CHAPTER XXXVIII.

CYCADOPHYTA.

+I. Cycadean Stems other than Cycadeoidea.+

Most of the stems now under consideration are represented by casts or impressions and afford no information with regard to anatomical characters. They are in many cases more slender and less tuberous than typical Cycadeoideas, and a few are characterised by an irregular form of branching, as is shown in some specimens of pith-casts from Wealden strata in Tilgate Forest figured by Mantell[1318] and now in the British Museum. The genus _Wielandiella_[1319] (fig. 566) is an altogether distinct type represented by flowers as well as vegetative organs. Several generic names have been proposed for Cycadean stems agreeing with those of many recent Cycads in the possession of an armour of persistent leaf-bases, but distinguished from _Cycadeoidea_ in the absence of any fertile lateral shoots intercalated among the petiole-bases. It is, however, impossible in most cases to give any satisfactory definition by which these genera can be distinguished from one another; the characters employed by Carruthers[1320], Saporta[1321], and other authors are of comparatively little importance as trustworthy criteria and to a large extent are merely the expression of different states of preservation or of differences in age. Attention has elsewhere been called to the absence of any clear dividing line between stems referred to _Bucklandia_, _Yatesia_, _Fittonia_ and _Cylindropodium_. The species _Cycadeoidea gigantea_ described on a previous page affords an instructive example of the difficulty of drawing a generic distinction between certain types of Cycadean stems: in habit, in the form and structure of the leaf-bases, and in the ramenta this species is identical with other species of _Cycadeoidea_, but it differs in the absence of lateral fertile shoots, a feature that may have no morphological significance. It has already been pointed out that the absence of flowers intercalated among the leaf-bases may simply mean that the plant had not reached the stage of flower-production, or their absence may be due to some unfavourable conditions. Similarly the stems for which Saporta proposed the generic name _Clathropodium_ agree in every respect with _Cycadeoidea_ except in the absence, apparent or real, of lateral flowering branches. Such types as _Clathropodium foratum_ Sap. and _C. sarlatense_ Sap.[1322], the latter probably from Upper Jurassic beds and the former from an unknown locality, should be included in the genus _Cycadeoidea_. The stem referred by Saporta to his genus _Platylepis_ as _P. micromyela_[1323] was originally assigned to _Cycadeoidea_ and more recently Lignier has wisely adopted the original generic name. The generic term _Bolpopodium_, also instituted by Saporta[1324], is applied to small tuberous stems which appear to be identical with the _Cycadeoidea_ type.

Having regard to the meagre data supplied by casts of stems preserved in various stages of defoliation, and in view of the impossibility of drawing other than purely arbitrary generic distinctions, it is preferable to employ one generic name in a liberal sense for stems that there is good reason to regard as plants that cannot reasonably be referred to _Cycadeoidea_. The name _Bucklandia_ is thus employed, and a few examples are described in illustration of the external features of stems that are undoubtedly Cycadean but in most cases readily distinguished from _Cycadeoidea_. There are substantial grounds for stating that plants which bore flowers of the _Williamsonia_ type possessed stems having the characters of _Bucklandia_. It should, however, be remembered that we cannot always draw a clearly defined distinction between flowers included in _Williamsonia_ and _Bennettites_, or _Cycadeoidea_, particularly when they are represented only by detached ovulate strobili as in _Cycadeoidea_ (_Bennettites_) _Morierei_ and _Williamsonia scotica_.

=BUCKLANDIA.= Presl.

_Bucklandia_ was proposed by Presl[1325] for a plant described by Mantell[1326] from the Wealden of Tilgate Forest and compared by him to the Euphorbiaceae and arborescent Ferns; the same generic name was given by Robert Brown in 1832 to a recent member of the Hamamelidaceae. Stokes and Webb[1327] referred the same fossil to _Clathraria_, a name applied by Brongniart[1328] to certain types of Sigillarian stems and afterwards adopted by him for the Tilgate Forest species, _C. Lyelli_. Presl ‘with remarkable discrimination’ recognised the Cycadean nature of the specimen. Carruthers[1329] in his definition of _Bucklandia_ includes a statement as to the nature of the carpellary leaves and suggests that a cone associated with the stems may be a staminate strobilus: the cone is undoubtedly a megastrobilus of an Araucarian plant and there is no evidence with regard to the nature of either the male or female reproductive organs in the material that he describes though, as already pointed out, there are reasons for believing that _Williamsonia_ flowers were borne on branches of _Bucklandia_ stems. The flowering shoots were not short and intercalated among the petiole-bases as in _Cycadeoidea_ with the strobili barely projecting beyond the surface of the leaf-base armour, but they formed comparatively long branches, sometimes forked, at the apex of the main stem (_cf._ figs. 541–543).

_Bucklandia_ is usually represented by casts, from Rhaetic to Lower Cretaceous strata, differing from _Cycadeoidea_ in the absence of numerous axillary short fertile shoots, in the more slender form and greater length of the stems, and in the less uniform size of the persistent leaf-bases which assume various forms. Some of the specimens reach a length of 4 feet and afford evidence of occasional branching: the surface is covered with leaf-bases preserved as imbricate, broad, and obtuse or truncate scales (fig. 575), or as slightly convex polygonal areas in some cases showing a tendency towards an irregular zonal arrangement of larger and smaller leaf-bases (fig. 576). Within the armour of leaf-bases there may be a cast of the large pith the surface-features of which are practically identical with the cast of a recent pith reproduced in fig. 398. Casts of the pith preserved as separate fossils are included in the genus _Cycadeomyelon_.

