ORGANIZATION OF THE FOSSIL PLANTS OF THE COAL-MEASURES. 749 
There is every reason to believe that the tracheal elements of the primary wood 
were all tracheides and not true vessels. No signs of transverse walls are ever 
observed, while sometimes we find the very tapering ends of the elements. Their 
length must have been great, as terminations are by no means frequently seen. 
The trachez of the interior of the wood are always pitted. Where they are in 
contact with one another the pits, which are very numerous and hexagonal in outline, 
are most distinctly bordered, as shown in fig. 25. The borders of the pits, which 
have not been described before, can only be seen in well-preserved specimens ; in 
many cases the borders have perished, leaving behind an apparently reticulate 
sculpturing. On the other hand really simple pits occurred on the surface of contact 
between trachez and parenchyma. 
The most important point, however, connected with the primary wood is the 
structure of its peripheral region, for it is here that a clear analogy with Lygino- 
dendron shows itself. 
The “coalesced clusters of small vessels which occupy the peripheral portion of the 
vasculo-cellular medullary axis” were described in the first memoir on Heterangium, 
and their continuity with the leaf-trace bundles was also observed.* 
We have paid special attention to the structure of these peripheral “clusters ” 
of trachez, which in many of the specimens are very distinct. A good typical example 
is shown, in transverse section, in Plate 26, fig. 24. The close resemblance to a 
perimedullary xylem-strand of Lyginodendron becomes evident if the figure cited be 
compared with Plate 21, fig. 2. In Heterangium, as in the former genus, the smallest 
elements of the xylem-strand are situated in its interior, though at no great 
distance from its peripheral surface. In both genera these elements are accompanied 
by parenchymatous cells. Indications of the spiral thickening of the smallest 
tracheides can sometimes be observed even in the transverse sections (see fig. 24, pz.). 
To determine their true nature, however, recourse must be had to the corresponding 
radial sections, such as the section represented in fig. 25, which should be compared 
with Plate 22, fig. 4, from Lyginodendron. Here we see that the smallest tracheides 
have a loosely spiral thickening and evidently constitute the protoxylem of the strand. 
The elements to the exterior of the protoxylem («’.), which form the centrifugal 
part of the primary wood, are more densely spiral. In other cases they are somewhat 
reticulate. They are quite distinct from the secondary xylem, which in the part 
figured is represented by a ray (x*r.). On the interior of the protoxylem we tind 
some parenchyma and then the large pitted tracheides (a.) of the centripetal wood. 
The same structure has been found in all the well-preserved specimens, so that 
there can be no doubt that the peripheral strands of xylem are ‘“mesarch” in 
the sense defined above (p. 713), and so far resemble the xylem of the foliar bundles 
* Wixuiamson, “ Organization,” Part IV., p. 401. 
