Dinoflagellate Cysts from the Upper Cretaceous (Upper Campanian to Lowermost Maastrichtian) of the Middle Vistula River Section, Poland

ABSTRACT In this article, the most representative dinoflagellate cyst genera and species recognised in the rich palynomorph assemblages of the upper Campanian–lowermost Maastrichtian succession of the Middle Vistula River section (central Poland) are treated taxonomically: in particular, six genera and 16 species are considered. Oligosphaeridium araneum sp. nov., which possesses processes with relatively long and slim stems and perforate or fenestrate terminations, is described as new. Glaphyrocysta pala comb. nov. and Hystrichosphaeridium brevispinum stat. nov. are proposed. Glaphyrocysta pala comb. nov. has a dorso-ventrally compressed central body and lacks mid-ventral processes connecting the central body with the membrane, suggesting its affinity with Glaphyrocysta, rather than Riculacysta; and H. brevispinum stat. nov. is raised to species rank on the basis of the distinct morphology of its processes. The tabulation pattern and plate arrangement are determined for the first time in Amphorosphaeridium and revised in Callaiosphaeridium. Both genera have a sexiform hypocystal configuration, L-type ventral organisation, and neutral torsion, which indicates their inclusion in the subfamily Leptodinioideae. The species-level taxonomy of the genera Hystrichosphaeridium and Samlandia is discussed. The transfer of Hystrichosphaeridium proprium to Hystrichokolpoma is rejected, and Hystrichosphaeridium? recurvatum is questionably left in Hystrichosphaeridium, although it is characterised by a commonly larger number of processes per plate and the lack of a preapical process.


Introduction
The rich and well-preserved organic-walled dinoflagellate cyst assemblages recovered from the Middle Vistula River section were found to include many biostratigraphically critical forms, which enabled the development of a refined dinoflagellate cyst-based biostratigraphical framework (Niechwedowicz and Walaszczyk 2021). The dinoflagellate cyst record, based on a dense sample set, revealed a narrow interval, immediately below the Campanian-Maastrichtian boundary, characterised by a distinctive taxonomic turnover in dinoflagellate cyst assemblages. This interval is marked by the extinction of a number of typically Campanian (and pre-Campanian) genera and species (Xenascus and some species of Odontochitina), and the appearance of typically Maastrichtian and younger (Paleogene) forms (of the genera Cladopyxidium and Glaphyrocysta). The documented Middle Vistula River dinoflagellate cyst succession (Niechwedowicz and Walaszczyk 2021) is consistent with both Boreal (Belgium and the Netherlands) and Tethyan Realm successions (southern Germany, south-west France, the northern Apennines), enabling reliable correlations. Of particular importance is the correspondence of the Polish material with the dinoflagellate cyst record documented in Tercis les Bains, south-west France [the Global Stratotype Section and Point (GSSP) for the base of the Maastrichtian Stage; see Odin and Lamaurelle (2001)], supporting the correlation between the Boreal and Tethyan Realms inferred from macrofossils (Walaszczyk et al. 2002;Walaszczyk 2004;Machalski 2012a).
In total, 129 species-or subspecies-level dinoflagellate cyst taxa have been identified from the upper Campanian-lowermost Maastrichtian interval of the Middle Vistula River section (see Niechwedowicz and Walaszczyk 2021). Not all are treated herein. The present paper focuses on those species or genera that, in light of the present material, appear new to science, are in need of supplementary taxonomic treatment, or represent biostratigraphically critical forms. The 16 species and six genera discussed herein include the description of a new species, Oligosphaeridium araneum sp. nov., a discussion on some stratigraphically important species of the genus Samlandia, the transfer of a Riculacysta species to Glaphyrocysta, a discussion on tabulation patterns and systematic positions in the genera Amphorosphaeridium and Callaiosphaeridium, and species-level taxonomy in the genus Hystrichosphaeridium. Three additional new dinoflagellate cyst species from the Campanian-Maastrichtian boundary interval of the Middle Vistula section, Odontochitina dilatata, Callaiosphaeridium bicoronatum, and Samlandia paucitabulata, were recently described (Niechwedowicz 2018a, andWalaszczyk 2021).

Stratigraphical framework
The Middle Vistula River composite section consists of middle Albian to Danian strata and is accessible in a series of outcrops located along the Middle Vistula River valley (Po_ zaryski 1938;Marcinowski and Radwa nski 1983;Walaszczyk 1992Walaszczyk , 2004. The Upper Cretaceous succession is quite complete, with small hiatuses mostly present in the lower part (Marcinowski 1980;Walaszczyk 1987). The upper Campanian-lowermost Maastrichtian interval treated in this study is exposed near the town of Solec nad Wisłą (Figure 1). Due to the rather monotonous siliceous marl (referred to locally as opoka; see Jurkowska et al. 2019) facies development of the succession, lithological correlations between particular sections in this interval are not reliable: the sole exception is a single marly layer, the so-called 'boundary marl' at the Campanian/Maastrichtian boundary ( Figure 2; Walaszczyk 2004). Fortunately, the succession is highly fossiliferous, yielding a variety of biostratigraphically significant groups, including both macro-and microfossils (Walaszczyk 2012(Walaszczyk , 2015. The modern biozonal schemes developed for the section, based on inoceramid bivalves (Walaszczyk 2004), ammonites (Machalski 2012b), belemnites (Keutgen et al. 2012;Remin 2012Remin , 2015, foraminifers (Dubicka and Peryt 2012;Dubicka in Walaszczyk et al. 2016), and dinoflagellate cysts (Niechwedowicz and Walaszczyk 2021), provide reliable stratigraphical control (Figure 2). Beyond biostratigraphy, the succession also contains good d 13 C (Voigt in Keutgen et al. 2012) and palaeomagnetic signals (Plasota et al. 2015). For a more detailed historical review of stratigraphical studies conducted in the area, the reader is referred to Walaszczyk (2004Walaszczyk ( , 2012.

Material and methods
The sampled portion of the Campanian-Maastrichtian boundary interval of the Middle Vistula River composite section consists of the following exposures (in ascending stratigraphical order): Piotrawin, Raj, Podole, Raj North, Kłudzie North, Kłudzie South, and Dziurk ow ( Figures 1C and 2). Their characteristics are summarised in Table 1. A total of 183 samples were collected, taken at 1 m intervals on average; denser sampling (down to 25 cm resolution) was conducted in the Campanian-Maastrichtian boundary interval and in strata critical for the inter-section correlation of particular exposures ( Figure 2).
