Tables of Contents:
Volume I
Volume III
Volume II
Atlas    Index
All Four Volumes:
Scans and Webpage
©George Langford III, 2011



Remains of plants referable to this order are abundant in the coal measures, in fragments of ribbon-like leaves, most rarely found in connection with the stems. For since the discovery by Sternberg of a stem bearing leaves, which was used by Corda for the analysis and illustration of his Flabellaria (Cordaites) borassifolia, I do not know that, until recently, any other has been found except the one represented, Atlas, Plate LXXVII, f. 3, discovered years ago in the upper Salem vein, near Pottsville, Penn'a. Now we know the Cordaiteae by splendid materials, stems bearing leaves and flowers, fruits attached to stems, etc., published two years ago in the Flore Carbonifere of Grand'Eury, and for this continent by those not less valuable to science, obtained especially by Mr. I. F. Mansfield, in his coal bed of Cannelton. These American specimens have afforded the means of comparing and confirming the observations of the celebrated French author, and of adding to the Cordaiteae some new generic divisions.

Cordaites, and the fruits and flowers, Cordaicarpus, Cordaianthus, Cordaistrobus, I admit in this order as generic divisions, Dicranophyllum, Grand'Eury, Desmiophyllum, and Lepidoxylon. The relation of these last genera is as yet indicated merely by the ribbon-like shape of the leaves, a character which may be conventional only; for, further discoveries may supply specimens indicatlng by new and more important characters a relation to different orders of plants. This has been the case already with the Taeniophyllae, formerly described in the Cordaiteae, and now forcibly placed with the Lycopodiaceae on account of their fructifications.

Cordiateae are now generally referred to the Dicotyledonous Gymnosperms, as intermediate in characters between the Cycadeae and the Conifers. Corda, to whom we owe the anatomy of Cordaites, from a species described in Beitr., p. 44, Pl. XXIV and XXV, as Flabellaria borassyolia, compared the structure of the plant to that of Dracaena, and to Lomatophloios especially, from the characters of the pith or medular cylinder.

Sternberg had already examined this plant and referred it to Palms, while somewhat later, another specimen of this kind, with narrower leaves, is mentioned by the same author as related to the
Cycadeae, Fl. d. Vorw., 1, p. 39, Pl. XL. From the apparent cylindrical form of the leaves this fragment seems to have the character of Taeniophyllum.

Brongniart, in his tableau des genres, p. 65, proposed a new genus,
Pychnophyllum for that Flabellaria borassifolia, and considering its structure as related to that of the Sigillariae, the Cycadeae, and the Conifers, he placed it in the Noeggerathiae, between these two last families.

Goeppert, admitting the
Noeggerathiae as Monocotyledonous, is however of the opinion of Goldenberg, that from their flowers and fructifications they should be placed between the Cycadeae and the Conifers.

Schimper places in the
Cycadeae, as genera of uncertain relation, Pychnophyllum, Brgt., for the description of Cordaites species, and a new genus, Psygmophyllum for that of the Noeggerathiae.

Weiss (1869) says that the inflorescence and fruits of these plants relate them to the
Cycadeae and the Conifers, but that by their leaves their affinity is with the monocotyledonous, while Heer (1876) admits them in the Conifers, and Grand'Eury, to whose admirable work we mostly owe our acquaintance with these two orders of plants, the Noeggerathiae and Cordaiteae, places them in the Gymnosperms, after the Sigillariae, thus remaining in concordance with Brongniart's opinion.

In a late work just published,
Structure Comparee, etc., Nouv. Arch. du Museum, II, 2d series, a work as remarkable by the precision of the anatomical details as by the extent and accuracy of the comparative researches, the author, M. B. Renault, resumes the results of the microscopical analysis of the Cordaites in an exposition of the characters of their woody cylinder and pith, their leaves, their flowers, their fruits, and concludes, p. 323, that these plants belong to the order of the Cycadeae, of which they constitute a distinct family. The exposition of the characters and the deduction derived from them are so clear that this conclusion seems indisputable.


Trunks of large size, irregularly branching, formed of a large medular canal or pith, marked on the outer surface by transverse narrow parallel simple ribs rarely joined by divisions, covered by double or triple layers of wood and bark converted by fossilization into thin layers of coal; leaves in spiral order, more or less distant, ribbon-like, of various length and width, linear or more generally gradually enlarging upward, obtuse, entire or undulate and split at the apex; borders curving to the sessile or semi-embracing somewhat inflated base; surface marked lengthwise by primary and secondary parallel simple nerves, generally more distant in the middle of the leaves and slightly inflated towards the base; flowers in racemes from the axils of the leaves; fruits generally oval, sessile, of various size.

The flowers, when found separately, represent the old uncertain genus,
Antholites; the fruits have been referred, until now, to different species of Carpolithes, Trigonocarpus, etc. Both these kinds of organism are further considered under the name of Cordaianthus and Cordaicarpus.

The decorticated stems of
Cordaites have been often described under the generic name of Artisia and Sternbergia.

Cordaites were, generally at least, or for some of their species, arborescent plants of great size. Grand'Eury has reconstructed, on a splendid plate, Fl. Carb., Pl. D, a number of trees which he estimates twenty to forty meters in altitude, their trunks forty to fifty centimeters in diameter. The divisions are oblique, the branchlets small, with crowded, imbricated short and small leaves, Atlas, Plate LXXVIII, f. 2, showing how far different are the leaves upon branches of different age, and how great is the difficulty of their identification when they are merely represented by fragments, or even when entire but separated from their supports.

In this case the nervation of the leaves is the essential character applied to the specific determination of the cordaites. But, as the distance between the primary nerves and the number of the intermediate veins vary, not only in leaves of different ages, but even in the same leaf, this character is not entirely reliable.

The stratigraphical distribution of
Cordaites like that of Lepidodendron and Sigillaria, is from the lowest sub-carboniferous coal measures to the base of the Permian.

