4 BULLETIN 1484, U. S. DEPARTMENT OF AGRICULTURE 
to give up moisture and its content of plant food material when 
needed." Skinner and Demaree (9) found that any of the ordinary- 
methods of increasing the supply of organic matter in soils tended to 
correct pecan rosette and stated that it was not definitely known 
whether the improved condition resulted from the additional plant 
food or from the greater ability of the soil to retain moisture. Rand 
(8) in 1922 concluded from study of the histology and cytology of the 
disease that "it bears more similarity to the known infectious chlo- 
roses than to those caused by soils or climatic factors" but presented 
as unanswered the question whether the factors responsible for the 
derangement must be introduced from without or whether they origi- 
nate within the plant itself. 
The limited evidence now available on the crazy-top disorder of 
cotton points rather to the possibility of environmental influences or 
to the existence of a type of contagium which does not change the 
structure and metabolism of plants that are functioning in a favor- 
able environment. 
The scarcity of crazy- top on fine-textured alluvial soils and the 
decided contrasts in the prevalence of the disease in contiguous areas 
that have received different cultural treatments indicate that plants 
which are subjected to stress conditions are more susceptible (or less 
resistant) to the factor or factors which produce the disorder than 
are plants which are continuously afforded favorable conditions for 
growth and development. 
PERSISTENCE OF THE DISEASE IN CERTAIN SOIL AREAS 
The recurrence of spots of diseased plants in the same location 
from year to year had been reported to the writers by various 
farmers, and steps were taken in 1923 to determine whether such 
persistence was characteristic. Two spots in a field of Hartsville 
cotton near Casa Grande, Ariz., and one spot in a field of Acala 
cotton near Mesa, Ariz., were located and recorded by actual 
measurement. These spots were selected because of their rather 
definite margins and their isolation from other areas of abnormal 
plants. 
In 1924 and 1925 the two crazy-top spots near Casa Grande held 
the same relative positions as formerly, but one of them had expanded 
considerably in all directions. The other showed no material change 
on the eastern boundary and had expanded very little except 
on the west side. In 1925 this spot was smaller than in 1924 and 
the symptoms were less severe. It was noted that the soil on the 
east side of this spot was of a different character from that within 
and west of the spot, containing more sand and being much more 
penetrable by water and by soil-sampling tools. No diseased plants 
were found in the area to the east of this spot, and none were noted 
in another field 75 feet east of the spot and planted to cotton for 
the first time in 1925. The spot located near Mesa maintained its 
position in 1924 and 1925, but since almost the entire field had become 
involved, with more than 60 per cent of the plants deranged by 
crazy-top, the spot could be distinguished only by reason of the fact 
that the plants in this area were taller, with darker foliage, and 
were more completely sterile than the surrounding plants. This 
seems to indicate a greater severity or earlier incidence of the 
