390 
Journal of Agricultural Research 
Vol. XXIII, No. 5 
INOCULATION EXPERIMENTS 
As has been stated above, positive inoculation results were made the 
final proof of the identity of the isolated organisms as Fusarium vasin - 
fectum Atk. Fusaria which from their morphology were judged to be 
the wilt organism were used to inoculate pots of soil sterilized in an 
autoclave. Com meal or cottonseed meal cultures were used in the 
earlier tests. Also, uninoculated meal was put in the control pots, but 
the meal made so favorable a medium for fungous and bacterial growth 
that plants in both the control and inoculated pots were killed by almost 
any organism which chanced to infect the meal. The use of even small 
quantities of com or cottonseed meal gave these results. The toxicity 
to cotton seedlings of the banana-wilt fungus as reported by Brandes 5 
can probably be explained as a similar case of “ damping-off.” This dif¬ 
ficulty was avoided by the use of sterilized cotton stem fragments carry¬ 
ing the inoculum, the infected material being placed under the surface 
layer of soil at the time the seed was planted. Apparently healthy 
acid-delinted seed was used for all plantings. 
The seedlings growing in soil infected with the wilt organism in some 
instances began to show the effects of inoculation in eight days after 
planting, while occasional plants lived for several weeks or months after 
the other plants growing in the same pots had died of wilt. There was 
apparently much variation in the virulence of different cultures isolated, 
judging from the behavior of the plants in different pots in the same 
series of inoculations. If the seedlings did not wilt soon after germina¬ 
tion, they usually survived for a considerable time before showing signs 
of infection. No culture which had not been judged from its morphology 
to be of the wilt organism ever caused an infection resembling wilt, al¬ 
though all Fusaria obtained among the earlier isolations were tested for 
possible pathogenicity. The color reaction of the wilt fungus on rice 
also came to be considered an almost positive indication of its identity, 
and, in later series, no cultures were used in infection experiments which 
did not give the characteristic color reaction. While at first the organism 
was reisolated from all parts of the wilted plants, in later inoculations the 
wilting of the plants, coupled with the typical blackening of the xylem, 
was considered sufficient proof of the identity of the organism as Fusarium 
vasinfectum (PI. i, A). In the earlier series both uninoculated pots and 
pots inoculated with known wilt cultures were run as controls, but in 
the later series the known wilt controls were omitted. Table II summarizes 
the inoculation experiments. 
An examination of Table II will reveal a great variation in the length 
of time elapsing between the planting of the seeds in the inoculated soil 
and the isolation of the fungus from infected plants. This may have 
been due to various causes. In some cases the plants died without 
showing positive indications of wilt but merely rotted off at the soil line. 
During the winter months the growing conditions were very unsatis¬ 
factory, because of poor greenhouse facilities. Apparently there was a 
great variation both in the virulence of the cultures and in the resistance 
of plants. Practically every culture has been tried out two or three 
times, and some have been consistently highly destructive while others 
have been less so. In some cases a plant or two survived a long time 
6 Brandes. E. W. banana wii/r. In Phytopathology, v. 9, p. 339-389 • 5 fig., pi. 21-34 (partly col.) 
1919. Literature cited, p. 388-389. 