Goeppert describes
Cordaites principalis and Cordaites borassifolius from the Permian of Silesia. I have not seen in North America any specimens of Cordaites from above the Pittsburgh coal. Species of the genus are extremely abundant in some localities, where their leaves, heaped and compressed in mass, are the essential constituents of whole beds of coal or of shale. The absence of branches, of fragments of stems in connection with these leaves, indicates for the plants a growth of long duration. The leaves, flowers and fruits were easily, perhaps annually, detached from the trees.


CORDAITES VALIDUS, Lesqx., Plate LXXVI, Figs. 1-2a.

Proc Am. Phil. Soc., 1878, p. 317, Pl. XLVII, f. 1, 2.

Stem narrowly irregularly striate; leaves long, linear, half embracing the stem at base, slightly enlarging above the point of attachment marked by a subcordate scar, obscurely nerved on the upper surface; nerves simple or with a single intermediate vein.

The fragment of stem figured is coarsely nerved or thinly striate, the lines being irregular in size, here and there inflated, half to one millimeter distant. The leaves appear very long, as seen from divers fragments, one of which is thirty-five centimeters long, linear, five to eight centimeters broad. The primary nerves, three to five per millimeter, obtusely keeled, simple or with an intermediate secondary vein, are scarcely distinct on the upper surface of the leaves, but clearly seen upon the lower, where they appear more equal and more numerous. The coaly layer of the bark of the stem is about one millimeter thick. The same thickness of coal fills the place of the leaves upon their slightly concave impressions, between the upper and lower faces, and represents their thickness.

The specimen,
Atlas, f. 1, seems to represent the leaf as decurrent on the side. The branch is broken at the point of attachment, and the apparently prolonged side is merely the turning of the border toward the stem behind, from which the medial and lacerated part of the leaf is detached.  Atlas, f. 2 shows the base of the leaves as seen separated from the stem and flattened by compression. That part is deeply undulate-laciniate and the base of the laciniae slightly inflated by fascicles of veins confluent to it. Atlas, f. 2a represents the shape of the scars left by the base of the leaves upon large stems.

Grand'Eury seems to have seen leaves like those described above, for he remarks in a note on a sub-species of
Cordaites borassifolius, under the name of Cordaites crassifolius, Fl. Carb., p. 216: "We do not know yet if we can refer to this type leaves of a thicker consistence, one face of which is marked by stronger nerves and alternate finer veins, while the other is more equally and definitely lined." This remark applies exactly to the nervation of the leaves of this species.

Habitat—Cannelton, Penn'a., Mr. I. F. Mansfield.


Noeggerathia crassa ? Goepp., Uebergsg. fl., p. 280, Pl. XL.

Fragment of broad linear leaves, coarsely striate lengthwise; texture thick.

I refer to this a number of specimens whose facies resembles that of flattened stems of
Calamites, irregularly ribbed lengthwise, often very long, always without any trace of articulations. These fragments vary in thickness from two to five millimeters; the striae appear composed of fascicles of veins here and there inflated, sometimes buried under a thick epidermis.

I am unable to say if these fragments represent Goeppert's species, or even if they are referable to this genus. Years ago I sent to Prof. Brongniart some of these specimens, but do not know if they have been examined by the celebrated author.

Habitat —Especially found in the anthracite beds of Penn'a.



Proc. Am. Phil. Soc., 1878, p. 318, Pl. XLVIII, f. 1-2a.

Leaves large, of a strong texture, widening upwards and fan-like from a narrow semi-lunar base, round truncate or rounded and undulately lobed and split at the top; nervation double; primary nerves obtuse, dichotomous or splitting, inflated, and more distinct toward the base, with one often indistinct intermediate vein.

Of this species I have not seen any stems, and all the leaves which I had for examination have the same truncate narrow base, one of them only being cut at the point of attachment in the semi-lunar form of the leaves of
Cordaites. Among the fine specimens sent by Mr. R. D. Lacoe, most of which are too large for illustration in our plates, the outspreading upwards is marked in different degrees. One of the leaves, for example, thirty-eight centimeters long, is gradually enlarged to the rounded top, where it is sixteen centimeters wide, undulate and split in short laciniae like Atlas, f. 1. Another with the base six millimeters broad, truncate, but concave at the point of attachment, is thirty-two centimeters long and fifteen centimeters broad at the apex, or there nearly half as broad as long.

Among the specimens of this species, one bears with fragments of leaves a flowering raceme upon a long axis. The raceme is curved or flexuous in the middle, about eighteen centimeters long. The axis, two millimeters in diameter, is flat, slightly channeled in the middle, narrow, two millimeters, bearing opposite sessile oval or obovate gemmules, superposed to a few imbricated narrow scales which are easily detached, the lowest nutlets being all naked. The branch has twenty-two pairs of these fruits, the largest the basilar ones, four to five millimeters long, three and a half millimeters broad, mostly of the same size, the upper ones only slightly and gradually smaller. Besides the branch the same specimen bears scattered, nearly globular nutlets, seven to eight millimeters in diameter, slightly emarginate at the base, surrounded by a ring, and resembling
Diplotesta Grand'Euryi, Brgt., as figured by Grand'Eury., Fl. Carb., Pl. XXVI, f. 27, or nearly similar to f. 49-50, Atlas, Plate LXXXV. But all the figures of Grd'Ey., Pl. XXVI, of Cordaianthus baccifer represent the fruits close and alternately disposed in short racemes, either supported at the base by imbricate narrow scales or posed in the axils of long linear bracts.

As this branch is disconnected, its reference to
Cordaites grandifolius as its fructification is not positive. This Cordaites is, however, the only one of which leaves have been found at the same locality, in a bed of shale covering a very limited area, about one square rod only.

Habitat—Sub-conglomerate ledge of Pittston, Pa., All the specimens have been found and communicated for examination by Mr. R D. Lacoe.


This section might be sub-divided into two groups, one for the species with large leaves, more generally found in the middle coal measures; the other for the narrow-leaved species, which appear related to those described by Grand'Eury, under the name of Poa-Cordaites. Though I cannot find either in the nervation or in the basilar form of the point of attachment of the leaves, any persistent characters which could enable me to distinctly separate them, these sub-divisions are preserved for convenience.

