Rostellum

Eventhough in the majority of orchids, rostellum in its complex form is present, complete absense of such a structure as well as transitional forms are met with in some of the primitive species. Rostellum is popularly believed to be the third stigmatic to be modified, although there are controversies regarding this subject. Anyway, it separates the functional stigma from the anther and helps in attaching the pollinia on to the insect's back, by means of adhesive secretions. Apostasieae and Cypripedieae, where all the three stigmas are fertile are devoid of a rostellum. In Cephalanthera of the tribe Neottieae also, all the three stigmatic lobes are fertile. But here the insect, on its way inside the flower, brushesagainst the pollen masses, which are glued on to its back thereby. In more advanced forms, particular regions of the rostellum are concerned with the secretion of the viscous matter. But there isnot yet any contact between the pollinia and the viscous matter. In the slightly more advanced forms like Calanthe, a portion of the rostellum is differentiated into the viscid pad which is actuall attached to and is removed along with the pollinium. In its most advanced state, as it is seen in Sarcanthinae and Oncidiinae, this viscidium is very sharply circumscribed and is attached to the pollinia by means of a sterile strap of tissue called the stipe. According to Dressler and Dodson, this differentiation of a highly complex rostellum is independently achieved along different phyletic lines in subtribes like prasophyllinae, Genyorchidinae, Spiranthinae etc.

According to Holttum (1960) there has been simplification of the rostellar apparatus in the history of its evolution, as there had been attainment of complexity. In the subtribe Spiranthinae of Neottieae, the pollinia have a distinct viscidium, whereas in the more advanced Dendrobiinae, the pollinia are naked. Holttum is of opinion that Dendrobium might have lost its pollinial appendages in the course of evolution.

In many genera of Orchideae, a 'rostellum' is present in addition to the three functional stigmas. Vermulen (1959) explained this by postulating that the rostellum in Orchideae has an origin independant of the other groups. But according to Garay (1960) Orchideae does not possess a rostellum comparable to the other tribes, but what is usually designated as rostellum in Orchideae is merely a connective tissue binding the two thecae of the anthers. He further explains that the column in Orchideae is not derived from any other group but is a product of independent evolutionary line emerging from a polyphyletic complex.