606 Osborn. — Lateral Roots of Amy elon radicans , Will., 
form, the rootlet that was to become the main one of the group pierced the 
cortex. It seems to have had its origin in the pericycle opposite one of the 
protoxylem groups. PL XLVI, Fig. i, shows the root-trace passing through 
the secondary wood from opposite one of the protoxylems of a triarch 
root. This rootlet, which may be regarded as the main one of a group, 
appears to have developed a very wide cortex zone, which is seen in the 
section ; and, in some cases, the periderm of the root is continued into the 
rootlet with probably one cork cambium. 
Branching of this rootlet was very frequent, but there is no sign of 
definite arrangement of the branches. As the rootlet bearing them was 
diarch they were probably arranged in two rows, as in the case of the 
tubercles of Podocarpus (Van Tieghem, ] 9 ) though this is not apparent 
from the sections. These branches of the main rootlet divided again almost 
immediately (PI. XLVI, Fig. i) ; possibly more than once, for there is always 
a large number of rootlets near the point of origin of a group. Once the 
tuft was produced, the individual rootlets seem to have run some distance 
(certainly some 2 or 3 cms.) without branching again. In the last part of 
their course they appear to have run in and out of the mass of debris with 
which they are found in the manner of Stigmarian rootlets. They may 
constantly be seen burrowing through fragments of stems and even older 
Amyelons. In general appearance it will have been seen that the rootlets 
are not unlike those of Stigmaria ; which resemblance Williamson com- 
mented on, both in his memoir and in his catalogue. 
The origin of the periderm is stated by Williamson to have been 
in a peculiar manner — namely, by radial elongation of the cells of the outer 
cortex, which subsequently formed numerous radial walls. This appears at 
first sight to be possible, if not probable, when an old Amyelon is examined. 
It is not, however, borne out by the younger stage. The periderm has been 
stated by Renault to have been superficial in its origin (loc. cit.). Scott, 
however, states it to be deep-seated in Amyelon radicans , and my observa- 
tions confirm this. The rootlet, when the metaxylem was formed, produced 
a cork cambium immediately outside the endodermis, as is usual in Gymno- 
sperms. The whole of the cortex would then be sloughed, the inner 
mycorhizal zone included (PL XLVI, Fig. 4). This is exactly what occurs 
in Podocarpus. 
The fungus is apparent in the cortex as hyphae, with a diameter 
of about 4 /x. Septa are either absent or very rare. In the outer cortex 
the hyphae are to be seen best in longitudinal section, when they are 
noticed running apparently from the exterior to the interior of the cell 
(PL XLVI I, Fig. 5). The inner cells may be divided, as Werner Magnus (6) 
has shown, into host cells (. Pilzwirthzellen ), and digestive cells ( V erdauungs- 
zellen). In the former the hyphae are noticed to coil around the exterior of 
the cell (PL XLVII, Fig. 6). In places the walls are dilated to about ij times 
