172 M . Chevely Rayner. 
tissue. In another region there is accumulation of food-material 
brought in by the fungus and the mycelium is ultimately digested. 
The production of flowers and the nutrition of the offsets is entirely 
dependent upon this and the plant must be regarded as completely 
parasitic on the fungus during part of its life-history. 
4. The fungus is entirely unmodified by the association and 
produces normal fructifications, a condition unique among fungi 
forming endotrophic mycorhiza. 
5. The use of the term ‘mycorhiza’ must be extended if it is 
to cover a case of this kind, in which an association is formed 
between a fungus mycelium and the shoot -tissues of a flowering 
plant. 
Kusano suggests that in Gastrodia there is represented the 
first step in the formation of a special mode of parasitic life of the 
fungus under special circumstances, which gives rise ultimately to 
a reciprocal exchange of food-material of benefit to the host. 
He considers that in the evolution of mycorhiza, the Gastrodia 
arrangement is probably a primitive stage. His conclusions in this 
respect are, however, open to criticism in view of the extreme 
specialization and reduction of the plant. They are indeed difficult 
to maintain after his demonstration that the role of * host ’ is played 
by the fungus and not by the tuber. 
Nothing is at present known of the behaviour of the seeds of 
this remarkable orchid at germination. In view of the dependance 
of the majority of orchid seeds upon infection at a critical stage of 
germination and of the observed facts of vegetative infection of the 
Gastrodia tuber the results of experimental seed cultures will be 
awaited with particular interest. 
Much has been, of necessity, omitted from this review of recent 
work, and much yet remains to be learned from experimental 
research. The view of Frank—that special relations might be 
expected to exist in certain families of plants—has proved to be 
well-founded. Since these relations, though equally complex, are 
strikingly different in the two groups for which they have been 
demonstrated, they have probably been independently evolved and 
present different physiological problems. Demonstration of an 
obligate relation between plant and fungus in the orchids and in 
Ericaceae introduces a new conception of the physiological relation 
and it is evident that the work of the past fifteen years has made it 
possible to define the term symbiosis, as used by Frank, with 
greater precision. Those cases, for which more specialized 
relations have been demonstrated experimentally, may be regarded 
