558 Heritable Spiral Torsions. 
inflorescences at the top. In this respect also, my results 
confirm those of DELPINO. 
External conditions exert a great influence on these 
secondary anomalies just as they do on the main torsions. 
The more favorable the conditions, the rarer are the in- 
dividuals, all of whose branches and leaves are normal. 
Obviously this fact suggests that the factor for the mal- 
formation must be present in all of them, 
Brief mention should also be made of the so-called 
local torsions. They occur occasionally in all twisted 
races. Fig. 132 represents an instance of them in J \ilc- 
rlana officinalis. It flowered in the same year and on the 
same bed as the completely twisted stem shown in Fig. 
122 (p. 528). 
In Dipsacus laciniatus the malformation was confined 
to these, and in Dipsacus syk'estris torsus I observed 
them, under very special conditions of culture, on erect 
and otherwise decussate stems ; but on the lateral branches 
of twisted individuals, especially on the strongest ones, 
such as those which arise from the axils of the radical 
leaves or from the middle of the twisted stems, they are 
always seen to be most profusely produced. 
As in the case of fasciations, forms intermediate be- 
tween these local torsions and the normal arrangement 
of the leaves, are relatively rare. This is not true, how- 
ever, if the ternary whorls, split leaves, and the local 
torsions on the lateral branches are included, but only if 
we confine our attention to the torsions on the main 
stems. Unfortunately the exact measurement of the 
part bearing leaves in a spiral is a matter of considerable 
difficulty : inasmuch as the spiral begins inside the ro- 
sette at a time when the oldest leaves have already rotted 
off and disappeared. Nevertheless I have recorded the be- 
