4 ERNST ANTEVS, LEPIDOPTERIS OTTONIS (GÖPP.) SCHIMP. AND ANTHOLITHUS ZEILLERI NATH. 
Frond (bi- to) tripinnate, narrowing sharply towards both ends. Main rachis 
broad, with close, rounded tubercles. Pinnae of first order close, sub-opposite to 
alternate, long and not very broad, lanceolate or all but linear; rachis with tubercles. 
"Zwischenfiedern”. Pinnae of second order close, linear to lanceolate, narrowing 
towards the apex. Pinnules small, very close, linear, with rounded apex, confluent 
at the base. This segmentation rather seldom fully developed; representing usually 
an intermediate stage between bi- and tripinnate division with oblongly triangular, 
more or less lobed segments. Pinnules sometimes a little rough. Venation not dis- 
tinguishable in unmacerated fronds; alethopteridic. ' Consistence thick and firm. 
Epidermis with isodiametrically polygonal, at times somewhat oblong cells, with 
thick, straight walls. Stomata on both sides, though rather few in number on the 
upper one; surrounded by 4 to 6 regular subsidiary cells with cuticular lobes all but 
closing the entrance; guardian cells sunk below the epidermis. 
The genus Lepidopteris, the principal representative of which is this very spe- 
cies, was instituted in 1869 by SCHIMPER (1869, p. 372) for some fossil plants from 
the Keuper, characterized, as he thought, and as the name indicates, by their scaly 
rachises. Already long before different writers had pointed out that the plants in 
question were very characteristic on account of that supposed scaliness, and that 
nothing similar was ever found in other fossils. They seem, however, to have at- 
tached no systematic value to this fact, but the fossils were classified with those 
genera with which they showed the greatest habitual conformity: Aspidoides JAEG., 
Pecopteris BRNGN., Aspidites GÖPP., and Alethopteris STERNB. It is NATHORST (1886, 
pl. 26, figs. 8—10) who first speaks of tubercles in the epidermis in these plants 
instead of scales, and thus gives the right explanation of the characteristic struc- 
tures. Recently GOTHAN (1909, No. 109) also remarked that he had not been 
able to find the smallest vestige of scales in either one or the other species. The 
explanation of the tubercles which he gave, he has, however, according to a kind 
communication by Professor NATHORST, later on given up (in a letter to Professor 
NATHORST). Finally, ZEILLER (1911, p. 3) has adopted NATHORST's explanation after 
examining Lepidopteris stuttgardiensis (JAEG.) SCHIMP. 
As will appear from the drawings and the photographs (pl. 1; pl. 2, figs. 5—38), 
tubercles occur on the rachises of the first order as well as on those of the second 
order. As their formation evidently requires that the rachis has reached a certain 
breadth, they are always, or almost always, missing from a rachis of higher order. 
At times the lamina has a touch of roughness, too. A narrow rachis has only one 
row of tubercles. It has a strange, articulated appearance, caused by the cireumstance 
that sharp folds were formed, when the tubercles were pressed down. . On a rachis 
of greater width there are several rows of similar tubercles parallel with-each other. 
Both with regard to the size and the form of the tubercles, there prevails a 
certain difference between the two sides. On one side (pl. 2, fig. 8), probably the 
upper one, they are rather large, rounded, or more or less oblong, with the greatest 
diameter in the transverse direction, while those on the other side (pl. 2, fig. 7) are 
circular, smaller and therefore more numerous. Between the large tubercles small 
