822 W. T. GORDON ON RHETINANGIUM ARBERI, A NEW GENUS OF 
similar when viewed in transverse section. In each case we find a mixture of tracheidal 
and parenchymatous tissue in the axis, while secretory cells are also seen in the paren- 
chyma. In longitudinal section, on the contrary, marked differences are at once visible. 
It is true that the peripheral groups of scalariform or sub-spiral elements are similar in 
position in each, but here the similarity ends. The remainder of the tracheidal tissue 
in Megaloxylon consists of short, broad elements with multiseriate bordered pits on 
their walls, while the secretory cells are also short. In Rhetinangium the tracheides 
are long, reticulately thickened, elements, and the secretory bodies are either short cells 
or long ducts (Pl. III. fig. 25, s.c. and s.s.). The distinction between the primary wood 
in these two genera, then, consists essentially in the type of tracheide. 
The secondary xylem is of a similar type in each case, but in the new genus it 
appears to be less compact than in Megaloxylon. An interesting point is the clearly 
marked ring of narrow elements in the secondary wood. Professor Sewarp records the 
sporadic occurrence of such tracheides in a similar position in his genus. 
Although the leaf-trace in Professor Sewarp’s plant is not known outside the zone of 
secondary xylem, yet its resemblance, in that area, to the stem of Heterangiwm (except 
for the exarch xylem) is not without interest, for the trace in Rhetinangium under the 
same conditions would have a similar form. In PI. I. fig. 2, C, the departing trace in 
Rhetinangium is merely a group of xylem strands, with conjunctive parenchyma as in the 
stem of Heterangium, and there cannot be any doubt that if the two ends of the trace 
B in that figure or in those shown in PI. III. figs. 16, 17, and 18, were bent inwards 
until they met, it would also form a leaf-trace similar to the stem of Heterangivum. 
(The position of the protoxylem groups, of course, is different.) It is just possible, 
also, that the breadth of the zone of secondary xylem and its compact nature in Mega- 
loxylon, may account for the rounding off of the petiole-trace in that genus, while it is 
traversing this tissue, and that when the trace became free it would flatten out in a 
similar fashion to the trace in Rhetenangium. In both forms several adjacent exarch 
strands coalesce to produce the trace, the protoxylem groups being situated abaxially. 
All these structural resemblances in the two stems show that they were probably 
related, but Megaloxylon appears to be the more specialised type. The peculiar short 
tracheides in the primary wood of the latter are, no doubt, as Professor SewaRp has 
pointed out, an adaptation for water-storage. 
The affinities of Rhetinangium with other members of the Pteridospermex are not 
so obvious. ‘The type of the primary xylem cylinder and the monostelic character of 
the stem, at first sight, suggest affinities with Heterangium, but the detailed structure 
of the stele and also of the petiole-trace proves that the resemblances are more apparent 
than real. In the primary vascular cylinder of the new genus the tracheidal groups are 
large and few in number, while in Heterangium (PI. III. fig. 22) the reverse is the case. 
Then, again, the exarch xylem and the numerous secretory elements in Rhetonangvwm 
at once distinguish it from Heterangium. It must be borne in mind, however, that 
certain species of the latter genus have protoxylem elements situated much nearer the 
