Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
- 55 - PALEOZOIC PLANTS AND THEIR ENVTBOMMENI/LL RELATIONS By Chester A. Arnold (University of Michigan) I, Introduction The fundamentals of the paleoecology of Paleozoic plants are essen- tially the sane as those of any other era as far as they concern relation to environment. The reasons for recognizing a separate problem for the Paleozoic are that special consideration should be given to certain limit- ing factors, and that some things have to be dealt with from a standpoint based on an intimate knowledge of Paleozoic plants. Any consideration of ecology of past ages must depend upon present day ecology for guidance. For example, nothing can be postulated about the temperature relations of extinct plants without first knowing something of the thermal requirements of living plants. In a like manner we decide whether a given fossil form was adapted to a terrestrial or aquatic exist- en«e by comparison with familiar living fozms showing the same or similar adaptations. Comparison with living forms is, therefore, the starting point for paleoecology. The problan of environmental adaptations of Paleozoic plants differs materially from that of Mesozoic end Tertiary plants in that all Paleozoic types are extinct and their nearest living relatives are so remote that ecological comparisons are difficult to make. Predominant among the Paleo- zoic plant types were. the pteridosporms, or Cycadpfilices, which before the discovery of their seeds, were assumed to be ferns. In the living flora ferns are most abundant in the tropics, and consequently a tropical steamy atmosphere was once postulated as the type of environment in which existed the forests of the Carboniferous coal swamps. But these pteridosperms have no close modern relatives and we cannot assume that their ecological re- quirements were similar to those of ferns. Ferns did exist in the Paleozoic in considerable diversity of form and habit but they were not closely re- lated to the living ones, nor did they exist in sufficient numbers to establish the existence of tropical surroundings. The culmination of a prominent Paleozoic gymnosperm line was realized in the Cordaitales. While these may resemble modern conifers to a slight extent their relationship to them is remote, and the predominance of cordaitean forms in a flora is not necessarily an indication of upland conditions without the support of evidence furnished by other forms. Some of- the cordaitean forms have ecological significance in that they show structures suggestive of adap- tations to certain conditions. The Paleozoic Lycopbdiales, which culminated in such genera as Lepido- dendron and Sigillaria, have as modern representatives a few straggling and slow growing remnants as Lycopodium and Selaginella. These latter are widely distributed and exist in a variety of habitats, not. being confined to situa- tions such as must have prevailed in the coal swamps. The same thing is true
- 56 - of the Paleozoic calamites andi their modem remnant, the genus Equisetum. This brief resume of the predominating plant forms of the Paleozoic is sufficient to show conclusively that we can tell little or nothing about the ecology of that era by knowing what plants existed then. They are too remote from modern plants to permit ecological comparison. Also, the ex- tent of relationship between these early and later plants is often poorly understood. Furthermore, members of the major groups as the Cycadofilices and the Cordaitales were so widely diversified and lived under such a va- riety of habitats that generalizations concerning their ecology are diffi- cult to formulate. Another item of major significance in securing ecological data from Paleozoic plants is that the entire plant can seldom be reconstructed. Be- cause of the size and shape of the plant body, the mode of attachment of the organs, and the frequent and periodical shedding of leaves, bark, cones, seeds, etc., most paleobotanical information is secured from detached parts. These parts which make up tiie great bulk of Paleozoic plant material furnish valuable data on phylogeny and classification, but us'jally have limited ecological significance. One of the lines of approach to the subject of Paleozoic plant ecol- ogy is through the study of structurally preserved remains which involves some knowledge of the relations between structure and environment. In liv- ing plants there is frequently a distinct correlation between foliar struc- tures, and moisture and light requirements. The difficulty, however, in applying the same inferences to Paleozoic plants is that very pronounced structural adaptations are not common anong plants growing under normal sur- roundings. It is only when plants are subject to extreme conditions, as aridity, salinity, etc., that