The taxonomic status of the genus Rubidgea

The genus Rubidgea Tate of the fossil family Glossopteridaceae was reduced to a synonym of Glossopteris by Seward (1907). Seward’s conclusion is now confirmed by a study of a wide range of imprints from a quarry near Hammanskraal, South Africa. The upper and lower surface imprints of a single leaf found on a split fragment of carbonaceous shale provides the main evidence presented. The finely striated upper surface imprint of the leaf could be identified with Rubidgea, whereas the lower surface imprint represents the typical strong venation of a Glossopteris. The type species of Rubidgea is transferred to Glossopteris as G. mackayi (Tate) Kovacs comb. nov. The characteristics of upper and lower surface imprints of a number of Glossopteris species are discussed.


IN TRO D U C TIO N
The family Glossopteridaceae appeared approxi mately in the Upper Carboniferous and disappeared more or less synchronously in the Lower Triassic.Though the species changed in time due to evolution and adaptation to changing climatic conditions, they are nevertheless generally recognizable as glossopterids.All evidence points to the family Glossopteri daceae of the Order Pteridospermales being a monophyletic and natural group of plants.As in most fossil groups, the classification presents considerable difficulties.Recently doubts have been expressed about the feasibility of a classification of the Glossop teridaceae based on leaf impressions when no cuticles or attached fructifications are available.In this and other papers, the author defends the point of view that taxa, at least at the specific level, may be distin guished by normal taxonomic methods, provided that sufficient good material, which is synchronous, is available and is studied in detail.The basic venation patterns of leaves are typical for all members of the group, but venation characters display definite and useful differences.The fact is that all fossil plants represent once living plants and this realization is necessary for real progress to be made.A knowledge of living extant plants should form an essential part of the equipment of the palaeobotanist.
In the investigation here reported the author has attempted to apply these views to the study of a rich deposit of glossopterid fossils.Special attention is given to the taxonomic status of the genus Rubidgea Tate.The material for this study was collected on the farm of Mr J J Brits, 30 km north of Pretoria near Hammanskraal.The carbonaceous shales, in which the leaf impressions are preserved, are of Ecca age.All the specimens studied are deposited in the offices of the Geological Survey, Pretoria, South Africa.

TH E ID EN TITY O F THE G EN U S R U B ID G E A
The genus Rubidgea was described by Tate in Q. J1 Geol.Soc.Lond. in 1867.The description is as follows: " Rubidgea Mackayi, gen.et spec.nov.
Frond oblong, obovate, rounded and obtuse at the apex; secondary veins very slender, very much crowded, dichotomous, oblique.In the preamble to this description Tate, however, says the following: " W ith the above-mentioned specimens from Bloemkop are some of an apparently, at first sight, second species of Glossopteris', these do not exhibit fructification.Dr Rubidge however, has communicated a drawing (by M r McKay) of a specimen of this species obtained by Mr McKay near East London: and I find that it presents characters generically distinct from those of Glossopteris-, for the position of the fructification is indicated by a few large elevated rings, arising from many veins, and somewhat regularly arranged in a row coincident with the margin, and not by numerous spots, small in size, supported by one vein, distributed over much of the surface o f the frond" .
Whereas the description of the new genus and species is clearly based on the leaf, Tate's comments in the second paragraph quoted here, show that it is only because of the so-called fructifications that he considered creating a new genus.It so happens that the two specimens cited by Tate have, in spite of several searches, not been traced.A comparison of the speci mens with the McKay drawing has thus not been possible.It seems clear that Seward (1907), who regards Rubidgea as a synonym of Glossopteris, also did not see Tate's syntypes.Since the specimens are missing, there seems to be no alternative but to accept the drawing by McKay published by Tate as Plate 5, Fig. 8, as the lectotype of Rubidgea Tate.It has already been pointed out by Seward (1907) that the so-called fructifications depicted by McKay, are artefacts and I agree with this view.

