158 
but yet no sound ones. I fully believed that all of tlie plants would 
be smutty, until on September 10, the lirst sound panicle appeared. On 
October 1 were counted 30 sound panicles and 140 smutty ones; and 
finally, on October 15, the proportion was 180 diseased, smutty plauts to 
70 sound ones. In plants which conceal the germ of destruction we 
find slight traces of the fungous threads only in the nodes and in the 
growing points, and in the latter they do not attain further develop¬ 
ment until the ovaries are formed. They then proceed to the formation 
of spores in this place only, not in the leaves, where they remain 
sterile and do not produce a single smut spore. The ovaries swell 
mightily with the rapid and abundant development of the fungus in 
them, and finally, like the horns of ergot, grow to be many times their 
natural size, projecting far out of the panicles. Finally, after the com¬ 
plete spore formation of' the fungus, they break up and allow the spores 
to dust away. In this stage scarcely a trace of the mycelium of the 
fungus is to be found in the host plant. 
The behavior of corn smut is directly opposed to that of the smut 
forms which inhabit the grain exclusively. This form can produce its 
smut beds on any part of the host plant, and in the strange and repul¬ 
sive similitude of cancriform swellings and ulcers. 
For infection experiments with smut germs the big corn plant is an 
ideal object. All parts of the maize, from the seedling to the inflores¬ 
cences and fruit-spikes, are developed on a large scale, and are easily ac¬ 
cessible for each form of the experiment. The corn smut itself, Ustilago 
maydis , is also a smut form especially suitable for the infection. 
C. I began infection experiments with corn smut in the spring of 1885. 
The spores of Ustilago maydis do not germinate in water, or do so very 
sparsely only after some years. In nutrient solutions they germinate 
without exception and immediately. They are therefore consigned to 
nutrient solutions or nutrient substrata, and not to mere water, for full 
germination. They produce an endless quantity of sprout conidia, and 
still more rapidly than the two forms of Ustilago previously mentioned, 
U. carlo and U. cruenta. The conidia are thrown down as a white, 
grauulous precipitate, which appears even whiter than the sediment of 
Ustilago cruenta. But in this case the sprouting of- the conidia takes 
place upon the nutrient solution, where mold-like pellicles are formed, 
from which the conidia can easily dust off through the air.* 
I. In the first series of experiments, in 1885, I infected only young 
seedlings in different stages of germination. In more than ten distinct 
sets of experiments the seedlings were copiously sprayed with conidia 
and were afterwards set out in the field. 
After 15 days, very scattering signs of smut were visible among the 
plants which had been infected iu the earliest stage of germination. 
Below, upon the axes, a smut swelling was developed, in consequence 
* This formation of sprout conidia in the air is likewise peculiar to a number of 
other Ustilaginece e. g., Ustilago bromivora and Ustilago destruens , also Tolyposporium 
junci, etc. 
