Fossil plants, Vol. 2

Frond bipinnate, broad rachis giving off branches at an acute angle; pinnules broadly linear, slightly falcate, with several slightly divergent forked veins.

A frond very similar to the Lower Lias specimen from Dorsetshire represented in fig. 362 was described by Leckenby as Ctenis Leckenbyi (Bean MS.) from the Inferior Oolite of Yorkshire[1466]. Leckenby recognised the possibility of a Cycadean affinity, but regarded the bipinnate habit as an objection. The branched fronds of the Australian Cycad Bowenia supply an answer to this objection. Several good examples of Ctenopteris cycadea are figured by Schenk[1467] from Rhaetic rocks of Persia. Zeiller’s Tonkin Rhaetic species, C. Sarrani[1468], affords a striking illustration of the difficulty of drawing a clear line of separation between Ctenopteris and some species of Thinnfeldia.

Ctenopteris is in all probability very closely related to Thinnfeldia and Ptilozamites.

Dichopteris.

This genus was proposed by Zigno[1469] for some large specimens from the Jurassic plant-beds of Northern Italy.

The bipinnate leaves are characterised by the great breadth of the rachis which is dichotomously branched in the distal region (fig. 363); the linear pinnae reach a considerable length. Pinnules relatively small, oblong and slightly contracted at the base; the decurrent and confluent lamina forms a narrow wing to the main axis. Veins slightly divergent and forked, as in Ptilozamites.

Dichopteris visianica, Zigno. Fig. 363.

A specimen of this species in the Padua Museum has a total length of 83 cm. It has been elsewhere suggested[1470] that a fragment figured by Zigno as a fertile example of this type is probably part of a frond of the Osmundaceous fern Todites. Since this opinion was expressed I have had an opportunity of examining the actual specimen at Padua: the circular patches described by Zigno as sori appear to be irregularities in the matrix and not an original feature.

Brongniart[1471] instituted the genus Pachypteris for some imperfectly preserved English Jurassic fossils from Whitby, which he described as P. lanceolata. Specimens have since been described[1472] from the Inferior Oolite rocks of the Yorkshire coast. Brongniart described the pinnules as being without veins or as possessing only a midrib. It is almost certain that the apparent absence of veins in most specimens[1473] is due to the fleshy nature of the segments and that the species P. lanceolata should be transferred to Dichopteris.

Krasser[1474] has described a species from Cretaceous rocks of the island of Lesina, off the Dalmatian coast, as Pachypteris dalmatica which is very similar in habit to the English specimens and to Zigno’s Dichopteris visianica. One of Krasser’s specimens is practically identical with Dichopteris lanceolata (Brongn.), while in others the small pinnules are replaced in some of the pinnae by a continuous lamina with a few distal serrations. The latter form a link between the Dichopteris and Thinnfeldia type of segment. Krasser gives a full résumé of opinions expressed by other authors in regard to the position of Pachypteris (= Dichopteris) and decides in favour of a Cycadean alliance.

A French Jurassic plant which Saporta[1475] made the type of a new genus Scleropteris, and described as S. Pomelii, appears to be indistinguishable from Dichopteris.

Dichopteris, though conveniently retained as a distinct genus, agrees so closely, in the broad and forked rachis and in the fleshy pinnules, with Thinnfeldia that it would seem reasonable to regard the two genera as members of the same group.

Several authors have drawn attention to the striking resemblance in form and venation between the fronds of the Palaeozoic genus Odontopteris and those of Ctenopteris and Thinnfeldia. In Odontopteris, as in Neuropteris, another Palaeozoic genus, the rachis occasionally bifurcates as in Thinnfeldia and Dichopteris, and the ultimate segments of some species of Odontopteris (fig. 366, A) are practically identical with those of Thinnfeldia and Ptilozamites.

Odontopteris is probably a Pteridosperm. There is no adequate reason for supposing that this group of plants which played a prominent part in the Permo-Carboniferous floras was no longer in existence during the Mesozoic era.

Odontopteris.

Brongniart[1476] instituted the genus Odontopteris for compound fronds from the Coal-Measures characterised by pinnules attached by the whole breadth of the base and traversed by numerous forked veins. Odontopteris is very rare in British Carboniferous rocks and “appears to be restricted to the Middle and Upper Coal-Measures[1477].”

