256 TAKING UP OF POLLEN BY INSECTS. 



it, and, on the other, to the pollinia surmounting the rostellum, and, hardening 

 almost instantaneously, cements them together. The flowers are visited by small 

 Hymenoptera belonging to the genera Cryi^tus, Ichneumon, and Tryphon, and 

 still more frequently by little beetles of the genus Grammoptera. When an 

 insect of any one of these genera lands on the labellum and proceeds to lick 

 the honey-secreting furrow from the bottom to the top, it finds itself, towards 

 the conclusion of its meal, in contact with the projecting edge of the rostellum. 

 In a moment the pollen-masses are cemented to the forehead of the insect in 

 the manner described, and are subsequently carried away upon a visit to another 

 flower. 



Strange to relate, the viscid masses sometimes adhere to the eyes of insects, 

 although there can be no doubt that their power of vision is thereby curtailed. 

 This occurs, in particular, in those Orchids where the anther-loculi and pollinia 

 diverge from one another towards their bases and are connected with two 

 separate viscid masses pertaining to the rostellum. In the flowers of Habenaria 

 montana the pollinia are inclined to one another at an angle of 70° and form a 

 kind of yoke under which it is necessary for butterflies to insert their heads if 

 they want to suck the honey from the long spur. Thus the viscid discs, and, 

 through their intervention, the pollinia are certain to attach themselves to either 

 side of the intruder's head, and the eyes are frequently the spots where the 

 adhesion happens to be effected. The genus Habenaria is also of interest 

 inasmuch as it illustrates the fact that the particular minor variations in floral 

 structure which are used to diflerentiate species always possess some special 

 significance in relation to the visits of insects. The Habenaria Hookcri of the 

 New World differs from the Habenaria montana of the Old World in having 

 in the middle of the stigma a projecting lobe, the presence of which results in 

 the formation of two enti-ances to the spur containing the honey. A butterfly 

 visiting the flower only inserts its proboscis into one of the two passages, and 

 therefore comes into contact with only one of the viscid discs and carries away 

 but a single pollinium. Yet another arrangement is found in Habenaria bifolia, 

 the Lesser Butterfly Orchis, which is widely distributed in Europe and Asia. In 

 this species the pollinia lie nearly parallel and above the enti'ance, and they 

 usually adhere simultaneously to one eye of the Sphingidse which visit them 

 (see fig. 258 '\ p. 227), or to the base of the proboscis in the case of nocturnal 

 Lepidoptera (NoctuEe of the genera AgroHs, Hadena, and Plusia). In the 

 various species of Gymnadenia the pollinia adhere to the sides of the probosces 

 of the small Noctuse which suck their honey, whilst in the Musk Orchis (Her- 

 minium Monorchis) they become attached to the front feet of such little 

 Hymenoptera and beetles as come to lick their sweet store. A long series of 

 contrivances showing a wonderful coiTelation between the forms of flowers and 

 those of flower-seeking insects might be added to these examples. 



At the time when insects visit the flowers, the Orchids hitherto referred to, 

 all of which have erect inflorescences, have their labella turned towards the 