G L O S S O P T E R IS
When studying the fossil flora of the Hammanskraal quarry, many leaf impressions were found which show the characteristic venation described and figured by Tate for his genus Rubidgea.A venation, such as figured by Tate, is difficult to interpret morphologigically.The very fine veins arise from the median line of the leaf, but are apparently not joined into a true midvein.A closer examination of these "Rubidgea" leaves from Hammanskraal shows that some of them exhibit areas of stronger venation with anastomoses typical of the genus Glossopteris (Fig. 3).It seems fairly certain that in these leaves part of the upper surface has decayed or peeled off revealing the nerva tion of the lower surface.In other instances both kinds of " venation" are seen on the leaf as if super imposed.On these imprints dense and fine lines are visible between the typical Glossopteris veins (Fig. 4).Amongst the many slabs of slate split, one showed an almost entire leaf with clear upper and lower surface imprints.These imprints are dissimilar, one being typical of Glossopteris, the other of Rubidgea (Figs 1  and 2).It is obvious also that the impression of the lower surface, which shows Glossopteris characters, represents the imprint of the venation.The very slender lines of the Rubidgea imprint, which do not match the nervation of the lower surface, have to be interpreted as fine grooves found on the upper surface of some Glossopteris species.The leaves of many extant species of plants would show feature less upper surface imprints, because their veins are not raised.There are many examples of these three states, viz.parts of Glossopteris leaves which show Rubidgea characters or vice versa and both kinds of "venation" next to one another on the same leaf.
These observations indicate that the genus Rubidgea and some species of Glossopteris were based on impres sions of upper and lower surfaces of leaves respec tively, Rubidgea representing impressions of the upper and Glossopteris of the lower surface.There is real proof, therefore, that the genus Rubidgea must be regarded as a synonym of Glossopteris as concluded, on rather slender evidence, by Seward.The study of several glossopterid species of the Hammanskraal fossil flora showed that in all species investigated, except for G. indica, the upper and lower surfaces differed in a characteristic and specific manner.In G. indica both surfaces were more or less similar.Of the species examined, all leaves with "Rubidgea" features proved to belong to the same species (Kovacs, 1976).It is possible, however, that other species may also exhibit a "Rubidgea" type of upper surface.
As I mentioned before, in some instances the imprint of what appears to be both upper and lower surfaces, can be seen on the same leaf impression (Fig. 4).In such cases there are thinner lines between the strong, anastomosing veins.This feature merits further discussion.Pant (1958)  On the evidence of the leaves from Hammanskraal, I interpret such fibres as being the markings of the upper (or " Rubidgea") surface showing through on the lower surface.This interpretation is to some extent confirmed by Pant (1958).He mentions that G. fibrosa and G. hispida, which both show the so-called fibres in the vein meshes, have rather thin laminae.Prior to these studies, differences between upper and lower surfaces of a Glossopteris leaf were already reported for G. browniana by Dana (1849, p. 717).
In the light of the foregoing, it is of some interest to establish how it has come about that the upper and lower surface leaf impressions were described as separate genera and why, one hundred years later these genera were still being recognized and expanded.
The description and identity of Rubidgea have already been discussed.It is clear, as stated before, that Tate believing these plants to be ferns, separated the new genus from Glossopteris mainly because of the presence of so-called " fructifications" .In his diagnosis of the species he does not comment on the presence of a midrib, but implies this by stating: " secondary" veins very slender, very much crowded.The " secon dary" veins arise from the midline of the blade and not from the base, which logically indicates a mid vein or a structure resembling a midvein.
Feistmantel's (1881, p. 91) statement when com paring Palaeovittaria with Rubidgea, viz." this latter (i.e.Rubidgea) showing no indication of a midrib in the lower part" , is therefore incorrect.On the drawing by McKay, published by Tate, the lower part of the leaf is missing but, as already indicated, shows the situation where the " secondary" veins arise from the median part of the leaf.In spite of this, Rubidgea is still usually characterized by the absence of a midrib as was done recently by Maheshwari (1965, p. 37).Maithy (1965) accepted Rubidgea as a generic entity, but amended it to accommodate several species found in the Karharbari Beds of the Ciridih Coalfield in India.Maithy (p.42) also starts the description of the venation with the statement "devoid of a midrib" .The characterization of taxa such as Gangamopteris and Rubidgea by the "absence of a midrib" , and the failure of authors to refer back to the type specimens, contributed to the confusion which arose.The morphological terms applied to extant plants are not always appropriate to extinct groups: " midrib" does not mean the same in Glossopteridaceae as it does, for example, in Angiosperms.Glossopteris leaves have a bundle of veins down the middle, but no clear single midrib.When Maithy writes of " numerous veins arising from the median longitudinal position of the frond, occasionally simulating a false midrib" , he accurately describes Tate's drawing.Later in his paper, Maithy refers to median veins, however, and omits the important character mentioned in Tate's description, that the secondary veins are very slender and closely crowded together.Ignoring this character, he described two new species, Rubidgea obovata and R. lanceolatus, both of which have rather strong veins.On the photo graphs of the leaf impressions of the two species, three of the four leaves have parallel veins fanning out from the narrowed base, yet Mainthy himself chracterizes Rubidgea by " veins arising from the median longitudinal position" .
At the end of his paper, Maithy indicates that the leaf of Lanceolatus palaeovittarius Plumstead "appears to be a Rubidgea because no midrib is evident, and the median region of the leaf seems to be occupied by the subparallel veins".In Feistmantel's time the more crowded subparallel median veins of glossopterid leaves were described as a midrib, and it is possible that this condition is found in Palaeovittaria Feistm.(1876, p. 368).

