ART. XXXVII.-Rhætic Plants from Honduras; by J. S. NEWBERRY. With Plate VIII.

IN 1886 Mr. Chas. M. Rolker, a mining engineer and graduate of the School of Mines, brought from San Juancito, Honduras, among other geological specimens, a piece of metamorphosed shale containing several impressions of the fronds of cycads. As no Mesozoic fossils had before been found in this region the discovery interested me much and I have since made earnest efforts to obtain other specimens from the same locality. Mr. Rolker kindly seconded these efforts and wrote to Mr. T. H. Leggett, E.M., who was located at San Juancito, giving him all the information he possessed in regard to the locality where the fossils were found, and soliciting his coöperation. This was cordially granted, with interesting results.

The specimen brought by Mr. Rolker was a loose piece picked up on the surface and nearly two years passed before its place of origin was ascertained. In January last I received a letter from Mr. Leggett announcing the discovery of the plant beds and the shipment to me of a box of fossils. These were exhibited at a meeting of the N. Y. Academy of Sciences. on Jan. 30, 1888, and were briefly described in the Transactions of the Academy, vol. vii, p. 113. Since that time Mr. Leggett has returned to New York bringing another box of fossils from the same place, among which are some additional species.

From the notes furnished me by Mr. Leggett it appears that the plant beds of San Juancito form part of a series of argillaceous shales now converted into hydromica schists several hundred feet in thickness. Below these, limestones crop out which are said to contain Carboniferous fossils, while above them are heavy masses of eruptive rock. The plant-bearing shales are much disturbed and metamorphosed, and are cut by a series of silver-bearing veins which have been worked for many years with considerable success. The outcrops which contain the plants are much decomposed and few good specimens have been obtained from them, but the number of species represented is large and it is evident that further excavation would result in the accumulation of much interesting material. The age of the deposit as indicated by its plants is plainly Upper Triassic and the flora as a whole has a great resemblance to that of the coal-bearing strata on the Yaki river in Sonora, Mexico, described by me in the Report of the San Juan Exploring Expedition, and to that described by Schenk in Die Fossile Flora der Grenzschichten des Keupers und Lias

Frankens and to Nathorst's Florau vid Bjuf; a number of the species being identical and others closely allied.

The localities nearest to Honduras where fossils indicative of Triassic age have been before discovered were in Sonora, Mexico, 2000 miles north, and in the Andes of Peru, where Triassic rocks were found by David Forbes, 2000 miles south. The plants contained in the collection made by Mr. Leggett are here briefly described:

Zamites (Pterophyllum) Rolkeri Newb. Figs. 1, 2.

Frond a foot or more in length by two inches in width; rachis chaffy; pinnules diverging from the midrib at an acute angle, alternate, closely set, attached by the entire breadth of the base to the upper surface of the midrib, which they completely cover; summit rounded or blunt-pointed; nerves fine, parallel.

This plant has the general aspect of those described by Heer in the Flora Arctica, vol. iii, Pl. XIV, XV, XVI, under the names of Zamites speciosus, Z. borealis and Z. acutipennis, and should be placed in the same genus. It also still more closely resembles Pterophyllum Nerbuddaicum Feistmantel (Flora of the Jabalpur Group, p. 14, Pl. VI, figs. 9, 9a), and "Zamites obtusifolius" and "Pterozamites gracilis" of Emmons (Amer. Geol. Rt., vi, p. 118, 119, figs. 85, 86).

It is evident that all these plants should be separated from Zamites if Z. Feneonis Brongt. be taken as a type of that genus. And they cannot be included in Pterophyllum if, with Schimper and Schenk, who have given us the latest and best classification of fossil cycads, we restrict that name to those in which the pinnules are set at a right angle with the rachis and are connate at their bases. Some writers would put them in Ctenophyllum and they are certainly closely allied to the group of cycads typified by Pterophyllum pecten of Lindley and Hutton (now Ctenophyllum), but have little in common with the gigantic Ctenophyllum grandifolium of Fontaine (Older Mesozoic Flora of Virginia), in which the pinnules are set on the side of the rachis, have the bases dilated and are sometimes a foot in length by half an inch in width.

