114 MR. W. G. P. ELLIS ON A TRICHODERMA 
only difference in fact being that, while in the thallus in light 
the brown patches of cells produced by the infection were close 
together, making indeed a continuous though rather diffuse 
pateh, in the thallus kept in darkness such brown cells were 
more separated, owing possibly to the elongation of the thallus 
due to the absence of light and the consequent attenuation of its 
individual parts (see Pl. 6. figs. 6-8). I carefully sterilized some 
fine camel-hair brushes by boiling them for some time in a test- 
tube. I allowed the fine point to just touch the spore-cluster, 
the spores thus removed being painted on the thallus, and 
each brush boiled before again being used. This precaution of 
removing spores only is very necessary because, especially if taken 
from a culture, a certain amount of some nutrient substance might 
otherwise be taken, and, as de Bary * has shown, saprophytic 
fungi started in nutrient solutions may become parasitic. I tried 
this effect in several cases, and found that where the spores were 
mixed up with a drop of gelatine before applying them to the 
Pellia, the thallus became diseased much earlier, the infection 
therefore being hastened by this nursing up of the spores. 
The time necessary for the manifestation of the disease as & 
result of direct infection varied considerably : thus while at the 
end of May four days (26th to 30th) sufficed for the appearance 
of the characteristic brown colour of the diseased thallus, 
towards the end of June fourteen and even sixteen days were 
necessary for a similar development of the disease. Can it 
be that this fungus is specially virulent in the early days of its 
parasitic phase? The original pan seems to point to some such 
loss of vital activity, for while at an early period the rate of 
extension of the disease was rapid—the diseased area having 
increased by an inch radially between Saturday, May 30th, and 
Monday, June 1st—during July there was very little change in 
appearance, the fungus in faet seeming to have exhausted itself, 
and the host actively rejuvenating. 
My chief diffculty was the procuring of a section that would 
show the actual infection of the host. Pieces of thallus we 
carefully infected with spores only, and after the third day 1 cut 
day by day transverse sections of the thallus and horizont 
sections, removing the superficial cells so as to try to get the 
point of infection in section and in surface view. I found that 
many spores had not germinated by the fourth, eighth, and eve? 
fourteenth day, and the irregularity of germination reduced mj 
* Comp. Morph. & Biol. of Fungi, Mycetozoa, and Bacteria, p. 381. 
