (viz. C. aff. fontensis, C. robiacensis, A. saturnin- um and A. quercyi). Immediately behind the postorbital process the anterior (frontal) edge of the temporal fossa 394 J. J. HOOKER BARTONIAN MAMMALS OF HAMPSHIRE BASIN 395 Text-figure 58 Left orbital regions of A, Acotherulum saturninum Gervais (holotype of Leptacotheru- lum cadurcense Filhol) from the Phosphorites du Quercy (MNHN Qu16366) (right side reversed); and B, Cebochoerus minor Gervais from Bed C, Barton Clay, Barton Cliff (M26649). Both x 1-3. Abbreviations: F = frontal; f = fossa (?for origin of the ventral oblique muscle of eyeball); J = jugal; L = lacrymal; If = lacrymal foramen; Ip = lacrymal process; M = maxilla; mc = maxillary canal; oosc = orbital opening of supraorbital canal; P=palatine; ppf= opostpalatine foramen; spf = sphenopalatine foramen. has a transverse orientation. This is like all cranially known Acotherulum and Cebochoerus species except C. lacustris, where the frontal extends backwards from the postorbital process almost to a position vertically above the anterior edge of the squamosal (see Pearson 1927: text-fig. 22). The anterior edge of the very shallow narial incision is preserved on the right premaxilla and on a tiny fragment of the right nasal. The configuration is very similar to that of C. lacustris, the only other cebochoerid where this part is known. The maxillary—premaxillary suture is not visible, apparently owing to fusion of these bones. Part of the edge of the right anterior palatine foramen is preserved, a feature covered by matrix/plaster on the C. lacustris cranium and unknown in any other specimen. SIGNIFICANCE OF THE P? CHARACTERS IN C. minor. The difference in protocone size between C. minor on the one hand and C. ‘helveticus’ and C. lacustris on the other is not taken here to be a specific character because of variation in the relevant occluding parts of P4s of C. minor: in particular two jaws with lightly worn premolars, one from Lamandine (MNHN Qu67) and another with no detailed provenance but of very similar preservation (MNHN Qu79). In Qu79 the P, paracristid has a doubly concave buccal phase facet which would have occluded with a P? postparacrista with small metacone. A lingual phase facet is present near the tip of the protoconid, sloping mesiolingually at a shallow angle; this would have made contact with a small distolingual P? protocone such as occurs on the Barton cranium. Qu67 (Filhol 1877b: pl. 14, figs 288-290) has a similar but more worn and flatter buccal phase facet on the P, Plate 31 Light macrographs of Cebochoerus minor Gervais, x 1:6. Fig. la—d, partial cranium (M26649) from Barton Cliff; in lateral (a, left cheek tooth row only), ventral (b), dorsal (c) and left dorsolateral (d) views. Fig. 2, holotype left M’ ? (MNHN Qu60b) from Lamandine, in occlusal view. Fig. 3, right M'? with roots of P** (reversed) (MNHN Qu71) from Lamandine, in occlusal view. See p. 391. 396 J. J. HOOKER ginv gdnv_ gtv BARTONIAN MAMMALS OF HAMPSHIRE BASIN 397 paracristid and another buccal phase facet near the tip of the large metaconid caused by contact with the postprotocrista of a P?. Also on the paracristid but lingual to the buccal phase facet is a long lingual phase facet sloping steeply in a mesial direction. This would have been caused by a larger, less distal P* protocone. Whether this protocone would resemble that of C. lacustris or C. aff. fontensis remains unclear, but P? and P;_, appear to be rather variable teeth in C. minor. Similar potential variation is demonstrated by two Mémerlein mandibles, one belonging to the cranium and an immature one consisting mainly of symphysis (M32172) with an unerupted P;. In the latter, P; has a prominent metaconid distolingual to the protoconid, whereas in the former the metaconid is smaller and distal to the protoconid. DISTINGUISHING FEATURES OF C. minor, C. ‘helveticus’ AND C. lacustris. The only observable difference between C. minor and C. lacustris, apart from size, is in the anterior edge of the temporal fossa. C. *helveticus’ is intermediate in size between the two, no cranial characters are known and it resembles C. Jacustris in its partly unreliable P°. Sudre’s (1978b: 75—76) diagnosis and discussion of C. lacustris does not help as it is mainly concerned with distinguishing it from ‘G.’ minor, not C. helveticus. Recorded contemporaneity of C. minor and C. lacustris applies to the type locality of the former (Lamandine) and relies on Stehlin’s (1908: 742) determination of specimens in Montauban and Munich museums. The quoted measurements indeed compare well with those of the Mémerlein skull of C. lacustris. The measurements which Sudre (1978b: 75) quoted for the now lost holotype maxilla of C. lacustris (8 mm for each of the three molars) are taken from Gervais’ (1876: 48) type description of C. minor, not C. lacustris. Therefore his maxilla from Fons 6 and the tooth from Sosis can reasonably be reidentified as C. minor. Gervais (1859: 197-198, text-fig. 20), in the first meaning- ful description of the holotype of C. lacustris, gave M‘~? length as 27mm, width of M! as 21 mm (which must be an error) and P*—M? length as 41 mm. Text-fig. 57, p. 392, shows length and width of upper molars for these three species. It shows the Barton cranium to plot easily within the range of variation of the type assemblage of C. minor. The known specimens of C. ‘helveticus’ (type assemblage only) overlap with C. minor and approach C. lacustris. There are several possible conclusions on the status of C. *helveticus’ from these data: 1, it consists of a mixed assemblage of C. minor and C. lacustris; 2, it is a species distinct from either, differing in characters as yet unknown; or 3, it is conspecific with C. minor, representing a degreee of intraspecific variation resulting from geographical and/or strati- graphical differences. Text-fig. 60A is a cumulative graph of coefficients of variation for the measurements given in Text-fig. 57. A variable but low v for seven C. minor specimens from Lamandine is increased to 6-4 in length and 7-7 in width when C. ‘helveticus’ measurements are added. These coefficients are still insufficient to demonstrate a specific separation of C. ‘helve- ticus from C. minor. They support the choice of conclusion 3 above, although conclusion 2 can never be excluded. The addition of measurements of specimens from Sosis, Le Bretou, Fons 6 and Barton reduces the coefficient, but the addition of two specimens of C. lacustris raises it to a level which suggests that on size alone more than one species is now represented. It is possible that speciation resulted from isolation during the Bartonian. Perhaps the Swiss and French assemblages gave rise to C. lacustris by slight size increase; English C. minor remained unchanged but spread to France and Spain after the end of the Bartonian. Text-figure 59 Partial reconstruction of skull of Cebochoerus minor Gervais. Cranium based on M26649 from Bed C, Barton Clay, Barton Cliff; mandible on MNHN Qu67 from Lamandine. x 1:3. Abbreviations: apf = anterior palatine foramen; F = frontal; fg = frontal groove; gdnv = groove for dorsal nasal vein; gfvy = groove for facial vein; glnv = groove for lateral nasal vein; iof = infraorbital foramen; J = jugal; L = lacrymal; lf = lacrymal foramen; lp = lacrymal process; M = maxilla; mc = maxillary canal; N = nasal; oosc = orbital opening of supraorbital canal; Pa = palatine; Pm = premaxilla; ppf= postpalatine foramen; sof =supraorbital foramina; spf = sphenopalatine foramen. A, dorsal view of cranium, B, ventral view of anterior end of palate and C, left lateral view of cranium and mandible. 398 J. J. HOOKER Cebochoerus robiacensis Depéret 1917 (Pl. 32, figs 2, 4, 5; Text-fig. 60B) v. 1977b Cebochoerus cf. ruetimeyeri Stehlin; Hooker: 141. v. 1980 Cebochoerus cf. ruetimeyeri Stehlin; Hooker & Insole: 44. Lectotype. Right maxillary fragment with M?-* (NMB Rb52). Marinesian, Robiac (figured Stehlin 1908: text-fig. 109); selected Sudre, 1969a: 125. RANGE. Marinesian, England and France. MATERIAL. Right M* (M36803); right DP* (M37498); left M,/. (M37501); 2 right M, 5 trigonid fragments (M36860, M37504); and possibly a distal fragment of left DP, (M37503). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. D1aGnosis. Small Cebochoerus (mean length of M' = 5:3mm). Upper molars with: protoloph- ule not joining paracone; and lingual crown height > buccal crown height. P* width/length ratio 1-1. P* with protocone lingual to paracone; and outline not narrowing mesially. P? with protocone. Supraorbital foramina above level of M?. Frontals not extended posteriorly from postorbital processes. 7 (ie) ° 105 ‘ + e 7 od 7 * aia : a * * | ~ natn eee ee Se N A iE SS T 4 6 B ! Text-figure 60 A, cumulative curves of coefficients of variation for lengths and widths of M' of the Cebochoerus minor/helveticus/lacustris complex. 1—7 from Lamandine; 8—10 from Mormont; 11 from Sosis; 12 from Le Bretou; 13 from Fons 6; 14 from Barton; 15-16 from Souvignargues and Mémerlein. B, scatter diagram of length (1) against width (w) in P* and M? of Cebochoerus robiacensis Deperet and Acotherulum campichii (Pictet). A = C. robiacensis from Lautrec (measured from cast); © = C. robiacensis from Caylux; © = C. robiacensis from Robiac; [] = A. campichii from Mormont (measured from Stehlin 1908: pl. 14, fig. 6); 42 = C. robiacensis and A. campichii from Creechbarrow. Symbols solid above = P*; completely solid = M?. Measurements in millimetres. Lines join teeth of one individual. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 399 DESCRIPTION AND DISCUSSION. Sudre (19786: 66) synonymized this species with Acotherulum campichii (Pictet) comb. nov. (p. 400), which he referred to the genus Cebochoerus. In discussing his identification of specimens as ‘C. campichii, he noted that the maxilla from Mormont figured by Stehlin (1908: pl. 14, fig. 6) differed from the Robiac material in having broader P* and shorter M®. He considered that these characters had no evolutionary value and constituted simply individual variation. Three more characters — small upper molar paraconule, mesio- lingual P* protocone and weak mesial and distal cheek tooth cingula and crests — also dis- tinguish the Mormont maxilla from the Robiac material. These last characters are constant for three P*s and five upper molars from Creechbarrow. These specimens are thus referred to A. campichii (see p. 400) and demonstrate the taxonomic importance of these characters in the distinction of C. robiacensis from A. campichii. The Creechbarrow M? (M36803, PI. 32, fig. 4) can be distinguished further from those of A. campichii by being slightly larger and by the mesial margin being convex, because of the bulging paraconule. Its paracone is broken buccally but a buccal cingulum appears to have continued round this cusp from the metaconal region. The cingula of the lectotype, according to Stehlin’s figure, have a similar extent; but further specimens from Robiac, Lautrec (Sudre 1978b: text-pl. 6, figs 1, 3) and Caylux (PI. 32, fig. 8) have weaker buccal cingula, showing this to be a variable feature. The DP* (PI. 32, fig. 2) has thinner enamel than the M?; an almost complete strong buccal cingulum, only slightly interrupted at the paracone; shallow concave valleys; and a strong postprotocrista which almost joins the premetaconule crista. The lower molars can be distinguished from those of A. campichii by having the mesial and distal crests of protoconid and hypoconid meeting at acute angles. M37501 has a mesoconid but this is a variable feature, common in Cebochoerus, less common in Acotherulum. It also has the hypoconulid at a more longitudinal orientation to the postcristid; this is another feature which is variable but more common in Cebochoerus than Acotherulum (see Pl. 32, figs 5b, 6b). The character of the hypoconulid is also visible on the DP, fragment. C. robiacensis has not previously been given a character diagnosis. Athors have tended only to seek recognition of it either as the ancestor of C. minor (Depéret 1917: 113) or the descen- dant of *C.’ suillus (Stehlin 1908: 745). Although relatively easily distinguishable from A. campi- chii, as shown above, it is less easily separated from the Auversian C. ruetimeyeri. C. ruetimeyeri is usually slightly larger and tends to have stronger buccal and lingual cingular developments, but there is much overlap. The most reliable difference appears to be lower lingual crown height of upper molars in C. ruetimeyeri, this being approximately equal to the buccal crown height. Genus ACOTHERULUM Gervais 1850 [incl. Metadichobune and Leptacotherulum Filhol 1877a, Cebochoerus (Gervachozrus) Sudre 1978b in part, and ?Moiachoerus Golpe-Posse 1972] TYPE SPECIES. A. saturninum Gervais 1850. Ludian; La Deébruge, Vaucluse, France (?including Leptacotherulum cadurcense Filhol 1877a). INCLUDED SPECIES. A. quercyi (Stehlin 1908) Sudre 19786; A. campichii (Pictet 1857) comb. nov.; A. pumilum (Stehlin 1908) Sudre 1978b; A. sp. (Bembridge Limestone). RANGE. Marinesian to late Ludian, England; Marinesian to Ludian, France and Switzerland; Ludian, Spain; and early Stampian, Belgium. EMENDED DIAGNOSIS. P+ large and caniniform; C, incisiform. Upper cheek teeth with weak mesial and distal cingula and crests. Upper molars with: small paraconule; lingual crown height > buccal crown height; fused or unfused lingual roots; postprotocrista restricted to distal wall of protocone; protolophule not joining paracone. P* protocone mesiolingual to paracone. P? protocone lingual to paracone; P? outline narrowing mesially. P? protocone absent. Lower molars with angles of protoconid and hypoconid crests right-angled to obtuse. Supraorbital foramina above level of M?. Adult horizontal mandibular ramus scarcely deeper below M, than below P,. 400 J. J. HOOKER Acotherulum campichii (Pictet 1857) comb. nov. (Pl. 32, figs 1, 3, 6, 7; Text-fig. 60B) v. 1977b Cebochoerus cf. campichii (Pictet) Stehlin; Hooker: 141. v. 1980 Cebochoerus cf. campichii (Pictet) Stehlin; Hooker & Insole: 44. Ho.otyre. Right mandibular ramus with P,—M,. Sidérolithique; Mormont, Canton Vaud, Switzerland (figured Pictet 1857: pl. 4, figs 5-9), now lost according to Stehlin (1908: 729). De la Harpe (1869: 466) recorded the species (as Dichobune Campichii) from only one of the Mormont localities - Gare d’Eclépens, so it can be concluded that the missing holotype came from there. RANGE. Marinesian; England, France and Switzerland. MATERIAL. Three right P*s (M36191, M36807, M37497); two left M?s (M36802, M37499): two right M*s (M36804, M37500); mesial fragment of right upper molar (M36805); left M,; (M37502); and left M, lacking trigonid (M36183). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. EMENDED DIAGNOSIS. Medium-small Acotherulum (length of M'=5-6mm). Upper molar lingual roots fused. M,/, more than half as wide as long. DESCRIPTION. Size appears to be slightly less than in the Mormont material (see Text-fig. 60B and Stehlin 1908: 731; pl. 14, fig. 6). P*: M37497 is very worn and shows few detailed features. M36191 shows a small metacone immediately behind the paracone. M36807 has no metacone. Both M36191 and M37497 have suggestions of a distal accessory conule and M37497 also a paraconule. In M37497, distinction from C. robiacensis can still be made by the position of the protocone and the low mesial and distal crests and cingula which have hardly been affected by the heavy wear which has removed most of the paracone and protocone (cf. Sudre 1978b: text-pl. 6, figs 1b, 3). M?: Variation is mainly in outline shape and distribution of cingula. The terminal bulge of the outline is midway along the metacingulum in M37499 and M36804 and is in the metastylar region in M36802 and M37500. The cingulum is continuous round the metacone in M37499, M36804 and M36802. In M37500, the precingulum does not extend as far lingually as in the other three. M36804 is the only specimen with even the slightest development of a postcingu- lum. M36802 and M36804 each have a very small accessory conule. This is absent in M37500 and the area is too worn to tell in M37499. All have small paraconules, which distinguish them from C. robiacensis from the same locality (see p. 398). M,).: M37502 is significantly narrower relatively than the same tooth attributed herein to C. robiacensis; it has no mesoconid; the paracristid is bent asymmetrically; the hypoconulid merges gradually with the postcristid; and the tooth as a whole appears less robust. M;: What remains of M36183 is fairly worn. It appears to have been relatively narrow; there is a lingual accessory conule between the entoconid and hypoconulid; and the cingula are weak. Size and the accessory conule are more consistent with attribution to A. campichii than to C. robiacensis. PHYLOGENETIC IMPLICATIONS WITHIN THE CEBOCHOERIDAE. Table 34 compares characters of the various species of Cebochoerus and Acotherulum. Primitive versus advanced states have been Plate 32 Light macrographs of cheek teeth of Cebochoerus and Acotherulum, x 5. Views are buccal (a), occlusal (b or unsuffixed) and lingual (c). Figs 1, 3, 6, 7 Acotherulum campichii (Pictet) from Creechbarrow. Fig. 1a, b, right P* (reversed) (M36807). Fig. 3a, b, right M® (reversed) (M37500). Fig. 6a—c, left M,). (M37502). Fig. 7a—c, left M3 talonid fragment (M36183). See above. Figs 2, 4,5 Cebochoerus robiacensis Depéret from Creechbarrow. Fig. 2a, b, right DP* (reversed) (M37498). Fig. 4a, b, right M? (reversed) (M36803). Fig. Sa—c, left M, 2 (M37501). See p. 398. Fig.8 Cebochoerus robiacensis Depéret from Caylux. Right maxilla with M‘~> (reversed) (M7528). BARTONIAN MAMMALS OF HAMPSHIRE BASIN 401 402 J. J. HOOKER 6(142)am = oO o C ~ = 5 2 3 = E = Me 5 3 > 7) 2 3 = > ° ¢ Q 5 Q = 5 9 5 ® & 5 @ E iS wy o a e I = 3 % 3 % 3 ° 5 s ~ 9 ° Q 6 > 2 2 = £ w < x x < (3) Ss) So (S) (3) 16 15 314) a i] 1 i] \ i] ! 1 8 Text-figure 61 Cladogram of species of Acotherulum and Cebochoerus. This is the most parsimonious possibility. Character 7 is apparently the only parallelism. Because of the incompleteness of certain species, some of their character states are unknown (see Table 34 opposite). Two of these characters (15 and 16) are each only known in a single species, unknown in closely related species and define terminal dichotomies. Their occurrence in the poorly-known, less related forms is unlikely, but if shown to occur they would be incongruent with other characters and would thus be parallelisms. Most of the other incompletely known characters are more fundamental and are shown hatched. Character 2 is poorly known in Acotherulum but is tentatively considered consistently primitive, as in its advanced state it is distinctive of only two species of Cebochoerus. The problems posed by characters 1, 3 and 4 could be overcome by excluding the very poorly known A. pumilum from the cladogram. C. ruetimeyeri is variable for two characters which are constant but different in the two terminal groups of Cebochoerus. Using only the advanced states of these characters, C. ruetimeyeri can be related equally to the C. robiacensis/fontensis group or to the C. minor/lacustris group, unless in the future it is found to have the advanced state for character 14. This would relate it more to the latter group, but it remains otherwise totally primitive within the genus Cebochoerus. Lack of knowledge of certain characters renders the cladogram very tentative, but it may be hoped the gaps will be filled by the anterior dentition and skull parts eventually becoming known for the incomplete species. By applying the cladogram to the occurrence of the species in the stratigraphical record, the phylogeny shown in Text-fig. 62 is produced. (« = greater size.) calculated by outgroup comparison with the rest of the Dichobunoidea, except character 14, which was by ingroup comparison as it is relatively poorly known outside the Cebochoeridae. C. suillus (Gervais 1852) Stehlin 1908, C. dawsoni Sudre 1978b and C. jaegeri Sudre 1978b are omitted since distinctive characters such as small premolariform P, and upper molar post- protocrista joining metaconule indicate that a new genus is required to include them. One way of displaying these character distributions is in the form of a cladogram (Text-fig. 61). Cebochoerus ruetimeyeri shares incipient development of advanced character 10 with the C. robiacensis/fontensis group and of advanced character 6 with the C. minor/lacustris group. It is otherwise primitive with respect to both, which makes C. ruetimeyeri from Egerkingen « a good candidate for their common ancestry. These groups correspond partially to Sudre’s subgenera Gervachoerus and Cebochoerus respectively, but are much more restricted. A maxilla from Egerkingen £ was referred to C. ruetimeyeri by Stehlin (1908: pl. 14, fig. 55). It is intermediate between the latter and C. minor in size and has an upper molar protolophule joining the paracone, as in C. minor. A phylogeny is proposed in Text-fig. 62. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 403 Table 34 Characters and character states of species in the genera Cebochoerus and Acotherulum. Charac- ters are estimated as primitive (—) or advanced (+) based on outgroup comparison with the rest of the Dichobunoidea. Lack of information on a character indicated by ‘?’. ied ss = 2 cS es iS Ss S oS S a 5 > 2 Se S Se es § 5 Character Primitive Advanced Se x oO © S) oO 6 1. P+ caniniform no yes + + ?+ + + + tPF 2. P? protocone absent present ? ? — ? + + = = = 3. P® protocone position lingual distolingual - 7) =F + regarding paracone 4. P* protocone position lingual mesiolingual + + + 2+ regarding paracone 5. Upper molar paraconule large small + + a5 3F = = = = = 6. Upper molar protolophule no yes = ~ — _ _ _ +/— + joining paracone 7. Upper molar lingual: = >1 + + + + ar 5 = = = buccal crown height 8. Upper molar postproto- no yes ar af =f te + + + ae + crista separate from metaconule 9. Upper molar lingual fused separate = + af ar roots 10. M!'/P* size ratio high low - =- = - a 11. P, paraconid high no yes = = = + 12. Angle of lower molar acute obtuse sb + + protoconid and hypoconid crests 13. Lower cheek teeth wide narrow - = = + 14. Supraorbital foramina M? M3 Rp = ? = = ? wR dk above M? or M? 15. Frontal projects behind no yes 2 — — 2 - = i = 4F postorbital process 16. Jugal shallow deep ? = oF ? - = ? - = 17. Horizontal mandibular shallow deep ? ae + + + + Tamus Stampian IE A. quercyi & cf. Ludian piri C. aff. fontensis A. saturninum . lacustris = t a C. fontensis C. minor A. campichii C. robiacensis Marinesian | | \ > ~ u YK Sx SS SZ Gi c. © Egerkingen 8) Auversian Ne NS ee “ : Ie C. ruetimeyeri Lutetian Se — ae Text-figure 62 Proposed phylogeny of Acotherulum and Cebochoerus in the Eocene of Europe; based on the cladogram of Text-fig. 61 and stratigraphical occurrence. 404 J. J. HOOKER Suborder ANCODONTA Matthew 19296 Superfamily ANTHRACOTHERIOIDEA Gill 1872 Note that the Choeropotamidae are included here, not in the Dichobunoidea. Their differences from the Cebochoeridae are fundamental and I follow Stehlin (1908) in considering the family close to Haplobunodon. Family HAPLOBUNODONTIDAE Pilgrim 1941 (sensu Sudre 1978b) TYPE GENUS. Haplobunodon Depéret 1908. INCLUDED GENERA. Rhagatherium Pictet 1857; Amphirhagatherium Depéret 1908; Lophiobuno- don Depéret 1908; Anthracobunodon Heller 1934; Masillabune Tobien 1980; ?Thaumasto- gnathus Filhol 1890a. RANGE. Early Lutetian to late Ludian, Europe. DIAGNOsIs. See Sudre (1978b: 90). Genus HAPLOBUNODON Deperet 1908 [The main early generic and specific synonymies are complex and are dealt with by Stehlin (1908: 752-765) ] Type SpEcIES. H. lydekkeri Stehlin 1908. Lower Headon Beds; Hordle Cliff, Hampshire. Nomen- clatural background is given by Stehlin (1908: 752-754). INCLUDED SPECIES. H. solodurense Stehlin 1908; H. muelleri (Rutimeyer 1862) Stehlin 1908; ?H. ruetimeyeri (Pavlow 1900) Stehlin 1908; and H. venatorum sp. nov. RANGE. Late Lutetian to Ludian, Switzerland; Marinesian to early Ludian, France and England. EMENDED DIAGNOSIS. P+ present to absent. No diastema between P, and P,. P* * with outline as an isosceles triangle and without metacone. Upper molars semibunodont with mesostyles less than half the height of the paracone; paraconule smaller than protocone and not separated from it by deep fissure; premetaconule crista short, often joined to protocone. M? distal edge straight. P, _, without hypoconid. P, often with large paraconid. Lower preultimate molars with postcristid joining hypoconulid usually independently of entoconid. [The otic region has been described by Pearson (1927: 438-440) for H. lydekkeri, but for no _ other haplobunodontid]. Discussion. From Pavlow’s (1900: pl. 5, figs 6-7) figures (fig. 6 reproduced by Stehlin, 1908: 773, text-fig. 112), H. ruetimeyeri, with large mesostyle, large paraconule and strongly oblique buccal edge distal to the mesostyle, fits Sudre’s (1978b: 94-99; pl. 4, figs 7-8; pl. 5, fig. 5) concept of Anthracobunodon better than Haplobunodon. It could even be conspecific with A. louisi Sudre 19786 from Grisolles. Moreover, from outline shape and cusp pattern, the teeth in Pavlow’s fig. 6 are more likely to be DP*—M! than M?~?. Haplobunodon venatorum sp. nov. (PI. 33, figs 1-5; Pl. 34, figs 1-4; Text-figs 63-64; Table 35) v. 1977b Haplobunodon sp.; Hooker: 141. vp. 1980 Haplobunodon spp.; Hooker & Insole: 44. Name. Latin, ‘of the hunters’, in allusion to the remains of a Mediaeval hunting lodge on the summit of Creechbarrow. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 405 HooryPe. Right P,, M37540. PI. 34, fig. 2. PARATYPES. Two right P*s, one fragmentary (M37712, M37528); right M!? associated (M37529); two fragmentary M'/?s (M37530, M37713); two left M*s (M37531-2); right M? (M37533); three fragmentary upper molars (M37535-7); left P, (broken}-M, (broken) (M37716); right P,; (M37539); right P, (M37541); right M,,. (M37542); fragment of right mandibular ramus with M, (M37544); three left M3s, all lacking trigonid (M37545, M37714—5); and right lower molar fragment (M37543). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DiaGnosis. Large Haplobunodon (length of M' = 8-6mm). P* broader than long, subtriangular in outline, with large protocone and strong ectocingulum and pre- and post-cingula. P, with metaconid and trifurcating postprotocristid. P, more than half as broad as long with meta- conid and small entoconid present; and paraconid subequal in height to metaconid. Horizontal mandibular ramus not deepening beneath M, 3. DIFFERENTIAL DIAGNOSIS. H. muelleri is much smaller and has a mandibular ramus deepening from M, 3; P34 are unknown. H. lydekkeri is slightly smaller; has a P* with an equilateral triangular outline, a smaller protocone, weaker ectocingulum and no pre- and post-cingula; a P, without metaconid and with single postprotocristid; and a P, with a large metaconid subequal with protoconid and a smaller paraconid. H. solodurense is slightly smaller, has a P; without a metaconid and a narrower P, with a paraconid but no metaconid or entoconid. Table 35 Length (1) and mesial (w,) and distal (w,) width measurements of Haplobunodon venatorum from Creechbarrow. Two width measurements are only given for lower molariform teeth. Measurements in millimetres. No. Tooth ] Wi W> M37712 RP* 7-4 8-5 M37529 RM! 8-6 (10-3) M37529 RM? 9:3 11:8 M37532 LM? 9-6 = M37531 LM? 9:5 12:4 M37533 RM? 9:5 11-4 M37716 LP, = 3 7/ M37716 LM, 8-5 6:2 6:3 M37716 LM, = 79 = M37539 —s RP, M37540 RP, 8-1 5-4 M37541 RP, 8-3 Sy M37542 RM, 8:5 6-1 6:6 M37544 RM, 13-5 79 6:9 M37545 LM, = = 6:2 M37714 LM, = = 6:6 M37715 LM, = = 1F22 M37543. -RM,,»3 2 6-7 2 DescrIPTION. P*: M37712 is quite worn. The cingula are faintly papillate. The ectocingulum, though strong, is interrupted between the midpoint on the paracone and the extensively crested parastyle. A small area of exposed dentine on the postparacrista indicates the presence of a tiny metacone. Its presence or absence in this state is intraspecifically variable in H. muelleri (see Stehlin 1910b: pl. 17, figs 35, 37). There is a worn swelling midway along the distal cingulum, which may have once been a small metaconule. There is a separation between this point and the postprotocrista which stops short, joining neither distal cingulum nor postparacrista. J. J. HOOKER 406 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 407 13 a] 105 —@ w w © © 2) ite Uppers Lowers T = Soo T oases ret | 4>—— T af =e =~0 T T 7 7 10 11 1 10 14 Text-figure 63 Scatter diagrams of length (1) against width (w) in upper and lower cheek teeth of species of Haplobunodon. Measurements in millimetres. Width of lower molariform teeth is talonid width. © = H. venatorum sp. nov. from Creechbarrow; [] = H. aff. venatorum from Mormont; O = H. lydekkeri Stehlin from Hordle Cliff. Symbols solid above = P4; solid on left = M1; solid on right = M2; completely solid = M3; outline = M1/2. Lines join teeth of one individual. M37528 has all the distal half and the lingual edge broken away. It resembles M37712 in what remains of the cingula, although the parastyle appears more weakly crested and it is less worn. A very worn right P* of Haplobunodon from Mormont (LGM 9157: LM 797, figured Stehlin, 1908: pl. 13, fig. 36) resembles M37712 except in that it is slightly larger, has more papillate cingula and the ectocingulum is more continuous. M'*: Two preultimate molars which were found in the same bag of matrix, and whose interstitial facets match (M37529) are the only ones identified with certainty as being either tooth. M? is larger than M! and less worn. The ectocingulum is so weak on M7? that it is nearly missing (M' is abraded in this area). The lingual cingulum in both is restricted to a short stretch between the protocone and metaconule. M37713 is broken but almost unworn, has a small ?hypostyle at the distal end of the postmetaconule crista, an almost complete cingulum round the metaconule and a low ridge in the valley, between the protocone and metaconule, which may represent a prehypocrista. Of three M'/*s of Haplobunodon from Mormont (figured Stehlin, 1908: pl. 13, figs 16, 36 and 41) the last has thin enamel and is low-crowned, and is thus here reidentified as a DP*. Of the remaining two, LGM 9166: LM 852 is the same length but slightly broader than the M37529 M! and has an almost complete ectocingulum; and LGM 9146: LM 854 is larger in both dimensions than the M37529 M?, has a complete papillate ectocingulum, a faint postcingulum and signs of a prehypocrista. M?: This tooth varies in outline shape from being like the M!' *s (M37529) to distally attenuated (M37531). The lightly worn example, M37529, can be distinguished from M'’s in Plate 33 Light macrographs of cheek teeth of Haplobunodon and Dacrytherium from Creechbarrow, x 3. Figs 1a, 2a, 3a, Sa, 6a, 7a are buccal views. Figs 1b, 2b, 3b, 4a, Sb, 6b, 7b are occlusal views. Figs 4b, Sc are lingual views. Figs 1-5 Haplobunodon venatorum sp. nov. Fig. 1a, b, right P* (reversed) (M37712). Fig. 2a, b, right M!? (reversed) (M37529). Fig. 3a, b, right M? (reversed) (M37533). Fig. 4a, b, right P, (reversed) (M37541). Fig. Sa—c, left M, talonid fragment (M37545). See p. 404. Figs 6-7 Dacrytherium elegans (Filhol). Fig. 6a, b, right M'~*, M! fragmentary (reversed) (M37717). Fig. 7a, b, left M!/? (M36814). See p. 410. 408 J. J. HOOKER having the metacone lower than the paracone. M37531 also has a buccally salient metastyle. There is also variation in strength of the ectocingulum. P,: M37539 preserves only the distal half. Quite a large metaconid appears to have been present. The postprotocristid trifurcates, only the convex buccal limb reaching the distal tooth margin, where there is a small cusp. H. solodurense is similar but has no metaconid (Stehlin 1908: pl. 14, fig. 48; 1910b: pl. 20, figs 10, 62). H. lydekkeri (BM(NH) 29713) has no metaconid and only a single straight postprotocristid. The characters, inasmuch as they are constant for three specimens of H. solodurense, are considered to have specific value. P,: The two complete specimens both have very small entoconids but vary considerably in size of paraconid and metaconid although retaining the same relative sizes of these cusps. In the holotype (M37540), the paraconid is a large swollen cusp, slightly worn, but which was prob- ably nearly as tall as the protoconid; the metaconid is a much narrower cusp; and the valley between this and the paraconid is very narrow. M37541 is more worn, but the metaconid forms a small distolingual bulge from the protoconid; the paraconid is much less inflated than on M37540: and the valley between the paraconid and metaconid is wide and very shallow. The M37716 P, has the paraconid broken away but has a small metaconid like M37541 and a slightly larger entoconid. Three P,4s of H. solodurense are constant for the specific differences from H. venatorum but vary in the size and position of the paraconid. In NMB Ef245 it is nearly as large as in M37541 but less lingual with respect to the protoconid (see Stehlin 1908: pl. 14, figs 35, 48; 1910b: pl. 20, fig. 55). Lower molars: No tangible morphological differences between this and the other Hap- lobunodon species have’ been detected. In H. venatorum there is slight variation in size of the ectostylid. In M, there is also slight variation in the orientation of the mesial hypoconulid crest (longitudinal or oblique) and the position of its attachment mesially: midway between the hypoconid and entoconid in M37544—5 and M37715, or directly to the hypoconid by way of the postcristid in M37714. Mandible: The M,, M37544, is one of very few specimens found in the field during excava- tion work at Creechbarrow. It is associated with a shattered piece of mandibular ramus, which shows the configuration in the region of M; to have been like that of H. lydekkeri (see Text-fig. 64 and Cooper 1928: pl. 2, fig. 2) and H. solodurense. Text-figure 64 Medial view of right mandibular fragment with M, of Haplobunodon venatorum sp. nov. (M37544) from Creechbarrow. Arrow indicates course of dental canal. x 2:3. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 409 Discussion. Considering the intraspecific variation in the Creechbarrow assemblage, it is prob- able that most of the Mormont material is conspecific with it, although exact identity is not demonstrable. It will thus be referred to as H. aff. venatorum. Its age is considered to be Marinesian, based on one specimen labelled ‘Eclépens’ and none being labelled “Entreroches’. The small upper molar from Entreroches (NMB Mt373; see Stehlin 1908: pl. 13, fig. 33; note that this is mislabelled in his plate legend as LM 853) is outside the range of size variation expected for H. venatorum and coincides closely with the size of H. solodurense. The species morphologically closest to H. venatorum appears to be H. solodurense from the late Lutetian—Auversian of Egerkingen. It would be possible to derive the former from the latter by increase in overall size and also in premolar molarization, including the curious enlargement of the paraconid. However, there appear to be no temporal trends within the Egerkingen assemblages towards H. venatorum. Ludian H. lydekkeri appears more primitive in its para- conid and is unlikely to be the descendant of H. venatorum. H. muelleri appears to be more distantly related, as its jaw characters are not shared with the other three species. On this basis and on the presence of P, in H. muelleri, Stehlin (1910b: 1112) considered that it might eventually be placed in a separate genus. Unfortunately, all others except the type species are too incomplete to know whether they had Pt. Family ?>CHOEROPOTAMIDAE Owen 1845 (sensu Sudre 19785) Genus ?CHOEROPOTAMUS Cuvier 1822 ?Choeropotamus sp. indet. vp. 1980 Choeropotamus spp.; Hooker & Insole: 44. MATERIAL. Fragment of right P?? (M37563) and fragment of right P*? (M37566). HORIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. DESCRIPTION. M37563 is broken mesially and distally, has a narrow wedge-shaped outline, one large cusp bordered by strong cingula and a very rugose surface. M37566 could be a distal fragment of P*; it is more finely rugose than M37563, has a strong papillate cingulum contin- uous throughout the preserved edge and a slightly papillate crest which might be a postpara- crista. Neither fragment fits any other taxon recorded from Creechbarrow but both resemble Choeropotamus in size, apparent bunodonty, enamel rugosity and probable outline. Size would correspond with C. lautricensis from the Marinesian of France. The identification is intended to be only very tentative, however. Both teeth are so fragmentary that no definite identification should be made even to ordinal level. Superfamily ANOPLOTHERIOIDEA Bonaparte 1850 (rank emend. Romer 1966) (sensu Sudre 19785) Family ANOPLOTHERIIDAE Bonaparte 1850 (sensu Viret 1961) DIAGNOsIS. See Viret (1961: 935). Subfamily DACRYTHERIINAE Depéret 1917 (rank emend. Viret 1961) TYPE GENUS. Dacrytherium Filhol 1876a. INCLUDED GENERA. Catodontherium Depéret 1908, Leptotheridium Stehlin 1910b and Tapirulus Gervais 1850. RANGE. Middle Lutetian—early Stampian, Europe. DiaGnosis. See Viret (1961: 936). 410 J. J. HOOKER Genus DACRYTHERIUM Filhol 1876a [incl. Plesidacrytherium Filhol 1884; for more detailed synonymies see Stehlin, 1910b: 882-900. ] Type spEciES. D. cayluxi Filhol 1876a. Early Ludian limestone; Lamandine, Quercy, France. Following Stehlin (1910b: 843) this species is almost universally synonymized by authors with D. ovinum (Owen 1857b), from the Lower Headon Beds of Hordle Cliff, Hampshire (not the Isle of Wight as originally stated by Owen, 1857b, and followed by Stehlin, 1910b, and Sudre, 1978b). INCLUDED SPECIES. D. saturnini Stehlin 1910b; D. priscum Stehlin 1910b; D. elegans (Filhol 1884) Stehlin 1910b; and D. cf. elegans. RANGE. Middle Lutetian to Ludian; England, France and Switzerland. DIAGNosIs. See Sudre (1978b: 110). Dacrytherium elegans (Filhol 1884) Stehlin 1910b (Pl. 33, figs 6-7; Text-fig. 65) v. 1977b Dacrytherium sp.; Hooker: 141. v. 1980 Dacrytherium sp.; Hooker & Insole: 44. Hotorypee. Cranial fragment with P?-M?. Early Ludian limestone; Lamandine-Haute, Quercy, France (Filhol 1884: 191-192; pl. 10, figs 5—6). RANGE. Marinesian, England, France and Switzerland; early Ludian, France. MATERIAL. Distal fragment of right P? (M37564); right M! (fragmentary)-M? (M37717); left M!/? (M36814); right lower canine? (M37548); left P, (M37718); and fragment of lower decid- uous premolar? (M36197). HoRIZON AND LOCALITY. Creechbarrow Limestone Formation, Creechbarrow. EMENDED DIAGNOSIS (essentially following Stehlin’s 1910b description). Small Dacrytherium (mean length of M' = 8-Omm). Upper molars strongly dilambdodont, with mesostyles broadly looped. Preorbital fossa shallow and restricted in extent. Discussion. The important character of the preorbital fossa obviously cannot be recognized when only teeth or jaws are available. This is the case with the Creechbarrow material. The pre-Marinesian species D. priscum and D. cf. elegans can be excluded from consideration since they are less dilambdodont. The teeth of the remaining species, however, appear to be only separable on size (see Sudre 1978b: 110). Text-fig. 65 is a scatter diagram of length against width of various upper molars of strongly dilambdodont species (mainly from published measurements). The Creechbarrow molars fit well with D. elegans from both the type locality and Eclépens-Gare. There is overlap, especially of M', with the slightly smaller D. saturnini, which appears otherwise to differ only in low size increase from M! to M? and in low length and high width variation of M'. This species, according to Stehlin (1910b: 929), has a deep preorbital fossa like D. ovinum, but the material used here for size comparison is from Ste Néboule (Sudre 1978a: 278), where no specimen with a preorbital fossa preserved is recorded. Plate 34 Light macrographs of cheek teeth of Haplobunodon and Dichodon from Creechbarrow, x 3. Views are buccal (a), occlusal (b or unsuffixed) and lingual (c). Figs 1-4 Haplobunodon venatorum sp. nov. Fig. 1a—c, distal half of right P, (reversed) (M37539). Fig. 2a-—c, holotype right P, (reversed) (M37540). Fig. 3a—c, left P, (paraconid broken), M, and trigonid fragment of M, (associated) (M37716). Fig. 4a—c, right M, (reversed) (M37544). See p. 404. Fig.5a—c Dichodon cf. cervinus (Owen). Left M, (hypoconulid broken) (M37549). See p. 413. Fig.6 Dichodon sp. indet. Distal fragment of left DP? (M37562). See p. 414. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 411 412 J. J. HOOKER 105 T T T 1 14 Text-figure 65 Scatter diagram of length (1) against width (w) in upper molars of species of Dacrytherium. Measurements in millimetres. Large rectangles delimit D. saturnini Stehlin from Ste Neéboule (from Sudre 1978a: 278). © = Lamandine; [] = Fons 4; © = Fons 5; A = Perriére; Y = Euzet (these three localities from Sudre 1978b: tab. 11); [>\= Eclépens B (from cast); (all D. ovinum (Owen) and D. sp. nov.?). ;y = Creechbarrow; 2 = Eclépens-Gare (from cast and Stehlin 1910b: pl. 16, fig. 10); 1S x ~ = = = = = 3 a S = S is) > ~ Pay =~ = > S S > AS Po) v 3S a) = x is} is) S) S = oS 3 = S 3 = = = TS i 3 = 3 5 = = = Ste) 6Se) cS ae Se Ses ans: >< x << a a A x a a Creechbarrow Limestone, Creechbarrow x x x x x Barton Clay, Hengistbury Gi Calcaire de St Ouen, Paris and Grisolles x cf. Sables de Cresnes, Berville x ‘Sidérolithique’, Eclépens-Gare x aff. x x (ex x (32) x x Lower Calcaire de Fons, Robiac x x x x x x x Sables du Castrais, Le Castrais x x x x x ‘Sidérolithique’, Le Bretou x x x ‘Sidérolithique’, Heidenheim 422 J. J. HOOKER the uniqueness of its occurrence and problems over its exact stratigraphical position require its exclusion. Anchilophus desmarestii was considered by Sudre (1969a) to be a characteristic Mari- nesian species, but it also occurs, admittedly with a cf. determination, at Egerkingen where it overlaps with the eocaenum-rhinocerodes Zone taxa. A number of fragmentary specimens of a large Lophiodon species, usually referred to L. cf. lautricense, occur at more than one level in marginal calcareous facies cited as being lateral equivalents of the Sables Moyens. They are from Arcis-le-Ponsart (Louis 1976) and Nogent-l’Artaud (Morellet & Morellet 1948), and their horizons have been correlated with the Facies (Horizon’) du Guépelle (Morellet & Morellet 1948) which contains an eocaenum-rhinocerodes Zone fauna. The specimens have never been figured or described in detail. Some overlap of the zones may occur, or the records may not represent L. lautricense but a different species, or the lithostratigraphic correlation of these marginal facies may not be accurate. A potential immediate ancestor for L. lautricense is not known. L. rhinocerodes has been discarded from this role by Sudre (1971) and his alternative (L. cuvieri) is considered herein (p. 375) to be more closely related and probably ancestral to L. thomasi. See Table 37 for distribution of taxa defining the zone. Eclépens-Gare is the only locality where all the taxa defining the zone occur together. The zone ranges from the Calcaire de St Ouen to the Sables de Cresnes in the stratified sequence of the Paris Basin. The faunal change immediately above this zone is well documented in a stratified non-marine sequence in the Alés Basin, southern France (Garimond et al. 1975). There may be a gap between the lautricense—siderolithicum Zone and that of Thalerimys [Isoptychus] headonensis (Bosma 1974) Vianey-Liaud 1979, defining the base of the Headonian in southern England. There is room in the earlier Ludian for the faunas of Fons (U. Calcaire de Fons) (Garimond et al. 1975). The lautricense—siderolithicum Zone seems to mark a particularly distinctive phase in Euro- pean Eocene mammal faunas. Of the total of over 100 named and unnamed species-group mammal taxa occurring in this Zone, only 3 range above and below; 14 range only below; 33 range only above; whilst 53 are unknown outside the zone. Of the 53 restricted forms, 8 have likely ancestors in the zone below; 13 have likely descendants in strata immediately above; whilst 2 are sandwiched between respective ancestors and descendants (see Table 38). Table 38 L. lautricense—L. siderolithicum Zone species-group taxa and main localities. The range column indicates occurrence extending below @, above @, above and below @, or restricted to the zone O. ?, aff., or cf. applied to one side or other of the symbol refers to qualification of the earlier or later record(s). Note that Amphiperatherium ?lamandini may be the same as A. aff. goethei, and that Berville, Heidenheim and Hengistbury (not listed) only contain Lophiodon lautricense (or cf.). Brackets indicate doubtful provenance. Sources as for Tables 36-37 with addition of Crochet et al. (1981). z 5 2 E a 9 = g =) Sie aes 2.» See mes ac SI = 5 5 Dye. ee GR US. Ooms Ss = 3 m xs (3) ° cP) is) ° c) oe O Ce aa a aq A we 2 Amphiperatherium giselense 2 _ — x = = = 3 A. minutum ©) = x = x A. bourdellense e@ — = x = x A. goethei © aff. = = a = = = = we A. lamandini rq) = = x = = A. fontense @) x — = = = = = x x Peratherium sudrei © = = x a = = a x be. P. bretouense O — — = = = = — x P. lavergnense ) ~ = = = x Saturninia mamertensis 2© — — = = = = = = x BARTONIAN MAMMALS OF HAMPSHIRE Table 38 (continued) S. grisollensis S. hartenbergeri S. beata S. grandis S. intermedia Gesneropithex grisollensis G. figularis G. sp. (Robiac) Rhinolophus sp. 1 Vespertiliavus gracilis V. wingei V. sp. 1 Nannopithex quaylei N. sp. 1 Pseudoloris crusafonti P. parvulus Necrolemur zitteli Microchoerus sp. (Grisolles) M. wardi M. creechbarrowensis Europolemur collinsonae Leptadapis magnus Adapis sudrei 2A. laharpei Plesiarctomys curranti P. hurzeleri Ailuravus stehlinschaubi Gliravus robiacensis Sciuroides rissonei S. russelli S. siderolithicus Treposciurus helveticus preecei Suevosciurus authodon Paradelomys crusafonti Elfomys tobieni Pseudoltinomys mamertensis Theridomys varleti Remys minimus Heterohyus sudrei H. nanus H. morinionensis Cynohyaenodon cayluxi Hyaenodon heberti Simamphicyon helveticus Quercygale angustidens Vulpavoides cooperi Propalaeotherium parvulum Lophiotherium siderolithicum Anchilophus desmarestii A. gaudini A. dumasii ge ~ 08000 G00 C800 COGOOOD00 GOOD D00 SOOODSBOOOSSBOOOSHOSEO Ran FR g eee g, 9 = BASIN 423 3 eg 2 BB S Gas r Sa le eG ee Ba Ss S 152) ie up tle mee iS Claes he Bw Gece baa cieome Oly) Clas cae Sea) Sota AS RS ei — te x = 9 a ae ee ey eS x = = = x = = x ) — = = x Lg 23 A Serle E ne = fe x = as Ls ye — — — — _ as = x = = = = = = = = = = x — — - - = - - -—- — ef. = = = x = = = = x x — x — — — — _ — — — — cf. _ x - = _ _ ~ — _ x — — x - = = - = x x ke = x ates f. ts ae td Dl = x x —_— _— — — — — — — — x Zz reg | hee =a eee = aff. — - -—- = — - - = - - - cf. x _ _ _ — — — x — — — — x = = = = (f) = = = = x _ — = = (x) x — x — x — — — — - _ aff -— — = - = x x x = ah = = SA 6g Fe ls = = x =) ee = ee se ae - - — - = x = - x x x = ad Yn ey oa = ae eee = S = = a = eee a = - = f& - — — = uch.) «ch = = ie x = — - ? - = ~ - — — aff. = = x yee = he ps = = - = = — - = x cf. cf. ~ x - = _ - = x — aff. ~ - - = — —- = x = x — — _— — — — — — — _ — — - = ~ - -- — x — _ - = = - - -— aff —- = — — _ _ x — - x x _ ~ - — x — = x x x ae oS pel 5 = Rap Re ta za aff. - - -— — aff. - = \= cf. x — - = = x - = x = = — — ee Fe alixs) - — ef. x = _ - - = x cf. — _— ef. x — — - = = x =" = Tele ick 424 Table 38 (continued) J. J. HOOKER A. depereti Plagiolophus curtisi curtisi P. curtisi creechensis P. cartailhaci P. annectens Leptolophus nouleti L. stehlini Palaeotherium ruetimeyeri . pomeli . castrense castrense . c. robiacense . siderolithicum . lautricense . muehlbergi Lophiodon lautricense L. thomasi L. tapiroides Chasmotherium cartieri Cebochoerus minor C. robiacensis Acotherulum campichii Mixtotherium gresslyi M. infans Haplobunodon venatorum Anthracobunodon louisi Choeropotamus lautricensis Dacrytherium elegans Catodontherium robiacense C. ?paquieri Leptotheridium lugeoni L. traguloides Tapirulus schlosseri Robiacina minuta R. weidmanni Xiphodon castrense Paraxiphodon cournovense Haplomeryx picteti Dichodon cervinus D. subtilis Pseudamphimeryx pavloviae P. renevieri P. valdensis 4s) as) ae} ae) ae) ‘as! ge 08 G0G00000 0800000088008 8000808080800 80008 Ran <) ae Es 2 oO = ee = oO 2 S| | 9 ‘ss E 3 = Some 2+ one) aapetaaene = o Oo re ee a ee = ~ - = = — ci = efits = x = ~~ es Se rar 2 a x = x Sh aff. aff. — aff. aff. — — — — — _ x — —_— — _ — — ~ -- (x) x _ x _ = = = = = = - — aff —- - x x x — — — — — — — x _ — _— — aa - x = - _ x x — — - _ - = x x — x cf. _ = - - = cf. x =| 8S — — aff. = _ - ef — x x - x - — — — — x< — _ — — — DG _ _ — — = x — - = x x — — a x _ - = = — x — x — x _ x - = x — — x — aff. — - -— — aff. ~ - = — — cf. - = = — x - - -— — aff. - - = — = = x = = = zs Sey be x = x = x - - -—- = x - -—- — aff. aa x L238 x = _— — — _— _— —_— x — — — 2s ae a Oe x —. =a = — - - = = - = -ss0 ich ylaft. — - - = = x - = x x — - - = = - - = x x — — — —_— _ x —_— —_ — —_— - x x = x x 2 = me | = = ey jes x = — = - - = x - = x x cf. =_ - - — . cf? = = = — = a 5 ean x = os ~ — — - =- = ~ - = x x — = - = = x - —- aff —- = = a x 2/7 =—— = ii. Discussion. An attempt has been made in each zone to find the same genera represented by different species, and use them as zonal indices, with varying degrees of success. In some cases it is thought that the species represent stages in evolutionary sequences — e.g. Lophiotherium pygmaeum to L. siderolithicum (see Savage et al. 1965, and herein); Palaeotherium eocaenum to P. lautricense (see Franzen 1968); and Haplobunodon solodurense to H. venatorum (p. 409). In the former case, transitional ‘advanced’ forms of L. pygmaeum occur at Egerkingen f (see BARTONIAN MAMMALS OF HAMPSHIRE BASIN 425 Stehlin 1905b). In the case of Ailuravus stehlinschaubi and Lophiodon thomasi the probable ancestral species (A. picteti and L. cuvieri) occur in the eocaenum-rhinocerodes Zone, but also range down into earlier Lutetian strata. Three of the lautricense—siderolithicum Zone index species (L. siderolithicum, Acotherulum campichii and Plesiarctomys hurzeleri) also have prob- able descendant species (L. cervulum, A. saturninum and P. gervaisii) in Ludian strata. In the former case, the Alés Basin has yielded a series of transitional forms from the Fons localities, linking L. siderolithicum from Robiac to L. cervulum from Euzet (Remy 1967, Garimond et al. 1975). The species Lophiodon lautricense, Chasmotherium cartieri, Leptolophus stehlini and Ailu- ravus stehlinschaubi last occur at the top of the lautricense—siderolithicum Zone, marking also the last appearance of the family Lophiodontidae, the family Helaletidae in Europe, the genus Leptolophus and the subfamily Ailuravinae in the stratigraphic record. The principal and most widespread taxa as well as those forming stratigraphical morpho- clines have been chosen to define the two zones erected here. Numerous other lesser-known taxa with assumed similar evolutionary relationships occur through this period of time. iv. Further subdivision by mammalian biostratigraphy. It is difficult to locate the Lutetian— Auversian boundary within the eocaenum-rhinocerodes Zone. Changes from one fauna to another in the Egerkingen fissures are highly relevant to this problem but the stratigraphical order of these fissures is unknown, being postulated on evolutionary grade. It is probable that Egerkingen f is the most recent; Egerkingen « appears more advanced than Egerkingen y but less so than Egerkingen f (see Hartenberger 1972). It is possible that with the probable chro- nological range involved these fissures straddle the Lutetian—Auversian boundary. A study in progress on early species of Plagiolophus may cast further light on this problem. Detailed correlation between the Hampshire and Paris basins has already been shown to be a problem and has ramifications in any attempt to arrange the various Bartonian mammal localities in succession. Text-fig. 67, however, attempts this by piecing together the various partial correlations which have been made using different groups of organisms. Hartenberger & Louis (1976) placed the mammal fauna of Grisolles, which occurred in about the middle of the Calcaire de St Ouen in the east of the Paris Basin, later than Robiac and earlier than Fons 4 in the scale of niveaux-reperes, all three being included in the Marinesian. This contrasts with most authors who place Fons 4 near the base of the Ludian (e.g. Garimond et al. 1975) and is contrary to evidence from the charophytes (Grambast 1962, 1964, 1972), which correlates Robiac with higher, not lower, parts of the Calcaire de St Ouen. Moreover, Grisolles contains a number of species typical of the lautricense—siderolithicum Zone but which do not occur in strata immediately overlying those from which the Robiac fauna has been obtained (Garimond et al. 1975): i.e. Pseudoloris crusafonti, Adapis cf. sudrei and Acotherulum campichii. The rest of the fauna too is very different from those of Fons 1, 2, 5, 6 and 7, and also from Le Bretou (except for some long-ranging species), which occupy Hartenberger & Louis’ (1976) postulated level for Grisolles in the Ales Basin (see Sigé 1976, Sudre 1978b, Louis & Sudre 1975, Crochet 1979, 1980 and Crochet et al. 1981 for faunal details). Grisolles, which is demonstrably well within the Bartonian, is therefore considered to be earlier than Robiac which is dated as late Marinesian because of its position immediately below the L. lautricense— L. stehlini extinction datum. The reasons given by Hartenberger & Louis (1976) for their dating of Grisolles relative to strata in the Ales Basin was based on evolutionary grade in certain rodents. The arguments ran thus: 1, Theridomys varleti and Pseudoltinomys sp. are more primitive than T. euzetensis and P. mamertensis from Fons 4; 2, the Pseudoltinomys is very similar to that from Le Bretou, a fissure locality considered slightly younger than Robiac; and 3, Suevosciurus russelli and Gliravus aff. robiacensis are more advanced than S. romani and Gliravus robiacensis from Robiac. The Pseudoltinomys by their own admission is very poorly known; the intrageneric relationships of Suevosciurus russelli (shown on p. 300 to belong to Sciuroides) with S. romani (synonymized with Sciuroides siderolithicus, p. 300) are obscure; it seems likely that S. siderolithicus gave rise to S. ehrensteinensis via transitional forms recorded at Weidenstetten (Schmidt-Kittler 1971a) and Eclépens B, whereas S. russelli is closer to S. rissonei, which occurs at Creechbarrow 426 J. J. HOOKER z fo > £ he eee] - 9 = (>) 3 a a E re) 3 Qa Eo ~ fs <= Ce) o 2 ° cw D = o a) 2 2 o 2m o PS he ° tS oO QO & = oO ppc =—1 Text-figure 67 Correlation of European Bartonian strata. Stratigraphical columns range into strata higher and lower than Bartonian in most cases. For key to lithological symbols, see Text-fig. 2, p. 201. Abbreviations of lithostratigraphy are given on left of columns and abbreviations (mammals) and symbols (other biota) of biostratigraphy are given on right of columns. Jagged break in Alum Bay column indicates the omission of c. 9m of Barton Clay. Abbreviations of lithostratigraphy: BC = Barton Clay Eg = Egerkingen; FM = Formation de Mortefontaine; Formation; BeS = Becton Sand Formation; FRR = formation rouge de Robiac; LCF = lower Cal- BMB = Bournemouth Marine Beds; BoS = Boscombe caire de Fons, LH =Lower Headon Beds; Sands: CC = Calcaire de Champigny; CD = Calcairede MH = Middle Headon Beds; ML = Marnes a Lucina Ducy; CG =Calcaire Grossier; CL = Creechbarrow inornata; MP=Marnes a Pholadomya ludensis, Limestone Formation; CO =Calcaire de St Ouen; SC=Sables de _ Cresnes, SD = Selsey division; CPa = Calcaire a Potamides aporoschema; SE= Sables d’Ezanville; SM = Sables Moyens; DPC = Dorset pipe clays; E=marnes d’Euzet,; UCF = upper Calcaire de Fons. EcB = Eclépens B; EcG = Eclepens-Gare; associated with a typical lautricense-siderolithicum Zone fauna. This leaves only the Gliravus at Grisolles supporting a younger age than Robiac. The coefficients of variation given by Harten- berger & Louis (1976: 89, tab. 4) for the lengths of upper and lower preultimate molars of Gliravus aff. robiacensis from Grisolles exceed 9, which is a good indication that more than one species is represented. In view of these problems, it is considered that more confidence can be Studley Wood |, Whitecliff Bay BARTONIAN MAMMALS OF HAMPSHIRE BASIN . Paris Basin ; N.W. Paris [ 10 metres Oo Text-figure 67, continued. Abbreviations of mammal taxa: Ac = Acotherulum campichii; Cr = Cebochoerus ruetimeyeri; Chd = Choeropotamus depereti; Chl = C. lautricensis; Chs = C. sudrei; Dr = Dichobune rober- tiana; Ll = Lophiodon lautricense; Lr = L. rhinocerodes; Lmc = Lophiotherium cervulum; Lmp = L. pygmaeum; Lms = L. siderolithicum; Pcr = Palaeotherium castrense robiacense; Pe=P. eocaenum; Plcc = Plagiolophus curtisi curtis: PIGcGa— curtisi creechensis; Tf = Thalerimys fordi; Th = T. headonensis. Biotic symbols: Dinocysts: & = first occurrence of Areoligera undulata and Areosphaeridium multicornutum: } = first occurrence of Rhombodinium draco; = base zone BAR-2 of Bujak et al. (1980); © = first occurrence of R. porosum; @=base zone BAR-4; © = first 427 Switzerland Alés Basin occurrence of Homotryblium floripes and Kisselovia cla- thrata angulosa. Charophytes: © = zone of Nogent- PArtaud; @; = zone of Chéery-Chartreuve; @ = zone of St Ouen; @ = zone of Robiac; © = zone of Verzenay (base). Nummulites: @ = N. prestwichianus; ® = N. rectus first appearance. Coral: = Paracyathus. Gas- tropods: g = Ectinochilus planum; | = Potamides apo- roschema. Ostracod: © = Cytheretta cellulosa. Sources: Bosma (1974), Bujak et al. (1980), Chateau- neuf (1980), Costa et al. (1976), Curry (1942, 1977), Deperet (1910), Ginsburg et al. (1965, 1977), Grambast (1962, 1964, 1972), Keen (1978), Kemp et al. (1979), Lemoine & Abrard (1926), Louis & Sudre (1975), Roman (1904), Stehlin (1903, 1904b, 1905b, 1906, 1908, 1918), Sudre (1969a, 1978b). attached to other elements in the mammal fauna and to the charophytes than to rather generalized ideas of evolutionary grade in those rodent species whose exact relationships are still unclear. Mammals recorded from other Calcaire de St Ouen localities include Palaeotherium castrense robiacense Franzen 1968 from about the middle of the formation in Paris (Lemoine & Abrard 1926). This subspecies is recorded from the top of the lautricense—siderolithicum Zone at Robiac, 428 J. J. HOOKER associated with the Robiac charophyte zone. Fossils characterizing the latter zone are recorded from immediately below the Marnes a Pholadomya ludensis in the northern Paris Basin, thought to be equivalent to the Sables de Monceau in the region of Paris (Cavelier 1979: 157). P. c. robiacense appears thus to occur at a lower horizon here, but the problems posed for correlation within the Paris Basin by the condensed sequences and the lateral facies changes near the basin margins make this by no means the only interpretation. The relationship to Grisolles, also occurring at about the middle of the formation, is equally a problem, especially as there is no mammalian faunal overlap between the two localities. In England, only two taxa are common to the Creechbarrow Limestone and the Barton Clay: Microchiroptera gen. et sp. indet. 1 (p. 241) and Plagiolophus curtisi (p. 353). The slight differences of size and morphology between the two assemblages of the latter have resulted in two subspecies being described. On the theoretical concept of evolutionary grade, P. curtisi creechensis from Creechbarrow is considered slightly more primitive than P. curtisi curtisi from Barton bed D/E, and therefore slightly earlier. Creechbarrow and Grisolles share only two short-ranged species: Pseudoloris crusafonti (cf. determination at Creechbarrow) and Acotheru- lum campichii. Ideas of evolutionary grade in two other genera give contradictory results: Creechbarrow Sciuroides rissonei appears in some respects to be more primitive than Grisolles S. russelli, whereas the Grisolles Microchoerus is more primitive than and intermediate in size between Creechbarrow M. wardi and M. creechbarrowensis. The choice is difficult but the evidence from Microchoerus is more convincing as it fits better with the subsequent sequence of events in Microchoerus evolution; in contrast, the relationships of S. russelli are more obscure. Grisolles is thus tentatively considered to be slightly earlier than Creechbarrow, which in this case must be Marinesian rather than late Auversian. The various mammaliferous localities of the Castrais region of the Aquitaine Basin (Sables du Castrais) are thought to vary somewhat in age (see Richard 1946 for faunal lists; Sudre 1969a). They were considered by Remy (1965) to be earlier than Robiac because they contained a lower-crowned and therefore more primitive species of Leptolophus, L. nouleti (Stehlin 1904a) Remy 1965, than at Robiac (L. stehlini). Increase in crown height is almost universally con- sidered to be an advancement, but at other localities in Le Castrais L. stehlini also occurs. It is most likely that the latter are younger than the localities yielding L. nouleti. Castrais L. stehlini, moreover, has slightly lower crown height than Robiac L. stehlini, although the difference is not nearly so marked as with L. nouleti. L. stehlini from Mormont (Eclépens A and probably, although not labelled, Eclépens-Gare) is very similar in crown height to that of Le Castrais, suggesting close correlation. An M,,, from the Auversian of Le Guépelle was recorded as ‘Plagiolophus taille de P. annectens’ by Ginsburg et al. (1965). It more closely resembles Lepto- lophus in morphology and is at least as low-crowned as L. nouleti. It is thus possible that the Castrais localities with L. nouleti are as old as Auversian. Other elements in the Castrais and Eclépens-Gare faunas have been shown to be slightly different from those of Robiac, indicating that they are probably slightly older (e.g. Sudre 1969a: Franzen 1968), but relationships with localities such as Creechbarrow and Grisolles are less easy to envisage. Creechbarrow and Eclépens-Gare share a number of short-ranged species (viz. Lophiotherium siderolithicum, Haplobunodon venatorum and Ailuravus stehlinschaubi), but there are subtle differences in these, e.g. in size, which could just as easily result from geographi- cal as from stratigraphical distance. The possibility of Creechbarrow Sciuroides rissonei being the ancestor of Eclépens-Gare S. siderolithicus suggests that Creechbarrow is earlier than Eclépens-Gare. Le Bretou has been considered to be slightly younger than Robiac but still Bartonian, on the basis of the occurrence of Remys minimus, Robiacina minuta, Xiphodon castrense and Simamphicyon helveticus (Hartenberger et al. 1974; Sudre 1978b: 185). This is also supported by the presence of Sciuroides siderolithicus, probably that recorded as Suevosciurus (Treposciurus) mutabilis by Hartenberger (1973: 65) and as Suevosciurus cf. romani by Hartenberger et al. (1974). Below is the suggested succession of European Bartonian mammal localities. It conforms with correlation based on other groups. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 429 England Paris Basin Switzerland S. France Le Bretou Paris? Robiac Barton C-F Berville Eclépens-Gare Le Castrais (Lautrec etc.) Creechbarrow Paris? Grisolles Hengistbury Sergy/Latilly? Egerkingen $? Le Castrais (Viviers)? Le Guépelle : or jen Egerkingen «? Lissieu? The solid line indicates the boundary of the two zones and the dashed line the Auversian— Marinesian boundary. The ‘? indicates doubt regarding position and, in the case of Paris, two possible positions for the Calcaire de St Ouen specimens are suggested. Other localities like La Liviniére in southern France and Laguarrés in Spain are thought to be early Bartonian but have so far yielded sparse mammal faunas, or only endemic taxa or those which are not distinctive of age (see Hartenberger 1973: 67 for lists). Lophiodon rhinocerodes and Palaeo- therium r. ruetimeyeri have recently been recorded from the Grés de Brenne at Fonliasmes, near Poitiers, France (Brunet & Gabilly 1981) and attributed an Auversian age. The two taxa certainly correspond to the eocaenum-rhinocerodes Zone but could be either late Lutetian or Auversian. It is worth noting that the junction of the two mammal zones must be close to the base of the Rhombodinium draco dinocyst zone and thus the mammal evidence supports Curry’s (1981) proposed modification of the base of the Bartonian. Evidence from Hampshire Basin sedimentation West of Whitecliff Bay, the beginning of a series of facies changes from predominantly argilla- ceous to predominantly arenaceous strata of less marine aspect affects most of the post-London Clay and pre-Fluviomarine strata. The result is that most fossils disappear and dating the western sequences is difficult. At intermediate localities such as Afton, Alum Bay, Barton Cliff and Bournemouth, thinning, facies interdigitation and the persistence of some groups of organisms (e.g. dinocysts) as far west as Bournemouth allows some dating of problem strata from the late Lutetian into the Bartonian (Eaton 1976; Costa et al. 1976). Actual facies change trends are observable at Hengistbury where there is an increase in sand content and abundance of pebble stringers in a westerly direction along the section in the Barton Clay (‘Hengistbury Beds’) (Hooker 1975). Fisher (1862: 86-88) noted an increase in abundance of pebbles westwards in the basal pebble bed of the Barton Clay at High Cliff, and (1862: 89, figs 3-4) suggested correlation of what is now referred to the upper part of the Boscombe Sands at High Cliff with what is herein referred to lower parts of the Barton Clay Formation in Alum Bay. Curry (1977), from recognition of two nummulite horizons in the lower Barton Clay (as ‘Hengistbury Beds’), correlated an overlying sand at Hengistbury with Bed A3 (High Cliff Sands and Clays of Wright, 1851) at High Cliff and Barton Cliff. Because of the interruption in the coastal exposure at Poole harbour west of Bournemouth, combined with more facies change and reduction in fossil content, the Creechbarrow sequence is very difficult to match with those further east. The only key seems to be the giant flints which occur at various horizons from c. 7m below the Creechbarrow Limestone downwards. Flirts of comparable size are recorded from the Boscombe Sands of Bournemouth (Gardner 1879: 218; Prestwich 1849: 46-47), the slightly diachronous basal Barton Clay pebble/cobble beds at Hengistbury and Alum Bay. They thus appear to occupy a relatively narrow span of strata east of Poole Harbour and it is logical to suggest that they do the same at Creechbarrow. If the Creechbarrow Limestone represents the end of a regression like similar limestones in the later Fluvio-marine series then the stratigraphically closest regression in the marine sequence, sand- 430 J. J. HOOKER wiched between the Boscombe Sands and Barton Clay Bed D (from faunal evidence; see Text- fig. 67, pp. 246-7) is that of Barton Clay Bed A3. If the progressive reduction of clastic sediment in the Creechbarrow sequence as the limestone is reached represents a raising of the water table resulting from a transgression, then a correlation with Barton Clay Bed B (the nearest trans- gressive stratum) might be more appropriate. Palaeoenvironments, palaeoecology and palaeogeography Depositional environments The marine province. The Barton Clay and Becton Sand Formations in Christchurch Bay have been shown in the stratigraphical section, pp. 197-212, to have been deposited in a series of three cycles. The erosive base of each cycle is taken to indicate the rapid transgression of a shelf sea, followed by a gradual regressive (progradational) phase, ending in shoreface sands. At the western end of Christchurch Bay at Friar’s Cliff and Hengistbury, there is evidence in lower parts of the sequence of two more cycles. These occur in the upper part of the Boscombe Sands, which here underlie the Barton Clay Formation. The upper of these cycles has a basal pebble bed and passes from non-glauconitic clayey sand to non-glauconitic sand, both with Ophiomorpha and plant debris. At Hengistbury this cycle spans Boscome Sands beds 2 and 3 (Hooker 1975: fig. 3). At Friar’s Cliff it spans the ‘Highcliff Sands’ of Gardner from the pebble bed upwards (see Hooker 1975: fig. 2) or beds 2 and 3 of Fisher (1862: 87, fig. 2). The lower of these cycles at Hengistbury has a basal pebble