Rhagasostoma (Bryozoa) from the Late Cretaceous of Eurasia: taxonomic revision, stratigraphy and palaeobiogeography

1 Borissiak Paleontological Institute of the Russian Academy of Science, Profsoyuznaya st. 123, Moscow, 117997, Russian Federation. 2 Department of Earth Sciences, Natural History Museum, Cromwell Road, London SW7 5BD, UK. 3 Senckenberg Forschungsinstitute und Naturmuseun, Sektion Marine Evertebraten III (Bryozoologie), Senckenberganlage 25, 60325 Frankfurt am Main, Germany. 4 The Sedgwick Museum of Earth Sciences, Department of Earth Sciences, University of Cambridge, UK.


Introduction
The cheilostome family Onychocellidae Jullien, 1882 includes well over 500 species, as well as numerous post-Cretaceous to Recent species. The huge number (over 200) of species assigned to the type genus Onychocella Jullien, 1882 is in part because the genus has been interpreted broadly and lacks a clear diagnosis (Favorskaya 1985;Gordon & Taylor 1999

Study methods
The specimens from the NHMUK and SM were studied uncoated using a LEO 1455-VP scanning electron microscope (SEM) operating at low vacuum and imaged using backscattered electrons. Specimens from the PIN and TsNIGR Museum were imaged without coating using the backscattered electron detector (BSE) of a Tescan Vega XMU SEM operated at low vacuum (10 Pa) with an accelerating voltage of 30 kV. Specimens from the SMF were coated in gold/palladium (20:80) and imaged with a Camscan CS 24 SEM using secondary electron images.
Measurements were taken from SEM images and are given in microns (μm) as range and number of measurements, in brackets; arithmetic mean ± standard deviation.

Description
Colony usually erect, with fl attened bifoliate branches, fragments 2.0-7.0 mm long by 2.5-5.5 mm wide; encrusting sheet-like colonies, measuring up to 5.0-7.0 mm long by 3.5-7.0 mm wide. Ancestrula and early astogeny not observed. Autozooids variable in shape, often broad, 6-sided and rhomboidal with rounded distal ends, zooidal boundaries raised. Gymnocyst lacking. Cryptocyst extensive, fi nely pustulose, depressed or slightly convex centrally, sometimes with proximal or proximolateral peripheral caverns, up to 0.17 mm long (Fig. 3D-G;andTaylor 1991, 2002). Opesiae terminal, rarely subterminal, semielliptical with narrow shelf in the distal part, formed by two walls, a thin inner wall delimiting the distolateral part and a salient, thickened outer wall delimiting the proximolateral part (Fig. 3J). Proximal edge of opesia straight, smooth, thickened, with outgrowths near the two proximolateral corners delimiting, small round opesiules (outgrowths are often broken, thus giving the opesiules the appearance of opesiular indentations or they may be obscured by sediment infi lls of the opesia). Septula not observed. Ovicells endozooidal, brooding cavity located within the proximal part of the distal zooid (Figs 3C, I, 4A-D); ooecium is formed by the distal zooid, well-recognizable, with cryptocyst-like surface and arch-like proximal edge with elongated proximolateral processes extending along the cryptocyst of the maternal zooid. Avicularia interzooidal, longer than autozooids, elongate (Figs 3C, I-J, 4F). Rostrum channelled, with elevated wing-like walls and pointed apex, conical in outline, asymmetrical, dextral or sinistral. Proximal part rounded, shorter and wider than rostrum. Cryptocyst pustulose, concave, sometimes with proximal peripheral caverns (Fig. 3E). Opesia large, usually roundish, rarely oval, with thickened distolateral edge and with thin articular ridges bearing two short teeth in the proximal margin and a slitlike opesiular indentation between the teeth; opesiules lacking. Kenozooids round, located at edges of colonies ( Fig. 4A, C-D). Cryptocyst fi nely pustulose, peripheral caverns not observed. Opesia roundish, small. Intramural reparative budding of autozooids and avicularia sometimes observed (Fig. 4C, F), all with the same polarity as the host zooid. Closure plates and reparative budding kenozooids not observed.
Rhagasostoma inelegans can be easily distinguished from R. brydonei sp. nov., R. minuens Brydone, 1936 and R. operculatum sp. nov. in having an avicularian rostrum conical in outline rather than spade-shaped or falciform. It further differs from R. brydonei sp. nov. by having endozooidal ovicells instead of immersed ovicells. Rhagasostoma inelegans differs from R. angliae in that the avicularian cryptocyst has a large roundish or oval opesia lacking opesiules, instead of a small subcircular opesia with two opesiules. It differs from R. aralense sp. nov., which also has an avicularian rostrum conical in outline, by the rostrum being asymmetrical instead of symmetrical, and the avicularian opesia roundish rather than egg-shaped.

