358 
Journal of Agricultural Research 
Vol. II, No. 5 
Numerous floral inoculations, undertaken early in these investigations, 
also failed to give results supporting this view. These were carried out 
in several different seasons and at various stages of development in the 
ovary. Dry spores of the head smut were placed in a paper bag and 
shaken into the flowers by inverting the bag over a head and shaking 
thoroughly; sometimes they were placed inside the glumes with a camePs- 
hair brush. Some of the spores were germinated before applying them, 
and were sprayed into the flowers with an atomizer either by opening the 
glumes with forceps, or in the early morning while the plant was still in 
bloom; some of the heads were not covered, but some were kept covered 
for a time with paper bags or with a large lamp chimney to keep them 
moist. This was an extremely difficult matter, however, owing to the 
high winds and to the consequent rapid rate of evaporation, which, from 
an open water surface, often exceeds half an inch in 24 hours at Amarillo. 
While there was occasionally a rather high percentage of infection in the 
resulting plants, this was not the uniform result of any particular method 
of inoculaton; nor was it sufficiently large to obviate its explanation by 
infection of the plants during development in the field, in view of the fact 
that it did not occur consistently. 
ENVIRONMENTAL, EXPERIMENTS 
In addition to the negative results of inoculations, it was found that 
seed from the same lots when planted at various points in the United 
States, or in different seasons at Amarillo, gave very different amounts 
of infection in the plants produced, while in plants from different lots of 
seed, grown at the same station, no consistent differences could be 
observed. 
A preliminary experiment was carried out in 1908. The plants were 
all grown from the same lot of seed, yet those grown at Amarillo were 
7.7 per cent smutted and those at Chillicothe, Tex., were 2 per cent 
smutted, while those at McPherson, Kans., were not affected at all. In 
1910 a new series was begun. Plantings were made from two lots of 
seed at eight different stations, including Amarillo and Chillicothe, 
Tex., St. Paul, Minn., and Arlington, Va. Of these two lots, that from 
Chillicothe happened to develop the greater percentage of smutted plants 
at Amarillo, and the seed grown from it was therefore used for the 
plantings in 1911. In this and subsequent seasons the intention was to 
plant at each station seed from each of the places concerned and to use 
only seed descended from the original lot and grown in consecutive sea¬ 
sons at the same station. This was usually done, but, owing to various 
mishaps, the plantings at the -four stations named were the only ones 
which were carried completely through the experiment as intended. 
The data from these four stations thus form a complete series and are 
summarized in Table IV. They involved in each case from about 150 
to 800 plants; usually about 300. 
