The rostellum, or middle portion, is a tall membranous projection of a whitish colour, formed of square cells, and is covered with a thin layer of viscid matter: it is slightly concave posteriorly, and its crest is surmounted by y minute tongue-shaped mass of viscid matter. The column, with its narrow pocket-like stigma and with the rostellum above, is united on each side to a green membranous expansion, convex exteriorly and concave interiorly, of which the summits on each side are pointed, and stand a little above the crest of the rostellum. These two membranes sweep round (see back views, Figs. C and D), and are united to the filament or base of the anther; thus forming a deep cup or clinandrum behind the rostellum. The use of this cup is to afford protection, as we shall immediately see, to the pollen-masses. When I have to treat of the homologies of the different parts, it will be shown by the course of the spiral vessels that these two membranes, forming the clinandrum, consist of the two upper anthers of the inner whorl, in a rudimentary condition, but utilised for this special purpose.
In a flower before it expands, a little mass or drop of viscid fluid may be seen on the crest of the rostellum, rather overhanging its front surface. After the flower has remained open for a little time, this drop shrinks and becomes more viscid. Its chemical nature is different from that of the viscid matter in most Orchids, for it remains fluid for many days, though fully exposed to the air. From these facts I concluded that the viscid fluid exuded from the crest of the rostellum; but fortunately I examined a closely-allied Indian form, namely, the Microstylis Rhedii (sent me from Kew by Dr. Hooker), and in this, before the flower opened, there was a similar drop of viscid matter; but on opening a still younger bud, I found a minute, regular, tongue-shaped projection on the crest of the rostellum, formed of cells, which when slightly disturbed resolved themselves into a drop of viscid matter. At this age, also, the front surface of the whole rostellum, between its crest and the pocket-like stigma, was coated with cells filled with similar brown viscid matter; so that there can be no doubt, had I examined a young enough bud of Malaxis, I should have found a similar minute tongue-shaped cellular projection on the crest of the rostellum.
The anther opens widely whilst the flower is in bud, and then shrivels and contracts downwards, so that, when the flower is fully expanded, the pollinia are quite naked, with the exception of their broad lower ends, which rest in two little cups formed by the shrivelled anther-cells. This contraction of the anther is represented in Fig. D in comparison with Fig. C, which shows the state of the anther in a bud. The upper and much pointed ends of the pollinia rest on, but project beyond, the crest of the rostellum; in the bud they are unattached, but by the time the flower opens they are always caught by the posterior surface of the drop of viscid matter, of which the anterior surface projects slightly beyond the face of the rostellum; that they are caught without any mechanical aid I ascertained by allowing some buds to open in my room. In Fig. E the pollinia are shown exactly as they appeared, but not quite in their natural position, when removed by a needle from a specimen kept in spirit of wine, in which the irregular little mass of viscid matter had become hardened and adhered firmly to their tips.
The pollinia consist of two pair of very thin leaves of waxy pollen; these four leaves are formed of angular grains (each apparently subdivided into four granules), which never separate. As the pollinia are almost loose, being retained merely by their tips adhering to the viscid fluid, and by their bases resting in the shrivelled anther-cells, and as the petals and sepals are so much reflexed, the pollinia would be exposed in a remarkable degree when the flower is fully expanded, and would be liable to be blown out of their proper position, had it not been for the membranous expansions on each side of the column forming the clinandrum, within which they safely lie.
When an insect inserts its proboscis, or head, into the narrow space between the upright labellum and the rostellum, it will infallibly touch the little projecting viscid mass, and when it flies away it will withdraw the pollinia, already attached to the viscid matter, but otherwise loose. I easily imitated this action by inserting any small object into the tubular flower between the labellum and rostellum. When the insect visits another flower, the very thin pollen-leaves attached parallel to the proboscis, or head, will be forced in, and their broad ends will enter the pocket-like stigma. I found pollinia in this position glued to the upper membranous expansion of the rostellum, and with a large number of pollen-tubes penetrating the stigmatic tissue. The use of the thin layer of viscid matter, which coats the surface of the rostellum in this genus and in Microstylis, and which is of no use in the transportal of the pollen from flower to flower, seems to be to keep the leaves of pollen, when brought by insects, in a proper position for entering and remaining in the narrow stigmatic cavity. This fact is rather interesting under a homological point of view, for, as we shall hereafter see, the primordial nature and purpose of the viscid matter of the rostellum is that common to the viscid matter on the stigmas of most flowers, namely, the retention of the pollen, when brought by any means to its surface.
The flowers of the Malaxis, though so small and inconspicuous, are highly attractive to insects; for the pollinia had been removed from all the flowers on each spike, excepting from one or two close under the buds. In some old flower-spikes every single pollinium had been removed. Insects sometimes remove only