A. Latifoliae.
CORDAITES BORASSIFOLIUS, (St.) Unger, Plate LXXVI, Figs. 3-3b.

Flabellaria borassifolia, St., Fl. d. Vorw., II, p. 27, Pl. XVIII.

Cords. Beitr., p. 44, Pl. XXIV, XXV.

Pychnophyllum borassifolium, Brgt., Tabl. d. genres, p. 65.

Schp., Paleont. veget., II, p. 190.

Cordaites borassifolius, Ung., Gen. and Sp., p. 277.

Lesqx., Proc. Am Phil. Soc., 1878, p. 319, Pl. XLVII, f. 3-3b.

Leaves generally large, from five to eight centimeters broad in the middle where they appear the widest, gradually narrowing both ways, upward to the obtuse or truncate apex generally more or less deeply split, downwards to the slightly contracted semi-lunar somewhat inflated base. Nervation indistinct to the naked eyes, close, five to seven primary nerves in one millimeter, and generally one intermediate thin veinlet; surface marked by cross wrinkles, more distinct than in the former species.

As figured by Corda, who has given the characters of the nervation, and of the areolation, the leaves are all shorter than I have generally found them. But the branch which the German author has figured is a young one, bearing top-leaves only. I have seen in Kentucky, near Amanda furnace, a bed of clay iron ore composed mostly of remains of this species, where, amongst an immense number of fragments, there were well preserved leaves five to six centimeters broad, some very obtuse, half round at the apex, some also split into laciniae in the middle, others narrowed at the top like that figured in Atlas. This one is cut in two, the middle part being left out from want of space. It measures, in its whole, forty-five centimeters in length and six centimeters in width in the middle, where, on the figure, the intermediate lines mark the diameter of the leaf.

Habitat—Found in most of the beds of the carboniferous measures, from the Millstone Grit to the Pittsburgh coal. Not rare at Cannelton, Pa.


Fl. Carb., p. 218, P1. XX.

Leaves large, gradually widening from the base to a very obtuse apex, Ungulate or spatulate in outline; primary nerves distinct, at equal distance through the whole width of the leaves; intervals obtusely keeled, transversely rough, without intermediate veins.

The leaves which I refer to Grand'Eury' s species are generally long, fifty centimeters or more, seven centimeters broad in the largest part toward the top, at twelve centimeters distance from the round apex and thirty centimeters from the base, which is two centimeters in width. This measure gives the exact form and dimensions of nearly all the numerous leaves which I have seen of this species. The author describes and figures it with much shorter leaves, but remarks that they are often as long as sixty centimeters. His description of the nerves, as fine and sharply defined, nearly equal, is applicable to the American specimens. In some of these, the epidermis renders the surface quite smooth and the veins obsolete in some places; but they are always distinct and sharply marked under the easily removed epidermis. The fruiting racemes of this species, attached to the axils of the leaves and generally covered by their base, are short, six centimeters long, and bear alternate round. nutlets, two to three millimeters in diameter, four millimeters distant, each in the axil of a long linear bract inclined upwards, two and a half centimeters long. They have the characters of the branch of
Cordalanthus baccifer, Grand'Eury, l.c. Pl. XXVI, f. 15, a little more distinctly round than oval.

Habitat—Boston mine, B and C seam, two and a half miles south-east of Pittston, Pa. Mr. R. D. Lacoe.


Proc. Am. Phil. Soc, 1878, p. 320.

Leaves of various size, generally smaller than those of Cordaites borassifolius, more rapidly narrowed to the base, distinctly nerved, though covered with a thick epidermis.

The largest leaf seen of this species is twenty-two centimeters long, thirty-seven millimeters broad near its top where it is broken, fifteen millimeters broad just above the point of attachment, with borders generally recurved. The primary nerves obtuse and less distinct on the upper surface than on the lower, generally three in a space of two millimeters, are separated by two to four intermediate secondary veins, a character which easily separates this species from the others of this group.

One of the specimens examined is a stem with some leaves and distinct semi-lunar scars of others already detached. It bears also a young branch with shorter closely imbricated. leaves, two centimeters long. It is in the same position and direction as the one in Atlas,
Plate LXXVIII, f. 2, and bears also apparently from the axils of the leaves, small gemmules (only one is distinct) covered with imbricated scales representing in a very diminutive shape Cordaianthus gemmifer (Atlas, Plate LXXVI, f. 5, 5a). These gemmules are not larger than those f. 4 of the same plate. Their mode of attachment is not seen. The stem of the species is covered by a coating of coal about one millimeter thick, with leaf-scars distinct.

Habitat—Clinton, Mo. Dr. J. H. Britts.

CORDAITES LACOEI, Sp. nov., Plate LXXXVII, Figs. 2-4, (this vol.)

Leaves small, elliptical, obtuse or rounded to an obtuse acumen, with a comparatively broad, inflated base; nerves distinct; texture thick, subcoriaceous.

Of these leaves I have seen four, none attached to a stem. They vary in size from three to twelve centimeters long and from one and a half to five centimeters broad. In the largest leaves, the semi-lunar base is two centimeters across, in the others a little less. The venation is very distinct. The primary nerves are obtusely keeled, three fourths to one millimeter distant, with four to six very thin intermediate veins, as seen
this volume, f. 3a, enlarged. The species is closely allied to Cordqites foliolatus, Grand'Eury, Fl. Carb., p. 219, Pl. XXI, f. 3a, differing in the comparatively broader size of the leaves, the larger base, and the nervation. The substance of these leaves is also different; for, according to the author, it is thinner in his species than in Cordaites lingulatus, while in the American specimens, the leaves are comparatively thick—nearly coriaceous.

Habitat—Pittston shale above coal E.  Mr. R. D. Lacoe.

B. Angustifolia.


Proc. Am. Phil. Soc., 1878, p. 320, Pl. XLVIII, f. 3, 3a.

Cordaites angustifolius,* Lesqx., Geol. Rept. of Ill., IV, p. 420.

* Name preoccupied by Dawson, Can. Natur., 1861, p. 10.

Leaves narrow, linear, half embracing at the point of attachment, imbricated at base, curved backward along the stem, upraised in tufts at the top; primary nerves very distinct, generally at equal distance, with intermediate very thin veins; surface flat, epidermis thick.