The palynological material was extracted from 95-105 g of dry sediment; in rare cases, only 20-50 g of sample were available. The samples were processed using a standard palynological preparation technique, including 30% hydrochloric and 70% hydrofluoric acid treatments, with repeated decantation; oxidation, ultrasonic treatment, and heavy-liquid separation were not used. The extracted organic residues were sieved through a 15 lm nylon mesh and concentrated via centrifugation. A drop of residue from each sample was mounted on a slide using glycerin jelly, overlain with a  Po_ zaryski (1974)), and locations of the studied sections (C) within a geological map of the upper Campanian-lowermost Maastrichtian part of the Middle Vistula River section (modified from Walaszczyk 2004). 20 Â 20 mm coverslip, and sealed with clear varnish; at least two slides per sample were prepared. Identification of palynomorphs was conducted under a transmitted light microscope (TLM). The most important specimens were documented using both a TLM (equipped with a digital video camera) and scanning electron microscopy (SEM), where possible; SEM was performed at the NanoFun Cryo-SEM Laboratory (Faculty of Geology, University of Warsaw, Warsaw, Poland), with Zeiss AURIGA 60 and Zeiss SIGMA VP scanning electron microscopes. The captions for each illustrated specimen include both slide number and microscope coordinates. The coordinates were provided from the vernier scale of the microscope (0.1 mm accuracy), consisting of a pair of horizontal/vertical coordinates measured from the 0/0 reference point (the bottom left corner of the coverslip marked on each slide). England Finder (EF) coordinates were computed using the England Finder Calculator (see Gonz alez 2012). For specimens mounted on SEM stubs, only EF coordinates are provided. The taxonomy is after Fensome et al. (1993) and Fensome et al. (2019a), with additions from Pearce and Williams (2018) and Niechwedowicz and Walaszczyk (2021). The descriptive terminology follows Evitt (1985), Williams et al. (2000), and Fensome et al. (2009). The 'para' terminology is not used here for the reasons suggested by Fensome et al. (2009, p. 8). The organic residues, palynological slides, and SEM stubs are lodged at the S.J. Thugutt Geological Museum, Faculty of Geology, University of Warsaw, Warsaw, Poland (MWGUW).
Remarks. The diagnosis of the genus is here emended to emphasise the leptodinioid tabulation pattern recognised in the type species (see Figure 3; Plate 1, figures 1-12). Although Davey (1969) did not recognise the tabulation pattern in A. fenestratum, and it cannot be clearly inferred from illustrations of the holotype alone (Davey 1969, pl. 3, figs 1, 2), he noticed some process alignment in the type species (Davey 1969, p. 31). In overall morphology, Amphorosphaeridium most resembles Exochosphaeridium Davey et al. 1966, which, however, belongs to the subfamily Cribroperidinioideae (see Helenes 2000, p. 137, 138, 140;Fensome et al. 2009, p. 31). Peyrot (2011, p. 284) synonymised Amphorosphaeridium fenetratum of Davey 1969, pl. 3, figs 1, 2 (the holotype of the species and the type species of the genus) with Exochosphaeridium majus (Lejeune-Carpentire 1940) Peyrot 2011, indicating the synonymy of the two genera, although not expressis verbis. However, the two species differ significantly: the processes in E. majus are solid or hollow (more commonly solid), whereas A. fenestratum generally bears hollow processes (only the slimmest, mainly sulcal processes may be solid distally, but proximally they are hollow). Furthermore, in E. majus the antapical prominence is never developed, while A. fenestratum may bear prominences at both poles.
Comparison. Similarly to Amphorosphaeridium, the genus Exochosphaeridium has fibrous (or fibro-pitted) central body wall and processes, but the two genera differ in the development of processes and their distribution. In Exochosphaeridium the processes are generally more slender; they are commonly acuminate or hair-like and generally solid. Hollow processes, when present, are usually hollow only proximally. Occasionally, Exochosphaeridium possesses an enlarged apical process, sometimes irregularly branched distalward (see Plate 1, figure 16). Furthermore, the processes in Exochosphaeridium are generally nontabular, rarely contabular, vaguely indicating the tabulation (recognised in only single species, E. alisitosense; see Helenes 2000). In contrast, processes in Amphorosphaeridium are more robust, distinctly hollow, and less fibrous, and their arrangement more clearly indicates tabulation. Moreover, Amphorosphaeridium may possess prominences at both poles. Fibrocysta Stover & Evitt 1978 has an ellipsoidal central body, generally horn-like protrusions at the cyst poles, and more slender processes. Cordosphaeridium Eisenack 1963b has fewer processes (1 meso-to obtabular process per plate; Fensome et al. 2009). Operculodinium Wall 1967 has generally solid processes, which usually are uniform in size and shape, and tend to be isolated. Turbiosphaera Archangelsky 1969a has fewer, intratabular (one per plate), wider, and taeniate processes.  Figure 3); arrows indicate a partial fusion of antapical processes. 9-12. MWGUW ZI/90/Dz9/ 0a, 1/13.2, EF L17/0; right lateral view of right lateral (9, 10) and left lateral (12)  neighbouring processes may be joined. Apical and antapical prominences are occasionally present, expressed by enlarged processes. The tabulation is indicated by the archaeopyle and arrangement of the processes. The archaeopyle is precingular, type 1 P (3 0 '), operculum is free or attached.

Amphorosphaeridium fenestratum
Emended description. Chorate cysts of intermediate size, with ovoidal central body and numerous (c. 60) robust processes. Central body wall and process walls are more or less fibrous (or fibro-pitted). The processes are constant in length (c. 1/3-1/2 of central body diameter), but vary in width; generally tapering and hollow throughout entire length, rarely acuminate and distally solid (most commonly sulcal processes). The processes are relatively wide proximally, typically with circular bases, and tapering, distally only slightly expanded, terminating with aculeate tip. The processes are intratabular, commonly arranged in latitudinal and meridional rows. The processes arising from the larger pre-and postcingular plates are commonly penitabular. Process distribution is more or less uniform; the number of processes per plate (1-5) depends on the plate size and position on cyst: one process per 1p plate and each sulcal plate, 1-3 processes per smaller plate of other types (6 0 ', 1 0 '', cingular, and apical plates), 3-5 processes per larger plate (rest of pre-and postcingulars, and antapical plate). Neighbouring processes occupying particular plates sometimes are connected proximally, or are completely fused. Apical and antapical prominences are occasionally evident, expressed by enlarged (widened) process. The tabulation formula is 4 0 , 6 0 ', 6c, 6 0 '', 1p, 5 s (ps, ls, rs, ras, as), 1 0 ''', more or less clearly expressed by the distribution of the processes, indicating leptodinioid pattern (sexiform antapex, L-type ventral plate arrangement, and neutral torsion of the hypocyst). The archaeopyle is precingular, type 1 P (3 0 '), operculum is free or attached. Remarks. The ovoidal shape of the central body, the presence of the 1 P archaeopyle and the distribution of processes (arranged in latitudinal and meridional rows, sulcal processes typically slimmer than others and distally solid, except for those occupying plates as and ps) make the cyst orientation easy to determine. The processes in A. fenestratum are of constant length, but occasionally exhibit significant variability in width on a single specimen. Essentially, the processes have circular bases, and their number per particular plate is constant. However, some of the neighbouring processes occupying the same plate area may connect proximally or may be almost completely fused; such processes have wider bases (oval, elliptical, or subangular in cross-section), and have two or more tips (Plate 1, figures 9-11). The subspecies differentiation of Davey (1969) is confusing, since in both subspecies the processes may be branched/connected (see Davey 1969, p. 31-33). The degree of process fusion in A. fenestratum is considered intraspecifically variable and, consequently, the two subspecies are synonymised herein. Peyrot (2011, p. 284, 289) included the specimens of A. fenestratum that lack enlarged antapical process (including the holotypesee Davey 1969, pl. 3, figs 1, 2) into Exochosphaeridium majus, significantly broadening its concept. The concept of Amphorosphaeridium refers to the presence of prominences at both poles (Davey 1969, p. 30), but this is not the most distinctive feature of A. fenestratum, in which the prominences may be absent or difficult to recognise. Based on the material studied herein, a single enlarged process at the apex may be one of the apical processes (e.g. the process occupying plate 2 0 , as in the specimen in Figure 3; Plate 1, figures 4-7), which does not occupy the central polar position. The antapical pole of the same specimen is characterised by the presence of four processes, none of which are enlarged (significantly wider) or occupy the central polar position. The distinction of the antapical prominence in this case may be marked by a partial fusion of processes (Plate 1, figures 6, 7). It appears that the processes occupying plate 1 0 ''' may be completely fused, forming one relatively wide process. This would explain the presence or absence of an antapical prominence in the type material (see Davey 1969, pl. 1, fig. 6, pl. 2, fig. 4, pl. 3, figs 1-3). In contrast, Exochosphaeridium may bear only one prominence (an enlarged apical process, commonly branched or expanded distally; see Plate 1, figure 16); but, as in Amphorosphaeridium, it is not always present. Thus, the most characteristic feature of A. fenestratum is the development of processes and their arrangement, in which it clearly differs from E. majus (Plate 1, figures 13-16). Consequently, the synonymy proposed by Peyrot (2011) is rejected.
