63 
U. maydis on Zca mays. The first two smut forms will answer as types 
for smut fungi living* solely in tbe grain, the last form as a type for 
smut forms which may* appear not only in the grain but also in every 
part of the host plant, from the period of its earliest youth to complete 
development. All these forms belong to the genus Ustilago , and yield by 
the cultivation of their spores in nutrient solutions an endless sprouting 
of yeast conidia as experiment material. 
The agriculturists, who listen to me, will perhaps ask themselves the 
question, Why not then make the experiments with the stone smut of 
wheat, the most important smut form upon the grains of this country ? 
To this I will reply at once that stone smut was purposely neglected for 
reasons not far to seek. The stone smut of wheat, belonging* to the 
smut genus Tilletia , is not a suitable object for experiment, first, because, 
as infective material, its conidia, cultivated in nutrient solutions and 
“ reproduced mold-like in the air ” can not be distributed so evenly and 
well upon the host plants, on account of their difficult dissemination 
and use in fluids, as can the the sprout conidia of the genus Ustilago 
which aie developed under fluids; second, because in Tilletia , especially^ 
the observation of the penetration of the germ into the host plant, and 
the further develox^ment of this within the host, is surrounded with the 
greatest difficulties on account of the extraordinary minuteness of the 
Tilletia germs and mycelia. Finally, it is yet to be added that in its 
appearance as a parasite, in the exclusive formation of the smut beds 
in the ovaries of the wheat, the stone smut agrees throughout with oat 
and millet smut, so that the results obtained in infection experiments 
with these plants will also unquestionably answer for the stone smut of 
wheat. 
I will now first describe in detail the execution of the infection experi¬ 
ments and will add to this the results which were obtained in the differ¬ 
ent series of experiments with the isolated smut forms named. 
In order to procure sufficient infective material, by the cultivation 
of smut spores in nutrient solutions, I proceeded in the following man¬ 
ner: Having the year before with the greatest care procured pure and 
ripe smut masses, I allowed single spores of Ustilago carbo , U. cruenta , 
and U. maydis to germinate in March or early in April in nutrient 
solutions on glass slides. After I was convinced, by exact observation, 
of the entire purity of the cultures and of the sprout conidia developed 
in them, I introduced a few of these conidia by means of a flamed 
needle into small, shallow flasks, with broad flat bottoms and short 
necks, constructed for the purpose, in which I had previously care¬ 
fully sterilized thin layers of nutrient solutions. The sprout conidia 
so transferred which, in these flasks, were under the most favorable 
conditions for their increase by sprouting, exhausted the nutrient solu¬ 
tions in the flasks in three days’ time and then accumulated on the bot¬ 
tom as a distinct precipitate. By samples taken out it was easily possi¬ 
ble to convince one’s self of the continued purity of the culture, because 