There is evidence of the occurrence of more than one zone of vascular tissue in a stem from Lower Greensand beds of Bedfordshire described by Carruthers as _Yatesia Morrisii_[1330] (= _Bucklandia Yatesii_), and Dr Stopes[1331] has recently described a species, _B. buzzardensis_ (fig. 578), with several zones of conducting tissue. This feature has not so far been satisfactorily demonstrated in _Cycadeoidea_. An Indian species, _Bucklandia indica_, shows that the secondary xylem is more compact than in typical _Cycadeoidea_ stems, and the tracheids have multiseriate pitting.

Two long and narrow stems figured by Nathorst from the Rhaetic of Scania as _Bucklandia Saportana_[1332] differ from other species in the irregular arrangement of the leaf-bases which in certain regions are crowded as in the typical example of the genus shown in fig. 576, but in the intervening portions of the stem they are few in number and widely separated by the finely striated bark. This type, though similar to some specimens of English, Mexican, and Indian Bucklandias in the zonal differences in the leaf-bases, represents an extreme case of the alternation of smaller and crowded and larger and scattered leaf-scars. It is by no means unlikely that _Bucklandia Saportana_ forms a transition between _Bucklandia_ and the stem of _Wielandiella_ described by Nathorst from the same region: in _Wielandiella_ the leaf-scars are concentrated at the region of forking but a few occur elsewhere: in _B. Saportana_ there is no evidence of branching and in this respect it differs from _Wielandiella_.

_Bucklandia anomala_ (Stokes and Webb).

This species, from Wealden beds in Sussex, was first described by Stokes and Webb[1333] as _Clathraria anomala_, and the same type was figured by Mantell and other authors as _Clathraria Lyelli_. The specimens referred by Carruthers to _Bucklandia anomala_ and _B. Mantelli_[1334] do not exhibit any well defined specific differences, and there would seem to be no reason for retaining both specific names. The petiole-bases are usually sub-rhomboidal in form and convex or flat, but in some stems more of each petiole is preserved and the surface is covered with broad imbricate scales (fig. 575) similar to some of the detached scales described under the generic name _Cycadolepis_. Pith-casts occur both in connexion with the _Bucklandia_ stems and as detached specimens. The leaf-bases often show an irregular zonation of smaller and larger rhomboidal areas. The pith-cast in the lower part of the specimen from the Wealden of Cuckfield in Sussex shown in fig. 575 is 5 × 3·5 cm. in diameter.

[Illustration: Fig. 575. _Bucklandia anomala._ (British Museum, No. V. 3690; rather less than natural size.)]

_Bucklandia Ruffordi_ Seward.

This species, from the Wealden beds on the Sussex coast[1335], was originally described as _Fittonia Ruffordi_, but in the absence of any well defined distinctive features that can be regarded as of morphological significance it is better to include it in _Bucklandia_. The species affords a good example of a long and narrow type of stem, one specimen reaching a length of nearly 80 cm. with a breadth of about 10 cm.; the surface is covered with persistent leaf-bases 1·7 cm. in depth with a scar agreeing in size and shape with the base of a frond of _Otozamites Goeppertianus_ (Dunk.)[1336] found in the same beds. There is no indication of any alternation of large and small leaf-bases, and the species is characterised by the uniform size and relatively greater depth in a vertical direction of the leaf-base areas. In all probability the stem bore fertile branches similar to those of _Williamsonia gigas_ with flowers of the _Williamsonia_ type: the fronds may have been those known as _Otozamites Goeppertianus_, but this has not been demonstrated. A stem described by Carruthers from the Lias of Lyme Regis as _Yatesia gracilis_[1337] and afterwards included by me in _Cycadeoidea_[1338] is very similar to _B. Ruffordi_ in its long and narrow form and in the shape of the leaf-bases; it should be transferred to _Bucklandia_ as _B. gracilis_ (Carr.).

_Bucklandia Milleriana_ Carruthers.

This species was founded on a cast from Lower Oolite beds at Brora in Sutherlandshire[1339] characterised by leaf-bases very similar to those of _B. anomala_ but smaller. Casts from the same locality were named by Carruthers _Yatesia crassa_ and _Y. Joassiana_[1340], but an examination of specimens in the Dunrobin Museum leads me to regard these forms as indistinguishable from _B. Milleriana_. The specimen reproduced in fig. 576 from the Great Oolite of Brora illustrates the external characters of a typical stem and shows the variation in the size of the leaf-bases. A portion of the pith-cast is exposed in the lower part of the stem.

[Illustration: Fig. 576. _Bucklandia Milleriana._ Brora, Scotland. (Manchester Museum, L. 7229. _ca._ ½ nat. size.)]

_Bucklandia Yatesii_ (Carruthers).

This type from the Lower Greensand of Bedfordshire (fig. 577) was described by Carruthers as _Cycadeoidea Yatesii_ and subsequently named _Yatesia Morrisii_[1341]. Ward expressed the opinion that the name should be _Yatesia Yatesii_, but as _Bucklandia_ is now used to include _Yatesia_ this combination is fortunately avoided. The stem is cylindrical, 20–30 cm. long and 12 cm. in diameter, covered with rhomboidal leaf-bases separated from one another by a ramental reticulum. There are two concentric vascular cylinders as stated by Carruthers. In a recent account of this species Dr Stopes[1342] adds further details: the xylem-cylinders are 5–8 mm. wide and the tracheids occur in single rows or there may be bands 4–5 elements broad; the circular bordered pits are uniseriate or in two alternate series. The medullary rays are broad but the cells are not preserved. The pith-cast is of the usual Cycadean type.

[Illustration: Fig. 577. _Bucklandia Yatesii._ Type-specimen in the British Museum.]

The type-specimen was presented by the Cirencester College to the British Museum.

_Bucklandia buzzardensis_ (Stopes).

