Re-study of some dinoflagellate cysts from the Oligocene and Miocene of Germany

The type material of six dinoflagellate cyst species from the Late Oligocene to Middle Miocene of northwest Germany, described originally by Gerlach (1961), is reillustrated and redescribed. It is shown to include representatives of nine species. Areosphaeridium (ex: Baltisphaeridium) pectiniforme is found to be a senior synonym of Areosphaeridium multicornutum Eaton. Systematophora placacantha is considered to be a senior synonym of Cleistosphaeridium (ex: Baltisphaeridium) panniforme (Gerlach). The new combination Rhynchodiniopsis tenuitabulata (Gerlarch) is proposed. Revised diagnoses for these three species and for Leptodinium membranigerum (Gerlach), Achomosphaera triangulata (Gerlach) and Lejeunecysta hyalina (Gerlach) are proposed. The morphology of a form described here for the first time, and tentatively attributed to Phthanoperidinium, is considered perhaps to imply a separate origin for the Phthanoperidiniaceae: for that reason familial, rather than tribal, rank is preferred for that group. The stratigraphical ranges of the nine species here recognised and of two others of Gerlach’s species redescribed in earlier papers are detailed; elimination of misattributed forms means that these ranges are shorter than the published literature suggests.


INTRODUCTION
The history of early studies of Oligocene and Miocene dinoflagellate cysts has been summarised in an earlier paper (Sarjeant, 1983). Whilst Dorothea Maier was completing her studies of cysts from borehole cores and outcrops in the Niederrheingebiet, Nordrhein-Westfalen, Schleswig-Holstein and Jutland (Denmark) at the University of Kiel, Ellen Gerlach was examining assemblages from two boreholes through sediments of similar age in the Emsland region of northwest Germany at the University of Tubingen. Maier found extreme difficulty in dealing with the diverse morphologies with which she found herself confronted and her research supervisor, Walter Wetzel, had too poor a comprehension of these microfossils (as his own papers make evident; see Wetzel, 1952Wetzel, , 1955 to give her the guidance she needed. In contrast, though Ellen Gerlach's research supervisor, Otto Schindewolf, was not himself a micropalaeontologist, she had the great advantage of being able to consult Alfred Eisenack for advice when in difficulties. In consequence, whilst Maier's work (1959) was to occasion confusion among other palynologists and to be the focus for much criticism, Gerlach's single paper (1961) was of higher quality and has proved of greater utility.
Four new genera were proposed by Gerlach. One of these, Pentudinium, has come to be regarded as of great importance in Tertiary stratigraphy (see Benedek, Gocht &Sarjeant, 1982 andEdwards, 1982, for recent interpretations). Another, Lejeunia, proved to be the junior homonym of a modern liverwort (see Artzner & Dorhofer, 1978). However, under the substitute name Lejeunecysta Gerlach's genus remains one that is widely recognised; its type species is reillustrated, and its distribution discussed, herein. Her third genus, Membranophoridiurn, fell into disuse but has been effectively reinstated by Stover & Evitt (1978), under a revised definition. A misunderstanding concerning the morphology of her fourth genus, Emsfundiu, caused it to be formally abandoned for a while; however, recently it has been likewise reinstated under a revised definition (Benedek & Sarjeant, 1981).
In contrast, the species assigned by Gerlach to genera already proposed before 196 1 have falled largely into disuse; only one of them, Spiniferires (ex: Hystrichosphueru) cornutus (Gerlach, 1961) Sarjeant, 1970, has been regularly reported. There are three reasons for this: first, because the advances in knowledge of dinoflagellate cysts since 1961 mean that her diagnoses can be no longer considered adequate for confident identification of those taxa; second, because her illustrations were produced at too small a magnification; and third, because her holotypes and paratypes are, in some instances, markedly dissimilar in morphology.
My work on Gerlach's holotypes was undertaken during visits to the University of Tubingen in 1979 and1981. During those visits, I received continuous and courteous co-operation from Dr. Hans Gocht, many of whose percipient comments are reflected in the pages that follow. Six of Gerlach's species are reillustrated, and five of them redescribed, herein; nine taxa are recognised among them. I hope to complete work on her other types during a future visit to Tiibingen.