Fronds large, bipinnate or tripinnate, the main rachis, which may be dichotomously branched, bears long linear pinnae with broadly linear or deltoid pinnules, acute or blunt, attached by the whole of the base; the lower margin of the lamina, which is usually entire and rarely lobed (e.g. Odontopteris osmundaeformis)[1478], is often decurrent on the axis of the pinna. The basal pinnule of each pinna is frequently attached by a contracted base, and the lamina may differ in form from that of the normal segments. Pinnules often occur on the main rachis, and in some species the petiole bears modified pinnules which are larger than the ultimate segments of the pinnae and in some cases Cyclopteroid in shape. The pinnules are traversed by numerous dichotomously branched veins; if a midrib is present it dies out in the basal part of the lamina. In some species (genus Mixoneura) pinnules of the Neuropteroid type, characterised by a well-defined midrib, occur in association with typical Odontopteroid pinnules on the same pinna.

The species represented in fig. 364, C, D, from the Middle Coal-Measures of Barnsley, Yorkshire, illustrates the form and venation of the Odontopteris type of pinnule. Another species, O. Reichiana Gutb.[1479], is also recorded by Kidston from the Lower Coal-Measures of Lancashire. Some unusually good specimens of the type-species of the genus Odontopteris minor, Brongn., have been figured by Zeiller[1480] from the Coal-Measures of Blanzy (fig. 365) which show the dichotomy of the main axis and the occurrence of Aphlebiae on the petiole. The late Dr Weiss[1481] divided Odontopteris into two sections, Xenopteris and Mixoneura, the pinnules of the former having the form shown in fig. 364, D; while in species of the latter sub-genus some of the pinnules are identical in form and venation with those of Neuropteris except that they are attached by the whole breadth of the base. Zeiller[1482] employs Mixoneura as a generic designation. In an American species O. Wortheni Lesq.[1483] the pinnules bear numerous hairs like those on some species of Neuropteris (fig. 373, p. 570). The large size of the fronds of Odontopteris suggested to Weiss[1484] that they were borne on the stems of tree-ferns, but Grand’Eury’s[1485] examination of specimens in the Coal-beds of central France led him to picture the plant as bearing a tuft of leaves on a short subterranean stem. Renault and Zeiller[1486], on the other hand, obtained evidence in the Commentry Coal-field of fronds borne on elongated stems which grew on the ground and were supported by stronger plants. Stur[1487] was the first to suggest that Odontopteris should be excluded from the ferns. Grand’Eury’s[1488] supposed fertile pinnules of Odontopteris do not afford any satisfactory evidence of the sporangial nature of the small swellings which he figures at the ends of the veins. This author pointed out several years ago that the petioles of some species of Odontopteris possess the anatomical features of Myeloxylon, a type of leaf-stalk which is now known to belong to Pteridosperms. In a recent paper Grand’Eury[1489] records the association of Odontopteris fronds with small seeds (Odontopterocarpus), a discovery which leaves little or no doubt as to the Pteridospermic nature of the genus. The fronds of Odontopteris are very similar in habit to those of Neuropteris, another Pteridospermic genus.

The similarity between some Odontopteris and Thinnfeldia leaves, to which attention has already been called, is well illustrated by O. genuina Grand’Eury[1490], a pinnule of which is represented in fig. 366, A. Odontopteris is a fairly widespread genus in Upper Carboniferous and Lower Permian rocks, and is recorded also from Triassic strata: it is represented in the Coal-fields of North America and in several parts of Europe[1491].

In some fronds included in Odontopteris the pinnae are characterised by a broad irregularly lobed lamina which also forms a winged border to the rachis. Examples of this form are afforded by Odontopteris Browni Sew.[1492] from the Burghersdorp Series (Triassic?) of Cape Colony, and O. Fischeri described by Brongniart[1493] from the Permian of Russia. The Russian species would perhaps be more appropriately placed in the genus Callipteris, as Weiss[1494] suggests; the absence of venation in O. Browni renders generic identification unsatisfactory.

Callipteris.

Brongniart[1495] instituted this genus for certain species of supposed ferns previously referred to the genera Pecopteris, Alethopteris, and Neuropteris. Callipteris is a characteristic Permian plant which is almost certainly a Pteridosperm. Zeiller has pointed out that such descriptions of fertile specimens as have been written are unsatisfactory. A few years ago, however, Grand’Eury[1496] recorded the occurrence of seeds in association with Callipteris fronds in the Autun district, and in some cases they were found attached to the pinnae and rachis. The seeds are ovoid or spherical (5–10 mm. broad) and smaller than those of Neuropteris. The drawings of fertile segments published by Weiss[1497] afford no indication of reproductive organs. Potonié[1498] figures some pinnules of Callipteris conferta in which the thick lamina is covered with sinuous grooves probably made by some insect larvae: as he suggests, similar markings may have been mistaken for the remains of sori. The occurrence of Callipteris fronds recorded by Weber and Sterzel[1499] in association with Medullosa stems in the Lower Permian of Saxony is in accordance with Grand’Eury’s conclusion.

