292 T. W. W oodhead and M. M. Brierley. 
around it and held it firmly. The rest of the branch, however, did 
not twine around this support; instead, it continued to twine 
independently of it, and eventually formed a spiral of two dextrorse 
turns above the point of attachment. A careful examination was 
made of this plant with the following result:— 
1. Number of branches which turned out of leaf axil 
and grew obliquely across the main axis. 
Sometimes both branches of a node crossed 
to opposite sides of the stem ... ... 24 
2. Number of wavy branches forming a more or less 
loose coil, some showing distinct signs of 
torsion ... ... ... ... ... ... 14 
3. Number of branches which after forming half a 
turn became sharply reflexed ... ... 8 
4. Number of branches forming tight coils clasping 
the base of the subtending leaf ... ... 8 
5. Number of branches forming tendril-like coils ... T2 
making a total of 66 branches on one plant showing the climbing 
habit. 
Occasionally this tendency appeared in the petioles of the 
leaves. In some of these the petiole curved in such a way as to 
sharply reflex the blade, in others it twisted so as to carry the blade . 
across the stem to the opposite side; in each case secondary 
thickening fixed the leaf in the assumed position. 
We thus see that from some doubtful cause these plants have 
developed a tendency to twine. These variations were confined 
almost entirely to the white variety, but of two specimens of a 
coloured variety (pink striped with dark red) grown at the end of 
the same border as the white ones, one developed four twining 
branches and the other five. None of the other coloured varieties 
in other parts of the garden exhibited a trace of it. 
That cuttings taken from the twining variety should have in 
turn continued the climbing habit was to have been expected and 
that they have done so, and in an increased degree, we have already 
seen. It was a matter of interest, however, to determine whether 
the next generation grown from seeds of this variety, would do the 
same. To test this point, seeds collected from the climbing form 
last autumn, were sown in March this year. They germinated 
freely and batches of seedlings were planted out in various situations. 
The majority of these, however, have grown slowly and produced 
few lateral branches of sufficient length to determine yet whether 
or not this habit will be perpetuated. One batch of seedlings was 
planted in a frame in the garden at Doe Royd, Almondbury, some 
