DESCRIPTION OF ADDITIONAL SPECIMENS. 300 
rather obscure on the surface of the cortex, and may be entirely subcor- 
tical. , 
Concerning the cicatricial traces within the leaf scar itself, little can 
yet be said. What appears to be a vascular trace is observable in many 
instances; also certain obscure depressions in the bolster, which occupy 
the position of the respiratory appendages at the base of the leaf, seem 
to be present, but I am far from certain that these appearances may not 
really be due to accident or other causes. 
In the slab, Museum register 5636, photographed in plate 22, figure 1, 
we have a large forked segment showing on the left the full width of the 
branch, the cast of the lower portion of which is still in place. The 
upper part or impression is the mold or counterpart of the fragment, 
plate 21, figure 4, just discussed. The similarity of the preservation on 
the lower left to that found in the lower right on the same slab is at once 
apparent. ‘The middle portion of the branch on the right presents, how- 
ever, the same characters as the lower portion of the other branch. In 
fact, we have at once on this specimen impressions of the large central 
convex boss of the type originally described as Lepidodendron cyclostigma, 
the quadrangular compressed bolsters, and the flattened bosses showing 
distinctly the details noted on the surface of the preserved stem. I am. 
not absolutely certain whether in this slab we have a dichotomizing 
stem or trunk or whether possibly two trunks are superimposed. The 
facts that the bolsters below are in accordance as to direction ; that those 
on the right of the angle change the direction of curve, as is natural at a 
dichotomy, and that I find no intercalated or separative zone or mate- 
rial, save numerous plications of the cortical tissues, make it seem most 
probable that the two branches are in actual union. Such plications are 
usually found in the angle of compressed Lepidodendroid stems, and they 
are especially to be looked for in those in which the cortical tissues are 
evidently spongy, and therefore subject to displacement under pressure. 
The above notes cover the essential details of the species, so far as I 
can discern them in the compressed specimens before me; but mention 
should be made of a number of other peculiarities in this singular if 
not problematic tree. 
To illustrate one of these I have partially represented in plate 21, 
figure 5, an impression or a mold to which the epidermis adheres. The 
margins of the outer boss appear to come nearly in contact with the 
margins of the upper part of the bolsters (although the latter can easily 
be traced to the apex) before curving inward and slightly downward, 
while becoming obscure, to meet the oval boss a little below its apex, 
thus producing a somewhat cordate effect. This aspect of the bolster 
and bosses is suprisingly like that figured as the type of Lepidophloyos 
