In situ gymnosperm pollen from the Middle Jurassic of Yorkshire

Several new male fructifications were found and described: Hastystrobus gen. nov. was erected for male conesyielding the Eucommiidites type of pollen grains. This genus is monotypic and the type species Hastystrobus muirii yielded pollen grains that agree with Eucommiidites troedssonii. Hastystrobus muirii very probably has Cycadalean affinities, because the whole abaxial surface ofthe microsporophylls is covered with sporangia. For the first time the male fructification of Ginkgo huttoni (Heer) Sternberg is described. It resembles in general the male fructification of the recent Ginkgo biloba L., and the pollen grains agree with those of Ginkgo biloba. Male cones associated with Brachyphyllum crucis Kendall have been found and described. They yielded pollen grains that after short maceration were identifiable as Circulina, while after prolonged maceration they could be assigned to Classopollis multistriatus Burger. Brachyphyllum crucis is provisionally assigned to the Hirmerellagroup on the basis of its


INTRODUCTION
The Yorkshire Jurassic flora is a classical Jurassic flora. The investigations on this flora have already begun in 1822, and many palaeobotanists have made contributions to our knowledge of this flora. Seward (1900) and Harris (1961) Acta Bot. Neerl. 20(1), Febr. 1971 2 Recently, I have examined the morphology of the pollen grains yielded by male gymnosperm fructifications from the Jurassic flora of Yorkshire. Recent pollen morphology proves that also taxonomic data can be obtained from it, and this was one of the reasons for choosing this subject.
Most descriptions of male gymnosperm fructifications only give slight information on the pollen grains. Couper (1958) was the first to give clear morphological descriptions. But nowadays many more male fructifications are known than when he published his paper. Also Potoni£ (1962,1967) gave pollenmorphological descriptions of some Yorkshire male fructifications, but he took the descriptions mainly from the literature, and dit not reach any new conclusions himself.
The aims of the present study are to examine: 1. Morphology of pollen grains yielded by male gymnosperm fructifications from the Jurassic of Yorkshire (in situ pollen grains), and if possible draw taxonomic conclusions from the pollen morphology. 2.
In connection with this direct comparison with other male fossil fructifications.
3. Also comparison with recent pollen grains from the same or similar taxa (if possible).
4. Comparison with dispersed pollen grains from the same horizons as the in situ material. This was also done by Couper (1958) and, to a lesser degree, by Potonié (1962Potonié ( , 1967.
The aim of this is to find out the affinities of dispersed pollen grains. Because dispersed pollen is much more commonly found than macrofossils, it might be possible through this method to extend our knowledge of the distribution of Jurassic plants on the basis of the distribution of their pollen grains.

GEOLOGICAL BACKGROUND
The Middle Jurassic of Yorkshire includes both marine and continental deposits.
For the continental deposits the term "Deltaic Series" is normally used, although the official term still is "Estuarine Series". But this latter name is very unfortunate, because the character of the rocks really is deltaic, as was already recognized by Simpson (1868). Later the term Estuarine Series came into use wrote historical surveys of the investigations; I do not want to repeat them here.
I wish only to name some of the most eminent investigators who devoted much of their life to the study of this flora; A. C. Seward, H. Hamshaw Thomas and T. M. Harris.
The best known parts of the plants are the shoots and the leaves, on which the majority of the research has been done. Cuticle analysis has added much to our knowledge of the leaves. Male and female fructifications are also quite well known, especially in their general morphology. But, on material such as wood, seeds and pollen grains much less research has been done.

IN SITU GYMNOSPERM POLLEN FROM THE MIDDLE JURASSIC OF YORKSHIRE
3 Acta Bot. Need. 20 (1), Febr. 1971 (Fox-Strangways 1892), but the work by Kendall and Wroot (1924), Black (1928Black ( , 1929Black ( and 1934 and Wilson, Hemingway and Black (1934) made it clear that we are dealing with deltaic sediments. Table I gives the stratigraphical classification of the Middle Jurassic rocks of Yorkshire in general, as proposed by Hemingway (1949) and adopted by Harris (1952). The terms between brackets are those proposed by Sylvester-Bradley (1949).
The word "Series" is in the literature on the Jurassic rocks of Yorkshire used in more than one sense: They are partly formations and partly facies.
Although the use of "Series" in this way is not completely correct, the present author uses it, because it is in general use in the literature on the Yorkshire Jurassic.
The three marine formations between the Deltaic Series are all built of several marine horizons, hence the term "Series" instead of "Beds" came into use.
The Ellerbeck Bed consists of ironstones and shales, the Millepore Oolite and the Scarborough Limestone of limestones and shales, and all bear faunas of Molluscs and Brachiopods.
The plant beds that made the Middle Jurassic rocks of Yorkshire so famous, are found in the marine "Dogger" and in the four non-marine horizons of the Deltaic Series.  (1), Febr. 1971 4 In Yorkshire the term "Dogger" is used to describe the marine beds between the Lias and the Lower Deltaic (see Hemingway 1949). These beds vary in facies from limestones and coarse calcareous sandstones with marine invertebrate fossils on the coast at Whitby, through chamositic oolites (ironstone) to fine silts and shales which overlie the Lias unconformably at Hasty Bank (see locality map, text- fig. 1). The latter are rich in plant remains and form the well-known plant beds of Hasty Bank and Roseberry Topping; Roseberry Topping formerly believed by Thomas (1915) to be Liassic in age and continental in origin.
Their marine natur was postulated by Harris (1964), who sought to explain the erratic distribution of Pachypteris papillosa, by suggesting that this plant lived in a salt marsh environment at sea level, and that salt water had, at least, some effect on the whole assemblage. This supposition was strenghened by the discovery of marine microfossils (Tasmanites and Dinoflagellates) in every locality where the Pachypteris assemblage is found (M.D. Muir-pers. comm.).
Famous localities of the Lower Deltaic are the plant beds of Whitby, Marske Quarry, and Haiburn Wyke (see locality map). They are of two types: In situ plant beds (grey clays) with for instance Williamsonia and Zamites, and drifted plant beds (channel sands with washed-in fossils, e.g. Brachyphyllum) (see also Black 1928).
The Lower Middle Deltaic (Sycarham) consists of dark grey clays disturbed by many minor tectonic movements. Their strong carbonization is thought to be a result of these movements. This is the reason why only plant fragments are found. More complete material is found in ironstones, found as loose blocks at these localities. These blocks, however, belong probably to the marine Ellerbeck Bed and not to the Sycarham Series.
The Upper Middle Deltaic (Gristhorpe) consists of grey clays. The famous plant beds of this Series -Gristhorpe bed (Cayton Bay) and Cloughton Wykeare in situ plant beds and very rich in extremely well preserved plant remains.
Most of the reproductive organs we know are from these plant beds.
The plant beds of the Upper Deltaic (Scalby Ness, Scalby Wyke) are definitely drifted plant beds (see Black 1929) and consist of channel sands from siltstone to coarse sandstone.
(For geological comments on all the four Deltaic Series, see Wilson 1958).
Plant beds: According to Harris (1952) there are five different types of plant beds in the Yorkshire Jurassic: 1.
Truly autochtonous beds (i.e. beds with plants preserved in the position of growth). For example Equisetum columnare roots and stems in the Lower Deltaic.