Etymology
In honour of Reginald Marr Brydone (1873Brydone ( -1943, a prolifi c author of bryozoan species from the English Chalk and collector of the type material of the new species.

Type locality and horizon
United Kingdom, England, Hampshire, south of Alton, Froxfi eld, Kings Lane; Sternotaxis plana Zone, late Turonian.

Remarks
We have restudied Brydone's examples of Rhagasostoma inelegans in the SM collection. The specimen identifi ed by Brydone as this species (Brydone 1930: pl. 25, fi gs 11, 12) does not belong to Lonsdale's species and is here regarded as a new species, R. brydonei sp. nov.
The new species differs from the closely related species Rhagasostoma inelegans, R. minuens, R. angliae and R. aralense sp. nov. in the avicularian rostrum being spade-shaped rather than having a conical outline shape or falciform. Moreover, the ovicells of R. brydonei sp. nov. are immersed. Rhagasostoma brydonei sp. nov. differs from R. angliae in having an avicularian cryptocyst with large oval or roundish opesia lacking opesiules, instead of a small subcircular opesia and two opesiules. The new species differs from R. operculatum sp. nov., which has a similar spade-shaped avicularian rostrum, by the rostrum being narrowed at the base, enlarged centrally and with a pointed tip, instead of being the same width along almost its entire length and having a pointed or rounded tip.

Description
Colony encrusting, sheet-like, fragments up to 10 mm in diameter, or erect with fl attened, bifoliate branches, fragments up to 6.0-7.0 mm long by 3.0-5.0 mm wide. Ancestrula ( Fig. 8B) about 0.33 mm in diameter, rounded rhombic, surrounded by six periancestrular zooids, fi ve zooids possibly budded directly from the ancestrula. Autozooids variable in shape, often broad and roughly rectangular with rounded distal ends; zooidal boundaries raised. Gymnocyst lacking. Cryptocyst extensive, fi nely pustulose, depressed or slightly convex centrally, sometimes with peripheral cavern surrounding the cryptocyst (Fig. 8C). Opesia terminal or subterminal, semielliptical with shelf distally, formed by two walls, a thin inner wall delimiting the distolateral part and a salient, thickened outer wall delimiting the proximolateral part (Figs 8E, H, 9C). Proximal edge of opesiae straight, smooth, thickened, with outgrowths near the two proximolateral corners delimiting small, roundish opesiules (outgrowths are often broken, giving the opesiules the appearance of opesiular indentations, or they may be obscured by sediment fi llings inside the opesia). Septula not observed. Ovicells endozooidal, ooecium is formed by the distal zooid, ill-recognizable, with cryptocyst-like surface (Fig. 9E). Avicularia interzooidal, longer than autozooids, elongate. Rostrum channeled with elevated wing-like walls and pointed tip, conical in outline or sometimes falciform, asymmetrical, dextral or sinistral. Proximal part rounded, shorter and slightly wider than the rostrum. Cryptocyst pustulose, concave with depression centrally, sometimes with peripheral cavern surrounding the cryptocyst (Fig. 8C, E, H), having three openings: a distal small subcircular opesia; two lateral and parallel slit-like opesiules formed by long teeth of articular ridges, which grow together with proximal edge of opesia; a long and slit-like opesiular indentation proximally between the teeth (Figs 8E, H, 9C, E, G). Cryptocyst between openings frequently destroyed, causing the openings to coalesce into a single opening of variable outline that is sometimes tulip-shaped. Kenozooids rounded, very rare, located at the margins of a colony or between zooids (Fig. 9D). Cryptocyst fi nely pustulose. Opesia roundish, small. Closure plates, intramural reparative budding in autozooids and avicularia not observed. Brydone (1936) introduced the subspecies Rhagasostoma inelegans angliae for unilaminar colonies lacking ovicells. We have restudied Brydone's syntypes in the SM collection, choosing a lectotype (the specimen fi gured by Brydone 1936: pl. 35, fi g. 4). The specimen fi gured by Brydone (1930: pl. 26, fi g. 5) was not restudied. Brydone (1936) compared his specimens of Rhagasostoma inelegans angliae from the Trimingham Chalk with the bilaminar, ovicellate Onychocella dichotoma sensu Levinsen (1925) from the early Maastrichtian white Chalk of southeastern Denmark and northern Jylland. However, Levinsen's drawing shows that his species has an avicularian opesiae similar to Rhagasostoma inelegans incarcerata sensu Brydone (1930). The avicularia of this subspecies apparently contain two pairs of lateral, parallel, slit-    Brydone, 1936. For each parameter the range is given with the number of measurements in brackets. The arithmetic mean is given ± standard deviation. All measurements in μm. like opesiules in the central part of the cryptocyst, whereas angliae has only one pair of lateral, parallel, slit-like opesiules. Voigt (1949: 26) mentioned one specimen from the late Campanian (Belemnitella lanceolata Zone) of Hemmoor. Our study shows that the avicularian opesia in this specimen are similar to angliae, while Voigt (1949)  Rhagasostoma angliae can easily be distinguished from R. inelegans, R. brydonei sp. nov., R. minuens, R. aralense sp. nov. and R. operculatum sp. nov. as the avicularian cryptocysts have small, subcircular opesia and two opesiules instead of large, roundish or oval opesia without opesiules. Rhagasostoma rowei  and R. mimosa (Brydone, 1930) also have avicularian cryptocysts with small subcircular opesia and opesiules, but R. angliae differs from these species in having articular ridges bearing teeth with proximal opesiular indentation between the teeth and interzooidal, and elongate avicularia with asymmetrical rostra instead of vicarious, rhomboidal avicularia with symmetrical rostra.