Fronds reaching 1 metre in length, bipinnate or tripinnate, main rachis frequently exhibiting a combination of dichotomous and pinnate branching. Pinnae linear, usually crowded, decurrent on the rachis; the pinnules on the lower side of the pinnae are continued on to the rachis. Pinnules of the Pecopteroid type, entire or slightly lobed, or of the Sphenopteroid type and more or less deeply dissected (fig. 366 C, D), the lamina of adjacent pinnules concrescent; on the lower pinnae the lamina may be continuous as in an Alethopteris pinnule. A midrib may extend almost to the bluntly rounded apex of the ultimate segments, giving off oblique, simple, or forked veins, the lowest of which arise directly from the rachis; in the Sphenopteroid forms the lateral veins are given off at a more acute angle.

A striking feature of the genus is the occurrence of pinnules on the main rachis, as in Odontopteris. Zeiller has wisely extended the application of Callipteris to fronds possessing this character irrespective of the entire or lobed form of the ultimate segments. He found among the numerous examples of the genus obtained from Autun[1500] and Lodève[1501] transitional forms connecting such species as C. conferta (fig. 367) and C. Pellati Zeill. (fig. 366, C) in which the Pecopteroid pinnules are slightly lobed, with C. lyratifolia (Goepp.) (fig. 366, D), C. flabellifera[1502] (Weiss), and C. Bergeroni Zeill. characterised by deeply lobed Sphenopteroid segments.

Callipteris conferta (Sternberg)[1503]. Fig. 367.

This polymorphic species (fig. 367) is one of the most characteristic Permian plants. The oval-linear pinnules, attached by the whole base, occur on both pinnae and rachis; this feature, the thick texture of the lamina, and the linear, obliquely set, pinnae render the fronds easily recognisable. The fronds bore seeds.

In a recent account of some Permian plants from Germany, Schuster[1504] refers a portion of a frond to Callipteris conferta (Sternberg) var. polymorpha Sterzel, which is characterised by unusually large and polymorphic pinnules. In size and shape the pinnules recall those of Neuropteridium validum Feist.

Callipteridium.

The name Callipteridium, created by Weiss[1505] as a sub-genus of Odontopteris, is applied by Zeiller and other authors to a few Upper Carboniferous and Permian species characterised by the occurrence of simply pinnate pinnae on the main rachis between the bipinnate primary pinnae. Single pinnules are borne directly on the rachis of the primary pinnae between the pinnate branches. The form and venation of a typical pinnule are shown in fig. 366, B. Callipteridium pteridium, originally recorded by Schlotheim as Filicites pteridius[1506], has been fully described by Renault and Zeiller from unusually large specimens found in the Commentry Coal-field[1507]. This species illustrates the peculiar morphological features of the genus. The main rachis of the tripinnate fronds, several metres long, shows a combination of dichotomous and pinnate branching; from the zigzag and forked axis are given off bipinnate pinnae and, between these, shorter pinnate branches. The pinnules closely resemble those of Callipteris conferta but reach a greater length; the pinnules borne on the rachises of the lateral branches differ from the others in their broader base and more triangular lamina.

No fertile specimens have been found. It is probable that Callipteridium was not a true fern, and that White[1508] is correct in including it among the Pteridosperms.

Archaeopteris.

In 1852 Forbes[1509] published a brief description of some supposed fern fronds, found by the Geological Surveyors of Ireland in Upper Devonian rocks of Kilkenny, under the name Cyclopteris hibernica. The Irish specimens were more fully described by Baily[1510] in 1858. Fronds of the same type were referred by other authors to Cyclopteris, Adiantites or Noeggerathia, until Schimper[1511] proposed the generic name Palaeopteris on the ground that the fronds described by Forbes and Baily are distinguished by the nature of their fertile pinnae from the sterile leaves included in Brongniart’s provisional genus Cyclopteris. The earlier use of Palaeopteris by Geinitz for an entirely different plant led Dawson[1512] to institute the genus Archaeopteris. The genus Archaeopteris may be defined as follows:

Fronds bipinnate, reaching a considerable length (90 cm.); the stout rachis bears long linear pinnae; sterile pinnules obovate or cuneate with an entire, lobed, fimbriate, or laciniate lamina traversed by divergent dichotomously branched veins. The fertile pinnae usually occur on the lower part of the rachis; pinnules with a much reduced lamina bear numerous fusiform or oval exannulate sporangia (fig. 369, A, E, H), sessile or shortly stalked, singly, or in groups of two or three. The base of the petiole is characterised by a pair of partially adnate stipules (fig. 369, C, D), and single pinnules or scales occur in some species on the rachis between the pinnae and on the petiole.