they develop specialized structures which are conspicuous enough to be recognized in Paleozoic plants and a similar en- vironment postulated for than. Moreover, certain specialized structures were developed in some Paleozoic plants that do not exist in living plants. This is shown by the extraordinary development of surface periderm in sane species of Lepidodendron or the internal periderm in stans of Medullosa. In Lepidodendron this external periderm served for support of the trunk since the vascular cylinder was small, but the ecological significance of this has not been fully explained. Another difficulty is the different structural adaptations which different species show. If all species re- acted the same toward their environment it would be much easier to determine anatomical adaptation in plants, extinct and living. II. Paleozoic plants in relation to the sediments Fossil plants are usually preserved in water deposited sediments and hence removed from the environment in which they grew. They might have been removed by streams, ocean currents, wind, or by any number of agencies. For this reason a fossiliferous deposit often contains plant de'bris from a variety of habitats, dry uplands as well as the more moist lowlands, and the remains occur together in inseparable confusion. It is usually to be ex- pected, however, that the lowland species will predominate in deposits formed
- 57 - under such conditions as probably existed in the Carboniferous coal swamps, and this fact has doubtless influenced attempts to determine climatic con- ditions of that period. Since a deposit of fossil plants is likely to contain a mixture of types the problem 'of elucidating the environment under which the plants grew is quite different from that relating to aquatic or- ganisms which were fossilized in their natural surroundings. Land plants and land vertebrates present many paleoecological problems in common. When the site of deposition is far removed from the locality in which the plants grew, little can be determined concerning the natural en- vironment other than what can be gathered from the structure and morphology of the plants themselves. In a few important instances, however, plant remains have been found in place and still rooted in the original soil where they grew. Such fossil deposits often reveal much concerning the surroundings, not only at the moment of deposition of the enclosing sedi- ments but also those under 7*1 ich the plants existed as living organisms. An example of such is the Rhynie chert deposit near Aberdeen, Scotland (11).* Here many sttans are still standing erect and attached to horizontal under- ground rhizomes. The chert has every appearance of having been a peat bed which was saturated with water highly charged with silica of volcanic or- igin. The small plants, some of which were leafless, suggest vegetation growing in the vicinity of volcanic fumaroles and hot springs. Other ex- amples of plants buried in place are the Lepidodendron trunks in the vol- canic ash beds on the island of Arran (Seward, 15, Fig. 51) and those which are so admirably exposed in Victoria Park in the city of Glasgow. Here are suggestions that during Lower Carboniferous times conditions prevailed which were quite similar to those in the Great Basin and on the Columbia Plateau regions during Tertiary times. The well known Gilboa trees, of Hamilton age, in:. eastern New York afford an excellent example of a forest submerged by an encroaching sea. It is probable that here the forest covered a swampy lowland as is indicated by the layer of black shale in â¢stiich the upright trunks are still rooted (8). Instances such as those mentioned in which plants are preserved more or less in their natural environment are comparatively rare; usually the plant fossils are fragmentary remains of detached parts which had been transported for some distance. Remains of land plants may occur with marine invertebrates, which means no more than that the plant remains were derived from a nearby land source. The difference between two fossil floras may be as much due to geo- graphical, i.e., environmental, factors as to time differences. During the process of transportation'of plant remains from the place where they grew to the place of final deposition a kind of "sorting out" process may occur, and not all of the components of the flora of a given lo- cality will always remain in recognizable form in the ultimate deposit. Such factors as distance from shore, direction of currents, presence of petrifying minerals, rate of deposition of sediments, and a-number of other things usually exert a selective influence by setting up an environment of deposition suitable for the preservation of some plants but not for others of the seme plant association. * See bibliography, page 64.