The seventeen samples from which Gerlach's assemblages were obtained stemmed from two shallow borings, Emsbiiren 7 (mapsheet Lohne 3509, R = 2586100/H = 54812830) and Emsbiiren 9 (mapsheet Lohne 3509, R = 2588090/H = 5808343), put down through the Oligocene and Miocene sediments of Hannover in the triangle of land bounded by Nordhorn, Lingen and A B C D Rheine. Full details of the petrography of the samples and of the method of preparation are given by Gerlach All described specimens are lodged in the Gerlach collection, Institut fur Palaontologie, University of Tubingen. Their present condition is highly variable. Some remain in as good order as when Gerlach described them, but others have split open or suffered in varying degree from bacterial or fungal attack, and two could not be located on the slide supposed to contain them. Details are given in the ensuing pages and lectotypes designated when necessary. (1961, pp. 146-149).
Paraplate 4' is quite large and asymmetrically pentagonal, having a fairly long boundary with a quadrate 6"; this boundary intersects that of 1' in a position anterior to that of 1' with the sulcus. All precingular paraplates are larger than their postcingular equivalents. The right boundary of the small, elongate paraplate 1"' is poorly marked; for this reason the sulcus, which broadens posteriorly, may appear to have the form of an inverted, broad-hafted axe with its blade to the right. Faint lines divide the sulcus into one anterior, at least four median and one (or two?) posterior paraplates. Two posterior intercalary paraplates of similar size separate the sulcus from the rather small antapical paraplate. The cingulum is of moderate breadth, forming a feeble laevorotatory spiral such that its two ends differ in anteroposterior position only by the cingulum's breadth. Surface of phragma uniformly granulate.
Archaeopyle single-plate precingular (type P) formed by the opening or loss of paraplate 3"'.
Holotype. Preparation 1170/14(676), illustrated by Gerlach, 1961, pl. 26, figs. 1-3, lodged  Remarks. The holotype was not found, the presence of what appears to be the collapsed and decayed residue of a cyst suggesting that it has disintegrated. Of the two paratypes cited by Gerlach, one (Pr. 1170/15) could not be identified with certainty: three examples of L . rnernbranigerurn, all in poor state, were present on this slide but none of the three could be identified with Gerlach's illustration (1961, pl. 26, fig. 4). The other paratype was located and is here reillustrated, but it has collapsed and is in lateral view (PI. 2, figs. 1-2, Fig. 1C-D). Moreover, it contains the detached operculum of another dinoflagellate cyst with dissimilar ornament (see PI. 3, fig. 3). For these reasons, it was considered unsuitable to serve as lectotype. Instead, another specimen figured, but not named as a paratype, by Gerlach (1961, pl. 26, fig. 7) was chosen. It is in dorsoventral view and, though split open at the left side, remains as well preserved as when she illustrated it.
This species was allocated tentatively to Impagidinium by Stover & Evitt (1978); but the relative size and the shapes of paraplates 4' and 6", and their relation to paraplate l', do not accord with the diagnosis of that genus. Yet the occurrence of a species of Leptodinium in sediments as young as the Late Oligocene is surprising. Since L . membranigerum has not yet been reported from Tertiary sediments elsewhere, the possibility that these specimens were reworked from more ancient strata cannot be ruled out. However, since this species has not been reported from earlier stratigraphic levels and since no other reworked species are present, there is as yet no good reason to doubt that it is indigenous.
