PART III. | THE MECHANISMS OF FLOWERS. 529 
labellum becomes full of loose cells, like flour, which probably 
attract insect-visitors (Fritz Miiller). 
Epidendrum.—tIn South Brazil there are several plain green 
species of this genus, quite without perceptible perfume, which 
secrete abundant honey, and must be fertilised by insects. This is 
an illustration of how flowers need not make themselves perceptible 
at a distance to their visitors by means which affect our senses. 
An observation of my brother’s in the island of St. Catharina 
in South Brazil shows how a plant may propagate itself by self- 
fertilisation when deprived of the insects for which its flowers were 
adapted. On this island an Epidendrum occurs, whose flower con- 
tains three perfect anthers; the two lateral anthers effect self- 
fertilisation regularly, while the pollen of the third can only be 
removed by insects, an event which seems to be exceedingly rare. 
The flower is almost scentless, At Itajahy an Epidendrum occurs 
which resembles the triandrous Epidendrum of St. Catharina almost 
exactly, save that it is monandrous and has a strong, aromatic 
scent. The triandrous variety can only be looked upon as descended 
from specimens of the monandrous form which got transported to 
St. Catharina. There the insects adapted for the flower must have 
been absent or very scarce, so that the power to reproduce by self- 
fertilisation became desirable. Accordingly, whenever the two 
lateral stamens appeared as an abnormality (as often happens 
in other Orchids) they had the greatest possible chance of being 
perpetuated by Natural Selection; the perfume of the flowers, on 
the other hand, being useless and therefore removed from the 
influence of Natural Selection, was lost (533). 
Tribe Neottiew. 
Spiranthes autwmnalis, Rich., has been observed by Darwin to 
be fertilised by humble-bees (155A, p. 127). 
380. ListerA ovaTa, R. Br—Sprengel has described the 
fertilisation of this flower as he observed it in his garden, but he 
-was not acquainted with the insect-visitors. Darwin has given an 
admirable description of the flower, and mentions two Hymenoptera 
(Hemitiles and Cryptus) which he saw attaching the pollinia to their 
foreheads ; as he caught the insects, he did not witness the placing 
of the pollinia on the stigma. My own observations may help to 
confirm and complete the accounts given by Sprengel and Darwin. 
‘On a sunny afternoon in May, 1867, I watched the insect-visitors 
M M 
-~ 
