752 PROFESSOR W. C. WILLIAMSON AND DR. D. H. SCOTT ON THE 
A local increase in the thickness of the secondary wood is sometimes found near 
the bases of adventitious roots (see Plate 10, fig. 28, x”). 
The wood has the same structure as in Lyginodendron, consisting of radial rows of 
tracheides, with rays between them (see Plate 26, figs. 24 and 25). The tracheides 
have numerous bordered pits on their radial walls similar to those of the primary 
wood. It is probable that a few pits also occurred on the tangential walls, as was 
certainly the case in H. tilieoides, in which the preservation is more perfect. 
The rays consist of thin-walled cells often showing in radial section a characteristic 
muriform arrangement.* 
Fig. 25 shows in radial section the inner part of a ray at a place where the 
secondary wood is still very thin. The ray is as yet scarcely differentiated from the 
pericyclic cells by the division of which it arose. Some of the rays are continuous 
with the conjunctive parenchyma of the primary stele (see fig. 24). We may call 
these the primary or principal rays, but they do not differ essentially from those 
which abut on the tracheides of the primary wood. Secondary rays were occasionally 
intercalated, where the wood attained a sufficient thickness.t 
Secretory sacs, which often run in a radial direction, are especially frequent in 
the rays of the Dulesgate specimens. A figure in a previous memoir gives a sufficient 
idea of the appearance of the secondary wood as seen in tangential section.{ The 
rays, however, are often of greater breadth than those shown in that figure. 
Cambium and secondary phloém are never very perfectly preserved in our 
specimens of H. Grievit, though, as we shall find, their preservation in H. tiliwoides 
is perhaps more perfect than in any other known fossil. In some of the best specimens 
of H. Grievii the tabular form of the cambial cells can be made out in transverse 
section (see especially Plate 26, fig. 24, cb.; also Plate 27, figs. 26 and 29). As 
regards the phloém, the transverse sections, as shown in the figures just cited, prove 
that it consisted of thin-walled elements of smaller diameter than the adjacent cells 
of the pericycle. Some of the longitudinal sections show that the phloém-elements 
were much elongated.§ 
There is no doubt whatever, that here, as in Lyginodendron and in Heterangium 
tulwordes the cambium was a perfectly normal one, forming wood on its inner and 
bast on its outer surface. 
Where the leaf-trace bundle passes out from the stele a parenchymatous gap is 
left in the secondary wood, corresponding to the “‘trace-gap” in Lyginodendron.||__ 
_Although we sometimes find traces of tissue-formation by tangential cell-division 
* See Wintaamson, “ Orgunization,” Part 1V., Plate 29, fig. 33. 
+ As in O.N. 1915 C. 
~ Wittiamson, “ Organization,” Part IV., Plate 29, fig. 33a. 
§ See especially C.N. 1268 A. and 1915 G. and R. 
|| See Witttamson, “ Organization,” Part IV., Plate 28, fig. 30, m'”. 
