( 426 ) 
and the basipetal impulse was visible in some of these leaves from 
the fact that stream-loops were formed. From the young prolifications, 
namely, some currents were seen going to the top of the leaf, which 
however later suddenly returned with a very sharp bend and then 
went back straight to the leaf-stalk. So gravity had first deviated them, 
but the continuous counteracting influence of the basipetal impulse, 
which was evidently. felt in every point of the leaf, had at last 
overcome gravity and got the upper hand. These loops had in large 
leaves a leneth of 5 to 10 mm. 
All these changes took place before at the apical side rootlets 
developed. This occurred finally with very many leaves; the earliest 
appeared after 9 days, the majority came later, but after almost four 
weeks they had not yet developed in all of them. That most of them 
had formed rootlets at the extremity of the leaf-stalk much earlier, 
proves that they possessed to the fuli the power of forming them. 
That the presence of rootlets at the leaf-stalk was no impediment 
for the development of new rootlets elsewhere, appeared also from 
the fact that, with respect to these latter, no difference could be noticed 
between leaves with and without rootlets at the stalk. 
If now these leaves were planted in mud with their apical rootlets 
(which, however, were hardly ever placed exactly at the top, but 
at a smaller or greater distance from it) the prolifications grew on 
or, if they had not been present beforehand, they always appeared 
after this. A connection was generally formed between the stream- 
bundle issuing from them and the rootlets and so a plant was obtained 
in which, under the impulse proceeding from the prolification, the 
nutritive current had developed in a direction opposed to the impulse 
existing in the leaf. 
Here also it could be proved in the same way, 
A as before by means of a cross-wound (as in fig. 2, 
in which the basal half of a leaf was planted 
upside down), that this bundle in the leaf obeyed 
indeed the impulse of the prolification A, since 
© from this side the thrust occurred. 
Yet here also the reversion appeared to be 
only local. A small prolifieation 4 had namely been 
formed below the cross-wound, after this had 
been made (consequently at the side of the apical 
rootlets). This new leaflet in its turn formed a 
Fig. 2. small stream-bundle of which some thinner currents 
went in the direction of these rootlets; one thicker current however 
took his course athwart alongside the wound, turned at the end with 