The figured specimen has been mentioned already as the first fragment found in this country of a stem of
Cordaites bearing leaves. The figure represents the base of the leaves a little too large, at least as seen from other fragments from different localities. The leaves are, however, extremely variable. One of them, from Clinton, Mo.. measures seven millimeters at the point of attachment, is immediately enlarged above it to eleven millimeters and gradually widens upwards to seventeen millimeters at the point where it is broken, eleven centimeters from the base. Other leaves from the same locality are exactly linear in their whole length, seventeen millimeters in diameter, while others still, fifteen millimeters broad just above the point of attachment, gradually decrease in width upward. These measurements show the variations in size of these leaves not merely in regard to their relative position, but comparatively to the different parts of their length.

To this species I refer a large number of separate leaves obtained at the same locality as the specimen f. 3. Some are flat, linear, others have the borders distinctly curved down or are half cylindrical; others still are true cylinders, not larger than goose quills, seemingly coming out of a common pedicel, being often found in bundles, enlarging upwards in proportion as they open and the borders become more and more flattened. The cross section of these leaves is a circular or an oval line. Though the surface of the leaves is covered with a thick shining epidermis, the primary nerves, three or four in one millimeter, are always so distinct that they are easily counted without a magnifier. But the intermediate veins, three or four, very thin, are scarcely perceptible without a strong glass.

Habitat—Upper coal measures of Penn'a, especially the South Salem vein of Port Carbon and around Pottsville. Low coal measures of Missouri, Clinton. Middle and lower coal of Ill. Abundant at Colchester, St. John, and Duquoin.

CORDAITES MANSFIELDI, Lesqx., Plate LXXVI, Figs. 4-4b; Plate LXXVIII, Figs. 1-2; Plate LXXXVII, Fig. 18, (this vol.)

Proc. Am. Phil. Soc., 1878, p. 321, Pl. XLVII, f. 4-4b; XLIX, f. 1, 2.

Stem with a thin bark of polished coal, indistinctly marked by the scars of the convex base of leaves, either close or more or less distant, disposed in spiral; leaves long, erect, nearly exactly linear, gradually diminishing near the top to an obtuse point, averaging fifteen millimeters in diameter, distinctly and distantly nerved; primary nerves fifteen to eighteen in one centimeter, with two to four intermediate veins; flowers composed of four sepaloid involucres, borne upon simple flexuous pedicels to which they are attached by short peduncles; fruits in large nuts sessile upon separate branches.

As seen from the splendid specimen figured here and from a number of others quite as remarkably well preserved, the species is characterized by its long erect linear leaves whose surface is marked by a distinct nervation, (
Atlas, f. 1a enlarged double, f. 1b enlarged four times.) The leaves, rounded and narrowed to the point of attachment reduced to half their diameter, are perfectly entire and obtusely acuminate. The stems are covered with a thin coating of coaly, shining bark, where the leaf-scars are indistinctly marked as they are also upon the subcortical surface. The branches apparently form the axils of the leaves, one of which is seen Atlas, f. 2, bear closely imbricated leaves proportionate in size to the length and age of the divisions. They are linear-lanceolate, obtusely pointed, with a nervation of the same character, reduced, of course, to proportionate dimensions by the size of the leaves. Another specimen bears a branch two to three centimeters thick, diverging in the same degree as the one figured, twelve centimeters long, with leaves also of proportionate size, the largest already fifteen centimeters long, all imbricated, linear-lanceolate, with the borders incurved, especially towards the top which thus appears acuminate. The nervation has equally the character of the larger leaves, the primary veins being one half to three fourths of a millimeter distant, with two or three intermediate veinlets.

Under the name of
Cordaianthus simplex, I refer the stem-bearing flowers, Plate LXXVI, f. 4, 4a, to this species, especially because the specimen was found in the same shale and in proximity to those bearing leaves, though not in connection with them. This stem is somewhat different by its slightly thicker coaly surface. The leaf-scars are indistinct, as is generally the case for this species, so that their relative position is rarely appreciable. The racemes of flowers, flattened by compression, are irregularly flexuous, with their vascular filaments distinct, as if the branches had been in an advanced state of maceration. The flowers, which appear to be male flowers, are borne upon short peduncles and pending, composed, as seen Atlas, f. 1a, of three or four involucral thick lanceolate acute scales. The point of attachment of the elongated narrow racemes is round, inflated in the lower part, as seen Atlas, f. 4b. Their position in regard to the leaves is not possibly seen.

All the flowering racemes of
Cordaites figured by Grand'Eury have the flowers, either sterile or fertile, sessile upon the branchlets. It is the same with those figured by Dawson, under the name of Trigonocarpum racemosum, Quat. Jour. Geol. Soc., vol. XVIII, p. 324, Pl. XVI, f. 47, which are referable to Cordaianaus baccifer, of Grand'Eury and with those of Weiss, Foss. fl., p. 195, f. 1, representing Cordaianthus gemmifer. A point of likeness only is found for the inflorescence of Cordaicarpus Mansfieldi, by the form of the flowers attached to a short pedicel and the thick raceme, in St. Fl. d. Vorw., I, Pl. XXVI, f. 2. This figure, though described without reference as plantula debilis, p. 33, represents evidently the flowering branch of a Cordaites.

As seen from the specimen bearing fruit,
Plate LXXXVII, f. 8 (this vol.), representing a branch or piece of bark of this species, well characterized by its thin shining coaly bark and the indistinct leaf-scars, the fruit is comparatively very large, obovate, six and a half centimeters long, four and a half centimeters broad above the middle. The outer envelope, testa, is a double coating of coaly matter, reduced, upon the flattened surface of the fruit, to a thickness of one millimeter, but originally twice as thick, as seen from the flattened borders of other specimens. The surface, either covered or naked, is distantly obscurely striate, more opaque and more distinctly lined under the testa. This fruit is evidently of the same kind as that figured under the name of Rhabdocarpus Mansfieldi, Atlas, Plate LXXXV, f. 21. I have seen a number of these fruits, which differ only by their size, and this even in a small degree. The last-mentioned figure is overturned. The oval depression at the top marks the point of attachment distinctly seen upon the figured specimen recently obtained, and still attached- to the slightly inflated basilar impression of the stem. This last specimen bears scattered leaves of Cordaites Mansfieldi, separated from the branch; but the peculiar character of its stems clearly indicates the relation of the fruit to this species. This fine Cordaicarpus is related to Rhabdocarpus multistriatus, St.

