PECAN ROSETTE. 21 
cells than rounded leaves of the same plant, but the narrower form 
of such leaves is due to a larger number of cell divisions with spin- 
dles parallel to the long axis. 
In pecan rosette the general shape of the leaf seems to follow this 
rule, depending upon the orientation of the cell divisions rather than 
upon differences in the size or the shape of the cells. That is, the 
linear shape is not due to the development of cells elongated parallel 
to the midrib, but to a difference in the number of cell divisions in 
the two axes; nor are reduction in both length and breadth of leaf 
caused by a decrease in the size of the cells, but rather to a decrease 
in the number of cell divisions in both axes. On the other hand, 
there is often a large and localized change in both the size and the 
shape of the diseased cells, but not as related to leaf shape nor neces- 
sarily to leaf diameter. In the linear type of leaf the cells are as 
likely to be enlarged parallel to the short as to the long axis of the 
blade, and in portions of leaves profoundly reduced in both length 
and breadth the palisade and the spongy cells are at the same time 
often considerably enlarged in all three dimensions. 
As a result of a considerable number of measurements of the 
thickness of leaves from healthy and resetted trees of like age and 
variety, striking differences were found associated with the disease. 
(Table II.) Comparisons were made between leaves collected from 
the north and from the south sides of trees, but no constant differ- 
ences were found which could be referred to situation, since all 
leaves were taken from the lower, outstanding branches, and under 
these conditions those on the north received nearly or quite as much 
light as those, on the south periphery. 
In each case the figures are based on 10 to 15 measurements of 
the thickness of each of several sections from comparable parts of 
each of 10 or more leaves. From such measurements it was found 
that the average variation from the greatest thickness in normal, 
individual leaves was 18 per cent, with extremes varying between 
10 and 22 per cent. In the various types of rosetted leaves the ex- 
treme differences in thickness varied between 11 and 62 per cent 
of the greatest thickness. The least variation was found in the 
nonmottled linear or aborted leaves, while the greatest differences 
occurred in mottled leaves. Extreme variations in thickness of 
normal leaves of the Frotscher variety were 131 to 187 microns, 
while in diseased leaves of the same variety the range was from 70 
to 234 microns. The smallest measurements were taken at the thin 
places in the leaves where tissue differentiation was lacking. The 
Van Deman specimens examined had slightly thicker leaves, but the 
same relations in thickness were found to hold between the healthy 
and the rosetted leaves of this and several other varieties. 
76289°— 22 1 
