34 2 Marshall Ward. — On a lily -disease. 
which is placed on the cover-slip and examined with the 
microscope : if it only contains one spore, it is selected — if it 
contains more than one, the whole is rejected, and a new 
cover-slip and drop taken, and so on. Having obtained a 
satisfactory drop of water with one spore, I then placed in the 
drop also a thin section of either the ovary from a young 
lily-bud, or of the central scales of a lily-bulb, cut with a 
perfectly clean razor. Such sections may be cut quite clean 
and free from foreign spores, etc., if care is taken in removing 
the outer coverings, and I found that such sections of the 
bulb shaken up in freshly distilled water to remove some of 
the starch-grains could be kept clean in the cultures for more 
than a week. 
The most interesting results were obtained from such 
cultures. The pieces of bulb yielded to the water sufficient 
nutriment to start the germinating fungus-spore, and a normal 
mycelium was generally obtained on the third day. 
In such cultures I have over and over again traced the 
hyphae growing across the field until their tips reach the 
piece of bulb, and observed that as growth proceeds the thin 
cellulose-walls of the bulb become swollen and evidently 
softened. I have also over and over again watched the tips 
of the hyphae enter into the substance of the cellulose-walls, 
and continue their growth in the plane of what would be the 
‘middle lamella 5 if such could be distinguished in these very 
thin walls. Moreover, such hyphae occasionally bore through 
from cell to cell, as shown in Figs. 57 and 58, thus placing 
beyond all cavil the significance of these observations. To 
describe a concrete case : — In Fig. 58 the hypha was observed 
to gradually approach the edge of the section of lily-bulb, 
and to come vertically in contact with the cell-wall, figured 
at a , at 2.55 p.m. ; the onward growth of the hypha continued, 
pressing the tip against the surface of the wall and deflecting 
it slightly, as seen in c, and d. The progress of this 
mechanical effect was quite visible at intervals of two or three 
minutes, and in fact b was drawn at 2.58 p.m. ; c at 3.0 p.m. ; 
and d at 3.2 p.m., which of course implies that the changes 