Habitat—Cannelton. Mr. I. F. Mansfield.

CORDAITES GRACILIS, Lesqx., Plate LXXVII, Fig. 4-4b.

Proc. Am. Philos. Soc., 1878, p. 322, Pl. XLVIII, f. 4-4b.

Stem slender, with a rugose somewhat thick bark; leaves narrowed at the base, open or inclined backwards, distant, gradually enlarged from the base to the apex, sub-linear, obtusely truncate; nervation obscured by a slightly rough epidermis; primary nerves variable in distance from one fourth to one millimeter, with one to four intermediate very thin veins.

Species allied to the former by the nervation, but distinct  from it by the shape, the size, the relative positions, the direction of the leaves and the thick rough bark of the stem. The leaves are nine centimeters long, five millimeters broad at the base, one centimeter at the obliquely truncate apex. The point of attachment,
Atlas, f. 4b, is slightly tumescent, reniform.

Habitat—Morris, Ill., Mr. S. S. Strong. One specimen with broken leaves, on which the point of attachment is distinct, is from Cannelton, Pa.

CORDAITES RADIATUS, Sp. nov., Plate LXXXVII, Figs. 5-7, (this vol.)

Branches small; leaves short, narrow, linear, obtuse, placed in right angle and star-like around the stems; primary nerves strong, obtuse, variable in distance; intermediate veins two to five, very thin.

There are, in the Mansfield collection, four specimens representing this remarkable species. They are all fragments of stems of various thickness, three to ten millimeters in diameter, with leaves disposed star-like and flattened upon the stone all around the axis. The smallest branch has the leaves two centimeters long and three millimeters broad, the largest has them six to eight centimeters long and six millimeters broad; another,
Atlas f. 5, is intermediate, and like the large ones also, is part of a stem covered by leaves horizontally diverging, so that each section of the stem shows them placed exactly llke the rays of a star. The primary nerves are very distinct, cylindrical, and immersed into the epidermis under which their impressions are concave. They vary in distance and are separated by very thin veins in number proportionate to the intervals, one to three. These veins are not discernible upon the leaves covered by the epidermis, but distinct underneath. By the nervation this fine species is related to a costatus. Its primary nerves, separated by two or three veinlets, are not as thick; and the texture of the leaves is thin, the stem not costate, etc.

Habitat—Cannelton. Mr. I. F. Mansfield.

CORDAITES COSTATUS, Lesqx, Plate LXXX, Figs. 1-3; Plate LXXXVII, Figs. 1, 2, (this vol.)

Proc. Am. Phil. Soc., 1878, p. 323, Pl. LI, f. 1-3.

Cordaicarpus costatus, Lesqz., ibid., 1878, p. 222, Pl. III, f. 1. 2.

Stems irregularly costate by the decurring prolongation of the tumescent leaf-scars; leaves erect, narrow, sublinear, slightly enlarging upwards; primary nerves at unequal distance, three to five in a space of two millimeters; intermediate veins three to four; surface wrinkled across; male flowers attached to narrowly oval sessile flexuous axillary racemes; fruits large, oval, slightly contracted to the tumescent point of attachment upon a narrow branch.

The stems seen of this species are not large, two to four centimeters; the leaves narrow, five millimeters above the base, gradually enlarging to one centimeter at the point where the longest (ten centimeters) is broken, none being preserved in its whole length. They are erect from a tumescent reniform point of attachment, which, narrowed and continuous downward, constitute, after disruption of the leaves, narrow carinate interrupted ridges which become sharply keeled upon old stems, as in
Atlas, f. 3, or form under the cortex elongated linear furrows.

The male flowers,
Atlas f. 1b, narrowly oblong, composed of narrow linear-lanceolate scales, imbricating at the base, are sub-sessile upon long cylindrical simple peduncles, slightly inflated or as if articulate at the point of attachment.

The fruit,
Cordaicarpus costatus, Lesqx., is a large oval nut, three centimeters long from its base to the very obtuse top, twenty-three millimeters broad, including the testa represented by the inflated border and a thin coaly layer upon the surface. The base is rounded in narrowing to the point of connection to the stem, evidently a separate branch one centimeter broad, flattened. The inflated supports of the fruits are exactly similar to those of the leaves of Cordaites costatus. This is the only proof I have of their reference to this species. But as stems and leaves of this species are common in the same shale where the fruit was found, and as this character is not remarked upon any other species from Cannelton I consider the reference as authorized.

Habitat—Cannelton. Mr. I. F. Mansfield.


Stems narrow, flexuous or serpentine, flattened at its top prolonged into a broad leaf.

CORDAITES SERPENS, Lesqx, Plate LXXIX, Figs. 1-4.

Proc. Am. Phil. Soc., 1878, p. 324, PI. L, f. 1-4.

Stem slender, flexuous or serpentine, abruptly rounded or truncate at the top in passing to a broad terminal long flat leaf-like prolongation; lateral leaves in right angle to the stems, sub-linear, narrowed to the point of attachment; nervation distinct.

The peculiar conformation of the stems is especially marked on
Atlas, f. 4, reduced to one sixth of the natural size. Besides the figured specimens, a number of others have been received and examined at Cannelton with the flexures of analogous shape.

The lateral leaves, five  to ten millimeters in diameter, narrowed in a curve to a semi-lunar point of attachment, are nearly linear, short, the upper ones cut in flexuous laciniae (flagellate),
Atlas, f. 1. The primary nerves, about at equal distances, are generally effaced near the borders of the leaves, three or four in a space of three millimeters, with three, sometimes five, thin intermediate veins. The nervation of the terminal leaves is of the same character.

