594 
Journal of Agricultural Research 
Vol. XXVI, No. 
A study of Table V shows that there is but little difference between 
forms III and XIX in their effect upon the stomata. On the whole, 
form III produces a slightly stronger reaction. This is particulerly no¬ 
ticeable in Baart, where only 8 per cent of the stomata show any trace of 
the reaction with form XIX even after eight days, and nearly half of the 
stomata are affected with form III after six or seven days. The differ¬ 
ence between the two rusts on the stomata of the other hosts is scarcely 
perceptible, although the two rusts differ markedly in their ability to 
grow upon these hosts. The effect of these two rusts upon the stomata 
bears no obvious relation to immunity or susceptibility and would appear 
to be independent of it. 
On comparing the effect upon the three hosts, however, marked differ¬ 
ences are seen. With both forms the effect on the stomata is least in 
Baart, intermediate in Kanred, and greatest in Mindum. So far as the 
microscope reveals, the effect is the same in kind on the three hosts, 
but differs in degree. This difference can be attributed only to differ¬ 
ences in the hosts. 
It seems fairly certain that the appressoria secrete some substance 
which penetrates the guard cells. It is natural to suppose that this may 
affect the entry of the fungus, but no such relationship is obvious from 
the data at hand. The effect on the stomata is least in Baart with form 
XIX and greatest in Mindum with form III; yet the percentage of en¬ 
tries in the two cases is about the same. Of form XIX, 75 per cent have 
entered Baart four days after inoculation, although in that material there 
were no visible effects upon the guard cells. Of form III on Mindum, 
72 per cent have entered at the end of four days, and in that material only 
8 per cent of the guard cells are unharmed, 58 per cent are partly killed, 
and 34 per cent are dead. 
The percentage of entries in the various cases is decidedly uneven. 
Take, for example, the two lots of Baart inoculated with form XIX. 
The two sets of seedlings were grown side by side in the greenhouse in 
December, 1920, given parallel treatment, and both were inoculated on a 
dark, rainy day; yet, for some reason, one showed 75 per cent of entries 
and the other, 30. Moreover, as already pointed out, different parts of 
the same leaf may vary in this respect. The percentage of entries in 
Kanred also varies, but within much narrower limits, and the entries are 
always markedly fewer than on either of the other hosts. 
DISCUSSION 
It has been shown that changes occur in the guard cells of stomata oc¬ 
cupied by appressoria. The wall just below the appressorium is first 
affected, then the adjoining cell contents, then the inner walls of the 
guard cells just below the appressorium, then the ends of the guard cells, 
and sometimes even the nearer parts of cells beyond. This process sug¬ 
gests that the fungus is secreting some substance or substances, perhaps 
enzymatic in character, which diffuse into the host tissue, even spreading 
into neighboring cells. 
It is possible that only one substance is secreted and that it first softens 
the walls and then kills the cell contents. Brown (9) has shown that a 
single enzym secreted by the germinating spores of Botrytis dissolves the 
cellulose walls of its host and then kills the cell contents. It has no action, 
however, on the cuticle. Only when the germ tube has ruptured the 
cuticle by mechanical pressure is the enzym able to produce results. 