- 58 - An exanple of such a sorting out of species is shown by the occur- rence of Archaeopteris and Callixylon in the Upper Devonian beds of New fork and Pennsylvania. Archaeopteris is commonly cited as an index fossil of the Upper Devonian, and in eastern North America it is supposed to characterize the uppermost teds commonly designated as Catskill and Chemung. Since it has recently teen shown, however, that the so-called "Catskill" represents an environment of deposition which extended from Hamilton to the end of Devonian time instead of representing a distinct time interval at the top, it follows that the occurrence of Archaeopteris in these de- posits, or any other deposits for that matter, proves very little concern- ing their exact place in the Upper Devonian. Archaeopteris occurs in por- tions of the so-called "Catskill" which are now known to belong to the Hamilton, and it may occur at any level above. In the westward marine ex- tensions of these beds, rihich in the east bear Archaeopteris, calcified remains of Callixylon occur (3). That these two genera were contemporan- eous is most certain, but in no single instance have they been found to- gether where the identification of both is beyond question. Archaeopteris, is known only from its leaves and sporangia, and such destructible material is seldom preserved in recognizable form at any great distance from its source. It usually occurs in shallow water deposits of mud and fine sand. The trunks and tranches of Callixylon were preserved only when they fell in dfieeper water which carried them away to places where limey or siliceous muds were accumulating. In shallow waters favorable for the preservation of the delicate impressions of the leaves of Archaeopteris the stranded Callixylon trunks mostly decayed. Other possible instances of this se- lective tendency are known. Because of the structure of certain organs partial decomposition of the plant may have a marked influence on the external appearance of the fossil. Failure to take this into account has been responsible for many superfluous generic names. The cases of Knorria, Aspidaria and Bergeria, being the different stages in decortication of Lepidodendron, or the Syringodendron condition of Sigillaria, are too familiar to require ex- planation. Seward (16), taking this state of affairs into account, re- cently described under a single specific name a considerable assemblage of sigillarian remains from Persia. All the material came from the same horizon and there was no convincing evidence for the presence of other types. In the Antrim, New Albany and Ohio shales 'are some rather spec- tacular jointed fossils resembling slender carbonized stems. These had been variously identified as Calamites, Pseudobornia and Callixylon, but a careful examination showed them to be indistinguishable from water soaked woody plant debris which had become cross-cracked as a result of frequent wetting and drying (5). Without this simple and rather obvious interpre- tation an indefinite number of attempts might be made to explain these curious structures. . . The examples given in the foregoing paragraphs seem to indicate two points: First, that a so-called flora may represent a sequence of similar environmental conditions rather than a definitely fixed time in- terval; and second, that the environment of deposition may so alter an organism that it may present a deceptive appearance.
- 59 - III. The problem of the relation of structural responses to environment It is Impossible to consider the entire realm of paleoecology without delving Into the problem of Paleozoic climate. A full discussion of this subject will not be undertaken but certain features having a direct bearing on fossil plant structures will be treated briefly. Before the study of fossils had developed into a science speculations were made concerning the climates under iftiich the organisms existed, and in recent times much has been written about Paleozoic climates. In 1882 Seward (14) wrote a prize-winning essay entitled: "Fossil Plants as Tests of Climate.' Even here an enormous amount of data is presented. Our knowledge of Paleo- zoic plants has probably doubled since Seward wrote his essay, but with re- spect to the climates that existed during this era most of the recent con- tributions have centered about arguments designed to show the fallacy of previous interpretations. The warm steamy atmosphere highly charged with carbon dioxide as postulated ty earlier writers has given way to the con- ception of more temperate conditions with even an occasional glimpse of a glacier. Most investigators, it seems, have largely given up the notion of world wide tropical conditions during the Paleozoic, but there still exists muah disagreement about the degree of seasonal fluctuations. Considerable assistance has been derived from a study of recent peat bogs which has dis- pelled the ideas that any era had but one set of climatic conditions through- out, or that if there were changes it was one uniform and continual change from the beginning until the end. Peat bogs have shown us that within the brief space of human occupation there have teen marked variations in the climate of North America. When we once adopt that point of view for the Paleozoic some progress may follow. The difficulty attending the problem of deciphering Paleozoic climates is that the objects upon which the most profound imprints were made - the plants - are extinct. Many of the earlier assumptions that the Paleozoic climate was tropical were based upon the supposed prominence of ferns, but the discovery of the pteridosperms induced certain authors to stress the fact that we know nothing of the temperature requirements of the prominent Paleo- zoic genera. It has also been pointed out that luxuriant vegetation is by no means proof of tropical conditions. The reduced and less ornate foliage of some of the plant types that survived the Permian glaciation has been re- ferred to as an example of a response to a climatic change, but here the evi- dence furnished by the flora is largely substantiated by that of the sediments themselves. 