65 
Incarvillea Delavayi , Franck. 
as has been described by Dr. Ogle in the case of the Foxglove (13), and 
as a result of this twisting the anthers take up their characteristic position, 
two on each side of the style (Fig. 2, d). Not only .is this the case in 
Incarvillea , but the anthers are pressed closely against the style, a protu- 
berant pad from the connective forming the actual point of contact and 
preventing the anther-lobes from touching (Fig. 2, f). We have, therefore, 
two pairs of opposed pads, touching the style, which act as buffers when the 
pollinating insect presses against the downwardly protruding anther-prongs. 
A few belated flowers opened at the top of the axis ; they had but short 
stalks, and other, unopened buds were placed just above them. In these 
circumstances they were unable to twist round and take up their usual 
position, and it was found that the stamens did not change from the juvenile 
position. These latter changes are therefore probably under gravitational 
influence. Although these flowers never completely opened, both stigma 
and pollen matured, as both organs were capable of functioning; and the 
ovary, after fertilization, gave rise to a fruit. 
The first prong of an anther that an insect would touch on coming into 
the flower is so arranged that a slit in the anther-lobe is placed above it and 
runs to the end of the lobe. The other thorny projection has the slit 
situated just below it (Fig. 2, E and f). In each case the surface of the lobe 
on either side of the slit is not a gently curved one, but is raised up in places 
and depressed in others. The significance of this configuration I hope soon 
to make clear. On the insect’s back touching the first pair of prongs, one on 
each side of the style, the pressure is communicated to the pads and the two 
stamens and the style are kept rigidly in position. The sides of the slit or 
anther-cleft are pressed closely together and are therefore kept shut. When 
the insect touches the next pair of prongs, which will of course be on the 
same two stamens (and I imagine that a large insect like a humble-bee 
would touch both pair of prongs at the same time), not only are the pads 
brought into play, to cause the whole apparatus to remain in position, but 
as they are of some size and would not suffer much compression, in spite of 
their spongy nature, the sides of the anther-lobes are kept away from the 
style and do not touch it. This allows of considerable deformation taking 
place in the second pair of anther-lobes, and in consequence of this the slit 
opens and a certain amount of dry pollen is shed, passively, on the creature’s 
back. The prongs, however, are very elastic and are attached each to 
a massive base composed, in great part, of thick-walled cells that resist 
a change of shape ; therefore, on the insect passing, the projections are set 
free with a reverberation, which, if sufficient, causes a further and wider 
shedding of pollen, and, in any case, would redistribute the pollen-grains in 
the sac. It is seen, therefore, that the inequalities on the surface of the sac 
serve a double function. They prevent the pollen in the sac from being all 
shed at once, so allowing of several visits to the flower, and they also bring 
F 