bed, resting on laminated lignitic clays and consists almost entirely of light non-glauconitic sand with Ophiomorpha, with an impersistent seam of pipe-clay at the top (Boscombe Sands bed 1 of Hooker, 1975: fig. 3; the underlying ‘?Bournemouth Marine Beds’ were incorrectly identified). Only the upper part of this cycle is exposed at Friar’s Cliff, where it consists of a thicker unit of light-coloured sands (‘Highcliff Sands’ of Gardner below the lower pebble bed shown in Hooker 1975: fig. 2). The equivalent in Alum Bay is considered to be the sands of bed 28 of Prestwich (1846). When traced eastwards these sandy units successively pass laterally into glauconitic sandy silts and silty clays of lower parts of the Barton Clay Formation (Huntingbridge division, including Elmore Member in part). The latter at Afton and Whitecliff Bay seem to continue up into time-equivalents of the lower part of the lower cycle of the Barton Clay in Christchurch Bay, entirely within Elmore facies. They cannot here be subdivided into cycles. These recogni- tion problems are likely to be because successive cycles, up to the lower one of the Barton Clay in Christchurch Bay, transgress each other. The progradational (regressive) phases of the first two did not develop seaward sufficiently to cause facies changes in the more offshore sequence further east (e.g. Afton and Whitecliff Bay). The cycles will here be numbered 1 to 5. The lower, middle and upper of the Barton Clay and Becton Sand Formations in Christchurch Bay, described on pp. 206-7, are numbered 3, 4 and 5 respectively. The lower and upper of the upper part of the Boscombe Sands are numbered 1 and 2 respectively. See Text-fig. 68 for the model. The glauconitic facies of the cycles are best developed in the west of the area (Christchurch Bay, Alum Bay). Glauconite is considered to form in the sea under conditions of some turbu- lence, with low sedimentation rates and some organic matter (Reineck & Singh 1980: 151). Porrenga (1967) noted that it was being formed today in the tropics at depths of 125—250m, controlled by temperature (10°-15°C). However, Bell & Goodell (1967) recorded authigenic glauconite in shelf environments at various latitudes, ranging in depth from 20-1780 fathoms (= 37-3255 m), being formed often inside foraminiferid shells or replacing coprolites or mica flakes; they found that it tends to form in a band parallel to the coastline. Murray & Wright (1974: 50-51), on the basis of foraminiferids, considered beds B-E at Barton Cliff (glauconitic beds of cycle 4) to have formed under water 50-100 m deep, and beds A2 and F-lower H (non-glauconitic clays of cycles 3 and 4) no deeper than 50m. Lower and higher strata were referred to an intertidal or estuarine environment and they concluded that ‘the Barton Beds at their type locality represent a full marine cycle from continental deposits, 431 (LI61 ‘O16T) JOxRUY AM Pur (OPT) YIIMISAIg (S61) 1940OH ‘(Zr61) AtiND :eIep Jo saoinog [Oz ‘d ‘7 “BY-1Xa]_ 9a8 ‘sjoquiAs [eoIsojoyI] 0} Aoy JOY ‘uINjep [eJUOZzIIOY do} ay} se pasn udaq sey spog UOpRa}] JIMOT 94} JO aseq 9Y} JUaSaId d1OY MA “A}I[LOO] USAID B ]L BILIIS PAP1OIAI JO JU9}X2 9}BOIPUI SMOPUIM anbI]gO JO sieq [edIVJOA ‘sofoho puv suONGLysIp SIDR] MOYS 0} “}SOM-YINOS/}se9-YyjJOU ATa}VUIXOIdde SuIpuss) ‘e}eINs URIUOJIEg UIseg dITYysdUIe PH JO UOT}DaS dI]s DI}BUIOYOS }RYMOWIOS gg oANSY-}xo 7, BARTONIAN MAMMALS OF HAMPSHIRE BASIN NOtrs, A\l a y i= ———— | NEMS ST GG jee eee ee AN Ol 0 : = ) = = Sie w —NOMLYWEOS | fe Ga ivae Gh oS: 3 === f= aes Es ove =e NOILVWuod 7 3NOLSaWI1 MO¥YVSHDIIND MS sq3@_ NOdGV3H 43aMO1 IN = = oO < z = RF i x = z= saj2A> fa) z 2 3 Sage geet he dg : i = a a $s % & 3% Ey 8 = ¥ 5 oF ea g a o RS = 4 8 z 432 J. J. HOOKER through marsh, inshore and offshore shelf, and back to shallow water conditions’. The close lithological, and to an extent faunal, similarities between cycles 3 and 4 suggest that their ‘full marine cycle’ may have been repeated. The uniqueness of their agglutinated foraminiferal fauna from Al (suggesting to them hyposaline tidal marsh) by their own admission takes no account of either the absence by leaching of calcareous fossils in this bed, or the probable absence of agglutinated forms from glauconitic beds in cycle 4 by disintegration on burial (Murray & Wright 1974: 48, 50). See also comments by Curry (1977: 404). Burton (1933) noticed that certain molluscs occurred in the Lower and Upper Barton Beds, but not in the Middle. These are in fact almost entirely restricted to the sands and clayey sands of Beds A3 and H and thus predominate at homologous stages in cycles 3 and 4. Represent- ative species are: Vepricardium porulosum, Glans oblonga, Calliostoma nodulosum, Homalaxis sp., Cordieria semicostata and Tornatellaea simulata. The very rare brachiopod Argyrotheca piperipyxis has a similar distribution. The gastropod genus Terebellum also has a similar occurrence but is represented by different species (T. fusiforme in A3; T. sopitum in upper H). Further repetitions occur between the sandy parts of cycles 4 and 5, e.g. Glans oblonga at Barton. Further east this extends also to Chama squamosa, V epricardium porulosum, Crassatella tenuisulcata and Terebellum sopitum, which occur in cycle 5 in Whitecliff Bay. This list is from Gardner et al. (1888: 605) who called the horizon at Whitecliff Bay the ‘Chama Bed’, but according to the dinocysts (Bujak et al. 1980) and the lithostratigraphy it belongs to cycle 5S. The true Chama Bed of Wright (1851) at Barton belongs to cycle 4. There are conflicting ideas on water depth in lower parts of the sequence in the east of the area. Murray & Wright (1974: 21) could find no foraminiferans in most of the weathered, poorly-exposed cliff section in Whitecliff Bay and, from the descriptions of the succession by Gardner et al. (1888) and White (1921a), concluded that clays above the N. prestwichianus bed represented ‘a return to intertidal, fluvial marsh, and channel environments’. This interpretation is not corroborated by the fully marine mollusc faunas from the ‘Chama Bed’ (Gardner et al. 1888) or in blue clays with N. rectus (personal observation). Likewise, the Elmore Member (Fisher beds XVIII-lower XIX) was considered to be intertidal (Murray & Wright 1974: 21), whereas an almost identical lithology (beds S—W) at Fawley (Curry et al. 1968: 195) was considered to have accumulated in deeper water than had the Selsey division (beds B—R) and under low oxygenation conditions with limited current movement, which were unfavourable to a rich benthos, in particular molluscs. It is probable that cycles 3 and 4 represent upward shallowing sequences from richly (more onshore shelf) or poorly (more offshore shelf) glauconitic greenish clays and silts (50-100m depth) through non-glauconitic blue or grey clays (maximum 50m depth) to (in the west only) lower shoreface sands (with Solen, Panopea, Pinna, see Curry 1977, and possibly Ophiomorpha in cycle 3). Cycle 5 represents a similar upward shallowing sequence but from non-glauconitic clays (c. 50m depth?) through shoreface sands to shelly storm washover sands (at Barton Cliff, Plint 1984). Both molluscs (Gardner et al. 1888: 592) and foraminiferans (Murray & Wright 1974) in the upper sandy part of the clays indicate the beginning of a major salinity reduction towards the brackish and freshwater environment of the overlying Lower Headon Beds which conclude cycle 5. At this same lithological horizon in Whitecliff Bay, however, the salinity appears to have been more normally marine with the ‘Chama bed’ type fauna mentioned above. At Barton Cliff archaeocetes are known from both the glauconitic and non-glauconitic clay facies and these presumably lived or at least died in this environment. The few land mammals are mainly medium- to large-sized, probably a feature of collecting availability and bias; the bat (p. 242) is an unusual exception, found admittedly by sieving techniques, this group being very rare in marine sediments and rare even at Creechbarrow. The skull of Plagiolophus curtisi is unusual in having the cranium and mandible in association, indicating a fairly nearshore site of deposition. The non-marine province. Knowledge of this province is restricted to strata on Creechbarrow Hill. The sands, clays and conglomerates beneath the Creechbarrow Limestone Formation are poorly exposed and little knowledge has been gained of sedimentary structure. Nevertheless the BARTONIAN MAMMALS OF HAMPSHIRE BASIN 433 poor sorting, the occasional conglomeratic horizons, and the unrolled state of the flints (indicating a nearby source) all suggest that they may be distal alluvial fan deposits (see Blissenbach 1954, Hooke 1967, Reineck & Singh 1980). A different and more complex depositional interpretation is required to account for the various features of the Creechbarrow Limestone. Indications of three different environments have been found: 1. Fluvial: . abundant quartz grains in an otherwise fine-grained limestone; a b. terrestrial mammal remains showing evidence of water transport (see below); 2. Lacustrine: a. freshwater gastropod molluscs; b. cyanophyte algal oncoliths often with laminae thin on one side; c. calcareous ?chironomid tubes associated with the algae; (Evidence of either fluvial or lacustrine aquatic environment is given by pantolestids, chelonians, croco- dilians, characoid fish (Gaudant 1979), unionids and quartz-grain tubes of ?Trichoptera). 3. Terrestrial: a. ?insect burrows (for aestivation, hibernation or pupation etc.); b. land snails and slugs; c. well-preserved associated teeth of terrestrial mammals showing little evidence of prede- positional transport. Examples from each of the three groups can be found together in the same hand specimen, so it is unlikely that the disturbed nature of the strata with consequent mixing of layers of different environmental composition can account for this phenomenon. Although a water body of some sort is definitely suggested, there is no evidence that this was very deep. In fact the rarity of pantolestids, crocodilians and chelonians (and small size of the last two, ?juvenile), normally abundant in the vertebrate faunas of the dominantly fluvio-lacustrine (Insole 1972) Headon and Osborne Beds, Bembridge Marls and Hamstead Beds, suggests that it may not have been deep enough to support such a community. Alternatively, their potential food supply might have been excluded by cyanophyte-produced toxins (see Shelubsky 1951). i. NOTES ON ASPECTS OF THE BIOTA. Cyanophyta: Recent fluvial and lacustrine cryptalgal struc- tures have identical morphology, according to Jones & Wilkinson (1978), but if only infrequent- ly agitated by waves or currents as in a lake, their banding is thinner on the side of the oncolith in contact with the substrate. Some of the Creechbarrow oncoliths show this form (Hudleston 1903: pl. 11, figs 2-4). In temperate climates (Lake Michigan) Jones & Wilkinson noted a seasonal growth pattern, with most activity in the summer, causing annual layers. In sub- tropical climates (Florida Everglades) a more complex pattern is produced by short wet and dry periods superimposed on overall wet summer and dry winter seasonality (Monty & Hardie 1976). The Creechbarrow cryptalgal structures are laminated irregularly and in places show scattered branched and unbranched radial filaments (PI. 35, fig. 1c). In gross structure they are nearly all oncoliths ranging from a few millimetres to a few centimetres in diameter. Most commonly they encrust the fresh-water snail shells (Preece 1980a: pl. 4, figs g, h), but they also occur without an obvious nucleus or as apparently hollow cylinders, now filled with limestone matrix and perhaps once formed round twigs, stems or roots, growing or loose (PI. 35, figs 1—2). One in particular is a straight symmetrical cylinder with faint longitudinal striations on the inside wall and with the encrustation recrystallized to radiating fibrous calcite crystals which give the outer surface a knobbly appearance (PI. 35, fig. 2). Only one fragment of cushion- shaped stromatolite has been found. When complete it would have been no more than about 10cm across and it appears to have had more than one gastropod nucleus. Daley (1974) described well-preserved material of two new genera of encrusting cyanophytes from the brack- ish Bembridge Marls of the Isle of Wight. ?Trichoptera larval cases (PI. 35, fig. 3): Both specimens are fragments of curved wall com- posed of a mosaic of subangular quartz grains 0-2-0-8 mm in diameter. They had formed nuclei for algal growth and were only discovered by random breakage in processing (or subaerial weathering) of oncoliths. The curvature suggests an original diameter of c. 4mm. The larger of the two fragments has an incomplete minimum length of 6mm. Polychaetes (e.g. Sabellaria) are J. J. HOOKER 434 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 435 known to construct similar tubes but have generally not become established in fresh-water environments (Chamberlain 1975). Vialov & Sukatsheva (1976) have reviewed the geological record of caddisfly larval cases. Organic-walled types are known from the Palaeogene of western Europe in the Middle Eocene of Geiseltal (D.D.R.) (Haupt 1956) and the Palaeocene of the Isle of Mull (Zeuner 1941). Caddis larvae that construct sandy tubes are more typical of lotic (e.g. fluvial) environments, although some occur in more lentic areas (e.g. lakes) (Chamberlain 1975). Whereas there is insufficient evidence to assign these fragments to Tri- choptera with certainty, there is nothing about the specimens to prevent such an attribution (personal communication, P. Barnard 1982). ?Chironomidae larval tubes (Pl. 35, fig. 1b): Randomly orientated, calcite-lined, sparite- infilled tubes occur within oncoliths. Similar structures in marine Triassic stromatolites have been described as the problematicum genus Microtubus Flugel 1964, a probable polychaete (see also Wright & Mayall 1981). Modern associations of chironomid larvae and cyanophytes led Anadon & Zamarreno (1981) to recognize as chironomids similar tubes in non-marine Palaeo- gene oncoliths from Spain. In a modern fresh-water environment with cyanophytes, chirono- mids are the most likely producers of such tubes, but there is limited morphological support for this attribution (personal communication, P. Cranston 1982). ?Insect burrows (PI. 35, fig. 4): These also occur in the Bembridge Limestone, where they were originally described as eggs of either turtles or the land snail ‘Bulimus’ ellipticus (Edwards 1852). White (1921a: 117) called them ovoid concretions. Keeping (1910) was the first to record them from Creechbarrow. Preece (1980a: 177) called them oviform bodies and compared their spiral grooves to ‘the mandibular pattern present in wasp nests’. Their calcite composition appears to constitute a geodal filling of a cavity. Ratcliffe & Fagerstrom (1980) noted that cydnid hemipterans make cell-shaped burrows without access shafts and several other insect groups make vertical or oblique burrows with ovoid terminal cells (trace fossil Amphorichnus Myannil 1966). None of the Creechbarrow or Bembridge Limestone bodies appear to have had an access shaft and their orientation in the rock is not recorded as they are normally found weathered out. Their presence nevertheless appears to suggest subaerial conditions. Tubular burrow: A single slightly flexed tubular limestone burrow fill with no surface struc- ture has been found. It is 45mm long with a diameter of 10-12 mm. Fresh-water molluscs: Daley (1972a: 27), reviewing earlier studies, stated that Viviparus and Melanopsis (probably a modern analogue of Coptostylus) live today in rivers, ponds, lakes and canals, grazing bare, soft bottom muds. He noted that whereas Melanopsis appears to prefer areas free from aquatic vegetation, Viviparus will live in marshes; and that whereas Viviparus is sensitive to salinity increases above 3%o, Melanopsis is tolerant of brackish and fresh water. The occurrence of Viviparus at Creechbarrow is an important indication of fresh water, although the abundance of Coptostylus brevis suggests near optimum conditions for this species. Preece Plate 35 Light macrographs of miscellaneous Creechbarrow biota. Figs 1—3, 4b are not sprayed with ammonium chloride. Fig. la—c Transverse section of cyanophyte oncolith (V.61740); a, complete view of polished face, showing concentric laminae and thinning on one side, x1; b, details of laminae and associated ?Chironomidae tubes, x 4; c, filaments (one branched), x 500. Fig. 2a, b Regularly cylindrical oncolith recrystallized to radiating acicular crystals but with a few concentric laminae remaining (V.61741); a, side view, x 1-5; b, polished face, x 3. Fig. 3 Fragment of oncolith coating part of ?Trichoptera larval case, composed of sand grains, (In.64609), viewed from inside, x 6. Fig. 4a,b ?Insect burrow (GG21472); a, side view, x 2; b, broken end showing geodal calcite infilling, x 4. Fig.5a—c Tooth of characoid fish (P61171), x 7;.a, buccal view; b, occlusal view; c, lingual view. Fig.6 Crocodilian tooth (R10006), x 5-5. Fig.7 Bone fragment of indeterminate ?mammal with gnaw marks (M37572), x 6°5. Fig.8 Fragment of mammalian scapula with gnaw marks (M37573); medial view with glenoid to left, x 3. 436 J. J. HOOKER (1980a: 178) suggested poorly vegetated or bare lime mud surfaces to account for the abundant assemblage of Coptostylus, the absence of Planorbidae and near absence of Lymnaeidae. Absence of ostracods and charophytes: This may be diagenetic rather than original. Re- crystallization in the limestone has destroyed the fine detail on the larger shells and it is likely that small calcareous organisms would have been obliterated completely. See also comments by Preece (1980a: 178). ii. PROBLEM OF DISTINCTION OF STROMATOLITES FROM CALCRETES. This subject has been discussed by Read (1976). Insole (1972) described various limestones in the Osborne Beds and Bembridge Limestone, which he considered to represent calcretes. In the case of the limestone in the Osborne Beds at Headon Hill, a sequence of marls passing up through nodular limestone is superficially similar to that at Creechbarrow. The Creechbarrow sequence differs in the presence of oncoliths rather than nodules. The limestones themselves also differ. The limestone in the Osborne Beds is vertically fissured and brecciated and the cavities have surface rinds, which is consistent with the structure of calcretes (see Hay & Reeder 1978). Goudie (1973: 23) noted that an important component of calcrete is silica (average 12:3%), including both quartz grains and amorphous silica cement. The Creechbarrow Limestone, in its lack of brecciation, fissuring and amorphous silica cement (the latter was not found in insoluble residues — see methods section, p. 194), is more consistent with an algal than a calcrete origin for the oncoliths. It cannot be excluded, however, that above the modern erosion plane at Creech- barrow, higher parts of the succession now missing may have shown evidence of pedogenesis. This can be associated with terminal regressive phases of lacustrine sequences (e.g. Engesser et al. 1981, Goudie 1973) and could account for the recrystallization and poor or non- preservation of small calcareous fossils in the Creechbarrow Limestone. ili. CONCLUSIONS. Preece (1980a) discussed the contrasting views of Hudleston (1902a), who thought the Creechbarrow Limestone represented deposition in a lake, and Bury (1934), who considered it a swamp, and partially favoured Bury because of the rarity of land shells in modern lake deposits. Bury was comparing the Limestone with the nearby Holocene tufa of Blashenwell. Preece (1980b) concluded that the Blashenwell tufa was deposited on marsh- ground with some water flow and boggy pools. In his list of recorded gastropods, nearly all are terrestrial and none are totally aquatic, a situation different from Creechbarrow. Moreover, his suggestion that ‘the nodular or pisolitic lithology and even encrustation of gastropod shells is suggestive of flowing water or seepages’ is not corroborated by Jones & Wilkinson (1978), who demonstrated that pisolitic growth can take place, admittedly at a slower rate, on the underside of the body in contact with the sediment. Several of the features of the Creechbarrow Limestone are consistent with environments found in the ‘interior marshes’ of the Florida Everglades (see Monty & Hardie 1976). Modern environments in this area have several times been invoked as possible analogues to various late Eocene/early Oligocene deposits of southern England (e.g. Machin 1971, Keen 1977, Collinson 1983). Salient points drawn from Monty & Hardie (1976) which fit the data from Creechbarrow are: the total fresh-water nature of the area; rainy summers and dry winters with superimposed smaller scale wet and dry periods favouring growth of cyanophytes but discouraging aquatic higher plants; cyanophytes able to grow in depths from a few cm to a few tens of cm; environment very quiet. The main features which do not fit the data from Creechbarrow are: the more continuous nature of the algal mat; the absence of sandy detrital deposition; and the calcareous nature of the bedrock substrate. At Creechbarrow the calcium carbonate is likely to have been derived from erosion of nearby Chalk, as the unworn flints and derived Bryozoa (p. 420) testify; the more restricted growth of algae and the detrital elements suggests that the environment was only intermittently quiet. : A Palaeogene non-marine to paralic complex in the Ebro Basin, Spain, with one facies similar to that at Creechbarrow, has been described by Anadon & Zamarrefio (1981). This was interpreted as a ‘palustrine setting in relation to alluvial fans’. The facies consisted of calcareous mudstones and argillaceous or sandy limestones; the algal bodies were mainly small asym- BARTONIAN MAMMALS OF HAMPSHIRE BASIN 437 metric oncoliths (up to 3cm in diameter and 10cm in length) with chironomid tubes. This was interpreted as representing marshes and shallow lakes, strongly influenced by sheet floods from the alluvial fans. Thus it is possible to envisage for Creechbarrow a shallow fresh-water lake to marsh environment, moderately favourable to the growth of oncolitic cyanophytes and to the exis- tence of mud-browsing gastropods and unionid bivalves. This would be subject to periodicity: a, flooding, which introduced clastic detritus and land-derived vertebrates as thanatocoenoses; and b, lowering of the water level, which inhibited algal growth and provided emergent margin- al areas which were burrowed by insects and colonized by land gastropods and presumably some land plants. The environment of deposition between Creechbarrow and Hengistbury will remain unknown as modern erosion has penetrated to much lower strata over the whole area. However, the evidence of progradation of coastal sand in the upper part of the cycle 3, and brackish elements in the otherwise typically marine A3 fauna in Christchurch Bay, e.g. Theo- doxus spp., Paludestrina sp., Melanopsis cf. subfusiformis, Batillaria cf. calcitrapoides and Pot- amides cf. variabilis (see Burton 1933), suggests that there may have been an intervening hyposaline lagoon. Tectonic hypothesis. Alluvial fans are known to develop from abrupt topographical features, often fault scarps (Reineck & Singh 1980: 298). If the poorly sorted sands, clays and conglomer- ates beneath the Creechbarrow Limestone are correctly interpreted as alluvial fan deposits (see above), then we might expect some such feature to have existed in Bartonian times, roughly to the west of Creechbarrow. Faulting is present in east Dorset south of Creechbarrow in the form of the Ballard Down Fault which trends east-west through the Isle of Purbeck. Small (1980: 52, fig. 1) intepreted this as having been active before the deposition of the Creechbarrow Limestone, on the grounds that ‘Creechbarrow type’ flints occur in solution hollows in the Chalk south of Creechbarrow. He thus ignored Arkell’s (1947: 241) caveat that the flints could have been redeposited long after initial deposition in the Eocene. East—west en echelon folds affecting the Hampshire Basin Palaeogene deposits have been attributed by Shearman (1968) to NW-SE dextral wrench faulting in the Palaeozoic basement, following pre-existing joint directions, which can be seen at the surface today in Devon and Cornwall. For details of these structures in SW England see Dearman (1964) and Palmer (1975), and references therein. Shearman considered that the movements in the basement were transmitted upwards into the Mesozoic and Cainozoic cover in the form of these folds accom- panied by a décollement. He repeated an experiment by Pavoni (1961) in which wet tissue was placed over two boards. When the boards were slid past each other dextrally in a SE direction, en echelon, sigmoidal folds trending east-west were produced in the tissue across the line of displacement. No northerly inclined asymmetry, as occurs in the Hampshire Basin, was produc- ed, but Shearman accounted for this by décollement, which did not occur in the experiment. All these authors accepted a Neogene date for the faulting and folding, i.e. long after deposi- tion of the Creechbarrow Limestone. Important lines of evidence are: 1, the youngest beds preserved in the Palaeogene sequence (Oligocene Hamstead Beds) are affected by the east—west striking folds; 2, the Oligocene deposits of the Bovey Basin are affected by NW-SE dextral wrench faulting; and 3, penecontemporaneous Palaeogene folding and facies lineation in the Hampshire and Paris Basins have a NW-SE trend (see Fig. 69; Murray & Wright 1974: figs 25E-F, 28; Daley & Edwards 1971; Pomerol 1967). This argues against the model of Small (1980) and suggests that a NW-SE trending feature in the area of central Dorset may have propagated the Creechbarrow fans. No superficial faults of this trend have been detected in this area but a feature of similar orientation occurred a little further to the east in late Albian and Cenomanian times. This is known as the Mid-Dorset Swell. It caused a drastic thinning of deposits and hardgrounds in its vicinity and its existence has been attributed to movements in Palaeozoic basement structures (Drummond 1970). Plint (1982) has provided sedimentological evidence in support of Eocene activity along the east-west fault and fold axis of Purbeck and 438 J. J. HOOKER the Isle of Wight. His evidence, however, could equally be used to support a NW-SE fold structure, which would better fit the other evidence. Moreover, his palaeocurrent direction for M. Eocene meandering rivers is to the north-east (Plint 1983: fig. 12). If NW-SE normal faulting with downthrow to the NE had occurred in the Palaeozoic basement in later Eocene times in central Dorset, slightly SW of the Mid-Dorset Swell, a NE-dipping monocline would have been produced in the Mesozoic and Tertiary cover. The monocline would have been obliterated by the much severer folding and faulting of the Neogene. Such a tectonic model could not only account for the Creechbarrow alluvial fans but also the successive rapid marine transgressions by intermittent activation of the monoclinal structure causing subsidence on the down-dip side. The clastics would thus have been available to fill much of the marine as well as the alluvial part of the basin between each transgression. The glauconitic nearshore facies is more restricted in the Barton Clay Formation than in the Bracklesham Group. Deeper water clays are reached more rapidly in an offshore direction and facies change takes place over a shorter distance (e.g. interdigitation of J with K over a distance of c. 1km at Barton). Moreover, the braiding of the rivers presumed to have formed the alluvial fans contrasts with the meandering rivers considered responsible for the earlier fluvial western facies coeval with the marine Bracklesham Group (Plint 1983). All this suggests greater topo- graphic relief, a more rapidly prograding shoreline with a steeper palaeoslope, and stronger current action during early stages of the cycles. See Text-fig. 69 for Hampshire Basin Bartonian facies map. If Shearman’s experiment is repeated with the NE plank at a slightly lower level than the other, simulating primary normal faulting, the folding produced by secondary dextral wrench faulting is overturned to the north on the downthrow side. This simulates the asymmetry of the Purbeck monocline without the necessity of a décollement. Palaeoecology of the Creechbarrow mammal fauna. Introduction. The restricted total outcrop, poor exposure and disturbed state of the Creech- barrow Limestone Formation make most aspects of taphonomic and palaeoecological analysis difficult or impossible. On the positive side, unlike most other English Palaeogene deposits, no surface collecting could be done and all but three specimens, found during excavation work, were recovered by systematic sieving techniques. Of the matrix collected, about 1600kg has been completely processed (with the exception of the larger limestone fragments), the remainder being initially searched for only larger mammalian specimens. The proportional abundances of the mammal specimens recorded from this 1600 kg sample can therefore be viewed with some confidence. Immediate areas of bias are likely to be against: 1, those mammals without teeth like the pangolins (Pholidota) and anteaters (Xenarthra), both of which are known from the Lutetian of Messel, W. Germany (Storch 1978, 1981), but unknown at Creechbarrow (see p. 441); and 2, the larger herbivores and carnivores because of the restricted outcrop and poor exposure (see p. 207). Less accountable problems are the impossibility of knowing how accu- rately the fossil sample reflects the composition of the living assemblage. For a review of these types of problems and bias in assessing modern mammalian populations, see Western (1980). Preservation. The mammal remains consist predominantly of teeth with relatively few bony skeletal parts being found. The teeth range in appearance from fresh, with mahogany-coloured enamel and slightly darker dentine, to worn and battered with an overall dark brown colour. Some apparently unabraded teeth are almost white; these seem to belong to the first preser- vational category, but recent weathering has leached them. Manganese is frequently found coating the teeth and is probably also responsible for the dark staining. In preservation bones range from unworn but weathered and white, to worn and of a dark brown colour. Some bones can be worn and light-coloured and have probably undergone leaching like the light-coloured teeth. No preservation is restricted to any particular taxon unless it happens to be very rare or unique. It is considered that, whereas fluvial transport probably brought most of the specimens into the depositional environment, the more worn and dark-coloured specimens have been carried 439 BARTONIAN MAMMALS OF HAMPSHIRE BASIN (4 PTZ61) MUA PUL (LT6T O16) J9ABUYM “(4 “PZOG6T “868T) POM (SL61) J2YOOH (Zp6T) AlIND : B1ep Jo sao1nog ‘(Aeg younyosiyD wor Aeme sjuajeatnba [eoisojoyy] pue) | peq pueg uojdag Jo syrum] A[Ja}sea paye[njsod = 9g ‘¢ 0} | sa[oKo Jo spuvs aarssasBaI JO WU] Wso}sva-YIIOU payepnjsod = ¢ ‘(Keg YoINYyo}slIyD wo AeMe syua|eAInbo jeorBojoy ly pur) ¢ pag Ae[D uojzeg Jo s}IWT ysoM-YyJJOU pue }soM-YINOS (peysep) poyejodenxa pue (snonuNUOS) UMOUY = p :Spaq dITUOONRIS Jo jUNT] ISoM-YINOS a[qeqoid = ¢ ‘ouT[aIOYs UNUIXeUL Jo UONISOd payeinjsod = 7 :auoouow surddip jsea-yjiou jeonayjodAy = | “AydesZoasoarjed pur sory uvluojieg ulseg ouysduiey Jo dey 69 onsy-}x9], ut ae oe OL (e) 440 J. J. HOOKER some distance, whilst the fresher specimens probably died much closer to the depositional site. A scatological origin is thought to be unlikely or at least an unimportant factor for two reasons. Firstly there is a complete gradation between the end members of the two preser- vational types, suggesting continuously varying distances of origin. Secondly the worn appear- ance of the tooth crowns coincides in most cases with loss of roots, indicating fluvial transport (see Korth 1979), and differing from the enamel-less but rooted teeth considered by Fisher (1981a, b) to have been excreted by crocodilians. This accords with the rarity of crocodilians at Creechbarrow. Moreover, the preservation lacks the characteristic corrosion typical of certain avian predators (Mayhew 1977). Some of the better-preserved teeth have been found to be associated, according to their matching interstitial facets as well as their similar natural wear and preservation (e.g. Lophio- therium siderolithicum M36177 and M37705, p. 351; and Haplobunodon venatorum M37716 and M37529, p. 405). M!