[Illustration: Fig. 578. _Bucklandia buzzardensis._ Rough sketch of a block of wood showing parts of at least eight concentrically arranged rings of secondary wood. (After Stopes; × ⅔.)]

This species, from Lower Greensand beds at Leighton Buzzard and believed to be derived from Wealden strata, is described by Dr Stopes[1343] as _Cycadeoidea buzzardensis_. Though agreeing generally with _B. Yatesii_, the stem is specifically separated on the ground that the petiole-bases are more expanded laterally and because of the occurrence of several vascular cylinders (fig. 578), sometimes as many as eight, each with a maximum diameter of 1 cm. Dr Stopes thinks it possible that _B. buzzardensis_ is an older form of _B. Yatesii_.

_Bucklandia squamosa_ (Brongniart).

Sternberg first described this species as _Conites Bucklandi_[1344] and regarded it as a cone bearing large imbricate cone-scales; it was named by Brongniart _Bucklandia squamosa_[1345] and Carruthers[1346] retained this designation. The type-specimen, in the Oxford Museum, from the Stonesfield Slate is 18 cm. long, showing in the lower part a cast of the pith. The surface of the stem is covered with thick imbricate petiole-bases very like those on the stem of a recent _Encephalartos_.

_Bucklandia_ (_Fittonia_) _squamata_ (Carruthers).

Carruthers[1347] founded the genus _Fittonia_ on a single specimen from the Wealden beds of the Isle of Wight, separating it from _Bucklandia_ on the ground of the occurrence on a portion of the stem of large imbricate leaf-bases which are at first reflexed and then ascending; the stem is also broader and more tuberous than most species of _Bucklandia_. The type-specimen, in the Museum of the Geological Survey (Jermyn street), bears a close resemblance to a trunk of a recent _Encephalartos_, but the part of the stem from which the imbricate stumps have fallen is practically identical with a _Bucklandia_. As in certain recent Cycads the surface-features probably changed with the age of the plant; when the foliage-leaves were first shed a portion of the ascending petiole remained on the stem, and at a later stage this was cut off leaving a clean-cut rhomboidal scar like those on the _Bucklandia_ shown in fig. 576. The difference between _Fittonia_ and _Bucklandia_ may, therefore, be a question of age. While substituting _Bucklandia_ for _Fittonia_ as the generic name the latter designation is added in parentheses to denote the possession of certain features which, though possibly of generic value, are not regarded as sufficiently important morphologically to warrant generic recognition.

The type-specimen of Saporta’s species _Fittonia insignis_[1348], in the Paris Museum, from the Oxfordian of Poitiers, appears hardly distinguishable from _F. squamata_ Carr. Another type with broader imbricate petiole stumps is described by Saporta from the Portlandian near Boulogne as _Fittonia Rigauxi_[1349].

[Illustration: Fig. 579. _Bucklandia indica._ A, side-view showing the leaf-bases and attached petioles of _Ptilophyllum_, also (to the left) a fragment of a detached _Ptilophyllum_ leaf. B, transverse section showing the pith, xylem cylinder, cortex, and sections of petioles. (British Museum; nat. size.)]

_Bucklandia indica_ sp. nov.

Oldham and Morris[1350] and subsequently Feistmantel[1351] described some specimens of Cycadean stems from the Rajmahal Hills of India of Lower Jurassic age: the latter author regarded them as stems of _Williamsonia_ because of their association with flowers of that type, a conclusion fully justified by the evidence. Feistmantel also called attention to the resemblance of the Indian stems to specimens described from British strata as _Bucklandia_ and _Yatesia_. Although the Indian examples are very similar to stems from Mexico discovered by Wieland[1352] and to some of the English types, it seems desirable to refer to them under a specific name and I therefore suggest the institution of the specific name _indica_, the type-specimen being that represented in fig. 579. This specimen is particularly interesting because it affords some information as to anatomical features and is one of the few fossil stems preserved in organic connexion with leaves (fig. 579, B). A short account of it was published in 1900[1353] and more recently Miss Bancroft[1354] has made a fuller investigation of this and other Indian specimens. The stem shown in fig. 579 from the Rajmahal Hills, and now in the British Museum, bears fronds of _Ptilophyllum cutchense_ Morr., a type that appears to be indistinguishable from _P. pecten_; and with similar stems from the same beds are associated flowers of _Williamsonia_. Miss Bancroft describes a bract-covered shoot which agrees very closely with those of English stems reproduced in figs. 541, 542. In addition to the evidence based on close association, there is the more important argument furnished by the discovery of ramental hairs like those on the bracts of _Williamsonia scotica_ and of anatomical characters in the bracts similar to those in the Scotch strobilus. The persistent leaf-bases are far from uniform in size; in this respect and in their form they agree closely with those on _Bucklandia_ stems from English and Mexican localities. The secondary wood is more compact than in recent Cycads or in _Cycadeoidea_, though it resembles that of _Cycadeoidea micromyela_; the medullary rays are uniseriate and the tracheids have multiseriate bordered pits on their radial walls instead of the scalariform pitting in the majority of _Cycadeoidea_ stems. Secretory canals are abundant in the parenchymatous ground-tissue; the cambium and phloem are not preserved[1355].

The transparent nature of the silicified material rendered very difficult the examination of the tissues, but enough was discovered to show that these Indian stems are characterised by certain features, the more compact nature of the secondary xylem and the presence of multiseriate pitting, which distinguish them from the _Cycadeoidea_ type. Further knowledge of the anatomical features of the _Williamsonia_ (_Bucklandia_) stems from other localities might enable us to recognise these or other peculiarities as constant distinguishing characters of _Bucklandia_ in contrast to the Cycadeoidea stems which bore the _Bennettites_ type of flower.

