92 West. — On Stigeosporiu m Marattidcearum and 
2. The host-plant (Phanerogam) derives the entire, or by far the greater, 
advantage from the association. 1 
3. The advantages of the association are entirely, or almost entirely, on 
the side of the fungal partner. 2 
B. The present investigation . 
The biological relations between the marattiacean roots and the 
endophyte which inhabits them will now be briefly considered in the light 
of previous work on mycorrhizal fungi. 
Hyphae are never found leaving the host-root. This is shown by the 
fact that few hyphae are found in the outer cortical layers, and these are 
obviously only the infecting hyphae, since they are relatively older than the 
mycelium found in the inner cortical layers of the host-root. This fact may 
be urged against the hypothesis that the fungus absorbs humus products or 
mineral salts from the surrounding soil and liberates them, with or without 
previous modification, in the tissues of the host. This also militates against 
the view that the mycelium of the endophyte takes the place of root-hairs 
in the higher plant. Moreover, the distribution of the latter bears no 
relation whatever to the presence or absence of the fungus. Again, root- 
hairs are sometimes relatively abundant on young infected roots. 
To the question whether the fungus is a facultative anaerobe capable 
of fixing free nitrogen from the air, no definite answer can be given in the 
absence of experimental evidence. 
But whilst admitting that the fungus is a true parasite, it cannot be 
said to cause any obvious injury to the plant attacked (cf. the fungus found 
in the seed and vegetative organs of Lolinm temulentum , Hanausek (24), 
Nestler (33)). Kusano (1. c.) has proved conclusively that the facultative 
parasite Armillaria mellea may behave towards the Orchid Gastrodia elata 
either as a true parasite or as an endotrophic symbiont. There is no reason 
for doubting that the host-plant utilizes part at least of the products of 
disorganization of the ‘ arbuscules ’, but there is no intimate relation between 
the endophyte and the nucleus of the mycorrhizal cell, such as Groom (1. c.) 
claims for the mycorrhiza of Thismia Aseroe. 
Now, apart from its role as habitat (and possible shelter from excess of 
oxygen) for the endophyte, the host-plant provides the latter with all, or 
at least the greater part, of the food material it requires, because the 
mycelium in the outer cortical layers of the root, which provides the only 
connexion between the living mycelium in the inner cortical layers of the 
root and the free hyphae in the soil, is usually dead and without contents, and 
therefore cannot function as a channel for the passage of food-material. More- 
over, starch invariably disappears from the invaded cells ; all stages in the 
solution of the starch granules, which do not reappear even after the complete 
collapse of the ‘arbuscules 1 (cf. Groom, 1. c., p. 348, and Kusano, 1. c., 
1 Gastrodia elata , Kusano (1. c.). 2 Gallaud (1. c.). 
