482 THE FERTILISATION OF FLOWERS. [PART III. 
bite a hole in the superior side of the tube immediately over the 
nectary and steal the honey. Dr. Ogle found 90 per cent. of the 
flowers bitten through, and I have always found the flowers bitten 
through and robbed of their honey by an Alpine robber-bee, Bombus 
mastrucatus, Gerst. % (609). 
Salvia nilotica, Vahl. (345, figs. 24, 25)—The main point in 
which the flowers differ from those of S. officinalis is that the two 
inferior anther-lobes lie freely side by side, and the connectives 
can be caused to revolve separately. Hildebrand saw them visited 
by bees. 
S. verticillata, L. (345, figs. 26-30; 172), has immovable con- 
nectives, but an upper lip which folds back when touched by an 
insect-visitor, and exposes the two superior anther-lobes to contact 
with it. The style, which would hinder this movement if in its 
usual place, is lower down. Hildebrand saw this species also 
_visited by bees. According to Delpino, the anthers, like those of 
Sideritis and of S. officinalis, are provided with sticky glands (178, 
p. 145). S. verticiilata is visited by numerous species of bees and 
humble-bees (590, III. ; 609). 
In S. patens, Cov. (172, 345, fig. 31), the anthers project partly © 
or entirely beyond the upper lip. The connectives are versatile, 
and the lower anther-lobes are metamorphosed into lamellw; the 
style is so fixed between the upper arms of the connectives that 
when the connectives revolve it is carried forwards and downwards 
with them, and its stigma, which projects beyond the anthers, is 
thus brought first in contact with the insect’s back. Ogle’s account 
(631) agrees with Hildebrand’s. But while Hildebrand considers 
self-fertilisation and cross-fertilisation as alike possible, Ogle shows 
that cross-fertilisation is insured. When an insect-visitor strikes 
the lower arm of the connective, its back is touched by the anthers 
and a little farther back by the stigma, and the space between the 
two points is increased by the lower stigma being very short in 
comparison with the upper. As the insect passes further in, 
anthers and stigma rub along its back, but no pollen from the 
anthers reaches the stigma of the same flower. As the insect 
draws back anthers and stigma return to their places below the 
upper lip; in the next flower the stigma comes in contact with 
a part of its back already dusted with pollen. In several flowers 
Dr. Ogle found the style shorter than the anthers; in such 
flowers insects might lead to self-fertilisation as well as cross- 
fertilisation. Ogle states the very surprising fact that the honey — 
in S. patens is not secreted by the base of the ovary as is usually 
—— 