Etymology
The species is named after its type locality, the region south of the former Aral Sea in the Republic of Karakalpakstan, Uzbekistan.    (2) 140.00 ± 14.14 OpD 20-30 (2) 25.00 ± 7.07

Description
Colony erect, fl attened, bifoliate, fragments 3.0-5.0 mm long by 2.0-3.0 mm wide. Ancestrula and early astogeny not observed. Autozooids variable in shape, often broad and subrectangular with rounded distal ends; zooidal boundaries raised. Gymnocyst lacking. Cryptocyst extensive, fi nely pustulose, convex centrally, slightly depressed around edges, with proximolateral peripheral caverns; lateral wall of some caverns ribbed (Fig. 10B, E-G). Opesia terminal, semielliptical without shelf in the distal part, formed by two walls, a thin inner wall delimiting the distolateral part and a salient, thickened outer wall delimiting the proximolateral part (Fig. 10C, F). Proximal edge of opesia straight, crenulated, thickened, with outgrowths near the two proximolateral corners delimiting small, roundish opesiules (outgrowths are often broken thus giving the opesiules the appearance of opesiular indentations, or they may be obscured by sediment infi lling the opesiae). Septula not observed. Ovicells endozooidal, ooecium is formed by the distal zooid, seen as a low swelling of its proximal cryptocyst (Fig. 10G). Avicularia interzooidal, larger than autozooids, elongate. Rostrum channeled with elevated wing-like walls and a pointed tip, conical in outline, almost symmetrical. Proximal part rounded, shorter and wider than rostrum. Cryptocyst pustulose, concave, without peripheral caverns. Opesia egg-shaped with the narrow end pointing upwards and with thin articular ridges bearing two short teeth proximally and a short or long, slit-like opesiular indentation between the teeth; opesiules lacking. Kenozooids rounded, located at the margins of a colony. Cryptocyst fi nely pustulose. Opesia roundish, very small. Closure plates, intramural reparative budding of autozooids, avicularia and kenozooids not observed.

Remarks
Rhagasostoma aralense sp. nov. differs from R. brydonei sp. nov., R. minuens and R. operculatum sp. nov. in having avicularian rostra that are a conical in outline rather than spade-shaped or falciform. The new species differs from R. inelegans and R. angliae, which also have avicularian rostra conical in outline, in the rostrum being almost symmetrical and the avicularian opesia egg-shaped. The crenulated proximal edge of the autozooidal opesia is clearly another important character distinguishing this species from others. However, it is possible that the ribs are sediment particles or diagenetic crystals.

Distribution
Early to middle Campanian Uzbekistan: between Chimboy/Shımbay (Чимбой/Шымбай) and the Aral Sea in the Republic of Karakalpakstan.

Etymology
The species is named from the Latin 'operculum' because the opesiae of autozooids are presumably closed by calcifi ed opercula.