2.
Lagoon and sluggish river channel beds (fine mud). The richest is the Gristhorpe bed; Mainly large delicate leaves (suggesting that the plants grew near at hand) mixed with some water-worn material.

3.
River channel beds (fine sand). The richest is the Whitby plant bed. They have a somewhat higher proportion of waterworn plants than group 2, but  (1), Febr. 1971 there also occur well preserved leaves and reproductive organs (type 2 and 3 intergrade). 4. Drifted plant beds (sand), like Black's drifted plant bed at Scalby Wyke (Black 1929). Here all but the smallest plants are severely waterworn (type 3 and 4 intergrade).
5. Redeposited plant beds. Such beds have not been described, but are probably common. They consist mainly of though cuticles.
Plant distribution : One of the remarkable things of the Deltaic Series is, that it is impossible to make a fine zonation in each of the four Series. Another pecularity is that each local plant bed may have its own special flora. Often a rare species (with one or a few known localities) is locally very abundant. This uneven occurrence is probably quite normal in floras from deltas.
Apart from the rare species of which both range and frequency are necessarily ill known, the flora can be divided according to its range and frequency into the following groups (Harris 1952): 1. A few species range all four Deltaic Series without any striking change in abundance, for example Brachyphyllum mamillare.
2. A few species range the Lower three Series, but are absent from the Upper Deltaic (Equisetum columnare).
3. Some species seem to be confined to the Lower Deltaic or the Dogger, and are absent from the other Series.
For example: Pachypteris papillosa is only found in the Dogger and is very common there, but nowhere else.

