530 Heritable Spiral Torsion. 
race of Dipsacus sylvestris torsus, during over thirteen 
years, together with a long series of further observations 
on the inheritance of this anomaly in other plants, have 
proved that this character is as heritable as other anomalies 
are. Plate VI gives a view of a culture of this race, re- 
produced from a photograph of one of my beds. 
Real spiral torsion only occurs in those species which 
normally have a decussate or whorled disposition of the 
leaves. It consists in the substitution of a spiral arrange- 
ment for this. The leaves arise from an unbroken spiral, 
along which they are attached to one another more or 
less closely to their bases (Fig. 123). This close spiral 
is sometimes interrupted and normal internodes are inter- 
calated in the twisted part. Not infrequently the torsion 
is limited to a greater or a lesser part of the stem (com- 
pare, e. g., below Dianthus, Fig. 129). Indeed no sin- 
gle stem is completely abnormal from the very beginning. 
As might be expected the fusion of the base of the 
peduncles into a continuous band results in an inhibition 
of the longitudinal growth of the stem. The internodes 
cannot elongate normally, and as they strive to extend, 
they partly unwind the leaf spirals. In consequence the 
spiral becomes steeper and not infrequently unwinds so 
much in the upper parts of the stem, i. e., those parts 
which normally grow in length, as to become a straight 
line. When this occurs the leaves and their axillary buds 
arise in a longitudinal series on one side of the stem 
(Fig. 122). Obviously this can only be reached by the 
stem twisting itself in the opposite direction to that of 
the leaf-spiral (Fig. 123). Inside the twisted stem, 
if it is hollow, the diaphragms which normally occur at 
the nodes, do not exist as such, but are united together 
