754 PROFESSOR W. C. WILLIAMSON AND DR. D. H. SCOTT ON THE 
in Plate 26, fig. 21. The continuity of the cortical tissues of the branch with those of 
the main axis is evident. Although the branch shows no leaf-traces, it is certainly a 
stem-structure. It is not a root, for its origin is clearly exogenous. It is not a leaf, 
for the stele has the same structure as that of the stem, and has even formed some 
secondary wood. The structure of the cortex of the branch is identical with that of 
the main stem. We have other sections of the same specimen, one of which shows 
the branch at a point where it has just become free from the parent axis.* 
B. THe Lear. 
1. Connection between Leaf and Stem. 
Several specimens show the bases of petioles in connection with the stem. Thus, 
in the series of sections previously figured the bases of two petioles are shown.t 
The same series contains a transverse section of another petiole, which has become 
free from the stem.{ 
We have figured in Plate 26, fig. 22, a longitudinal section of a stem, passing through 
the insertions of two petioles (pt.). These examples show that the leaves were inserted 
alternately on the stem. The length of the internodes shown in fig. 22 appears 
to have been only about 5 millims. In the series first mentioned the length of the 
internodes must have been much greater, for the bases of two leaves only were met 
with in a piece of stem nearly two inches long. Fig. 22 shows that the cortex of the 
petiole contained the same horizontal sclerotic bands as that of the stem. In most 
specimens it is evident that only a single bundle entered the leaf from the stem, and 
not a double bundle, as in Lyginodendron. The only exception is one of the 
Dulesgate specimens, in which two quite distinct bundles are seen in. transverse 
section in the cortex, which appear from their position to be destined for the same 
petiole.§ 
In most respects the leaf-traces agree closely with those of Lyginodendron. 
The specimens showing petioles in connection with the stem enable us to identify 
other petioles, which have become free. Their cortical structure is quite charac- 
teristic, owing to the horizontal sclerotic plates of the inner cortex and the 
“ Dictyoaylon” structure of the outer zone. The only difficulty is the possibility of 
confusion with the petioles of Lyginodendron. Two distinctions, however, exist. 
1. The characteristic emergences, to which the petioles of Lyginodendron owed 
their former name of Rachiopteris aspera, are absent from the petioles of Heterangium 
Grievit. 
* C.N. 1885 H. 
+ Wittiamson, ‘ Organization,” Part TV., Plate 30, figs. 42 and 44, 
t ON, 1244, 
§ C.N. 1915 M. 