Comparison. Exochosphaeridium majus (Lejeune-Carpentier 1940) Peyrot 2011 (see Plate 1, figures 13-16) most resembles Amphorosphaeridium fenestratum, both in overall appearance and in its fibrous (or fibro-pitted) wall. What differentiates the two species is the process development. In A. fenestratum, processes are usually relatively wide and distinctly hollow throughout their entire length; by contrast, E. majus bears solid and hollow processes, but the latter are commonly acuminate, being hollow only proximally (if at all). Furthermore, the processes in E. majus are not arranged in a manner indicating tabulation. Amphorosphaeridium fenestratum may bear apical and antapical prominences, while E. majus may bear only the apical one. Pervosphaeridium elegans Louwye 1997 and P. tubuloaculeatum Slimani 1994 also resemble A. fenestratum in possessing hollow processes. Pervosphaeridium elegans (Plate 1, figures 17, 18) differs in having more slender and relatively longer (c. 3/4 of central body diameter) processes, constant in width on individual specimens (except for the sulcal processes) and never connected proximally. Pervosphaeridium tubuloaculeatum (Plate 1, figures 19-21) is significantly smaller, and it can be distinguished by the morphology of its processes (degree of tapering of the processes and their termination type). Furthermore, Pervosphaeridium has an archaeopyle of type 2 P. Emended diagnosis. Gonyaulacacean (leptodinioid) chorate cysts with subspheroidal to ovoidal central body, bearing gonal (processes) and sutural (ridges, septa, processes) features, clearly reflecting plate arrangement. The processes are of two types: (i) six prominent equatorial processes (tubular, broad, distally open) in gonal positions; and (ii) simple processes (solid, slender, or taeniate) in gonal or intergonal positions, located elsewhere on epi-and hypocyst excluding cingulum. Processes of both types commonly are connected by sutural septa of variable height (not higher than the processes), with distal margin entire or irregular; septa sometimes are distally supported by thickening (transverse ridge parallel to plate suture). The archaeopyle is epicystal.

Recorded
Emended description. Chorate cysts with subspheroidal to ovoidal central body, bearing gonal (rarely intergonal) processes, sutural ridges and septa; suturocavate. Epi-and hypocyst are of similar size, or epicyst is slightly smaller. The central body surface is internally smooth, and externally it is smooth, scabrate, or rugulate, with plate margins smooth and often perforate in regular manner (lines of perforations parallel to plate boundaries); septa and processes are generally smooth. The processes are of two types: (i) six equatorial processes, tubular, broad, located in gonal positions between cingular and postcingular plate series; and (ii) slender processes, predominantly gonal, rarely intergonal, positioned elsewhere on epi-and hypocyst, except for the anterior and posterior margin of the cingulum. The equatorial processes are distinct, tubiform, distally flared, and open, terminating with few spines, or with low crests with incorporated numerous short spines. The equatorial processes bear more or less expressed longitudinal striae (three or four) extending from postcingular and cingular sutures. The slender gonal processes are generally solid (sometimes partially hollow), resulting from merging of sutural septa at gonal points; distally they are commonly furcate, occasionally with additional bifurcation of second order. The intergonal processes, if present, are taeniate. Equatorial and slender processes are connected by a more or less complete network of sutural ridges and septa of variable height (not higher than the processes); equatorial processes are sometimes interconnected basally by low septa or septum-like hollow membranous structures. Septa have entire or irregular distal margin, in the latter case forming taeniate intergonal processes; distally septa may be supported by thickening (transverse ridge parallel to plate suture), crowned with short spines incorporated into low crests. The tabulation formula is 4 0 , 6 0 ', 6c, 6 0 '', 1p, 5 s, 1 0 ''', expressed by the archaeopyle and arrangement of gonal and sutural features, indicating leptodinioid pattern (sexiform hypocystal configuration, L-type ventral organisation, and neutral torsion of the hypocyst; see Figure 4); sutural ridges between apical and sulcal plates may be reduced or absent. The archaeopyle is epicystal (formula A 1-4 0 þ P 1-6 0 ' ), operculum is free or attached ventrally.
Remarks. The present emendation emphasises the leptodinioid tabulation pattern, the gonal nature of the tubular processes, and variability in the development of septa and processes in Callaiosphaeridium. The sutural features indicating tabulation in Callaiosphaeridium may be present at almost all plate boundaries, but may be only partially developed or absent in the sulcal and apical regions (see also Duxbury 1983, pl. 5, fig. 12, text- fig. 18; Khowaja-Ateequzzaman and Garg 2004, pl. 1, fig. 7). The sutural ridge is also absent from the anterior margin of the cingulum (the boundary between the precingular and cingular series is defined by the archaeopyle suture), but contrary to Davey and Williams (1966, p. 103) and Evitt (1985, p. 244), the posterior margin of the cingulum and the boundaries between particular cingular plates are clearly defined by low sutural ridges ( The processes in Callaiosphaeridium are mainly gonal (including the large tubular ones), and rarely intergonal (these are present only in C. trycherium). The characteristic six equatorial processes in Callaiosphaeridium are not intratabular (as suggested before, e.g. Davey and Williams 1966, p. 103;Stover and Evitt 1978, p. 202;Below 1981, p. 27;Evitt 1985, p. 181, 188, 243, 245), but occupy gonal positions (compare also Duxbury 1980, p. 114;Fensome et al. 2009;p. 17, 2016, p. 30). This is indicated by the distribution of faint sutural ridges on the cyst surface: tubular processes arise at locations where three or four sutural ridges meet (see Figures 4 and 5; Plate 3, figures 5-7; Plate 5, figures 8-10). These ridges separate postcingular and cingular plates, and extend onto the tubular processes in the form of longitudinal striae (Plate 2, figures 7, 10; Plate 3, figure 11; Plate 4, figure 17). Since the tubular processes in Callaiosphaeridium are gonal and positioned between the cingular and postcingular series, they should rather be referred to as equatorial, not cingular. The gonal position of such processes is not an unusual feature; e.g. gonal, hollow, and distally open processes are known in Spiniferites pseudofurcatus (see Evitt 1985, p. 228;Riding and Lucas-Clark 2016, p. 45, pl. 8, figs 1, 2), and S. procerus (see Marheinecke 1992, p. 29).
