230 
When young the concave surface is on the proximal side of 
the pulvinus, later it may be on the dorsal, ventral or even 
distal sides. As the pulvinus develops, it straightens spontane- 
ously, the difference between the relative irritabilities of the con- 
cave and convex surfaces and of the sides being best shewn when 
the tendril is nearly straight , as when already markedly curved, 
rubbing the sides produces a side curvature along with an 
increase of the original curvature and the pulvinus twists 
towards, instead of simply bending to, the side stimulated. 
Stimulation of the concave surface, when a marked curvature 
is already present, produces a less marked and rapid response, 
than when only a slight curvature is originally present. When 
the pulvinus is older, the dorsal, ventral or even the distal 
surface may become concave. The new concave surface posses- 
ses, when but a slight curvature is present, an almost imper- 
ceptibly greater irritability than the other sides do; practically 
the pulvinus is iso-irritable on all sides at this stage. 
Removal of the epidermis injuring the cortex may cause the 
pulvinus to die, but if the injury is only slight, it remains 
living and curves in one or more coils with the injured side 
innermost, but does not perceptibly harden or thicken. In this 
case an injury acts as a stimulus to curvature, but not to 
thickening, as it does for example in the hooks of Uncaria 
Artabotrys etc. 
It is only when a quite young pulvinus twines around a 
support that the leaflet portion is thrown off. Even then as 4 
general rule this only takes place, when the support is a thick 
one, and the pulvinus is not, or is barely, long enough to twine 
around it. Occasionally without any contact or perceptible 
cause the leaflets fall off from the young terminal pair of leaf 
segments, leaving the pulvini, which may remain irritable and 
capable of coiling for 3 weeks or more , as short thick curved 
“tendril-hooks”. If the leaflets are expanded when the pulvinus 
twines around a support, they remain attached, and at first 
may be upside down, but soon the pulvini of the leaflets twist 
around, as also does the distal portion of the petiole, $0 that 
