PART III. | THE MECHANISMS OF FLOWERS. 449 
all the anther-lobes end alike in short points, in Z. officinalis the 
upper lobe of each anther is blunt, but the lower is provided with 
a thin, stiff spine (3, Fig. 154). Of these four spines it is almost 
exclusively the two lower ones that come in contact with insects; 
for not only are they much longer than the upper ones, but the 
stamens which bear them are inclined at an angle of about 60° 
from the upper pair, and their anthers bend downwards so much 
that the spines stand well down in the mouth of the flower (1, 
Fig. 154). In Odontites serotina the pointed anther-lobes must 
themselves be touched for the pollen to be shaken out, and the 
broad filaments, placed close together and set with sharp points on 
their inner side, insure this, In JZ. officinalis, on the other hand, 
it is sufficient if one of the long spines on the lower anthers be 
touched, and accordingly the filaments curve widely apart, standing 
close to the wall of the corolla; and they are smooth and narrow, 
so that an insect can insert its head into the tube without hindrance. 
In doing so it touches the lower anther-spines, and shakes the whole 
system of anthers, causing the pollen to fall out. Rows of hairs 
on the upper anther-lobes prevent the pollen from being scattered 
at the sides, and insure its falling on the head of the insect. 
In the points described hitherto all my specimens of L. officinalis 
agree with one another. But in regard to the size and conspicu- 
ousness of the flowers, and the possibility of self-fertilisation, I 
have found two different forms of this plant—one large-flowered, 
which never fertilises itself, and one small-flowered, which in 
absence of insects fertilises itself regularly. Intermediate forms 
may probably occur, but were not present among my specimens. 
In the large-flowered form (4, Fig. 154), probably £. montana, 
Jord., the stigma protrudes from the flower before the anthers are 
| ripe, and even after the anthers are ripe it is the first part to be 
touched by insect-visitors, and is, therefore, regularly cross-fertilised ; 
' in absence of insects, it is incapable of self-fertilisation. 
In the small-flowered form, probably #. gracilis, Jord. (1, 
_ Fig. 154), when the flower opens, the stigma stands so far above 
and behind the anthers that it escapes being touched by insccts ; 
__.the style gradually elongates and carries the stigma downwards 
and forwards into a position where it will be touched before the 
anthers by an insect’s head, and where, in absence of insects, 
_ pollen can fall from the upper anthers upon it. 
i, Thus in case of insect-visits cross-fertilisation is insured m both 
forms, but in absence of insects self-fertilisation is only possible in 
_ the small-flowered plants. 
Ge 
