782 
Journal 0} Agricultural Research 
Vol. XXIII, No. xo 
While many more fertilizers and combinations of treatments and more 
extended cropping measures could have been tested, it seems reasonable 
to 'believe that the experiments carried on covered a sufficiently large 
number of possibilities to warrant the drawing of rather definite con¬ 
clusions therefrom. 
While the extent of rosette disease was influenced very little in the 
various experiments just enumerated, the proportion of recovery after 
the development of the disease was somewhat less in the untreated than 
in the treated and fallowed plots. 
In addition to the experiments on soil treatments and on rotation and 
tillage methods, soil disinfection experiments were carried on in connec¬ 
tion with the work on parasitic factors. These experiments, which are 
described later, involved the treating of infested soils with a dilute formal¬ 
dehyde solution previous to seeding in the fall. Such treatment resulted 
in a complete control of the disease. 
While little investigation seems to have been made on the chemical 
reaction of formaldehyde in the soil, it appears to be rather generally be¬ 
lieved that its action is confined largely to protein substances and that 
its chief effect in the soil is upon living organisms present. 
The results of the experiments upon soil treatment and cropping 
methods and those obtained from soil disinfection, together with all the 
indirect evidence presented, seem to warrant the conclusion that im¬ 
proper nutrition or toxic substances, which may be in the soil as a result 
of continuous cropping or from the decomposition of organic or inorganic 
matter, are not to be looked upon as primary causes of the rosette disease. 
PARASITIC FACTORS 
This factor has been given more serious consideration than any other 
for the reason that a greater mass of indirect evidence points in its direc¬ 
tion. 
In order to throw some light upon this phase of the problem, it was 
decided to conduct experiments along lines of soil disinfection. It was 
reasonable to believe that if the disease is caused by certain types of para¬ 
sites, soil disinfection should control the trouble. Two experiments were 
conducted on infested soil in the experimental plots located near 
Granite City, Ill. 
The first experiment was conducted in six standard No. oo galvanized 
steel garbage pails, 10 % inches in diameter by n inches deep. These 
pails were filled with heavily infested soil. The soil used in three of the 
pails was disinfected with a solution consisting of i part of 40 per cent 
formaldehyde and 49 parts water. This solution was sprinkled on thin 
layers of the infested soil as the soil was shoveled into the pails. Suffi¬ 
cient solution was used to produce a thick mud. This was done five 
weeks previous to sowing the seed. The three pails containing the un¬ 
treated soil were handled in the same manner except that water alone was 
added to the soil. Pails were allowed to stand in the open to air and dry 
out for about five weeks. Harvest Queen (white-chaffed Red Cross) 
wheat, a susceptible variety, from a field free from rosette disease, was 
sown in each pail on October 8, 1920. The pails were set on a frame 1 
foot from the ground and surrounded with sawdust held within a retain¬ 
ing wall (PL 5, A). This arrangement served to prevent unduly severe 
freezing and winterkilling. From 20 to 26 robust plants developed in 
each of the pails during the fall. No signs of the rosette disease devel¬ 
oped during the fall period. In the spring, however, the disease devel- 