A large group
of species occur commonly in the Lower Deltaic, are rare or absent in the next two divisions, but are common again in the Upper Deltaic (Pachypteris lanceolata, Ptilophyllum pectinoides). 5. A considerable group of species are more or less abundant in the two middle divisions, but rare or absent above or below.
I would like to add another group to those five: 6. Some species are only known from the Upper Deltaic and are common there (some members of the Ginkgoales).
sporophyll, these were macerated in a mixture of KCI0 3 and concentrated HN0 3 (commercial grade). The time of maceration is depending on the state of fossilization. After this, they were washed in H 2 0 and transferred to a dilute solution of NH 4 OH, until the brown colour vanished. The pollen sacs were then opened and the pollen grains mounted in glycerin jelly and sealed with paraffin (method after Punt 1962).
2. If it was impossible to obtain complete pollen sacs from a cone, the following method was used: Part of the cone was prepared free from the specimen and transferred to a centrifuge tube. KC10 3 and HN0 3 concentrated (commercial grade) were added, and it was heated until boiling. The tube was then placed in a ultrasone (Phillips PH2101/00 21kHz) and vibrated for 1 minute. Because of this treatment, the cone fragment fell into little pieces, the pollen grains were freed and immediately macerated. The material was then centrifuged and decanted, and dilute NH 4 OH was added. After 5 minutes the material was again centrifuged and decanted, washed with water, recentrifuged and decanted, and 50 % glycerin-water mixture was added. The material was again centrifuged and after decantation the tubes were placed upside, down for drying. After 20 minutes slides were made, usually six, by mounting the material in glycerin jelly and sealing with paraffin (method after Punt 1962).
Examination of the slides: The slides from the British Museum and the slides with the dispersed pollen grains were first examined and photographed using a Zeiss photomicroscope 0054 (property of Imperial College, London). Lateron Bot. Neerl. 20(1), Febr. 1971 borrowed the slides, and I examined them just as my own slides, using a Leitz Ortholux microscope, and made the photographs with the Leitz Orthomat camera in combination with the microscope. Some material has also been examined with the help of an electron scanning microscope (Cambridge "Stereoscan" mark II a), in the possession of Imperial College, London.
Nomenclature'. The majority of the male fructifications from the Jurassic of Yorkshire are classified in special organ genera, for instance Caytonanthus for male Caytonialean fructifications. The specific names used by Harris (1961Harris ( , 1964Harris ( , 1969 are, according to the present author, in agreement with the International Code of Botanical Nomenclature, and are thus followed by her. Some male cones (especially of Conifers) are simply named by the name of the species to which they belong, with the addition male cone, for instance Elatides williamsoni (Brgt) Sew. male cone. In these cases, the cones have either been found attached to the shoots, or have been definitely attributed to the plants mainly on the basis of agreement of cuticle structure, or on the basis of very close association. As to the nomenclatureof the dispersed pollen grains, Potonie (1956Potonie ( ,1958Potonie ( , 1960) was mainly followed. Some advices were given by Mrs. Dr. M. D. Muir. Pollenmorphological terminology. Erdtman's terminology (1952Erdtman's terminology ( , 1965 was used, especially with regard to the different layers of the exine. Although this terminology is mainly in use for recent (spores and) pollen grains, the present author found that it is quite adequate for fossil pollen grains as well. As to the terms used to indicate length, breadth etc. of colpate and saccate pollen grains Couper (1958) Thomas,p. 237,pi. 14 fig. 33, probably also figs. 34,35,38,40,42 (1941,1946,1964) gives 22 p (18-28 p). So they all agree more or less, except Harris, who gives a somewhat smaller size.
As to the structure of the sacci, most authors agree that they have an indistinct fine reticulum. Thomas (1925)  Winged pollen grains characteristic of Antholithusarberi occur on the stigmatic surface of Gristhorpia.
He also gives the arguments that both the male and female fructifications belong to plants which bore leaves known as Sagenopteris. Harris (1940) records that he has found in 38 preparations Caytonanthus pollen grains (and no other pollen grains) in the micropyle of Caytonia nathorsti and Caytonia sewardi seeds. In the case of Caytonia sewardi in seeds from intact "fruits" and also in isolated seeds, in the case of Caytonia nathorsti only in isolated seeds.
Since the attribution of Caytonanthus to the Caytoniales has been sufficiently proved, it has even been possible to attribute the different Caytonanthus species to species of Sagenopteris and Caytonia. Harris 1941 gives the following attributions: Caytonanthus arberi to Caytonia nathorsti (Thomas) Harris  There has been a long discussion about the systematic position of the Caytoniales, which until now is not completely finished; but nowadays almost everybody agrees as to the main points. In this paper the systematic position will only briefly be discussed. Thomas (1925) thought that the Caytoniales were very early and primitive Angiosperms, but far remote from the recent Angiosperms, and probably originated from Pteridosperms.
In 1931 he defends this view against Krausel's idea that the Caytoniales might be Gymnosperms (Krausel 1926 in Engler). Harris 1940 proves that the pollination of the Caytoniales was gymnospermous and not angiospermous as Thomas believed.
In 1937 Harris already presumed that the Caytoniales did not have a phylogenetic relationship with the Angiosperms. He then discussed the difference between Caytonanthus and any type of Angiosperms stamen or flower: "The 'anther' appears to be a radially symmetrical synangium of the Asterotheca type, while in almost all flowering plants it is a bilaterally symmetrical organ with a connective (occasionally strongly developed) and two lateral thecae.
There is no similarity between the way the anthers of Caytonanthus and the flowering plant are borne on their filaments, but, on the other hand, if they were borne on the lower side they had a considerable resemblance to the type of fructification seen in certain ferns and Pteridosperms". Harris (1951) gives a detailed account of the relationships of the Caytoniales.
He states that the Caytoniales resemble mostly the Pteridosperms, "especially in relation to the general form of leaf, microsporophyll and megasporophyll. The group of Pteridosperms most favoured is Thomas' Corystospermaceae, which agree in their pollen as well as in the size of their parts. Formerly there seemed to be several points of agreement with the Angiosperms, now there are only two-the structure of the stoma and the distribution of the cuticles in the seeds, in both of which they agree better with the Angiosperms than with any known Gymnosperm (including the Pteridosperms). This, however, would be more impressive if these aspects of the Pteridosperms had been more studied".
When we compare the pollen grains of Caytonanthus arberi, Caytonanthus oncodes and Caytonanthus sp. A., we see that they are quite similar in general, but that they differ in some minor aspects. Secondly the sacci of Caytonanthus encodes have a clearer reticulum than in Caytonanthus arberi, and much more than in Caytonanthus sp. A.; and the sacci bulge more in Caytonanthus oncodes than in Caytonanthus arberi. All these differences have also been seen by various other authors (e.g. Harris 1941, Townrow 1962and Couper 1958 Comparison with some other pollen-bearing organs There are some pollen-bearing organs of unknown affinities that show some similarity to Caytonanthus, and have often been placed in the Caytoniales: Pramelreuthiaand Harrisiothecium. Pramelreuthia haberfelneri Krasser was first described as a male fructification without knowing anything of its internal structure, nor of its pollen grains. Later it was thought to be a female fructification like Caytonia. Krausel's paper (1949)  160 [x, the corpus is smooth and the sacci are covered with an indistinct ornament (reticulum; muri ca. 3 |x wide, lumina ca. 7 p. wide). Potonie (1967) states that the total width of the pollen grain is normally less than 160 p (75-100 p). Harris (1951)  having disaccate, monolept (= monocolpate) pollen grains, but they differ considerably in size of the pollen grains and in shape and insertion of the sacci.
I agree with him that there does not appear to be any close connection between Pramelreuthiaand the Caytoniales. Klaus (1966) has re-examined the pollen grains of Pramelreuthia haberfelneri, and he comes to the conclusion that they are the same as his dispersed pollen species Lueckisporites junior (their overall breadth is according to him ca. 80 p, the length ca. 58 p). He states that these pollen grains are quite unlike those of the Caytoniales. Their affinities lie probably within the Pteridosperms.
Harrisiothecium marsiloides (Harris) Lundblad has disaccate pollen grains with a total breadth of about 40 p (Harris 1932, Townrow 1962, Potonie 1967 The systematical position of Harrisiothecium is uncertain. Normally it is provisionally placed in the Pteridosperms, but the organ branches in three planes, unlike most Pteridosperm fructifications and also unlike Caytonanthus. The pollen grains, however, do show a great deal of resemblance to those of Caytonanthus (and to a less extent also to Pteruchus and Pteroma). Townrow (1965) suggests that the difference between disaccate pollen grains with a distinct germinalum furrow (colpus) and those merely with a thin area (leptoma) between the saccus roots might not only be of generic value (Townrow 1962), but even useful at family level. Discussion: The ornamentation of the sacci is rather difficult to see because most of the specimens are poorly preserved; it is certainly not a clear reticulum.
The sulcus is only visible in a few specimens (plate II, fig. 1). The attachment of the sacci is very indistinct, giving rise to inaccurate measurements, and is in proximal view only slightly sineous.
General discussion and comparison with other Pteridosperm fructifications Pteroma was established by Harris (1964) as an organ genus for male Pteridosperm fructifications resembling Pteruchus (Southern hemisphere-Middle Triassic) and to a lesser extent Harrisiothecium (Lundblad 1961, Rhaetic of Greenland). Pteroma thomasi resembles most Pteruchus simmondsi (Townrow 1962) in appearance, but the pollen sacs in Pteruchus simmondsi point freely outwards, while they point slightly inwards in Pteroma.
In Hanisiothecium the branching of the organ is on all sides, while in Pteruchus and probably also in Pteroma it is pinnate (in one plane).
As to the pollen grains: Pteroma thomasi pollen grains are very much alike those of Pteruchus; they resemble especially Pteruchus dubius closely, but the lumina in the reticulum of the sacci of Pteroma are definitely smaller than in Pteruchus dubius. Also Pteruchus dubius has a distinct reticulum, while Pteroma pollen is obscurely reticulate.
Pteruchus simmondsi, also with pollen of similar size, has a much coarser reticulum, while Pteruchus africanus with smaller pollen grains also has a prominent reticulum (see Townrow 1962a, b, Potonie 1962and Thomas 1933 fig. 6) resembles Pteroma thomasi most, but there are some differences between them (see also Harris 1964): The first point is that the measurements are not exactly the same, for example the overall breadth in Alisporites thomasi is 60 p. (extremes 45-78 p,), while it is 76 p, (extremes 65-100 p.) in Pteroma, and the ratio length to breadth of the corpus is 1,28 (extremes 0,95-2,1) in Alisporites thomasi and 1,47 (extremes 1,89) in Pteroma.
The sacci show a coarser reticulum in Alisporites thomasi; the line of attachment of the sacci is clearer than in Pteroma and the corpus is almost smooth, while it has a pitted tectum in Pteroma.
Alisporites thomasi was first described from the Brora Coal, N. E. Scotland (Middle Jurassic), but was later recorded from more places, for example Tubbergen borehole, Netherlands,Upper Jurassic (Burger 1966) and Scania, Sweden, Lias (Nilsson 1958). It ranges at least from the Lower Lias to almost the top of the Cretaceous (Staplin, Pocock and Jansonius 1967). It has also been recorded from the Jurassic of Yorkshire by Dr. M. D. Muir (see Harris 1964). Those (1), Febr. 1971 24 but differ from the Conifers, in which the sporangia! wall is thin and in fossil specimens reduced to a delicate and translucent membrane.
The thickness of the wall is largely provided by the heavily thickened epidermal cells, which appear to agree in their strongly pitted walls with those of the living Cycads. Another point of agreement with the Cycads is provided by the inner cuticle of the sporangium, which it was found possible to demonstrate also in recent Encephalartos cone; I am unaware of its existence in Conifers, but it is seen also in certain Pteridosperms.
The cuticle of the microsporophyll provides stomata in which the subsidiary cells are arranged in a ring, having a different origin from the guard cells. This type is found in many plants, including nearly all gymnosperms except the Bennettitales. The thickenings of the surface of the guard cells themselves is also of the general gymnosperm type, but the scattering of the stomata and the rather irregular grouping of the subsidiary cells round the sunken stomata is a good deal more characteristic of the Cycads (and perhaps the Pteridosperms) than of the Ginkgoales or Conifers. The microspores agree with the type which is found throughout the Cycads, but also in many members of other Gymnosperm families, being that Wodehouse (1935) considers the primitive or generalised condition for this family.