Archaeopteris hibernica (Forbes). Figs. 368, 369, A–C.

The specimen from Kilkenny represented in fig. 368 has a length of over 80 cm. The upper pinnae bear numerous imbricate obovate pinnules (fig. 369, A, B) with an entire or very slightly fimbriate margin, while on the shorter lower pinnae the ultimate segments are reduced to a slender axis bearing numerous fusiform sporangia, 2–3 mm. in length. Kidston[1513] has pointed out that sporangia occasionally occur on the edge of ordinary pinnules, and he first recognised the stipular nature of the scale-like appendages which Baily noticed on the swollen petiole base (5 cm. broad) of the Irish species (fig. 369, C). Restorations of Archaeopteris hibernica have been figured by Baily[1514] and by Carruthers[1515], but the description of the fertile pinnae by the latter author requires modification in the light of Kidston’s description of the Dublin specimens.

Archaeopteris is recorded from Upper Devonian rocks of the South of Ireland, Belgium, Germany, Southern Russia, Bear Island, and Ellesmere Land in the Arctic regions, Canada, Pennsylvania, and elsewhere. Many of the specimens described under different names bear a close resemblance, which in some cases probably amounts to specific identity, to A. hibernica. A. Jacksoni originally described by Dawson[1516] and more recently by Smith and White[1517] from Devonian rocks of Maine, the Canadian type A. gaspiensis Daws., and some species figured by Lesquereux[1518] from Pennsylvania, are examples of forms which present a striking similarity in habit to the Irish species. The Belgian Devonian fossils named by Crépin[1519] Palaeopteris hibernica var. minor are regarded by him as probably identical with Goeppert’s species Cyclopteris Roemeriana from the neighbourhood of Aachen. Heer recorded Archaeopteris Roemeriana from Upper Devonian beds in Bear Island, and Nathorst[1520], who has published a more complete account of the Arctic forms, draws attention to the resemblance of some of them to A. hibernica. A species described by Schmalhausen[1521] from the Upper Devonian of Southern Russia as A. archetypus (fig. 369, D) appears to differ from A. hibernica in the slightly less reduced lamina of the fertile segments. This species has been more adequately illustrated by Nathorst[1522] from material collected in Ellesmere Land: he is unable to confirm Schmalhausen’s statement that the pinnae are spirally disposed.

The species A. fimbriata (fig. 369, G) described by Nathorst from Bear Island is characterised by the more deeply dissected lamina of the sterile pinnules. In A. fissilis Schmal. from Russia and Ellesmere Land the lamina (fig. 369, E, F) is cut up into filiform segments: a fertile pinnule of this species is represented in fig. 369, E.

Some sterile impressions figured by Krasser[1523] from Palaeozoic strata (Lower Carboniferous or Upper Devonian?) in the province of Nanshan in China as Noeggerathia acuminifissa are considered by Zeiller[1524] to be portions of an Archaeopteris or Rhacopteris frond. The resemblance to the former genus is however by no means close enough to warrant a reference to Archaeopteris. The sterile specimens described by Stur[1525] from the Culm of Altendorf as species of Archaeopteris are probably not generically identical with the Irish and Arctic species. The dichotomous branching of the rachis in A. Tschermaki and A. Dawsoni is a feature unknown in Archaeopteris. In the absence of fertile pinnae the separation of Archaeopteris from Rhacopteris is by no means easy.

Archaeopteris was regarded by Carruthers as a fern closely allied to recent species of Hymenophyllaceae, but this conclusion was based upon an interpretation of the fertile segments which Kidston[1526] has shown to be incorrect. The latter author regarded the presence of stipules and the structure of the exannulate sporangia as evidence of a Marattiaceous alliance. In a later reference to Archaeopteris, Kidston expresses the opinion that the genus is not a true fern but a member of the Cycadofilices or Pteridosperms, a view shared by Grand’Eury[1527] and doubtless by many other palaeobotanists. The sporangia of Archaeopteris appear to be of the same type as those of Dactylotheca (fig. 290, E, p. 399). Schmalhausen gave expression to his disagreement with Nathorst and other authors who referred Archaeopteris to the Marattiaceae by proposing the distinctive group-name Archaeopterideae.

