62 Transactions British Mycological Society. 
fection of the kind typical to the plant. In such seedlings 
infection by hyphae from the seed coats can commonly be 
observed and similar observations have been made on seeds 
germinating viviparously within the fruit by Ternetz(3) for 
Calluna and by me for Pernettya mucronata. Moreover, it is 
often possible to observe the presence of fine hyphae upon the seed 
coats of ripe seeds taken from unopened fruits, the observation 
of which may be facilitated by the use of careful methods of 
maceration. Interovarial infection can be demonstrated in 
sections of unripe fruits with the aid of a very careful technique, 
although such demonstration of infection of seeds within the 
ovary usually involves a relatively large amount of tedious and 
painstaking work. I am of opinion that extensive infection of 
the seed coats occurs only at a late stage, and is probably 
associated with death and desiccation of the tissues composing 
the fruit-wall and the increased mycelial activity subsequent 
to this. 
In my experience, very careful work and much patience are 
needed to obtain convincing proofs of the presence of mycelium 
within the ovary and young fruit. For this reason I do not 
attach great importance to Christoph’s failure to observe it. 
I have offered elsewhere what I believe to be convincing 
evidence on this matter (loc. cit.). 
The technique adopted by me to obtain seeds free from in- 
fection is as follows. Seeds, freed from all impurities, are wetted 
by centrifuging, they are then sterilized by immersion in I % 
mercuric chloride for three minutes and, after repeated rinsings 
in sterile water, sown on agar media containing sugar and 
peptone to facilitate the detection of micro organisms if present. 
In my original experiments, seedlings from such pure cultures 
were transferred to tubes of broth and bouillon kept at various 
temperatures, in order to test in the most drastic way possible 
their freedom from micro-organisms (loc. cit.). 
Seedlings raised in this way and planted in nutrient agar 
suitable for their growth do not develop a root system and 
eventually die; if inoculated at planting from a suitable pure 
culture of the endophyte zsolated from the ovary or unopened 
fruit, they develop normally in every way and the roots show 
typical infection. 
Every link in this chain of events has been demonstrated 
experimentally and there is no escape from the following con- 
clusions: 
(a) That under the experimental conditions described de- 
velopment of the Calluna seedling is bound up with infection. 
(b) That such infection takes place regularly from the testa 
of the seed at, or subsequent to, germination. 