Remarks
Rhagasostoma operculatum sp. nov. differs from R. brydonei sp. nov., another species with a spadeshaped avicularian rostrum, in the rostrum having almost the same width along the whole of their length and having a pointed or rounded rostral tip instead of the rostrum being narrowed at the base, enlarged centrally and with a pointed tip. The new species differs from the other species described herein in having a spade-shaped avicularian rostrum instead of one that is conical in outline or falciform. Moreover, opesiae of autozooids of R. operculatum sp. nov. are presumably closed by calcifi ed opercula.  (Voigt & Williams 1973;Voigt 1974;Di Martino & Taylor 2013;Koromyslova 2014b;Koromyslova & Shcherbinina 2015). Calcifi ed opercula have also been observed in several species belonging to the family Cribrilinidae Hincks, 1879 including Castanopora lambi Turner, 1975 from the Maastrichtian of USA (Turner 1975;McKinney et al. 2003;Taylor & McKinney 2006).

Description
Colony erect, with fl attened bifoliate branches (3-4 mm wide). Ancestrula and early astogeny not observed. Autozooids variable in shape, often broad and subrectangular with rounded distal ends; zooidal boundaries raised. Gymnocyst sometimes observed (Figs 12G, 13E). Cryptocyst extensive, fi nely pustulose, slightly depressed or slightly convex centrally, sometimes with proximal peripheral caverns (Fig. 13B, C). Opesia terminal, rarely subterminal, semielliptical with distal shelf, a thin inner wall delimiting the distolateral part of the opesia and a thickened, projecting outer wall delimiting the proximolateral part (Figs 12C,  13C). Proximal edge of opesia straight, smooth, thickened, with proximolateral outgrowths defi ning small opesiules at the corners; outgrowths often destroyed, giving opesiules the appearance of opesiular indentations. Septula not observed. Ovicells endozooidal, ooecium is formed by the distal zooid, wellrecognizable, with cryptocyst-like surface and arch-like proximal edge with elongated proximolateral processes extending along the cryptocyst of the maternal zooid (Figs 12B, D, G, 13B, D-E). Avicularia interzooidal, smaller than autozooids, elongate. Rostrum channeled, symmetrical, with elevated winglike walls and a pointed tip, rising above the surface of the colony and tilted to the right or left, rostral tip not reaching the opesia of the distal autozooid. Rostrum tilted over the proximal part of the distal autozooid or the ooecium of a maternal autozooid. Wedge-like structures present ( Fig. 12B-D), extending from the base to the middle of the rostrum and partially covering the opesia, but often these structures are broken (Fig. 12E). Proximal part rounded, shorter and wider than rostrum, cryptocyst pustulose, concave, without peripheral caverns. Opesia roundish, with two short teeth proximally, opesiules lacking (Fig. 12E). Kenozooids subcircular, located along the branch margins (Fig. 12A). Cryptocyst fi nely pustulose, opesia roundish (Figs 12C, 13A, D-E). Intramural reparative kenozooidal buds sometimes observed within host autozooids and avicularia (Fig. 13D-E), all with the same polarity as the host zooid. Closure plates, intramural reparative budding autozooids and avicularia not observed.

Remarks
The type material of Rhagasostoma gibbosum (Marsson, 1887) could not be found, although part of the Marsson Collection has recently been recovered (Martha 2014). From the similar species R. tchvanovi , R. gibbosum differs in having a rostrum that is short and tilted to the right or left instead of being long and almost straight. Rhagasostoma gibbosum can be distinguished from R. gibbosulum by the slightly convex border of the cryptocyst and the very rare presence of a gymnocystal portion.  Brydone, 1936 Fig. 14, Table 8 Rhagasostoma gibbosulum Brydone, 1936: 74, pl. 36, fi g. 1.   Brydone, 1936. For each parameter the range is given with the number of measurements in brackets. The arithmetic mean is given ± standard deviation. All measurements in μm.  Additional fi gured material BELARUS • Grodno Region; erratic block of ?late Campanian age in a quarry near Hrodna/Grodno (Гродна/Гродно); PIN 2922/242 (Fig. 14E-F). GERMANY • 4 specs; Schleswig-Holstein, Saturn quarry near Kronsmoor; early Maastrichtian; SMF 29908 ( Fig. 14G), 29909 (Fig. 14H), 29911 (Fig. 14I), 29912 (Fig. 14J).

Basin
Other material GERMANY • 2 specs; same data as for preceding; SMF 29910, 29913.

