3 88 
Journal of Agricultural Research voi xxm, no. 5 
organism was carried within the seed seemed possible. In 1921 in a small 
plot of cotton grown at Fayetteville, from thoroughly surface disinfected 
seed, two wilt plants appeared. The plot was far from the cotton area 
and had never before been planted to cotton. One of the two diseased 
plants died early, producing no bolls; the other lived throughout the 
season. These two cases strongly suggested the possibility that the 
cotton-wilt fungus may be carried on the inside of the seed coat, and 
the experiments here reported were directed primarily toward discovering 
whether or not such was the case. 
METHODS OF PROCEDURE 
Seed was carefully selected from wilt-infected cotton plants, delinted 
with concentrated sulphuric acid as a preliminary treatment, then further 
disinfected before planting. The seed was germinated under aseptic 
conditions, and as far as possible all fungi appearing on either viable or 
dead seed were identified. Fungi which from their morphology and their 
biochemical reactions were judged to be Fusarium vasinfectum Atk. were 
saved for further tests expected to establish their identity. No fungus 
was definitely accepted as being F. vasinfectum until successful inocula¬ 
tions of cotton plants had been secured. At first the plants which had 
died, apparently of wilt, were strongly surface sterilized, and the fungus 
was reisolated from all parts of the stem; but in the later tests the typical 
wilting, accompanied by the blackening of the xylem, was considered 
final proof of the identity of organism as F. vasinfectum . 
ISOLATION OF THE FUNGUS 
Seed from badly wilted cotton plants, in which the vascular infection 
could be traced into the bolls from which the seed was taken, was selected 
in the field September 21 to 24, 1921. The selections were made with 
particular care, which fact probably had much to do with the results 
subsequently obtained. After the longer lint had been hand picked 
from the seed, the seed was delinted with concentrated sulphuric acid, 
washed and dried, and put away in a cloth bag until used for plating. 
Each lot of seed was again subjected to about five minutes* treatment 
with concentrated sulphuric acid, washed, treated for two minutes in 
1 to 1,000 mercuric-chlorid solution, and washed in sterile water before 
plating. It would seem impossible for any organism external to the 
seed coat to survive this treatment, and all organisms isolated from seed 
so treated have been considered as coming from inside the seed coat. 
Plating was begun September 27, three days after the seed was gath¬ 
ered. Four series of platings were made within a month of the time the 
seed was gathered, and the wilt fungus was isolated 31 times from 524 
seeds, or from practically 6 per cent of the seeds plated. Within three 
months from the time the seed was gathered, the fungus had been iso¬ 
lated from 39 of 769 seeds plated. The first plating was made on corn- 
meal agar, but in all following series the plating was done on white blot¬ 
ting paper in Petri dishes sterilized in an autoclave. All fungi appearing 
on the seeds were saved for identification, at first by transferring them to 
corn or potato agar stants, later by transferring the suspected fungi to 
rice tubes for the color reaction of the fungi, which in the case of the wilt 
organism is rose pink, or a little darker. Germinated seeds which showed 