~3 of Dacrytherium elegans (M37177, p. 410) were found in a row, but slightly separated from one another and without the jaw, in a block of limestone. The main specimen of Plagiolophus curtisi creechensis (M36181, p. 364) consists of associated upper and lower jaws on both sides as well as numerous unreconstructable jaw fragments. It seems a remarkable coincidence that one of only two specimens of the same species from the marine Barton Clay should also have both upper and lower jaws associated. Single teeth in jaws have occurred in two cases: Haplobunodon venatorum (M37544, p. 408) found in the field; and Gesneropithex figularis (M35407, p. 234) found separated in residues, but where the broken root surfaces on tooth and jaw were found to match up. The few other jaws found were all edentu- lous and this condition may be due largely to recent weathering of the deposit. Three isolated teeth have been found in limestone blocks. They belong to Plesiarctomys curranti (M35450, p. 285; Pl. 13, fig. 7), Lophiotherium siderolithicum (M37446, p. 351) and Acotherulum campichii (M36802, p. 400). All are dark-coloured, battered and rootless. Some of the bones (both fresh and worn) show evidence of rodent gnawing (PI. 35, figs 7-8). In some: cases the marks are transverse to the long axis and occur as parallel grooves; in another they occur round a bite notch (see Bonnichsen 1979 and Morlan 1980 for modern examples). Those on a scapula fragment (PI. 35, fig. 8) are of appropriate size for the incisors attributed herein tentatively to Suevosciurus authodon. Those figured in Pl. 35, fig. 7, however, appear to have been made by much smaller teeth. If these were of a rodent, it must have been smaller than any recorded from teeth at this locality. Faunal Composition. Text-fig. 70A—B shows relative abundances based on the ordinal composi- tion of species and individuals respectively. They show that whereas the artiodactyls display the greatest diversity of species, the rodents make up over 50% of the individuals (more than a third belonging to Suevosciurus authodon). The apatotheres, whilst showing some diversity, are few in number of individuals. They were also found to form only c. 1% of North American Eocene faunas (West 1973). Creechbarrow seems notable for having representatives of groups normally rare in European Eocene deposits, i.e. one paroxyclaenid, three apatotheres, two Plesiarctomys and a third undet. manitshine, and Ailuravus. This may be more a feature of the bias against smaller mammals in old collections from the main Lutetian and Bartonian local- ities and the increasing rarity of these groups from late Bartonian times onwards. The latter feature is particularly relevant to the locality of Robiac. Text-fig. 70C shows log. of maximum numbers of specimens per species for the 1600kg sample plotted against log. of size based on M!? area (or the best approximation to it when this tooth was not available). Creighton (1980) showed that area (length x width) of M, correlated accurately with body size in modern mammals (except edentates). M' is preferred here because length and width are usually approximately equal and can thus be more easily extrapolated, when only lower teeth are available. In certain mammals, e.g. rodents, M' and M? are essen- tially the same size and their mixing for statistical purposes still produces a low coefficient of variation. In Text-fig. 70C, species considered on their dental morphology to be at least mainly fru- givorous or herbivorous plot in broad linear fashion, becoming less abundant with increasing Marsupicarnivora (2) elgg ?Proteutheria (1) Ze | i W ~7-— Lipotyphla (2) / | /' é /~_-— Chiroptera (2+) Artiodactyla (11) ——~ | | / Primates (8) Perissodactyla (3) ——_\ ; Ve SS \ ie / | << / Condylarthra (1) ew / | A Carnivora (2) L [ x i Apatotheria (3) —— ees Rodentia (7) Cebochoerus robiacensis ?Mixtotherium sp. indet./———~ OLD), ie] Met oth ect “- s Wi — ?Choeropotamus sp. indet. xtotherium aff. gresslyi Dacrytherium elegans Dichodon sp. indet. Amphiperatherium aff. goethei Amphiperatherium fontense ?Pantolestidae indet. ?Hyperdichobune sp. Y Hyperdichobune sp. 1 = Plagiolophus curtisi — Lophiotherium siderolithicum Propalaeotherium aff. parvulu Vulpavoides cooperi ?Miacinae indet. > 4 Gesneropithex figularis \ Scraeva sp. indet. Microchiroptera spp. indet. 1 - 4 ~— Nannopithex quaylei Nannopithex sp. 1 Pseudoloris cf. crusafonti Microchoerus wardi ?Miacis sp. indet. Heterohyus morinionensis Heterohyus aff. nanus’ | Heterohyus cf. sudrei/ Microchoerus creechbarrowensis Europolemur collinsonae Leptadapis aff. magnus Plesiarctomys hurzeleri Plesiarctomys curranti ?Manitshinae indet. Ailuravus stehlinschaubi Rodentia DNS Iaroidee rissonei Treposciurus helveticus 52 36 33 24 (e) fe) 49 f) 2:0 Gu 34 46 16 43 a oy He 4 OO 47 oe Haas gk? og ot! J 041 30 4 4 18 Log. size | © (2) o'4 <) 4 @10 1.0} COI 313 oF | 20 | ca CS eee Ve ®i2g 4 e 54g! 1 83, 5 e 24 C U .3 e 2 Ly ha ae La a ema Lad ed ga om) ae Gr ee Gee oe (0) 5 1.0 5 2.0 5 3.0 Log. N Text-figure 70 Diagrams of relative abundances of Creechbarrow mammals. A. Circular diagram showing proportional representation of species numbers per order. B. Circular diagram showing proportional representation of specimen numbers per species, grouped by order. C. Scatter diagram of log. size (based on M' area, or its extrapolation by squaring M, length) against log. numbers of individuals (based on maximum numbers of teeth). Bar on left gives range of size of modern arboreal herbivores from Eisenberg (1978) extrapolated to tooth size using Creighton (1980). Outline circles are probable herbivores; solid circles are probable insectivores/carnivores; numbers against symbols refer to taxa listed in Table 1, pp. 212-215. 442 J. J. HOOKER size. Those species considered on dental morphology to be mainly insectivorous or carnivorous are rarer in occurrence than the equivalent-sized herbivores. There is some overlap, however, the herbivorous ?Mixtotherium sp. indet. and cf. Manitshinae indet. being exceptionally rare and the carnivore ?Miacis sp. indet. and possible carnivore Amphiperatherium fontense being relatively common. The correlation of increasing rarity with increasing size in the herbivores may reflect gener- ally more scattered populations of animals with greater biomass, or it might reflect their lower birth and death rates (see Western 1980). This ignores possible gregarious versus non- gregarious habits, simple rarity or abundance of particular species, range of habitats sampled and many other unknowns. However, as a general rule, insectivore-carnivores tend to be rarer than frugivore-herbivores for a given size (Wolff 1975). Moreover, the consistency in the range of good to poor preservation for all the taxa suggests that even if areas at different distances from the depositional site have been sampled, the same fauna occurred throughout. The correlation of herbivore size and abundance plus the presence of the rhinoceros-sized herbivore Lophiodon cf. lautricense in nearly contemporaneous marine deposits at Hengistbury suggest that larger herbivores and probably carnivores may once have lived in association with the Creechbarrow fauna as known at present, but are missing on account of inadequate sampling. The range of larger perissodactyls and artiodactyls as well as the larger creodonts and carnivores present in such Bartonian localities as Eclepens-Gare and Robiac also supports this. If the linear plot of Text-fig. 70C is continued to the log. M' area of L. lautricense (3-34 calculated from Sudre’s 1971 measurements), it can be calculated that somewhere in the region of 30 tonnes of Creechbarrow Limestone would have to be processed in order to find even a single individual of this species (if present). Korth (1979) determined a sequence of fluvial deposition of different mammalian skeletal elements from settling and flume experiments. Particular elements grouped into five categories. Generally poor preservation of bone at Creechbarrow prevents a direct comparison with his categories, because of the likelihood of bias against those which were small and fragile and those that were large and little transported (i.e. less mineralized and liable to shattering to an unrecognizeable condition). Nevertheless a large range of different elements of different sizes and different preservation states are present. Complete or fragmentary rib (1), radii (2), ulna (1), calcanea (2), femur (1), mandibles (10), tibia (1), metapodials (6), podials other than calcaneum and astragalus (3), phalanges (5) and sesamoids (3), as well as the many teeth, were recovered from the 1600 kg sample. In addition, fragments of a vertebra, pelvis, scapula and astragali have been recovered from other samples. Notes on diet, habits and habitat. Table 39 lists the Creechbarrow mammal taxa with, for each, a suggested nearest living relative (column 1), modern taxon with similar teeth (if available) (column 2) and where possible a relevant nearest relative from the early Lutetian of Messel, B.R.D., where articulated skeletons often with soft parts and gut contents are known (column 3). In many cases the first column will be irrelevant, particularly in the case of totally extinct family or higher taxa. If, however, this is the same as, or closely related to, the taxon in the second column, the significance of the latter is reinforced. Column 3 is important either as a back-up or in the absence of a column 2 taxon. Comments on the possible ecology of the Creechbarrow mammals are given below, incorporating the Table 39 data. DIDELPHIDAE. Crochet (1980: 230) noted that an arboreal habit predominates in modern didelphids, but some are ground dwellers and one a specialized aquatic. He also noted that their diet is very varied from insectivorous, carnivorous to granivorous and frugivorous, and that whereas there are different dietary specializations between different species, the overlap is great. He suggested (1980: 231-232) that the structure of the humerus in the extinct Pera- therium might indicate less arboreality than modern forms; but preferred the idea that it reflected some unknown specialization of the European Tertiary forms. He also suggested that the teeth of the genus Amphiperatherium showed a tendency towards carnivory. Table 39 Nearest living relatives and modern dental analogues of the Creechbarrow mammals. The third column gives the relevant nearest relative in the Lutetian fauna at Messel, B.R.D. Creechbarrow taxon Amphiperatherium aff. goethei A. fontense ?Pantolestidae indet. Gesneropithex figularis Scraeva sp. indet. Microchiroptera indet. 14 Nannopithex quaylei N. sp. 1 Pseudoloris cf. crusafonti Microchoerus wardi M. creechbarrowensis Europolemur collinsonae Leptadapis aff. magnus Adapinae indet. Plesiarctomys curranti P. hurzeleri ?Manitshinae indet. Ailuravus stehlinschaubi Sciuroides rissonei Treposciurus helveticus preecei Suevosciurus authodon Heterohyus cf. sudrei H. aff. nanus H. morinionensis ?Miacis sp. indet. ?Miacinae indet. Vulpavoides cooperi Propalaeotherium aff. parvulum Lophiotherium siderolithicum Plagiolophus curtisi creechensis Hyperdichobune sp. 1 2H. sp. 2 Mixtotherium aff. gresslyi 2M. sp. indet. Cebochoerus robiacensis Acotherulum campichii Haplobunodon venatorum ?Choeropotamus sp. indet. Dacrytherium elegans Dichodon cf. cervinus D. sp. indet. Nearest living relative group Didelphini Didelphini 9 Erinaceidae? Erinaceidae? Microchiroptera Tarsius Tarsius Tarsius Tarsius Tarsius Lemuroidea Lemuroidea Lemuroidea Rodentia Rodentia Rodentia Rodentia Rodentia Rodentia Rodentia Ferungulata? Ferungulata? Ferungulata? Carnivora Carnivora Ungulata Equus Equus Equus Artiodactyla Artiodactyla Artiodactyla Artiodactyla Artiodactyla Artiodactyla Hippopotamidae? Hippopotamidae? Artiodactyla Camelidae? Camelidae? Modern analogue with some similar teeth Marmosa Marmosa Neotetracus Ptilocercus most insectivorous microchiropteran families Microcebus Microcebus Tarsius Alouatta Hapalemur Ratufa Ratufa Petaurista, Trogopterus Tomys Tomys Iomys Dactylopsila, Daubentonia Dactylopsila, Daubentonia Dactylopsila, Daubentonia see p. 446 Dendrohyrax Indri, Propithecus Cercocebus albigena Cercopithecus pogonias Cercopithecus pogonias Indri, Propithecus see p. 447 see p. 447 Nearest Messel relative Buxolestes piscator Pholidocercus hassiacus Microchiroptera Adapidae indet. A. macrurus Masillamys krugi Masillamys krugi Masillamys krugi Paroodectes feisti Paroodectes feisti Kopidodon macrognathus P. messelense P. messelense P. hassiacum Messelobunodon schaeferi Masillabune martini 443 444 J. J. HOOKER PANTOLESTIDAE? A semiaquatic mode of life for this family was originally suggested by Matthew (1909) when describing the type genus. This was confirmed by a skeleton of Buxol- estes from Messel, which has a tail with swimming adaptations and gut contents including fish (Koenigswald 1980a). The Creechbarrow specimens are only tentatively referred to the family but may also have been semiaquatic. LipoTYPHLA. The cheek teeth of Gesneropithex figularis are similar in cusp pattern to those of hedgehogs, but differ in the following important ways: metastylar wing is reduced along with the degree of buccal phase shear; protocone and hypocone crests are better developed along with greater degree of lingual phase wear; cristid obliqua is higher for lingual phase wear; and paracristid is reduced along with general buccal phase shear. The only extant erinaceid where there is any sign of reduction of the metastylar wing and paracristid on the molars is the galericine Neotetracus, which has fewer insect and more plant items in the diet than the others (Walker 1975). A proposed combination of insectivory and frugivory for G. figularis from modern dental analogy is supported by the record of plant and insect remains as stomach contents of the amphilemurid Pholidocercus from Messel (Koenigswald & Storch 1983: 492). Study of several complete skeletons from this locality led Koenigswald & Storch (1983: 492-493) to conclude that Pholidocercus was mainly a waterside ground dweller capable of shallow digging in a moist substrate. Gesneropithex is tentatively attributed similar habits, although no postcranial ele- ments are yet known. Scraeva and other nyctitheres are generally considered insectivorous, but it is difficult to find modern analogues with similar cuspate teeth that are not either nyctalodont or myotodont (Menu & Sigé 1971). The mainly insectivorous (Walker 1975) tupaiid Ptilocercus has both anterior teeth and cheek teeth with somewhat similar cusp patterns; but incisor denticulations are missing, intermediate conules are missing on the upper molars and the lower molars are broader, less high-cusped and the cristid obliqua is less oblique. Scraeva then appears more distinctly specialized for insectivory than Ptilocercus. MICROCHIROPTERA. Nothing more can be said than that the few teeth resemble those of insec- tivorous modern Microchiroptera and presumably had a similar diet. That there were Eocene insectivorous bats has been shown by Smith et al. (1979) for Messel specimens with gut contents. PRIMATES. The family Tarsiidae, being monotypic and relict today, cannot necessarily be expected to provide information on all the diverse members of the tarsiiform Omomyidae (see Szalay 1976). Of the Creechbarrow microchoerines, only Pseudoloris is really similar dentally to Tarsius (Teilhard 1921; Szalay 1976). It probably had a similar insectivorous/carnivorous diet. The species of Nannopithex have a modern dental analogue in Microcebus, which is mainly insectivorous (Walker 1975). N. sp. 1 (p. 251), however, unlike Microcebus has wrinkled enamel and may show a slight trend towards herbivory. Microchoerus was suggested by Szalay & Delson (1979: 265) to be ‘primarily folivorous’ because of the predominance of sharp cutting edges on its cheek teeth, maintained throughout wear, although they did find it difficult to decide whether the herbivorous emphasis was on fruit or leaves. The late advanced species M. edwardsi exhibits much cusp duplication and accessory crests which would fit an adaptation to folivory (cf. folivorous rodents discussed below). Except for the presence of a molar mesostyle, however, early species, as at Creechbarrow, are very similar to Necrolemur, which Szalay & Delson (1979: .262) considered ‘highly frugivorous’. Postcranial evidence of incipient typical modern tarsier locomotory specializations is scanty but widespread in the Omomyidae, includ- ing the Microchoerinae (Szalay & Delson 1979; Schmid 1979). The five Creechbarrow micro- choerines are therefore considered arboreal clinger-leapers with the variety of dietary specializations outlined above. Szalay & Delson (1979: 125) suggested a frugivorous-insectivorous diet for Europolemur (then monotypic). E. collinsonae, however, shows specializations paralleling those of Caenopithecus in buccal flexing of the molar centrocrista and premolar simplification. Szalay & Delson (1979: BARTONIAN MAMMALS OF HAMPSHIRE BASIN 445 143) suggested Caenopithecus was a folivore, comparing it with modern indriids. E. collinsonae, in its slightly less Caenopithecus-like specializations, was probably therefore frugivorous- folivorous. Leptadapis magnus was considered by Szalay & Delson (1979: 139) and Gingerich & Martin (1981: 254) to be probably folivorous. Its tooth pattern is somewhat like the modern folivore Hapalemur but the upper molars are relatively longer and the hypocones larger, whilst the lower molars have lower metaconid—metastylid and higher entoconid cusps (see Seligsohn & Szalay 1978). Fogden (1974) found that modern Tarsius bancanus inhabited mainly secondary forest with minor occurrence in the understory of primary forest, avoiding forest canopy, very dense vegetation and also rather open secondary vegetation with dense herbaceous cover. Whether or not extinct tarsiiforms had broader tolerances is not known. Koenigswald (1979) has demon- strated the presence of a typical lemuriform hind foot complete with groom claw in a Messel adapid. According to the evaluations of Decker & Szalay (1974) and Gingerich & Martin (1981: 255), it seems reasonable to consider both Europolemur collinsonae and Leptadapis aff. magnus as arboreal. RODENTIA. The dental diversity of the Lutetian—Bartonian rodent radiation is paralleled and exceeded by living members of the Sciuridae. For instance, great similarities are shown by manitshine paramyids to Ratufa; ailuravine paramyids to Petaurista and Trogopterus; pseudo- sciurids to Jomys; and even glirids to Exilisciurus. Wood (1962) considered that the manitshine genera Manitsha and Ischyrotomus were ground-dwelling and possibly subfossorial, but could not be precise on their individual special- izations. Unlike these genera, Plesiarctomys is unknown beyond the skull and dentition, so questions of arboreality or ground dwelling are impossible to answer. Its cranial and dental similarities to other manitshines make the latter more likely. P. curranti has broadly basined cheek teeth with some enamel wrinkling and nearly transverse buccal and lingual phase masti- cation. It is thus very similar to Ratufa which is mainly a frugivore. P. hurzeleri has larger blunt-crested cheek teeth, but both species may have been frugivorous. Arboreal adaptations have been suggested for Messel Ailuravus macrurus, including prehen- sility of the very long tail (40 vertebrae), by Weitzel (1949). Wood (1976a: 146) suggested a frugivorous diet for the genus. A. stehlinschaubi, however, has cheek teeth with many more accessory crests and wrinkles than has A. macrurus, and in these structures it is similar to modern Petaurista and Trogopterus. Petaurista was considered to have an arboreal rating of 4-5 (out of 5) and a herbivory rating of 3-5 (out of 5) by Eisenberg (1978) and was stated to have a diet of leaves, fruit and seeds by Muul & Lim (1978). Trogopterus is recorded as feeding on oak leaves (Walker 1975). Trogopterus has slightly higher-crowned and more elaborately sele- nodont and crested teeth, but is otherwise very similar in pattern to A. stehlinschaubi, having upper molar hypocones unlike Petaurista. Both modern genera, however, have less transverse motion in both buccal and lingual wear phases, although within the Sciurus type (see Butler 1980). It is concluded therefore that A. stehlinschaubi was arboreal and essentially a folivore. Iomys shares with the Creechbarrow pseudosciurids squared upper molars with large hypo- cone joined by metalophule to the metacone, and lower molars with hypolophulid. Of the three pseudosciurids in the fauna, Suevosciurus is closest in molar morphology to Jomys, but even so is less lophodont than this genus, and wear is less oblique. Jomys has a diet of fruit and seeds (Muul & Lim 1978) so it is probable that Sciuroides rissonei, Treposciurus helveticus preecei and Suevosciurus authodon were also frugivorous/granivorous. Suevosciurus is known from an almost complete articulated skeleton (S. ehingensis, M. Oligocene of Armissan; Lavocat 1955, as Pseudosciurus suevicus). Its fore and hind limb proportions and unfused tibia and fibula suggested to Lavocat a primitive scrambler, capable of living both on the ground and to a certain extent in trees, but with no specific arboreal adaptations. This was probably also the case with Sciuroides and Treposciurus and appears to have been the primitive rodent condition (Lavocat 1955, Wood 1962). The later (Oligocene) rather restricted distribution of Suevosciurus (mainly in southern Germany) suggests some specialization, but the ability of two lineages of this genus to cross the 446 J. J. HOOKER ‘Grande Coupure’ virtually unchanged suggests some versatility of habits or the occupation of a niche not exploited by the new immigrants. The drastic reduction of Sciuroides at the same time as the spread of theridomyids in the early Ludian does not support the theory that both shared similar ecological adaptations (see Schmidt-Kittler 1971a: 125-127). The slightly differ- ent emphasis of longitudinal (in Sciuroides) versus transverse (in Suevosciurus) crests is likely to reflect only minor dietary differences. APATOTHERIA. Stehlin (1916) compared the genus Heterohyus with the modern lemuroid Daub- entonia mainly in search of relationship, but in this he has subsequently been shown to be incorrect. West (1973: 23-24) compared Apatemys teeth functionally with those of the pha- langerid Dactylopsila and concluded that like the latter it too may have been arboreal and used its enlarged incisors for tearing bark open to hunt for insects. Cartmill (1974) compared skull structure in relation to food-gathering function in both Daubentonia and Dactylopsila and found similar specializations, including clinorhynchy, in both. Scott & Jepsen’s (1936: pl. 5) side view of the skull of the Oligocene apatemyid Sinclairella suggests a certain degree of clin- orhynchy but less than occurs in Dactylopsila. Skulls of Heterohyus are virtually unknown, but the incisors are robust and higher-crowned than in Sinclairella or Dactylopsila, although without the persistent pulps of Daubentonia. An arboreal mode of life hunting wood-boring insects is likewise envisaged for Heterohyus. PAROXYCLAENIDAE. Some similarities to the reduced bunodont to semi-bunodont cheek teeth of various viverrids (e.g. Arctogalidea) and procyonids (e.g. Bassaricyon, Nasua) are shown by the paroxyclaenid Paroxyclaenus. These viverrids and procyonids are mainly frugivorous and partly carnivorous (Walker 1975). Other paroxyclaenid genera (e.g. Vulpavoides, Kopidodon) have analogues in more insectivorous viverrids like Bdeogale, Crossarchus and Eupleres. Molar size reduction is accompanied by occasional increase in number in Kopidodon and also in the modern canid Otocyon (see Tobien 1969, Van Valen 1964). An approach to the semi- zalambdodonty of Vulpavoides is found in Eupleres, although here the teeth are excessively reduced in size and robusticity. No exact modern analogue has been found for the curious high-cusped M! of Vulpavoides cooperi. Otocyon eats insects, especially termites, but is also partly frugivorous and carnivorous; Eupleres catches adult and larval insects by digging and also eats small tetrapods (Walker 1975). It is possible that Vulpavoides combined insectivory, carnivory and frugivory in its diet in a similar way. PERISSODACTYLA. The nearest modern relative is obviously no guide to diet in the three Creech- barrow palaeothere genera. However, Propalaeotherium messelense, a very close relative of P. parvulum, is recorded from Messel with preserved gut contents consisting in one case of leaves of several dicotyledon families (Sturm 1978), and in another of Vitis (grape) seeds (Koenigswald & Schaarschmidt 1983). Of the leaves, species of the Lauraceae and Apocynaceae predominate, with Annonaceae or Symplocaceae, Myrtaceae, Juglandaceae and Moraceae as more minor constituents; of these, only the Annonaceae, Symplocaceae and ?Moraceae occur in the poorly- known flora of Bed A3 at Barton Cliff (penecontemporaneous with Creechbarrow). The closely related Lophiotherium siderolithicum shows trends in premolar molarization, perhaps towards a coarser diet, but may still have been folivorous-frugivorous. Plagiolophus curtisi is the largest species in the fauna so far known and has the highest- crowned teeth. These show, however, only incipient trends in hypsodonty and their pattern has been described as selenolophodont. A herbivorous diet, coarser than for either of the other palaeotheres, is suggested. The occurrence of upper and lower jaws preserved together at two localities suggests perhaps powerful jaw muscles. This together with the massive parallel- orientated procumbent canines and robust cheek teeth, unlike other species of the genus, Suggests that it may have had rooting habits. This species exceeds the size range for arboreal folivores (see Text-fig. 70C and Eisenberg 1978), and for this and family-wide osteological reasons all three palaeotheres were likely to have been ground dwellers. The probability of larger herbivorous perissodactyls like Palaeotherium and Lophiodon lautri- cense (from penecontemporaneous records from Hengistbury and Barton — see p. 214) having BARTONIAN MAMMALS OF HAMPSHIRE BASIN 447 occurred at Creechbarrow is high. The latter’s molar specializations resemble those of modern tapirs, which feed on leaves of aquatic and low-growing terrestrial plants, favouring the most abundant taxa (Walker 1975). Fischer (1964: 69-70) compared their mode of chewing with that of both tapirs and pigs. He then, however, stressed the differences from tapirs: e.g. absence of evidence for a proboscis, presence of conical incisors and larger canines in Lophiodon. He compared the postcranial skeleton with that of the extinct Pleistocene species of the rhino- ceroses Dicerorhinus and Coelodonta. ARTIODACTYLA. Estimating way of life of this group from either nearest living relative or modern dental analogue is difficult. There are two relevant Messel skeletons, Messelobunodon and Masillabune, with gut contents. Those of the dichobunid Messelobunodon schaeferi were found to be undetermined fungi along with decayed leaves and it was concluded that it acquired the former by rooting on the forest floor (Richter 1981). Hyperdichobune sp. 1 (p. 376), as it is a dichobunid, may have had similar habits, but the molar mesostyle and larger metacon- ule and incipient P* molarization might point to frugivorous or folivorous tendencies. The gut contents of the haplobunodontid Masillabune were found to be leaves (Tobien 1980). The probability that Haplobunodon venatorum was also highly folivorous is supported by its convergent dental similarities to Propalaeotherium aff. parvulum. Mixtotherium and Dacrytherium belong in different families but are dentally rather similar, being low-crowned and strongly selenodont with strong molar and, in the case of Mixto- therium, P* mesostyles. The closest modern analogues seem to be the indriid primates Pro- pithecus and Indri. Particular similarities are the configuration of the upper premolars and molars with prominent recurved parastyles, especially like Mixtotherium. The main difference is the weakness of the selenodonty in the lower molars and in the lingual cusps of the upper molars. Modern indriids are predominantly arboreal folivores (Kay & Hylander 1978) but the postcranial skeleton of Mixtotherium is unknown. That of anoplotheres is, however, typically artiodactyl and it is likely that both genera were ground-dwelling folivores. Dichodon had higher-crowned selenodont teeth than Dacrytherium and was larger, but like Mixtotherium P4 was molariform. Its blade-like anterior dentition without diastema was like Dacrytherium but appears to have no modern analogue. Modern tragulids and small cervids have higher-crowned molars and different food-gathering apparatus, with incisiform canines and a marked diastema. Modern tragulids are both frugivorous and folivorous but have simpler, less selenodont molars than Dichodon. Molars of modern small deer (browsers) are more comparable in their degree of selenodonty and development of upper buccal and lower lingual stylar cusps. It is considered likely that Dichodon was primarily folivorous. Similarities of the teeth of cebochoerids (Cebochoerus and Acotherulum) to cercopithecid primates misled Filhol (1877b) to suggest a relationship between these and suoid artiodactyls (cebochoerids were originally classified as pigs). Similarities in brachyodont bunodont dental pattern of the Creechbarrow cebochoerids to the dominantly frugivorous cercopithecid taxa are striking and suggest a similar diet for them. Once again, the postcranial skeleton is unknown but it is likely that they were ground-living like most other artiodactyls. Conclusions This probable community, which is rich in primates, arboreal species generally and species with frugivorous and folivorous diets, is typical of a forest environment, amongst available modern habitat types (see Davis 1962, Nesbit Evans et al. 1981). Non-mammal support for this environment is the high proportion of shade-demanding land gastropods (Preece 1980a). The number of mammal species (40 excluding bats) is close to that of a modern temperate forest (Davis 1962). However, in view of the likely incompleteness of the Creechbarrow record (see p. 438; and Wolff 1975), the postulated original number (nearly double that recorded) is closer to that of a modern tropical forest (Davis 1962). Buchardt (1978), Chateauneuf (1980), Collinson et al. (1981), Daley (1972b), Murray & Wright (1974) and Wolfe (1978, 1980) have nevertheless all pointed to a cooler than tropical climate in northern latitudes in the Eocene. The warmest temperatures are thought to occur in the earliest Lutetian or latest 448 J. J. HOOKER Ypresian, decreasing to a Palaeogene minimum by the end of the Priabonian (= late Ludian— early Stampian). The temperatures postulated on isotopic evidence by Buchardt (1978) for the age under discussion fall within the modern forest climate called paratropical by Wolfe (1979). A number of the nearest living relatives and nearest dental analogues of the Creechbarrow mammal assemblage are confined to tropical forest. This probably reflects the results of relic- tion and retreat from increasing seasonality (Daley 1972b) rather than an original geographical and climatic restriction. It is concluded that the Creechbarrow mammal fauna probably inhab- ited a forest, existing under a paratropical climatic regime. The diversity of tarsioids and of ground-dwelling frugivores and folivores suggests an abundance of low-growing shrubs but without dense herbaceous undergrowth. No flora is known from the Creechbarrow Limestone, presumably for preservational reasons, but its probable time equivalent the Barton Clay bed A3 at Barton Cliff has yielded some fruits and seeds (Chandler 1960). In terms of nearest living relatives this small flora is thought to include two coniferous trees, three aquatic or subaquatic herbs, three marginal aquatic herbs/ shrubs, six shrubs/small trees, two probable lianes and several unknowns (see Chandler 1960, 1964). Unfortunately this is a low-diversity assemblage dominated by those plant organs most likely to survive transport into a marine depositional environment. It is probably not a true representation of the vegetation of the area (Chandler 1964; M. E. Collinson, personal commu- nication 1982). The spatial relationships of the presumed forest to the shallow lake depositional environment of the Creechbarrow Limestone is unknown. In view of the unsorted appearance of the mammal assemblage it is unlikely to have been far away. Palaeogeography Faunal Provinces. Schmidt-Kittler & Vianey-Liaud (1975) and Schmidt-Kittler (1977b) divided the main European island area in the late Ludian into several faunal provinces based on rodents: 1, southern England plus Belgium and Germany west of the Rhine Graben; 2, southern Germany east of the Rhine Graben; 3, the Franco-Swiss area; and 4, Spain. At this time the inundated Rhine Graben (and possibly a non-marine area linking the latter to the Paris Basin, see Krutsch & Lotsch 1958) were important physical barriers. The former authors, however, considered that the main reasons for separation were ecological, a forested area existing in southern Germany east of the Rhine Graben, favouring such pseudosciurids as Pseudosciurus and Suevosciurus. Franzen (1968: 158-159) distinguished northern and southern faunal provinces in the Ludian based on the genus Palaeotherium. His northern province equates with 1 and 2 of Schmidt- Kittler & Vianey-Liaud (1975) and his southern province with their 3 and 4. He noted a tendency for species with short, broad feet to occupy the north and for species with longer and more slender feet to occupy the south. He suggested that the north might have been a damper, more wooded region and the south drier and more open. Sudre (1978b: 192-196) found some evidence from artiodactyls and primates to support Schmidt-Kittler & Vianey-Liaud’s (1975) rodent provinces. He also noted endemic forms in the southern English Ludian (see also Sudre 1974), Spain and southern Germany. Faunal provinces are less easy to envisage in the Bartonian because of the small number of localities, but there seems to be a tendency for endemics such as Leptolophus, Plagiolophus cartailhaci, P. aff. annectens, Elfomys, Necrolemur, Peratherium, Saturninia, Adapis, Gliravus, Paradelomys, Pseudoltinomys, Anchilophus, Xiphodon, and Haplomeryx to occur in the south. There is also a tendency for Haplobunodon, Plagiolophus curtisi, Treposciurus, Suevosciurus, Nannopithex, Microchoerus and Scraeva to occur in the north. An important zone of overlap occurs in the Mormont area of Switzerland. At the same time a number of taxa are widespread, e.g. Lophiodon lautricense, Lophiotherium siderolithicum, Ailuravus stehlinschaubi and Plesiarc- tomys hurzeleri. It seems that at least during some of the time there were no geographical barriers to dispersal in the later Bartonian. Evidence from southern Germany is restricted to one species (L. lautricense) from one locality (Heidenheim). Sediments of this age are sparse and BARTONIAN MAMMALS OF HAMPSHIRE BASIN 449 ee ea () 500 Lee gare v ee ee km eae Text-figure 71 The palaeogeography of Europe in the Bartonian with Lophiodon lautricense- Lophiotherium siderolithicum Zone mammal localities. Fine dashed lines show outlines of modern land masses. Heavy lines show Bartonian coastlines, with dots on the seaward sides. Stipple indicates principal areas of non-marine sedimentation. Two-way arrow indicates position of postulated Weald-—Artois land bridge. Abbreviations of localities: B = Barton; Be = Berville; Br = Le Bretou; C = Creechbarrow; Ca = Le Castrais localities; E = Eclépens-Gare and Eclépens A; G = Grisolles; Gu = La Guittardie; H = Hengistbury; He = Heidenheim; L = Latilly; P = Paris; R = Robiac; S = Sergy. Solid triangles indicate positions of localities. Sources of palaeogeographic data: Azzaroli (1981), Boccaletti & Manetti (1978), Krutsch & Lotsch (1958), Lemoine (1978), Ollivier-Pierre (1980), Plaziat (1981) and Pomerol (1973). non-marine in the Rhine Graben (Sittler 1969) and it seems unlikely that it could then have formed a physical barrier to the spread of faunas from the north-west. The English Bartonian mammals, except for the Barton and Creechbarrow endemics, must have reached the sites by means of an interchange with continental Europe. Heavy mineral evidence suggests that a land bridge from Dorset to the Cotentin peninsula existed in the Auversian (Morton 1982, i.e. top of his unit 7). It has also been suggested that the Weald—Artois Axis was active at this time (Pomeroi 1973) and that a land bridge existed here between Britain and the continent during the later Bartonian. No direct evidence exists, owing to erosion of 450 J. J. HOOKER Palaeogene sediments in the London Basin. Typical Barton Clay mollusc faunas, which might be expected to occur, are absent from Belgium to the east of the supposed bridge. The presence of continental European mammal species in the English later Bartonian strongly supports the presence of an at least intermittent Weald—Artois or Dorset—Cotentin land bridge. See Text-fig. 71 for palaeogeographic map of Europe in the Bartonian. Migrations after the Bartonian. Schmidt-Kittler (19776) thought that the primitive Treposciurus and Suevosciurus, being common to all provinces in the Ludian, may have represented ‘relics of an earlier fauna common throughout Europe’. However, they are unknown in pre-Ludian deposits except at Creechbarrow. It is more likely therefore that they are differentiated in the northern area and intermittently spread southwards. Treposciurus was first, appearing in the early Ludian of Eclépens B, Fons 4 and Roc de Santa. Suevosciurus came later in the late Ludian of Entreroches and Escamps and the middle Oligocene of Armissan. Whether these genera were already in Bavaria before the Ludian is unclear. Heterohyus morinionensis may also have originated in the north and spread south, giving rise to H. quercyi of the French Ludian. Leptadapis magnus (including aff.) is not recorded from pre-Ludian strata except at Creech- barrow, but some lower molars from old Quercy collections with similar primitive morphology to those from Creechbarrow may be Bartonian. Newcomers to the English early Ludian (for faunal lists, see Hooker & Insole, 1980) from the south are more numerous, e.g. Isoptychus, Plagiolophus annectens, Pseudoloris parvulus, Quercy- gale angustidens, Gesneropithex aff. grisollensis, Gliravus, Catodontherium and Anchilophus. English Ludian descendants of known English or northern endemic Bartonian taxa are restricted to Scraeva spp., Treposciurus spp., Suevosciurus sp. and Microchoerus spp., the list including genera that also spread southwards. The mammalian fauna at the base of the Lower Headon Beds, including its newcomers of uncertain provenance, is thus very different from that which went before. It is dominated by semi-hypsodont theridomyids, palaeotheriids and artiodactyls, at the expense of the more brachyodont taxa, especially pseudosciurids, paramyids, lophiodontids, apatemyids and paroxyclaenids, which became either reduced in abundance and diversity or extinct. Certain little-modified surviving genera often show trends from frugivory towards folivory (e.g. Microchoerus). This sudden faunal turnover is more likely to represent extinction and migration in relation to some important environmental change than an evolutionary jump. Such a turnover may be a critical point reaction to the more gradual floristic changes suggested by Collinson et al. (1981) in the Hampshire Basin coastal area in the late Eocene. Penecontemporaneous changes of a similar degree and nature occur in the mammalian faunas of southern France (Garimond et al. 1975), suggesting that the environmental change was on a European scale. Acknowledgements I am indebted to the following institutions (see abbreviation explanation on p. 195) and individuals for access to collections in their care: Dr H. de Bruijn (GIU), Dr J.