Marshall Ward. — On a lily -disease. 355 
possible to assure myself that the protoplasm of the conidium 
passed into the germ-hypha in the cell- wall, and the hypha 
branched and grew in the plane of the surface of the bud, leaf, 
etc. Moreover, transverse sections through the recent infec- 
tions show beyond all doubt the presence of the hyphae in the 
substance of the cell-wall, as seen in Figs. 55 and 56 : these 
preparations also show clearly that the cellulose in which the 
hyphae are running is swollen considerably, and it seems 
impossible to reject the explanation that this is due to the 
action of the ferment secreted by the fungus, and isolated by 
precipitation, as described, and that the change here is of the 
same nature as the swelling of the cellulose in the experiments 
referred to on pp. 348-353. 
Among other variations of the conditions of infection, I tried 
the effect of sowing the conidia in drops of Pasteur’s solution 
on the epidermis of the buds and leaves. The phenomena of 
infection were not markedly different, the chief noteworthy 
point being that the germ- hyphae gave rise to branches 
outside as well as inside the cuticle. These outside hyphae 
then repeatedly branch, and form organs of attachment which 
rapidly destroy the cuticle beneath, and enter the tissues : 
such a case is illustrated in Fig. 50, the specimen being ob- 
served from within. This exceptional mode of infection is 
peculiarly interesting, because it resembles one described by 
De Bary 1 as the only mode occurring with the mycelium 
derived from ascospores of Sclerotinia (Peziza) Sclerotiorum , 
where the mycelium is incapable of becoming parasitic unless 
it has previously been nourished as a saprophyte. Here, again, 
as will be shown later, the entrance of the hyphae into the 
tissues depends upon the excretion of a ferment which dis- 
solves cellulose. I shall, however, return to this point subse- 
quently. 
But perhaps the best and most convincing preparation 
showing the causal relation between the fungus and the spots 
was the one figured in Fig. 53. I had several times found that 
1 Bot. Zeit. 1886, Nos. 22-27. 
