68 
M. Cheveley Rayner. 
with them, or are specific to the plant, apart from its symbiotic 
relations with any other organism. 
It has already been mentioned (page 64), that the roots of 
seedlings when germinated are always infected with mycorhizal 
fungus. 
Microscopic examination of young seedlings, protected against 
infection from without during germination, shows that emergence 
of the radicle is accompanied by growth of fungal hyphae from the 
testa. Some of these hyphae grow across to the root and infect the 
seedling root, most commonly, via the cells of the root cap. This 
mode of infection appears to be very regular and characteristic. 
Bacteria are also associated with the seed-coat, but these are not 
conspicuous in seeds germinated under healthy conditions. 
Infection of the seed-coat by fungi and bacteria occurs com¬ 
monly in many seeds, such infection taking place after the seeds 
have been exposed to the air. 1 In the case of Calluna seeds, in 
addition to this superficial infection, a fungus is present, which is 
identical morphologically with the mycorhizal fungus present in 
healthy roots. 
So far as could be ascertained from examination of fresh 
material, endosperm and embryo are free from infection in the 
resting seed. Microtome sectons of the seed pointed to infection 
of the testa only, and therefore to the possibility of sterilizing the 
seeds. The delayed infection observable in heated seeds (page 64) 
also supported this view, which has now been confirmed as 
correct, by germinating seedlings in a sterile condition, after 
adequate sterilization of the seed-coat. 
It remains to enquire the source of infection and at what stage 
of seed development it takes place. 
Examination of the ovary of the flower after fertilization reveals 
the fact that, in common with other parts of the flower the ovary 
wall is infected with fungal hyphae. 
Infection of the young seed-coat takes place from this source 
while the seeds are still enclosed in the ovary. 
A full account of the details of this infection of the seed-coat, 
and of the mode of infection of the flower—whether from without 
at the period of flowering or from within by a more widespread 
distribution of the mycorhizal fungus in the plant than has been 
commonly assumed—is reserved for a subsequent paper. 
Fungal infection of the testa is difficult to observe in the 
1 Arcichovskij, V. “ L'eber die Methoden zur Gewinnung mikroor- 
ganismenfreier Saincn.” Centralb. f. Bakt., Bd. 36, 1912, Nos. 15—18. 
