Lycopodiales. 15 1 
size except Lepidodendron Harcourtii (where its occurrence is 
doubtful) and Bothrodendron munduin (14); secondary cortical 
tissues are present in all known species. The smaller species 
without secondary xylem may prove to be portions of bigger ones, or 
they may he really species without secondary wood. In Lepidodendron 
fuliginosum, a species so like Lepidodendron Harcourtii that the 
two were long confused, a small amount of secondary xylem was 
produced irregularly by an anomalous cambium. Dr. Scott points 
out that we may regard this species as exhibiting a primitive and 
rudimentary or a reduced form of secondary growth (14). But it is by 
no means certain that some forms of Lepidodendron Harcourtii itself 
had not some secondary xylem. Mr. Seward and Mr. Hill have 
written a paper on a Lepidodendron from Dalmeny, which they 
regard as probably identical with L. Harcourtii. They bring 
forward only fairly good arguments in support of this suggestion 
(20), but the geological horizons agree well. The Dalmeny stem, 
which whether or no it be specifically identical with L. Harcourtii, 
is clearly identical with another species, L. Wunschianum, has 
secondary xylem, and though the question must remain open, it is 
possible that a certain amount of secondary xylem was usually 
found in L. Harcourtii, but that its formation began very late, 
many examples being overtaken by fossilization before their 
primary structure had been modified. In any cases it seems not 
unlikely that the absence or the rudimentary development of 
secondary wood in L. Harcourtii is a primitive character. 
As regards the external markings of the stem, characters on 
which the division of the order into genera is founded, the scars 
left by the fall of the leaf are practically identical throughout the 
order, and consist of a mark of the vascular bundle, of the marks 
of the parichnos (two parenchymatous strands on each side of it), 
and of the scar of the ligule above the bundle. 
In Bothrodendron the scars are flush with the surface, but in 
Lepidodendron, Lepidophloios and Sigillaria, they are raised on 
cushions on varying forms. Further, there is Mr. Kidston’s genus 
of Archaeosigillaria, in which the scar of the leaf resembles that of 
Lepidodendron, except that the presence of a ligule has not as yet 
been demonstrated, and that there is no parichnos or leaf- 
cushion (8). It would seem natural to regard the various forms 
in which the leaves are seated on cushions as having originated 
independently of one another from smooth-stemmed types such as 
Bothrodendron and Archaeosigillaria, both relatively ancient genera 