Emended Diagnosis. Cyst proximate, holotabulate, apically cornucavate (monocornucavate). Ambitus spheroidal to broadly rounded-subpolygonal, with epitract and hypotract of almost equal shape and relative size. Apical horn short, tapering and blunt. Low, narrow crests, entire to undulate distally, delimit the paraplates: the crests on the horn may impart to it a trifid appearance. Paratabulation 2pr, 4', 6", 6c, 6"', Ip, l"", ?8s. Paraplate 4' is quite large and asymmetrically quadrate, having a long boundary with an almost rectangular 6"; their mutual boundary intersects that of paraplate 1' just anterior to the junction of the latter paraplate with the sulcus. Paraplate 1' broadens considerably in its posterior portion. Paraplate 3" is unusually small, whereas paraplates 2' and 3' are somewhat larger than usual. Paraplate 1"' is reduced and elongate and 2"' reduced and quadrate, both paraplates being separated from the antapex by paraplate lp. Cingulum of moderate breadth and degree of spirality, its two ends differing in anteroposterior position by 1 1/2 times its width. The sulcus is moderately broad, extending from mid-point on the epitract to the antapex; it is subdivided by faint lines into at least six, perhaps eight or more, small paraplates. Surface of phragma granulate. Archaeopyle single-plate precingular (type P), formed by loss of paraplate 3": operculum free. Type Horizon and Locality. Middle Oligocene, depth 179m. Emsburen boring no. 7, northwest Germany.

Remarks.
The holotype was not found; however, the slide supposed to contain it included a specimen sufficiently similar to it in general form, to have been probably considered referable to this species by Gerlach. It is of closely similar size and shape, but lacks any trace of the archaeopyle so prominently seen in Gerlach's figure (1961, pl 25, fig. 11). It is here redescribed as Phthanoperidinium sp. In consequence, there are two possibilities First, that the holotype is contained in that slide, but has either disintegrated to unrecognisability or was simply missed in traversing. Second, that the slide was mislabelled, the specimen here illustrated (PI. 3 figs. 4-5) being mistaken for the holotype at some stage of Gerlach's work. Whatever the reason, the holotype must be considered lost and a lectotype selected.
Gerlach designated two paratypes (ibid., p. 159), though neither of them was illustrated. Only one of these was found; it is here designated as lectotype. Nevertheless, it is far from being ideal, containing dark organic material and showing some evidences of fungal attack, as well as being slightly folded at left. In consequence, the paratabulation could not be determined with complete confidence, the form and relative size of paraplates 2'" and Ip and the exact number and shape of the midventral sulcal paraplatelets remaining uncertain, as is indicated by the broken lines in my diagram (Fig. 2).
Fortunately, the anterior ventral and dorsal paratabulation could be confidently elucidated. Unusual features are the relatively small size of 3" (lost in archaeoformation) and, as a result, the proportionate enlargement of 2' and 3' (see PI. 4, fig. 3). An unusually broad paraplate 1' and a long, quite broad sulcus furnish additional distinguishing characters. The diagnosis is emended to stress these characteristics and to include mention of the presence of preapical paraplates. The anterior ventral paratabulation demonstrates that this species belongs, not in the genus Millioudodinium as presently defined, but in Rhynchodiniopsis.
None of the specimens attributed to this species since Gerlach's time accords beyond doubt with the revised diagnosis. The German Upper Oligocene specimen compared with this species by Brosius (1 963) is the most similar, but the ambitus is rather more angular and the apical horn broader. The record from the early Miocene of the Grand Banks, offshore eastern Canada (Barss et al., 1979) is unsubstantiated by an illustration. Other records must be rejected. The Belgian Eocene specimen illustrated by De Coninck (1968) has an indistinct paratabulation, an ambitus and apical horn of dissimilar character and a much larger archaeopyle; it cannot be referable to this species. De Coninck's later record (1975) must, in consequence, also be rejected. Eaton's English Eocene specimen (1 976), illustrated in oblique ventral view, has a shorter apical horn, a small posterior ventral paraplate, a much larger paraplate 3" (lost in archaeopyle formation) and, in consequence, smaller dorsal apical paraplates. The specimens from the Early to Middle Miocene of Japan described by Matsuoka (1983) are too elongate, and have too dissimilar a ventral paratabulation, for retention in this species.
extraneous operculum that has become lodged within the split-open cyst ( x 1,000).
For the moment, Rhynchodiniopsis tenuitabulata can be considered, in view of the fact that two morphological types were assigned to it by Gerlach, to be known confidently only from the Middle Oligocene (with a questionable extension of range up to the Middle Miocene) and only from Germany.
