15 6 
B: For the infection experiments with millet smut, TJstilago cruenta , 
I selected the largest kind of millet, viz, Sorghum saccharatum (nigrum), 
because in the others the seedlings are entirely too small and therefore 
not suited for the experiments. Even the seedlings of the sugar millet 
are quite small in comparison with those of our own cereals. They have, 
on the other hand, the advantage that at first they grow much more 
slowly than the seedlings of oats and barley. 
The millet smut ( Rirsehrand ), like the dusty smut, appears in the fruit¬ 
ing spikelets, and the grains are changed into a black mass of smut.* * 
The spores germinate readily and produce sprout conidia in endless 
abundance. These are deposited in the nutrient solution as a precipi¬ 
tate, which differs strikingly from that of the dusty smut in its whiter 
color and the non-gelatinizing membrane of the conidia. 
I. The first series of experiments, in 1885, was reduced to 32 plants 
by a hail storm. The germinating embryos were infected with the 
sprout conidia of U. cruenta by means of the atomizer. Among the 32 
plants which remained there were, in autumn, 12 smutty and 20 sound. 
II. The next series of experiments was made in the following year 
by direct infection of seedlings, which, however, were not all of the 
same size or in quite the same stage of germination. In autumn the 
harvest of 270 plants yielded 120 sound and 150 smutty. 
Early experiments, with sufficient materials, where the seedlings were 
rigorously sorted according to their size, were not begun till 1887. 
III. First, the smallest plants, in which the growing point was just 
emerging from the grain, were picked out and infected. Here, in 
autumn, out of 250 plants were harvested 180 smutty and 70 sound.! 
IV. Next, seedlings were selected with shoots a centimetre long. 
Here, in autumn, from 150 infected plants were gathered only 24 smutty 
and 120 sound. 
V. Seedlings with shoots 1J centimetres long. Here, in autumn, 
from 190 infected plants, 12 diseased panicles were counted; 178 re¬ 
mained sound. 
VI. Seedlings with shoots 2 centimetres long and projecting from 
the sheath. In autumn, out of 220 plants only 4 were diseased, the rest 
were sound. 
VII. Seedlings with shoots which had grown through the sheath to 
a distance of 1 centimetre. Here, in autumn, no smutty plants appeared. 
VIII. As soon as the millet seedlings were large enough to be in¬ 
fected by spraying the germs into the heart from above, 192 plants, 
ago, in my first oulturo experiments with the dusty smut, I discovered that the smut 
spores from barley would not germinate even after one year, while those from oats 
still germinated readily after more than six years. 
* In more than three hundred smutted millet plants, for which I have to thank 
Prof. Julius Kuhn, I found the smut nowhere except in the ovaries. 
t There can be no doubt that the larger per cent of smutty plants in the millet 
as compared with oats, is referable to the slower growth of the millet soedlings. 
Otherwise, on account of their smallness, the seedlings are less favorable objects for 
infection than those of oats. 
