228 
PEOEESSOE W. C. WILLIAMSON ON THE OEGANIZATION 
who gradually raised the walls of his dome, but who, instead of leaving a central aper- 
ture for a circular skylight, built up his brickwork to the very keystone of the closed 
arch. So far as my specimens throw light upon this physiological feature, most 
of our Lancashire Dadoxylons have been constructed upon this type with imperforate 
pits. 
All the sections exhibit medullary rays. These are radially disposed lines of mural cells 
(tigs. 35 / 38/), which cells are often cubical, having a diameter of from -00125 to -0007. 
In the tangential section (tig. 40) we see that these cells are arranged in single vertical 
lines. The number of cells in each vertical series varies from one (fig. 40,/') to fifteen, 
such variable numbers being distributed indiscriminately over each section. Whilst 
the specimen under examination is remarkable for the preservation of the tissues just 
described, it is especially so in the case of the bark, which is unequalled by any other 
example of a Carboniferous Dadoxylon that I have met with ; at the same time some of 
its outermost tissues are most probably wanting. The entire bark, as it now exists, is 
about -05 in thickness, and is separable into two very distinct portions — an inner and 
an outer one. In the transverse section (fig. 34) the inner bark, d, appears as an indi- 
stinct tissue, in which traces of small cells, or of the orifices of oblique tubes, are visible. 
On the other hand, the outer layer (fig. 34, e ) consists of a strongly marked form of 
parenchyma composed of cells of various sizes, the majority of them being about from 
•002 to -0025. On turning to the vertical section (fig. 36) we see that the inner bark, 
d, contains numerous parallel longitudinal lines, which are evidently the thin walls of 
tubes in which I can occasionally trace transverse septa. In all probability some of 
these latter are the representatives of vasa propria. They appear to have a diameter of 
from -001 to -0006. Some of these apparent tubes may be resin-canals. Whatever 
their physiological nature, the abrupt transition from them to the outer bark, as seen 
in the same section, is very marked. The latter tissue (fig. 36, e) is composed, as I have 
just observed, of coarse parenchyma, the cells of which are arranged in irregular vertical 
lines which tend somewhat outwards as the linear rows ascend through the stem. 
Traces of divergent vascular bundles in this Dadoxylon will he considered after exa- 
mining some similar organs in other specimens. 
Fig. 41 is a vertical section of a specimen for which I am indebted to Mr. Butter- 
worth. It consists wholly of the Sternbergian medulla of a Dadoxylon, and is drawn 
of the natural size of the original. At its upper part it exhibits extremely well the 
regular arrangement and unbroken continuity across the medullary cavity of the suc- 
cessive layers of the medulla. When I published my original memoir on Sternbergia I 
was uncertain whether or not these medullary disks were perfectly continuous, or whether 
they were perforated along the centre of the medullary axis. It is now certain that 
they were continuous. This specimen is further peculiar because of the two thick 
lateral masses of undivided pith ( a , a), which, in this specimen, has evidently formed a 
rather thick layer of parenchyma lining the inner surface of the ligneous cylinder, the 
more central portion being split up into the characteristic Sternbergian disks. As 