Comparison. The possession of an epicystal archaeopyle, and more broadly an overall similarity in appearance, make Actinotheca Cookson & Eisenack 1960a, Avellodinium Duxbury 1977, and Heslertonia Sarjeant 1966b very similar to Callaiosphaeridium. Actinotheca differs in having the equatorial processes, or rather 'intergrown membranes' or 'equatorial row of somewhat boxlike chambers' (Evitt 1985, p. 243, 244), extended more in the equatorial plane. Additionally, it is characterised by the absence or weak development of epiand hypocystal sutural and gonal features (ridges, septa, processes), unlike Callaiosphaeridium. Avellodinium, similar to Callaiosphaeridium, has gonal and intergonal processes. However, its equatorial processes are not tubular, but are comparable in shape with other processes. Heslertonia has high and regularly developed sutural septa covering the entire cyst sutures. Morphologically, Callaiosphaeridium is a transitional form between Avellodinium and Heslertonia.
In ventral, dorsal, or lateral views, Callaiosphaeridium specimens with attached operculum resemble Spiniferites Mantell 1850. The gonal, distally furcate processes occupying the polar regions in Callaiosphaeridium (e.g. in C. asymmetricum) are very similar to those in Spiniferites (see Evitt 1985, p. 244). The latter genus, however, differs from Callaiospharedium in having S-type ventral organisation and a precingular archaeopyle.
Stratigraphical comments. Callaiosphaeridium is biostratigraphically important for the Campanian-Maastrichtian boundary; C. asymmetricum disappears in the uppermost Campanian (Habib and Miller 1989;Kurita and Mclntyre 1994;Nøhr-Hansen 1996;Slimani 2001;Skupien and Mohamed 2008;Niechwedowicz and Walaszczyk 2021), and the genus finally disappears in the lowermost Maastrichtian in Poland with its youngest representative, C. bicoronatum (Niechwedowicz and Walaszczyk 2021 The slender processes are connected by sutural septa with U-shaped and entire distal margin; the equatorial processes are occasionally connected to each other and to the slender ones by low septa. The archaeopyle is epicystal. Emended description. Chorate cysts with subspheroidal to ovoidal central body, bearing gonal (processes) and sutural (ridges and septa) ornament; suturocavate. Epi-and hypocyst are of similar size, or epicyst is slightly smaller. The central body surface is internally smooth, and externally bears rugulate ornament covering most of the plate areas; plate margins are smooth, and often regularly perforate (lines of perforations parallel to plate boundaries); septa and processes have smooth walls. The processes are of two types: (i) six equatorial tubular processes located in gonal positions between cingular and postcingular series; and (ii) slender gonal processes, positioned elsewhere on epi-and hypocyst, except for margins delimiting cingulum. The equatorial processes are distinct, tubiform, with oval to elliptical cross-sections (wider in equatorial plane), distally are flared and open, terminating with a few aculeate spines (c. 7-10 mm in length). The equatorial processes bear three to four longitudinal striae (faint, poorly expressed) extending from postcingular and cingular sutural ridges. The slender processes are generally solid (may be partially hollow), resulting from merging of sutural septa at gonal points; distally are furcate (usually trifurcate), with branches more or less perpendicular to process stem and parallel to sutures; occasionally with additional bifurcation of second order. The slender processes are connected by a network of U-shaped septa, variable in height, with entire distal margin; septa are equal in height to process length, with minimum height at half distance between the processes. Rarely, the equatorial processes are connected by low septa to the slender processes, or are interconnected basally by septa or septum-like hollow membranous structures. The tabulation formula is 4 0 , 6 0 ', 6c, 6 0 '', 1p, 5 s (ps, ls, rs, ras, as), 1 0 ''', expressed by the archaeopyle and arrangement of gonal and sutural features, indicating leptodinioid pattern (sexiform hypocystal configuration, L-type ventral organisation, and neutral torsion of the hypocyst). Sutural ridge between plates 1 0 and 4 0 may be reduced; ridges outlining plate 1 0 '' and sulcal plates may be reduced or absent. The cingulum is relatively narrow, laevorotatory, shifted by one cingulum width. The archaeopyle is epicystal (formula A 1-4 0 þ P 1-6 0 ' ), operculum is free or attached ventrally.
Remarks. The present emendation defines more precisely the distribution of the sutural and gonal features on the cyst surface and their correspondence to the tabulation pattern (see also Figure 4). Additionally, it highlights the nature of tubular equatorial processes that are in gonal positions in the type species of the genus, as are the slender ones. The central body wall in C. asymmetricum is relatively thin (c. 0.5-1 mm), but robust. It is homogeneous in structure with indistinguishable layering (Plate 2, figure 8). The surface of the central body wall is smooth, foveolate, or granular under TLM. SEM studies prove it to be rugulate, resembling the wrinkled skin of withered fruit (Plate 2, figures 5-7, 12; Plate 3, figures 6, 7, 9-12). Plate margins are smooth, with the characteristic lines of perforations, which is a manifestation of the suturocavate septal nature (Plate 2, figures 5-7; Plate 3, figures 5-7). The most characteristic features of this species are the epi-and hypocystal slender gonal processes connected by U-shaped sutural septa (Plate 2, figures 6, 7, 13). The slender processes are distally furcate (usually trifurcate), commonly with additional second-order bifurcation (Plate 2, figure 13; Plate 3, figure 12) similar to the processes in Spiniferites ramosus (Ehrenberg 1837) Mantell 1854. The six tubular (equatorial) processes in C. asymmetricum bear distally a few characteristic, aculeate spines (Plate 2, figures 5, 9, 10) that arise from the outer surface of the process wall. The spines arise perpendicularly to the process wall and curve distally towards the central body. Remarks. Similarly to Callaiosphaeridium asymmetricum, the sutural features (ridges and septa) are well expressed in C. bicoronatum and discernible under both TLM and SEM. The ridges bounding the cingular and postcingular plates are extended to six tubular (equatorial) processes ( Figure 5; Plate 4, figures 16, 17; Plate 5, figures 8-10), with each bearing a constant number (three or four, depending on their position on the cyst) of longitudinal striae (Niechwedowicz and Walaszczyk 2021, pl. 4, figs 13, 18, pl. 5, fig. 18). The presence of these sutures proves that the equatorial processes occupy gonal positions. As in C. asymmetricum, the sutural ridges between plates 1 0 and 4 0 may be reduced, and the ridges outlining plate 1 0 '' and the sulcal plates may be reduced or absent (Plate 4, figure 2; Plate 5, figure 14). The tabulation formula in C. bicoronatum (4 0 , 6 0 ', 6c, 6 0 '', 1p, ps, ls, rs, ras, as, 1 0 ''') is identical to that of C. asymmetricum (see Figure  4). The plate arrangement in C. bicoronatum, well demonstrated by the distribution of gonal processes, sutural ridges, and septa, reflects, as in the type species, a leptodinioid pattern (sexiform hypocystal configuration, L-type ventral organisation, and neutral torsion of the hypocyst; Plate 5, figures 5-12, 16, 18). The only difference in the plate arrangements is in the shape of the first apical plate: in both species, it is antero-posteriorly elongated, but in C. bicoronatum it is roughly rectangular (Plate 5, figure 4), while in C. asymmetricum it is subtriangular (with posteriorly directed acute angle; plate 1 0 is not in contact with plate 6 0 '; Plate 3, figure 4). Callaiosphaeridium asymmetricum differs from C. bicoronatum in having relatively low and U-shaped sutural septa. In the latter species, septa are of constant height and are supported by distal ridges (parallel to plate sutures) surmounted with short spines incorporated into low crests. In C. bicoronatum, the equatorial processes distally bear crests with spines similar to those terminating septa, while C. asymmetricum lack this feature.  Type. White 1842, pl. 4, fig. 11, as Xanthidium tubiferum var. complex (original type). Neotype: Davey & Williams 1966, pl. 7, fig. 1, as Oligosphaeridium complex, designated by Davey & Williams (1966, p. 71

Diagnosis.