Description
Colony erect, with fl attened bifoliate branches (1-6 mm wide). Ancestrula and early astogeny not observed. Autozooids variable in shape, often broad and roundish; zooidal boundaries raised. Gymnocyst? present proximolaterally, smooth (Fig. 14B, G, I), but usually not visible. Cryptocyst extensive, fi nely pustulose, usually slightly convex centrally. Opesia terminal or subterminal, semielliptical, delineated by two walls, a thin inner wall forming the distolateral rim and a projecting and thickened outer wall forming the proximolateral rim. Proximal edge of opesia straight, smooth, with outgrowths at the two proximolateral corners delineating small opesiules; outgrowths often destroyed, giving the opesiules the appearance of opesiular indentations (Fig. 14C, I-J). Septula not observed. Ovicells endozooidal, ooecium is formed by the distal zooid, well-recognizable, with cryptocyst-like surface and arch-like proximal edge with elongated proximolateral processes extending along the cryptocyst of the maternal zooid (Fig. 14I). Avicularia interzooidal, smaller than autozooids, elongate. Rostrum channeled, asymmetrical, rarely symmetrical, with elevated wing-like walls and a pointed tip, rising above the surface of the colony and tilted to the right or left, rostral tip not reaching the opesia of distal autozooid. Rostrum tilted over gymnocyst of the distal autozooid or ooecium. Wedge-like structures present ( Fig. 14I-J), extending from the base to the middle of the rostrum and partially covering the opesia, but often these structures are broken (Fig. 14D). Proximal part of avicularium rounded, shorter and wider than rostrum, cryptocyst pustulose, concave, but often not visible. Opesia round with two short teeth proximally, opesiules lacking (Fig. 14J). Kenozooids sometimes numerous, located along edges of branches, roundish (Fig. 14H). Cryptocyst fi nely pustulose. Opesia round. Intramural reparative kenozooidal buds in host autozooids sometimes observed (Fig. 14B), all with the same polarity as the host zooid. Closure plates and intramural reparative budding of autozooids and avicularia not observed.

Remarks
Rhagasostoma tchvanovi differs from the closely related species R. gibbosum in the avicularian rostrum being long and almost straight instead of short and tilted to the right or left. The specimen described by Brydone (1930: pl. 26, fi g. 12) as Onychocella gibbosum weybournensis differs from R. gibbosum (Marsson, 1887) in having a long and straight avicularian rostrum and is conspecifi c with R. tchvanovi. Specimen 32/12582 in the collections of the TsNIGR Museum and labelled as "O. subgibbosum Brydone, 1930" is not the one that was depicted by Favorskaya (1992: pl. 67, fi g. 3) and belongs to R. tchvanovi.

Description
Colony rigidly erect, bifoliate, multiserial, branches 2.0-10.0 mm wide. Ancestrula and early astogeny not observed. Colony formed by pyriform and ovate autozooids and vicarious avicularia. Autozooids subrectangular with rounded distal ends; zooidal boundaries raised. Pyriform autozooids with a narrow proximal end, widening distally, the widening usually starting from the proximal margin of the opesia. In ovate autozooids, the proximal end is partly overlapped by the avicularian rostrum and the autozooid begins to widen considerably below the proximal margin of the opesia, usually at the boundary with an avicularium. Gymnocyst lacking. Cryptocyst slightly granulated, slightly depressed centrally and occupying half or more of the frontal surface of the autozooid, peripheral caverns lacking. Opesia terminal or subterminal. Opesial rim elevated, formed by projecting cryptocyst, bell-shaped due to small lateral projections (occlusor lamina) and opesiular indentations, between which there is a short or long, tongue-like projection of cryptocyst (Figs 16A, C, E, 17G, 18H-I), often broken, in which case the opesiular indentations are poorly defi ned and the proximal edge of the opesia is almost straight. In the complete absence of lateral projections and tongue the opesiae are subcircular (Figs 17C-E, H, 18A-B, D-G). Distal walls with two septula (Fig. 16H); septula in lateral walls not observed. Ovicells immersed, ooecia vestigial formed by the distal zooid, vizor-like, with slightly granulated surface, not protruding above colony surface (Figs 16G, 17C, E, G, 18G, I). Avicularia vicarious, rhomboidal. Rostrum conical in outline and symmetrical, usually with straight lateral walls; indentations or projections present at the rostral base. Three types of avicularia differing mainly in the length of the rostrum and rostral apex which partly overlaps the proximal end of the distal autozooid: avicularia with long rostra and long troughlike apices (130-210 μm); avicularia with short rostra and short trough-like apices (0-130 μm); and avicularia with rostra without projecting apices. Proximal part short, narrowing downwards, with convex lateral sides. Entire frontal surface of avicularium occupied by a fragile, slightly granulated cryptocyst containing fi ve openings: two small, subcircular openings, distally and proximally, in between three slit-like, parallel openings (Figs 17I,18G). Cryptocyst between openings frequently destroyed causing them to coalesce into a single opening of variable outline. Intramural reparative budding of autozooids and closure plates not observed. Intramural reparative budding of avicularia may be present in some or all avicularia depending on the colony (see Koromyslova 2014a), they are observed within host avicularia, numbering as many as three intramural buds, all apparently having the same polarity as the host avicularium.
Kenozooids not observed. Both the inner and outer sides of branches of colonies show concentrations of autozooids in which the opesiae are subcircular and the proximal part is shortened or completely absent, and deformed avicularia that are irregularly ovate to rhomboidal in shape and possess an elongated opening at the centre.