-Y. Crochet (UM), Dr B. Engesser (NMB), Dr P. D. Gingerich (UMMP), Dr L. Ginsburg (MNHN), Dr M. Godinot (MNHN), Dr C. Guerin (FSL), Mr G. Gunnell (UMMP), Dr J.-L. Hartenberger (UM), Dr H. Haubold (GH), Dr E. Heintz (MNHN), Prof. J. Hiirzeler (NMB), Dr J.-J. Jaeger (UM), Mr D. J. Kemp (GM), Dr G. Krumbiegel (GH), Prof. H. W. Matthes (GH), Mr A. Morter (BGS), Dr C. G. Riimke (GIU), Dr D. E. Russell (MNHN), Dr P. Sartenaer (IRSNB), Mme M. Sirven (FSL), Mlle C. Sudre (Museum d’Histoire Naturelle de Toulouse), Dr J. Sudre (UM), Dr M. Weidmann (LGM) and Mr C. J. Wood (BGS). I am also indebted to M. M. Crochard, M. M. Girardot, Dr J. Herman, Mr R. L. E. Ford and especially Mr R. Gardner for access to their private collections. Mr & Mrs P. Clasby, Mr R. J. Curtis, Mr I. C. Daniels, Mr W. J. Quayle and Dr A. J. Rundle drew attention to rare Bartonian mammals they had collected and generously presented them to the BM(NH). Dr M. Godinot, Mr D. J. Kemp and Dr J. Sudre provided important casts. I am particularly grateful to Dr M. E. Collinson, Mr J. Cooper, Mr A. P. Currant, Dr R. Preece, Mr W. J. Quayle, Mr A. Rissoné, Mr D. J. Ward and Mrs A. Ward, who contributed much to the Creechbarrow excavations by their enthusiastic and untiring hard work. Mr P. Ensom, Dr A. N. Insole, Mr J. P. James, BARTONIAN MAMMALS OF HAMPSHIRE BASIN 451 Mr & Mrs J. Kermack, Mr D. N. Lewis, Mr P. Richens and Mr T. Windle also assisted in various aspects of the field work. Mr B. Jones and Mr Q. Palmer, of English China Clays (Wareham) Ltd., generously gave access to the Creechbarrow site, provided equipment and facilities and permitted data from a borehole to be used. My late parents provided facilities for the early stages of sediment processing. I would especially like to thank Dr A. A. Bosma, Prof. P. M. Butler, Dr M. E. Collinson, Dr J.-Y. Crochet, Mr A. P. Currant, Dr B. Gardiner, Dr A. W. Gentry, Dr M. Godinot, Dr C. R. Hill, Prof. K. A. Kermack, Mr C. King, Dr D. E. Russell, Dr A. J. Sutcliffe and Mr D. J. Ward for helpful comments and stimulating discussion. Dr M. Weidmann and Mr D. A. Rigassi kindly showed me the Mormont localities and provided important information on them. Drs M. E. Collinson and A. W. Gentry critically read the text. This report was submitted for the degree of Ph.D. in the University of London and I thank my supervisors Dr A. W. Gentry, Prof. K. A. Kermack and the late Dr W. R. Hamilton for their advice and encouragement during the project. Dr M. E. Collinson provided encouragement and inspiration through- out and helped in many different ways. The Photographic Studio (Dept. of Central Services, BM(NH)) took some of the photographs. Mr F. M. P. Howie, Mr B. Martin and Ms A. Longbottom took X-ray photographs. Mr C. H. Shute took the photograph reproduced in PI. 35, fig. 1c. I would like to thank Mrs P. Williams for typing. References Alston, E. R. 1876. On the classification of the order Glires. Proc. zool. Soc. Lond., 1876: 61—98, pl. 4. Anadon, P. & Zamarreno, I. 1981. Paleogene nonmarine algal deposits of the Ebro Basin, northeastern Spain. In Monty, C. (ed.), Phanerozoic Stromatolites: 140-154, 4 pls, 2 figs. Berlin, Heidelberg, New York. Andrews, C. W. 1907. Note on the cervical vertebra of a Zeuglodon from the Barton Clay of Barton Cliff (Hampshire). Q. J! geol. Soc. Lond., 63: 124-127, 1 fig. Arkell, W. J. 1947. The Geology of the Country around Weymouth, Swanage, Corfe and Lulworth. xix + 386 pp., 19 pls, 84 figs. Mem. geol. Surv. U.K., London. Astre, G. 1960. Quelques variants extremes chez Lophiodon. Bull. Soc. Hist. nat. Toulouse, 95 (3/4): 213-224, 2 figs. ’ Aubry, M.-P. 1983. Biostratigraphie du Paléogéne épicontinental de Europe du Nord-Ouest. Etude fondée sur les nannofossiles calcaires. Docums Lab. Geol. Fac. Sci. Lyon, 89. 317 pp., 38 figs, 8 pls, 7 tabs. Aymard, A. 1846. Essai monographique sur un genre nouveau de mammifére fossile trouve dans la Haute-Loire et nommé Entelodon. Annls Soc. Agric. Sci. Puy, 12: 227-267, 1 pl. — 1850. (Réponse a M. Robert.) Annls Soc. Agric. Sci. Puy, 14: 80-86. Azzaroli, A. 1981. Cainozoic mammals and the biostratigraphy of the island of Sardinia, Western Medi- terranean. Palaeogeogr. Palaeoclimat. Palaeoecol., Amsterdam, 36: 107—111, 1 fig. Beaumont, G. de 1966. Les Miacinae (Carnivora, Miacidae) de |’Eocéne de la Suisse. Bull. Soc. vaud. Sci. nat., Lausanne, 69 (6): 273-285, 1 pl., 2 figs. Bell, D. L. & Goodell, H. G. 1967. A comparative study of glauconite and the associated clay fraction in modern marine sediments. Sedimentology, Amsterdam, New York, 9: 169-202. Black, C. C. 1968. The Oligocene rodent Ischyromys and discussion of the family Ischyromyidae. Ann. Carneg. Mus., Pittsburgh, 39 (18): 273-305, 26 figs, 3 tabs. 1971. Paleontology and geology of the Badwater Creek Area, Central Wyoming. Part 7, Rodents of the family Ischyromyidae. Ann. Carneg. Mus., Pittsburg, 43 (6): 179-217, 73 figs, 2 tabs. Blainville, H. M. D. de 1864 (1839-64). Osteographie, ou Description Iconographique composée du Squelette et du Systeme dentaire de Mammiferes récents et fossiles pour Servir de base a la zoologie ou a la géologie. 8 vols, Paris. Blissenbach, E. 1954. Geology of alluvial fans in semi-arid regions. Bull. geol. Soc. Am., New York, 65: 175-190. Blondeau, A. & Pomerol, C. 1969. A contribution to the sedimentological study of the Palaeogene of England. Proc. Geol. Ass., London, 79 (4): 441-455, 4 figs. Bloomfield, A. H. 1911. Bembridge fossils on Creechbarrow Hill, Isle of Purbeck. Geol. Mag., London, (5) 8 (1): 46. —— 1913. Bembridge Limestone at Creechbarrow Hill. Geol. Mag., London, (5) 10: 45—46, 1 fig. Blumenbach, J. F. 1779. Handbuch der Naturgeschichte, 1. 448 pp., 2 pls. Gottingen. 452 J. J. HOOKER Boccaletti, M. & Manetti, P. 1978. The Tyrrhenian Sea and adjoining regions. In Nairn, A. E. M., Kanes, W. H. & Stehli, F. C. (eds), The Ocean Basins and Margins, 4B: 149-199. New York. Bonaparte, C.-L. 1850. 1. Conspectus Systematis Mastozoologiae — Editio altera reformata. In: Conspectus Systematum. 4 charts, Leiden. Bonnichsen, R. 1979. Pleistocene bone technology in the Beringian refugium. Pap. archaeol. Surv. Canada, Ottawa, (Mercury ser.) 89. xiii + 297 pp. Bosco, C. 1902. I] Lophiodon sardus delle ligniti di Terras de Collu (Sardegna). Atti Accad. naz. Lincei Rc., Rome, (5) 11 (2nd Sem.): 178-182, 3 figs. Bosma, A. A. 1974. Rodent biostratigraphy of the Eocene—Oligocene transitional strata of the Isle of Wight. Utrecht micropaleont. Bull. spec. Publs, 1. 128 pp., 7 pls, 36 figs. —— & Insole, A. N. 1976. Pseudosciuridae (Rodentia, Mammalia) from the Osborne Beds (Headonian), Isle of Wight, England. Proc. K. ned. Akad. Wet., Amsterdam, (B) 79: 1-8, 1 pl. Bowdich, T. E. 1821. An Analysis of the natural classifications of Mammalia, for the use of Students and Travellers. 115 pp., 15 pls. Paris. Bown, T. M. & Schankler, D. 1982. A review of the Proteutheria and Insectivora of the Willwood Formation (lower Eocene), Bighorn Basin, Wyoming. Bull. U.S. geol. Surv., Washington, 1523: 1-79, 10 pls, 18 figs, 8 tabs. Bravard, A. 1850. Notes sur les ossements fossiles de Débruge. In Gervais, P., Zoologie et Paléontologie frangaises ... (etc.), 2 (Explication des planches et divers mémoires) expl. pls 46-48: 2-3. Paris. (See Gervais 1852). Brisson, M. J. 1762. Regnum animale in classes IX. distributum, sive synopsis methodica. Editio altera auctior. (vill +) 296 pp. Leiden. Brunet, M. & Gabilly, J. 1981. Decouverte d’une faune de Vertébrés bartoniens dans le Tertiaire continen- tal du Seuil du Poitou. Bull. Soc. geol. Fr., Paris, (7) 23 (1): 95—100, 1 pl. Buchardt, B. 1978. Oxygen isotope palaeotemperatures from the Tertiary Period in the North Sea area. Nature, Lond., 275: 121—123, 2 figs. Bujak, J. P., Downie, C., Eaton, G. L. & Williams, G. L. 1980. Dinoflagellate cysts and acritarchs from the Eocene of southern England. Spec. Pap. Palaeont., London, 24: 1-100, 22 pls, 24 figs. Burton, E. St J. 1929. The Horizons of Bryozoa (Polyzoa) in the Upper Eocene Beds of Hampshire. Q. J geol. Soc. Lond., 85: 223-239. — 1933. Faunal horizons of the Barton Beds in Hampshire. Proc. Geol. Ass., London, 44: 131-167. Bury, H. 1934. “Creech Barrow”. Proc. Bournemouth nat. Sci. Soc., 26: 68-72. Butler, P. M. 1939. Studies of the mammalian dentition. Differentiation of the postcanine dentition. Proc. zool. Soc. Lond., 109: 1—36. 1947. An Arctocyonid from the English Ludian. Ann. Mag. nat. Hist., London, (11) 13 (106): 691-701, 2 figs. — 1952a. The milk-molars of Perissodactyla, with remarks on molar occlusion. Proc. zool. Soc. Lond., 121 (4): 777-817, 16 figs. — 1952b. Molarisation of the premolars in the Perissodactyla. Proc. zool. Soc. Lond., 121 (4): 819-843, 93 figs. —— 1956. The skull of Ictops and the classification of the Insectivora. Proc. zool. Soc. Lond., 126 (3): 453-481, 8 figs. —— 1967. Dental merism and tooth development. J. dent. Res., Baltimore, 46: 845-850. 1972. The problem of insectivore classification. In Joysey, K. A. & Kemp, T. S., Studies in Vertebrate Evolution: 253-265, 6 figs. Edinburgh. 1980. Functional aspects of the evolution of rodent molars. Palaeovertebrata, Montpellier, Mem. jub. Lavocat: 249-262, 7 figs. Cartmill, M. 1974. Daubentonia, Dactylopsila, woodpeckers and klinorhynchy. In Martin, R. D., Doyle, G. A. & Walker, A. C., Prosimian Biology: 655-670, 5 figs, 2 tabs. London. Casanovas-Cladellas, M. L. 1975. Estratigraphia y Paleontologia del yacimiento ludiense de Roc de Santa (area del Noguera—Pallaresa). Paleont. Evoluc., Sabadell, 10: 1-158, 11 pls, 12 figs, 28 tabs. Cavelier, C. 1979. La limite Eocéne—Oligocéne en Europe occidentale. Sci. géol. Inst. Geol. Strasbourg (Mém.) 54: 1—280, 6 figs, 40 tabs. —— 1980. Ludien. In Cavelier, C. & Roger, J., Les étages frangais et leurs stratotypes. Mem. Bur. Rech. geol. minier., Paris, 109: 237-244. & Pomerol, C. 1976. Les rapports entre le Bartonien et le Priabonien. Incidence sur la position de la limite Eocéne moyen—Eocene supérieur. C.r. somm. Seanc. Soc. géol. Fr., Paris, 1976 (2): 49-51, 1 tab. Chamberlain, C. K. 1975. Recent lebensspuren in nonmarine aquatic environments. In Frey, R. W. (ed.), The Study of Trace Fossils. A synthesis of Principles, Problems and Procedures in Ichnology: 431—458, 10 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 453 figs. Berlin, Heidelberg, New York. Chandler, M. E. J. 1960. Plant remains of the Hengistbury and Barton Beds. Bull. Br. Mus. nat. Hist., London, (Geol.) 4: 191—238, pls 29-35. —— 1964. A summary and survey of findings in the light of recent botanical observations. The Lower Tertiary Floras of Southern England, 4. xii + 151 pp., 6 figs, 4 pls. London, British Museum (Natural History). —— 1978. Supplement to the Lower Tertiary floras of Southern England, part 5. Tert. Res. spec. Pap., London, 4. iv + 47 pp., 20 pls, 6 figs. Chantre, E. & Gaillard, C. 1897. Sur la faune du gisement sidérolithique ¢océne de Lissieu (Rh6ne). Cr. hebd. Seanc. Acad. Sci., Paris, 125: 986-987. Chateauneuf, J.-J. 1980. Palynostratigraphie et paléoclimatologie de ’Eocéne supérieur et de l’Oligocéne du Bassin de Paris (France). Mem. Bur. Rech. géol. miniér., Paris, 116: 1-357, 31 pls. Chow Minchen 1964. A lemuroid primate from the Eocene of Lantian, Shensi. Vertebr. palasiat., Peking, 8: 257-262, 1 pl. [In Chinese and English]. Clarke, M. R. 1981. The sand and gravel resources of the country north of Bournemouth, Dorset. Miner. Assess. Rep. Inst. geol. Sci., London, 51. 11 + 128 pp., 12 figs, 1 pl. Clemens, W. A. 1966. Fossil Mammals of the Type Lance Formation, Wyoming. Part 2, Marsupialia. Univ. Calif. Publs geol. Sci., Berkeley, 62. vi + 122 pp., 77 figs, 24 tabs. 1979. Marsupialia. In Lillegraven, J. A., Kielan-Jaworowska, Z. & Clemens, W. A. (eds), Mesozoic Mammals, the first two thirds of mammalian history: 192-220, 7 figs. Berkeley, Los Angeles, London. Collinson, M. E. 1983. Palaeofloristic assemblages and palaeoecology of the Lower Oligocene Bembridge Marls, Hamstead Ledge, Isle of Wight. Bot. J. Linn. Soc., London, 86: 177-225, 68 figs. —— , Fowler, K. & Boulter, M. 1981. Floristic changes indicate a cooling climate in the Eocene of southern England. Nature, Lond., 291: 315-317, 2 figs. Coombs, W. P., jr & Coombs, M. C. 1977. The origin of anthracotheres. Neues Jb. Geol. Palaont. Mh., Stuttgart, 1977 (10): 584-599, 5 figs. Cooper, C. F. 1910. Microchoerus erinaceus (Wood). Ann. Mag. nat. Hist., London, (8) 6: 39-43, pl. 3. — 1928. Pseudamphimeryx hantonensis, sp. n., with notes on certain species of artiodactyls from the Eocene deposits of Hordwell. Ann Mag. nat. Hist., London, (10) 2: 49—SS, pls 2-3. Cooper, J. 1976. British Tertiary Stratigraphical and Rock terms formal and informal, additional to Curry 1958, Lexique Stratigraphique International. Tert. Res. spec. Pap., London, 1: 1-37. Cope, E. D. 1867. An addition to the vertebrate fauna of the Miocene period, with a synopsis of the extinct Cetacea of the United States. Proc. Acad. nat. Sci. Philad., 1867: 138-156. 1873. Third account of new Vertebrata from the Bridger Eocene of Wyoming Territory. Proc. Am. phil. Soc., Philadelphia, 12: 469-472. —— 1880. On the genera of the Creodonta. Proc. Am. phil. Soc., Philadelphia, 19: 76-82. —— 1881. A new type of Perissodactyla. Am. Nat., Salem, 15: 1017-1018. — 1883. The extinct Rodentia of North America. Am. Nat., Salem, 17: 43-57, 165—174, 370-381. —— 1884. The Vertebrata of the Tertiary formations of the West. Rep. U.S. geol. Surv. Territ., Washing- ton, (F. V. Hayden) 3: 1-1009, Atlas pls 1—75a. — 1889. The Vertebrata of the Swift Current River, II. Am. Nat., Salem, 23: 151-155. Costa, L. I., Downie, C. & Eaton, G. L. 1976. Palynostratigraphy of some Middle Eocene sections from the Hampshire Basin (England). Proc. Geol. Ass., London, 87 (3): 273-284, 3 figs, 1 tab. Cox, L. R. 1954. Obituary notice — Arthur George Wrigley. Proc. malac. Soc. Lond., 30 (6): 157-160. Cray, P. E. 1973. Marsupialia, Insectivora, Primates, Creodonta and Carnivora from the Headon Beds (Upper Eocene) of southern England. Bull. Br. Mus. nat. Hist., London, (Geol.) 23 (1): 1-102, 6 pls, 23 figs, 18 tabs. Creighton, G. K. 1980. Static allometry of mammalian teeth and the correlation of tooth size and body size in contemporary mammals. J. Zool., Lond., 191: 435-443, 2 figs. Crochet, J.-Y. 1977. Les Didelphidae (Marsupicarnivora, Marsupialia) holarctiques tertiaires. C.r. hebd. Seanc. Acad. Sci., Paris, (D) 284: 357-360, | pl. — 1979. Diversité systematique des Didelphidae (Marsupialia) européens tertiaires. Geobios, Lyon, 12 (3): 365-368, 16 text-figs. — 1980. Les Marsupiaux du Tertiaire d'Europe. 279 pp., 2 pls, 241 figs, 40 tabs. Paris (Edns Fond. Singer-Polignac). —— , Hartenberger, J.-L., Rage, J.-C., Remy, J. A., Sige, B., Sudre, J. & Vianey-Liaud, M. 1981. Les nouvelles faunes de vertébrés antérieurs a la “Grande Coupure’ découvertes dans les phosphorites du Quercy. Bull. Mus. natn. Hist. nat., Paris, (4) 3 (C, Sci. terre) (3): 245-266. Crusafont-Pairo, M. 1967. Sur quelques prosimiens de l’Eocene de la zone préaxiale pyrénaique et un essai 454 J. J. HOOKER provisoire de réclassification. Colloques int. Cent. natn. Rech. scient., Paris, Problémes actuels de Paléon- tologie — Evolution des Vertébrés, no. 163: 611-632. —— & Golpe-Posse, J. M. 1975. Les prosimiens de la région préaxiale pyrénaique. II Adapidae. Colloques int. Cent. natn. Rech. scient., Paris, Problémes actuels de Paléontologie — Evolution des Vertébrés, no. 218: 851-859, 3 figs, 1 pl. —— & Russell, D. E. 1967. Un nouveau Paroxyclaenidé de lEocéne d’Espagne. Bull. Mus. natn. Hist. nat. Paris, (2) 38: 757-773. Curry, D. 1937. The English Bartonian Nummulites. Proc. Geol. Ass., London, 48: 229-246, 2 pls. 1942. The Eocene succession at Afton Brickyard, I.O.W. Proc. Geol. Ass., London, 53 (2): 88-101, pl. 3. —— 1958a. Angleterre, Pays de Galles, Ecosse. Paleogene. Lexique Strat. Int., Paris, 1, Europe (3a xii). 82 pp. —— 1958b. Itinerary IV. The Barton Area. In Curry, D. & Wisden, D. E., The Geology of the Southampton Area. Geol. Ass. Guide, Colchester, 14: 8-12. — 1962. Sur la découverte de Nummulites variolarius (Lamarck) dans le Lutétien des bassins de Paris et du Hampshire. C.r. somm. Seanc. Soc. geéol. Fr., Paris, 1961: 247. 1967. Problems of correlation in the Anglo—Paris—Belgian Basin. Proc. Geol. Ass., London, 77: 437-467, 5 figs. — 1977. The age of the Hengistbury Beds (Eocene) and its significance for the structure of the area around Christchurch, Dorset. Proc. Geol. Ass., London, 87 (4): 401—407, 2 figs. —— 1981. Bartonian. In Pomerol, C. (ed.), Stratotypes of Paleogene Stages. Bull. Inf. Geol. Bassin Paris, (Mem. hors serie) 2: 23-36, 2 figs. ——, Adams, C. G., Boulter, M. C., Dilley, F. C., Eames, F. E., Funnell, B. M. & Wells, M. K. 1978. A correlation of Tertiary rocks in the British Isles. Spec. Rep. geol. Soc. Lond. 12. 72 pp., 5 figs, 2 tabs. —— , Daley, B. & Edwards, N. 1972. Itinerary I, Whitecliff Bay. In Curry, D., Daley, B., Edwards, N., Middlemiss, F. A., Stinton, F. C. & Wright, C. W., The Isle of Wight (3rd edn). Geol. Ass. Guide, Colchester, 25: 3—7, figs 2-3. —— & Edwards, N. 1972. Itinerary VIII, Alum Bay and Headon Hill. In Curry, D., Daley, B., Edwards, N., Middlemiss, F. A., Stinton, F. C. & Wright, C. W., The Isle of Wight (3rd edn). Geol. Ass. Guide, Colchester, 25: 21—23. —— , Hodson, F. & West, I. M. 1968. The Eocene succession in the Fawley Transmission Tunnel; with an appendix on the clay mineralogy of the strata by R. J. Gilkes. Proc. Geol. Ass., London, 79 (2): 179-206, 6 figs, 1 tab. —— , King, A. D., King, C. & Stinton, F. C. 1977. The Bracklesham Beds (Eocene) of Bracklesham Bay and Selsey, Sussex. Proc. Geol. Ass., London, 88 (4): 243-254, 6 figs. Cuvier, G. 1804-05. Sur les espéces d’animaux dont proviennent les os fossiles repandus dans la pierre a platre des environs de Paris (1-3 meém., | sect.). Annls Mus. Hist. nat. Paris, 3: 275-303, 364-387, 442-472, pls 1-13, 1(38)-6(43) (1804). (3 meém., 2 sect.): Restitution des pieds de devant. Loc. cit., 6: 253-283, pls 1-5(50—54) (1805). —— 1812. Recherches sur les ossemens fossiles de quadrupeédes ou l’on retablit les caracteres de plusieurs especes d’animaux que les revolutions du globe paroissent avoir détruites, 3. 394 pp., 68 pls. Paris. — 1822. Recherches sur les ossemens fossiles, ou l’on retablit les caracteres de plusieurs animaux, dont les revolutions du globe ont detruit les especes. (New edn), 2. 648 pp., 60 pls. 3. 412 pp., 80 pls. Paris. Daley, B. 1972a. Macroinvertebrate assemblages from the Bembridge Marls (Oligocene) of the Isle of Wight, England, and their environmental significance. Palaeogeogr. Palaeoclimat. Palaeoecol., Amster- dam, 11: 11—32, 4 figs, 2 tabs. —— 1972b. Some problems concerning the early Tertiary climate of southern Britain. Palaeogeogr. Palaeoclimat. Palaeoecol., Amsterdam, 11: 177—190. 1974. Shell encrusting Algae from the Bembridge Marls (Lattorfian) of the Isle of Wight, Hampshire, England. Revue Micropaleont., Paris, 17 (1): 15-22, 3 pls. — & Edwards, N. 1971. Palaeogene warping in the Isle of Wight. Geol. Mag., London, 108 (5): 399-405, 3 figs. —— & Insole, A. N. 1979. Lithostratigraphical nomenclature of the English Palaeogene succession. Geol. Mag., London, 116 (1): 65-66. Davis, D. D. 1962. Mammals of the lowland rain-forest of North Borneo. Bull. natn. Mus. St. Singapore, 31: 1-129, 23 pls, 20 figs. Dearman, W. R. 1964. Wrench faulting in Cornwall and south Devon. Proc. Geol. Ass., London, 74 (3): 265-287, 5 figs. Dechaseaux, C. 1965. Artiodactyles des Phosphorites du Quercy. 1. Etude sur le genre Dichodon. Annls BARTONIAN MAMMALS OF HAMPSHIRE BASIN 455 Paleéont., Paris, (Vert.) 51 (2): 91-108, 8 figs. Decker, R. L. & Szalay, F. S. 1974. Origins and function of the pes in the Eocene Adapidae (Lemuriformes, Primates). In Jenkins, F. A. jr., Primate Locomotion: 261-291, 11 figs. New York, London. Dedieu, P. 1977. Sur la Systematique des Tapiroidea (Mammalia) de l’Eocéne européen. C.r. hebd. Séanc. Acad. Sci., Paris, (D) 284 (22): 2219-2222, | fig. Dehm, R. 1937. Uber die altertiare Nager Familie Pseudosciuridae und ihr Entwicklung. Neues Jb. Miner. Geol. Palaont. BeilBd., Stuttgart, (B) 77: 268-290. —— 1970. Zur Bezeichnung fossil fuhrender Spaltenfiillungen. Mitt. bayer. St. Palaont. Hist. Geol., Miinchen, 10: 351—364, | fig. Deperet, C. 1894. Sur un gisement sidérolithique de mammiféres de Eocéne moyen, a Lissieu, prés Lyon. C.r. hebd. Seanc. Acad. Sci., Paris, 118: 822-823. 1901. Révision des formes européennes de la famille des Hyracothéridés. Bull. Soc. geol. Fr., Paris, (4) 1: 199-225, pls 4—S. —— 1906. Sur une nouvelle espece de Lophiodon (L. Thomasi) du Bartonien de Sergy (Aisne). Bull. Soc. geol. Fr., Paris, (4) 6: 688-692, 1 fig. —— 1908. L’histoire géologique et la phylogénie des anthracothértidés. C.r. hebd. Seanc. Acad. Sci., Paris, 146: 158-162. —— 1910. Etudes sur la famille des Lophiodontidés. Bull. Soc. géol. Fr., Paris, (4) 10: 558-577, pl. 7. —— 1917. Monographie de la faune de mammiferes fossiles du Ludien inférieur d’Euzet-les-Bains (Gard). Annls Univ. Lyon, (1) 40: 1-290, pls 1-25. -—— & Carriere, G. 1901. Sur un nouveau gisement de mammiféres de !Eocéne moyen a Robiac, prés Saint-Mamert (Gard). C.r. hebd. Seanc. Acad. Sci., Paris, 133: 616-618. Desmarest, A. G. 1822. Mammalogie, ou description des especes de mammiferes, 2: i—viii, 277-555. Paris. Dobson, G. E. 1875. Conspectus of the suborders, families and genera of Chiroptera arranged according to their natural affinities. Ann. Mag. nat. Hist., London, (4) 16: 345-357. Dorr, J. A. 1977. Partial skull of Paleosinopa simpsoni (Mammalia, Insectivora), latest Paleocene Hoback Formation, Central Western Wyoming, with some general remarks on the family Pantolestidae. Contr. Mus. Paleont. Univ. Mich., Ann Arbor, 24 (23): 281-307, 5 figs. Drummond, P. V. O. 1970. The Mid-Dorset Swell. Evidence of Albian-—Cenomanian movements in Wessex. Proc. Geol. Ass., London, 81: 679-714. Eaton, G. L. 1976. Dinoflagellate cysts from the Bracklesham Beds (Eocene) of the Isle of Wight, southern England. Bull. Br. Mus. nat. Hist., London, (Geol.) 26 (6): 225-332, 31 figs, 21 pls. Edwards, F. E. 1852. A Monograph of the Mollusca from the Eocene Formations of England. Part II, Pulmonata. Palaeontogr. Soc. (Monogr.), London: 57—122, pls 10-15. Eisenberg, J. F. 1978. The evolution of arboreal herbivores in the class Mammalia. In Montgomery, G. G. (ed.), The Ecology of Arboreal Folivores: 135-152, 6 figs, 6 tabs. Washington D.C., Symp. natn. zool. Park, Smithson. Instn. Elliott, G. F. 1954. New Brachiopoda from the Eocene of England, France and Africa. Ann. Mag. nat. Hist., London, (12) 7: 721-728, pl. 15. Engesser, B., Matter, A. & Weidmann, M. 1981. Stratigraphie und saugetierfaunen des mittleren Miozans von Vermes (Kt. Jura). Eclog. geol. Helv., Lausanne, Basel, 74 (3): 893-952, 29 figs, 1 tab. Fahlbusch, V. 1976. Report on the International Symposium on mammalian stratigraphy of the European Tertiary. Newsl. Stratigr., Leiden, 5 (2/3): 160-167, 1 tab. Feuillee, P. 1964. Historique de l'utilisation du terme ‘Bartonien’ dans le Bassin de Paris. Mem. Bur. Rech. geol. minier., Paris, 28: 37-46. Filhol, H. 1873. Sur un nouveau genre de lémurien fossile, recemment découvert dans les gisements de phosphate de chaux du Quercy. C.r. hebd. Seanc. Acad. Sci., Paris, 77: 1111-1112. —— 1874. Nouvelles observations sur les mammiféres des gisements de phosphate de chaux (Lémuriens et pachylémuriens). Annls Sci. geol. Paris, 5 (4): 1-36, pls 7-8. —— 1876-77. Recherches sur les Phosphorites du Quercy. Etude des fossiles qu’on y recontre et spéciale- ment des mammiferes. 1. Annls Sci. géol. Paris, 7 (7): 1-220, pls 10-36 (1876a). 2. Loc. cit. 8 (1): 1-338, pls 1-28 (1877b). Also: 1. Bibltque Ec. ht. Etud., Paris Sect. Sci. nat., 15 (4): 1-220, pls 6-32 (1876a). 2. Loc. cit. 16 (1): 1-338, pls 1-28 (18776). 1 + 2. These pres. a Fac. Sci. Paris (&c.). 561 pp., 55 pls. Paris (1877c). — 1876b. Mammifeéres fossiles nouveaux provenant des dépdts de phosphate de chaux du Quercy. Cr. hebd. Seanc. Acad. Sci., Paris, 82: 288-289. — 1877a. Considérations sur la découverte de quelques mammiferes fossiles appartenant a l’e@poque éocene supérieure. Bull. Soc. philomath. Paris, 7 (1): 51-54. 456 J. J. HOOKER —— 1877b, 1877c. See 1876-77. —— 1879. Etude des mammiféres fossiles de St-Gérand-le-Puy. Bibltque Ec. ht. Etud., Paris Sect. Sci. nat., 19 (1): 1-252, 30 pls. —— 1880a. Note sur des mammiferes fossiles nouveaux provenant des phosphorites du Quercy. Bull. Soc. philomath. Paris, 7 (4): 120-125. 18806. Sur la découverte de mammiferes nouveaux dans les dép6ts de phosphate de chaux du Quercy. C.r. hebd. Seanc. Acad. Sci., Paris, 90: 1579-1580. —— 1882. Mémoires rélatifs a quelques mammiferes fossiles provenant des dépdts de phosphorite du Quercy. Bull. Soc. Sci. phys. nat. Toulouse, 5 (1): 19-156, pls 1-10. —— 1884. Observations rélatives a des mammiferes fossiles nouveaux provenant des dépdts de phosphate de chaux du Quercy. Bull. Soc. Sci. phys. nat. Toulouse, 5 (2): 159-203, pls 1-13. —— 1888a. Etude sur les vertébrés fossiles d’Issel (Aude). Mem. Soc. géol. Fr., Paris, (3) 5 (1): 1-188, 21 pls. —— 1888b. Description d’un nouveau genre de Pachyderme provenant des depots de phosphate de chaux du Quercy. Bull. Soc. philomath. Paris, (7) 12 (4): 143-147. 1890a. Description d@un nouveau genre de mammifere. Bull. Soc. philomath. Paris, 8 (2): 34-38, 1 fig. —— 1890b. Description d’une nouvelle espéce de Lémurien fossile. Bull. Soc. philomath. Paris, (8) 2: 39-40. — 1890c. Description d’un nouveau genre d’insectivore. Bull. Soc. philomath. Paris, (8) 2 (4): 174-175, 3 figs. Fischer, J. B. 1829. Synopsis Mammalium. xlii + 752 pp., 4 tabs. Stuttgart. Fischer, K.-H. 1964. Die tapiroiden Perissodactylen aus der eozanen Braunkohle des Geiseltales. Geologie, Berlin, 45: 1-101, 22 figs, 10 pls. — 1977. Neue Funde von Rhinocerolophiodon (n. gen.), Lophiodon und Hyrachyus (Ceratomorpha, Perissodactyla, Mammalia) aus dem Eozan des Geiseltals bei Halle (DDR). 1 Teil: Rhinocerolophiodon. Z. geol. Wiss., Berlin, 5 (7): 909-919, 5 figs (1977a). 2 Teil: Lophiodon. Loc. cit., 5 (9): 1129-1152, 18 figs (1977b). Fisher, D. C. 1981a. Crocodilian scatology, microvertebrate concentrations, and enamel-less teeth. Paleo- biology, Chicago, 7 (2): 262-275, 4 figs. 1981b. Taphonomic interpretation of enamel-less teeth in the Shotgun local fauna (Paleocene, Wyoming). Contr. Mus. Paleont. Univ. Mich., Ann Arbor, 25 (13): 259-275, 4 figs. Fisher, O. 1862. On the Bracklesham Beds of the Isle of Wight Basin. Q. JI geol. Soc. Lond., 18: 65—94, 4 figs. Fitch, F. J., Hooker, P. J., Miller, J. A. & Brereton, N. R. 1978. Glauconite dating of Palaeocene—Eocene rocks from East Kent and the time scale of Palaeogene volcanism in the North Atlantic region. J. geol. Soc. Lond., 135: 499-512, 14 figs, 2 tabs. Flower, W. H. 1883. On the arrangement of the orders and families of existing Mammalia. Proc. zool. Soc. Lond., 1883: 178-186. — 1884. Catalogue of the specimens illustrating the osteology and dentition of vertebrated animals, recent and extinct, contained in the Museum of the Royal College of Surgeons of England. 2, Mammalia other than man. xliii + 779 pp. London. Fligel, E. 1964. Microproblematica aus den rhatischen Riffkalken der Nordalpen. Palaont. Z., Berlin, 38 (1/2): 74-87, 1 tab., 1 fig., pls 8—9. Flynn, J. J. & Galiano, H. 1982. Phylogeny of early Tertiary Carnivora, with a description of a new species of Protictis from the middle Eocene of northwestern Wyoming. Am. Mus. Novit., New York, 2725: 1-64, 9 figs, 2 tabs. Fogden, M. P. L. 1974. A preliminary field study of the western tarsier, Tarsius bancanus Horsefield. In Martin, R. D., Doyle, G. A. & Walker, A. C., Prosimian Biology: 151-165, 4 figs, 1 tab. London. Fortey, R. A. & Jefferies, R. P. S. 1982. The use of stratigraphy in determining phylogeny, a compromise approach. In Joysey, K. A., Problems of Phylogenetic Reconstruction: 197-234, 17 figs. London, Spec. Vol. Systs Assn. Fowler, K., Edwards, N. & Brett, D. W. 1973. In situ coniferous (taxodiaceous) tree remains in the upper Eocene of southern England. Palaeontology, London, 16 (1): 205-217, pl. 19, 1 fig. Franzen, J.-L. 1968. Revision der Gattung Palaeotherium Cuvier 1804 (Palaeotheriidae, Perissodactyla, Mammalia). 1, 181 pp., 20 figs; 2, 35 pls, 15 tabs. Freiburg, Inaugural-Dissertation zur Erlangung der Doktorwiirde der Naturwissenschaftlich-matematischen Facultat der Albert-Ludwigs-Universitat zu Freiburg 1. Br. —— 1972. Pseudopalaeotherium longirostratum n.g., n.sp. (Perissodactyla, Mammalia) aus dem unterstampischen Kalkmergel von Ronzon (Frankreich). Senckenberg. leth., Frankfurt a.M., 53 (5): BARTONIAN MAMMALS OF HAMPSHIRE BASIN 457 315-331, 8 figs, 2 tabs. —— 1980a. In Franzen, J. L. & Krumbiegel, G., Messelobunodon ceciliensis n. sp. (Mammalia, Artiodactyla), ein neuer Dichobunide aus der mitteleocdnen Fauna des Geiseltales bei Halle (D.D.R.). Z. geol. Wiss., Berlin, 8 (12): 1585-1592, 3 figs, 1 pl. 1980b. Das Skelett eines juvenilen Propalaeotherium isselanum (Mammalia, Equidae) aus dem mitteleozinen Olschiefer der Grube Messel bei Darmstadt. Dortmund. Beitr. Landeskde, 14: 85-102, 1 fig., 5 pls, 2 tabs. — 1981. Das erste Skelett eines Dichobuniden (Mammalia, Artiodactyla) geborgen aus mitteleozanen Olschiefern der ‘Grube Messel’ bei Darmstadt (Deutschland, S-Hessen). Senckenberg. leth., Frankfurt a.M., 61 (3/6): 299-353, 11 figs, 6 tabs, 11 pls. Freshney, E. C. 1978. Ramnor Inclosure Borehole. Jn Institute of Geological Sciences, I.G.S. boreholes 1977. Rep. Inst. geol. Sci., London, 78 (21): 5-6. — & Edwards, R. A. 1983. Christchurch borehole. Jn Institute of Geological Sciences, I.G.S. boreholes 1981. Rep. Inst. geol. Sci., London, 82 (11): 1-3. Gardner, J. S. 1879. Description and correlation of the Bournemouth Beds. Part 1. Upper Marine Series. Q. JI geol. Soc. Lond., 35: 209-228, 6 figs. —— , Keeping, H. & Monckton, H. W. 1888. The Upper Eocene, comprising the Barton and Upper Bagshot Formations. Q. JI geol. Soc. Lond., 44: 578-635. Garimond, S., Remy, J. A. & Sudre, J. 1975. Nouvelles données sur le renouvellement des faunes de mammiferes a l’Eocéne supérieur, d’aprés les gisements de Fons (Gard). Colloques int. Cent. natn. Rech. scient., Paris, Problémes actuels de Paleontologie—Evolution des vertébrés, no. 218: 611-625, 7 figs, 3 tabs. Gaudant, J. 1979. Sur la présence de dents de Characidae (poissons teleéosteens, Ostariophysi) dans les ‘Calcaires a Bithynies’ et les ‘Sables Bleutes’ du Var. Geobios, Lyon, 12 (3): 451—457, 1 fig., 1 pl. Gervais, P. 1849. Recherches sur les mammiferes fossiles des genres Palaeotherium et Lophiodon, et sur les autres animaux de la méme classe que l’on a trouves avec eux dans le midi de la France. C.r. hebd. Seanc. Acad. Sci., Paris, 29: 381-384. — 1850. Nouvelles recherches relatives aux mammiferes d’espéces éteintes qui sont enfouies auprés d’Apt, avec des Paleotheriums identiques a ceux de Paris. C.r. hebd. Seanc. Acad. Sci., Paris, 30: 602-604. —— 1852 (1848-52). Zoologie et Paleontologie frangaises (animaux vertebres) ou Nouvelles Recherches sur les Animaux Vivants et Fossiles de la France. 