A species of Oligosphaeridium with relatively slim and long (equal to, or slightly longer than, central body diameter) processes. Terminations of the processes are short (branching occurs in their distalmost portion), and wide, significantly expanded laterally, forming perforate or fenestrate platforms.
Description. The central body is subspheroidal, and consists of relatively thin (c. 0.5-1 mm) and homogeneous wall with rugulate outer surface; wall layering is not distinguishable, except at the centre of the plates where the periphragm forms the mesotabular processes, endophragm and periphram appressed elsewhere. The processes are relatively robust, slim, and long (up to c. 4/3 of central body diameter), hollow with round cross-sections; the process stems are long (c. 85% of total process length), approximately constant in width or slightly tapering. The process terminations are short (branching occurs in distal c. 15% of the processes), but wide, distinctly expanded laterally (process termination width ¼ c. 50% of total process length), forming perforate or fenestrate platforms that consist of networks of simple, branched, or interconnected spines, distally united by continuous trabecular ring bearing 5-9 short (c. 3-6 mm) free spines, perpendicular to process stems or slightly curved towards the central body ( Figure 6; Plate 6, figures 3, 10, 11). The processes are generally comparable in size and shape, but apical, ps, and 1p processes are commonly slimmer; the process formula is 4 0 , 6 0 ', 5 0 '', 1p, ps, 1 0 '''. The archaeopyle is apical (type A 1-4 0 ) with angular margin, operculum is usually free.

Remarks.
To date, surface tabulation of the central body in the genus Oligosphaeridium has only been recognised in O. abaculum Davey 1979. Its tabulation corresponds to the formula pr, 4 0 , 6 0 ', 6c, 6 0 '', 1p, 5 s, 1 0 ''' (Davey 1979, p. 430). As in O. abaculum, the process formula in O. araneum sp. nov. does not reflect the full tabulation: some processes are absent (pr, 1 0 '', all of the cingular and most sulcal processes; see Plate 6). The outer surface of the periphragm in O. araneum sp. nov. is finely rugulate (Plate 6, figure 16), but the rugulate nature of the ornament is only clearly visible under SEM (the wall surface may appear smooth to finely granular under TLM). Oligosphaeridium pulcherrimum in Herngreen et al. (1998, pl. 4, fig. 2) has slim, relatively long processes, the distal extremities of which terminate with a regular trabecular margin bearing short spines. These features do not correspond to the original concept of O. pulcherrimum (see Deflandre and Cookson 1955, p. 270, 271, pl. 1, fig. 8, text-figs 21, 22); instead, the specimen probably belongs to O. araneum sp. nov.
Comparison. Oligosphaeridium araneum sp. nov. is easily distinguished from other species of the genus by its slim and long process stems and characteristic process terminations. shorter than the central body diameter, with wider and shorter stems and a pronounced branching of processes starting at c. 1/2-2/3 of the total process length. Furthermore, while process extremities in both species are fenestrate, in O. pulcherrimum they terminate with an irregular margin, rather than with ring-shaped trabeculae as in the new species. Oligosphaeridium perforatum (see Plate 7, figures 9, 10) has relatively short processes, with their distal tips in the form of irregularly perforated platforms with a ragged margin devoid of spines. The processes in O. perforatum? colum have relatively short stems, and their distal extremities are significantly more expanded and highly fenestrated.

Recorded stratigraphical range.
Middle upper Campanian-lowermost Maastrichtian, 'Inoceramus' altus Zone-lower Endocostea typica Zone (species recorded in all sections studied). Other occurrences. The Netherlands: Curfs quarry, sample RGD08 (lowermost Danian)as Oligosphaeridium pulcherrimum in Herngreen et al. (1998, pl. 4, fig. 2 Remarks. The inner cyst wall in Samlandia may be very thick (up to 6 mm in the type species; see Eisenack 1954), and is described as either one wall (e.g. Eisenack 1954; Stover and Evitt 1978 specified it as autophragm), or as two wall layers (endo-and periphragm; e.g. May 1980;McMinn 1988). The inner wall in S. mayi, when observed in cross-sectional view under TLM, appears to consist of two layers (Plate 10, figures 10, 11), but in S. carnarvonensis a SEM study revealed the presence of only one layer (Plate 10, figures 17, 18), referred to here as an autophragm. The ectophragm in Samlandia is thin and fragile, and may be discontinuous (see Stover and Evitt 1978;Damassa 1984;McMinn 1988).
The nature of archaeopyle formation in Samlandia species studied herein (S. carnarvonensis, S. mayi, S. cf. vermicularia) is similar to that in Samlandia paucitabulata Niechwedowicz in Niechwedowicz & Walaszczyk 2021 (see Niechwedowicz and Walaszczyk 2021, pl. 1, figs 9, 10). The ectophragmal piece of the operculum ('ectoperculum') is smaller than a part of the operculum formed from the inner cyst wall (Plate 11, figures 17-22). Similarly, the ectophragmal edges of the 'ectoarchaeopyle' form a smaller opening than that formed in the inner cyst wall. Consequently, the ectophragmal edges of the archaeopyle are usually collapsed into the cyst (Plate 10, figures 7, 16-18; Plate 11, figure 4), making the observation of wall structure at the archaeopyle edge difficult.
Remarks. The inner wall in Samlandia carnarvonensis is relatively thick (c. 1.5-2 mm) and homogeneous, and appears to consists of an autophragm (Plate 10, figures 17, 18). Its outer surface is shagreenate to very finely granular, giving the wall a smooth appearance under TLM. The autophragm ornamentation consists of a dense network of low (c. 1-2 mm high) non-tabular crests and short pillars or processes (c. 1-3 mm high) arising from the crests (the total ornament height is c. 2-5 mm) and supporting the ectophragm (Plate 10, figure 19; Plate 11, figures 11, 12). The ectophragm is very thin (c. 0.1-0.2 mm) and fragile (e.g. Plate 10, figures 14, 16-19) and homogeneous to finely spongy, with a densely granular outer surface (granulae c. 0.1-0.2 mm in diameter; see Plate 11, figure 6). The apical horn is commonly present and well developed (4-6 mm in length; Plate 10, figures 14-18; Plate 11, figures 2, 4, 5). The cyst ornament in the equatorial area may occasionally be higher (Plate 11, figure 2), which suggests the position of the cingulum; however, no indications of a paratabular ornament alignment were observed.
The connection between the ectophragm and its supporting structures appears to be rather weak, since the species was occasionally found with partially or completely, possibly mechanically (?), detached ectophragms (Plate 11, figures 7-14). Such forms are very similar to Pyxidinopsis bakonyensis (G ocz an 1962) Stover & Evitt 1978;Pyxidinopsis Habib 1976 differs from Samlandia in the absence of the ectophragm and the apical horn (see e.g. Stover and Evitt 1978). The ectophragm in S. carnarvonensis is located very close to the autophragm; consequently, it may be difficult to differentiate under TLM. The issues associated with distinguishing between Pyxidinopsis and Samlandia have been discussed in previous works (Antonescu et al. 2001a, p. 239;Antonescu et al. 2001b, p. 258, 259). In light of the above considerations and the original illustrations, the forms assigned to Pyxidinopsis bakonyensis by Siegl-Farkas (2001, pl. 1, figs 10-12) and Antonescu et al. (2001a, pl. 2, figs 41-44) could possibly be conspecific with S. carnarvonensis. One of the specimens illustrated by Siegl -Farkas (2001, pl. 1, fig. 12) appears to possess an apical horn. P. bakonyensis was originally described from the upper Maastrichtianas opposed to the Campanian-Maastrichtian boundary intervalwhich may further reinforce the proposed reinterpretation.
