DETAILED DESCRIPTION OF LEPIDODENDRON CYCLOSTIGMA. 331 
the central area, one on either side of the central trace. The figure of 
this specimen is so placed on the plate that the central, obscurely sub- 
triangular area comes generally at a little above the middle. Thus the 
transverse side, which is seen in many cases either to be crescentic or to 
contain a very obtuse angle, is made to constitute the base. This ar- 
rangement, which seems to conform with that of figures 1 and 2, plate 
22, is made largely for the sake of preserving the morphologic similarity 
of the bolsters of this tree with those of the conventional type of Lepi- 
dodendron as represented, for example, in L. elypeatum. I confess, how- 
ever, a lack of assurance as to the actual attitude of some of the stems. 
I have attempted to make them conform in position to other better pre- 
served fragments, the orientation of which will be discussed farther on 
In the comparison of the trunks from Missouri with the Lepidoden-. 
droid type, the reader will recall that in the typical Lepidodendron the 
surface of the cortex, which comprises the bulk of the trunk, is marked 
by rhomboid or fusiform, more or less protuberant leaf-base cushions or 
“bolsters,” usually elongated and acute longitudinally, rounded laterally, 
compactly arranged in a complicated system of spirals. Near or above 
the middle of the bolster is a transversely rhomboidal cicatrix, variable 
in size and form, the scar left by the deciduous leaf. This “leaf scar” 
contains a central and two lateral punctiform traces or cicatricules, the 
central being the cross-section of the midrib or nerve of the leaf. The 
sometimes oval, lateral traces have been shown by Potonié to represent 
cross-sections of large-celled, lacunose, parenchymatous tracts considered 
as transpiratory. The latter reach the epidermis in two oblique, oblong 
mammille or “appendages,” one on either side of the median line, just 
below the lower margin of the leaf scar. The two appendages have also 
been interpreted by various paleobotanists as glandular or other orifices 
in the leaf base. On the median line of the bolster, above the leaf scar, are 
two small deltoid depressions. The lower, close above the leaf scar, is be- 
lieved by many to correspond to the ligular attachment in Selaginella, 
while, in continuation of the homology, the other deltoid depression, in 
the apex of the bolster, has been regarded by Stur and others as corre- 
sponding to the position of the sporophyll in Selaginella and Isoetes. 
Passing now to the originals described by Lesquereux, we find that 
the ficure given in the Coal Flora represents a small portion of an irreg- 
ular fragment 27 centimeters in length and 15 centimeters in width. 
In this fragment, which, like the one described above, is a mold or 
impression, the Lepidodendroid form of the bolsters is clear. In a por- 
tion of the slab one end, presumably the lower, of the bolsters is slightly 
truncated by pressure in fossilization. The convex central areas or com- 
pressed central bosses are mostly ovate-triangular or ovate, as is shown 
