366 
Journal of Agricultural Research 
Vol. II, No. 5 
growing point—was removed and preserved. The plants were then 
induced to produce a second growth from what remained. The meristem 
of those which developed head smut at maturity was then carefully exam¬ 
ined ; yet in none of the 16 plants which developed the disease could the 
hyphae be found in the parts preserved. 
In additon to the negative evidence of these dissections, Mr. Karl F. 
Kellerman, of the Bureau of Plant Industry, stated to the writer in recent 
conversation that he performed a number of experiments with this smut 
by artificial inoculations on sorghum in the greenhouse while working in 
Ohio with his father, Dr. W. A. Kellerman. The plants were in pots and 
were inoculated at stages varying from the time they first appeared above 
ground until they were about 5 inches high. The method used was to 
wash the soil away from the roots, sift dry spores over them, and re-cover 
with soil. While some indications pointed to infection through the roots, 
this was not definitely established. Whatever the mode of entry, how¬ 
ever, the parasite proved able under the conditions in the greenhouse to 
infect plants at all the stages at which they were inoculated. 
In the recent greenhouse experiments (Table V, plat F, No. 1) some of 
the plants were successfully inoculated after the first leaf had begun to turn 
green. But, most unexpected of all, after leaving these plants to grow all 
summer it was found in October that the control (plat F, No. 2) contained 
five smutted plants, whereas the original culms which developed in April 
showed no sign of the disease. Other plants, too, which had not been 
smutted in the spring had grown smutted culms by fall* While Hecke 
(1907, p. 572) has presented similar facts as proof of shoot or branch 
(“Trieb”) infection by Ustilago anther arum, in the case of sorghum, at 
least, there is some uncertainty as to the exact point of infection. The 
inoculation of the nodal buds has been tried a few times in the greenhouse 
without result. This does not preclude the possibility of such an infec¬ 
tion, however, and more careful work supported by histological observa¬ 
tions is needed. 
It does not seem that the spread of the disease from plant to plant 
under greenhouse conditions makes it probable that such an occurrence 
is at all common in the field, but it does add certainty to the conclusion 
that infection by this smut is by no means confined to the early seedling 
stage of the host. This, then, together with the sparse germination of the 
spores, readily explains the repeated failures to produce any appreciable 
amount of infection by inoculation of the seed. 
In Table V, plats C, D, and E, it will be observed that the same lot 
of seed, “Seed lot II,” previously treated with a 0.24 per cent solution of 
formaldehyde, was used for nearly all the inoculations. This seed pro¬ 
duced plants free from head smut at both Manhattan, Kans., and St. 
Paul, Minn, (plats D, Nos. 10, 11, 12, and 13, and E, Nos. 7 and 10), 
except when artificially inoculated; but at Amarillo all but one of the 