Samlandia carnarvonensis can be distinguished from S. mayi by its smaller size andusuallythe presence of a well-expressed apical horn. The ornament covering the inner wall in S. carnarvonensis forms a finer and almost regular 3 Plate 6. Dinoflagellate cysts from the upper Campanian-lowermost Maastrichtian of the Middle Vistula River section, central Poland. Photomicrographs 1-14 taken with a transmitted light microscope; 15-20 taken by scanning electron microscopy; scale bars ¼ 20 mm (unless otherwise specified). 1-20. Oligosphaeridium araneum sp. nov. 1-4. Holotype, MWGUW ZI/90/KS3/0, 2.1/11.7, EF N18/0; apical view of apical (1) and antapical (4) surfaces, and optical section (2, 3). 5-8. MWGUW ZI/90/Po11/0a, 17.4/12.8, EF M33/2; dorsal view of dorsal (5, 6) and ventral (7, 8) surfaces. 9, 10. MWGUW ZI/90/Po12/0, 6.6/11.7, EF N22/2; apical view of apical (9) and antapical (10)  Remarks. The inner cyst wall is thick (2-3 mm) and composed of either autophragm or appressed endo-and periphragm; further SEM study is needed to solve this issue. TLM preliminarily suggests that the endo-and periphragm are present, with the endophragm thinner (c. 1 mm) and homogeneous, and the periphragm thicker (c. 1.5-2 mm) and less homogeneous in structure (Plate 10, figures 10, 11), giving the wall a smooth appearance. The ectophragm is homogeneous to finely spongy in structure, with a smooth to shagreenate outer surface; it is very thin (c. 0.2-0.5 mm), fragile, and prone to folding (see Plate 10, figures 7-9). It is relatively distant from the inner cyst wall, and it is supported by irregularly arranged, c. 6-10 mm high, homogeneous to finely spongy discontinuous muri and processes, distally connected by an arch-shaped trabecula. As a result, the thin ectophragm has an irregularly wavy appearance (Plate 10, figures 2, 6, 7). The nature of the holocavate wall structure in S. mayi commonly gives the cyst a raspberry-like appearance under TLM (Plate 10, figures 1, 2, 12). The processes may be membranous or connected proximally (Plate 10, figures 8, 9), but unlike S. carnarvonensis the ornamentation does not form a regular reticulum covering the inner cyst wall. The ectocoel cavations are generally of uniform height, but occasionally may be higher at the apex, antapex, and equatorial area (Plate 10, figures 2, 3). The greater height of cyst ornamentation in the equatorial area may suggest the position of the cingulum, but no signs of paratabular ornament alignment were observed. As in other Samlandia species, the only feature indicating the tabulation is the archaeopyle (operculum free or attached). The apical horn is rarely distinguishable (Plate 10, figure 12), although its presence may be obscured by the characteristic nature of holocavation in this species.

Synonymy.
1973 Cyclonephelium expansum Corradini: p. 161, 162, pl. 24, figs 8a, b, text-fig. 7. 1992 Cyclonephelium? castelcasiense subsp. prominentum Marheinecke: p. 73, 74, pl. 15, figs 1-4. 1997 Glaphyrocysta expansa (Corradini 1973) Roncaglia & Corradini: p. 187,189,pl. 4,figs 5,[7][8][9] Remarks. Glaphyrocysta expansa is characterised by a wide membrane that arises from the lateral and antapical cyst margins; it is distinctly expanded antapically and laterally, but is commonly widest near the archaeopyle edge (Plate 12, figures 2, 5). The membrane is generally smooth, which is most evident at its distal portion (Plate 12, figures 2, 3, 6), although the membrane may be foveolate proximally, near the contact with the central body. The membrane is supported by the acuminate processes, which are apparently incorporated within it; the processes taper distally and gradually become lost in the membrane. The central body commonly bears two antapical lobes, of which the left one is often more expressed (Plate 12, figure 3). The cyst surface is finely granular, usually with a denser granulation found on the dorsal side along the plate boundaries (Plate 12, figure  1). Corradini's illustration (1973, text- fig. 7) of the paratype is probably the best figure to elucidate the most distinctive features of this species. Glaphyrocysta perforata Hultberg & Malmgren 1985 also possesses signs of tabulation on the dorsal surface of the central body and a smooth membrane. However, its membrane usually bears large perforations (holes) that are regularly distributedthat is to say, they only occur proximally at the contact of the membrane with the central body. If the perforations are numerous, the membrane appears to be connected to the central body by short simple or membranous processes.
Glaphyrocysta castelcasiensis prominenta (Marheinecke 1992) Michoux & Soncini in Fauconnier & Masure 2004, based on the description and illustrations of Marheinecke (1992), is a junior synonym of G. expansa. Marheinecke (1992, p. 74) compared the two species, but the differences he listed (degree of membrane expansion and its ornamentation, outline of antapex), while not inaccurate, are probably not substantial enough to differentiate these forms. For instance, G. expansa may also have a bilobated and asymmetrical antapex, in contrast to his suggestions (compare Corradini 1973, p. 161, 162).
Although G. expansa is similar to Glaphyrocysta pala comb. nov., it differs from the latter by possessing a smooth, unperforated membrane, which is distinctly expanded laterally (for a full comparison, see below).
Remarks. Kirsch (1991, p. 128) questionably referred this species to the genus Riculacysta; however, he did not specify what his concerns were with his tentative assignment. Fauconnier (in Fauconnier and Masure 2004, p. 473) suggested it was based on the dorso-ventral compression of the central body in his species. The central body in Riculacysta is subspheroidal, and connected with the membrane by processes arising from the antapical, lateral, and mid-ventral cyst areas (see Stover 1977, p. 76, 77, pl. 2, figs 22, 25, 28). In contrast, Glaphyrocysta has generally marginate processes and/or membranes occupying the antapical and lateral periphery of a dorso-ventrally compressed central body (Fensome et al. 2009, p. 32). The present specimens are quite comparable with  material, and confirm that Glaphyrocysta pala comb. nov. has a lenticular central body. Its antapical and lateral peripheries are connected with the membrane by short processes, but it lacks the processes on the mid-ventral cyst area. Consequently, it is transferred to the genus Glaphyrocysta.
Comparison. Glaphyrocysta pala comb. nov. is most comparable with Glaphyrocysta expansa (Corradini 1973) Roncaglia & Corradini 1997 and Glaphyrocysta semitecta (Bujak in Bujak et al. 1980) Lentin & Williams 1981. Glaphyrocysta expansa differs in having a smooth, unperforated membrane expanded antapically and laterally. In G. pala comb. nov., the membrane arises from the dorsal cyst side as in G. semitecta, and is typically shovel-shaped (not expanded near the archaeopyle edge; see Plate 12, figures 8, 9) and clearly perforate, with perforations irregular in size, shape and distribution (see Plate 12, figures 11, 13, 16). Additionally, G. pala comb. nov. has a characteristically low membranous flange surrounding the dorsal edge of the archaeopyle (Plate 12, figure 10). The dorsal surface of the central body is reticulate and generally lacks any sign of tabulationonly the cingulum is rarely discernible. The perforations in the membrane are a characteristic feature of G. pala comb. nov., and are similar to those present in G. semitecta, although in the latter the perforations are typically larger. Moreover, in G. semitecta the distalmost portion of the membrane is connected with the central body through long processes arising from the antapical and lateral cyst peripheries. Glaphyrocysta pseudoreticulata Vieira et al. 2017 differs from G. pala comb. nov. in lacking a well-developed membrane, and in possessing more numerous processes, that are grouped into linear or arcuate complexes, usually connected distally by irregular trabeculae.
