46 Snow .— The Conduction of Geotropic Excitation in Roots. 
in the elongating region. The four on the right, with tips replaced, have 
curved down to different extents. 
The above results should leave no doubt that replacement of the tip 
does lead to geotropic curvature. 
Attempts to produce geotropic curvatures by cutting off the tip, and 
attaching it in a horizontal position to a glass slide while the 
stump remained vertical and then replacing the tip, were not 
successful. This is, perhaps, not surprising, since the tips 
must have been under very abnormal conditions. Attempts 
were also made to present the tip only to gravity, without 
presenting the stump, by cutting off the tip obliquely and 
then rotating it through 180°, as shown in Fig. 2. But the 
results, though favourable, were not pronounced enough to 
-TIG. 2 # 4 
be convincing. Reasons will be given later for considering 
that the results of this section can scarcely be explained except as due to 
conduction of excitation from tip to elongating region. 
Fig. 3 
Section 2. Replacement of Tips in Anomalous Positions. 
5. Nine roots were decapitated and the tips stuck on again with 
gelatine in a position covering half the stump, as shown in 
Fig. 3. They were then kept vertical. After twenty-four hours, 
5 were still straight and 4 slightly curved towards the side 
covered by the tip. 
6. Six roots were decapitated, the tips killed by boiling 
and then stuck back, covering half the stump, as above. After 
twenty-four hours, 2 were still straight, 3 were curved slightly 
towards the side covered by the tip, and 1 strongly so curved. 
It would seem natural to explain this result as due to some 
substance diffusing from the dead tip, passing straight up the side of the 
stump covered by the tip, and retarding its growth. 
7. The results of experiments to be referred to later will show that 
geotropic response can be brought about by conduction of excitation along 
either the upper or the lower half of the root alone. In the case of the 
lower half, if such excitation travels back by a straight path from the tip 
(as we shall find reason to believe to be the case), then it must be of such 
a kind as to cause retardation of growth in the lower half of the responding 
zone, which is affected by it. 
If, then, the tip is stuck back so as to cover half only of the end of the 
stump (as in the last two experiments), and the root is then laid horizontally 
with that side of the stump that is covered by the tip uppermost (Fig. 4), 
we should expect a similar retarding influence to be transmitted now along 
the upper side of the stump, and so cause it to bend up. Unfortunately, it 