Very perfect and fairly complete agreement has thus been established between
Androstrobus and the Cycad male cone, but difference between it and the repro-  Thomas & Harris (1960) pointed out when they described and discussed Androstrobus prisma, that there are some differences with the other Androstrobus species; but they did not consider them of generic importance.
Also the type of pollen grains of Androstrobus prisma seems to be different from the type of pollen grains of the other Androstrobus species currently known. So Androstrobus prisma might belong to another group of Cycadalean plants different from the other Androstrobus species. It might be possible to distinguish between the pollen grains of the Nilssoniaceae and of the group to which Pseudoctenis belongs (which also includes the genera Ctenis and Paracycas, and possibly also Ctenozamitessee Harris 1964 Pollen grains monocolpate; outline in polar view ellipsoid, in equatorial view rounded rectangular; longest axis ca. 28 ;x; colpus extending over 7/8th of the length of the pollen grain, rounded at the ends; exine laterally thickest, up to grains were produced by a memberof some extinct gymnospermous group. The over-all shape of the grains resembles that of both living and fossil cycadophyte pollen. The form of the sulcus, particularly where it is rounded at its extremities, is closer to that of living cycad pollens than to most angiosperms pollen".
It now appears that also because of its exine ornamentation Clavatipollenites may have Cycadalean affinities, although of course the Angiosperms have to be considered as well.
Comparison with dispersed pollen grains Couper (1958)  So I think that it might be possible to divide Monosulcites minimus into three species (or sub-species), one presumably Ginkgoalean with elliptical outline and pointed ends, one presumably Cycadalean or Bennettitalean (see Williamso-niella) with elliptical to circular outline, rounded ends, and a pitted surface, and one presumably Cycadalean with an almost circular outline and a granulate surface (the Androstrobus prisma-type).
As far as I can see, it must be possible to distinguish between those types of dispersed pollen grains, although there will always be transitions and badly preserved specimens, in which case it will be very difficult to decide to which species they might belong.
Monosulcites carpentieri and Monosulcites subgranulosus are well-defined species ; they have a much larger size than Monosulcites minimus, averages resp. 58 and 60 p, and also than both Cycadalean and Ginkgoalean type pollen grains.
They match more the common Bennettitalean type of pollen grain. But we will come to this when describing and discussing the Bennettitalean pollen grains. Burger (1966) (1), Febr. 1971 The whole genus Chasmatosporites seems to me to be composed of both spores (monolete) and pollen grains (monocolpate and grains like Androstrobus prisma) and is thus rather a problematic genus. Pocock and Jansonius (1969) redescribed and emended the genus Chasmatosporites. They recognized (as well as other authors) its gymnospermous affinities, and it is now considered a genus of monosulcate pollen grains.
Many more authors described somilar pollen grains as Monosulcites or under various other names.
This type of pollen grain is quite common from the Triassic onwards.
Diagnosis: Microsporophylls borne spirally around the slender axis of the strobilus, which is at least 2 cm. long; distally exposed region of microsporophylls rhomboidal, 2-2,5 mm. broad, 1,5 mm. high; microsporophylls near apex slightly imbricate; under surface (abaxial side) of microsporophylls probably completely covered with sporangia; pollen grains as a ruletricolpate; one colpus more developed than the other two; colpi not symmetrically spaced on the pollen grain, the distance between the two shorter ones normally less than the distance between a short and the long one (Eucommiidites-type of pollen grain). Unfortunately, it has proved impossible to make cuticle preparations of the microsporophylls, although the specimens seemed well preserved. Apparently the cuticle is very thin, for it broke into tiny fragments, which gave no information at all. Luckily, however, the sporangia contained pollen grains. he believed it to have angiosperra affinities, and compared it with recent Eucommia pollen grains. Kuyl, Mullerand Waterbolk (1955) were the first to doubt this assignment. They stated: "One of the coipi is longer than the other two, and is provided with rounded-off ends. The distance between the short furrows may be smaller than between a long and a short furrow. Finely, the grains are not ellipsoidal, but flattened. Accordingly, radial symmetry which, in our opinion, is a most important feature of dicotyledonous pollen, is absent. The grains can perhaps best be considered as monocolpate with two additional furrows on the proximal side". According to them, the relationship might be sought with the Chlamydospermae. Couper (1956Couper ( , 1958 Thomas (1915) states that the pollen grains "appear circular or elliptical and were probably flattened spheres; their walls were thin and they were about 0,02 mm. in diameter".
Harris (1944) does not fully describe the pollen grains either; he only says: "Ripe oval pollen grains showing a longitudinal furrow lie scattered in pollen sacs and over the sporophylls, and others are found among the hairs of the bracts. Immature pollen, often in tetrads, occurs in the sporangia of many isolated sporophylls, and abnormal pollen was noted". I agree with him that many isolated microsporophylls contain unripe pollen. Couper's description (1958)  Florin (see Thomas 1915, Harris 1944, on grounds of association and because involucral bracts of Williamsoniella coronata were found bearing a diminutive Nilssonopteris vittata lamina. There is thus a transition between the bracts and the leaves. Also stems have been found that show scars, the size of which agrees very well with that of the (associated) flowers and leaves.
Williamsoniella papillosa is attributed to Nilssoniopteris major Harris (see Cridland 1957, and Harris 1969) on grounds of association.
When we compare the pollen grains of Williamsoniella coronata with those of Williamsoniella papillosa, we see that they are of the same type (Harris 1969 says that the pollen grains look the same): monocolpate grains with an elliptical to circular outline and a very thin wall. But the exine ornamentation is somewhat Weltrichia sol, as he is not absolutely certain that the Williamsonia gigas male flower described by Thomas (1915) (Nathorst 1909, Couper 1958, Potonie 1962, but according to Harris (1969) (1), Febr. 1971 Discussion : Nathorst (1909) was the first to describe the pollen grains: "Diese sind denjenigen von Williamsonia spectabilis ahnlich, sind aber etwas kleiner und messen meistens nur 36-44, bisweilen jedoch 50 ji." This description has been repeated by some subsequent authors (Couper 1958, Potonie 1962. Harris (1969) does not give a description, but merely a comparison with the grains of Weltrichia pecten. which are very similar (see also Weltrichia pecten).
General discussion Harris (1969) (1), Febr. 1971 In addition to the five species of Weltrichia from Yorkshire, Harris (1969) recognises 5 species that were already classified in Weltrichia: W. mirabilis  (1), Febr. 1971 crispa, but in his volume on the Bennettitales (1969) he says about this species: "The specimen Harris described as Wonnacottia crispa consists of a leaf 12 cm long in which every segment of the lamina is more or less strongly curled and bears, on the abaxial side, numerous round pouches formerly regarded as pollen sacs but now regarded as galls". And somewhat further in his account: "Most of the galls macerated yielded no pollen but one (unfortunately) was found to give about a hundred grains, and on this fact I concluded that they were sporangia. The grains are monocolpate as usual in the Bennettitales. Subsequently I found a few specimens which look like ordinary A. nilssoni leaves but bear just a few galls". So this species is now reinterpreted.