Family Spiniferitaceae Sarjeant, 1970, emend. Sarjeant & Downie, 1974 Genus Achomosphaera Evitt, 1963 Achomosphaera triangulata (Gerlach, 1961 to 28% of the cyst breadth, sometimes appearing longer because of distortions in the orientation of their branches. Gonal processes slender, arising from broad bases but of very constant thickness between the base and the position of branching; trifurcate, their branches long and directed almost parallel (or at only a low angle) to the phragma surface, with bifid terminations, the branchlets typically recurved. Intergonal processes present in situations corresponding to the sutures between precingular and between postcingular plates; slimmer than the gonal processes and bifurcate, with extremely slender branches that may be simple or bifid.
Paratabulation not directly indicated. Surface of phragma coarsely granulate to shagreenate or verrucate. Archaeopyle single-plate precingular (type P), formed by loss of a portion of the cyst wall corresponding to plate 3". Operculum shield-shaped, free or attached. Dimensions. Holotype: length of central body 58pm, breadth 53pm, length of processes c. 12-15pm. Paratype A: length of central body 60pm, breadth 48pm, length of processes c. 12-14pm. Paratype B: length of central body 53 pm, breadth 42pm, length of processes c. 9-1 1 pm. Range of dimensions: length of central body 36-60pm (mean 53wm), breadth 33-48pm (mean 44pm). 12 specimens measured (material 36 specimens). [Note: Gerlach quotes unrealistically high lengths for the processes, up to 22pm. No specimens seen had processes of such high proportionate length].

Remarks.
The Spiniferites/AchomosphaeralNematosphaeropsis group of dinoflagellate cysts is a particularly long-ranging one and known to be especially prone to intraspecific variation; this was recognised by Lejeune-Carpentier (1937a, b), in her restudies of Ehrenberg's type material of Spiniferites ramosus (then Hystrichosphaera ramosa) and given emphasis by Davey & Williams' work on two of those genera (1966). Nevertheless, a plethora of new species names continue to be proposed and the task of sorting out this taxonomic mess grows ever more formidable. It seems likely that future studies of variation in Achomosphaera ramulifera (Deflandre, 1937) Evitt, 1963 may show A . triangulata to merit merely subspecific or varietal status. For the moment, however, it is retained as a separate species and distinguished from A . ramulifera by the slimness of its spines and the length of their branches.
Although not reported as such from other localities since its first description, A . triangulata may be recognised under different names in illustrations by some other authors. Gocht's form from the Late Eocene of Germany (1 969) is illustrated in polar view, but seems to correspond with A . triangulata in all particulars.
Benedek's single specimen from the Middle Oligocene of Germany, described asA. aff. triangulata, is stated to have only two process branches ('hooklets'); however, in all other respects it is closely comparable and may well prove, on restudy, to have trifurcate gonal processes. One of the forms from the Late Eocene of Victoria, Australia, illustrated by Cookson & Eisenack (1974)    Processes ranging in length between about 35% to 45% of the cyst breadth. Each process is distally expanded and bifurcate (licrate); the attitude of the bifurcations ranges from patulate to recurved. The two branches are of variable relative and absolute length; they vary in breadth from slender, with a denticulate distal margin, to broad, with a denticulate or irregular distal margin and with some development of fenestration. Paratabulation 4', 6", 7c, 6"', Ip, 5pa, 1"". The paraplates, however, are not always fully represented; the processes equivalent to 6" and Ip, and up to three of the cingular processes, may be lacking and the preantapical processes are often very incompletely developed (as few as one may be present in some specimens). The antapex, with its single process, is typically offset to the right of the midventral line. Surface of phragma laevigate to finely or more coarsely granulate. Archaeopyle apical (type tA): operculum (or opercular pieces?) free. (290), illustrated by Gerlach, 1961, pl. 28, fig. 14, text-fig. 18, and herein, pl. 1, fig. 2; pl. 4, fig. 2; lodged in the Gerlach collection, University of Tubingen.
Dimensions. Holotype (in apical view): diameter of central body 32pm, length of processes c. 1 3 p m .