Zamites (Otozamites) Leggetti Newb. Figs. 3, 4.

Fronds linear, one and a quarter inches in width by eight to ten inches in length; pinnules alternate, crowded, set on the upper side of the rachis, their bases meeting above, closely approaching below; in form they are oblong or linear, obliquely rounded above, sometimes slightly rounded at the base; nerves fine, radiate below, parallel above.

This plant, like the preceding, was briefly characterized in the Transactions of the New York Academy of Sciences, vol. vii, p. 114. It differs from Z. Rolkeri in its shorter and broader pinnules and in its nervation, which is partly radiate. Better specimens than any yet obtained are needed before a complete description of it can be written, or its generic relations be accurately determined. Some of the pinnules appear to be attached by the entire base, while in others the attachment is only by the central portion and the nervation is more distinctly radiate. Like the preceding species it is only referred to Zamites provisionally, and it is almost certain that they will ultimately be assigned to different genera.

In general aspect this and the preceding species closely resemble Ptilophyllum acutifolium and Pt. Cutchense Morris, Fossil Flora of India, but are distinctly separated from them by the mode of attachment and the nervation.

It is evident that still another review of fossil cycads is needed although we have had so many already, but this is no place for it and it will be sufficient for our present purpose to refer the plants described above to Zamites, as Heer has done those with which I have compared them, leaving their final generic titles to be decided hereafter, by the study of more complete specimens.

Otozamites linguiformis, n. sp. Figs. 9, 10.

Fronds strong; rachis striate; pinnules one to two inches in length by half an inch in width, crowded, sometimes overlapping; below issuing from the rachis at right angles, above obliquely; outline long-tongue-shaped; summits evenly rounded, bases unequally cordate, attached by a single point at the center; nerves numerous, fine, simple or forked, radiating from the center of the base to all parts of the margin.

This handsome species is allied to, though quite distinct from, Otozamites Macombii N. from the Upper Triassic rocks of Sonora, Mexico. (Report of the San Juan Expedition, p. 141, Pl. Iv, figs. 1, 2). In that species the pinnules are broader, are abruptly rounded or truncated at the summit, and on the lower part of the frond are quadrate; while the corresponding pinnules of the plant now described are oval. Some of the upper pinnules are set obliquely on the rachis, are long-elliptical in outline, slightly curved or sigmoidal, are obliquely rounded at the base and attached by a single point, thus conforming strictly to the definition of Glossozamites Schimper, and serving as a connecting link between that genus and Otozamites. Whether the rachis in our plant is furrowed as in the type species of Glossozamites (G. Zittelli Schenk) is not certain, but apparently it is so. Figure 10 represents one of the longer pinnules detached.

Taeniopteris glossopteroides Newb.

Frond simple, six to twelve inches in length by one and a half to two inches in width, spatulate in outline, summit subacute, base long-wedge-shaped, the median nerve strong and smooth, the lateral nerves relatively sparse and distinct, frequently forked at base, more rarely above, the branches sometimes approaching, forming elongated areoles but never inosculating.

This species was first obtained from Sonora, Mexico, and was described in the San Juan Report, p. 147, Pl. VIII, fig. 2a. By an error of the draughtsman the lateral nerves were represented as inosculating, but recent observation has shown that they do not join. Owing to this error another plant was confounded with this, which strengthened the misapprehension in regard to its nervation. This latter plant is shown in fig. 2 of the plate cited, where it is given a strong midrib, though it had none, unless toward the base; but has a distinctly reticulated nervation. It therefore belongs to Feistmantel's genus Gangamopteris and I have called it G. Americanus.* The plant I have named Taeniopteris glossopteroides is represented by a number of specimens in the collection of Mr. Leggett, but unfortunately owing to the weathering of the rock they do not show the nervation well. The lateral nerves from base to summit leave the midrib at an acute angle, generally arching with a gentle curve to the margin, but sometimes nearly straight. They are often forked near the base, more rarely above, sometimes running near together, but apparently never join. This nervation is just that of Taeniopteris murantacea Presl. (Pecopteris macrophyllum Brongt., Stangerites manantacea Bornem), a plant made the type of his genus Danaeopsis by Heer; but in that plant the frond is pinnate, in ours it is symmetrically spatulate, gradually narrowed to the base, and was undoubtedly simple. In all the characters shown in the specimens before me this plant comes nearest to the group of Permian species to which Schimper restricts the name of