1st edn. 1 (text), 271 pp.; 2 (pl. expl.); 3 (atlas), 80 pls. Paris. See 1859. —— 1856. Sur les mammiferes fossiles que l’on a recueillis dans le département du Gard. C.r. hebd. Seanc. Acad. Sci., Paris, 42: 1159-1161. — 1859. Zoologie et Paléontologie frangaises ... (&c.). 2nd edn. 1 (text), viii + 544 pp., 51 figs; 2 (atlas), xil + 84 pls. Paris. — 1875. Palaeotherium du calcaire grossier. J. Zool., Paris, 4: 421-422. —— 1876. Zoologie et Paleontologie générales, nouvelles recherches sur les Animaux Vertebreés dont on trouve les ossements enfouis dans le sol et sur leur comparaison avec les espéces actuellement existantes. 2nd edn. 72 pp., 25 pls. Paris. Gilkes, R. J. 1968. Clay mineral provinces in the Tertiary sediments of the Hampshire Basin. Clay Miner., Oxford, 7: 351-361, 3 figs, 1 pl. — 1978. On the clay mineralogy of upper Eocene and Oligocene sediments in the Hampshire Basin. Proc. Geol. Ass. London, 89: 43-56, 7 figs. Gill, T. 1872. Arrangement of the families of mammals with analytical tables. Smithson. misc. Collns, Washington, 11 (230, Art. 1). vi + 98 pp. Gingerich, P. D. 1972. Molar Occlusion and Jaw Mechanics of the Eocene Primate Adapis. Am. J. phys. Anthrop., Washington, 36 (3): 359-368, 8 figs. 1974. Size variability of the teeth in living mammals and the diagnosis of closely related sympatric fossil species. J. Paleont., Tulsa, 48: 895—903. —— 1975a. Systematic position of Plesiadapis. Nature, Lond., 253 (5487): 111-113, 2 figs. —— 1975b. Dentition of Adapis parisiensis and the Evolution of Lemuriform Primates. Jn Tattersall, I. & Sussman, R. W. (eds), Lemur Biology: 65-80, 8 figs. New York. —— 1975c. A new genus of Adapidae (Mammalia, Primates) from the late Eocene of southern France, and its significance for the origin of higher primates. Contr. Mus. Paleont. Univ. Mich., Ann Arbor, 24 (15): 163-170, 3 figs. —— 1976a. Paleontology and phylogeny: patterns of evolution at the species level in early Tertiary mammals. Am. J. Sci., New Haven, 276: 1—28. — 1976b. Cranial anatomy and evolution of early Tertiary Plesiadapidae (Mammalia, Primates). Pap. 458 J. J. HOOKER Paleont. Mus. Paleont. Univ. Mich., Ann Arbor, 15: 1-140, 42 figs, 14 + 17 (appendix) tabs, 12 pls. 1977a. New species of Eocene primates and the phylogeny of European Adapidae. Folia Primatol., Basel, 28: 60-80, 8 figs, 2 tabs. —— 1977b. Radiation of Eocene Adapidae in Europe. Geobios, Lyon, (Mém. spéc.) 1: 165-182, 7 figs, 1 tab. —— 1981. Cranial morphology and adaptations in Eocene Adapidae. 1. Sexual dimorphism in Adapis magnus and Adapis parisiensis. Am. J. phys. Anthrop., Washington, 56: 217-234, 8 figs, 3 tabs. 1983. New Adapisoricidae, Pentacodontidae, and Hyopsodontidae (Mammalia, Insectivora and Condylarthra) from the late Paleocene of Wyoming and Colorado. Contr. Mus. Paleont. Univ. Mich., Ann Arbor, 26 (11): 227-255, 10 figs. — & Martin, R. D. 1981. Cranial morphology and adaptations in Eocene Adapidae. 2. The Cambridge skull of Adapis parisiensis. Am. J. phys. Anthrop., Washington, 56: 235—257, 11 figs, 4 tabs. & Russell, D. E. 1981. Pakicetus inachus, a new archaeocete (Mammalia, Cetacea) from the early— middle Eocene Kuldana Formation of Kohat (Pakistan). Contr. Mus. Paleont. Univ. Mich., Ann Arbor, 25 (11): 235-246, 5 figs. —— —— , Sigogneau-Russell, D. & Hartenberger, J.-L. 1979. Chorlakkia hassani, a new middle Eocene dichobunid (Mammalia, Artiodactyla) from the Kuldana Formation of Kohat (Pakistan). Contr. Mus. Paleont. Univ. Mich., Ann Arbor, 25 (6): 117-124, 2 figs, 1 pl. —— & Sahni, A. 1979. Indraloris and Sivaladapis: Miocene adapid primates from the Siwaliks of India and Pakistan. Nature, Lond., 279 (5712): 415-416, 1 fig., 1 tab. Ginsburg, L., Arques, J., Broin, F. de, Le Calvez, Y., Mouton, J., Obert, D., Prive-Gill, C. & Roucan, J.-P. 1977. Découverte d’une faune de mammiferes dans le Lutétien supérieur de la Défense, a Puteaux, prés Paris (Hauts-de-Seine). C.r. somm. Seanc. Soc. géol. Fr., Paris, 1977 (6): 311-313, 2 figs. , Montenat, C. & Pomerol, C. 1965. Découverte d’une faune de mammiféres terrestres dans les couches marines de I’Auversien (Bartonien inferieur) du Gueépelle (Seine-et-Oise). C.r. hebd. Seanc. Acad. Sci., Paris, 260: 3445-3446. Godinot, M. 1984. Un nouveau genre temoignant de la diversite des Adapinés (Primates, Adapidae) a lEocéne terminal. C.r. hebd. Seanc. Acad. Sci., Paris, (II) 299 (18): 1291-1296, 1 pl. —— 1985. Evolutionary implications of morphological changes in Palaeogene primates. In Cope, J. C. W. & Skelton, P. W. (eds), Evolutionary case histories from the fossil record. Spec. Pap. Palaeont., London, 33: 39-47, 5 figs. Golpe-Posse, J. M. 1972. Faunas de yacimientos con suiformes en el terciario espanol. Paleont. Evoluc., Sabadell, 2: 1-197, appendix 13 pp., 7 pls, 5 figs. Goudie, A. 1973. Duricrusts in Tropical Landscapes. 174 pp., 46 figs. Oxford. Grambast, L. 1962. Apergu sur les charophytes tertiaires du Languedoc et leur signification strati- graphique. C.r. somm. Seanc. Soc. géol. Fr., Paris, 1962 (10): 313-314. —— 1964. Indications fournies par les charophytes pour la stratigraphie du Paleogene. Mem. Bur. Rech. geol. minier., Paris, 28: 1009-1011. —— 1972. Principes de lutilisation stratigraphique des charophytes. Applications au Paléogéne d’Europe occidentale. Mem. Bur. Rech. géol. minier., Paris, 77: 319-328, 1 fig. Grandidier, G. 1904. Un nouveau Lemurien fossile de France, le Pronycticebus gaudryi. Bull. Mus. Hist. nat. Paris, 10; 9-13. Gray, J. E. 1827. Synopsis of the species of the class Mammalia as arranged with reference to their organisation, by Cuvier and other naturalists with specific characters, synonyma etc. In Cuvier, G., The Animal Kingdom arranged in conformity with its organization with additional descriptions of all the species hitherto named, and of many not before noticed, by Edward Griffith and others. 5, Mammalia. 391 pp. London. Gregory, W. K. 1910. The orders of mammals. Bull. Am. Mus. nat. Hist., New York, 27: 1-254. —— 1915. On the classification and phylogeny of the Lemuroidea. Bull. geol. Soc. Am., New York, 26: 426-446. —— 1920. Studies in comparative myology and osteology, no. 5 — On the anatomy of the preorbital fossae of Equidae and other ungulates. Bull. Am. Mus. nat. Hist., New York, 42 (3): 265-284, pl. 18. Gruas-Cavagnetto, C. 1970. Aperg¢u sur la microflore et le microplancton du Paleogene anglais. C.r. somm. Seanc. Soc. géol. Fr., Paris, 1970: 19-21. Guth, C. 1964. A propos de Miacis exilis des Phosphorites du Quercy. Mammalia, Paris, 28 (2): 359-365, pls 15-16, 1 fig. Haeckel, E. 1866. Generelle Morphologie der Organismen. Allgemeine Entwicklungsgeschichte der Organismen. Kritische Grundziige der mechanischen Wissenschaft von den entstehenden formen der Organismen, begriindet durch die Descendenz-Theorie. 2 vols, clx + 462 pp., 8 pls. Berlin. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 459 Halstead, L. B. & Middleton, J. 1972. Notes on fossil whales from the Upper Eocene of Barton, Hamp- shire. Proc. Geol. Ass., London, 83 (2): 185-190, 3 figs. Hardenbol, J. & Berggren, W. A. 1978. A new Paleogene numerical time scale. In Cohee, G. V., Glaessner, M. F. & Hedberg, H. D., Contributions to the geologic time scale. Stud. Geol. Am. Assoc. Pet. Geol., Tulsa, 6: 213-234, 8 figs. Harlan, R. 1834. Notice of fossil bones found in the Tertiary formation of the state of Louisiana. Trans. Am. phil. Soc., Philadelphia, (NS) 4 (3): 397-403, pl. 20. Harland, W. B. et al. 1972. A concise guide to stratigraphical procedure. J. geol. Soc. Lond., 128: 295-305. Harpe, P. de la 1869. Sur la faune du terrain sidérolithique dans le canton de Vaud. Bull. Soc. vaud. Sci. nat., Lausanne, 10: 457-467. & Gaudin, C. 1854. Sur les ossements fossiles trouvés par ces messieurs au Mormont, prés La Sarraz. Bull. Soc. vaud. Sci. nat., Lausanne, 3 (26): 117-128. Hartenberger, J.-L. 1968. Les Pseudosciuridae (Rodentia) de ’Eocéne moyen et le genre Masillamys Tobien. C.r. hebd. Seanc. Acad. Sci., Paris, (D) 267: 1817-1820. —— 1969. Les Pseudosciuridae (Mammalia, Rodentia) de ’Eocéne moyen de Bouxwiller, Egerkingen et Lissieu. Palaeovertebrata, Montpellier, 3 (2): 27-61. — 1971. La systématique des Theridomyoidea. C.r. hebd. Seanc. Acad. Sci., Paris, (D) 273: 1917-1920. —— 1972. Les mammiferes d’Egerkingen et l’histoire des faunes de ’Eocéne d'Europe. Bull. Soc. géol. Fr., Paris, (7) 12 (5): 886-893, 1 fig., 2 tabs. —— 1973. Etude systematique des Theridomyoidea (Rodentia) de ’Eocéne supérieur. Mem. Soc. geéol. Fr., Paris, (NS) 52 (Mem. 117): 1-76, 8 pls, 20 figs. —— 1975. Evolution des rongeurs primitifs de Ancien Monde. Colloques int. Cent. natn. Rech. scient., Paris, Problémes actuels de paleéontologie — évolution des vertébrés, no. 218: 777-790, 1 pl., 1 fig., 2 tabs. — & Louis, P. 1976. Nouveaux rongeurs dans l‘Eocéne supérieur de Grisolles (Aisne). Geobios, Lyon, 9: 81-95. —— & Schmidt-Kittler, N. 1976. Un nouveau genre de Theridomyoidea (Mammalia, Rodentia) dans les Phosphorites du Quercy. Mitt. bayer. St. Palaont. Hist. Geol., Mtinchen, 16: 65—69, 1 pl., 1 fig. —— , Sige, B. & Sudre, J. 1974. La plus ancienne faune de mammiferes du Quercy: Le Bretou. Palaeover- tebrata, Montpellier, 6: 177-196, 8 figs, 5 tabs. Haupt, H. 1956. Beitrag zur Kenntnis der eocanen Arthropodenfauna des Geiseltales. Nova Acta Leopol- dina, Halle a.S., (NF) 18 (128): 3-90, 106 figs. Haupt, O. 1921. Die eocanen Stisswasserablagerungen (Messeler Braunkohlenformation) in der Umgegend von Darmstadt und in ihr palaeontologischer Inhalt. Z. dt. geol. Ges., Berlin, 73: 175-178. 1925. Die Palaohippiden der eocdnen Siisswasserablagerungen von Messel bei Darmstadt. Abh. hess. geol. landesanst., Darmstadt, 6 (4): 1-159, 29 pls. Hay, O. P. 1902. Bibliography and catalogue of the fossil Vertebrata of North America. Bull. U.S. geol. Surv., Washington, 179: i-i11, 1-868. Hay, R. L. & Reeder, R. J. 1978. Calcretes of Olduvai Gorge and the Ndolanya Beds of northern Tanzania. Sedimentology, Amsterdam, New York, 25: 649-673. Hedberg, H. (ed.) 1972. Summary of an international guide to stratigraphic classification, terminology, and usage. Lethaia, Oslo, 5: 297-323. — (ed.) 1976. International Stratigraphic Guide: A Guide to Stratigraphic Classification, Terminology, and Procedure. xvii + 200 pp., 14 figs, 3 tabs. New York, London, Sydney, Toronto. Heissig, K. 1977. Neues Material von Cryptopithecus (Mammalia, Pantolestidae) aus dem Mitteloligozan von “Mohren 13” in Mittelfranken. Mitt. Bayer. Staatssamml. Palaont. & hist. Geol., Miinchen, 17: 213-225, 24 figs. Heller, F. 1930. Die Sdugetierfauna der mitteleozdnen Braunkohle des Geiseltales bei Halle a.S. Jb. halle. Verb. Erforsch. mitteldt. Bodenschatze, (NF) 9: 13-41, 5 pls. — 1934. Anthracobunodon weigelti n.g., n.sp., ein Artiodactyls aus dem Mitteleozan des Geiseltales bei Halle a.S. Palaont. Z., Berlin, 16: 147-263, 1 fig., 2 pls. — 1935. Amphilemur eocaenicus n.g. et n.sp., ein primitiver Primate aus dem Mitteleozan des Geiseltales bei Halle a.S. Nova Acta Leopoldina, Halle a.S., (NF) 2 (3): 293-300, 1 pl., 1 fig. — 1936. Neue Beuteltierreste aus der mitteleozanen Braunkohle des Geiseltales bei Halle (Saale). Nova Acta Leopoldina, Halle a.S., (NF) 4: 311-316, 1 pl. Hensel, R. 1856. Beitrage zur Kenntnis fossiler Sdugetiere. Z. dt. geol. Ges., Berlin, 8: 660—704. Hershkovitz, P. 1971. Basic crown patterns and cusp homologies of mammalian teeth. Jn Dahlberg, A. A. (ed.), Dental Morphology and Evolution: 95-150, 17 figs. Chicago & London. Hill, W. C. O. 1953-55. Primates: Comparative Anatomy and Taxonomy. 1, Strepsirhini. xxiii + 798 pp., 34 460 J. J. HOOKER pls, 199 figs (1953). 2, Haplorhini. 347 pp., 14 pls (1955). Edinburgh. Hooke, R. LeB. 1967. Processes on arid-region alluvial fans. J. Geol., Chicago, 75: 438-460. Hooker, J. J. 1972. The first land mammals from the marine Barton Beds (Upper Eocene) of Hampshire. Proc. Geol. Ass., London, 83: 179-184, 1 tab. 1975. Report of Field Meeting to Hengistbury Head and adjacent areas, Dorset; with an account of published work and some new exposures. Tertiary Times, Bromley, 2 (3): 109-121, 6 figs. —— 1977a. A mammal from the Upper Eocene of Hengistbury, Dorset. Tert. Res., London, 1 (3): 91-94, 1 fig., 1 tab. — 1977b. The Creechbarrow Limestone — its biota and correlation. Tert. Res., London, 1 (4): 139-145. —— 1980. The succession of H yracotherium (Perissodactyla, Mammalia) in the English early Eocene. Bull. Br. Mus. nat. Hist., London, (Geol.) 33: 29-42, 6 figs, 1 tab. (1982). Mammals from the Bartonian (Middle/Late Eocene) of the Hampshire Basin, southern England. PhD Thesis, Univ. London (unpubl). —— 1984. A primitive ceratomorph (Perissodactyla, Mammalia) from the early Tertiary of Europe. Zool. J. Linn. Soc., London, 82: 229-244, 22 figs. —— & Insole, A. N. 1980. The distribution of mammals in the English Palaeogene. In Hooker et al., 1980: 31-45, tab. 2. See next ref. ———— , Moody, R. T. J., Walker, C. A. & Ward., D. J. 1980. The distribution of cartilaginous fish, turtles, birds and mammals in the British Palaeogene. Tert. Res., Rotterdam, 3 (1): 1-45, 2 tabs. & Ward, D. J. 1980. Introduction and stratigraphy and list of localities. In Hooker et al. 1980: 1-12, tab. 1. See previous ref. Hudleston, W. H. 1902-03. Creechbarrow in Purbeck. Geol. Mag., London, NS (4) 9: 241-256, 4 figs (1902a). No. 2. Loc. cit. 10: 149-154, 197-203, pl. 11, 4 figs (1903). —— 1902b. Creechbarrow: an essay in Purbeck geology. Proc. Dorset nat. Hist. antig. Fld Club, Dor- chester, 23: 146-190, 1 pl., 10 figs. Hiurzeler, J. 1946. Gesneropithex nov. gen., ein neuer Primate aus dem Ludien von Gosgen (Solothurn). Verh. schweiz. naturf. Ges., Basel, 126: 126-127. 1947a. Gesneropithex Peyeri nov. gen. nov. spec., ein neuer Primate aus dem Ludien von Gosgen (Solothurn). Eclog. geol. Helv., Lausanne, Basel, 39 (2): 356-361, 4 figs. —— 1947b. Alsaticopithecus Leemanni nov. gen. nov. spec., ein neuer Primate aus dem unteren Lutetien von Buchsweiler im Unterelsass. Verh. schweiz. naturf. Ges., Basel, 127: 81—82. 1948a. Alsaticopithecus Leemanni nov. gen. nov. spec., ein neuer Primate aus dem unteren Luteétien von Buchsweiler im Unterelsass. Eclog. geol. Helv., Lausanne, Basel, 40 (2): 343-356, 12 figs. 1948b. Zur Stammesgeschichte der Necrolemuriden. Abh. schweiz. palaont. Ges., Basel, 66: 1—46, 41 figs, 2 tabs. Huxley, T. H. 1880. On the application of the laws of evolution to the arrangement of the Vertebrata and more particularly of the Mammalia. Proc. zool. Soc. Lond., 1880: 649-662. Illiger, C. 1811. Prodromus systematis Mammalium et Avium additis terminis zoographicis utriusque classis, eorumque versione Germanica. xviii + 301 pp. Berlin. Insole, A. N. (1972). Upper Eocene and Lower Oligocene Mammal Faunas from Southern England. PhD Thesis, Univ. Bristol (unpubl.). International Commission on Zoological Nomenclature 1985. International Code of Zoological Nomencla- ture, 3rd edn. 338 pp. London, Berkeley and Los Angeles. Jaeger, J.-J. 1971. La faune de mammiféres du Lutétien de Bouxwiller (Bas-Rhin) et sa contribution a lélaboration de l’échelle des zones biochronologiques de !Eocéne européen. Bull. Serv. Carte geol. Als. Lorr., Strasbourg, 24 (2/3): 93-105. Jepsen, G. L. 1934. A revision of the American Apatemyidae and the description of a new genus, Sinclairella, from the White River Oligocene of South Dakota. Proc. Am. phil. Soc., Philadelphia, 74: 287-305, figs 1-4, pls 1-3. Jones, F. G. & Wilkinson, B. H. 1978. Structure and growth of lacustrine pisoliths from recent Michigan marl lakes. J. sedim. Petrol., Menasha, 48 (4): 1103-1110, 3 figs. Kay, R. F. & Cartmill, M. 1977. Cranial morphology and adaptations of Palaechthon nacimienti and other Paromomyidae (Plesiadapoidea, ?Primates), with a description of a new genus and species. J. hum. Evol., London, 6: 19-53, 10 figs, 1 pl., 5 tabs. —— & Hylander, W. L. 1978. The dental structure of mammalian folivores with special reference to Primates and Phalangeroidea (Marsupialia). In Montgomery, G. G. (ed.), The Ecology of Arboreal Folivores: 173-191, 15 figs. Washington, D.C., Symp. natn. zool. Park, Smithson. Instn. Keen, M. C. 1977. Ostracod assemblages and the depositional environments of the Headon, Osborne, and Bembridge Beds (upper Eocene) of the Hampshire Basin. Palaeontology, London, 20 (2): 405-445, pls BARTONIAN MAMMALS OF HAMPSHIRE BASIN 461 46-49, 11 figs, 2 tabs. —— 1978. The Tertiary—Palaeogene. In Bate, R. & Robinson, E. (eds), A Stratigraphical Index of British Ostracoda. Geol. J., Liverpool, (Spec. Iss.) 8: 385-450, 8 figs, 12 pls, 5 tabs. Keeping, H. 1887. On the discovery of the Nummulina elegans Zone at Whitecliff Bay, Isle of Wight. Geol. Mag., London, (3) 4: 70—72. — 1910. On the discovery of Bembridge Limestone fossils on Creechbarrow Hill, Isle of Purbeck. Geol. Mag., London, NS (5) 7: 436-439, 1 fig., pl. 34. —— 1912. On the results of the further examination of Creechbarrow Hill. Rep. Br. Ass. Advmt Sci., London, (Dundee) 1912: 129-131. Kellogg, R. 1936. A review of the Archaeoceti. Publs Carnegie Instn, Washington, 482. xv + 366 pp., 88 figs, 37 pls. Kemp, D. J., King, A. D., King, C. & Quayle, W. J. 1979. Stratigraphy and biota of the Elmore Formation (Huntingbridge division, Bracklesham Group) at Lee-on-the-Solent, Gosport, Hampshire. Tert. Res., London, 2 (2): 93-103, 5 figs. Kielan-Jaworowska, Z., Bown, T. M. & Lillegrayen, J. A. 1979. Eutheria. In Lillegraven, J. A., Kielan- Jaworowska, Z. & Clemens, W. A. (eds), Mesozoic Mammals, the first two-thirds of mammalian history: 221-258, 22 figs. Berkeley, Los Angeles, London. Koenigswald, W. y. 1979. Ein Lemurenrest aus dem eozianen Olschiefer der Grube Messel bei Darmstadt. Palaont. Z., Berlin, 53: 63—76, 5 figs. —— 1980a. Das Skelett eines Pantolestiden (Proteutheria, Mammalia) aus dem mittleren Eozan von Messel bei Darmstadt. Palaont. Z., Berlin, 54 (3/4): 267—287, 20 figs, 2 tabs. —— 1980b. Fossillagerstatte Messel — Literaturtibersicht der Forschungsergebnisse aus den Jahren 1969— 1979. Geol. Jb. Hessen, 108: 23-38, 1 fig. —— 1983. Skelettfunde von Kopidodon (Condylarthra, Mammalia) aus dem mitteleozinen Olschiefer von Messel bei Darmstadt. Neues Jb. Geol. Palaont. Abh., Stuttgart, 167 (1): 1-39, 23 figs, 2 tabs. — & Schaarschmidt, F. 1983. Ein Urpferd aus Messel, das Weinbeeren frass. Natur Mus., Frankf., 113 (3): 79-84, 8 figs. — & Storch, G. 1983. Pholidocercus hassiacus, ein Amphilemuride aus dem Eozan der ‘Grube Messel’ bei Darmstadt (Mammalia, Lipotyphla). Senckenberg. leth., Frankfurt a.M., 64 (5/6): 447-495, 27 figs, 3 tabs. Korth, W. W. 1979. Taphonomy of microvertebrate fossil assemblages. Ann. Carneg. Mus., Pittsburgh, 48 (15): 235-285. Kowaleysky, W. 1874. Monographie der Gattung Anthracotherium Cuv. und Versuch einer nattirlichen Classification der fossilen Huftiere (2). Palaeontographica, Cassel, (NF) 2 (4): 211-290, pls 10-12. Kretzoi, M. 1943. Kochictis centennii n.g. n.sp., az egeresi fels6 Oligocénbél. Foldt. Kozl., Budapest, 73: 10-17, pl. 1. (Also in German: Kochictis centennii n.g. n.sp., ein alterttimlicher Creodonte aus dem Oberoligozan Siebenbiirgens. Loc. cit.: 190-195). Krutsch, W. & Lotsch, D. 1958. Ubersicht iiber die palaogeographische Entwicklung des zentraleuropais- chen Alttertiars (ohne Tethys-Raum). Ber. geol. Ges. D.D.R., Berlin, 3 (2/3): 99-110, pls 19-29. —— —— 1964. Propositions a lappui d’une tentative en vue de subdiviser les depots de l’Eocéne supérieur et ceux de l’Oligocéne inférieur et moyen et de mettre en paralléle ces depots d'Europe occidentale entre eux et avec ceux d'Europe centrale, et étude de la position a assigner 4a la limite entre lEoceéne et Oligocene dans ces régions. Mem. Bur. Rech. géol. minier., Paris, 28: 949-963, 1 tab. Kiihne, W. G. 1969. A multituberculate from the Eocene of the London Basin. Proc. geol. Soc., London, 1658: 199-202, | fig. Lapparent, A. de 1906. Traité de Géologie, 5th edn. 1288 pp., 583 figs. Paris. Laurillard, C. L. 1849. Lophiodon. In d’Orbigny, A., Dictionnaire Universel d'Histoire Naturelle, 7: 438-439. Paris. Lavocat, R. 1951. Revision de la faune des mammiferes oligocenes d’Auvergne et du Velay. 153 pp., 26 pls. Paris. —— 1955. Sur un squelette de Pseudosciurus provenant du gisement d’Armissan (Aude). Annls Paleont., Paris, 41: 75-89, 11 figs, pl. 4. Lawson, J. D. 1979. Stability of stratigraphic nomenclature. Newsl. Stratigr., Leiden, 7 (3): 159-165. Leach, W. E. 1814. The Zoological Miscellany; being descriptions of new, or interesting animals, 1. 144 pp., 60 pls. London. Leidy, J. 1856. Notices of the remains of extinct Mammalia discovered by Dr F. V. Hayden in Nebraska Territory. Proc. Acad. nat. Sci. Philad., 8: 88-90. — 1872. Remarks on fossils from Wyoming. Proc. Acad. nat. Sci. Philad., 1872: 277. — 1873. Contributions to the extinct vertebrate fauna of the Western Territories. Rep. U.S. geol. Surv. 462 J. J. HOOKER Territ., Washington, (F. V. Hayden) 1 (1): 1-358, 37 pls. Lemoine, M. (ed.) 1978. Geological Atlas of Alpine Europe and Adjoining Alpine Areas. 584 pp. Amsterdam, Oxford, New York. Lemoine, P. & Abrard, R. 1926. Sur la présence de Palaeotherium magnum Cuvier dans le calcaire de Saint-Ouen. Bull. Soc. géol. Fr., Paris, 26: 3—9, 1 fig., pl. 1. Lemoine, V. 1878. Communication sur les ossements fossiles des terrains tertiaires inferieures des environs de Reims faite a la Société d'Histoire Naturelle de Reims. Bull. Soc. Hist. nat. Reims, 2 (2): 90-113. ——— 1880. Sur les ossements fossiles des terrains tertiaires inférieures des environs de Reims. C.r. Ass. fr. Avance. Sci., Paris, 8 (Montpellier 1879): 585-594. 1891. Etude d’ensemble sur les dents les mammiferes fossiles des environs de Reims. Bull. Soc. geol. Fr., Paris, (3) 19: 253-290. Lewis, G. E. 1933. Preliminary notice of a new genus of lemuroid from the Siwaliks. Am. J. Sci., New Haven, (5) 26: 134-138. Lillegraven, J. A. 1969. Latest Cretaceous mammals of upper part of Edmonton Formation of Alberta, Canada, and review of marsupial—placental dichotomy in mammalian evolution. Paleont. Contr. Univ. Kans., Lawrence, 50 (Vert. 12): 1-122, 53 figs, 21 tabs. Linnaeus, C. 1758. Systema Naturae ... 10th edn, 1. 824 pp. Stockholm. Louis, P. 1976. Gisements de mammiféres bartoniens du Tardenois. Remarques sur la paléogéographie de PAuversien du sud de Fismes. Bull. Inf. Geol. Bassin Paris, 13 (2): 41—S8, 3 figs, 2 tabs. — & Sudre, J. 1975. Nouvelles données sur les Primates de l’Eocéne supérieur européen. Colloques int. Cent. natn. Rech. scient., Paris, Problémes actuels de Paléontologie — Evolution des Vertébrés, no. 218: 805-828, 5 figs, 2 pls. Lydekker, R. 1883. Indian Tertiary and Post-Tertiary Vertebrata. Pt 5 — Siwalik selenodont Suina. Mem. geol. Surv. India Palaeont. indica, Calcutta, (10) 2 (5): 143-177, pls 23-25. 1887. Catalogue of the Fossil Mammalia in the British Museum, 5 (Group Tillodontia, Orders Sirenia, Cetacea, Edentata, Marsupialia, Monotremata and Supplement). xxxv + 345 pp., 55 figs. London. Maack, G. A. 1865. Palaeontologische Untersuchungen tuber noch unbekannte Lophiodonfossilien von Hei- denheim am Hahnenkamme in Mittelfranken nebst einer kritischen Betrachtung sammtlicher bis jetzt bekannten Species des Genus Lophiodon. 76 pp., 14 pls. Leipzig. Machin, J. 1971. Plant microfossils from the Tertiary deposits of the Isle of Wight. New Phytol., London, Cambridge, 70: 851-872. MacIntyre, G. T. 1972. The trisulcate petrosal pattern of mammals. Evol. Biol., New York, 6: 275-303, 8 figs, 1 tab. McKenna, M. C. 1962. Collecting small fossils by washing and screening. Curator, New York, 5 (3): 221-235, 6 figs. —— 1963. Primitive Paleocene and Eocene Apatemyidae (Mammalia, Insectivora) and the Primate— Insectivore boundary. Am. Mus. Novit., New York, 2160: 1—39, 5 figs. Maier, W. 1979. Macrocranion tupaiodon, an adapisoricid(?) insectivore from the Eocene of ‘Grube Messel’ (West Germany). Palaont. Z., Stuttgart, 53 (1/2): 38-62, 8 figs, 1 tab. Major, C. J. F. 1873. Nagertiberreste aus Bohnerzen Suddeutchlands und der Schweiz. Palaeontographica, Cassel, (NF) 2 (2): 75-130. Marsh, O. C. 1871. Notice of some new fossil mammals and birds, from the Tertiary Formation of the West. Am. J. Sci., New Haven, (3) 2: 120-127. — 1872. Preliminary description of new Tertiary mammals. Am. J. Sci., New Haven, 4: 122-128, 202-224. Martini, E. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation. In Farinacci, A. (ed.), Proceedings of the II Planktonic Conference, Roma 1970, 2: 739-785, 4 pls. Rome. Matthes, H. W. 1952. Die Creodontier aus der mitteleozanen Braunkohle des Geiseltales. Hallesches Jb. mitteldt. Erdgesch., Halle, 1 (4): 201-240, pls 16—40, 4 figs. 1977. Die Equiden aus dem Eozan des Geiseltales. 1. Die Zahne. In Matthes, H. W. & Thaler, B., Eozane Wirbeltiere des Geiseltales: 5-39, pls 1-159, 176. (1, text, 149 pp.; 2, atlas, 176 pls). Halle, Martin-Luther Universitat, Halle-Wittenberg. Matthew, W. D. 1909. The Carnivora and Insectivora of the Bridger Basin Middle Eocene. Mem. Am. Mus. nat. Hist., New York, 9: 289-567, pls 42-52, 118 figs. —— 1910. On the osteology and relationships of Paramys and the affinities of the Ischyromyidae. Bull. Am. Mus. nat. Hist., New York, 28: 43-71. —— 1928. The evolution of the mammals in the Eocene. Proc. zool. Soc. Lond., 1927 (4): 947-985. —— 1929a. Critical observations upon Siwalik mammals. Bull. Am. Mus. nat. Hist., New York, 56 (7): 437-560, figs 1-55. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 463 — 1929b. Reclassification of the Artiodactyl families. Bull. geol. Soc. Am., New York, 40: 403-408. — 1929c. Preoccupied names. J. Mammal., Baltimore, 10 (2): 171. — & Granger, W. 1921. New genera of Paleocene mammals. Am. Mus. Novit., New York, 13: 1-7. Matthews, S. C. 1973. Notes on open nomenclature and on synonymy lists. Palaeontology, London, 16: 713-719. Mayer [Eymar], K. 1857. Versuch einer neuen Klassifikation der Tertiar-Gebilde Europa’s. Verh. schweiz. naturf. Ges., Basel, 42: 165-199. 1869. Tableau synchronistique des terrains Tertiaires inferieurs. 4th edn, Zurich. Mayhew, D. 1977. Avian predators as accumulators of fossil mammal material. Boreas, Oslo, 6: 25-31, 3 figs. Mein, P. & Freudenthal, M. 1971. Les Cricetidae (Mammalia, Rodentia) du Neogene Moyen de Vieux- Collonges. 1: Le genre Cricetodon Lartet, 1851. Scr. geol., Leiden, 5: 1-51, 6 pls, 13 figs. Menu, H. & Sigé, B. 1971. Nyctalodontie et myotodontie, importants caractéres de grades évolutifs chez les chiroptéres entomophages. C.r. hebd. Seanc. Acad. Sci., Paris, (D) 272: 1735-1738, 1 pl. Meyer, H. v. 1846. [Neue Saugthiere von Weisenau.] Neues Jb. Miner. Geol. Palaont., Stuttgart, 1846: 473-475. —— 1852. Mittheilungen an Professor Bronn gerichtet. Neues Jb. Miner. Geol. Palaont., Stuttgart, 1852: 831-833. Michaux, J. 1964. Diagnoses de quelques Paramyidés de |’Eocéne inférieur de France. C.r. somm. Seanc. Soc. geol. Fr., Paris, 1964: 153. 1968. Les Paramyidae (Rodentia) de Eocene inférieur du Bassin de Paris. Palaeovertebrata, Mont- pellier, 1 (4): 135-193, 10 pls, 4 figs, 2 tabs. ; Miller, G.S. 1923. The telescoping of the cetacean skull. Smithson. misc. Collns, Washington, 76 (5): 1-72, pls 1-8. — & Gidley, J. W. 1918. Synopsis of the Supergeneric Groups of Rodents. J. Wash. Acad. Sci., 8: 431-448. Misonne, X. 1957. Mammiféres oligocénes de Hoogbutsel et Hoeleden; 1 — Rongeurs et ongulés. Bull. Inst. r. Sci. nat. Belg., Brussels, 33 (51): 1-61. Monty, C. L. V. & Hardie, L. 1976. The geological significance of the fresh water blue-green algal calcareous marsh. In Walter, M. R. (ed.), Stromatolites. Devs Sedimentol., Amsterdam, 20: 447-488. Moody, R. T. J. 1980. Notes on some European Palaeogene Turtles. Tert. Res., Rotterdam, 2 (4): 161-168, 1 fig., 1 tab. Morellet, L. & Morellet, J. 1934. Que faut-il entendre par ‘Bartonien’ Mayer-Eymar 1857? C.r. somm. Seanc. Soc. géol. Fr., Paris, 1934 (5): 73-74. 1948. Le Bartonien du Bassin de Paris. Mem. Serv. Carte geol. det. Fr., Paris. 437 pp. Morlan, R. E. 1980. Taphonomy and archaeology in the upper Pleistocene of the northern Yukon Territory: a glimpse of the peopling of the New World. Pap. archaeol. Surv. Can., Ottawa, (Mercury ser.) 94. xxvii + 398 pp. Morton, A. C. 1982. Heavy minerals of Hampshire Basin Palaeogene strata. Geol. Mag., London, 119 (5): 463-476, 4 figs, 1 pl., 2 tabs. Munier-Chalmas, E. & Lapparent, A. de 1893. Note sur la nomenclature des terrains sédimentaires. Bull. Soc. geol. Fr., Paris, (3) 21: 438-488. Murray, J. W. & Wright, C. A. 1974. Palaeogene Foraminiferida and palaeocology, Hampshire and Paris Basins and the English Channel. Spec. Pap. Palaeont., London, 14: 1-129, 20 pls. Muul, I. & Lim, B. L. 1978. Comparative morphology, food habits, and ecology of some Malaysian arboreal rodents. In Montgomery, G. G. (ed.), The Ecology of Arboreal Folivores: 361—368, 7 figs, 6 tabs. Washington, D.C., Symp. natn. zool. Park, Smithson. Instn. Myannil, R.M. 1966. O vertikalnykh norkakh zaryvaniya v Ordovikskikh izvestiyakakh Pribaltiki. In Gekker, P. F. (ed.), Organizm i sreda v geologischeskom proshlom: 200-207, figs 1-2, pls 1-2. Moskva, Leningrad (Nauka). (Not seen). Nesbit Evans, E. M., Van Couvering, J. A. H. & Andrews, P. 1981. Palaeoecology of Miocene sites in western Kenya. J. hum. Evol., London, 10: 99-116, 11 figs, 1 tab. Niemitz, C. 1977. Zur Funktionsmorphologie und Biometrie der Gattung Tarsius Storr, 1780 (Mammalia, Primates, Tarsiidae). Cour. ForschInst. Senckenberg, Frankfurt a.M., 25: 1-161, 73 figs, 12 tabs. Noulet, J.-B. 1851. Note sur une nouvelle espéce de Lophiodon (Lophiodon lautricense). Mem. Acad. Sci. Inscript. Toulouse, (4) 1: 245-250. —— 1863. Etude sur les fossiles du terrain Eocéne supérieur du Bassin de l’Agout (Tarn). Mem. Acad. Sci. Inscript. Toulouse, (6) 1: 181-206. Novacek, M. J. 1976. Insectivora and Proteutheria of the later Eocene (Uintan) of San Diego County, 464 J. J. HOOKER California. Contr. Sci. Los Angeles, 283. 53 pp., 19 figs, 6 tabs. , Bown, T. M. & Schankler, D. 1985. On the classification of the early Tertiary Erinaceomorpha (Insectivora, Mammalia). Am. Mus. Novit., New York, 2813: 1—22, 11 figs, 1 tab. Odin, G. S. & Curry, D. 1981. L’echelle numérique des temps paléogénes en 1981. C.r. hebd. Seanc. Acad. Sci., Paris, (2) 293 (14): 1003-1006. & Hunziker, J. C. 1978. Radiometric dates from N.W. European glauconites and the Palaeo- gene time scale. J. geol. Soc. Lond., 135: 481—497, 2 figs, 4 tabs. Ollivier-Pierre, M.-F. 1980. Etude palynologique (spores et pollens) de gisements paleogénes du Massif Armoricain. Stratigraphie et Paleogéographie. Mem. Soc. geol. miner. Bretagne, Rennes, 25: 1—239, 33 pls, 29 figs. Osborn, H. F. 1910. The age of Mammals in Europe, Asia and North America. xx + 635 pp., 220 figs. New York. Owen, R. 1841. Description of some fossil remains of Chaeropotamus, Palaeotherium, Anoplotherium and Dichobunes, from the Eocene formation, Isle of Wight. Trans. geol. Soc. Lond., (2) 6 (1): 41-45, pl. 4 1845. Odontography; or a treatise on the comparative anatomy of the teeth; their physiological relations, mode of development, and microscopic structure, in the vertebrate animals, 3: xli-lxxiv, 289-655; pls i, 11, 62A, 90-150. London. —— 1846. A History of British Fossil Mammals and Birds. 560 pp., 236 figs. London. —— 1848a. On the fossil remains of Mammalia referable to the genus Palaeotherium, and to two genera Paloplotherium and Dichodon, hitherto undefined, from the Eocene Sand at Hordle, Hampshire. Q. J] geol. Soc. Lond., 4: 17-42, 6 figs, pls 3-4. 1848b. Description of teeth and portions of jaws of two extinct Anthracotherioid quadrupeds (Hyopotamus vectianus and Hyop. bovinus) discovered by the Marchioness of Hastings in the Eocene deposits on the N.W. coast of the Isle of Wight: with an attempt to develope Cuvier’s idea of the classification of pachyderms by the number of their toes. Q. JI geol. Soc. Lond., 4: 103-141, pls 7-8, 13 figs. — 1857a. On the Dichodon cuspidatus Owen. Q. JI geol. Soc. Lond., 13: 190-196, pl. 3. —— 1857b. Description of the lower jaw and teeth of an Anoplotherioid quadruped (Dichobune ovina, Ow.) of the size of the Xiphodon gracilis, Cuv., from the Upper Eocene Marl, Isle of Wight. Q. JI geol. Soc. Lond., 13: 254-260, pl. 8. Palmer, C. P. 1975. Neogene structures in South-West England. Tertiary Times, Bromley, 2 (4): 178-184, 2 figs. Parker, T. J. & Haswell, W. A. 1897. A Text-book of Zoology, 2. xx + 683 pp. London. Paylow, M. 1900. Etudes sur V’histoire paléontologique des ongulés. 7. Artiodactyles anciens. Bull. Soc. Nat. Moscou, (n.s.) 13: 266—328, 2 pls. Pavoni, N. 1961. Faltung durch Horizontalverschiebung. Eclog. geol. Helv., Lausanne, Basel, 54: 515-534. Pearson, H. S. 1927. On the skulls of early Tertiary Suidae, together with an account of the otic region in some other primitive Artiodactyla. Phil. Trans. R. Soc., London, (B) 215: 389-460, 53 figs. Peterson, O. A. 1919. Report upon the material discovered in the upper Eocene of the Uinta basin by Earl Douglas in the years 1908-1909, and by O. A. Peterson in 1912. Ann. Carneg. Mus., Pittsburgh, 12: 40-168, pls 34-47. Pictet, F.-J. 1857. Seconde Partie. Description des ossements fossiles trouvés au Mauremont. Jn Pictet, F.