Remarks. Even when Eaton first proposed his species
Areosphaeridium multicornutum, he was aware that it might prove to be a junior synonym of Gerlach's Baltisphaeridium pectiniforme, for he commented: "B. pectiniforme has been recorded by two writers, GERLACH (1 961) from the Middle Oliogocene of Germany, and BROSIUS (1963) from the Upper Oligocene of Germany. The specimens figured by these two writers have processes which are identical to those of A . [ reosphaeridium ] arcuatum [now A . dictyostilum (MenCndez, 1965) emend. Sarjeant, 19811 and A . multicornutum. It is possible that one of the species described from the Bracklesham Beds may be identical to B. pectiniforme, but the precise relationship between the English and the German forms cannot be determined until the number and distribution of the processes in B. pectiniforme is known" (Eaton, 1971, p. 364).
In the ensuing thirteen years, though the two species recognised by Eaton have been widely reported, Gerlach's material has remained unstudied. My examination of it shows Eaton's hesitancy to have been fully justified, since the two specimens reported by Gerlach represent, not one, but both species! The holotype, a thin-walled specimen in apical view and having numerous, very slender processes, corresponds in all particulars with Eaton's Areosphaeridium multicornutum. In consequence, the latter name must be regarded as a subjective junior synonym of Areosphaeridium pectiniforme and its use abandoned.
The paratype, in contrast, is thicker-walled, with more massive processes that are broader at base and tip than those of the holotype and are much less numerous. It is reattributed below to A . dictyostilum. Areaosphaeridium pectiniforme has a known range from middle Eocene to late Oligocene; specimens found in Miocene sediments by Barss et al. (1979) were considered to be reworked. It has been reported from Explanation of Plate 4 All figures are x 1000.
[A full synonymy for this species to 1980 is given by Sarjeant, 1981Sarjeant, , p. 1151Sarjeant, . 1981 Areosphaeridium arcuatum Eaton;Lentin & Williams,20. 198 1 Areosphaeridium dictyostilum (MenCndez); emend. Sarjeant, 1981, 115-116. Remarks. As noted above, the "paratype" of Gerlach's Baltisphaeridium pectiniforme proved to be a specimen of Areosphaeridium dictyostilum. It is in oblique polar view and is somewhat crushed and obscured (see PI. 4 fig. 4). This species has been reported hitherto only from the late Eocene (see Sarjeant, 1981, p. 116). Its occurrence in the Middle Oligocene may represent an extension of its range or (since only a single, poor specimen is present) may be a consequence of reworking. Certainly the record should be distrusted until further middle Oligocene specimens are reported.
Remarks. The holotype of Baltisphaeridium panniforme, as originally illustrated (preparation 11 70/55(315); Gerlach, 1961, pl. 28, fig. 13) was far from being an ideal specimen; it was badly obscured by adherent material and contained black pyritic inclusions. In the intervening years, it has deteriorated into unrecognisability and its exact character can no longer be determined. Two paratypes were named, though neither was illustrated. One of these, preparation 1170/56(585), though in the morphological range included in Gerlach's diagnosis, differs too markedly from the holotype to serve as a suitable lectotype (see P1. 2., fig. 4; PI. 4, figs. 1, 5, herein); it is discussed later. The other paratype, preparation 1170/56(624), though much obscured internally and externally by pyrite crystals, corresponds much more closely to the holotype. This is reillustrated here (PI. 3 fig. 1) and selected as lectotype.
The morphology of this specimen cannot be fully determined. Nevertheless, the nature of its processes and their arrangement into groups shows it to be a Systematophora; moreover, it is sufficiently similar in morphology to S. placacantha (Deflandre & Cookson) to be regarded as referable to that species. Accordingly, Cleistosphaeridium (ex: Baltisphaeridium) panniforme is here regarded as a subjective junior synonym of Systematophora placacantha and abandoned. May (1980) has proposed a revised diagnosis for S. placacantha which has tightened its definition considerably. This revision was based, not on the Australian Miocene type material, but o n latest Cretaceous (Maastrichtian) specimens from New Jersey. Whether all the specimens attributed to this species (the type material included!) accord with the emended diagnosis, remains to be determined.