Stratigraphical comments. Glaphyrocysta pala comb. nov. first appeared in the latest Campanian with some other species of the genus (G. expansa, G. aff. semitecta, and Glaphyrocysta sp. A; Niechwedowicz and Walaszczyk 2021). This event illustrates the earliest significant radiation of the genus, which otherwise is more typical of the Paleogene (e.g. Fensome et al. 2016Fensome et al. , 2019b. The wide geographical distribution of G. pala comb. nov. (see, e.g. Slimani 2001;Torricelli and Amore 2003;Mohamed and Wagreich 2013)  Remarks. Some of the Glaphyrocysta specimens studied herein strongly resemble G. semitecta (Bujak in Bujak et al. 1980) Lentin & Williams 1981. Such forms have a significantly dorso-ventrally compressed central body, which is circular in outline (with an absence of antapical or apical lobes), and is distinctly reticulate on both the ventral and dorsal surfaces.
The perforate membrane appears to arise from the dorsal cyst surface and turn towards the ventral cyst side. Its distal portion is connected with the central body by long processes (Plate 12,figures 17,18 (7), and optical section (8); note the lack of contact between the ectophragm (ec) and the processes arising from the autophragm (a). 9-12. MWGUW ZI/90/P9/SEM/2, EF E18/1; damaged specimen; oblique dorsal view of dorsal surface (9) and optical section (10), and oblique ventral view of ventral surface (11); close-up (12) shows the ornament of the outer surface of the autophragm (a dense and regular network of low muri and short processes); note the almost complete lack of ectophragm (ec), only partially preserved; note also the lack of any tabulation features other than the archaeopyle (indicated by arrows is a characteristic Eocene-Oligocene species (e.g. Bujak et al. 1980;Williams et al. 2004). The present specimens either prove that (i) G. semitecta first appeared in the Campanian (see also Mohamed and Wagreich 2013), or (ii) constitute a separate, although similar species. The latter option seems to be more reasonable in the presented estimation.

Remarks.
A single specimen from the collection studied represents a form different from all of the other known species of the genus, and is referred to here in open nomenclature as Glaphyrocysta sp. A. Its central body is circular in outline, with one weakly marked asymmetrical (left) indentation. It has a wide, highly perforate membrane similar to Glaphyrocysta semitecta, arising from a transverse furrow present on the dorsal surface of the central body, likely representing the cingulum. The membrane originates from the lower margin of the dorsal transverse furrow, surrounds the cyst antapex, and turns towards the ventral side, where it is the most distant from the central body. As in G. semitecta, the distal portion of the membrane is connected to the antapical and lateral cyst peripheries by long processes; however, in contrast to G. semitecta the connection is not formed by single processes, but by linear complexes composed of 3-5 solid processes (Plate 12, figures 21, 22). The membrane arising from the upper margin of the dorsal transverse furrow runs towards the dorsal archaeopyle edge, where it forms a membranous flange (Plate 12, figure 23) analagous to that present in G. pala comb. nov. or G. semitecta. The process complexes in Glaphyrocysta sp. A are probably penitabular and arise from the posterior and lateral margins of pre-and postcingular plates, partially reflecting the tabulation.
In contrast to the description of Fensome et al. (2016, p. 50), the processes in Hystrichosphaeridium may proximally cover most of the underlying plate, which is exemplified by Figure 7. Schematic drawings showing the distribution and morphology (base shape and width, and relative length) of the dorsal projections in selected pyrodinioid species. A, Hystrichosphaeridium tubiferum; B, Hystrichosphaeridium salpingophorum; C, Hystrichosphaeridium proprium; D, Hystrichosphaeridium brevispinum stat. nov. (morphotype B); E, Alisocysta circumtabulata.
Hystrichosphaeridium brevispinum  stat Remarks. This species was originally designated as a variety of Hystrichosphaeridium tubiferum (Ehrenberg 1837) , and was subsequently treated as a subspecies (Lentin and Williams 1973). Marheinecke (1992, p. 61) proposed a new combination of this taxon (as Hystrichosphaeridium proprium subsp. brevispinum), arguing it had a considerable morphological resemblance to his Hystrichosphaeridium proprium, particularly in the shape of its processes. It is worth noting H. proprium was transferred to Hystrichokolpoma by Foucher in Fauconnier and Masure (2004, p. 283, 284); however, it is retained in Hystrichosphaeridium here (see below).
Hystrichosphaeridium brevispinum  stat. nov. bears very characteristic processes, which are short (c. 1/3 of central body width) but relatively wide. These are the primary feature distinguishing between this species and H. tubiferum. The morphological distinction between these two species is even greater than that between H. tubiferum and H. salpingophorum, or between H. salpingophorum and H. proprium (see e.g. Marheinecke 1992, p. 58). There seem to be no objective reasons to treat this taxon a subspecies of H. tubiferum: consequently, it is raised here to species rank.
The two morphotypes of H. brevispinum stat. nov. differ significantly from each other: they represent either two separate species (see also Ioannides 1986, p. 25) or considerable intraspecific variation. Davey and Williams (1966, p. 58) did not provide a detailed description of H. brevispinum stat. nov., and the shape of the process cross-sections cannot be clearly inferred from illustrations of the holotype alone (Davey and Williams 1966, pl. 10, fig. 10; Bujak et al. 1980, pl. 8, figs 10-12;Fauconnier and Masure 2004, pl. 47, figs 1-4). A more thorough examination of the type material is needed before a final taxonomic decision is reached.

Recorded stratigraphical range.
Middle upper Campanian-lowermost Maastrichtian, 'Inoceramus' altus Zone-lower Endocostea typica Zone (recorded in all sections studied except for Kłudzie South). closed distally (e.g. Williams and Downie 1966, p. 176; circular bases, constant width and length (the latter roughly equal to the central body diameter), and terminate with 5-6 short (c. 5 mm) spines commonly curved towards the central body. The monotonous process development distinguishes H? recurvatum from H. tubiferum, H. proprium, and H. salpingophorum, in which the cingular, sulcal, and apical processes are slightly slimmer than the others; the difference is especially evident in comparison with the dorsal postcingular processes. Furthermore, H? recurvatum differs from other Hystrichosphaeridium in the number of processes. It may have more processes in particular latitudinal series, but the pr process is absent (Plate 9, figures 16, 17). In H? recurvatum, cingular processes are commonly more densely spaced on the dorsal cyst side (Plate 9, figure 13), while those on the ventral side (Plate 9, figure 14) are sparsely arranged, which may suggest the presence of two processes per cingular plate or a variable number of processes per plate (the process formula of the specimen illustrated in Plate 9, figures 13, 14 is 4 0 , 6 0 ', 8c, 6 0 '', 1p, 5 s, 1 0 '''). The number of processes in other plate series may also be higher (the specimen lacking operculum, illustrated in Plate 9, figures 9-12, has 35 processes). In Hystrichosphaeridium, all of the processes are mesotabular (one per plate), and their number is invariant from specimen to specimen. Hystrichosphaeridium? recurvatum is also very similar to Fetchamium prolixispinosum  Pearce & Williams 2018 in overall morphology: the latter also has processes of uniform size and morphology, but with a reduced number of cingular and sulcal processes (0-4c, 1-3 s; see Pearce and Williams 2018, p. 19). Consequently, H? recurvatum is questionably left here in Hystrichosphaeridium.