The stems all slender, one and a half to two centimeters in diameter, are irregularly varlable in thickness at divers parts of their length as seen.
Atlas f. 1, still more distinctly, Atlas, f. 4. The surface is covered by a double layer, the upper bark thin, easily detached, represented Atlas, f. 1, 2, by flakes of coaly matter; the woody cylinder, thicker, more compact and tenacious, covering the pith, Atlas f. 3, which is transversely ribbed with the characters of Artisia or Sternbergia.

This stem,
Atlas f. 3, has preserved its cylindrical normal shape. I have other fragments about of the same size, merely of the internal pith, also in a decorticated state, one among others representing its top abruptly flattened or as pinched to a flat prolongation, as would be the stem of Atlas, f. 2, without its bark and its terminal leaf. This abrupt termination of the cylinder, the flagellate sub-divisions of the upper lateral leaves, Atlas f. 1, the enlarging of the terminal one, and still more the serpentine conformation of the stems, indicate the mode of vegetation of these plants. They were evidently growing flat upon the mud, expanding their leaves all around, supported by a woody cylinder, which, in reaching water, was transformed into a large probably very long floating leaf serving as a support to the stem.

Prof. Dawson has represented in Geol. Surv. of Can., 1871, Pl. III, f. 28, a branch of
Artisia abruptly terminating into a short cone, as in the stems of this specles. He considers it as probably referable to a Dadoxilon or Conifer. He refers also to the same genus another fragment of stem of a same character, bearing on one side a piece of thick bark like that bordering the stem, Atlas f. 3. Some of the specimens from Cannelton, like the one, Atlas, Plate LXXXI, f. 2, represent larger stems referable to other species of Cordaites, identified as they are by the character of the fragments of bark partially covering them. There is no doubt whatever that the transversely ribbed cylinders generally described as Artisia or Sternbergia represent generally if not always the pith of plants of this family. This opinion has been discussed in the length and admitted from sufficient evidence by Renault.*

* Struct. Comp., Nouv. Archives du Mus. II, 2d Series, p. 285, etc.

Habitat—Cannelton. Communicated in numerous specimens by Mr. I. F. Mansfield.



Can. Natur., May, 1861, p. 8. Dev. Plants, Quat. Journ. Geol. Soc. (1862), p. 316, Pl. XIV, f. 31a, b, c.

Leaves elongated, parallel-sided, an inch or more in width, with very delicate equal longitudinal striae.

On this plant, the author remarks that the leaves are very unequal in size, variable in form, oblong, nearly obtuse when young, with numerous equal parallel nerves, and a smooth surface. It closely resembles
Cordaites borassifolius, differing by the perfectly equal nerves.

Habitat—St. John, New Brunswick. This species is not positively known from American specimens, for the author says of it, that he has seen in the New York State cabinet a fragment of leaf from the Hamilton group not sufficiently perfect to render its identification certain.


Can. Natur., l.c., p. 10. Dev. Plants, Quat. Journ. l.c., p. 318.

Leaves elongated, one tenth to one fourth of an inch wide, with delicate equal striae.

According to the author, it differs from the former, at least, in its proportionate narrowness and decided striation. He compares it to
Noeggerathia graminifolia, Ung., "which in form and dimension it much resembles."

Habitat—Found in the Marcellus shale of New York.

CORDAITES FLEXUOSUS, Quart. Journ. Geol. Soc., 1863, p. 462.

Leaves lanceolate-acuminate, broad at the base; nerves numerous, parallel, somewhat sinuous and uneven.

It has the general appearance of the leaves of Cordaites, but less distinct nervation, and apparently less rigidity than the other species.

Habitat—Perry, Maine.


When the paper containing a description of species of Cordaites found in the coal measures of North America was published, Proc. Am. Phil. Soc., 1878, we had stem-bearing leaves and racemes of flowers of species of Cordaites, but not any fruits in connection with their supporting racemes or branches. I then described the flowers under the generic name of Cordaianthus, proposed by Grand'Eury, and of Cordaicarpus, for the fruits found detached or isolated among leaves of Cordaites, which were then hypothetically considered as their fructifications. Since that time new discoveries have supplied materials sufficient for the determination of three kinds of racemes of male flowers attached to stems bearing leaves of Cordaites, as described above: Cordaites lingulatus, Cordaites Mansfieldi, and Cordaites costatus, with the fruits also of the last two species, found in their normal position, attached to their supports. The mode of inflorescence of this order of plants is thus clearly recognized from American specimens.

But generally the fruits and racemes of flowers of
Cordaites are found isolated or in fragments, and though their relation to stem bearing leaves is unknown for the present, it is advisable to name and describe them.

It is for fragments of this kind that I have preserved here the names proposed by the French author, not as generic, but as convenient appellations, substituted to the more general and ancient name of

CORDAIANTHUS (gemmifer), Plate LXXVI, Figs. 5-6b.

(Expl. of the Plates)

Grd'Eury, Fl., foss. p. 228, Pl. XXVI, f. 4-7.

Buds or gemmules sessile, composed of imbricated scales, often attached to the axils of a linear bract.

CORDAIANTHUS OVATUS, Sp. nov., Plate LXXVI, Figs. 5, 5a.

Gemmules ovate, sessile in the axils of a linear bract; scales closely imbricate, ovate-lanceolate, acute.

These gemmules vary in size from six to eight millimeters, being generally longer than broad with a conical apex, but also rounded or obovate. The number of scales is proportionate to their age, and the scales are either closely appressed in their whole length or half open.

In the analysis of a specimen from Illinois which I described under the name of
Schutzia bracteata, Geol. Rept. of Ill., IV, p. 427, Pl. XXI, f. 6-9, I found, under the scales, a yellowish transparent membrane, formed of elongated equilateral small meshes, inclosing or supporting very small granules of opaque brown matter. These granules, scarcely the hundredth part of a millimeter in diameter, of an irregular polygonal or round shape, were agglomerated together, separating with difficulty. From their size, the irregularity of their shape and their mode of agglomeration, I considered these corpuscles as pollinic grains, evidently not spores. I have not seen with them any filiform remains, or fragments which might be considered as stamens.

