I] SYMBIOSIS 17 



its morphological characters. The mycelium not only extends through the 

 roots and colourless parts, but grows into the subaerial tissues of the stem, 

 leaves, ilowers and fruit. Moreover the seedling is regularly infected on 

 germination by the hyphae which have penetrated the seed coat. When 

 deprived of its fungus by sterilization of the seed, the seedling fails to develop, 

 its root system in particular being inhibited; when subsequently infected 

 from artificial culture it renews its growth. If, however, a weak seedling is 

 inoculated from a vigorous culture it is completely parasitized and destroyed ; 

 the fungus in such a case may be regarded as having escaped from the con- 

 trol of its partner. 



The fungi present in the orchids investigated by Bernard were referred 

 to the form genus Rhizoctonia^ , certain species of which have been shown to 

 be conidial stages of the basidiomycete Hypochnus. It is significant that 

 conidia are never produced on a healthy host plant, though they can be 

 obtained when the mycelium is grown in culture; in the cells of the host a 

 tangled weft of hyphae develops and ultimately undergoes digestion, forming 

 an amorphous mass. In Odontoglossum the mycelium does not enter the 

 stem, and in Vanda and its allies it is confined to the perennial roots; in 

 the Cattleyeae the roots are deciduous and their disappearance is followed 

 by an autonomous phase; in Bletilla hyacinthina, also, the mycelium never 

 invades the green, superficial rhizome, but the young roots are regularly inx 

 fected as they reach the length of a few centimetres. In winter the orchid 

 is represented by its rhizome alone and its activity at the beginning of fresh 

 growth is consequently autonomous; the symbiotic phase follows on the 

 development of the new roots and lasts about six months, covering the period 

 of the maturation of the seeds. Duringthe autonomous phase the fungus vege- 

 tates in the soil and loses, to some extent, its "virulence" or power of infection. 



Bletilla differs from the cases previously described in that its symbiosis is 

 facultative and development can take place without infection, but the seed- 

 lings grow slowly and are delicate ; inoculation by a mycelium greatly reduced 

 in virulence has little effect, the hyphae enter a few cells and are at once 

 digested ; a mycelium, on the other hand, which has attained a high degree of 

 virulence, penetrates at once to the region of attachment of the suspensor and 

 instead of undergoing premature digestion spreads through the host cells 

 while the seedling grows rapidly and its lower part swells to form a protocorm. 



The restriction of the mycelium to the non-chlorophyllous regions of 

 these orchids is not accidental, for the contents of certain of the stem cells 

 have a poisonouseffecton thefungus. It follows that here the mycelium cannot 

 penetrate into the ovary and the infection of the germinating seed is conse- 

 quently a matter of chance; hence the difficulty frequently experienced in 

 inducing the germination of orchids. 



' Burgeff, in 1909, proposed the new generic name Orcheomyces for these fungi. 

 G.-V. 2 