Remarks
Eschara volgensis Eichwald, 1865 from the Late Cretaceous of Simbirsk (now Ulyanovsk, Russian Federation) was regarded as possibly conspecifi c with Rhagasostoma rowei by Voigt (1967). Voigt, who examined the Eichwald Collection housed at the Palaeontological and Stratigraphical Museum at the Faculty of Geology, St Petersburg State University, found that the original material of this species from Simbirsk, Ulyanovsk Oblast, Russia differs from R. rowei only in the opesia being smaller and having a slightly visible tongue-like projection of the cryptocyst. Unfortunately, the specimen of E. volgensis (PSM PSU 2/187 and PSM PSU 2/188) could not be found during our visit to the collections. However, a photograph of the specimen made by Prof. E. Voigt in 1963 is reproduced here (Fig. 16A). This species was redescribed by I.I. Lahusen (1873) whose collection is stored in the Museum of the Mining Institute, St Petersburg. The specimen of E. volgensis (MMI 15/49) could not be examined using SEM, but a microphotograph is shown instead (Fig. 16B). The Lahusen material of E. volgensis is from the vicinity of Yazykov village in Simbirsk Province, Ulyanovsk Oblast. He stated that the sample described by Eichwald (1865) was not completely identical with his samples. Our fi ndings show that the sample described by Lahusen is conspecifi c with Rhagasostoma rowei.
In a monograph on the development of different species of Woodipora Jullien, 1888 from the Coniacian to the Maastrichtian, Schubert (1986) transferred Brydone's species rowei to the genus Woodipora. He did not regard the presence of opesiules or opesiular indentations as being diagnostic for species and concluded that rowei and mimosa were synonymous, differing simply in the preservation of the opesiules (for a more detailed summary regarding a possible synonymy of rowei and mimosa, see the remarks for R. mimosa).
Rhagasostoma rowei has bell-shaped autozooidal opesia because of the small lateral projections and opesiular indentations separated by a short projecting tongue of cryptocyst. Opesia with a straight or rounded proximal edges are formed when the lateral projections and cryptocystal tongue become brokenoff. When the cryptocyst located between the opesiules in autozooids of R. mimosa is destroyed, the opesia takes a shape similar to that of R. rowei. Thus, this similarity is related to the state of preservation and it is evident that pristine colonies of R. rowei have opesiular indentations whereas those of R. mimosa have opesiules.
We have restudied the type material from the collections of R.M. Brydone and also material from the collections of I.I. Lahusen, E. Voigt and T.A. Favorskaya. The specimens labelled as O. mimosa (SM B36697) from the Brydone Collection and (TsNIGR Museum 32/9757) from the Voigt Collection belong to R. rowei because the autozooids lack opesiules. Usually R. rowei is characterized by colonies that widen distally with numerous avicularia. The large number of avicularia is due to the fact that avicularia occur at the start of the new zooidal rows needed to widen the branches. From these avicularia, ovate autozooids are budded distally, from which in turn pyriform autozooids are budded (Koromyslova 2014a). However, there is one specimen of R. rowei from the Maastrichtian of Turkmenistan ( Fig. 18E; fi gured by Voigt 1967: pl. 18, fi g. 2) that differs from the holotype and many other specimens in having almost parallel-sided branches and rare avicularia. Additionally, some large colonies are almost parallel-sided at the base with rare avicularia but widen distally where numerous avicularia are present (Figs 17A-B, 18F-G). It can be assumed that the samples described by Voigt (1967) and other similar specimens (Fig. 18D) were broken-off from the bases of such colonies and hence have few avicularia. Ovate autozooids, which are budded from avicularia, are also rare and often poorly expressed. Obviously, R. rowei was usually characterized by stem-like colonies, which are almost parallel-sided at the base and have few avicularia, but become fan-like distally where numerous avicularia are present. Furthermore, colonies of this species can occasionally have bifurcating branches (Figs 16B-C, F, 18A-B, F). Entire colonies are rare and usually we can see only a small part of colonies, which were broken-off.   A-B, D, H, J = 1 mm; C = 100 μm; E-G = 2 mm; I = 500 μm. (Brydone, 1930) Fig. 19, Table 11 Onychocella mimosa Brydone, 1930: 49, pl. 28, fi gs 1-2.