Grand'Eury has seen the same kind of organism, which he describes, l.c., p. 229, and finds in their conformation an affinity to the pollinic vesicles attached under the bracts of Conifers. They are represented in his flora, Pl. XXVI, f. 2.*

* Renault has described and admirably figured transverse and longitudinal sections of flowering buds of Cordaites, exposing stamens, anthers, grains of pollen either free or taken from an unopened anther, or contained in the pollinic canal. Arch. du Mus., l.c., Pl. XVII. He finds the inflorescence analogous to that of the Cycadeae.

The specimen from Illinois probably represents a different species. But it is not possible to find specific evidence from its outside characters, as except the gemmules cut for analysis, most of the others are deteriorated by maceration, some of them resembling small oval nutlets transformed into charcoal.

Habitat—This species is abundant at Cannelton.

Branches thick, dichotomous; gemmules sessile in the axils of linear bracts, crowded, narrowly oval; scales lanceolate, acute, few.

The essential character of this species is the dichotomous division of the branches which seems to have been derived from a fasciculate rather than from a simple axis. No species of
Cordaianthus with a ramification of that kind has been described by the authors. All those figured by Grand'Eury have simple racemes. The gemmules appear to be in different stages of maturation; some with imbricated scales are naked, the bracts being already detached; others are like unopened buds in the axils of the bracts.

Besides the ramification, this species differs from the former by narrower longer scales. The branches are marked by transverse lines, like scars of detached tubercles and, upon some parts of the branches, the gemmules are evidently crowded in spiral order.

Habitat—Clinton, Mo., Dr. J. H. Britts.

Of the Cordaianthus (gemmifer), described and figured by Grand'Eury, l.c.: Cordaianthus glomeratus, Cordaianthus foliosus, Cordaianthus gracilis and Cordaianthus circumdatus, none are in close relation to those considered above.

CORDAIANTHUS, (baccifer.)

Racemes bearing, at the axils of foliaceous bracts, small rudimental or immature ovules, either obtuse or pointed.

The ovules are solitary not aggregate, with an outside solid envelope. Two species of these have already been described, one with
Cordaites grandifolius, as probably referable to this species, the other with Cordaianthus lingulatus, in branches attached to a stem bearing leaves.

These fructifications are very rare in our coal measures. Under the name of
Antholithes, and being probably a branch in a more advanced stage of maturation of Antholithes Pitcairniae, Ll. & Hutt., Dr. Newberry, (Geol. Rept. of Ohio, I, Pl. XLI, f. 2), has figured a raceme with a thick, apparently cylindrical axis, three millimeters in diameter, bearing opposite, exactly ovate-obtuse ovules, six millimeters long, four wide, in the axils of short lanceolate acute foliaceous bracts. No description is given of this fragment. It is related to Cordaianthus Sub-Germarianus, Grand'Eury, Fl. carb., p. 232, P1. XXVI, f. 10, which has the ovules close, alternate, and the bracts longer and narrower. Another specimen, also figured by Dr. Newberry, same plate, f. 4, represents distant, nearly round, abruptly pointed nutlets, one centimeter long and nearly as broad, attached without bracts and sessile to two fragments of striate sterns, one centimeter in diameter. These ovules are alternate, apparently bordered by a flat margin, like the fruits generally described under the generic name of Cardiocarpus. The nutlets are similar to Cordaicarpus minor,* of the same author, Atlas, Plate LXXXV, f. 38.
* [In the referenced plate, it's actually called Cardiocarpus minor - GL,III, ed.]

A specimen from Cannelton, in the museum of Princeton College, may be referable to this group. It represents a flattened stem about two millimeters in diameter, of the same character as the stem of Cordaicarpus ovatus. The simple axis bears alternate or opposite nutlets inclined upward, the lower ones naked or deprived of bracts, the upper ones with short acute bractlets. The ovules, three millimeters long, half as broad, appear double or enclosed two together and appressed against each other into a membranaceous involucre, as seen when the fruit is detached or when the top of the involucre is destroyed. Some of the round involucres are evidently empty or mere bags; this is the case with the lowest upon the raceme. The small seeds are obovate, attached to the support by the acuminate base; they do not appear as surrounded by a membranaceous border as in Cordaicarpus.

Under the name of
Botryoconus, Grand'Eury has remarked and described as referable to the family of the Conifers, seeds which, if not of the same order, have an analogy to those of this specimen. He has seen them enclosed in scaly buds forming ears, which he considers as far different from those of Cordaites. The type of this new genus is Antholithes Pitcairniae, Ll. & Hutt., II, Pl. LXXXII, f. 1 & 2; also figured by Newberry, in Geol. Rept. of Ohio, Pl. XLI, f. 1, 3. The last figure is separated under the name of Antholithes priscus, but it appears to represent the same organism as that of the English author in f. 1; that is, in a more advanced state of maturity. Grand'Eury and also Carruther have described these seeds as short, pedicelled, enclosed three or four in the middle of opposite distichous buds attached to a striate not articulate axis. These seeds are oval, acute, surrounded by a narrow flat border, often split at the top like those in Atlas, Plate LXXXV, f. 45-50, and described by some authors as Samaropsis. I have considered them under the old name of CordaicarpusAntholithes Pitcairniae (Botrioconus, Grd'Ey), is not rare in the American coal measures. I have obtained some specimens of it from the coal shale of Pomroy, Ohio, always in flattened, poorly preserved fragments. Dr. Newberry has his Antholithes specimens from the low coal of Youngstown.


To this generic division are referable seeds of various size and shape, hypothetically considered until now by their association to Cordaites as the fructifications of these plants. Of these fruits the number is considerable, but except the nutlets figured by Newberry as Cordaicarpus, by Dawson as Trigonocarpum racemosum, (very small nutlet of Cordaianthus baccifer), by Weiss as Rhabdocarpus, and later by Grand'Eury, no seeds has been observed attached to stem or branches of Cordaites. Except the Rhabdocarpus of Weiss, all the fruits above mentioned are very small and merely in an imperfect state of development.