The German Lower Eocene specimens allocated to Impletosphaeridium panniforme by Morgenroth (1 966) have a circular opening and processes of a very different nature; they cannot be included into S. placacantha. If all other records listed in the synonymy prove acceptable on rescrutiny, then this species has a stratigraphic range from latest Cretaceous to late Miocene. Its geographic range is wide; it has been reported from Germany (Gerlach, 1961 ;Benedek, 1975), doubtfully from Belgium (De Coninck, 1967), more positively from France (Gruas-Cavagnetto, 1972), England (Bujak, 1980), Denmark (Piasecki, 1980), Maryland and New Jersey, U.S.A. (Benson, 1976;May, 1980), Argentina (Heisecke, 1970) and Australia (Deflandre & Cookson, 1953;Verdier, 1970). In addition, it has been recorded from submarine sediments in the North Atlantic Ocean (Ioannides & Colin, 1977), the Grand Banks and Scotian Shelf, offshore eastern Canada (Williams & Brideaux, 1975;Barss et al., 1979), offshore from the southeastern United States (Stover, 1977), in the southwestern Atlantic Ocean (Goodman & Ford, 1983) and from the Tasman Sea off Tasmania (Haskell & Wilson, 1975). All records appear to come from sediments laid down in cool to warm temperate marine waters; it has yet to be reported from polar or tropical marine sediments.
Sub-order Peridiniineae Fott, 1959, emend. Bujak & Davies, 1983Family Phthanoperidiniaceae Drugg & Loeblich, 1967, emend. Bujak & Davies, 1983 Remarks. Certain features in the morphology, not only of the specimen described below, but also of existing species of this genus, are considered by me to cast doubt on  belief that Phthanoperidinium should be considered an infrafamilial grouping within the Family Deflandreaceae. Thus, though their emendation is adopted, I prefer to continue to rank the Phthanoperidinaceae as a separate family rather than reducing it to the status of a tribe.

?(pars) 1975 Gonyaulacysta giuseppei m a j o r
Description. Proximate, holotabulate cyst, monocornucavate. Ambitus ovoidal to rounded-subpolygonal, with an apical horn of moderate length. Phragma two-layered, but relatively thin and delicate. Epitract and hypotract of about equal size. Epitract almost exactly conical, with only a slight inbulge about the base of the apical horn; horn broad-based and slightly rounded at the tip, formed from periphragm only. Crests delimiting paraplates low and entire. Paratabulation ?lpr, 4', 3a, 7", 6c, 6'", lp, 2"". Paraplate 1' is relatively narrow, its boundary with the sulcus not clear in the specimen but certainly anterior to the junction of paraplates 4' and 7"-a junction of moderate dimension. All dorsal precingulars are reduced to accommodate the three dorsal intercalaries, paraplate 4" especially so and much broader than long. Paraplate 2a is the largest of these inteicalaries, linteloid and with side HI less than half the width of H,, whilst H, and I& are relatively elongate (see Bujak & Davies, 1983, text- fig. 3 for explantion). Cingulum broad and strongly laevorotatory, its two ends differing in anteroposterior position by more than twice its width. Paraplate 1"' is elongate and narrow, having an oblique boundary with elongate and relatively narrow lp, which also separates the quadrate paraplate 2"' from the antapex. Paraplate 6"' is relatively small, whereas 4"' is the largest of all the paraplates. The two antapical paraplates are both comparatively small, their mutual boundary poorly seen in this specimen. Sulcus broad, extending from apex to antapex and undivided. Archaeopyle not developed. Remarks. This specimen is present in the preparation labelled by Gerlach as containing the holotype of Rhynchodiniopsis (ex: Gonyaulax) tenuitabulata. Though it is similar in size and ambitus to the holotype (as illustrated by Gerlach, 1961, pl. 25, figs. 1 0 , l l ) it is certainly not that specimen, since it has no archaeopyle and is conspicuously different in paratabulation.
However, it appears certain that Gerlach attributed this specimen to R . tenuitabulata and quite probable that, at some stage, the preparations containing it and the holotype came to be inadvertently transposed.
The typical paratabulation pattern of Phthanoperidinium, both according to the emendations of Edwards & Bebout (1981) and the more recent emendation of Islam (1982), contains only five postcingulars and no posterior intercalary paraplate. However, this difference is not so great as may appear at first sight. Some species of Phthanoperidinium , for example P. brooksii Edwards & Bebout 1981 (see especially their text- fig. 3) show an inbend to the left of the sulcal margin of paraplate 1"' equivalent of paraplate 1"' of this specimen); moreover, that paraplate does not impinge upon the antapex, being separated from it by a space comparable to that occupied by paraplate l p in the German specimen. It seems, therefore, that the two left ventral paraplates seen in this specimen have been lost in later species of Phthanoperidinium.
