( 427) 
a sharp bend and went to the old base of the leaf. In doing so this 
current crossed?) the bundle going from the other prolification to the 
apical rootlets, but even this did not cause it to change its direction. 
So here also a reversion was obtained, this time by the influence 
of gravity, but it also was proved to have a very local character. 
The inverted leaves gave me material for still another experiment. 
These leaves, as has been remarked, had at last for the greater 
part formed prolifications and rootlets towards the apical side. Now 
in some of them a prolification and a rootlet were found at about 
the same height, but the one on the left side of the leaf, the other 
on the right. 
What would happen now if this piece of the leaf, isolated from 
the other basal rootlets and prolifieations by a eross-wound, were 
planted separately ? Since a manifold direct, but feeble communication 
actually existed between the two organs by means of the numerous 
fine protoplasmic currents, it was possible that the direet communication 
would be strengthened and so a cross-current would arise, in the 
same way as above a large cross-wound. But it was also possible 
that the basipetal impulse of the prolification and of the piece ot 
the leaf would not admit a communication or not one in that 
direction. 
The three experiments for which suitable material was obtained, 
were not entirely at an end at my departure. Yet it then appeared 
already with perfect distinctness that nowhere a strengthening of 
the cross-communication had taken place. On the contrary, the currents 
everywhere went from the prolifications straight to the basal wound ; 
currents Communicating with the rootlet showed the same. Even an 
indirect Communication between the two, via the basel wound, was 
not established. 
In one of the leaves a young rhizome was formed beside the 
prolification at a distance of 1'’, millimetre from it; a communication 
between the two was established, but by a very roundabout way, 
mm. 
In another leaf a rhizome was developed near the rootlet and 
another a little above the prolification. Im both cases the communi- 
viz. via the basal wound, which lay at a distance of 6°/ 
2 
cation between each of the two groups was again established via 
the basal wound, but none between prolification and rootlet. The 
complete physiological separation between prolification and rootlet 
1) The current seemed to intersect the bundle, but as the currents proceed 
from beam to beam and often two currents are attached to one beam at different 
heights, these currents must have crossed each other. The same often happens with 
stream-loops. 