General discussion
The Order of the Bennettitales is distinguished from all other groups mainly by their stomata, which show a pair of subsidiary cells opposite the two guard cells, the whole group of cells looking as though formed by the division of a single cell (syndetocheilic arrangement). In almost all other Gymnosperms the subsidiary cells are differently arranged, not being sister cells of the guard cells, but having an independent origin (haplocheilic). Florin (1933) in particular has discussed this character. As to the taxonomic position of the Bennettitales, they are normally placed in the Cycadospermae, together with the Cycadales. However, Thomas (1950) states that these two groups are not so closely related as the similarity of their vegetative structure would suggest. There are great differences in the form of their reproductive organs. Harris (1969) goes even further: "Although the Bennettitales were originally classed as Cycads, and later linked with Cycads, it seems that any phylogenetic connection between them must be remote.
One can indeed imagine many alternative possible links between those two classes, but until there is good reason for preferring a particular link, it is best to consider them unconnected. This is true also of other classes; the Bennettitales are or seem remarkably isolated. Thus the term "Cycadophyte" which was intended as a major phyletic group means no more than "Gymnosperm with a pinnate leaf".

Diagnosis of male fructification:
Male fructification, about 10 mm long and 2 mm in diameter, consisting of a rather thin central axis and microsporophylls loosely placed around it; no special arrangement of the microsporophylls around axis discernable; form of microsporophylls rather indistinct, cuticle unknown; at each microsporophyll probably two pollen-sacs; pollen grains monocolpate; outline elongate-elliptical, with acute ends; longest axis 35 p (extremes 29-42 p); colpus (in the distal surface) extending about the whole length of the pollen grain, slitlike to broad; wall (exine) 1-1,5 p thick; consisting of two layers: nexine and sexine; nexine smooth, 0,5 p thick; sexine 0,5-1 p thick, not completely smooth, but structure rather vague.
Description of the material : There is only one specimen of the male fructification, and it is rather badly preserved. Because of this it appeared to be impossible to make any cuticle preparation.
The specimen was found at Scalby Ness, the Ginkgo bed, and is deposited in

1.
Association. The specimen is from Scalby Ness, Ginkgo plant bed, where Ginkgo huttoni is by far the most common fossil.
On the block the male fructification is associated with Ginkgo huttoni and Haiburnia blacki Harris.
Haiburnia blacki is a Conifer, and the structure of the fructification and the shape of the pollen grains do not suggest a coniferous origin for the fructification.
Consequently Ginkgo huttoni remains the most reasonable parent plant.