=Cycadeomyelon.= Saporta.

Casts of the pith-cavity of Cycadean stems, like that shown in fig. 575 projecting beyond the armour of leaf-bases, are occasionally found as separate fossils and cannot always be referred to a particular species of stem. For such detached casts Saporta[1356] instituted the name _Cycadeomyelon_: they are characterised by their comparatively large diameter and by the possession of surface-features similar to those on the corresponding cast from a recent Cycadean stem shown in fig. 398, namely spirally disposed, more or less prominent, lozenge-shaped areas formed by the sand or mud filling the cavities left on the decay of the parenchyma of the broad medullary rays of a manoxylic stem. Occasionally a slit at the lower end of a medullary ray area marks the position of the leaf-trace bending outwards from the lower angle of the mesh in the xylem-lattice[1357]. Lignier figures part of a pith-cast of _Cycadeomyelon Apperti_[1358] in which each medullary-ray area has a circular depression and not a slit extending from the lower angle: this may indicate that the surface shown on the cast is slightly external to the inner edge of the stele and in a plane where the leaf-traces were embedded in the parenchyma of the rays and free from the xylem-cylinder.

Large and branched examples of _Cycadeomyelon_ were figured by some of the earlier authors from English Wealden beds as species of _Clathraria_[1359] and in many cases these are undoubtedly pith-casts of _Bucklandia_ stems: a similar cast is figured under this name by Schenk from the Wealden of North Germany. From Liassic beds in Normandy Lignier figures two species of _Cycadeomyelon_, _C. Apperti_ and _C. densecristatum_. The surface-features of _Cycadeomyelon_ resemble those of the Palaeozoic genus _Tylodendron_ (see Vol. +iv.+), but in the latter genus the nodal swellings are a characteristic peculiarity. Though medullary casts of this type are of no great botanical importance and their specific distinctions are of little value, it is safe to assume that broad medullary casts with comparatively large lozenge-shaped areas belong to Cycadean stems, while narrower specimens with smaller lozenges are more likely to be pith-casts of Coniferous stems.

Lester Ward[1360] instituted the genus _Feistmantelia_ for some Lower Cretaceous casts from the Black Hills which he compared with an Indian fossil from Cutch described by Feistmantel as ‘the stem of a Coniferous plant[1361],’ and with pith-casts figured by Stokes and Webb as _Clathraria anomala_. It is impossible to determine the systematic position of such imperfect specimens as that on which Ward founded his species _F. oblonga_: they may, as Hollick and Jeffrey[1362] suggest, be casts of the bark of some Conifer; there is certainly no good reason for connecting them with Cycads.

=COLYMBETES.= Stopes.

_Colymbetes Edwardsi_ Stopes. This genus[1363] is founded on the inner portion of a petrified trunk which was probably cylindrical and more than 12 cm. in diameter, consisting of a pith, 7·5 cm. in diameter, and part of a vascular cylinder of remarkable structure. The type-specimen is of Aptian age and may have come from Leighton Buzzard (Bedfordshire). The pith (fig. 580, _p_) consists of large parenchymatous cells and numerous secretory canals: the perimedullary zone, _pm_, is characterised by the occurrence of loosely disposed tracheids in groups and radial rows pursuing a sinuous longitudinal course in the accompanying parenchyma. The tracheids in this region are small in diameter and have oval, scalariform, or circular pits. Abutting on the perimedullary zone is the secondary xylem the inner edge of which forms bays, and this is composed of alternating zones of vertical and horizontal tracheids (fig. 580, _y₁_–_x₅_; fig. 581) with bordered, scalariform, pits on their walls traversed by medullary rays generally biseriate and from 4 to 30 cells deep. The disposition of the tracheids is such as to render transverse and radial longitudinal sections practically identical in appearance; the first zone of secondary xylem with its bayed inner edge consists of vertically running elements; this is succeeded by a zone in which the tracheids pursue a horizontal course, and beyond this second zone is another band of vertical elements (fig. 581). ‘Where the one zone passes into the next, a curving of the elements is frequently evident, and in a few cases it is quite possible to trace a single radial series of tracheids through an angle of 90° running in the same section, first as a transverse and then as a vertical series. One and the same medullary ray also can sometimes be followed, first in transverse and then in radial longitudinal section, which later again turns to true transverse. The inference is therefore drawn that there was but a single cambium, which had periodic changes of direction.’ Leaf-traces (fig. 580, _lt_) are large and numerous; they are spirally disposed and pass nearly straight through successive xylem-zones: each trace consists of a small-celled ground-tissue including stone-cells and patches of tracheids in more or less regular radial rows. Tangential sections of the wood show that the tracheids follow a sinuous course forming loops enclosing numerous medullary rays.

[Illustration: Fig. 580. _Colymbetes Edwardsi_. Transverse section showing ten alternating zones of wood outside the perimedullary zone, _pm_; _y₁_, _y₂_, etc., vertically running xylem series; _x₁_, _x₂_, etc., horizontally running xylem series; _p_, pith; _lt_, leaf-traces. (After Stopes; × 2.)]

[Illustration: Fig. 581. _Colymbetes Edwardsi._ Diagram of stem in transverse (A) and radial longitudinal (B) section. _p_, pith; _pm_, perimedullary xylem; _b_, bays of first, vertically running, secondary xylem; _x₁_, _x₂_, etc., zones of horizontally running secondary xylem cut transversely in the radial and radially in the transverse section of the stem; _y₁_, _y₂_, etc., longitudinally running xylem cut transversely in the transverse and longitudinally in the radial section. (After Stopes.)]

