296 T. W. Woodhead and M. M. Brierley. 
In the pericycle, however, rather marked differences occur. 
Just within the endodermis one, sometimes two, small elements 
appear with distinctly thickened walls, these stand opposite to the 
phloem groups and are the first signs of pericyclic stereome while 
on the opposite (convex) side, there is no trace of them. In a fully 
coiled, mature branch these thickened elements have extended on 
the concave side to such an extent that they form a nearly 
continuous stereome band in the pericycle, whereas on the convex 
side no thickening whatever in this region occurs. In many cases, 
towards the end of the season, the coils had thickened considerably 
and a section through these thickened coils showed a very extensive 
development of xylem exactly similar to what occurs in many 
typical climbers. A complete cylinder of wood is developed and 
the coils become firm and 'rigid. This, which is part of the usual 
secondary thickening of the branch, becomes more pronounced in 
the region of the coil. 
Summary. 
The twining habit appeared in a striking manner in many 
plants. While numerous branches of an individual were affected 
others showed no external signs of it. We have also seen that 
there is a refreshing absence of uniformity in the manner of twining 
as well as in the region in which the coil is produced. 
Frequently twining takes place at the base of the shoot without 
reference to any object around which it could twine. In others the 
coil gripped tightly the base of its subtending leaf, modifications 
which are quite useless as a means of obtaining support. Still, 
may we not regard it as part of a variation in the race which in a 
modified form becomes useful ? In some, the coil at the base is a 
loose one, and may act as a hook and thus be useful for support, so 
also will be a more distant loose coil, especially during the swaying 
movements of the plants in a wind. 
Other coils may be completed within a single internode. Or they 
may be formed by the upper part of one internode and the lower 
part of the next internode above, thus involving a node. And this 
may occur in any portion of the branch from base to apex. 
Not uncommonly the flowering branches and the pedicels of 
the flowers show the same variation, but when flowering shoots 
developed the twining habit strongly, the flowers were usually 
aborted. 
Again the coils may completely encircle a support in a close 