There can be little doubt that the reproductive organs of Archaeopteris so far discovered are microsporangia, and that the plant bore seeds. The sporangia are larger than those of any known fern and, as Kidston points out, they are similar to those of Crossotheca which he has shown to be microsporangia of the Pteridosperm Lyginodendron. The presence of stipules in Archaeopteris hibernica, A. fimbriata, A. archetypus (fig. 369, D) and probably throughout the genus does not materially affect the question of taxonomic position. Stipules are a characteristic feature of Marattiaceae and, in a reduced form, of Osmundaceae, but similar appendages are borne at the base of the petiole of the Cycad Ceratozamia. The occurrence of Aphlebiae on the rachis of Archaeopteris is a feature shared by the fronds of Neuropteris and other Pteridosperms.

Neuropteris.

The fronds for which Brongniart[1528] created this genus, though suspected by Stur in 1883 as wrongly classed among the ferns, have only recently been shown to be the leaves of Pteridosperms. As yet only one case is recorded in which

Neuropteris pinnae occur in organic connexion with seeds[1529], but it is almost certain that the genus as a whole must be placed in this generalised group. Renault[1530] pointed out that the petioles of Neuropteris fronds from Autun had the anatomical features of Myeloxylon (petiole of Medullosa). Since Kidston’s important discovery of seed-bearing pinnae of N. heterophylla, Grand’Eury[1531] has recorded the association of Neuropteris fronds with seeds in French Coal-fields. By some of the older authors Neuropteris was compared with Osmunda because of a similarity in venation. In the frequent dichotomy of the frond and in the occurrence of pinnules on the rachis, Neuropteris closely resembles Odontopteris[1532]: there can be little doubt as to the close relationship of the Pteridosperms possessing these two types of foliage. Neuropteris may be defined as follows:

Fronds reaching a considerable size, probably in soma cases a length of 10 metres[1533]; bi- or tri-pinnate; the rachis may be dichotomously branched (figs. 354, D; 370); both rachis and petiole bear single pinnules, those on the latter frequently differ from the normal leaflets in their larger Cyclopteroid laminae (fig. 370). Pinnules entire, rarely slightly lobed, broadly linear, attached by a small portion of the base, which is usually more or less cordate. In N. Grangeri Brongn. the pinnules are attached by a short pedicel[1534]. The midrib always dies out before reaching the blunt or pointed apex of the lamina and gives off at an acute angle numerous secondary veins characterised by their arched course and repeated forking.

Potonié describes the secondary veins of the pinnules of Neuropteris pseudogigantea[1535] as occasionally anastomosing, a feature which may be regarded as a step towards the reticulate venation of the closely allied genus Linopteris.

Renault[1536] described some petrified pinnules of Neuropteris in which the mesophyll shows a differentiation into upper palisade tissue and lacunar tissue below; the lower epidermis is infolded at intervals where grooves (probably stomatal) occur like those on the leaves of an Oleander (Nerium oleander).

The rachises of Neuropteris fronds are described by Grand’Eury under the generic name Aulacopteris[1537].

Neuropteris heterophylla, Brongniart[1538]. Figs. 354, E; 371.

This species is characteristic of the Lower Coal-Measures of Britain; it occurs also in the Middle Coal-Measures and is a common type in Upper Carboniferous rocks in various parts of the world. The fronds are large and tripinnate, the rachis is often dichotomously branched and Cyclopteroid pinnules may occur on the petiole. The pinnules, 5–20 mm. in length and 3–8 mm. broad, have a rounded apex (fig. 354, E, p. 535).

As shown in fig. 371 which represents a primary pinna, the small pinnules on the lower branches are gradually replaced in the upper portion of the specimen by falcate segments.

Neuropteris macrophylla, Brongniart[1539]. Figs. 354, D, D′; 372.

The rachis of the large fronds of this species illustrates the dichotomous habit of many Neuropteris fronds, also the occurrence on the petiole of large Cyclopteroid pinnules (cf. fig. 370). The small piece of a pinna reproduced in fig. 372 shows the slender attachment of the segments, the blunt apex, and the Neuropteroid venation. Single pinnules of this species may be distinguished from those of N. Scheuchzeri by the blunter apex, the absence of the pair of small Cyclopteroid pinnules on the same branch and by the absence of hairs. N. macrophylla is characteristic of the Upper Coal-Measures of Britain.