-J., Gaudin, C. & de la Harpe, P. 1855-57. Memoire sur les Animaux Vertebres trouves dans le terrain siderolithique du canton de Vaud et appartenant a la faune eocéne: 27-120, pls 1-13. Geneva, Matériaux Paléont. Suisse (1) 2. & Humbert, A. 1869. Memoire sur les animaux vertebres trouves dans le terrain sideérolithique du Canton de Vaud et appartenant a la faune eocene. Supplement. Geneva, Matériaux Paléont. Suisse, (5) 2: 121-197, pls 14-28. Pilgrim, G. E. 1941. The dispersal of the Artiodactyla. Biol. Rev., Cambridge, 16: 134-163. Plaziat, J.-C. 1981. Late Cretaceous to late Eocene palaeogeographic evolution of southwest Europe. Palaeogeogr. Palaeoclimat. Palaeoecol., Amsterdam, 36: 263-320, 24 figs, 1 tab. Plint, A. G. 1982. Eocene sedimentation and tectonics in the Hampshire Basin. J. geol. Soc. Lond., 139 (3): 249-254, 3 figs. — 1983. Sandy fluvial point-bar sediments from the Middle Eocene of Dorset, England. Spec. Publs int. Ass. Sedimentol., Oxford, 6: 355—368, 12 figs. 1984. A regressive coastal sequence from the Upper Eocene of Hampshire, southern England. Sedimentology, Oxford, 31: 213-225, 9 figs. Pomel, A. 1847a. Note sur les mammiferes et les reptiles fossiles des terrains €océnes de Paris, inférieurs au dépot gypseux. Archs Sci. phys. nat., Genéve, 4: 326-330. BARTONIAN MAMMALS OF HAMPSHIRE BASIN 465 —— 1847b. Note critique sur le genre Palaeotherium. Bull. Soc. geol. Fr., Paris, (2) 4: 584-587. — 1848. [Sur la classification des mammiféres ongulés]. Bull. Soc. geol. Fr., Paris (2) 5: 256-257. — 1853. Catalogue methodique et descriptif des vertebres fossiles decouverts dans le bassin hydrographique superiur de la Loire et surtout dans la vallee de son affluent principal |’ Allier. 193 pp., Paris. Pomerol, C. 1961. Correlation entre le Lédien du Bassin de Bruxelles et le Lutétien supérieur du Bassin de Paris. C.r. hebd. Seanc. Acad. Sci., Paris, 253: 3839-3841. — 1964. Le Bartonien du Bassin de Paris: Interpretation stratigraphique et essai de corrélation avec les bassins de Belgique et du Hampshire. Mem. Bur. Rech. geéol. minier., Paris, 28: 153-173. —— 1967. Esquisse paléogéographique du Bassin de Paris a l’ere tertiaire et aux temps quaternaires. Revue Geogr. phys. Geol. dyn., Paris, (2) 9 (1): 55-86, 17 figs. — 1973. Stratigraphie et Paleogeographie. Ere Cenozoique (Tertiaire et Quaternaire). 269 pp., 235 figs. Paris. — 1980. Auversien. Marinésien. In Cavelier, C. & Roger, J., Les étages frangais et leurs stratotypes. Mem. Bur. Rech. geéol. minier., Paris, 109: 227-236. Porrenga, D. H. 1967. Glauconite and chamosite as depth indicators in the marine environment. Mar. Geol., Amsterdam, 5: 495-501. Pratt, S. P. 1835. Remarks on the Existence of the Anoplotherium and Palaeotherium in the Lower Freshwater Formation at Binstead, near Ryde, in the Isle of Wight. Trans. geol. Soc. Lond., (2) 3 (3): 451-453, 2 figs. Preece, R. 1976. A note on the terrestial Gastropoda of the Bembridge Limestone (Upper Eocene/Lower Oligocene) of the Isle of Wight, England. Tert. Res., London, 1 (1): 17-19, 1 tab. — 1980a. The Mollusca of the Creechbarrow Limestone Formation (Eocene) of Creechbarrow Hill, Dorset. Tert. Res., Rotterdam, 2 (4): 169-180, 4 pls, 2 figs. —— 1980b. The biostratigraphy and dating of the tufa deposit at the Mesolithic site at Blashenwell, Dorset, England. J. Archaeol. Sci., London, 7: 345-362, 5 figs, 6 tabs. — 1981. The occurrence of Proserpina in the British Tertiary, with the description of a new species (Prosobranchia: Helicinidae). Arch. Molluskenk., Frankfurt a.M., 111 (1/3): 49-54, 3 figs. Prestwich, J. 1846. On the Tertiary or Supracretaceous Formations of the Isle of Wight as exhibited in the sections at Alum Bay and White Cliff Bay. Q. JI geol. Soc. Lond., 2: 223-259. — 1847. On the probable age of the London Clay. Q. JI geol. Soc. Lond., 3: 354-377. —— 1849. On the position and general characters of the strata exhibited in the coast section from Christchurch Harbour to Poole Harbour. Q. JI geol. Soc. Lond., 5: 43-49, 4 figs. — 1857. On the correlation of the Eocene Tertiaries of England, France and Belgium. Part 2 — the Paris Group. Q. JI geol. Soc. Lond., 13: 89-134. Quayle, W. J. 1982. A new Eocene isopod (Crustacea) from the Hampshire Basin. Tert. Res., Rotterdam, 4 (1): 31-34, 1 pl. — & Collins, J. S. H. 1981. New Eocene crabs from the Hampshire Basin. Palaeontology, London, 24: 733-758, pls 104-105, 2 figs. Quinet, G. E. 1965. Un condylarthre de Hoogbutsel. Bull. Inst. r. Sci. nat. Belg., Brussels, 41 (15): 1—5, 1 pl. —— 1968. Les mammiféres du Landénien continental belge; 2, étude de la morpholgie dentaire comparée des ‘carnivores’ de Dormaal. Mem. Inst. r. Sci. nat. Belg., Brussels, 158: 1—64, 7 pls, 4 figs. Radinsky, L. 1964. Paleomoropus, a new early Eocene Chalicothere (Mammalia, Perissodactyla), and a revision of Eocene Chalicotheres. Am. Mus. Novit., New York, 2179: 1—28, 3 figs. Rasmussen, H. W. 1972. Lower Tertiary Crinoidea, Asteroidea and Ophiuroidea from northern Europe and Greenland. Biol. Skr., Copenhagen, 19 (7): 1-83, 14 pls, 5 figs. Ratcliffe, B. C. & Fagerstrom, J. A. 1980. Invertebrate Lebensspuren of Holocene floodplains: their morphology, origin and palaeoecological significance. J. Paleont., Chicago, 54 (3): 614-630, | pl., 4 figs. Read, J. F. 1976. Calcretes and their distinction from stromatolites. In Walter, M. R. (ed.), Stromatolites. Dev. Sedimentol., Amsterdam, 20: 55-71, 3 pls. Reid, C. 1898. The geology of the country around Bournemouth. Mem. geol. Surv. U.K., London, sheet 329. iv + 72 pp. 1902a. The geology of the country around Ringwood. Mem. geol. Surv. U.K., London, sheet 314. iv + 62 pp., 4 figs. 1902b. The geology of the country around Southampton. Mem. geol. Surv. U.K., London, sheet 315. iv + 70 pp., 21 figs. Reineck, H.-E. & Singh, I. B. 1980. Depositional sedimentary environments, with reference to terrigenous clastics. 2nd edn. 549 pp., 683 figs. Berlin, Heidelberg, New York. Remy, J.-A. 1965. Un nouveau genre de Paléothéridé (Perissodactyla) de l'Eocéne supérieur du Midi de la France. C.r. hebd. Seanc. Acad. Sci., Paris, 260: 4362-4364, | fig. 466 J. J. HOOKER —— 1967. Les Palaeotheridae (Perissodactyla) de la faune de mammiféres de Fons 1 (Eocéne supérieur). Palaeovertebrata, Montpellier, 1 (1): 1-46, 20 figs, 8 pls, 12 tabs. —— 1972. Etude du crane de Pachynolophus lavocati n. sp. (Perissodactyla, Palaeotheriidae) des Phos- phorites du Quercy. Palaeovertebrata, Montpellier, 5 (2): 45—78, 14 figs, 5 pls, 3 tabs. —— 1976. Etude comparative des structures dentaires chez les Palaeotheriidae et divers autres perissodactyles fossiles (Thése presentée pour le doctorat de 3éme cycle en sciences odontologiques, Faculte de chirurgie dentaire, Universite Louis Pasteur). 207 pp, 32 figs, 14 pls, 39 tabs. Strasbourg. Revilliod, P. 1917a. Fledermause aus der Braunkohle von Messel bei Darmstadt. Abh. hess. geol. Land- esanst., Darmstadt, 7 (2): 161-201, 1 pl., 18 figs, 4 tabs. 1917b. Contribution a l’étude des chiropteres des terrains tertiaires. Part 1. Abh. schweiz. palaont. Ges., Basel, 42: 1—S7, 13 figs, 1 pl. Rich, T. H. U. 1971. Deltatheridia, Carnivora, and Condylarthra (Mammalia) of the early Eocene, Paris Basin, France. Univ. Calif. Publs geol. Sci., Berkeley, 88: 1-72, 18 figs, 32 tabs. Richard, M. 1946. Contribution a l’étude du bassin d’Aquitaine. Les gisements de mammiferes tertiaires. Mem. Soc. geéol. Fr., Paris, (NS) 24 (52): 1-380, 52 figs. Richter, G. 1981. Untersuchungen zur Ernahrung von Messelobunodon schaeferi (Mammalia, Artiodactyla). Senckenberg. leth., Frankfurt, a.M., 61 (3/6): 355-370, 12 figs. Ride, W. D. L. 1964. A review of the Australian fossil Marsupials. J. Proc. R. Soc. West. Aust., Perth, 47 (4): 97-131, 13 figs, 3 tabs. Robinson, P. 1968. Nyctitheriidae (Mammalia, Insectivora) from the Bridger Formation of Wyoming. Contr. Geol. Univ. Wyo., Laramie, 7 (2): 129-138, pls 1-2. Roman, F. 1904. Contributions a l’étude des bassins lacustres de l’Eocéne et de l’Oligocéne du Languedoc. Bull. Soc. géol. Fr., Paris, (4) 3 (5): 546-616, 22 figs, pls 19-20. Romer, A. S. 1966. Vertebrate Paleontology. ix + 468 pp. Chicago and London. Russell, D. E. 1964. Les mammiféres paléocénes d’Europe. Mem. Mus. natn. Hist. nat., Paris, 13 (C, Sci. terre): 1-321, 16 pls, 73 figs. —— , Godinot, M., Louis, P. & Savage, D. E. 1979. Apatotheria (Mammalia) de l’Eocéne inférieur de France et de Belgique. Bull. Mus. Hist. nat., Paris, (4) 1 (C) (3): 203-243, 4 figs, 5 pls, 2 tabs. —— , Louis, P. & Savage, D. E. 1967. Primates of the French early Eocene. Univ. Calif. publs geol. Sci., Berkeley, 73. vi + 46 pp., 14 figs. 1973. Chiroptera and Dermoptera of the French early Eocene. Univ. Calif: Publs geol. Sci., Berkeley, 95: 1-57, 13 figs. ——-—— 1975. Les Adapisoricidae de l’Eocéne Inférieur de France. Réévaluation des formes considérées affines. Bull. Mus. natn. Hist. nat. Paris, (3) 45 (Sci. terre) (327): 129-193, 7 pls. — & McKenna, M. C. 1962. Etude de Paroxyclaenus, mammifére des Phosphorites du Quercy. Bull. Soc. géol. Fr., Paris, (7) 3: 274-282, 1 fig. & Sige, B. 1970. Revision des chiroptéres Lutétiens de Messel (Hesse, Allemagne). Palaeovertebrata, Montpellier, 3 (4): 83-182, 29 figs, 6 pls, 5 tabs. et al. 1982. Tetrapods of the Northwest European Tertiary Basin. Geol. Jb., Hannover, (A) 60: 1-74. Riitimeyer, L. 1862. Eocaene Saugethiere aus dem Gebiet des schweizerischen Jura. Neue Denkschr. Allg. schweiz. Ges. ges. Naturw., Zurich, 19: 1-98. 1888. Uber einige Beziehungen zwischen den Sdugethierstimmen alter und neuer Welt. Erster Nachtrag zu der eocanen Fauna von Egerkingen. Abh. schweiz. palaont. Ges., Basel, 15: 1-63, 1 pl. —— 1890. Ubersicht der eocinen Fauna von Egerkingen nebst einer Erwiederung an Prof. E. D. Cope. Zweiter Nachtrag zu der Eocanen Fauna von Egerkingen (1862). Abh. schweiz. palaont. Ges., Basel, 17: 1-24. —— 1891. Die Eocane Sdugethier-Welt von Egerkingen. Gesammtdarstellung und dritten Nachtrag zu den ‘Eocanen Saugethieren aus dem Gebiet des schweizerischen Jura’ (1862). Abh. schweiz. palaont. Ges., Basel, 18: 1-153, 8 pls. [N.B. Plates apparently published in 1892]. Saban, R. 1954. Phylogénie des insectivores. Bull. Mus. Hist. nat. Paris, (2) 26: 419-432. Savage, D. E. & Russell, D. E. 1977. Comments on mammalian paleontologic stratigraphy and geochro- nology; Eocene stages and mammal ages of Europe and North America. Geobios, Lyon, (Mém. spec.) 1: 47-56, 2 figs. P. 1965. European Eocene Equidae (Perissodactyla). Univ. Calif. publs geol. Sci., Berkeley, 56. v + 94 pp., 42 figs, 1 pl. 1966. Ceratomorpha and Ancylopoda (Perissodactyla) from the Lower Eocene Paris Basin, France. Univ. Calif. publs geol. Sci., Berkeley, 66. v + 38 pp., 28 figs. Schaub, S. 1958. Simplicidentata (Rodentia). In Piveteau, J., Traite de Paleontologie 6 (2): 659-818. Paris. Schlosser, M. 1884. Die Nager des europdischen Tertiars. Palaeontographica, Cassel, 31 (3): 1-143 (19- BARTONIAN MAMMALS OF HAMPSHIRE BASIN 467 161), pls 1-8 (5-12). —— 1886. Beitrage zur Kentniss der Stammesgeschichte der Huftiere und Versuch einer Systematik der Paar- und Unpaarhufer. Morph. Jb., Leipzig, 12: 1-136, 6 pls. —— 1887-90. Die Affen, Lemuren, Chiropteren, Insectivoren, Marsupialier, Creodonten und Carnivoren des Europaischen Tertiars. 1. Beitr. Palaont. Geol. Ost.-Ung., Wien, Leipzig, 6: 1-224, pls 1-5 (1887). 3. Loe. cit. 8 (1, 2): 1-106, 2 tabs (1890). Schmid, P. 1979. Evidence of Microchoerine Evolution from Dielsdorf (Ziirich Region, Switzerland) — a Preliminary Report. Folia primatol., Basel, 31: 301-311, 5 figs, 2 tabs. Schmidt-Kittler, N. 1970. Ein neue Pseudosciuride von Ehrenstein westlich Ulm. Mitt. bayer. St. Palaont. Hist. Geol., Munich, 10: 433-440. —— 197la. Odontologische Untersuchungen an Pseudosciuriden (Rodentia, Mammalia) des Alttertiars. Abh. bayer. Akad. Wiss., Munich, Math.-naturw. K1., (NF) 150: 1-133, 2 pls, 46 figs, 8 tabs. —— 1971b. Eine unteroligozane Primatenfauna von Ehrenstein bei Ulm. Mitt. bayer. St. Palaont. Hist. Geol., Munich, 11: 171—204, 33 figs, 5 tabs, pl. 13. — 1977a. Neue Primatenfunde aus unteroligozanen Karstspaltenfiillungen Stiddeutschlands. Mitt. bayer. St. Palaont. Hist. Geol., Munich, 17: 177-195. — 1977b. Some aspects of evolution and provincialism of rodent faunas in the European Paleogene. Geobios, Lyon, (Mém. spéc.) 1: 97-106, 3 figs. & Vianey-Liaud, M. 1975. Les relations entre les faunes de rongeurs d’Allemagne du Sud de France pendant l’Oligoceéne. C.r. hebd. Seanc. Acad. Sci., Paris, (D) 281 (9): 511-514, pl. 1. Schroeder, H. 1916. Eozane Saugetierreste aus Nord- und Mitteldeutschland. Jb. preuss. geol. Landesanst. BergAkad., Berlin, 37: 164-195. Schwartz, J. H. & Krishtalka, L. 1976. The lower antemolar teeth of Litolestes ignotus, a late Paleocene erinaceid (Mammalia, Insectivora). Ann. Carneg. Mus., Pittsburgh, 46 (1): 1-6, 4 figs. —— & Tattersall, I. 1985. Evolutionary relationships of living lemurs and lorises (Mammalia, Primates) and their potential affinities with European Eocene Adapidae. Anthrop. Pap. Am. Mus. nat. Hist., New York, 60 (1): 1-100, 40 figs. Scott, W. B. & Jepsen, G. L. 1936. The mammalian fauna of the White River Oligocene — Part 1. Insectivora and Carnivora. Trans. Am. phil. Soc., Philadelphia, (NS) 28 (1): 1-153, 7 figs, 22 pls. —— & Osborn, H. F. 1890. The Mammalia of the Uinta Formation. Trans. Am. phil. Soc., Philadelphia, 16: 461-572. Seeley, H. G. 1876. Notice of the occurrence of remains of a British fossil Zeuglodon (Z. Wanklyni Seeley) in the Barton Clay of the Hampshire Coast. Q. JI geol. Soc. Lond., 32: 428-432. Seligsohn, D. & Szalay, F. S. 1978. Relationship between natural selection and dental morphology: tooth function and diet in Lepilemur and Hapalemur. In Butler, P. M. & Joysey, K. A., Development, Function and Evolution of Teeth: 289-307, 6 figs. London, New York, San Francisco. Shearman, D. J. 1968. On Tertiary fault movements in North Devonshire. Proc. Geol. Ass., London, 78 (4): 555-566, 3 figs, pls 16-17. Shelubsky, M. 1951. Observations on the properties of a toxin produced by Microcystis. Verh. int. Verein. theor. angew. Limnol., Stuttgart, 11: 362-366, 3 tabs. Sige, B. 1975. Insectivores primitifs de ’Eocéne supérieur et Oligocéne inférieur d’Europe occidentale. Apatemyides et Leptictidés. Colloques Int. Cent. natn. Rech. scient., Paris. Problémes actuels de Paléon- tologie — Evolution des Vertébrés, no. 218: 653-672, 9 figs, 2 pls, 2 tabs. — 1976. Insectivores primitifs de Eocéne supérieur et Oligocene inferieur d’Europe occidentale. Nycti- thériidés. Mem. Mus. natn. Hist. nat., Paris, (C, Sci. terre) 34: 1-140, 111 figs, 36 tabs. —— 1978. 8, Insectivores et chiroptéres. In La poche a phosphate de Ste-Néboule (Lot) et sa faune de vertébrés du Ludien supérieur. Palaeovertebrata, Montpellier, 8 (2-4): 243-268, 2 figs, 2 pls, 5 tabs. Simons, E. L. 1961. Notes on some Eocene tarsioids and a revision of some Necrolemurinae. Bull. Br. Mus. nat. Hist., London, (Geol.) 5: 45—69, pls 12-14. 1962. A new Eocene primate genus, Cantius, and a revision of some allied European lemuroids. Bull. Br. Mus. nat. Hist., London, (Geol.) 7: 1-36, 3 pls, 4 figs. 1963. A critical reappraisal of Tertiary primates. Jn Buettner-Janusch (ed.), Evolutionary and genetic biology of primates, 1: 65-129. New York & London. Simpson, G. G. 1927. Mesozoic Mammalia VIII — Genera of Lance mammals other than multi- tuberculates. Am. J. Sci., New Haven, (5) 14 (80): 121-130. — 1928. A new mammalian fauna from the Fort Union of southern Montana. Am. Mus. Novit., New York, 297: 1-15. — 1931. A new classification of mammals. Bull. Am. Mus. nat. Hist., New York, 59: 259-293. — 1940. Studies on the earliest primates. Bull. Am. Mus. nat. Hist., New York, 77: 185-212, 8 figs. 468 J. J. HOOKER —— 1941. A giant rodent from the Oligocene of South Dakota. Am. Mus. Novit., New York, 1149: 1-16. —— 1945. The principles of classification and a classification of mammals. Bull. Am. Mus. nat. Hist., New York, 85. xvi + 350 pp. —— , Roe, A. & Lewontin, R. C. 1960. Quantitative Zoology, revised edn. 440 pp., 64 figs, 16 + v tabs. New York. Sisson, S. & Grossman, J. D. 1953. The Anatomy of the Domestic Animals, 4th edn, rev. 972 pp., 738 figs. Philadelphia & London. Sittler, C. 1969. L’Eocene dans le Fossé rhénan. In Colloque sur l’Eocéne, Paris, Mai 1968, 3. Mem. Bur. Rech. geol. minier., Paris, 69: 371-383, 1 tab. Small, R. J. 1980. The Tertiary geomorphological evolution of south-east England: an alternative inter- pretation. In Jones, D. K. C., The Shaping of Southern England: 49-70, 6 figs. London, &c.; Inst. Br. Geogrs spec. Publ. 11. Smith, J. D., Richter, G. & Storch, G. 1979. Wie Fledermause sich einmal ernahrt haben. Umschau, Frankfurt a.M. and Leipzig, 79 (15): 282-284. Springhorn, R. 1980. Paroodectes feisti, der erste Miacide (Carnivora, Mammalia) aus dem Mittel-Eozan von Messel. Palaont. Z., Berlin, 54 (1/2): 171-198, 10 figs, 2 tabs. 1982. Neue Raubtiere (Mammalia: Creodonta et Carnivora) aus dem Lutetium der Grube Messel (Deutschland). Palaeontographica, Stuttgart, (A) 179: 105-141, 15 pls, 9 text-figs, 5 tabs. Stamp, L. D. 1921. On cycles of sedimentation in the Eocene strata of the Anglo-Franco—Belgian basin. Geol. Mag., London, 58: 146-157. Stehlin, H. G. 1903-16. Die Saugetiere des schweizerischen Eocaens. Critischer Catalog der Materialen. 1. Die Fundorte, die Sammlungen — Chasmotherium — Lophiodon. Abh. schweiz. palaont. Ges., Basel, 30: i-vi, 1-153, figs 1-5, pls 1-3 (1903). 2. Palaeotherium — Plagiolophus — Propalaeotherium. Loc. cit. 31: 155-258, figs 6-13, pls 4~7 (19046). 2 (cont.). Palaeotherium — Plagiolophus — Propalaeotherium. Loc. cit. 32: 259-445, figs 14-30, pls 8-9 (1905a). 3. Lophiotherium — Anchilophus — Pachynolophus. Nachtrage — Schlussbetrachtungen uber die Perissodactylen. Loc. cit. 32: 447-595, figs 31-64, pls 10-11 (19055). 4. Dichobune — Mouillacitherium — Meniscodon — Oxacron. Loc. cit. 33: 597-690, figs 65—98, pl. 12 (1906). 5. Choeropotamus — Cebochoerus — Choeromorus — Haplobunodon — Rhagatherium — Mixtotherium. Loc. cit. 35: 691-837, figs 99-133, pls 13-14 (1908). 6. Catodontherium — Dacrytherium — Leptotheridium — Anoplotherium — Diplobune — Xiphodon — Pseudamphimeryx — Amphimeryx — Dichodon — Haplomeryx — Tapirulus — Gelocus — Nachtrage — Artiodactyla incertae sedis — Schlussbetrachtungen Uber die Artio- dactylen — Nachtrage zu den Perissodactylen. Loc. cit. 36: 839-1164, figs 134-243, pls 15—20 (1910b). 7(1). Adapis. Loc. cit. 38: 1165-1298, figs 244-289 (1912). 7(2). Caenopithecus — Necrolemur — Micro- choerus — Nannopithex — Anchomomys — Periconodon — Amphichiromys — Heterochiromys — Nachtrage zu Adapis — Schlussbetrachtungen zu den Primaten. Loc. cit. 41: 1299-1552, figs 290-371, pls 21-22 (1916). —— 1904a. Sur les mammiféres des sables bartoniens du Castrais. Bull. Soc. géol. Fr., Paris, (4) 4: 445-475, 2 figs, pls 11-12. —— 1904b-08. See 1903-16. — 1910a. Remarques sur les faunules de Mammiferes des couches €océnes et oligocénes du Bassin de Paris. Bull. Soc. geol. Fr., Paris, (4) 9: 488-520. — 1910b-16. See 1903-16. — 1918. Le Pernatherium rugosum P. Gervais. Bull. Soc. géol. Fr., Paris, (4) 18: 123-138. — 1940. Zur Stammesgeschichte der Soriciden. Eclog. geol. Helv., Lausanne, Basel, 33: 298-306, 5 figs. — & Schaub, S. 1951. Die Trigonodontie der simplicidentaten Nager. Abh. schweiz. palaont. Ges., Basel, 67: 1-385. Stinton, F. C. 1970. Field Meeting in the New Forest, Hants. Proc. Geol. Ass., London, 81: 269-274. 1975-80. Fish otoliths from the English Eocene. 258 pp., 16 pls. Palaeontogr. Soc. (Monogr.), London. — & Curry, D. 1979. Lithostratigraphical nomenclature of the English Palaeogene succession. Geol. Mag., London, 116: 66—67. Stock, C. 1934. New Creodonta from the Sespe upper Eocene, California. Proc. natn. Acad. Sci. U.S.A., Washington, 20: 423-427, 1 pl. Storch, G. 1978. Eomanis waldi, ein Schuppentier aus dem Mittel-Eozan der ‘Grube Messel’ bei Darmstadt (Mammalia, Pholidota). Senckenberg. leth., Frankfurt, a.M., 59 (4/6): 503-529, 5 pls, 5 figs. — 1981. Eurotamandua joresi ein Myrmecophagide aus dem Eozan der ‘Grube Messel’ bei Darmstadt (Mammalia, Xenarthra). Senckenberg. leth., Frankfurt, a.M., 61 (3/6): 247-289, 14 figs, 3 pls. Sturm, M. 1978. Maw contents of an Eocene Horse (Propalaeotherium) out of the oil shale of Messel near Darmstadt. In Kvacek, Z. & Schaarschmidt, F. (eds), International Symposium Advances in Angio- BARTONIAN MAMMALS OF HAMPSHIRE BASIN 469 sperm Palaeobotany, Liblice (CSSR), June 13-17, 1977. Cour. ForschInst. Senckenberg, Frankfurt a.M., 30: 120-122. Sudre, J. 1969a. Les gisements de Robiac (Eocene supérieur) et leurs faunes de mammiféres. Palaeoverteb- rata, Montpellier, 2: 95-156, 21 figs. 1969b. Acquisitions réecentes pour la faune des Mammiferes de Robiac. C.r. somm. Seanc. Soc. geéol. Fr., Paris, 1969 (4): 125-126, 1 fig. 1971. Etude de variabilité chez Lophiodon lautricense Noulet. Palaeovertebrata, Montpellier, 4 (3): 67-95, 11 figs, 4 pls, 5 tabs. —— 1972. Révision des artiodactyles de !Eocéne moyen de Lissieu (Rh6ne). Palaeovertebrata, Montpel- lier, 5 (4): 111-156, 17 figs, 7 tabs. — 1973. Un Dichodon géant de La Débruge et une nouvelle interprétation phyletique du genre. Bull. Mus. natn. Hist. nat. Paris, (3) 25 (Sci. terre) (133): 73-78, 1 fig. 1 tab. —— 1974. L’endemisme des faunes de mammiferes du Sud de l’Angleterre a ’Eocéne supérieur. Reun. a. Sci. Terre, Paris, 2: 362. 1978a. 9. Primates et Artiodactyles. In La Poche a phosphate de Ste-Néboule (Lot) et sa faune de vertebrés du Ludien supérieur. Palaeovertebrata, Montpellier, 8 (2-4): 269-290, 5 figs, 6 tabs. 1978b. Les artiodactyles de !Eocéne moyen et supérieur d’Europe occidentale (systematique et évolution). Mem. Trav. Inst. Montpellier, 7: 1-229, 23 figs, 33 pls. —— 1980. Aumelasia gabineaudi n.g. n.sp., nouveau Dichobunidae (Artiodactyla, Mammalia) du gisement d’Aumelas (Herault) d’age Luteétien terminal. Palaeovertebrata, Montpellier, Mem. jub. Lavocat: 197— 211, 3 figs, 2 tabs. Swindler, D. R. 1976. Dentition of Living Primates. xvii + 308 pp., 55 figs, 214 tabs. London, New York & San Francisco. Szalay, F. S. 1968. Origins of the Apatemyidae (Mammalia, Insectivora). Am. Mus. Novit., New York, 2352: 1-11, 4 figs, 1 tab. —— 1969. Mixodectidae, Microsyopidae, and the insectivore—primate transition. Bull. Am. Mus. nat. Hist., New York, 140 (4): 163-330, pls 17-57, 28 figs, 21 tabs. 1971la. The European Adapid Primates Agerina and Pronycticebus. Am. Mus. Novit., New York, 2466: 1-19, 13 figs, 1 tab. — 1971b. Relationships of the alleged primate Gesneropithex peyeri Hiirzeler, 1946. J. Mammal., Balti- more, 52 (4): 824-826, 1 fig. — 1974. New genera of European Adapid Primates. Folia Primatol., Basel, 22 (2—3): 116-133, 10 figs. — 1975. Phylogeny, adaptation and dispersal of the tarsiiform primates. In Luckett, W. P. & Szalay, F. S. (eds), Phylogeny of the Primates: A Multidisciplinary Approach: 357-404, 19 figs. New York. —— 1976. Systematics of the Omomyidae (Tarsiiformes, Primates); taxonomy, phylogeny and adapta- tions. Bull. Am. Mus. nat. Hist., New York, 156 (3): 157-450, 178 figs, 29 tabs. & Delson, E. 1979. Evolutionary History of the Primates. xiv + 580 pp., 263 figs. New York, &c. Tattersall, I. & Schwartz, J. H. 1984. A new genus and species of adapid primate from the middle Eocene of Alsace, and a new genus for ‘Adapis’ ruetimeyeri Stehlin 1912. Folia Primatol., Basel, 41 (3-4): 231-239. Tawney, E. B. 1882. On the Upper Bagshot Sands of Hordwell Cliffs, Hampshire. Proc. Camb. phil. Soc. math. phys. Sci., 4: 140-155. Teilhard de Chardin, P. 1915. Les carnassiers des Phosphorites du Quercy. Annls Paleont., Paris, 9 (3-4): 101-191, pls 12—20, 13 figs, 8 tabs. — 1921. Sur quelques primates des Phosphorites du Quercy. Annls Paleont., Paris, 10: 1—20, 6 figs, pls 1-2. 1922. Les mammiferes de l’Eocéne inférieur frangais et leurs gisements. Annls Paleont., Paris, 11: 1-116, 42 figs, pls 1-8. — 1927. Les mammiféres de l’Eocéne inférieur de la Belgique. Mem. Mus. r. Hist. nat. Belg., Brussels, 36: 1-33, 29 figs, 6 pls, 1 tab. Thaler, L. 1965. Une échelle de zones biochronologiques pour les mammiféres du Tertiaire d'Europe. C.r. somm. Seanc. Soc. géol. Fr., Paris, 1965: 118. 1966. Les rongeurs fossiles du Bas-Languedoc dans leurs rapports avec Vhistoire des faunes et la stratigraphie du Tertiaire d'Europe. Mem. Mus. natn. Hist. nat. Paris, (C, Sci. terre) 17: 1-296, 25 figs, 27 pls, 13 tabs. — 1972. Datation, zonation et mammiferes. Mem. Bur. Rech. géol. minier., Paris, 77: 411—424, 4 tabs. Thomas, H. 1906. Note sur un gisement de Lophiodon a Sergy, au Sud-Est de Fére-en-Tardenois (Aisne). Bull. Soc. géol. Fr., Paris, (4) 6: 686-688. 470 J. J. HOOKER Tobien, H. 1961. Ein Lophiodon-Fund (Tapiroidea, Mamm.) aus den niederhessischen Braunkohlen. Notizbl. hess. Landesamt. Bodenforsch. Wiesbaden, 89: 7-16, 1 fig., 1 pl. 1969. Kopidodon (Condylarthra, Mammalia) aus dem Mitteleozan (Lutetium) von Messel bei Darm- stadt (Hessen). Notizbl. hess. Landesamt. Bodenforsch. Wiesbaden, 97: 7-37, pls 1-3, 7 figs, 9 tabs. —— 1980. Ein anthracotherioider Paarhufer (Artiodactyla, Mammalia) aus dem Eozan von Messel bei Darmstadt (Hessen). Geol. Jb. Hessen, 108: 11—22, 1 fig., 2 pls. Trouessart, E. L. 1879. Catalogue des mammiféres vivants et fossiles. Insectivores. Revue Mag. Zool., Paris, (3) 7: 219-285. —— 1885. Catalogue des mammiferes vivants et fossiles (Carnivores). Bull. Soc. Etud. scient. Angers, 1884 (Suppl.): 1-108. True, F. W. 1908. The fossil cetacean Dorudon serratus Gibbes. Bull. Mus. comp. Zool. Harv., Cambridge, Mass., 52 (4): 65—78, 2 figs, 3 pls. Tullberg, T. 1899. Uber das System der Nagethiere, eine phylogenetische Studie. I-II. Nova Acta R. Soc. Scient. upsal., (3) 18 (1) (sect. 2): 1-328. Vail, P. R. & Hardenbol, J. 1979. Sea-level changes during the Tertiary. Oceanus, Woods Hole, Mass., 22 (3): 71-79, 11 figs. Van Valen, L. 1964. Nature of the supernumerary molars of Otocyon. J. Mammal., Baltimore, 45 (2): 284-286. —— 1965. Paroxyclaenidae, an extinct family of Eurasian mammals. J. Mammal., Baltimore, 46 (3): 388-397, 4 figs, 1 tab. —— 1966. Deltatheridia, a new order of mammals. Bull. Am. Mus. nat. Hist., New York, 132: 1-126, pls 1-8, 17 figs. — 1967. New Paleocene insectivores and insectivore classification. Bull. Am. Mus. nat. Hist., New York, 135 (5): 217-284, pls 6-7, 7 figs, 7 tabs. 1978. The beginning of the Age of Mammals. Evol. Theory, Chicago, 4: 45-80, 4 pls, 7 figs. Vialov, O. S. & Sukatsheva, I. D. 1976. [Fossil caddisfly cases (Insecta, Trichoptera) and their strati- graphical value.] In Kramarenko, N. N., Paleontologiya i Biostratigrafiya Mongolu. Trudy sovmest. sov.-mongol’. paleont. Eksped., Moscow, 3: 169-232, 7 figs, 9 pls. [In Russian]. Vianey-Liaud, M. 1979. Evolution des rongeurs a lOligocéne en Europe occidentale. Palaeontographica, Stuttgart, (A) 166: 136—236, 64 figs, 17 tabs. Viret, J. 1958-61. Perissodactyla. In Piveteau, J., Traite de Paleontologie 6 (2): 368—475, 112 figs (1958). Artiodactyla. Loc. cit. 6 (1): 887-1084, 166 figs (1961). Paris. Wahlert, J. H. 1974. The cranial Foramina of Protrogomorphous Rodents; An Anatomical and Phylogen- etic Study. Bull. Mus. comp. Zool. Harvard, Cambridge, Mass., 146 (8): 363-410, 13 figs. Walker, C. A. 1980. The distribution of birds in the English Palaeogene. In Hooker et al. 1980. Tert. Res., Rotterdam, 3 (1): 25-30. Walker, E. P. 1975. Mammals of the World, 3rd edn. 1500 pp. Baltimore, London. Ward, D. J. 1980. The distribution of sharks, rays and chimaeroids in the English Palaeogene. In Hooker et al. 1980. Tert. Res., Rotterdam, 3 (1): 13-19. Watelet, A. 1864. [Sur des Lophiodons découverts a Jouy (Aisne)]. Bull. Soc. geol. Fr., Paris, (2) 21: 298-300. Weigelt, J. 1933. Neue Primaten aus den mitteleozanen (oberlutetischen) Braunkohle des Geiseltals (geborgen 1932 in der Gruben Cecilie und Leonhardt). Nova Acta Leopoldina, Halle a.S., (NF) 1: 97-156, 11 pls, 3 figs. Weitzel, K. 1933. Kopidodon macrognathus Wittich, ein Raubtier aus dem Mitteleozan von Messel. Notizbl. Ver. Erdk. Darmstadt, (5) 14: 81—88, pl. 7. —— 1949. Neue Wirbeltiere (Rodentia, Insectivora, Testudinata) aus dem Mitteleozan von Messel-bei- Darmstadt. Abh. senckenb. Naturforsch. Ges., Frankfurt a.M., 480: 1—24, 5 pls. West, R. M. 1973. Review of the North American Eocene and Oligocene Apatemyidae (Mammalia, Insectivora). Spec. Publs Mus. Tex. Tech. Univ., Lubbock, 3. 42 pp., 20 figs, 8 tabs. —— 1974. New North American Middle Eocene Nyctithere (Mammalia, Insectivora). J. Paleont., Tulsa, 48 (5): 983-987, 1 pl. Western, D. 1980. Linking the ecology of past and present mammal communities. In Behrensmeyer, A. K. & Hill, A. P. (eds), Fossils in the Making. Vertebrate Taphonomy and Paleoecology: 41—S4, 6 figs. Chicago, London. Whitaker, W. 1910. The water supply of Hampshire (including the Isle of Wight), with records of sinkings and borings. Mem. geol. Surv. U.K., London. vi + 252 pp., 1 pl. 1917. Some Hampshire wells. Pap. Proc. Hamps. Fld Club, Southampton, 8: 40-64. White, H. J. O. 1915. The geology of the country near Lymington and Portsmouth. Explanations of sheets BARTONIAN MAMMALS OF HAMPSHIRE BASIN 471 330 and 331 (mainland). Mem. geol. Surv. U.K., London. v + 78 pp., 15 figs. —— 192la. A short account of the geology of the Isle of Wight. Mem. geol. Surv. U.K., London. v + 219 pp. —— 1921b. Appendix II. Report on sections exposed in excavations for sewers in the parish of Brock- enhurst, Hants. Mem. geol. Surv. Summ. Prog., London, 1920: 81-84. Wilson, J. A. & Szalay, F. S. 1976. New adapid primate of European affinities from Texas. Folia Primatol., Basel, 25: 294-312. 1977. Mahgarita, new name for Margarita Wilson & Szalay, 1976, non Leach 1814 (1819, Lea, 1836, 1838, Lister, 1894). J. Paleont., Tulsa, 51: 643. Wolfe, J. A. 1978. A paleobotanical interpretation of Tertiary climates in the Northern Hemisphere. Am. Scient., New Haven, 66: 694-703. —— 1979. Temperature parameters of humid to mesic forests of eastern Asia and relation to forests of other regions of the northern Hemisphere and Australasia. Prof. Pap. U.S. geol. Surv., Washington, 1106: 1—37, 8 figs. 1980. Tertiary climates and floristic relationships at high latitudes in the Northern Hemisphere. Palaeogeogr. Palaeoclimat. Palaeoecol., Amsterdam, 30: 313-323, 5 figs, 1 tab. Wolff, R. G. 1975. Sampling and sample size in ecological analyses of fossil mammals. Paleobiology, Chicago, 1: 195-204, 5 figs, 1 tab. Wood, A. E. 1959. Rodentia. In McGrew, P. O., The geology and paleontology of the Elk Mountain and Tabernacle Butte area, Wyoming. Bull. Am. Mus. nat. Hist., New York, 117: 157-169, figs 14-27, tabs 7-9. — 1962. The Early Tertiary Rodents of the Family Paramyidae. Trans. Am. phil. Soc., Philadelphia, (NS) 52: 1-261. — 1970. The European Eocene Paramyid Rodent, Plesiarctomys. Verh. naturf. Ges. Basel, 80: (2): 237-278, 15 figs, 5 tabs. 1976a. The Paramyid Rodent Ailuravus from the Middle and late Eocene of Europe, and its relationships. Palaeovertebrata, Montpellier, 7: 118-149, 6 figs, 2 tabs. — 1976b. The Oligocene rodents Ischyromys and Titanotheriomys and the content of the Family Ischyromyidae. In Churcher, C. S. (ed.), Athlon. Essays in Palaeontology in honour of Loris Shano Russell: 244-277, 9 figs. Toronto, R. Ontario Mus. Life Sci. Misc. Publs. 1980. The Oligocene rodents of North America. Trans. Am. phil. Soc., Philadelphia, 70 (5): 1-68, 1 tab. & Patterson, B. 1970. Relationships among hystricognathous and hystricomorphous rodents. Mam- malia, Paris, 34 (4): 628-639. & Wilson, R. W. 1936. A suggested nomenclature for the cusps of the cheek teeth of the rodents. J. Paleont, Menasha, 10 (5): 338-391. Wood, S. V. 1844. Record of the discovery of an alligator with several new mammalia in the freshwater strata at Hordwell. Ann. Mag. nat. Hist., London, (1) 14 (92): 349-351. —— 1846. On the discovery of an Alligator and of several new Mammalia in the Hordwell Cliff, with observations upon the Geological Phenomena of that Locality. Lond. geol. J., 1: 1-7, pls 1-2. Woodburne, M. O. & Bernor, R. L. 1980. On superspecific groups of some Old World hipparionine horses. J. Paleont., Tulsa, 54 (6): 1319-1348, 7 figs. Wortman, J. L. 1901. Studies of Eocene Mammalia in the Marsh Collection, Peabody Museum (contd.) Am. J. Sci., New Haven, (4) 12 (68): 143-154, figs 18-30. Wright, T. 1851. A stratigraphical account of the section of Hordwell, Beacon and Barton Cliffs, on the coast of Hampshire. Ann. Mag. nat. Hist., London, (2) 7: 433-446. Wright, V. P. & Mayall, M. 1981. Organism—sediment interactions in stromatolites: an example from the Upper Triassic of South-West Britain. In Monty, C. (ed.), Phanerozoic Stromatolites: 74-84, 6 figs. Berlin, Heidelberg, New York. Wrigley, A. & Davis, A. G. 1937. The occurrence of Nummulites planulatus in England with a revised correlation of the strata containing it. Proc. Geol. Ass., London, 48 (2): 203-228, figs 18-20, pls 17-18. Young, C. C. 1944. Note on the first Eocene mammal from South China. Am. Mus. Novit., New York, 1268: 1-4, 1 fig. Zdansky, O. 1930. Die alttertiaren Saugetiere Chinas nebst stratigraphischen Bemerkungen. Palaeont. sin., Peking, (C) 6 (2): 1-87, 5 pls. Zeuner, F. E. 1941. The Eocene insects of the Ardtun beds, Isle of Mull, Scotland. Ann. Mag. nat. Hist., London, (11) 7: 82-100. Zittel, K. A. v. 1893. Handbuch der Palaontologie, 4. 799 pp. Munich & Leipzig. 