I have described above from specimens obtained at Cannelton two
Cordaicarpus referable to their species of Cordaites, by the characters of the stem, and I have figured, Plate LXXXIII, f. 6-11, the fruits most commonly found in the shale of the Cannelton coal, where the remains of stems and leaves of Cordaites are most abundant. These I refer to Cordaicarpus. But of all the other species, Cordaicarpus, Trigonocarpus, Carpolithes, some of them considerd as referable to Cordaites, there are none, or scarcely any, at Cannelton, and until their relation is more positively established, I describe them separately under these last generic names, generally admitted by the authors.

CORDAICARPUS GUTBIERI, (Gein.), Grand'Eury, Plate LXXXIII, Figs. 8-11.

Cordaicarpus Gutbieri, Grand'Eury, Fl. carb., p. 236, Pl. XXVI, f. 19.
Gein. Verst., p. 39, Pt. XXI, f. 23-25.

Fruit oval or subcordiform, slightly truncate at one end, and pointed at the other, or broadly obtuse at both ends, narrowly margined; surface smooth, pericarp transformed into a thin coating of coaly matter.

Seeds of this character vary in size from one and a half to two and a half centimeters in transverse diameter, generally longer than broad, sometimes circular. They are abundant in the shale of Cannelton. One of the specimens seems to be attached to a branch, but the point of connection is not clear. As represented by Geinitz, l.c., these fruits are sometimes subcordate or emarginate at the base, as in his f. 25. I have not seen any of this form. They are generally exactly oval, broadly obtuse or even round, sometimes obtusely acute at one end, point of junction to a stem, as in
Atlas, f. 8, which corresponds in character with f. 24, of Gein., and f. 19, of Grand'Eury, l.c. The middle part of the fruit is generally somewhat inflated, as by an oval, nearly central nucleus (Atlas, f. 9), often surrounded by concentrical circles.

It is on the characters of fruits of this kind, when exactly round, that the genus
Cyclocarpus, Goepp., has been established.

Habitat—Cannelton, Pa.


Grand'Eury, Fl. carb., p. 236, Pl. XXVI, f. 20.

Fruit smaller, oval, depressed, or slightly emarginate at the point of attachment, covered with a coaly envelope, narrowly striate lengthwise.

The specimen represents two seeds upon a narrow cylindrical flexuous peduncle two millimeters in diameter, branching in the middle, so that one of the fruits is lateral, curved downward and short pedicelled, while the other is terminal on a longer subdivision of the peduncle. These seeds measure sixteen millimeters in length, and thirteen in width. The cortex is thicker than in the former species, and regularly minutely striate. In Grand'Eury, f. 20, l.c., the fruit is a little narrower towards the apex, more distinctly cordate at base and the short pedicel is scaly. The difference in shape is evidently the result of compression of the nutlets, which is oblique in our specimen, and vertical in that of the French author. The ramification of the peduncle in the American specimen gives to the fruit a different aspect. No trace of a border or of a wing is seen upon it.

Habitat—Cannelton, Penn'a. Other specimens have been communicated without pedicel, showing a close analogy to the former species.


* These three species are given, in the explanation of the plates, as Cordaicarpus Gutbieri.

Seeds of small size, oval, rounded at base, abruptly acuminate, bordered by a comparatively large ring, marked in the center by an oval, central ovule.

These seeds, six to ten millimeters in transverse diameter, ten to eighteen millimeters long, are distantly related to
Cordaicarpus congruens, Grand'Eury, l.c., Pl. XXVI, f. 21, and compare well enough by their shape and size with Rhabdocarpus lineatus, Goepp. and Berg., De. fruct., Pl. I, f. 15 & 16, figured from better specimens in Gein., Verst. Pl. XXI, f. 19-20. They might be referable to two species on account of their different size. But it is now certain. that the small ovules, which we see attached to their branches in an incipient state, are ovaries, which become with age, some of them at least, of very large size. This difference is not more remarkable than what is seen now between the ovaries and the mature seeds of a Juglans, or, to take a comparison nearer to the point, between the ovaries and the ripe fruit of a Torreya or a Ginkgo.

Habitat—Cannelton, Penn'a. Mr. I. F. Mansfield.


Strobile cylindrical, tapering to a blunt acumen, covered, by transversely rhomboidal scars placed in spiral, bearing narrow linear leaves, with the characters form and nervation of leaves of Cordaites.

The plant from which this genus is established might have been described as
Cycadoidea or Mantellia, generic divisions established for the description of stems of Cycadeae, which are mostly globular, or conical obtuse, or nest-form, all referable to a more recent formation, the Permian. As the leaves are of a different character, as also the reference of this cone to the Cycadeae is not positively ascertained, it is more rational, considering the characters of the leaves, to describe it in a separate genus, indicating its relation to Cordaites.


Lesqx., Proc. Amer. Phil. Soc., 1878, p. 328, Pl. LIII, f. 3.

Cone cylindrical from the base to the middle, narrowed upward and acuminate, borne upon a somewhat thick pedicel or axis equally striate in the length; scars transversely rhomboidal, inflated in the lower part, topped by diminutive leaf scars of the same shape, points of attachment of short linear obtuse leaves.

The leaves are mostly broken above, one only is preserved whole on one side; on the other side, four remain, close to each other, all evidently attached to the rhomboidal leaf scars which show the spiral disposition of the leaves. Their nervation, seen in
Atlas, 4a, is that of the Cordaites. The primary nerves are close, especially toward the borders, separated by one or two intermediate veins. The axis of the cone, is covered by a coaly bark, more than half a millimeter thick. It is deeply and regularly striate, the striae being also obscurely seen along the middle of the cone, even to its point, by compression of the scars, as represented upon the figure.

This cone seems to indicate, more than any other of the organs described, the relation of
Cordaites to the Cycadeae. By the leaves it is a true Cordaites; by the scars and their disposition it represents a small stem or a cone of Cycas. It is, however, difficult to explain its true nature. It does not look like a fruiting cone, and all that is known until now of the stems of the Cordaites is without relation to this organism. Another specimen of the same character is cylindrical in its whole length; but the top seems to have been destroyed by maceration. The shape and distribution of the leaf-scars are quite distinct.

Habitat—Cannelton. Mr. I. F. Mansfield.