2.
Agreement in structure with the male fructification of the recent Ginkgo biloba L.
Although the microstrobilus of Ginkgo biloba is somewhat larger than the specimen here described, it agrees with it in almost all essential characters: The microstrobilus of Ginkgo biloba is built of a central axis with irregularly arranged microsporophylls around it just as observed in Ginkgo hut toni; although bad preservation might, in the case of the latter, have obscured any arrangement of the microsporophylls.
In Ginkgo biloba the microsporophylls bear 2 (sometimes 3 or 4 or even more) V. 27499a. This slide was also studied by Couper (1958), but he states that "the pollen grains are still in the pollen-sac, making it difficult to obtain adequate  (1), Febr. 1971 illustrations". Harris (1948 and personal communication) definitely says that the grains are not in pollen sacs but on bud scales, and he is not even sure that they are really pollen grains of Ginkgo huttoni: " Associated pollen : The inner (triangular) bud scales have pollen grains sticking to them; one has a few grains, another nearly a hundred, some on both sides.
The pollen grains are of uniform size and show a well-developed longitudinal groove and a rather thin, smooth extine. This pollen is of an unspecialized Gymnosperm type, occuring in Ginkgo and many other genera, and there is no evidence, apart from its association, for refering it to this species.
In view of the fact that the pollen is sometimes met on leaf cuticles of the same species (e.g. Cayttonanhus pollen on Sagenopteris), it would be most interesting if this same association were met elsewhere in G. huttoni. If it were, it would provide evidence that G. huttoni, like G. biloba, shed its pollen shortly after its resting buds opened".
The present author also examined this slide. She quite agrees with Harris about it being bud scales, and she also thinks that the pollen grains very probably do belong to Ginkgo huttoni, for they agree in all aspects with those here described from the microstrobilus, and on grounds of association. Couper Cycadalean and Bennettitalean grains also fall within this species (see the chapters on those groups). Most authors (Couper 1958, Potonie 1962, Nilson 1958, Burger 1966, Staplin, Pocock and Jansonius 1967 ( 1), Febr. 1971 "None of the present specimens is as richly provided with pollen grains as the Greenland ones (especially " Microcheiris ") but all those that still possess the fibrous inner membrane show moderate numbers of grains scattered over it. The specimen shown in fig. 26 has, in addition, several compact masses of pollen situatedjust below the embryosacs, in positions which correspond with those of "Microcheiris", and there are similar pollen masses among the broken fragments from inside other specimens. It is very difficult to make out the form of the pollen grains in these masses as they overlap; the only thing certain is that there are many large grains, some about 50 ;x wide, with rather thin, very finely punctate walls. I am not able to say whether wings and triradiate scars are absent, but Thus, the present author is inclined to think that Leptostrobus cancer (and with it probably the whole Czekanowskia-group) has more relationships to Conifers than to Ginkgoales. Krasilov (1968)  granula about 1 p in section; brim of sexine around the nexine (if the two layers are separate) 6,5 p (extremes 4,5-10 p).
Discussion: Brachyphyllum mamillare was attributed by Kendall (1949) to the Araucariaceae mainly on the basis of the structure of the female cone-scales (including their cuticle) and seeds, which typically agree with those of recent Araucaria.
There is remarkably close agreement between the cuticle of the outside of the cone-scale in Araucarites phillipsi Carruthers Kendall (1949) was also the first to describe male cones of Brachyphyllum mamillare. Some cones were attached to the shoots, others were found separately but in close association with Brachyphyllum mamillare. With respect to the pollen grains she states: "Pollen grains are obtainable in small numbers at least from every cone macerated. They are small, round, and 60-80 |x in diameter.
The wall has been thrown into a series of irregular folds on collapsing. Pits and granules are absent from the exine".
Couper (1958)  The present author also examined Kendall's slide, which only contains six pollen grains and is rather deeply stained (pi. XI, fig. 1).
My own material consisted of several attached and separated cones. All of them yielded some pollen grains, and one yielded over 50 grains out of a few microsporophylls.
The pollen grains appeared to be of two types:  (1), Febr. 1971 middle thinner area, probably corresponding to the sulcus, according to him.
He also states: "One grain shows a separate round body, 30 p wide, with a well-developed wall. This feature has not been seen in any of the other specimens, and its nature is unknown". This certainly recalls Brachyphyllum mamillare.
2. Gamerro (1965a) described male cones from the Lower Cretaceous Apterocladus lanceolatus Arch. 1966, which yielded pollen grains that agree in some respects with those of Brachyphyllum mamillare. The pollen grains are somewhat smaller (equatorial diameter 45 p); they have a circular thinner area (colpus?), and are normally round with a finely granulate exine. Sometimes there is a kind of inner body with a brim around it, with a tendency to become trilobate, and in some cases there seem to be three rudimentary sacci (see Gamerro 1965 a andArchangelsky 1966). This certainly looks like the pollen grains of Brachyphyllum mamillare, but since I did not study the material, I can say nothing for certain. Apterocladus lanceolatus was provisionally attributed to the Podocarpaceae by Archangelsky (1966), mainly on the basis of the leaf cuticle, but there are several points of disagreement between Apterocladus and the Podocarpaceae. The pollen grains resemble, according to him, the pollen grains of the Araucariaceae but there are some differences in the structure of the exine. They differ from those of the Podocarpaceae, in not having clearly saccate grains; but Saxegothaea of the Podocarpaceae does not have sacci either. However, its exine structure is rather different. Therefore, it seems that the pollen grains have some points of agreement with those of Araucariaceae, and it would be worth to re-examine the material.
N.B. Gamerro believes that the saccus (as he calls it) arises through a separation of the extine and intine around the colpus.
3. Townrow (1967) described a Jurassic male cone from Antarctica: Masculostrobus warrenii. which yielded pollen grains that also have some resemblance to Brachyphyullm mamillare grains. He described the grains: "Pollen of more or less rounded outline, diameterabout 110 fx (78 to 153 p, amb equilateral triangular, but with well rounded corners, of side about 88 p (62 to 110 p), and usually reaching outside edge of grain at three places. Grain with three sacci, rarely confluent. Sacci about 16 p (3 to 24 p) wide, not always of equal size.
Sacci sometimes showing small radial foldings. Exine of amb 1,5 p thick, at saccus edge (exoexine) 2p; sometimes very faintly showing more or less reticulate pattern, and sometimes faint triradiate mark, with arms running more or less straight and reaching almost to the equator".
Those pollen resemble to some degree the trilobate form of the Brachyphyllum mamillare pollen, but as the other types of the Brachyphyllum mamillare pollen have not been found in these cones, this is the only point of agreement between these two species. Masculostrobus warrenii is thought to belong to the Podocarpaceae, also on the basis of the structure of Nothodacrium warrenii Townrow, the species to which Masculostrobus warrenii is attributed.
Comparison with recent pollen grains I studied the pollen grains of some recent species of the Araucariaceae and found University, Utrecht) showed the same two types of pollen grains as Brachyphyllum mamillare (PI. XII, figs. 1-4). Here the majority of the grains had the nexine loose from the sexine, and to some degree shrunk. When the nexine was shrunk, especially if it was considerably shrunk, the brim of sexine was wrinkled and scalloped (PI. XII, figs. 1, 2). Where the nexine was only slightly shrunk, the sexine is almost unfolded (PI. XII, fig. 4). But in all cases the total diameter of the pollen grain (i.e. the diameterof the sexine) is about the same. The diameter of the nexine is of course quite variable. Here again, as in Brachyphyllum mamillare, the nexine is smooth and the sexine granulate. I measured 100 pollen grains of Araucaria araucana and found that 42 % were of the normal type, in 39 % the nexine was somewhat shrunk, and in 19 % the nexine was considerably shrunk.
I also noticed a triradiate mark on some of the pollen grains (PI. XII, fig. 1). So we see that the dimorphism of the pollen grains of Brachyphyllum mamillarealso occurs in the recent Araucaria araucana. As a dimorphism like this one is not very common, we can say that either Araucaria araucana must have originated from Brachyphyllum mamillare or a related plant, or that in the Araucariaceae there is a trend in the pollen grains to separate the nexine from the sexine and produce pollen grains that appear to be monosaccate.