Remarks. Hystrichosphaeridium salpingophorum has a subspheroidal to ovoidal central body. The processes are mesotabular, with expanded bases and tips; in individual specimens, they are comparable in length (c. 2/3 of central body width) but variable in width (as in H. tubiferum or H. proprium). The process width generally depends on the plate series from which they derive. The slimmest are the processes of the sulcal, cingular, and apical series (especially 1 0 and 4 0 ), as well as the processes corresponding to plates 1 0 '' and pr. The rest of the postcingular and precingular processes are roughly comparable in width (3 0 ''-5 0 '' are usually the widest processes; see Plate 8, figure 5). The shapes of the process bases and terminations roughly reflect the shapes of the plates, so the pre-and postcingular processes are subangular or subquadrate, while the cingular processes have oval cross-sections ( Figure 7B; Plate 8, figure 5). Distally, the processes bear spines that are prolonged from the faint longitudinal striae, with the most prominent spines (3-5 mm in length) arising from the corners of subquadrate process termination (Plate 7, figures 21, 24).
The main feature that differentiates this species from H. tubiferum is the subquadrate termination of some of its processes (e.g. Deflandre and Cookson 1955;Davey and Williams 1966;Ioannides 1986;, although the distinction may not necessarily be obvious (see Davey and Williams 1966, p. 62;Ioannides 1986, p. 26). In the present material, the process stems and terminations were commonly deformed (withered in appearance) in both species, which obscures their true shape and renders their differentiation impossible. In this respect, the material studied here is quite comparable with that of Marheinecke (1992). However, the illustration of the holotype [Deflandre 1937, pl. 13 (also labelled pl. 10), fig. 1] clearly indicates that the process bases and terminations are of the same shape. The pre-and postcingular processes in H. salpingophorum have subquadrate or subangular bases, whereas they are rounded in H. tubiferum ( Figure 7A,B). Thus, the shape of process bases is considered here the most valuable feature in distinguishing between these two species (compare also Wrenn and Hart 1988, p. 355;Marheinecke 1992, p. 59, 63).
Remarks. Hystrichosphaeridium tubiferum is very similar to H. salpingophorum. Both species have processes of constant length, varying in width on a single specimen (see above) with expanded bases and tips: the main distinction is that all of the process bases (and terminations) in H. tubiferum are rounded (see Figure 7A). While the process bases are also rounded in H? recurvatum, its processes are relatively longer (as compared with the central body diameter), comparable in width, and have a different style of terminations. Furthermore, H? recurvatum differs in the number of processes and lacks the pr process (see above).
Under TLM, the central body wall of H. tubiferum is smooth to granular (as in H. salpingophorum), but SEM study proves that it is homogeneous to spongy in structure: it could be described either as an autophragm or as an endoand periphragm. The innermost part of the central body wall (c. 0.1-0.2 mm) appears to be more homogeneous, forming a continuous level that could be referred to as endophragm (see Plate 7, figure 20). If this is the case, then the thinwalled (c. 0.2-0.5 mm) processes in H. tubiferum are not formed from the entire periphragm (c. 1-1.5 mm thick), but only from its outermost portion. In any case, the process walls in H. tubiferum are distinctly thinner than the central body wall (compare, e.g. with the thickness of walls in Kleithriasphaeridium loffrense; Plate 9, figures 18-21).

Conclusions
The new dinoflagellate cyst species Oligosphaeridium araneum sp. nov. is consistently present throughout the succession, appearing first in the middle upper Campanian ('Inoceramus' altus Zone). Outside central Poland, the species (illustrated as Oligosphaeridium pulcherrimum in Herngreen et al. 1998) is known from the lowermost Danian of the Netherlands and potentially has stratigraphical significance.
In the new combination Glaphyrocysta pala comb. nov., the species pala is transferred from the genus Riculacysta to Glaphyrocysta, based on its lenticular (dorso-ventrally compressed) central body and the absence of mid-ventral processes connecting the central body with the membrane. The lenticular shape of the central body was already noticed by Kirsch (1991, p. 128), apparently leading him to provisionally attribute the species to Riculacysta (Fauconnier in Fauconnier and Masure 2004, p. 473). Differentiating between Samlandia carnarvonensis and S. mayi has been troublesome to date (Antonescu et al. 2001a(Antonescu et al. , 2001b. Using both TLM and SEM microscopy techniques, it is possible to identify the most evident features differentiating the two species: namely, the height of the ectocoel cavations and the inner cyst wall ornament. In S. mayi, the ectophragm is significantly more distant from the inner cyst wall (? endo-and periphragm, or an autophragm) and connected to it by the discontinuous muri and long processes distally joined by arch-shaped trabeculae, giving the impression of an imperfect (commonly discontinuous) and sparse reticulum. In contrast, the inner cyst wall (an autophragm) in S. carnarvonensis is covered by a dense network of low crests forming a clear and regular reticulum. The ectophragm, which in S. carnarvonensis is located distinctly closer to the autophragm (the holocavate nature of the wall structure may be difficult to recognise under TLM), is supported by thin and short processes arising from the muri. The connection between the ectophragm and the process tips in S. carnarvonensis is, however, rather weak. Consequently, the ectophragm can be mechanically separated. Such specimens lacking the ectophragm may be confused with Pyxidinopsis bakonyensis. Samlandia mayi is more similar to Samlandia paucitabulata, another taxon characteristic of the upper Campanian. Samlandia paucitabulata differs from S. mayi in having signs of tabulation reflected by the arrangement of wall features, while the latter is atabulate. Furthermore, S. paucitabulata possesses a distinctly thicker, more robust, and regularly outlined ectophragm.
The distribution of processes, sutural ridges, and septa in Callaiosphaeridium, documented here using TLM and SEM, revealed that the tabulation pattern in the genus is characterised by a sexiform antapex, L-type ventral configuration, and neutral torsion of hypocyst, which enables the attribution of the genus to the subfamily Leptodinioideae. Moreover, all of the processes (including the tubular ones) in Callaiosphaeridium are shown to be gonal in position. A leptodinioid tabulation pattern was also recognised in Amphorosphaeridium, based on the distribution of the processes in A. fenestratum. Amphorosphaeridium and Callaiosphaeridium, and their type species, are emended accordingly.
An analysis of the genera Hystrichosphaeridium and Hystrichokolpoma suggests that the species proprium should be re-transferred back to Hystrichosphaeridium, which accords with the original affiliation proposed by Marheinecke (1992). The arrangement of processes in this species clearly reflects a pyrodinioid tabulation, rather than a cribroperidinioid pattern (the latter is typical for Hystrichokolpoma). Consequently, the transfer of H. proprium to Hystrichokolpoma proposed by Foucher (in Fauconnier and Masure 2004) is here rejected. Additionally, one new status, H. brevispinum stat. nov., is proposed. This taxon is raised to the species level on the basis of the distinct morphology of its processes, which concurs with features widely used to differentiate species within the genus. Hystrichosphaeridium? recurvatum, another morphologically related chorate species, is questionably left in Hystrichosphaeridium, although it possesses a variable number of processes between individual specimens, the common presence of more than one process per plate, and the lack of a preapical process. These features are not typical for Hystrichosphaeridium.