'AS with certain living species, of which the bracken fern is a stock example, it is quite possible that certain Paleozoic species were able to tolerate a considerable range of temperature, although specific examples cannot be cited. The presence or absence of growth rings- or "annual rings" in Paleozoic ' stems is often employed as evidence concerning climate. In modern woody plants of the temperate zones the seasonal fluctuations are often strongly re- corded by marked difference in cell size. This.reaction is so strong in sane trees, as Finus ponderosa and .Sequoia gigantea, of the southwestern United States, that they reveal climatic cycles for centuries past (2, 10). Many
- 60 - things may influence ring development, of which temperature and moisture are most important. Fires, defoliation, late or early frosts, direction of wind, shading, production of seed, etc., are factors known to be influ- ential. The woody dicotyledons react in the same way as do conifers in the temperate regions but the resulting rings are often less distinct because of differences in structure of the wood. These differences are differences in degree only. Most north taaperate zone species react to seasonal changes in the same way, but in the tropics some are different. The Araucarian conifers, for example, do not respond as markedly to seasonal fluctuations as do some other species in the same habitat. Seasonal fluctuations of either moisture or temperature have little effect on the development of the rings although indistinct rings seem to form in these trees under most any natural outdoor â¢onditions. On the other hand, the common persimmon, Diospyros virginiana, has been observed to develop absolutely uniform wood in an unchanging cli- mate, but when the same species is subjected to seasonal fluctuations well marked rings will form (1). From such observations as these it "becomes evident that in judging climatic conditions from growth rings it is necessary to know how the par- ticular species in question reacts to seasonal changes. The Sequoiae and the pines of the southwestern states react to a decided degree but it is known that other species in other places do so to a lesser extent. It is quite possible that some species, when subjected to such extreme seasonal changes of tanperature as prevail in Arizona and the High Sierras, gradu- ally become more responsive. But from observations on some tropical spe- cies it is apparent that the presence or absence of rings in a tree is not governed wholly by seasonal fluctuations. The physiological sensitivity of the plant is as responsible for the obvious anatomical differences as the vigor of the environmental change. Growth rings occur in some Paleozoic woods. They appear frequently in the Permian and to a somewhat lesser extent in the Pennsylvanian. They are quite recognizable in several Devonian forms. It seens that a greater proportion of Devonian woods show them than do the Pennsylvanian types. Neither the Pennsylvanian nor the Devonian rings show the same prominence of development as do those of our north tanperate zone coniferous species but they are more comparable to those of certain tropical trees. The rings of the fossil woods are usually several millimeters wide and the summer wood zone seldom exceeds three or four cells. In some cases, the rings ex- tend only part way around the stem. Without making a careful examination of the fossil wood it is sometimes difficult to distinguish a growth ring from a zone of crushed cells, and it is quite probable that such appearances have misled some investigators. Probably the best example of a Devonian plant showing growth rings is Callixylon. A stem of C_. erianum from the Genesee shale and which has a radius of E3 mm. shows five complete rings (3). Similar rings are shown in £. Newberryi from the New Albany shale. Calamopitys eupunctata from the Portage of NPW York also shows definite rings (13). Rings have been observed in the Lower Carboniferous Pitys antique from Scotland (9).
- 61 - In rocks of Pennsylvania!! age Cordaites_ Michiganensis from the Saginaw formation (Pottsville) of Michigan shows no ring development in a wood cylinder 1 em. thick (4). Dadoxylon roraingerianum, probably of the Conemaugh group, from Coshocton, Ohio, shows absolute uniformity of growth in a radial extent of 5 cm. (4). From Kansas some interesting facts are available. A specimen as Cordaites materiarutn (17) from the Des Moines series shows distinct layers which vary from 3.5 am. to 6.5 mm. in width. From the Virgil series Dadoxylon douglasense is without growth rings, although to the unaided eye a transverse section of the wood shows indefinite zones which might be mistaken for such (17). At a higher level, at the base of the Big Blue series, a specimen iden- tified as Cordaites recentium (7) shows rings varying in width from 3 to 8 mm. In the Penno-Carboniferous growth rings are common, as is evident from even a casual examination of the literature. Bre examples just noted of growth rings in Paleozoic woods are rela- tively recent discoveries and are not mentioned in the older literature from which