Comparison with dispersed pollen grains
When we compare the pollen grains of Brachyphyllum mamillare with dispersed pollen, we see that type 1 of Brachyphyllum mamillare grains resembles Araucariacites australis Cookson (PI. XII, fig. 5).
Araucariacites australis was described by Cookson from the Tertiary of the Kerguelen, but the Jurassic and Lower Cretaceous grains are indistinguishable from this species, and are thus included in it (see Couper 1958). According to Couper Araucariacites australis grains are comparable in all respects with those of Brachyphyllum mamillare. I would like to qualify this, that they may only be compared with Brachyphyllum mamillare grains type 1. Araucariacites australis is recorded from Jurassic times from all over the world. There are some other species of the genus but they are of minor importance.
When we come to Brachyphyllum mamillare type 2 pollen grains, we meet much more difficulties. They resemble pollen grains that fall within the genus Applanopsis Doering. This is a genus that has caused much confusion, especially in its nomenclature. A short review of the most important literature will be given here. Balme (1957) fig. 1).
2. Zonalapollenites trilobatus: Amb rounded triangular. Pollen grains complex, consisting of a sub-triangular central body with three equatorially attached bladders, sometimes fusing to form a single trilobate bladder constricted at the apices of the central body.
No visible germinal mechanism, but a vestigial triradiate scar is sometimes present.
Exine of central body 1-2 jx thick, rugose or with a wrinkled appearance.
3. Zonalapollenites segmentatus: Amb circular, outline indented. Pollen grain complex, consisting of a circular central body surrounded distalo-equatorially by a narrow bladder. No germinal aperture or tetrad markings observed.
The three species have much in common, and apparently might form together a good genus. The genus Zonalapollenites which was described in 1953 by Thomson and Pflug,  resp. Applanopsis lenticularis and Triangulopsis discoidalis. He considered them to be plankton species, but the publication is effective.
Also in 1961, Dev transferred Z. dampieri, Z. trilobatus. and Z. segmentatus to his new genus Callialasporites. He stated that Zonalapollenites was a synonym of Tsugaepollenites. While these three species, however, do not fall within Tsugaepollenites because they do not have the type of velum characteristic for this genus, they had to be transferred to a new genus (in casu Calliasporites). Dev described, besides the already known species, a new species C. monoalasporus, with an unfolded "saccus" and a very clear boundary between "corpus" and "saccus". Unfortunately his paper appeared some months after Doring's, so consequently Applanopsis has priority over Callialasporites. Pocock (1962)  Callialasporites rimalis (with a thick brim around a "body" with folds) and C. triletes (with a clear tetrad scar). Muir (1964) fig. 2). Goubin et al. (1965) discussed in detail the taxonomic position of Z. dampieri and Z. trilobatus and came to the conclusion that both should be placed within Doring's genus Applanopsis. They emended Applanopsis to include also Triangulopsis. Burger (1966) combined Applanopsis dampieri and A. segmentatus to one species (A. dampieri) because he found in the Dutch Upper Jurassic and Lower Cretaceous many transitions between them. Reiser & Williams (1969) said that there appears to be a gradation from Tsugaepollenites segmentatus through T. dampieri to T. trilobatus. Norris (1969) emended Callialasporites dampieri to include pollen grains with a "saccus" that is slightly lobed, but not completely trilobate (like some of the grains of Brachyphyllum mamillare). When we compare the pollen grains of type 2 of Brachyphyllum mamillare with Applanopsis, we see that the majority resembles most Applanopsis dampieri, but some pollen grains come quite near Applanopsis trilobatus, while a few are more like Appl. segmentatus or Tsugaepollenites sp. A. This is also in agreement with Horowitz (1968), who says that Appl. dampieri is more common in the Upper Jurassic of Israel than both Appl. trilobatus and Appl. segmentatus. This has also been recorded by various other authors.
We therefore,, conclude that from one cone (Brachyphyllum mamillare) pollen grains come that when found dispersed should fall not only into different species (Appl. dampieri, Appl. trilobatus and App. segmentatus), but even into different genera (Araucariacites and Applanopsis) (as stated above, I believe that Tsug. sp. A also belongs to Applanopsis).
This again is in agreement with the fact, that all these grains are found together within one cone.
It should be pointed out that not all Applanopsis species might have belonged to the Araucariaceae, but that at least some grains of this genus did originate from an araucariaceous male cone. The pollen grains that were described by Cooper (1958) under the name Tsugaepollenites mesozoicus, and were transferred by Nilsson (1958) to Cerebropollenites mesozoicus are, in my opinion, unrelated to Applanopsis and the Araucariaceae.
When we compare the pollen grains of the male cones that showed some resemblance with Brachyphyllum mamillare with dispersed pollen, we see that the grains of Brachyphyllum mirandai agree with Araucariacites australis, except for the pollen grain with the "inner body".
So I believe that there must be some relationship between these three cones and Brachyphyllum mamillare.

Distribution of the fossil Araucariaceae
The fact that these two types of pollen grains have been found within one cone, has also consequences for the distribution of the fossil Araucariaceae. Seward & Conway (1934) published a detailed account on the fossil distribution of this family. Pollen grains monoporate or inaperturate; outline almost circular, but often folded; diameter 52 p (extremes 42-62 p); porus, if present, with a weak annulus; exine consisting of two layers: nexine and sexine; nexine 0,5-1 p thick, smooth; sexine ("perine") loosely fitting, normally wrinkled or torn, very thin (less than 0,5 p), scabrate; size of pollen grains without sexine 39 p (extremes 30-47 p).
Discussion: Phillips (1875) was the first to describe a male cone of Elatides williamsoni but he did not describe its pollen grains, neither did Seward (1900Seward ( , 1919. Harris (1943)  The affinity of Elatides williamsoni was discussed in detail by Harris 1943. He concluded that Elatides williamsoni agreed on all major points with the Taxodiaceae (shoots, microsporophylls, pollen grains, female cone-scales and seeds): "It is concluded that Elatides williamsoni is certainly a member of the Taxodiaceae and is nearest to Cunninghamia in several respects but not ancestral to it".

Comparison with recent pollen grains
When we compare the pollen of Elatides williamsoni with those of the recent

Elatides Heer
Taxodiaceae, we see that they arc very close to the grains of Cunninghamia lanceolata (PI. XIV, fig. 1) having also a loosely fitting sexine. They are also similar to grains of other genera of the Taxodiaceae, like Sequioa and Cryptomeria.
Fragments of male cones of Hirmerella airelensis were described by Muir & v. Konijnenburg (1970) as having microsporophylls bearing probably two pollen sacs.
They attributed the material from Cape Twt described by Harris (1957) as Cheirolepis muensteri and also having only two pollen sacs on each microsporophyll, to this species.
Of the male cone associated with Brachyphyllum scotti (Kendall 1949) we know little; only the pollen grains are well known, and they are considered below.
We also know very little about the male cone associated with Pagiophyllum connivens (Kendall 1952).
The pollen grains are compared below.
Masculostrobus rishra (Barnard 1968 As only the female cone-scales of both the Hirmerella species are known, we can only attribute these species with certainty to the Hirmerellagroup; but there is no reason why the other species would not belong to this group (and indeed it is highly probable that they will belong to it), only we have not the proof yet.
When we compare the male cones of these 7 species with each other, we see that they fall into two groups on base of the number of pollen sacs on each microsporophyll. There is a group with a number of pollen sacs (6-8 for Masculostrobus rishra, and up to 12 for Hirmerella muensteri), and a group with only 2 pollen sacs ( Hirmerella airelensis and Brachyphyllum crucis). We do not know the number of pollen sacs of Pagiophyllum connivens and Brachyphyllum scotti; and Tomaxellia biforme will probably fall within the first group. The pollen grains will be considered later in this section.
There is another point of agreement between those 7 species, namely the type