As the pith and xylem are the only tissues preserved it is on their structure that any speculation as to affinity must be based. The close arrangement of the leaf-traces (about 1 cm. apart), as Dr Stopes says, indicates small leaf-bases, assuming that each leaf received a single trace. In some respects the xylem and medullary rays resemble those of Cycads, and the author of the genus includes it in the Cycadophyta; but as she points out there are many peculiar features, and it is clearly impossible to assign the new type to a more precisely defined position. The possibility of any purely mechanical explanation of the course of the tracheids in the alternating zones is ruled out by the straight course of the outgoing leaf-traces, and it would seem that the cambium must have turned over at right-angles at regular intervals during the growth of the stem.

=Cycadolepis.= Saporta.

This name was used by Saporta[1364] for linear-lanceolate scales from Upper Jurassic rocks in France which he compared with bud-scales of recent Cycads. The imperfect scale described as _Cycadolepis villosa_ bears a striking resemblance to the hairy bracts of _Williamsonia_ and may well belong to that genus. Saporta’s term may be usefully employed in a more extended sense, including not only lanceolate scales but larger and much broader scales resembling the flattened petiole-bases on stems of _Macrozamia_, _Encephalartos_, and some other recent genera, as well as detached carpellary scales, other than _Cycadospadix_, and microsporophylls which cannot be assigned to a particular stem. Two qualifying subgeneric terms have been proposed[1365]:

i. _Cycadolepis_ (_Dory-Cycadolepis_). Scales more or less linear-lanceolate like those described by Saporta and a specimen from Jurassic rocks of India named by Feistmantel[1366] _Cycadolepis pilosa_. This type of _Cycadolepis_ may be identical with the bracts of _Williamsonia_ flowers, though in the absence of any definite evidence of such affinity the provisional generic name is more appropriate.

ii. _Cycadolepis_ (_Eury-Cycadolepis_). Broadly oval or orbicular thick scales (figs. 582, 583), the broadest part being frequently nearer the distal than the proximal end. These larger scales though usually found as detached fossils have in one instance been obtained attached to an imperfectly preserved stem.

[Illustration: Fig. 582. _Cycadolepis_ (_Eury-Cycadolepis_) sp. From the Wealden beds of Sussex. (Brit. Mus. No. V. 2799.)]

[Illustration: Fig. 583. _Cycadolepis_ (_Eury-Cycadolepis_) sp. Scale from the Wealden beds of Sussex. (British Museum, No. V. 2131_a_; nat. size.)]

_Eury-Cycadolepis_ sp.

This type of scale, represented by specimens from the Wealden of Sussex[1367] (figs. 582, 583), reaches a length of 13 cm. and a breadth of 7 cm. and is sometimes almost orbicular. The lamina is convex but shows no definite venation and bears a close resemblance to the scale-like petiole-stumps on an old stem of _Macrozamia_. On some of the smaller specimens (fig. 583) several forked veins extend vertically from the broad base. Since these specimens were first described additional examples have been discovered in the Wealden beds of Sussex, some of which are attached to a piece of stem[1368] in such a manner as to give support to the view that they are leaf-bases very similar to those on such fossil stems as _Bucklandia_ (_Fittonia_) _Rigauxi_ (Sap.)[1369] and _B._ (_Fittonia_) _squamata_ (Carr.)[1370]. One partially carbonised scale yielded pieces of cuticle showing numerous stomata similar to those of recent Cycads and the outlines of very thick-walled epidermal cells[1371].

_Eury-Cycadolepis Jenkinsiana_ (Tate).

The large and approximately orbicular or broadly ovate scales so named are believed to be identical with Tate’s _Cyclopteris Jenkinsiana_[1372] from the Uitenhage series of Cape Colony (Wealden). The scales reach a length of 12 cm. and were attached by a broad base; the lamina, which may be strongly bent as though folded over some immature organ as a protective bract, shows numerous repeatedly forked veins of the _Cyclopteris_ type and several anastomosing and irregular lines between the veins suggesting that the scales were tomentose.

+II. Reproductive Organs of Cycadean Plants other than those of the Bennettitales.+

The fact that practically all known Cycadean stems bore flowers either of the _Bennettites_ or _Williamsonia_ type prepares us for the scarcity of reproductive organs like those of recent Cycads. No specimens have been discovered in a petrified state affording any evidence of their close affinity to the cones, sporophylls, or seeds of the Cycadales. Such genera as _Cycadospadix_, _Androstrobus_, and _Zamiostrobus_, as the following descriptions show, are founded on material that is too imperfect to throw much light on their true morphological nature. The probability is that some at least of the specimens included in these genera are the reproductive organs of Cycadean plants more closely allied to the existing Cycads than to the Bennettitales. Among the numerous fossil seeds referred to such genera as _Cycadeospermum_ and _Cycadinocarpus_ there are but few that can confidently be assigned to the Cycadales rather than to the Ginkgoales or Coniferales. While the seeds of the Bennettitales are clearly distinguished by their much smaller size from those of modern Cycads, many of the latter agree in size and form with those of some other Gymnosperms and in the absence of anatomical details could not easily be identified as fossils. Some of the examples included in the miscellaneous collection described by authors as species of _Carpolithus_ or _Carpolithes_ agree closely in external features with the seeds of modern Cycads, but it is seldom possible to accept them as undoubted records of Cycadalean plants.

The general conclusion is that such meagre evidence as we possess affords strong confirmation of the conclusion based on stems and foliage from Jurassic and Cretaceous strata, namely that the present representatives of the Cycadophyta are a relatively late product of evolution, though retaining in their anatomical features many survivals from a remote antiquity. The occurrence of Cycadean characteristics in the vegetative organs of the Medulloseae and the recurrence of what may be called the Cycadean seed-plan, with certain more or less striking peculiarities reminiscent of earlier stages of evolution, in several types of Palaeozoic seeds such as _Cycadinocarpus_, _Stephanospermum_, _Lagenostoma_ and others bear testimony to the antiquity of the Cycadean stock.