most conclusions pertaining to Paleozoic climates have been drawn. Most of the best known Coal Measures plants as Lepidodendron, Sigillaria, Mesoxylon, Lyginopteris and Medullosa are typically without any indications of growth responses, but other forms such as those just mentioned, Which were nearly contemporaneous with them, show rings in unmistakable fashion. It is significant that the rings usually occur in forms which are believed to repre- sent upland types, of which Callixylon, Pitys, and various specimens assigned to Cordaites and Dadoxylon are examples. Furthermore, it is believed that the Cordaitales were represented during the Carboniferous by both swamp inhabiting and upland species. The apparent occurrence of growth rings in some Paleozoic woods and their absence in others might also have some correlation with long range cli- matic fluctuations within that era, and it is believed that many all inclusive statements about Carboniferous climates do not take into account this probabil- ity. It is inconceivable that a lapse of time as long as the Carboniferous should have a uniform climate throughout or that conditions during that time could have been uniform over the whole earth. Coleman (6) has warned against sweeping conclusions concerning Paleozoic climates from the limited amount of available data. Although growth rings in Paleozoic woods cannot be interpreted as con- stituting evidence of marked seasonal fluctuations within that era, it is im- possible to explain them on other grounds than that they reflect seasonal changes of some sort. Without seasonal variations of some kind it is impossi- ble to account for the origin of regular growth rings in the first place. It is reasonable to assume that the most primitive plants with secondary growth developed uniform wood and year after year each successive addition of cells was indistinguishable from the previous one. Only when vascular plants had become more specialized both in structure and in their ability to respond to the environment did growth rings appear. Considering all the information available it is believed that the follow- ing conclusions are justified: (1) that the occurrence of growth rings in some
- 62 - of the Paleozoic woods is not a strong argument in favor of marked season- al fluctuations during that era, nor do their absence in others prove uni- founity of climate; and (2) that growth rings such as do occur are in all likelihood responses to seasonal fluctuations of some sort, to which cer- tain Paleozoic species were not sensitive. Various attempts have been made to correlate' other structures shown by Paleozoic plants with the environment. Bilignea solida, from the cal- ciferous sandstone (Lower Carboniferous) and Megaloxylon scottii from the Lower Coal Measures show large cavities in the pith which have teen inter- preted as water storage structures. Whether they served this purpose or not is entirely a matter of conjecture. Unless' there was an inadequately developed root system it is difficult to understand the utility value of such structures except for plants subject to extreme aridity. However, if the plants grew in situations where there was a marked seasonal lowering of the water level the accompanying dryness might account for the presence of such cavities. Investigators have long hoped that an examination of structurally preserved leaves -.Tould ultimately thro1.? more light on the question sur- rounding Paleozoic ecology, but the results obtained have not come up to expectations. Such features as epidermal outgrowths, size and number of intercellular spaces, arrangement and position of the stomata, cuticulari- sation and enrolling of the margin are known definitely to have some corre- lation with environment, but the degree of correlation is usually peculiar to the species. Only under extreme aridity, salinity, exposure to high winds or other strenuous conditions do any of these features show specializa- tion to the extent of being conspicuous. Dre tolerance .of structural var- iations under nomal conditions of temperature and moisture is large, and different species in a similar habitat may show considerable variety of structure with respect to the features mentioned. Most Paleozoic leaves of which the structure is known could apparently survive very well under north temperate zone conditions such as prevail in the eastern or southern portions of the United States or along the Pacific coast as far north as Alaska. They show no special adaptation to subdued light, high temperature, or to excessive humidity. Seward's statement is appropriate: "Broadly speaking we see no indication that these leaves were exposed to any condi- tion of climate other than such as now obtain." (14, p. 72) IV. Environment and the origin of floras characteristic of certain periods The assumption that an assemblage of species, or a certain species regarded as an "index species," furnishes definite proof of the age. of a deposit sometimes leads to erroneous conclusions. To accept a given species or type as an absolute and undisputable time marker is frequently a con- fession of ignorance concerning its ecological relations. A given species exists in a given deposit largely because the environment was suitable for its existence up to the time of deposition, and with the approach of less favorable conditions it became less abundant or extinct. In this way fossils may best be regarded as "environment markers" rather than strict time mark- ers. The occurrence of Archaeopteris throughout the landward phase of the Upper Devonian delta on the northern Appalachian! region is a specific example.
