The Spiral Disposition of the Leaves. 533 
to one of the ordinary types of leaf arrangement, (e. g., 
In the rosettes of radical leaves, where the internodes 
do not lengthen, the spiral disposition of the leaves does 
not involve any further disturbance. The leaves simply 
grow out and retain their original position. But when, 
in the second year, the young internodes begin to elon- 
gate, this cannot happen equally on all sides of the stem, 
because the line of attachment of the leaves acts as a 
check. In consequence of this the stem must twist and 
unwind the leaf-spiral, the angle of divergence between 
successive leaves becoming gradually smaller. The num- 
ber of windings decreases, and on the other hand, the 
numbers of leaves on a single section of the spiral (i. e., 
from a given point on the stem to another vertically 
above it) increases, as our Fig. 123 clearly shows. 
Spiral torsion is, therefore, a mechanical result of the 
loss of a single character, the decussate arrangement of 
the leaves. Once this is lost, the ancestral spiral dispo- 
sition steps in, but now accompanied by peculiarities in 
the structure of the basal parts of the leaves, which on 
normal plants never occur independent of an arrangement 
in whorls, since they can only, so to speak, agree with the 
normal structure of plants with this arrangement. There- 
fore, in my opinion spiral torsion is due to a retrogressive 
transformation of the decussate arrangement of leaves. 
That it is a mechanical result can be proved by ex- 
periment; all that is necessary is to remove the cause of 
the twisting at an early stage, by cutting through the 
leaf spiral. If this is done carefully, the general growth 
of the plant is not interfered with, but the torsion will 
be locally inhibited, or, more strictly speaking, does not 
appear at the place operated on. Thus a straight inter- 