=CARPOLITHUS[1373].= Linnaeus.

This generic name, as Nathorst[1374] has recently pointed out, was used by Linnaeus in 1768 for ‘Phytolithus fructus’ and has therefore priority over Sternberg’s genus _Carpolites_ employed in 1825. Lester Ward[1375] attributes _Carpolithus_ to Stokes and Webb (1824) and states that in the plural form the name was used by Walch in 1771. _Carpolithus_ is a convenient term to apply to fossil seeds that cannot be assigned to a particular group of plants and which do not exhibit any peculiarities of form sufficiently striking to deserve generic recognition. Pomel[1376] proposed the genus _Ulospermum_ but it never came into general use. Schimper’s genus _Cycadinocarpus_ and Saporta’s _Cycadeospermum_ (preferable in its morphological implication), though useful in the case of detached seeds of undoubted Cycadean affinity, can seldom be employed without an admission that they may imply a relationship that cannot be absolutely established. In the great majority of cases the better plan is to be content with the more non-committal term _Carpolithus_ with the addition of a family-name when there are reasonably good grounds for a more definite reference. No useful purpose would be served by attempting a complete survey of the numerous casts and impressions of supposed Cycadean seeds recorded in palaeobotanical literature, but a few types are briefly described as examples of specimens with fairly well defined characters, which are in all probability Cycadean.

_Carpolithus conicus_ (Cycadales?) Lindley and Hutton.

[Illustration: Fig. 584. _Carpolithus conicus._ The type-specimens in the Manchester Museum; nat. size. (_a_, No. 361; _b_, _c_, No. 360.)]

[Illustration: Fig. 585. _Carpolithus conicus._ From a specimen in the Malton Museum; nat. size.]

The original specimen figured by Lindley and Hutton[1377] from the Coralline Oolite of Malton, Yorkshire, as _Carpolithes conica_ and now in the Manchester Museum, is represented in fig. 584. A second ‘species,’ _Carpolithus Bucklandi_ Lind. and Hutt. ex Will. +MS.+[1378], from the same locality is no doubt specifically identical with _C. conicus_. The seeds are conical, broadly truncate at one end, presumably the base, and tapered to a blunt apex; the broad end is characterised by the presence of three ridges or in some specimens by a single median ridge illustrating an oscillation between the radiospermic and platyspermic form similar to that in _Ginkgo biloba_. As usually obtained the seeds are probably nucules or casts showing the surface-features of the inner wall of the sclerotesta, the sarcotesta having been destroyed before fossilisation: the irregular marginal teeth at the truncate end suggest casts of vascular bundles in the integument. The scattered tubercles on the sides of some of the seeds (fig. 584, _a_) are probably casts of holes in the shell bored by insects and comparable with those occasionally preserved on the casts of _Trigonocarpus_. A specimen in the Malton Museum shown in fig. 585 which may be an example of this species illustrates the occurrence of an internal cast enclosed by the remains of a thick testa. These Jurassic casts resemble the seeds of _Macrozamia Fraseri_, but it is impossible to determine their systematic position with confidence.

_Carpolithus_ sp. (Cycadales?) Seward.

An unusually well preserved specimen from the Wealden beds of the Sussex coast described under this name in 1895[1379] consists of a kernel and mould, 1·8 × 1·1 cm. The mould from which the kernel is readily removed is lined with a thin structure representing part of the testa and between this and the surrounding rock is a layer of coal. On the surface of the kernel, probably the cast of the seed-cavity, a reticulum of narrow grooves indicates the course of the vascular bundles over the surface of the nucellus.

_Carpolithus_ (Cycadales?) _Pomelii_ (Saporta).

The specimen from the Upper Corallian of Châteauroux (Indre) on which this species was founded by Saporta[1380] under the name _Cycadeospermum Pomelii_ is a large ovate cast, 5·5 cm. long and 3·5 cm. broad, closely resembling some of the larger recent Cycadean seeds: it cannot be accepted as a true record of the group without reservation.

Saporta describes other species of _Cycadeospermum_ but none of them are of any real importance from a botanical point of view: the same remark applies to the seeds referred by Fontaine[1381] from Potomac beds to the same genus, also to many other recorded examples of seeds that afford no decisive evidence of affinity.

Some specimens described by Compter[1382] from the Lettenkohle of Apolda (Thuringia) as Cycadean fruits—too imperfect to be determined with accuracy—furnish an additional illustration of the slender foundation on which many of the records of supposed Cycadean reproductive organs are based.

=CYCADOSPADIX.= Schimper.

This name was proposed by Schimper[1383] for some French Jurassic fossils, described by Pomel[1384] as _Crossozamia Hennocquei_ and _C. Moraeana_, on the ground that they bear a close resemblance to the megasporophylls of _Cycas_. Their occasional association with _Otozamites_ fronds suggested a reference to the same parent-plant, but such data as we have point to _Otozamites_ fronds having been borne by plants with the _Williamsonia_ type of flower. Schenk[1385], who figured a specimen of _Cycadospadix_ from France as the inflorescence of a Cycad, expresses the more probable opinion that it belonged to a plant with _Cycadites_ fronds. A Permian species described by Renault as _Cycadospadix Milleryensis_ is now transferred to the genus _Strobilites_[1386].

_Cycadospadix Pasinianus_ Zigno.