* The distinction between Gangamopteris and Glossopteris is not quite as clear as it might appear from a comparison of the specimens figured and described by Feistmantel in his Flora of the Talchir-Karharbari Beds; the diagnostic character of Gangamopteris, the absence of a midrib, being variable, as shown in his Glossopteris decipiens (op. cit., p. 17, Pl. XVIII, figs. 3-5), in which the midrib fades out in the upper part of the leaf. This I find also to be true of a number of Australian specimens in my possession. These include many leaves of small size having the spatulate outline and rounded summit of G. Browniana, and labeled as such, and others, much larger, of G. ampla Dana. Both these forms show a midrib only near the base, and have a nervation but sparingly reticulate, with elongated meshes. Other specimens representing Dana's G. elongata and G. reticulum (Geol. U. S. Exploring Expedition Atlas, Pl. XIII, figs. 2, 3, 4) have the midrib persistent to the summit and the nervation strongly reticulate throughout. These two forms might stand for Gangamopteris and Glossopteris if to the former genus a midrib were conceded for part of the length of the frond.

Taeniopteris, viz: T. multinervis Weiss, T. Eckardi Germ, etc. More and better specimens must, however, be obtained from Honduras before the relations of these plants can be accurately determined.

Encephalartos? denticulatus, n. sp. Fig. 5.

Size of frond unknown, rachis smooth or finely striated longitudinally; pinnules diverging at an angle of about 45°, lanceolate, 30mm long by 6mm wide, acute, gradually narrowed to the point, abruptly narrowed at the base which is attached by its entire breadth; margins set with numerous, spiny teeth nerves fine, mostly parallel, somewhat radiate from the base and many terminating in the teeth of the margin.

;

Of this remarkable cycad only a single specimen has yet come into my hands. This is a fragment apparently from the middle of a frond showing three complete pinnules and the bases of two others. Its general characters are well shown in the figure. So far as known this is the first instance of the discovery of a cycad with denticulated pinnules in American Mesozoic rocks, and among foreign cycads only a group of species of Sphenozamites have pinnules with toothed margins. It is not uncommon to see this character in living cycads, particularly in Encephalartos and Zamia. In the latter genus the nerves are parallel and terminate in closely approximated marginal teeth or notches toward the upper extremity. In the former genus, however, forms occur which are almost exactly like those presented by the fossil under consideration, viz: fronds bearing pinnules obliquely inserted, contracted at the base, lanceolate in outline, having fine mostly parallel nerves and margins set with spiny teeth, e. g., Encephalartos Altensteinii Lehmann, Cape of Good Hope. This correspondence in the form of the pinnules is so close that I felt warranted in, placing our fossil provisionally in the genus Encephalartos. The fructification will of course be necessary for a demonstration of generic identity and has not yet been obtained.

Only one fossil species of Encephalartos has yet been described and that is E. Gorceixianus Saporta, from the Miocene Tertiary of Koumi, Greece. This has lanceolate, acute entire pinnules, two inches long, somewhat constricted at the base and slightly decurrent; nerves parallel. It is supposed to represent an extinct species of Encephalartos similar to E Lehmani of South Africa, but the similarity between these fossil species is scarcely as great as between the plant under consideration and E. Altensteinii Lehm., with which I have compared it.