174 W. T. GORDON 
mixed pith divides into two large masses connected by a thin neck, or, in other words, 
becomes figure-of-eight-shaped, the outer xylem still remaining elliptical. As a result 
of this, the large elements previously noted are found towards the outside of the narrow 
neck connecting the two masses of the mixed pith (Pl. I. figs. 4 and 5). The pith then 
divides into two separate masses and the outer xylem becomes figure-of-eight-shaped, 
with large xylem elements in the constriction of the eight (Pl. I. fig. 6); and finally 
the two parts separate. The division is into two equal parts, z.e. it is an equal 
dichotomy. During the division there is no appearance of a branch gap, 2.e. the 
departure of the branch is protostelic. Thus all emissions from the stem, whether of 
petiole, root, or branch-trace, have proved to be protostelic. ‘The small solid branch— 
if it be a branch—is of interest because it is quite similar to a case noted in Diplolabis 
rémert (Solms), where the inner wood almost entirely disappeared at a bifurcation of 
the stem.* Such a reduced branch also might be expected to show primitive character- 
istics and to give some indication of the race from which the plant had sprung. The 
evidence from this specimen points to an ancestor with a solid stele, and this is in 
harmony with the evidence from a study of the petiole. 
HisroLoGy oF THE PETIOLE. 
It is generally admitted that, at the junction of the petiole with the stem, ancestral 
characters may be expected, so it is very important that the changes in the petiole-trace 
near the base should be carefully noted. In making a minute examination of the xylem 
tissue in this region, it will be found more convenient to work up the petiole, and not 
down as in the general description. 
The first figure to be noted is Pl. II. fig. 16, which represents a transverse section — 
of the stem in a rather flattened condition. At a a short arm of parenchymatous tissue 
may be seen stretching from the central mixed pith into the outer xylem. Near the 
end of this radial arm are some rather small elements which constitute one of the 
protoxylem groups of a petiole-trace. The other protoxylem group is not differentiated 
until later, so that, at this early stage, the difference in phase of the two ends of the 
petiole-trace is quite marked. In the next section of the series (PI. II. fig. 15) the 
arm @ is much longer, and the protoxylem group (p7'a') is at the end of the arm furthest 
from the inner zone of mixed tracheides and parenchyma. ‘There is still no sign of 
the second protoxylem group. Fig. 14, which follows, shows one group of protoxylem 
elements isolated in the outer xylem zone, the arm of parenchyma and its accompanying 
small tracheides having disappeared. On the inner margin of the outer wood, a short 
distance round from where the first medullary arm appeared, a second small sinus 
containing protoxylem may be seen at a’. The next section, represented by fig. 13, 
shows the first protoxylem group (p7z') well out in the outer zone and the second 
group (pra) quite distinct. 
In Pl. I. fig. 12 both groups are clearly seen, and in the next two sections (figs. 11 
* Gorvon, “On Diplolabis rémeri (Solms),” Trans. Roy. Soc. Hdin., vol. xlvii. p. 720, 1911. 
