DARWIN ON FERTTLTZATTON OF ORCHIDS. 875 



Gymnadenia eonopsea and alhida differ iu detail only from Orchis. 



Habenaria clilorantha has a drum-like viscid disc of great func- 

 tional importance, but its structure and action are far too complicated 

 to be abridged here. 



H. hifolia is found to differ in so many characters from H. chlo- 

 rantha as to be considered an undoubtedly good species, and further 

 it is fertilized in a totally different maimer. 



^pipactis palustris. The labelliun is of peculiar structure, the 

 distal half being hinged on the other so lightly that a fly depresses 

 it. An insect entering the flower depresses the distal portion (which 

 closes after it), and reaches the nectarial cayity without touching the 

 rostellum ; but, in backing out, the action of raising the said distal 

 portion forces the insect against the rostellum, when it removes the 

 pollen. There is no movement of depression required ; for, on the 

 entrance of the insect in another flower, the polhnia it bears are 

 brought into immediate contact with its stigmatic surface. In 

 E. latifolia, the distal portion of the lip is not flexible, and the 

 operation is more simple. 



Geplialantliera grandijlora presents the all but unique case of an 

 Orchid wanting the rostellum {Cypripedium being the only other); 

 its pollen grains are separate and spherical. Here perpetual self- 

 fertiKzation is imperfectly secured by the friable pollen grains reach- 

 ing the stigma at a very early period indeed ; but the structure of 

 the flower and relations of the parts are such that insects must help, 

 so that the flowers are partially fertilized by their own poUen and 

 partly by that of other flowers. The details are very intricate, and 

 the discussion highly interesting and curious. 



Goodytra repens is one of the most interesting British Orchids, as 

 connecting several distinct forms ; iu the development of a caudicle 

 and cohesion of the pollen grains, it approaches the tribe Ophrece; in 

 other respects it is allied to JEpipactis, Spiranthes, and Orchis, 



Spiranthes autumnalis. The rostellum here bears an erect boat- 

 shaped disc, filled with a viscid fluid, and decked with a membrane 

 endowed with the power of fissiu-ing on the slightest stimulus (but 

 not spontaneously). The poUinia consist each of two brittle, leaf- 

 like laminae, and are exposed by a contraction of the anther case. 

 The lip, at an early period, moves away from the rostellum, leaving a 

 narrow passage to the nectary. The flowers are visited by bees, 

 which touch the rostellum with their proboscis, causing the boat's 

 deck to burst and expose the viscid fluid which attaches the pollinia 

 to their proboscis. But at the period when the flower is open 

 enough to allow of bees removing the pollinia, the aperture is not 

 sufiiciently wide to allow this to be applied to the stigma. The 

 flower thereafter opens wider by the further movement of the la- 

 bellum ; hence it happens that fully expanded flowers are fertilized by 

 newly expanded ones. The analysis of the whole operation is most 

 graphically given by Mr. Darwin. 



Malaxis paludosa has flowers with the lip turned upwards, owing 