Maithy also mentions Gangamopteris obovata
Carruthers as a possible Rubidgea, because "secon dary veins emerge from median veins" .Carruthers's figure (1869) depicts a whole leaf and the equally strong subparallel veins radiate from the base.Dr C. R. Hill of the British Museum of Natural History checked the holotype (v.229,Geol. Surv.,British Museum Nat.Hist.) at my request.He found the drawing remarkably accurate, except in showing the anastomoses between the veins.
The confusion that exists in the taxonomy of Glossopteris and related genera can be attributed to poor descriptions and circumscription of new taxa; to the difficulties inherent in the interpretation of what represents a " midrib" of the leaf of the Glossopteridaceae; and to the fact that the differences in the upper and lower surfaces of glossopterid leaves are usually ignored.
If progress is to be made with the taxonomy of the Glossopteridaceae, these pitfalls will in future have to be avoided.

A NEW C OM BINA TION IN G L O S S O P T E R IS
In order to comply with the Rules of Botanical Nomenclature the formal transfer of the type species of Rubidgea to Glossopteris is hereby made.

A
CK N O W LED G EM EN TS The author is grateful to Dr S. H. Haughton for critical reading of the manuscript; to Dr B. de Winter thanks are due for useful comments and interesting discussions; to Dr S. Endrody-Younga for help with the taxonomy; to Dr A. W. Keyser for his support.The photographs were taken by Dr R. P. Stapleton, Dr H. C. Klinger and Mrs Ruth Fregona.
Localities.Bloemkop, near the Sunday's River, G raaff Reinet (D r Rubidge); East London, at the mouth o f the Buffalo River (Mr. M cKay)." There is no indication o f anastomosis o f * Geological Survey, Private Bag 112, Pretoria.the veins.

brow- niana var. indica {G. indica), G. angustifolia, G. brou- niana var. australasica, (G. browniana) and G. com munis
described two new species, namely Glossopteris fibrosa and G. hispida, based mainly on cuticular examinations.He mentions that G. fib ro sa : " looks much like many of the figures given by various authors under the names G. . . .All are clearly different because they lack fibres in vein meshes" .
C a r r u t h e r s , Q. J l Geol.Soc.Lond.23: 140-149.