Description
Colony rigidly erect, bifoliate, multiserial, branches 2.0-10.0 mm wide. Ancestrula and early astogeny not observed. Colony formed by pyriform and ovate autozooids and vicarious avicularia. Autozooids subrectangular with rounded distal ends; zooidal boundaries raised. Pyriform autozooids with a narrow proximal end but widening distally, the widening usually starting from the proximal margin of the opesia. In ovate autozooids, the proximal end is partly overlapped by the avicularian rostra and the autozooid begins to widen considerably below the proximal margin of the opesia, usually at the boundary with an avicularium. Gymnocyst lacking. Cryptocyst slightly granulated, slightly depressed centrally and occupying half or more of the autozooidal frontal surface, peripheral caverns lacking. Opesia terminal or subterminal. Opesial rim elevated, formed by a projecting, thickened cryptocyst, subcircular, a pair of opesiules located 70-90 μm proximallateral of opesia and divided from the latter by a cryptocystal tongue. Fusion of cryptocyst distal of opesiules and tongue of cryptocyst incomplete. Septula in walls not observed. Ovicells immersed, ooecium is formed by the distal zooid, triangular, vestigial, with slightly granulated surface, not protruding above the colony surface ( Fig. 19A-B, D-F). Avicularia vicarious, rhomboidal. Rostrum conical in outline and symmetrical, usually with straight lateral walls; indentations or projections present at base of rostrum. Usually two types of avicularia differing mainly in the length of the rostrum and rostral apex, which partly overlaps the proximal end of the distal autozooids: avicularia with short rostra and short, trough-like apices (0-130 μm), and avicularia with long rostra without projecting apices. Proximal part short, narrowing downwards, with convex lateral sides. Entire frontal surface of avicularium occupied by a fragile, slightly granulated cryptocyst containing fi ve openings: two small, subcircular openings, distally and proximally, in between three slit-like, parallel openings. Cryptocyst between openings frequently destroyed causing them to coalesce into a single opening of variable outline. Closure plates, kenozooids and intramural reparative budding in autozooids and avicularia not observed.