Comparison with Classopollis and Circulina
A short account of the most important literature on Classopollis is given first. Reissinger (1950) described dispersed pollen grains like those found in the male cones of Hirmerellamuensteri under the name Pollenites torosus. He also mentioned some other grains that were rather like Pollenites torosus but not identical according to hem, and he did not name them. Pflug (1953) described those grains (so far as I can judge from the illustrations they are the same) under the new name Classopollis classoides. He thought them to be tricolporate or sometimes tetracolporate and with a so called "rimula". He misinterpreted the grains completely,but the name is valid, and so long as it is not proved that Pollenites torosus is the same as Classopollis classoides, the specific epithet must be maintained and the latter species is the type species of the genus Classopollis. Couper (1955) gave a much better morphological description and compared the grains with those from the Pagiophyllum connivens cone. He thought them to be identical. In 1958 the same author emended the genus Classopollis, and he took Classopollis torosus (Reis.) Couper as type species, because according to him it is conspecific with Classopollis classoides (PI. XV, fig. 5). Pocock & Jansonius (1961) in their revision of Classopollis emend the genus again, and they take Classopollis classoides as type species, for the same reason as discussed above by the present author. They give an elaborate morphological description and interpretation, saying among others: "Exoexinal ornament appears to consist of small pits". They describe three new species: Classopollis belloyensis, Classopollis minor and Classopollis pflugii. Classopollis belloyensis was found in the Permian, and there are only a few rather badly preserved specimens. Classopollis minor is a rather small form of Classopollis with prominent striations. The Lower Cretaceous Classopollis pflugii has "the exoexine loosely enveloping the intexine, which forms a more or less spherical central body". Chaloner & Clarke (1962) (Harris 1957) which was later assigned to Classopollis harrisii (Muir & v. Konijnenburg 1970). They state: "The exine is tegillate (tectate) but in addition the ectonexinous layer bears a series of relatively large inwardly-directed rods. The sexine and ectonexine are of homogenous composition and beneath them a lamellatedendonexine may also be present.
The wall structure proves to be as distinctive as the general morphology of the pollen, and in its complexity is unmatched even among living Angiosperms". Maljavkina (see Klaus 1960) because it had an equatorial girdle, and not in Classopollis because it did not have striations on this girdle. So they widened the diagnosis of Classopollis to include species like this. Burger (1965Burger ( , 1966  Description of the material : There is only the type specimen. It was found at the Gristhorpe bed and is in good condition. However, the cuticle of the microsporophyll broke into tiny fragments, which gave no information at all.
Because there is only one specimen, I did not section it to study the number of pollensacs. So far as I could see, there are two pollensacs on each microsporophyll, but there might be more.
Discussion : This cone was discovered during a routine examination of male Discussion: Although the cones are in bad condition, they were described because they yielded disaccate pollen grains, which until now have not been found in a Jurassic Conifer cone. Some pollen grains are definitely disaccate (PI. XVI, fig. 6) while others appear to be trisaccate (PI. XVI, fig. 4). The cones were included in the genus Pityanthus because they produced saccate pollen grains (see Barnard 1968  The pollen grains were also compared to recent Cycad pollen grains, and it was concluded that they agree in general respects but differ in detail especially in exine ornamentation. In any case, fossil Cycad pollen has wall structure, and thus, it is possible that fossil monocolpate pollen is not necessarily of Angiospermous origin as was formerly thought (see for instance Clavatipollenites) but may also point to Cycadalean origin.
Dispersed fossil Cycad pollen grains fall within Cycadopites ( = Monosulcites) minimus , although the present author is inclined to think that it should be possible to distinguish between the different types of Cycadalean pollen grains within this species.
The new genus Hastystrobus is described. This is a genus for fossil male cones which yield Eucommiidites type of pollen grains. It is the first time that a cone yielding Eucommiidites pollen had been described. Only the type-species is known: Hastystrobus muirii (yielding Eucommiidites troedssonii) which very likely has Cycadalean affinities, as probably the whole under (abaxial) surface of the microsporophylls is covered with sporangia, a feature that is known only from the Cycadales. The pollen grains are normally tricolpate, with a large main colpus and two smaller additional ones. Some pollen grains show only the main colpus, or the main colpus plus one additional one. Thus, it seems likely that Eucommiidites troedssonii must have evolved from monocolpate pollen grains, like those of the Cycads.
There are other Eucommiidites species known, and some of them are recorded from the micropyles and pollenchambers of seeds (E. delcourtii in Spermatites pettensis, E. minor in Spermatites patuxensis and E. troedssonii (although the present author believes that these pollen grains belong to a new species and not to E. troedssonii) in Allicospermum retemirum), proving that Eucommiidites has gymnospermous affinities, and is not of Angiosperm origin as was thought by Erdtman. Although Hastystrobus muirii and therefore the E. troedssonii associa-71 ted with it is very likely of Cycadalean origin, it is possible that the other Eucommiidites species do not have Cycadalean affinities.
Bennettitales: Within the Bennettitales two types of pollen grains were recognized : 1.
Dispersed pollen grains of type 1 fall within Cycadopites minimus while those of type 2 agree with either Cycadopites carpentieri or Cycadopites subgranulosus.
Ginkgoales: For the first time the male cone of Ginkgo huttonihas been described. The pollen grains yielded by it are of the normal Ginkgoalean type. The pollen grains were compared with those of the recent Ginkgo biloba and they agree closely. Dispersed they resemble Cycadopites minimus, but the present author believes that the C. minimus pollen grains of Ginkgoalean origin can be distinguished from C. minimus grains of Cycadalean or Bennettitalean origin because of the more elongate outline and acute ends of the pollen grains.
Doubtful Ginkgoales : Within the female fructification Leptostrobus cancer pollen grains were found in compact masses near the micropyles of the seeds.
Although it is not definitely proved that those pollen grains belong to it, it seems highly probable. Thus, pollen grains yielded by one cone, may dispersed be classified not only in different species but also in different genera. N.B. Araucariacites and Applanopsis are often recorded together. Because the three Applanopsis species and Araucariacites have been recorded from all over the world in Jurassic times, it seems very likely that the Araucariaceae were universally distributed in those days.
Taxodiaceae: The pollen grains of Elatides williamsoni were already well known, and nothing new was found. Dispersed they resemble closely Perinopollenites elatoides.
Hirmerella-group: For the first time male cones of Brachyphyllum crucis were described and they yielded pollen grains of the Circulina and Classopollis type. When the pollen grains were only slightly macerated, they resembled Circulina but after prolonged maceration exine ornamentation including the striae on the equatorial belt became apparent and the grains were comparable to Classopollis multistriatus. Because of its pollen grains B. crucis was provisionally assigned to the Hirmerella-group. It is suggested by the author that all Conifers that belong to the Hirmerella-group have a cuticle with a special type of stoma: the stomatal pit is surrounded by a thickened "ring" formed by the subsidiary cells. The ring may bear papillae. Also the periclinal walls of the subsidiary cells may have striae.

Coniferospermae incertae sedes:
Masculostratus harrisii is described for the first time; its pollen grains are of the Inaperturopollenites type. The only difference between the male cone of Elatides williamsoni and Masculostrobus harrisii are the pollen grains. Otherwise they agree in almost all respects. Pityanthus scalbiensis was also described for the first time; it yielded disaccate pollen grains, that are rather primitive because the sacci are almost without structure. As the cones were badly preserved not very much information could be gleaned from them.
I would like to finish this paper by making some suggestions for future work on male cones and their pollen grains: 1. Fresh examination of the present material may reveal more information.
2. Further collecting is absolutely necessary.
3. Examination with electron microscope and scanning electron microscope will yield much new information.
4. Chemical tests, for instance fluorescence microscopy and staining, are promising.
5. Study of the wooden axes of cones and comparison with other fossil wood may gain an insight into their taxonomic position and relationship.