24 BOTANICAL GAZETTE [ JANUARY 
The stabbed ones did not curve so much as the control, and a 
larger percentage of them straightened within twenty-four hours. 
A possible source of error is seen here, since, if observations are 
not made within a few hours after branding, the root may have 
already curved and straightened. This may account for the 
negative results in some of the earlier lots. 
If we leave out the first two lots of sixteen roots each in 
which the results are so low, we get exactly the same percentage 
of curved roots, 68.1 per cent., in both those that were branded 
on the side opposite to the stab and those that were branded on 
the side 90° from the stab. This is only 11.9 per cent. less than 
the control roots which were branded and not stabbed. This 
difference might easily be due to the inhibiting or paralyzing 
action of the injury. 
Putting together the results of these experiments, it is seen 
that the transmission of the stimulus past the cut depends very 
largely on how much of the cortex is present. The lowest per- 
centage of curved roots (20.4 per cent.) was found in those that 
had the axis complete, but much of the cortex cut off (table IV). 
The highest percentage of curved roots, 60 per cent. and 68.1 
per cent., was found among those in which the axis was com- 
pletely cut off, but the cortex mostly present (table V). The 
results in tables I, II, and III show a curving in proportion, 
roughly speaking; to the amount of cortex present. When the 
axis, as well as a large part of the cortex, was left intact 59.3 
per cent. curved on both sides of the cut, only 1.2 per cent. less 
than when the axis was cut off completely. 
We are therefore justified in drawin 
g the following conclu- 
sions : : 
1. The stimulus can be, and is, transmitted in the cortex. 
2. While it is not absolutely demonstrated t 
cannot be transmitted in the axial c 
able that it is not transmitted there. 
3. The impulse that affects the part of the root above the 
cut can be transmitted in the cortex of either the concave or the 
convex side, 
hat the impulse 
ylinder, it seems very prob- 
. 
