Holden. — Some Fossil Plants from Eastern Canada. 245 
A considerable number of Tylodendron pith casts and specimens of 
petrified wood were collected. All the wood from Gallas Point, St. Peter’s, 
Holland Cove, and most of that from the vicinity of the Sable River, proved 
to be of different species of Cordaites . Certain specimens from Sable River, 
however, presented slightly different features, and will be referred to later. 
Some of the pith casts presenting the characteristics of Tylodendron , 
Weiss, and Tylodendron Baini, Dawson, are shown in PL XXII, Figs. 1-9. 
Figs. 1-4 represent a particularly good specimen. It shows one of the 
swellings which recur at regular intervals, probably marking the nodes, and 
also the rhomboidal scars which completely cover the surface. These scars are 
of variable length, being considerably longer in the internodes than at the 
nodes. In the more highly magnified views (Figs. 3 and 4), it may be seen 
that the lower half of each rhombic area is characterized by a slit. This 
cast was found near Cannoe Cove, almost entirely buried in the mud. Figs. 5 
and 6 represent a similar specimen from the same locality. Here, too, the 
cleft ridges are well marked, but there is less of a nodal swelling. The 
upper part of Fig. 6 shows the pith to be crossed by transverse diaphragms 
similar to those of Sternhergia . Figs. 7-9 are different magnifications of 
another specimen. This specimen includes the region of the branch whorls, 
and also demonstrates the presence of diaphragms. Fig. 10 shows a petrified 
stem. This piece is of especial value, since it contains the pith, and also 
has a wound cap. 
The nature of such stems was for a long time misunderstood. Weiss, 1 
who originally defined the genus, considered them branches denuded of 
leaves. The rhombic areolae he took for the scars left by leaf-bases, and 
the split he believed to be caused by a resin canal. Potonie, 2 however, 
showed them to be pith casts, bearing the same relation to Tylodendron as 
Sternhergia to Cordaites. The surface configuration he showed to be caused 
by the course of the protoxylem strands. The material from Prince Edward 
Island confirms these statements in detail. PI. XXIII, Fig. 13, a transverse 
section of the stem shown in Fig. io, shows the instanding protoxylems, 
causing the ridges and furrows on the stem. The manner of exit of the 
leaf-trace is elucidated by Figs. 15 and 16, cut from the same stem. When 
a protoxylem strand is going to the supply of a leaf, it starts from the 
furrow at the base of one of the rhombic scars. At first, it pursues a course 
which is chiefly upward and only slightly outward. Accordingly, it splits 
the scar, which does not become entire until about half-way up, when the 
trace turns and passes directly out. Fig. 15 shows one of these strands 
during its upward journey in the act of splitting the ridge ; Fig. 16 shows 
it at the point where it begins to pass out. That its path is almost exactly 
1 Weiss : Die fossile Flora der jiingsten Steinkohlenformation it. des Rothliegenden im Saar- 
Rheingebiete, p. 182. 
2 Potonie, H. : Ueber die fossile Pflanzen-Gattung Tylodendron. Jahrb. der Konigl. Preuss. 
Geol. Landesanstalt, 1887, p. 311. 
