66 Cutting . — On the Pollination Mechanism of 
about a wider shedding of the grains on to the insect’s back than would be 
the case if the surface were smooth. It is possible that the creature makes 
several contacts with any one prong : this would result in several separate 
sheddings. On the insect passing to the next pair of stamens the same 
procedure takes place, so that on an insect entering the flower only the 
inner lobes of the anthers shed pollen on to it, and these probably do 
so on about the same spot. I am, of course, quite unable to be sure of this ; 
the movement in or out of a flower can be imitated on the anthers by pass- 
ing a finger along the points of the prong-like projections, but the exact 
details of pollination can be made out only in the natural habitat of the 
plant, and by an examination of the pollinating insect itself. When the 
insect makes its exit from the flower the reverse process takes place ; only 
the outer lobes of the anthers open and pollen is shed from them. The 
figure of the anther given in Engler (14) shows a quite similar set of prongs, 
arranged as in Incarvillea Delavayi and Incarvillea grandiflora (a few 
flowers of this species were examined by me), and the connectival pad 
is also present, but the figure given of the flower of Incarvillea Sinensis 
shows the stamens emerging from the mouth of the corolla tube ; if this is 
a normal flower of the plant, the pollination mechanism cannot be the same 
as in the two species here described. The flower, however, is drawn almost 
in a vertical position, and in two flowers of I. Delavayi that developed 
in a similar manner, owing to the flower-stalk being short and the other buds 
being crowded around it, the anthers were notable to take up their normal, 
adult form. This may have been the case in the specimens figured in 
Engler. It would seem probable that neighbouring genera have similar 
mechanisms and would repay any observation of them. Schumann (14) 
describes and figures the lobes of Amphicome as having thorns and a leaf- 
like broadened connective. Unfortunately I have been unable, so far, 
to obtain specimens of this genus. 
The possibility that some stimulus mechanism might be connected with 
pollen discharge in Incarvillea was not left untested. It was found that 
anthers which had been placed in alcohol for a period of at least a year 
were able to behave in a similar manner to the live ones. Sections of 
the anthers showed that they are provided with the usual fibrous cells ; 
these, most probably, are concerned with the preliminary opening of the 
slit, and may also function in other ways : for example, in assisting in 
maintaining the necessary slight resistance to change of shape on the part 
of the anther. A microscopic examination of the stigmatic lobes of In- 
carvillea did not show the presence of any Incarvillea pollen, but there 
was plenty of foreign pollen, and an examination of this indicated that 
it was mostly that of the Rhododendron— some few plants of which grew 
in the garden near the experimental plants. As these are humble-bee 
flowers, and as I have seen these insects visiting the Incarvilleas. I have. 
