8 T. G. HALLE, LOVER DEVONIAN PI.ANTS FROM RÖRAGEN IN NORWAY. 



sufficiently sharp distinctions on which a subdivision into smaller specific units can 

 be founded. While the wliole material will therefore be treated as belonging to 

 one species, it has been thought convenient to distinguish a few characteristic 

 types and describe these separately. In the following description only the outward 

 features of the specimens, sueh as the general appearance and the branching of the 

 stems, the disposition and the shape of the leaves etc., will be considered. >Such 

 evidence as Ave possess regarding the structure of the axis will be dealt with under a 

 separate heading. 



A. -Specimens r e s e m b 1 i n g Dawson's t y p e-s p e c i m e n s (Nat- 

 horst, 1913, pl. 4, figs. 8, 9; this p a p e r, pl. 1, figs. 1 — 7; pl. 4, figs. 1 — 3. — 

 These specimens are stem-like impressions, measuring up to 1 cm. in breadth. Despite 

 their flattened condition it is evident, from the radial arrangement of their leaf-like projec- 

 tions that these specimens were originally cylindrical. In the following pages f urther consi- 

 deration will be given to the probable morphological natnre of these projections as well 

 as of the axis from which they arise. For the present the fossils may be spöken of as 

 leaf-bearing stems. The leaves are best seen at the outlines of the impressions where 

 they appear in profile, but they can be traced also on the surface of the impressions them- 

 selves, either as imprints in the matrix or as elevations of the mineral substance which 

 has replaced the organic tissue. It is evident that the leaves ha ve a radial arrangement, 

 but in the majority of the specimens there cannot be traced any regularity in other 

 respects: no distinct ortostiches or parastiches can be observed, but it is probable that 

 the disposition was a spiral one. In some cases, notably in the specimen in Nathorst's 

 pl. 4, fig. 9, there appears to be a tendency towards a kind of pseudo-verticillate arrange- 

 ment in some places. In the specimens figured by Nathorst the leaves form a nearly 

 right angle to the axis. They ha ve a thickened basal part, which was evidently of a coni- 

 cal shape in the natural state and passed gradually över into the stem. From this 

 thick base springs a straight narrowly linear, subulate, upper portion which is not sharply 

 defined from the basal expansion but passes insensibly över into it. The length of the 

 whole leaf is about equal to the breadth of the stem on the impression or sometimes a 

 little greater. In a fairly large number of specimens, the leaves — though in other re- 

 spects answering to the description just given — form a somewhat oblique angle to the 

 stem or are more or less strongly falcate, with the apices directed forward. Such is the 

 case, for instance, in the specimens shown in pl. 1. fig. 1, and in pl. 4, figs. 1 — 3. 



In places where the preservation is good, there can sometimes be noted a very fine 

 but distinct vein-like line running through the leaf. It springs from the surface of the 

 stem at a right angle and is especially marked in the thick basal part of the leaf; in the 

 narrow upper part it gradually disappears. Though no vascular elements can be observed, 

 this line no doubt represents a vein or vascular strand. The fact that this narrow 

 strand has caused a distinct marking in the matrix, in spite of the thickness of the leaf- 

 base in which it is embedded, indicates that the leaves were not härd and spine-like -- at 

 least not in the lower part — as might be supposed. In the specimens with falcate leaves 

 the latter usually have a distinct spiral arrangement and are often more densely placed 