- 63 - Another instance bearing out the same idea is the recent discovery in eastern Kansas (12) of an assemblage of plants consisting of Walchia, Taeniopteris and other typical "Permian" forms in a deposit of undisputed upper middle PennsyIranian age. A competent paleontologist, basing his as- sumption upon the floral assemblage alone, maintained that the deposit is unquestionably of Permian age. A study of its stratigraphic position, how- ever, completely invalidates this view. The flora represented here is re- garded as an upland one, and although contemporaneous or nearly so with the swamp floras of other parts of North America, developed along different lines. Then, with the approach of more favorable Permian conditions this upland flora survived and spread, and the descendants became the typical Permian types as they are now known. Here, as with Archaeopteris, what would othervrise be a stratigraphic anomaly can be satisfactorily explained when the environment and the adaptations to it are taken into consideration. . V. Summary and conclusions Any treatment of the paleoecology of Paleozoic plants is largely a consideration of problems and appraisals of previously offered theories. While the fundamental principles are probably the same as those applying to recent plants they are difficult to apply because the environmental relations of most Paleozoic plants are unknown. The same adaptations shown by recent plants cannot always be applied to Paleozoic plants "because of the remoteness of the relation between them, and also by the fact that there is a great dif- ference in sensitivity to adjustment anong different species. Sufficiently pronounced structural features in Paleozoic plants may indicate adaptations of a certain kind if comparable adaptations can be recognized in living forms. Environmental conditions varied in .time and place during the Paleozoic, and sweeping conclusions regarding temperature, seasonal fluctuations and humid- ity should not be drawn from plants within a limited area or restricted horizon. The old idea of a graduated climate from hot to cool within the Paleozoic has been gradually abandoned, but there still exists too great a tendency to re- gard the eras or periods as climatic units. The visible differences between fossil floras may be due as much to the various agencies in operation at the time of deposition, or to geograph- ical location, as to differences in time. Also, because of the complicated structure-of the plant body, the appearance of the resulting fossil may pre- sent a deceptive appearance. Lastly, fossils may be regarded as markers of environmental units as well as time units, and plant associations which developed within one per- iod but within a restricted range may have survived to a later time in which they spread when favorable conditions developed.
- 64 - References 1. Antevs, E. The climatologic significance of annual rings in fossil wood. Am, Jour, Sci., vol. 209 (1925), pp. 296-300. 2. Antevs, E. The big tree as a climatic measure. Carnegie Inst. Publ. No. 253 (1925), pp. 117-153. 3. Arnold, C. A. The genus Callixylom from the Upper Devonian of central and western New York. Pap. Michigan Acad, Sci,, Arts and Lett,, vol. 11 (1930), pp. 1-50. 4. Arnold, C. A. Cordaitean wood from the Pennsylvanian of Michigan and Ohio. Bot. Gaz., vol. 91 (1931), pp. 77-87. 5. Arnold, C. A. The so-called "branch impressions of Callixylon newberryi (Dn) Elkins and Wielond and the conditions of their preser- vation. Jour. Geol., vol. 42 (1934), pp. 71-76. 6. Coleman, A. P. Late Paleozoic climates. Am. Jour. Sci., vol.209 (1925), pp. 195-203. 7. Goldring, W. Annual rings of groTrth in Carboniferous wood. Bot. Gaz., vol. 72 (1921), pp. 326-330. 8. Goldring, W. The Upper Devonian forest of seed ferns in eastern New York. New York State Mus. Bull 251 (1924). 9. Gordon, W. T. The genus Pitys, Witham, emend. Trans. Roy. Soc. Edinburgh, vol. 58 (1935), pp. 279-311. 10. Huntingtcn, E. Tree growth and climatic interpretations. Carnegie Inst. Publ., no. 352 (1925), pp. 157-204. 11. Kidston, R. and W. H. Lang. On Old Red sandstone plants showing struc- ture, from the Rhynie chert bed, Aberdeenshire. Part V. Trans. Roy. Soc. Edinburgh, vol. 52 (1921), pp. 588-597. 12. Moore, R. (C., M. K. Elias and-N. D. Newell. A "Permian" flora from the Pennsylvania rocks of Kansas. Jour. Geol., vol. 44 (1936), pp. 1-31, 13. Thomas, D« E. A new species of Calamopitys from the American Devonian. Bot. Gaz., vol. 97 (1935), pp. 334-344. 14. Seward, A. C. Fossil plants as tests of climate. London (1882) 15. Seward, A. C. Plant life through the ages. New York and Canbridge (1931). 16. Seward, A. C. A Persian Sigillaria. Phil. Trans. Roy, Soc. London, B., vol. 221 (1932), pp. 377-390. 17. Steidtmann, W. E. Cordaitean wood from the Pennsylvanian of Kansas. Amer. Jour. Bot., vol. 21 (1934), pp. 396-401.
