608 Osborn. — Lateral Roots of Amyelon radicans, Will ., 
occurred (cf. Weiss). Again, the hyphae are confined to one region of the 
roots, namely to the inner cortex. Traces of the hyphae are, it is true, to 
be found in the epidermis and the outer cortex, but they are comparatively 
rare. The fungal filaments, when seen there, occur singly, and appear to be 
running from the exterior inwards, or longitudinally with the root. There 
are, as it has been pointed out, no dilations or branchings in this region. 
Moreover, I have never observed them in the phloem region or in the 
wood. It should also be remembered that they are absent from all the 
tissues of the main roots that have undergone secondary thickening. 
The possibility of parasitism has also to be considered. Oliver (10) 
has described parasitic fungi from the Palaeozoic era. The argument of 
restriction of infection, cited against saprophytism, also holds here. A para- 
sitic fungus might be expected either to ramify over the whole of the tissues, 
or to attack such a region of soft tissue as the bast, as in the case of 
Armellaria melleus , to take a common instance. This, however, is not the 
case. 
The constancy with which the fungus appears is also against its 
being of the nature of a disease. In all cases where the preservation is 
sufficiently good in material, from both different blocks and different 
localities, hyphae are to be observed. It is hardly to be expected that 
a fungus would have so widespread an occurrence, unless there were 
a great epidemic, or else some vital reason for its presence. 
This last explanation, that of a symbiosis, seems to be the most 
probable relationship. First the fact, that the root has the tendency to 
branch so frequently forming definite tubercles, is strongly reminiscent 
of Myrica , Alnus , &c. The division of the cortex into infected and non- 
infected regions, and the occurrence of knots of hyphae in some only of the 
cells, reminds one of similar cases described by Frank (2), Warlich (20), 
Groom (3), Werner Magnus (6), Shibata (15), &c. Weiss, in describing 
Mycorhizonmm , has commented upon the frequency with which some form 
of symbiotic union is to be met with among recent plants ; instancing Janse’s 
paper (5) in the Buitenzorg ‘ Annales ’. The structure described by Weiss 
is, however, the only recorded case among Palaeozoic plants. 
In the case of recent Gymnosperms some union of the spermophyte 
with a thallophyte is of common occurrence. The case of Cycas is well 
known. Dacrydium and Podocarpus have tubercles, the Coniferae generally 
have an ectotropic mycorhiza. It need not, therefore, be a matter of 
surprise that their Palaeozoic ancestors should have had a similar relationship. 
The very interesting question of the biologic conditions under which 
Cordaites lived appears to be somewhat elucidated by these observations. 
It has been pointed out that in some respects the fungus has a resemblance 
to that found in Podocarpus. There is, however, no formation of nodules^ 
as in the latter plant, but rather a production of numerous short and much- 