This species, first described from Jurassic strata in Northern Italy[1387], is recorded also from the Kimmeridgian of France[1388] and Scotland[1389]. Zigno’s figures give a fairly accurate representation of the type-specimens in the Padua Museum. The megasporophylls, almost identical in shape with those of some recent species of _Cycas_ (figs. 381; 392, A–C), consist of a broadly lanceolate or triangular limb with deeply laciniate sides terminating a pedicel, or the distal expansion may be preserved without the stalk from which it was no doubt easily detached as in certain recent Cycads (_cf._ fig. 392, A). In the specimen from Scotland there are no clear indications of veins in the lamina, which may have been woolly as in _Cycas_. In some of the specimens figured by Saporta[1390] and now in the École des Mines, Paris, the stalk is absent, but in pedicellate examples scars occur on the sides of the narrow axis and casts of seeds are found in the same beds. A good example of _Cycadospadix Hennocquei_ is figured by Saporta from a drawing supplied by Schimper showing two seed-scars near the base of the lamina: the same specimen, as figured by Saporta and Marion[1391], bears a seed, but this is presumably a partial restoration. The occurrence of _Cycadites rectangularis_ Brauns at Hettange in association with _Cycadospadix_ strengthens the conclusion, based on the form of the megasporophylls, that some of the Jurassic Cycads bore megasporophylls very similar to those of existing species of _Cycas_.

_Cycadospadix integer_ Nathorst.

This Rhaetic species from the south of Sweden[1392] was instituted for an imperfect broadly lanceolate lamina recalling the distal end of the megasporophyll of a _Cycas_: the discovery of a more complete example[1393] justifies Nathorst’s use of the name _Cycadospadix_, though without further evidence one hesitates to regard the species as a thoroughly trustworthy record of a Cycadean fertile leaf. The species is characterised by the entire margin of the broad and relatively short and thick terminal limb borne on a broad stalk with alternate lateral projections presumably marking the position of the seeds.

These species of _Cycadospadix_ are particularly interesting as evidence—though not amounting to demonstration—of the production by some Jurassic and Rhaetic plants of fertile leaves agreeing closely with those of _Cycas_. It would seem from the abundance of Bennettitalean flowers and the very scanty remains of fertile leaves or cones like those of modern Cycads that the existing type was exceptional in Mesozoic floras.

=BEANIA.= Carruthers.

_Beania gracilis_ Carruthers.

[Illustration: Fig. 586. _Beania gracilis._ (After Carruthers; ½ nat. size.)]

The generic name _Beania_[1394] was given to a branched fertile shoot (fig. 586) from the Middle Jurassic beds at Gristhorpe, Yorkshire, characterised by loosely disposed sporophylls bearing two sessile seeds: each sporophyll is given off at a wide angle from a fairly stout axis and the seeds are borne on the adaxial side of a peltate distal expansion. Carruthers compared the type-species with a cone of _Zamia_ with which it agrees in the general plan of construction but differs in the more open habit and in the longer and more slender seed-bearing pedicels. The same type of shoot was figured by Lindley and Hutton[1395] as _Sphaereda paradoxa_. _Beania_ is generally regarded as a Cycadean reproductive shoot, but there is no doubt that the majority of Jurassic Cycadophyta possessed flowers of the _Bennettites_ types, and it is clear that _Beania_ differs considerably from _Bennettites_ and _Williamsonia_. Another suggestion is that _Beania_ may belong to some member of the Ginkgoales[1396]: though very different from the normal ovuliferous shoot of a _Ginkgo_, it resembles some abnormal forms (_e.g._ fig. 631, D) in which the ovules occur on elongated pedicels, but they are borne singly and the micropyle is directed outwards, while in _Beania_ the ovules are attached in pairs to the inner face of a distal expansion. There is no conclusive evidence in support of either interpretation, though the general agreement between the Jurassic type and the cones of recent Cycads would seem to favour the inclusion of _Beania_ among the Cycadophyta.

A specimen described by Nathorst[1397] from Upper Jurassic rocks in the North of Scotland as _Beania Carruthersi_ closely resembles the type-species, differing chiefly in its smaller size and in the rather closer arrangement of the sporophylls. The seed-like bodies borne in pairs on the adaxial side of the terminally expanded pedicels are covered with small granulations which Nathorst thinks may be clusters of microspores, the apparent seeds being ‘antherangia.’ The granulations are, however, very similar to those on the larger detached seed-like bodies originally described by Nathorst from Rhaetic beds in Sweden as _Antherangiopsis rediviva_[1398]: subsequent examination of that species demonstrated that the granulations are due to the presence of resinous bodies in the tissues of true seeds[1399], and it is not improbable that a similar interpretation may hold for the surface-features in the supposed male organs of _Beania Carruthersi_. Pending further evidence it may be suggested that _Beania Carruthersi_ is like _B. gracilis_ a seed-bearing shoot. The Rhaetic specimens described by Nathorst as _Stenorrachis scanicus_[1400] are similar in habit to _Beania_ but differ in the forking of the sporophylls (fig. 656) and in the absence of any terminal swelling on which the seeds are borne: Nathorst considers that _Stenorrachis_ may be the female organ of a _Nilssonia_ and it is not improbable that that genus and _Beania_ are closely allied types. We have no definite information with regard to the reproductive organs of the Nilssoniales: the closer resemblance which their fronds bear in the structure of the epidermal cells to those of recent Cycads is consistent with the view that their fertile shoots were also more like those of existing types. It is, however, still an unsettled point whether _Beania_ is more closely allied to the Cycadophyta or to the Ginkgoales, but the balance of opinion is in favour of the former alliance.

=Zamiostrobus.= Endlicher.

=Cycadeostrobus.= Carruthers.

