34 THE JOURNAL OF BOTANY 



close against the stem, the circular shield-like laminae, which are 

 attached by the centres of their backs to the apices of the petioles, 

 are vertical and face inwards, and the tentacles are tightly inflexed, 

 so that they are entirely within the hollow lamina. As the stem 

 elongates the petiole gradually bends away until it becomes hori- 

 zontal or inclined downwards. Simultaneously by a movement 

 of the petiole just below its apex the lamina is carried completely 

 over, so that when it comes to rest (at about the same time as the 

 petiole does) its face is towards the earth. Still a third motion 

 proceeds almost simultaneously with the other two ; as soon as 

 the petiole lias moved a little away from the stem a few tentacles 

 of the outer row of the then upper edge of the lamina unfold, and 

 are soon extended in the plane of the rim. These movements are 

 all illustrated in Plate 523, fig. 2. 



At this stage of its life the plant sways and gyrates in the 

 gentlest breeze. If in its gyrations its tentacles come in contact 

 with a twig, tree-trunk, the horizontal wire of a fence, or even a 

 flimsy straw (the swaying stem of another example of the species 

 is not rejected), they immediately take firm hold — it is remark- 

 able how firmly they adhere even after only momentary contact. 

 Having caught hold, the tentacles bend outwards until they lie in 

 close contact with the back of the lamina ; their apical halves then 

 bend forward sharply, and ultimately lie upon the lower halves. 

 The final position is: the lower half of the tentacle is pressed 

 close upon the lamina, and the upper half is pressed closely on 

 the lower half (fig. 1). The advantage of this arrangement is 

 obvious : the plant is fixed to its support by spring attachments. 

 When the movement of the lamina is complete, these attachments 

 stand upon its inner edge, next the axis of the plant. Being 

 centrally attached, the lamina itself has considerable freedom of 

 motion. In clinging to its support by the inner edge of such a 

 lamina the plant secures, I think, the maximum benefit from its 

 spring attachments ; I feel certain that attachment to the outer 

 edge would mean considerably more risk of an occasional dead 

 pull. The perfection of the arrangement is remarkable, and might 

 well be thought to have been developed especially to assist the 

 plant in climbing ; but the laminae and tentacles of D. stricticaulis 

 behave similarly, and will hold on to objects (a neighbouring tree- 

 trunk, e. g., is often so held), though the plant is not a climber. 



I see no reason for supposing that this latter species has 

 arisen from D. macrantha, and inherited from it the tentacle 

 movements. My impression is that both species have arisen 

 independently from some older form. It seems probable that 

 D. macrantha had practically all the machinery for climbing at 

 the start ; in other words, the perfection of the arrangement is 

 largely fortuitous. I say largely fortuitous, because I think some 

 of the movements have become more definite and intense in 

 D. macrantha than in D. stricticaulis ; but there are several non- 

 climbing allies of these species, which I must fully examine before 

 I can speak definitely. 



The tentacles which unfold first stand at the morphological 



