i 74 M. Clievely Rayner. 
of plants in the partially sterilized soil which he used is easily 
interpreted in the light of recent work on soil sterilization. 
It is still only possible to speculate on the evolutionary history 
of the mycorhiza habit in plants. The facts, as they are known at 
present, fit into such a scheme as the following. The roots of 
vascular plants grow ordinarily in the soil—a medium teeming with 
micro-organisms. There is evidence that bacteria, lower algal 
forms, and fungi tend to aggregate about these roots,—owing, 
possibly, to the excretion of minute quantities of organic material. 
This chemotactic influence may operate more effectively in some 
cases than in others and determine the entry of mycelium into the 
roots and, hence, the occurrence of plants possessing endotrophic 
mycorhiza. Whatever the ultimate cause, it is certain that internal 
infection of the root-cells by fungal hyphae is a common and wide¬ 
spread phenomenon among flowering plants. In most of these 
cases the condition is limited to the formation of mycorhiza, and 
the mycelium is rigidly excluded from the chlorophyllous tissues. 
In the simpler cases described by Gallaud, there is evidence of 
exchange of nutritive material with a balance of profit ultimately on 
the side of the plant. The fungus invades the root as a parasite, but 
the resistance of the root-cells to invasion is such that the parasitic 
habit is held in check and the mycelium grows for a time as a kind 
of internal saprophyte in the root-tissues. Ultimately, by digestion 
of the invading mycelium (mycolysis), the plant recovers the material 
of which it has been robbed and may profit otherwise to a variable 
extent depending on the metabolic activities of the fungus. It is 
possible, for instance, that in some cases the endophyte can utilize 
atmospheric nitrogen, whereas in others, e.g., the orchids, this 
power seems to be absent (see Burgeff, loc. cit.). It is not very clear 
that the individual mycelium can profit in the long run, but it is 
sheltered temporarily from competition and from desiccation, and 
parts survive which are set free by the eventual decay of the roots. 
In certain plants, e.g., the orchids and Ericaceae, such relations have 
become more specialized and more intimate, and have resulted in 
the condition of obligate symbiosis described—obligate for the plant, 
but not for the fungal partner. 
The problem of ensuring infection has been solved very 
differently in the two groups, and each has its attendant risks and 
advantages. The evolutionary history must have been long and 
complex in both, and cases representing side-lines of development 
might he expected to occur. The case of Gastrodia among the 
orchids may he of such a kind and others will doubtless be brought 