Remarks
Rhagasostoma mimosa was introduced by Brydone (1930) for species from the Porosphaera beds of Trimingham very much resembling R. rowei  but differing from the latter in having two opesiules. He considered the possibility that colonies of R. rowei were badly preserved samples of R. mimosa but concluded that the two must be different species because colonies of R. rowei are more slender and fragile, avicularia of R. mimosa have a central cavity enclosed by a marked wall, and R. rowei and R. mimosa never co-occur in the Chalk of England.
Voigt (1967) re-examined R. rowei and R. mimosa from Central Asia and rejected the morphological arguments used by Brydone (1930) to separate the two species. He concluded that except for R. mimosa having opesiules, the two species are completely identical, but as no intermediate stages occur, he regarded a separation of R. rowei and R. mimosa as justifi ed. He argued that calcifi cation of the frontal wall of R. rowei during the Late Cretaceous resulted over time in the formation of two opesiules completely separated from the opesia as passages for the bundles of parietal muscles. A similar development has also been observed in Rhagasostoma disparile (d 'Orbigny, 1851) and Rhagasostoma strumulosum Marsson, 1887(cf. Schubert 1986, which are indeed closely related to R. rowei and R. mimosa. Furthermore, Voigt (1967) assigned mimosa to the genus Woodipora Jullien, 1888 because of the opesiules, but kept rowei in Onychocella.
In his Woodipora monograph,  regarded rowei and mimosa as two different phenotypes of the same species, which he assigned to Woodipora. The length and width measurements that he conducted on autozooids and avicularia and their opesiae revealed no signifi cant distinction between rowei and mimosa. Furthermore, he argued that colonies in the mimosa stage often have autozooids with and also without opesiules (cf. Schubert 1986: pl. 8, fi g. 3). As the zooids without opesiules are of primary origin and cannot have been destroyed during preparation or other processes, he concluded that the occurrence of opesiules is not species-specifi c and cannot be considered an argument for the existence of mimosa as an independent species. He regarded the mimosa stage as an 'end-member' of a parallel development by progressive calcifi cation of the cryptocyst from the Coniacian to the Maastrichtian resulting in the formation of two opesiules. Zooids with advanced calcifi cation of the cryptocyst, however, did not completely replace the rowei stage, the two stages thus co-occurring in single colonies. Voigt (1991) again regarded mimosa and rowei (and also R. disparile and R. strumulosum) as separate species. He saw the four species as an example of an intermediate status between onychocellid pseudomalacostegans and microporid or thalamoporellid coilostegans, placing disparile and rowei in Onychocella and strumulosum and mimosa in Woodipora. Thus, the microporid or thalamoporellid genus Woodipora would have originated several times by convergent evolution.
Without description or fi gures,  mentioned Onychocella mimosa from the Maastrichtian of the southern Aral Sea Region in Uzbekistan and from Hanga-Baba (урочище Ханга-баба), 30 km east of Fort Shevchenko (Форт-Шевченко) in the Mangystau Region, Kazakhstan.
A comprehensive description of morphological differences between R. rowei and R. mimosa from the Grodno quarry in Belarus was undertaken by Koromyslova (2014a). As she showed, the two species are not conspecifi c and differ in the shape of the autozooidal opesia and in the cryptocyst that develops opesiular indentations in R. rowei and opesiules in R. mimosa. Furthermore, specimens of R. rowei show intramural reparative budding in avicularia, which has not been observed in colonies of R. mimosa. Based on the length of the rostrum and the apex, three different types of avicularia have been observed in R. rowei, while R. mimosa shows only avicularia with short rostra and short trough-like apices (Koromyslova 2014a). Avicularia of R. rowei and R. mimosa are indeed very distinct from most other species within the genus Rhagasostoma. Centrally located, there is an opesia and three opesiules in the avicularian cryptocyst.
We have restudied Brydone's syntypes in the SM collection, choosing a lectotype (fi gured by Brydone 1930: pl. 28, fi gs 1-2), as well as the material from the collection of E. Voigt (Fig. 18I, 19C) and T.A. Favorskaya (Fig. 19D-F). The specimens labelled as Onychocella mimosa (SM B36697) from the Brydone Collection and (TsNIGR Museum 32/9757) from the Voigt Collection belong to R. rowei because the autozooids lack opesiules. As already discussed by Brydone (1930) and Voigt (1967), R. mimosa and R. rowei are very similar, differing only in R. mimosa having opesiae that are round instead of bell-shaped and opesiules instead of opesiular indentions. The presence of autozooids with opesiular indentations in colonies of R. mimosa in most cases can be explained by post-mortem destruction of the cryptocystal tongue that separated the opesiules from the opesia, or by the fact that the opesiae and opesiules of these autozooids were not formed fully and they were in the rowei stage (cf. Schubert 1986: pl. 8, fi g. 3).
Coniacian of England; R. minuens -Coniacian, late Santonian, early Campanian and early Maastrichtian of England, early Campanian of Germany, and Campanian of France; R. angliae -middle Campanian and early Maastrichtian of England, late Campanian of France, early Maastrichtian of Germany; R. aralense sp. nov. -early to middle Campanian of Uzbekistan; and R. operculatum sp. nov. -early to middle Campanian of Turkmenistan. Rhagasostoma inelegans is excluded from the list of common onychocellids in the Late Cretaceous of Europe and Central Asia. This species and R. brydonei sp. nov., R. minuens, R. angliae, R. aralense sp. nov. and R. operculatum sp. nov. have narrow stratigraphical and geographical distributions.
As a result of studying new material of Rhagasostoma gibbosum and R. gibbosulum, their distributions have been extended (Fig. 20); R. gibbosum is now known from the late Campanian of Belarus and Kazakhstan, the early Maastrichtian of England and Germany, and the Maastrichtian of Denmark;