472 J. J. HOOKER Index New taxonomic names and the page numbers of the principal references are in bold type. An asterisk (*) denotes a figure. Abbreviations 195 Acotherulum 192, 215, 389-91, 391*, 393, 395, 399-400, 401*, 402*, 403, 403*, 447 campichii 215, 398*, 399-400, 401*, 403, 415, 420-1, 424-5, 427*, 428, 440, 443 pumilum 399, 402*, 403 quercyi 391*, 393, 399, 403 saturninum 393, 395*, 399, 403, 425 aff. saturninum 393 sp. 399 acritarchs 203 Adapidae 225-6, 265, 267, 443, 445 Adapinae 267, 274, 280*, 282 Adapis 267, 269, 274, 277, 448 laharpei 277, 423 magnus 274-5, 277 var. leenhardti 277-9 aff. magnus 277-8 parisiensis 274, 277 stintoni 274 sudrei 277, 423 cf. sudrei 425 Adapisorex 226 gaudryi 226 Adapisoricidae 224-6, 228-9, 231, 234, 237 Adelomys depereti 301 (Sciuroides ) fontensis 301 Aeluravus 291 Aethomylos 327 Afton 205, 429, 430 Agerina 267 Agerinia 267, 269 Ailuravinae 282-3, 291, 295, 425, 445 Ailuravus 213, 282-3, 291, 293, 295, 440 macrurus 291, 293, 295, 443, 445 michauxi 291, 293 picteti 291, 293-5, 420, 425 remensis 291 stehlinschaubi 213, 287*, 291, 292*, 293-5, 420-1, 423, 425, 428, 443, 445, 448 sp. 295 Albian 437 Alés Basin 422, 425 Alouatta 443 Alsatia 267 dunaifi 271* Alsaticopithecus 223, 225 Alum Bay 200, 203-6, 415-16, 419, 426*, 429-30 Amphichiromys 329 Amphidozotheriinae 239 Amphilemur 223, 225-6, 228-9, 231, 234, 239 eocaenicus 226, 228, 234, 237, 238* leemanni 234, 236, 379 sp. 225, 234 Amphilemuridae 223-5, 228-9, 231, 239, 444 Amphiperatherium 212, 215-16, 217*, 219, 220*, 442 ambiguum 216 bastburgense 216, 221 bourdellense 216, 422 brabantense 216 exile 216 fontense 212, 216, 217*, 218, 220, 220*, 422, 442-3 frequens 216 giselense 216, 219, 422 goethei 216, 219, 220*, 221, 422 aff. goethei 212, 216, 217*, 218-9, 220*, 422, 443 lamandini 216, 219-20, 422 maximum 216 minutum 216, 219, 422 sp. 216, 220 Amphirhagatherium 404 Amphorichnus 435 anagalids 340 Anaptomorphinae 226, 247 Anathana elliotti 229 Anchilophus 342, 352-3, 448, 450 depereti 424 desmarestii 422-3 dumasii 423 cf. dumasii 352 gaudini 423 Anchomomys 192, 223, 225, 267, 269 grisollensis 225 latidens 225 pygmaeus 225 Ancodonta 404 Ancylopoda 374 Androconus 222 Annonaceae 446 Anoplothertidae 382, 409, 447 Anoplotherioidea 409 anteaters 438 Anthracobunodon 404 louisi 404, 424 Anthracotherioidea 382, 404 Antoingt Zone 324 Apatemyidae 327-9, 446, 450 Apatemys 327, 336, 446 Apatotheria 223, 327, 440, 446 Apocynaceae 446 Aquitaine Basin 428 Archaeoceti 342, 432 Archaeonycteris 243 Archygromia durbani 420 - Arcis-le-Ponsart 239, 295, 375, 422, 429 Arctocyonidae 222, 339 Arctogalidea 446 Areoligera undulata 427* Areosphaeridium dictyoplokos 417 multicornutum 427* Argenton 343, 345*, 346 Argile d’Asse 416 Argyrotheca piperipyxis 432 Arisella 274 Armissan 445, 450 Arthropoda 420 Artiodactyla 225, 376, 377*, 414*, 442-3, 447-8, 450 indet. 414 Arvaldus 240 ascophorans 420 asteroids 203 Aubrelong 2 333-4 Aumelasia 376 Auvers 416 Auversian 197, 205, 224, 239, 249, 274, 277, 282, 295, 299, 306, 371, 374-6, 383-4, 399, 409, 415-17, 425, 428-9, 449; lower limit 416 Avenay 291, 329 Bach 391* Bagshot 415 Ballard Down Fault 437 Barnard, P. 435 Barton, Barton Cliff passim, esp. 199-207 Beds 199-200, 205-6, 415-17, 430, 432 Clay 192, 197, 199-200, 201*, 202*, 203-5, 207, 242, 242*, 342, 371, 416-18, 426*, 428-30, 439*, 440, 448 Formation 196*, 197, 198*, 199, 203-6, 344, 356, 373-5, 392, 416-17, 426*, 429-30, 438 Member 206 Formation 200, 203, 206 Sand 200, 203, 205-6 series 200 Bartonian passim, esp. 415-29; cycle 200 Basilosauridae 342 Basilosaurus 214, 342 sp. 214, 342 Bassaricyon 446 Batillaria cf. calcitrapoides 437 Bdeogale 446 Beacon Bunny Beds 205 Cliff 199, 205-6 Beauchamp 416 Becton Bunny 199, 205; Beds 206 member 205 Sand Formation 196*, 197, 199, 201*, 202*, 203, 205-7, 415-17, 426*, 430, 439* Member 206 Belgium 325* Bembridge 361, 362* Limestone 207, 240, 419, 435-6 Marls 361, 433 Bembridgia 210 cincta 420 Bernloch 1A 326, 1B 327 Berville 421-2, 429, 449* Binstead 413 Blackwell, Paddy 342 Blashenwell 436 Bond’s brickyard 211 Bos taurus 393 Boscombe Sands 196*, 197, 199, 204, 426*, 429-30 Bouldnor Cliff 275, 275* Bournemouth 342, 429 Marine Beds 426*, 430 Bouxwiller 216, 220*, 252*, 271*, 273, 287*, 329, 334, 340, 341*, 374, 379 Bovey Basin 437 Bracklesham 204, 415 Beds 197, 200, 204, 416 Group 196*, 204, 206, 438 Braconnac 376 Bransgore 205 brick clay 211 Brockenhurst 206 Bed 417 Buchsweiler 267 Buff marl 210 Bulimus ellipticus 435 Burgmagerbein 2 327 Buxobune 376 Buxolestes 444 piscator 443 Cadurceryx sp. 420 Caenopithecus 267, 274, 444-5 sp. 270 caillasses 416 Calcaire a Potamides aporoschema * de Champigny 426* de Ducy 375, 417, 426* de Fons 220, 330, 422 de St Ouen 256, 417, 422, 425, 426*, 427, 429 Grossier 374, 382, 416, 420, 426* calcretes 436 Calliostoma nodulosum 432 Camelidae 443 canid 446 Capulus 203 Carnivora 212, 327, 336, 338, 438, 442-3 Castrais 288, 428 Castres 288 Catodontherium 410, 450 paquieri 424 robiacense 424 Caylux 359*, 391*, 398*, 399, 401* Cebochoeridae 352, 382, 389-90, 393, 395, 400, 402-4, 415, 447 Cebochoerus 192, 215, 352, 389-93, 391-2*, 395, 398-400, 401*, 402-3, 402-3*, 447 anceps 390 campichii 399 cf. campichii 400 dawsoni 402 fontensis 390-1, 391*, 393, 402-3, 402* aff. fontensis 390-1, 391*, 393, 397 helveticus 390-1, 391—2*, 393, 395, 397, 398* Jaegeri 402 lacustris 390-1, 392*, 393, 395, 397, 398*, 402-3, 402* minor 215, 390-1, 391—2*, 393, 394*, 395, 395—-6*, 397, 398*, 399, 402-3, 402*, 424 robiacensis 215, 390, 391*, 393, 398-400, 398*, 401*, 402-3, 402*, 415, 420, 424, 443 ruetimeyeri 390, 399, 402-3, 402*, 420-1, 427* cf. ruetimeyeri 398 suillus 399, 402 sp. 391 (Gervachoerus ) 390, 399 Cenomanian 437 Cercamonius 267, 269 Cercocebus albigena 443 cercopithecid 226, 447 Cercopithecus pogonias 443 cervid 447 Cetacea 212, 342 Chalk 436 Chama Bed 204-7, 432 Chama squamosa 432 Chamblon 344 Characoidei 420, 433, 434* Chardinyus 329 sp. 331 charophytes 417, 425, 427-8, 427*, 436 Chasmotherium cartieri 424-5 Chattian 339, 376 Chelonia 420, 433 Cherry Marl 212 Chewton Bunny 198*, 199, 342, 392 Chironomidae 433, 434* larval tubes 420, 435 Chiroptera 241, 243 Choeromorus 352, 389-90 helveticus 352 jurensis 352 Choeropotamidae 404, 409 Choeropotamus 215, 409 depereti 427* lautricensis 409, 424, 427* sudrei 427* sp. 215, 409, 443 Chorlakkia 376 Christchurch Bay 199-200, 201-2*, 203, 205-6, 417, 430, 437, 439* chrysochlorids 223 Church Knowle 364 cladogram 269*, 402* Clasby, Mr & Mrs Paul 342, 392 Clausilia 420 Cochlostoma heterostoma 420 Coelodonta 447 Collinson, M. E. 270, 448 Columbomyinae 296 concurrent range zones 419-20 Conde-en-Brie 291 Condylarthra 212, 222, 327, 338, 340, 371 Cooper, J. 340 Coptostylus 435-6 brevis 420, 435 coral 427* Cordieria semicostata 432 correlation 415, 418 cranium 228 Cranston, P. 435 Crassatella tenuisulcata 432 Creech 364 Creechbarrow, passim, esp. 207-12 Bed 12 210 Hill 207, 432 Limestone 193, 207, 212, 240, 371, 428-9, 433, 436-7, 442, 448; Formation, passim creodonts 327, 340, 442 Cresnes 422, 426* Cretaceous 210 cricetid 298-9 Crocodilia 420, 433, 434*, 440 Crossarchus 446 crustaceans 203 Cryptadapis 267, 269, 274 cryptalgal structures 433 Cryptopithecus 222 major 223, 223* sideroolithicus 223 Cuis 329 Curtis, R. J. 356 Cyanophyta 420, 433-37 cyclothems 206 Cynohyaenodon cayluxi 423 Cytheretta cellulosa 427* Dacrytheriinae 410 Dacrytherium 215, 406*, 410, 412*, 447 cayluxi 410 elegans 215, 406*, 410, 412*, 424, 440, 443 cf. elegans 410, 412 ovinum 410, 412, 412* priscum 410 saturnini 410, 412* sp. 410, 412* Dactylopsila 443, 446 Daniels, I.C. 373 Daubentonia 443, 446 Dendrogale melanura 229 Dendrohyrax 443 dental analogues 443 nomenclature 222*, 296*, 343*, 377* depositional environments 430 diacodexids 415 Dibden 196* Dicerorhinus 447 Dichobune 376 Campichii 400 langi 381 leporina 347 robertiana 420-1, 427* spinifera 376 sp. 225 Dichobunidae 225, 376, 381-2, 415, 447 Dichobunoidea 376, 402, 404 Dichobunus 419 leporinus 347 Dichodon 215, 411*, 412-14, 447 cartieri 412 cervinus 412-14, 424 cf. cervinus 215, 411*, 413-14, 443 cuspidatus 412-14 frohnstettensis 412 lugdunensis 412 ruetimeyeri 412 simplex 412 stehlini 412-13 subtilis 412, 424 sp. 215, 411*, 413-14, 443 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 473 Dictulumus 419 leporinus 347 Didelphidae 215, 442 Didelphinae, Didelphini 215, 443 Didelphis 215 Didelphoidea 215 diet 442 dinocerates 340 dinocysts 203, 416-18, 429, 432 Donrussellia 267, 269 Dormaal 338 Dorset Pipe Clay series 207, 426* Dorudontidae 342 Dulcidon 339 Ebro Basin 436 Echinosorex 229 gymnurus 229 Eclépens 290, 309, 310-11*, 350, 409, 449* A 290, 428 B 238*, 239, 256*, 290, 301, 306, 308-9, 311, 412*, 426, 426*, 450 Eclepens-Gare 288, 290, 293, 300-1, 324, 345*, 346-7, 350-2, 400, 410, 412*, 414, 421-2, 424, 5 426*, 428-9, 442, 449* Ecole des Mines 344 Ectinochilus planum 427* Egerkingen 239, 249, 267, 273, 291, 311, 344, 345*, 346-7, 352, 356*, 375, 381, 382*, 383, 385, 389, 409, 412, 418, 422, 425, 426* a 275*, 277, 344, 347, 373, 375, 383, 402, 421, 425, 429 B 402, 421, 424-5, 429 y 271*, 299, 344, 370-1, 383, 385, 388-9, 420-1 Ehingen 1 324, 327 12 326 Ehrenstein 275—-6* 1A 259, 263, 275, 278, 303, 307, 323-4, 327 1B 326-7 Elfomys 448 tobieni 423 Elmore 194, 196*, 197, 199,203-5, 212, 344, 346, 372*, 374-6, 418, 430 Formation 196*, 197, 204-6 Member 199, 344, 374-5, 416-17, 430, 432 England 325*, 429 English China Clays 210-11, 234 Entreroches 260*, 264, 288, 324, 327, 376, 378-9, 409, 414, 450 eocaenum-rhinocerodes Zone 422, 425, Eochiromys 327 Eomoropidae 374 Equidae 361-2, 371 Equoidea 342, 361 Equus 361, 443 Erinaceidae 223-5, 228, 443 erinaceoid 225, 229 Erinaceomorpha 223-4, 228 erinaceotans 226 Erinaceus europaeus 227*, 228 Escamps 450 Estellomys 296-7 Estes, R. 419 Eupleres 446 Europolemur 192, 213, 267, 270, 271*, 273-4, 444 collinsonae 192, 213, 267, 270, 271—2*, 273-4, 423, 443-5 klatti 267, 270, 271*, 273 aff. klatti 267, 271*, 273 Eutheria 212, 222, 340 Euzet 259, 275-6*, 277, 347, 390-1, 391*, 393, 412*, 415 Exilisciurus 445 474 faunal composition 440 provinces 448 Fawley 196*, 205-6, 432; Transmission Tunnel 206 Ferungulata 443 Filholia elliptica 419 laevolonga 420 Florida Everglades 433, 436 Fluvio-marine series 200, 210, 241, 259, 429 Fonliasmes 421, 429 Fons 422, 425, 426* 1 301, 350, 425 2 425 3 425 4 220-1, 220*, 315, 391*, 412*, 425, 450 5 412*, 425 6 392*, 397, 398*, 425 7 425 foraminifera 203, 416, 430, 432; agglutinated 432 Ford, R. L. E. 275 Formation de Mortefontaine 426* France 429 Freshney, E. 206 Friar’s Cliff 203, 430 galericines 228-9, 444 Galerix exilis 226 Gardner, R. 228, 256, 370 Gardonyus 329 gastropods 203, 209-10, 419, 427*, 432-3, 436-7, 447; see slugs, snails etc. Gehlbergschichten 375 Geiseltal 238*, 252-3*, 267, 273, 339, 435 Gentilly 345* Germany 325* Gervachoerus 390-1, 399, 402 campichii 391 fontensis 391 minor 397 Gesneropithex 192, 213, 223-6, 228-9, 230*, 231, 234, 238*, 239, 444 figularis 192, 213, 224, 226, 227*, 230*, 232—3*, 234-5, 238*, 239, 423, 440, 443-4 grisollensis 224, 234, 236-7, 238*, 239, 423, 450 aff. grisollensis 226, 227*, 228, 230*, 236, 238*, 239 latidens 224, 228-9, 230*, 234, 236, sks, 2) peyeri 224-6, 228, 234, 238*, 239 sp. 224-5, 234, 238*, 239, 423 Glans oblonga 432 Gliravus 426, 448, 450 robiacensis 423, 425 aff. robiacensis 425-6 glirids 445 gnaw marks 434* Gosgen Pumpstation 224, 238* Grande Coupure 275, 324, 326, 446 Grauves 329 Grés de Brenne 429 Grisolles 219, 236, 238*, 256-9, 256*, 260*, 261, 263-5, 270*, 297, 404, 421, 424-6, 428-9, 449* Gunville 205 Gypse 415 habitat 442 habits 442 Hampshire Basin 196*, 425, 431*, 439*: sedimentation 429 Hamstead Beds 433, 437 Hapalemur 443, 445 Haplobunodon 215, 404-5, 406*, 407-8, 407*, 411*, 448 lydekkeri 404-5, 407-9, 407* muelleri 404—5, 409 ruetimeyeri 404 J. J. HOOKER solodurense 404-5, 407-9, 420-1, 424 venatorum 192, 215, 404, 405, 406-7*, 408-9, 408*, 411*, 420-1, 424, 428, 440, 443, 447 aff. venatorum 407*, 409 sp. 404 Haplobunodontidae 404, 447 Haplomeryx 383, 412, 415, 448 picteti 424 Hartenberger, J.-L. 295 Headon 308, 433 Beds 199-200, 203, 240, 300, 303, 324, 326-7, 431* Hill 240, 260*, 275-6*, 278*, 284*, 287, 415, 436 Limestone 275 Sands 200, 203, 206, 415-16 Headonian 419-20, 422 hedgehogs 227*, 228, 444 Heidenheim 421—2, 448, 449* Helaletidae 425 Helmstedt 375 Hemiacodon 226 Hengistbury 194, 199, 203-5, 212, 418, 421-2, 429-30, 437, 442, 446, 449* Beds 199, 205, 429 Head 199, 205 Herrlingen | 326 Heteraulacacysta porosa 417 Heterochiromys 329 Heterohyus 214, 327-9, 331, 332*, 333-4, 335*, 336, 446 armatus 328-30, 334 europaeus 328-30, 334 gracilis 328-30 heufelderi 328-30 morinionensis 192, 214, 328-31, 332*, 333-4, 335*, 423, 443, 450 nanus 328-30, 333-4, 335*, 423 aff. nanus 214, 330, 331, 332*, 333-4, 335*, 443 quercyi 328-30, 333-4, 336, 450 sudrei 328-30, 333-4, 335*, 423 cf. sudrei 214, 330, 332*, 333-4, 335*, 443 sp. 328-31, 334 (Chardinyus ) sp. 329, 331 (Gardonyus) 329 heteromyid 298 High Cliff 204, 429; Sands and Clays 429 Highcliff Sand and Clay 205 Sands 199, 430 Hinton Admiral 205 hipparionine 362 Hippopotamidae 443 hipposiderid 243, 245 Hobby, Ken 342 Holocene 436 Homalaxis sp. 432 Homotryblium floripes 427* Hoogbutsel 327 Hordle Cliff 199, 219, 223, 226, 228, 230*, 237, 239, 256-7, 256*, 260*, 261, 263-5, 275, 275—6*, 277-9, 278*, 321, 327, 329, 334, 356*, 364, 364*, 368*, 369-70, 373-4, 404, 407*, 410, 412, 414; Mammal Bed 226, 227*, 238* Hordwell 415 horseshoe concretions 209 Huerzeleris 267 Huntingbridge 196*, 204-5 bed(s) 205 division 196*, 197, 204, 206, 416, 430 Formation 417 Huppersand 249, 252*, 267, 382*, 383, 421 Hyaenodon heberti 423 Hylomys 228-9 suillus 229 Hyperdichobune 215, 376, 379, 381-2 hammeli 376, 379 langi 376, 379, 381 nobilis 376, 379, 381 spectabilis 376, 378-9 spinifera 378-9 sp. 215, 376, 378-9, 380*, 381, 386*, 415, 443, 447 Hyracotherium 293, 344, 371 leporinum 362 siderolithicum 347 Indraloris 267 Indri 443, 447 indriids 445, 447 insects 420, 435, 437; burrows 433, 434*, 435 Insectivora 212, 223, 225, 229, 339 Tomys 443, 445 Ischyromyidae 282-3 Ischyromys 283 Ischyrotomus 445 compressidens 291 Isectolophidae 374 Isoptychus 450 headonensis 422 Issel 343 Issiodoromyinae 296 Jepsenella 327 Jouy 374 Juglandaceae 446 Jumencourt 356*, 365* Kallo 416 Kemp, D. J. 199, 375 Kisselovia clathrata angulosa 427* Klikia vectiensis 420 Kochictis 338-9 Kopidodon 338-40, 446 macrognathus 443 La Bouffie 220*, 221, 329-30, 382 La Deébruge 264, 356*, 364, 365*, 371, 390, 399, 413 La Guittardie 424, 449* La Liviniere 429 La Millette 288 Labidolemur 327 Lacertilia 419-20 Lacey’s Farm Quarry 275, 275*, 324, 326-7 Laguarreés 429 Lamandine 390-1, 391—2*, 393, 394*, 395, 396*, 397, 398*, 410, 412* Lantianius 376 Latilly 429, 449* Lauraceae 446 Lautrec 376, 398*, 399, 429 lautricense-siderolithicum Zone 422, 425-7 Lavergne 221 Le Bretou 256*, 392*, 397, 398*, 421, 424-5, 428-9, 449* Le Castrais 421, 424, 428-9, 449* Le Guépelle 418, 421-2, 428-9 Le Quercy 382 Lee-on-Solent 197, 199, 374-5 Lemuriformes 267, 445 Lemuroidea 443, 446 Leonhardt Mine 267 Leptacodon 239 Leptacotherulum 399 cadurcense 393, 395*, 399 Leptadapis 213, 267, 269, 274-5, 275-6*, 277-9, 278*, 282 capellae 274 magnus 274-5, 275*, 277-9, 278*, 282, 423, 445, 450 var. leenhardti 277, 282 aff. magnus 213, 278-9, 278*, 280-1*, 443, 445 priscus 274, 277 Leptadapis(continued) ruetimeyeri 274, 275-6*, 277, 279, 282 stintoni 274-5, 275—-6*, 277-8, 278*, 2 Leptolophus 343, 425, 428, 448 nouleti 424, 428 stehlini 420-1, 424-5, 428 Leptotheridium 410 lugeoni 424 traguloides 424 Leptotomus bridgerensis 291 costilloi 291 grandis 291 huerfanensis 291 leptodus 291 mytonensis 291 parvus 291 sciuroides 291 Les Pradigues 221 Lignite Bed 200, 206, 275 Ligurian 415 Lipotyphla 223-4, 228-9, 340, 379, 444 Lissieu 249, 252*, 299, 306, 345*, 346, 379, 381, 382*, 383, 418, 421, 429 Litolestes 231, 237 London Basin 196* Clay 199-200, 429 Long Bank 199 Long Mead End 205; Sands 206 Lophiaspis 374 Lophiobunodon 404 Lophiodon 214, 343, 372*, 374-6, 422, 447 buchsowillanum 374 compactus 374 cuvieri 374-6, 420, 422, 425 cf. cuvieri 214, 372*, 374-5 isselense 374-5 lautricense 374, 376, 420-2, 424-5, 427*, 442, 446, 448, 449* cf. lautricense 214, 376, 422, 425, 442 leptorhynchum 374 parisiense 374-5, 420 remensis 374 rhinocerodes 374, 420-2, 427*, 429 sardus 374 tapiroides 375, 424 tapirotherium 374 thomasi 374-5, 420, 422, 424-5 Lophiodontidae 374, 425, 450 Lophiotherium 214, 332*, 342, 347, 350, 352-3, 390, 419 atavum 347 cervulum 347, 350-3, 425, 427* aff. cervulum 350 geiseltalensis 347 magnum 347 progressum 347 pygmaeum 347, 350-2, 420-1, 424, 427* robiacense 345*, 347, 350-2 siderolithicum 214, 332*, 345*, 347. 350, 352-3, 354—-5*, 420-5, 427*, 428, 440, 443, 446, 448, 449* transiens 347 voigti 347 Lower Bagshot Beds 207, 419 Lower Brown Coal 267 Lower Calcaire de Fons 288 Lower Freshwater Formation 206 Lower Hamstead Beds 275 Lower Headon Beds 199-200, 206-7, 219, 257, 260*, 261, 275, 277, 287, 404, 410, 412, 414, 417, 426*, 432, 450 Ludian 192, 199, 205, 207, 212, 219-21, 224, 239-40, 246, 252, 256-9, 263, 267, 274, 278-9, 283, 290, 300-1, 307-8, 311, 315, 322-4, 329-30, 339, 343, 347, 350, 370-1, 373, 376, 382, 384, 390-1, 399, 403-4, 409-10, 412-15, 417, 419, 422, 425, 446, 448, 450 Lutetian 216, 219, 224, 239, 249, 267, 274, 282, 291, 295, 297-9, 311, 329, 339, 342-4, 347, 353, 370-1, 373-6, 382, 384, 390, 403-4, 409-10, 412-13, 415-16, 418-20, 425, 429, 438, 440, 442-3, 445, 447 Lymnaea cf. longiscata 420 Lymnaeidae 436 Macrocranion 228-9, 231 tupaiodon 226, 229 Mahgarita 267, 269 Malperié 329-31, 335* Mammal Bed 239, 256, 256*, 321, 370 Manitsha 283, 445 Manitshinae 282-3, 284*, 290, 442-3, 445 Manitshini 282 Marchwood 196*, 205 Margarita 267 Marinesian 197, 205, 219-20, 249, 252, 257-8, 264, 278-9, 282, 288, 290, 295, 300, 307-8, 315, 329-30, 339, 347, 350, 373-4, 376, 384, 391, 398-400, 404, 409-10, 413-18, 422, 425, 428-9; boundary in England 417 Marls with Pholadomya ludensis 416 Marmosa 443 Marnes a Lucina inornata 417, 426* Marnes a Pholadomya ludensis 415-17, 426*, 428 Marnes d’Euzet 426* Marsupialia 215 Marsupicarnivora 215 Masillabune 404, 447 martini 443 Masillamys krugi 443 Maurimontia 291 picteti 293 Megachiromyoides 291 Megalocochlea pseudoglobosa 420 Megatarsius 267 Melanopsis 435 cf. subfusiformis 437 Meldimys 282 Meliceritites sp. 420 Memerlein 330, 392*, 393, 397, 398* Meniscodon 376 menotyphlan insectivores 229 Mesonychidae 340 Messel 224, 339, 438, 442-4, 446-7; adapid 445 Messelobunodon 376, 415, 447 schaeferi 443, 447 Metadichobune 399 Metatheria 215 Metriotherium 376 Miacidae 336; indet. 336 Miacinae 336, 337*, 338, 443 Miacis 214, 336, 338 exilis 338 sp. 214, 336, 337*, 338, 442-3 Miacoidea 336 Microadapis 267, 269, 274, 282 Microcebus 443-4 Microchiroptera 241, 243, 244*, 245, 428, 443-4 gen. et sp. indet. 242* microchoerid 226 Microchoerinae 195, 245-8, 265, 267, 269-70*, 444 Microchoerus 213, 225, 245, 247, 255*, 256-9, 260*, 261, 263-5, 267, 269, 269-70*, 428, 444, 448, 450; Bed 240 creechbarrowensis 192, 213, 257-8, 260*, 261, 263-5, 266*, 268, 423, 428, 443 BARTONIAN MAMMALS OF HAMPSHIRE BASIN 475 edwardsi 257-9, 260*, 261, 263-5, 268, 444 erinaceus 257-9, 260*, 261, 263-5, 268 edwardsi 259 erinaceus 259 ornatus 257-9, 260*, 263-4 wardi 192, 213, 255*, 257-9, 260*, 261, 262*, 263—S, 268, 423, 428, 443 sp. 257-8, 261, 264-5, 268, 270*, 423 Microsuevosciurus 315 microsyopid 225 Microtubus 435 Mid-Dorset Swell 437-8 Middle Headon Beds 199, 417, 426* Milford-on-sea 199 Milner, A. C. 419 Miocene 216, 267 Mixtotheriidae 381 Mixtotherium 215, 379, 381, 382*, 383, 388-9, 447 cuspidatum 381-2, 384, 388-9 depressum 381-2 gresslyi 381—3, 382*, 385, 388-9, 424 aff. gresslyi 215, 382*, 383-4, 386—7*, 388, 443 cf. gresslyi 382 infans 381-4, 382*, 389, 424 leenhardti 381-3 priscum 381-3, 382*, 389 ef. priscum 382-3 quercyi 381-2 sp. 215, 386*, 389, 442-3 Moiachoerus 389, 399 Monera 420 Mont St Martin, Formation de 416 Montmartre 353, 371, 418 Moraceae 446 Mormont 284, 284*, 287-90, 338, 344, 346-7, 350, 352, 378, 382-3, 382*, 388-9, 391—2*, 398*, 399-400, 407, 407*, 409, 414, 448 Mortefontaine 420* Mouillacitherium 239, 376 aff. elegans 225 cf. elegans 378 Mudeford 203 Mull, Isle of 435 Mutigny 329, 336 Myrtaceae 446 Naish Estate 373 Nannopithex 192, 213, 245, 247, 249, 250*, 251—2, 252—-5*, 257, 265, 267-9, 269*, 444, 448; fold 2228 V 23 filholi 249, 251, 252*, 267-8, 269* pollicaris 249, 251 quaylei 192, 213, 249, 250*, 251, 252-5*, 268, 269*, 423, 443 raabi 249, 251, 252—3*, 268, 269* sp. 213, 249, 250*, 251, 252-4*, 423, 443-4 Nasua 446 Necrolemur 245, 257-9, 265, 267, 269, 269-70*, 444, 448 antiquus 258, 268, 270* filholi 249 parvulus 252 zitteli 258, 268, 270*, 423 cf. zitteli 268 Necrolemurinae 245 Necrosorex 329 Neomatronella 231 Neotetracus 228-9, 443-4 sinensis 229 niveaux reperes 418, 425 Nogent-l’Artaud 422 Nordshausen bei Kassel 256* Nummulites 416, 427*, 429 Nummulites prestwichianus 197, 200, 202*, 203-4, 417-18, 427, 432 rectus 204, 427, 432 variolarius 206, 416 476 Nyctitheriidae 224, 239, 241, 444 Nyctitheriinae 239 Nyctitherium 239 Obergosgen 371 Oligocene 216, 219, 239, 315, 327, 336, 436-7, 445, 450 Oltinomyinae 296 Oltinomys 297 Omomyidae 245, 246-7, 267, 269*, 444 Omomyinae 247 Omomyoidea indet. 241 Omomys 226, 247 oncoliths 209-10, 433, 434*, 435-7 Oodectes 336 Ophidia 419-20 Ophiomorpha 205, 207, 430, 432 Opsiclaenodon 222 Orlinger Tal 315 Osborne Beds 275, 308, 324, 413, 433, 436 ostracods 203, 417, 427*, 436 Otocyon 446 Pachynolophus 342 lavocati 362, 370 livinierensis 362, 370 Palaearctonyx 336 Palaechthon 225, 228 nacimienti 229 Palaeocene 216, 222, 239, 327, 435 Palaeochiropterygoidea 243 Palaeodonta 376 palaeoecology 430, 438 palaeoenvironments 430 palaeogeography 430, 448 Palaeoglandina costellata 420 Palaeomarmota 291 Palaeophyllophora 243 palaeoryctids 327 palaeoryctoids 340 Palaeotheriidae 342, 343*, 353, 371, 446, 450 palaeotherioid 352-3 Palaeotherium 214, 342, 353, 357, 361, 364, 371, 372*, 373-4, 419, 446, 448 castrense 371, 373-4 castrense 424 robiacense 424, 427-8, 427* crassum 371, 373 crusafonti 371 curtum 371, 373 duvalii 371 eocaenum 371, 420-2, 424, 427* isselanum 343 lautricense 371, 420-1, 424 magnum 371, 373 stehlini 374 medium 371, 373 minus 353 muehlbergi 371, 373, 424 praecursum 373 aff. ?muehlbergi 214, 371, 372* pomeli 371, 373, 424 renevieri 371, 373 ruetimeyeri 371, 373, 424 ruetimeyeri 373, 429 siderolithicum 371, 424 tapiroides 374 sp. 214, 371, 372*, 373 Palaeoxestina occlusa 420 Palenochtha 228 Paleomoropus 374 Paloplotherium 353, 361 Paludestrina sp. 437 pangolins 438 Panopea 432 Pantolestidae 222, 223*, 433, 443-4 Pantomimus 222 Paracyathus 427* Paradapis 274 Paradelomys 297-301, 448 crusafonti 297, 301, 423 J. J. HOOKER depereti 297, 300-1 quercyi 297 spelaeus 297 Paralophiodon 374 Paralophiodontinae 374 Paramyidae 282-3, 286*, 291, 292*, 295, 298, 445, 450 Paramyinae 282, 295 Paramys delicatior 291 delicatus 291 francesi 291 Paraplagiolophus 342, 356*, 365* codiciensis 356*, 363—4, 365* Paraxiphodon 412 cournovense 424 Pargny 376 Paris 421, 424, 427-9, 449* Basin 224, 415-17, 420-1, 425, 428-9, 437 Parisian 415 Paroodectes 336 feisti 443 Paroxyclaenidae 338, 340-1, 440, 446, 450 Paroxyclaenus 338, 446 Pelycodus 226 Peratherium 215, 442, 448 bretouense 422 lavergnense 422 sudrei 422 sp. 219 Periconodon 225, 267, 269, 282 Perissodactyla 342, 442, 446 Perriére 256*, 259, 331, 382, 412* Petaurista 443, 445 phalangerid 446 Pholadomya ludensis 416-17 Pholidocercus 223-5, 228, 234, 444 hassiacus 443 Pholidota 438 Phosphorites 282, 330 du Quercy 228, 230*, 252, 274, 278-9, 338, 356*, 361, 363*, 367*, 369, 381, 382*, 395*; see Quercy Pinna 432 Pivetonia 245, 249, 269* isabenae 249, 267 Plagiolophus 214, 343, 353, 356*, 357, 359-60*, 361-3, 363*, 364-5, 364-5*, 367—-8*, 369-71, 425, 428 annectens 353, 356*, 357-8, 361—S*, 368*, 369, 371, 424, 428, 450 aff. annectens 448 cartailhaci 353, 356, 358, 361—2, 424, cartieri 353, 371 curtisi 192, 214, 353, 357, 363, 370-1, 428, 432, 446, 448 creechensis 192, 214, 356*, 358*, 364, 367, 367—-8*, 369, 370*, 371, 424, 427*, 428, 440, 443 curtisi 214, 356, 356*, 358—60*, 362-6*, 367, 368*, 369, 371, 424, 427*, 428 fraasi 353, 356*, 357, 359*, 361-2, 362-3*, 367*, 369-70 javalii 353 lugdunensis 353 minor 356*, 357, 361, 364, 370-1 sp. 356, 356*, 360*, 364, 371 planktonic foraminiferal zones 415-16 Planorbidae 436 Pleistocene 198*, 210, 216, 447 plesiadapiform 228 Plesiarctomys 213, 282-5, 287-8, 290-1, 440, 445 curranti 192, 213, 283-5, 284*, 286-7*, 287-90, 423, 440, 443, 445 cf. curranti 284* gervaisii 283-5, 288-90, 425 hartenbergeri 283—5, 287-9, 287* hurzeleri 213, 283-5, 284*, 287-90, 292*, 420-1, 423, 425, 443, 445, 448 savagei 283-4, 287, 289 spectabilis 283-5, 287, 289, 420 sp. 287*, 288, 290 Plesidacrytherium 410 Pleurocyon 336 Podogymnura 228-9 truei 229, 230* Poitiers 429 Pont d’Assou 274 Pontifactor 239 Poole harbour 429 Potamides aporoschema 426-7* cf. variabilis 437 Poulner 205-6 Premiere Masse du Gypse 353, 371 preservation 438 Priabonian 212, 415, 448 Primates 212, 224-6, 228-9, 231, 245, 444, 447-8 procyonids 446 Pronycticebus 267, 269 sp. 249 Propachynolophus 342 maldani 344 Propalaeotherium 214, 342-4, 347, 371, 446 argentonicum 343 hassiacum 343, 443 helveticum 343 hengyangense 343 isselanum 343 messelense 343-4, 443, 446 parvulum 343-4, 345*, 346, 423, 446 aff. parvulum 214, 343, 345*, 346, 348-9*, 443, 447 cf. parvulum 343 rollinati 343 sinense 343 Propithecus 443, 447 Proserpina cf. woodwardi 420 Prosimii 225, 228-9, 231, 245 Protadelomys 298-9 alsaticus 297, 299 cartieri 299, 311 lugdunensis 299, 306-7 Proteutheria 222-3, 327 Prothryptacodon 222 Protoadapis 267, 269-70, 274 klatti 270 sp. 270 Protodichobune 376, 415 Protrogomorpha 282-3 Pseudamphimer yx pavloviae 424 renevieri 424 schlosseri 420-1 valdensis 424 Pseudoloris 213, 245, 249, 251-2, 254*, 256*, 265, 267, 269, 269*, 444 crusafonti 252, 256, 256*, 268, 269*, 420, 423, 425 cf. crusafonti 213, 254*, 256, 428, 443 parvulus 230*, 249, 256, 268, 423, 450 reguanti 252, 256, 256* sp. 249 Pseudolorisinae 245 Pseudoltinomys 425, 448 mamertensis 423, 425 sp. 425 Pseudopalaeotherium 343 Pseudoparamyinae 282 Pseudoparamys 282 Pseudorhinolophus 245 weithoferi 245 Pseudosciuridae 288, 296, 296*, 297-301, 307-8, 317, 320, 445, 448, 450 Pseudosciurinae 296-7 Pseudosciurus 296, 298-9, 308, 448 suevicus 445 Pteridium aquilinum 208* Ptilocercus 443-4 lowi 229 Pugiodens 338-9 mirus 339 sp. 340 Purbeck 437; limestone 211 Isle of 437 Purgatorius 228 Quayle, W. J. 199 Quercimys 296 Quercy 260*, 265, 270*, 275—-6*, 288, 382-3, 391, 410 phosphorites 228, 267, 275, 301, 385; see Phosphorites du Quercy Quercygale angustidens 423, 450 radiometric dates 418 Ratufa 443, 445 Raynal 274, 278, 278* Reithroparamys debequensis 291 Remyinae 296 Remys minimus 423, 428 Rhagatherium 404 sp. 379 Rhenanian 419 Rhine Graben 448-9 Rhinocerolophiodon 374 rhinoceros 447 Rhinolophus sp. 423 Rhombodinium draco 415-7, 427*, 429 intermedium 417 porosum 417, 427* Rigassi, D. 309 Rissoné, A. 301 Robiac 224-5, 234, 238*, 239, 284*, 288-9, 297, 301, 330, 335*, 345*, 347, 350-1, 353, 373, 398-9, 398*, 413, 418, 421, 424-9, 426*, 440, 442, 449* Robiacina minuta 424, 428 weidmanni 424 Roc de Lunel 421 Roc de Santa 450 Rodent Bed 238*, 239 Rodentia 212, 282, 287*, 295, 440, 443-5, 448 Ronheim 1 324, 326 Rupelian 212 Russellites 338-40 simplicidens 340 sp. 340 Sabellaria 433 Sables a Batillaria bouei 295 dAuvers 295, 418 de Beauchamp 417 de Cresnes 417, 422 d’Ezanville 417, 426* de Monceau 428 de Mortefontaine 417 du Castrais 376, 428 Moyens 416, 420, 422, 426* San Cugat de Gavadons 256* Sands above the Hengistbury Beds 206 Saturninia 239, 241, 448 beata 241, 423 gracilis 241 grandis 241, 423 grisollensis 241, 423 hartenbergeri 241, 423 intermedia 423 mamertensis 422 Scandentia 229 Schelklingen | 326 sciuravid 298 Sciuridae 445 Sciurognathi 282 Sciuroides 214, 297-301, 303, 305-7, 315, 425, 445-6 ehrensteinensis 300, 303, 305-6, 425 fontensis 301 BARTONIAN MAMMALS OF HAMPSHIRE BASIN intermedius 300-1 quercyi 300-1 rissonei 214, 287*, 300-1, 302*, 303, 305-6, 306—7*, 320, 423, 426, 428, 443, 445 ruetimeyeri 300-1 russelli 300-1, 303, 306, 423, 426, 428 aff. russelli 301 siderolithicus 297, 300-1, 303, 305-6, 423, 425, 428 aff. siderolithicus 301, 306 sp. 288-9, 305 Sciuroidinae 296-7, 300 Sciurus 445 ruetimeyeri 301 siderolithicus 300 Scraeva 213, 239-41, 444, 448, 450 hatherwoodensis 240-1 woodi 240-1 sp. 213, 240*, 241, 443 Sea Road Gap 205 Selsey division 197, 199, 204, 206, 416, 432 Sergy 375, 424, 429, 449* Sidérolithique 309, 400 Siluriformes 420 Simamphicyon helveticus 420-1, 423, 428 Sinclairella 327, 446 Sivaladapis 267 slugs 420, 433; plates 209-10 snails 209, 419, 433, 435; opercula 210 Solen 432 solenodontids 223 soricids 223 Sosis 392*, 397, 398* Southampton 206 Souvignargues 390, 392*, 398* Spaniella 338 Sparnacian 216, 267, 338-9 St Loup 347 St Ouen 422; see Calcaire de St Ouen Stampian 212, 267, 274, 298, 315, 325*, 343, 353, 390, 399, 410, 448 Ste Neboule 410, 412* Stehlinella 327 Stone Band 207 stromatolites 433, 435-6 structure 212 Stubbington 197 Studley Wood 196*, 204-5 Suevosciurus 192, 195, 214, 298-300, 303, 310, 315, 316*, 317, 320-3, 325-6*, 327, 445-6, 448, 450 authodon 192, 214, 309, 315, 316*, 317, 318*, 319, 321-2, 322-3*, 327, 423, 440, 443, 445 ehingensis 315, 322-4, 326-7, 326*, 445 fraasi 315, 321, 323-4, 325*, 326-7, 326* minimus 315, 323—4, 325*, 327 aff. minimus 315 mutabilis 303, 428 palustris 315, 324, 325*, 326-7 romani 297, 300-1, 303, 425 cf. romani 428 russelli 300, 303, 425 sp. 315, 321, 450 ( Microsuevosciurus ) 315 ( Treposciurus ) mutabilis 428 ( Treposciurus ) romani 300-1 Suidae 352 suoid artiodactyl 447 superficial deposits 211 Switzerland 325*, 429 Symplocaceae 446 Synaphodus 376 Taddiford Gap 205 talpids 223 Tapirulus 410 schlosseri 424 Tapocyon 336 Tarnomys 296-9 quercinus 297 Tarsilidae 225, 246-7, 444 Tarsiiformes 245, 444—5 tarsioids 448 Tarsius 246-9, 248*, 253*, 256, 443-4 bancanus 246-8, 253*, 445 borneanus 246-8, 248* natunensis 246, 248* spectrum 247-8 syrichta 246-8, 248*, 253* Taylor, P. D. 419 teeth 195, 229; curation of small 194 Teilhardina 247 tenrecids 223 Terebellum 432 fusiforme 432 sopitum 432 Tetonius 247 Thalerimys fordi 427* headonensis 422, 427* Thaumastognathus 404 Theodoxus 437 Theria 215 Theridomyidae 296~7, 299, 446, 450 Theridomyinae 296 Theridomyoidea 295-8 Theridomys euzetensis 425 varleti 423, 425 Titanotheriomys 283 Tornatellaea simulata 432 Totland Bay 206, 415 Totton 205 Toulouse 376 tragulid 447 Treposciurus 192, 214, 298-301, 303, 307-8, 310-11, 315, 316*, 327, 445, 448, 450 helveticus 192, 214, 307-8, 311 helveticus 214, 308, 310—-11* preecei 214, 299, 308-9, 310-11*, 313, 314*, 316*, 327, 423, 443, 445 intermedius 307-11 mutabilis 303, 307-9, 311, 428 helveticus 308 mutabilis 308 romani 300-1 sp. 308 Trichoptera 433, 434*, 435 larval cases 420 Trogopterus 443, 445 Tupaia glis 229 tupaiids 225, 228-9, 444 Turritella 373 elongata 417 turtles 203, 435 Tylopoda 412 Uintacyon 336 Uintan 267 Ulm 307 Ungulata 212, 231, 443 Unio brightoni 420 unionid 437 Upper Bagshot Beds 196*, 197 Upper Bagshot Sand 206 Upper Headon Beds 277 Urogale cylindrura 229 Utrecht 195 Vassacyon 336 Vepricardium porulosum 432 Vespertiliavus gracilis 423 wingei 423 sp. 423 Vincularia 420 Vitis 446 viverrids 446 Viviers 429 Viviparus 435 angulosus 420 cf. lentus 420 477 478 = ‘ulpavoides 192, 214, 338-40, 341*, 446 cooperi 192, 214, 339-40, 341*, 342, 423, 443, 446 germanica 339-40 simplicidens 339-40, 341*, 342 Vulpavus 336, 339 Ward, Mr D. J. and Mrs A. 259 Wareham 331 Warwick Slade 206 Weald—Artois Axis; land bridge 449, 449* J. J. HOOKER Weidenstetten 301, 306, 425 Weidmann, M. 350 Weissenau 216 Weissenburg 2 324 6 323, 327 8 256, 256*, 324 Whitecliff Bay 200, 203-6, 429-30, 432 Wight, Isle of 210, 240, 259, 275, 287, 300, 303, 315, 322-4, 410, 413, 419, 433, 438 W oodipora sp. 420 Xenarthra 438 Xiphodon 412, 448 castrense 420-1, 424, 428 Xiphodontidae 383, 412, 415 Xiphodontoidea 412 Ypresian 207, 246-7, 249, 283, 291, 328-9, 343, 374, 376, 415, 419, 448 Zeuglodon Wanklyni 342 zones, definition 420 Zygorhiza 214, 342 wanklyni 214, 342 Accepted for publication 19 December 1984 NK British Museum (Natural History) The relationships of the palaeoniscid fishes, a review based on new specimens of Mimia and Moythomasia from the Upper Devonian of Western Australia B. H. Gardiner There are some 25,000 species of acinopterygians, far out-numbering the tetrapods and accounting for over half of all the living species of vertebrates. The Actinopterygii contain such diverse forms as polypterids, sturgeons, paddle fishes, gars, the bowfin and the enormous variety of teleosts. The Devonian actinopterygians described in this monograph stand near the base of the Osteichthyes, a natural group including the coelacanths, lungfishes and tetrapods. This unique Upper Devonian fish fauna was discovered in 1963 by Mr Harry Toombs of the British Museum (Natural History) while on a collecting trip to Australia. The majority of the material was subsequently collected in 1967 by a joint expedition from the British Museum (Natural History), the Western Australian Museum and the Hunterian Museum, Glasgow. The discovery of these superbly preserved Devonian palaeoniscids at Gogo, Western Australia, has not only greatly increased our knowledge of the early actinopterygian condition but has also enabled us to appreciate the original osteichthyan plan. In other words, many of the characters of the Gogo palaeoniscids are primitive for all bony fishes, and this in turn allows comparison with the other gnathostome groups. These Devonian species also provide detailed information which illuminates many problems in the structure and evolution of the actinopterygians. This knowledge has permitted a more comprehensive and soundly based account of the history of that _ group. The relationships are expressed in a new classification of gnathostomes in which the placoderms are considered to be the sister-group of the osteichthyans, the acanthodians the sister-group of those two and the chondrichthyans the primitive sister-group of other gnathostomes. The Porolepiformes are considered to be the sister-group of the choanates. Finally the interrelationships of actinopterygians are reviewed and the great range of fossil forms presently included in the Palaeonisciformes discussed. Bull. Br. Mus. nat. Hist. (Geol.) 37(4): 173-428 £39.00 Titles to be published in Volume 39 Upper Cretaceous ammonites from the Calabar region, south-east Nigeria. By P.M. P. Zaborski Cenomanian and Turonian ammonites from the Novo Redondo region, Angola. By M. K. Howarth The systematics and palaeogeography of the Lower Jurassic insects of Dorset, England. By P. E. S. Whalley Ne Nhe a SS a i es ol Mammals from the Bartonian (middle/late Eocene) of the Hampshire Basin, southern England. By J. J. Hooker ee est a i ¥ Typeset by Santype International Ltd., Salisbury, Wilts. Printed by Oxford University Press. Se hs & BOUND 20 JUL 1988 ya ips ahy cle ise tes Bs : i y, ne hs o Of RE Aisa! ANG Ket He ia iy ee i aptre eee Ac Tene Pita te KO Pee panes =< Pay ey ak Ste =