Though instituted by Endlicher[1401] for a cone figured by Lindley and Hutton as _Zamia macrocephala_[1402] which is almost certainly Abietineous and has no claim to be included in the Cycadales, the genus _Zamiostrobus_ has been adopted by many authors for Cycadean ovulate cones, not only such as are believed to be closely allied to those of _Zamia_ but for Cycadean cones generally. Carruthers[1403] suggested _Cycadeostrobus_ as a more suitable name on the ground that it is less limited in its implication of affinity; but, as Fliche points out, Endlicher’s generic name has been widely adopted in a comprehensive sense as standing for Cycadean megastrobili, excluding the supposed _Cycas_-like megasporophylls, included under _Cycadospadix_.

Many of the specimens described as species of _Zamiostrobus_ are of little or no value as records of Cycadean plants, _e.g._ _Zamiostrobus orientalis_ Heer[1404] from the Jurassic beds of Amurland. A Lower Cretaceous (Albian) species described by Fliche[1405] as _Zamiostrobus Loppineti_, though not entirely satisfactory, is more likely to belong to the Cycadales. The type-specimen is an elliptical strobilus, 5·5 cm. × 3·2 cm., consisting of an axis bearing at right-angles numerous small, contiguous, peltate megasporophylls each with two small seeds on the lower surface. The figures given by Fliche are, however, not convincing. An examination of specimens in the British Museum, from Wealden and Jurassic rocks, described by Carruthers as species of _Cycadeostrobus_, convinced me that several are undoubtedly Araucarian cones[1406]. Solms-Laubach[1407] called attention to the Araucarian appearance of _Cycadeostrobus Brunonis_, a cone from an unknown locality, and this with other species, _e.g._ _C. elegans_, _C. sphaericus_, _C. truncatus_, etc., may safely be referred to _Araucarites_. The specimen figured by Lindley and Hutton[1408] as _Zamia crassa_ from the Inferior Oolite of Towcester (Northamptonshire) affords no satisfactory evidence of Cycadean affinity. The Lower Cretaceous Bohemian specimens described by Corda[1409] and Velenovský[1410] as _Microzamia gibba_ should not be included in a genus implying Cycadean affinity: though Velenovský states that the megasporophylls bear a pair of seeds his illustrations do not afford any satisfactory evidence of this Cycadean character. Similarly the fossil regarded by Carruthers[1411] as probably a male flower of _Bucklandia_ is almost certainly an Araucarian cone. A small cone from the Lower Miocene of Armissan (Aude) named by Schimper _Zamiostrobus Saportana_ and figured by Saporta and Marion[1412] may, as Solms-Laubach says, be Cycadean, but we have no information with regard to the internal structure or as to the presence or position of the seeds.

=Androstrobus.= Schimper.

Schimper[1413] instituted this genus for ‘amenta, cycadeacea antherifera, cylindrica, e squamis imbricatis, latere postico antheras sessiles ferentibus efformata.’ It may conveniently be applied to fossils which resemble the male cones of recent Cycads sufficiently to justify the use of a name implying relationship. As so defined, _Androstrobus_ is used in a more restricted sense than the word suggests, just as _Masculostrobus_[1414] has been employed for fossils that are believed to be the corresponding organs of Conifers. Among the few species assigned to Schimper’s genus reference may be made to _A. Balduini_ Sap., originally named by Schimper _A. zamioides_, from the Upper Bathonian of Etrochy, and _A. Guerangeri_ (Brongn.), another French type[1415]. Heer’s species _A. sibirica_[1416] of Jurassic age is represented by a slender axis bearing numerous appendages which in surface-view have the form of polygonal discs: there are no indications of microsporangia and the evidence of Cycadean affinity is far from convincing. Nathorst’s Rhaetic species _A. borealis_[1417] is no more satisfactory as a record of a Cycadean strobilus. A fossil from the Lower Cretaceous of Bohemia described as _Zamites familiaris_ and regarded by Corda[1418] and Carruthers[1419] as a male flower of a Cycad though not above suspicion may be included in _Androstrobus_.

Under the name _Fričia nobilis_ Velenovský[1420] described some cones from the Lower Cretaceous plant-beds of Bohemia which he regards as male strobili of some Cycadean plant: the cone, as shown in Velenovský’s restoration, is 10 cm. long and 5 cm. in diameter; it bears a close superficial resemblance to a large cone of _Zamia_ and consists of a stout axis bearing contiguous peltate, hexagonal, scales gradually contracted towards the proximal end, similar to those of _Androstrobus_. The evidence on which this species is identified as a male cone rests on the occurrence of numerous pits on the surface of the scales; but no spores or sporangia were found and the pits as shown in the published figures do not present the appearance of scars of sporangia.

The Rhaetic specimen originally named by Nathorst _Androstrobus Scotti_ and afterwards transferred to the genus _Lycostrobus_ was described in Volume +ii.+[1421]

_Androstrobus Nathorsti_ Seward.

The type-specimens were obtained from the Wealden beds of Sussex: they were referred to the genus _Androstrobus_ on evidence which cannot be regarded as decisive[1422]. A fairly stout axis, 6·5 cm. long, bearing spirally disposed sub-triangular scales hexagonal in section and attached to the axis by a broad base; the scales, or sporophylls, are 1–1·5 cm. long and gradually tapered towards a pointed or slightly rounded apex. Near the proximal end of some of the sporophylls there are regularly arranged polygonal depressions which may be impressions of microsporangia. The regular disposition of the depressions is a striking feature and in contrast to the less regular reticulum exposed after the removal of the sporangia from the microsporophyll of a recent Cycad. An examination of the cuticular membrane of the microsporophylls shows that the epidermal cells have thick and straight walls[1423], characters consistent with the supposed Cycadean affinity of the strobilus.