If this interpretation is correct, then it is possible that Phthanoperidinium was not a product of the Deflandreoid lineage, but was instead a "parallel evolution", derived directly from the Gonyaulacoids. This might account for the fact, remarked on by Bujak & Davies (1983, p. 134), that its paratabulation is so unusually clear; and it calls into question the intrafamilial hierarchy proposed by those authors to include this genus. For that reason, I prefer to continue to treat the Phthanoperidiniaceae as a separate family, rather than reducing them to the rank of a tribe within the Subfamily Palaeoperidinioideae, as advocated by Bujak & Davies. Since only a single specimen was available for study, however (and in particular because its style of archaeopyle is unknown), no new generic or specific name is proposed for it. Remarks. The longitudinal surficial lines that characterise this species are difficult to interpret. They are too narrow, and too uniform, to seem a likely product of post-mortem shrinkage. These lines are absent from the hollow between the two antapical horns, nor do they seem to traverse the cingulum. It is likely that they include, or mask, traces of a paratabulation and it is possible that they reflect growth stages of the motile theca.
The style of archaeopyle formation in L . hyalina remained long a source for controversy.. Had Gerlach's illustrations been reproduced at larger scale, this might have been avoided, for an intercalary archaeopyle is present in the lower (dorsal) surface of the holotype itself (see P1. 3 fig. 2).
The Swedish Upper Cretaceous specimen attributed to Lejeunecysta (then Lejeunia) hyalina by Kjellstrom (1972) has a very much larger intercalary archaeopyle and appears to be tricornucavate. I consider it referable to Phelodinium rnagnificurn (Stanley, 1965) Stover & Evitt, 1978. The Late Cretaceous specimen figured by Wilson (1978) also corresponds better withPhelodinium magnificum . The Japanese Miocene forms described by Matsuoka (1983) are too elongate, and have antapical horns that are too short, for retention in L . hyalina; in ambitus they are perhaps closest to L . beninensis Biffi & Grignani, 1983, but they lack the tiny antapical hornlets which characterise that species. It is likely that Matsuoka's specimens should be placed into a new species of Lejeunecysta. Vozzhennikova's record (1 967) is unaccompanied by illustrations and must be distrusted accordingly. The NorwegiadGreenland Sea forms of Middle Oligocene to Early Miocene date reported by Manum (1976) are, from his illustration, too dissimilar in morphology to be retained in this species.

STRATIGRAPHICAL RESULTS
In Table 1 is given a list of species described originally by Gerlach and redescribed in this paper or earlier ones by Benedek & Sarjeant (1981) and Benedek, Gocht & Sarjeant (1982). For simplicity of reference, the arrangement adopted is alphabetical. (The justifications for the ranges quoted are set forth above or in the earlier papers). Species in Gerlach's assemblage that have not yet been re-examined are omitted from this Table. Two species, Leptodinium membranigerum and Rhynchodiniopsis tenuitabulata, and the form here styled ?Phthanoperidinium sp. are known as yet only from Gerlach's assemblage. One species, Systematophora placacantha, is long-ranging and another, Achomosphaera triangulata, potentially so. All other species appear to be of stratigraphical utility in the identification of middle to late Palaeogene strata at outcrop or in subsurface.

ACKNOWLEDGEMENTS
I am deeply indebted to the University of Tiibingen for making Gerlach's type material available for restudy and to Dr. Hans Gocht for his co-operation and valuable comments, both when we were working together and subsequently. I travelled to Europe and undertook my subsequent research with the financial support of National Research Council Operating Grant A8393 and a Staff Travel Grant from the University of Nottingham. The photographs were taken in Tiibingen with Dr. Gocht's assistance and were processed at the University of Saskatchewan by Miss Robin Currie; Mrs. Linda Dietz helped in the preparation of the manuscript.