186 
C. M. Gibson. 
found already dead and shrivelled, and this is very frequently the 
case on the third day. But however long they keep up the battle, 
or however vigorous they appear at first, the hyphae invariably lose 
in the struggle for existence. In all the preparations of these 
infections that I have examined I have never found a single 
haustorium. 
The facts seem to suggest that the death of the entering 
hyphae is not due so much to starvation as to some poisonous 
substance emitted by the cells ; for on examining an epidermis 
freshly stripped from a leaf inoculated three days before, the hyphae 
outside are seen to be alive and unshrunken while a great number 
of those inside are already dead and shrivelled. Also in microtome 
sections of the same date after inoculation the swellings and hyphae 
inside the leaf were often shrivelled and empty, while the hyphae 
outside were still full of protoplasm (Fig 20). Sowings of 
uredospores from Chrysanthemum were also made in water to 
compare with those made on the leaf and after three days fixed and 
stained; it was difficult to keep the culture any longer than this as 
it became so over-run with Penicillium. In the case of those 
examined the protoplasm had disappeared from most of the hypha, 
being confined to a short distance behind the tip; in most cases 
however it had not degenerated, and had preserved its original 
structure, although the nuclei, when seen at all, often shewed signs 
of degeneration. There was certainly not as much shrivelling as is 
often seen in the hyphae inside the plant. The whole question 
is however difficult to decide, as the hyphae seem to vary in their 
powers of resistance, dying and vigorous ones being found in the 
same infection spot (Figs. 21 and 22). The leaves also seem to vary 
in their power to kill, one set of sections containing only dead and 
shrivelled hyphae and the next set of the same age containing 
mostly living ones. 
Although the observations recorded above were for the most 
part made on stained microtome sections of hardened material, 
parallel observations were made on the living leaves by stripping 
off the epidermis and examining its under surface. In all cases 
in which the result is given as positive the germ-tubes have been 
seen projecting through the stomata and forming beneath them 
swellings shewing the characteristic colour of the protoplasm of the 
particular spore used for infection in that case. All substomatal 
swellings not shewing the characteristic colour were ignored. 
From the above results it is evident that infection is a much 
more complicated matter than the mere entry of the stoma by the 
