84 THE FERTILISATION OF FLOWERS. [PART III. 
it forces the humble-bees to suck the honey by the way that alone 
leads to fertilisation. 
The two upper petals serve a very different purpose. In each the 
hollow pointed end of the spur (a?, 3, 5), which is directed back- 
wards and is inclosed in the hollow sepaline spur, secretes honey 
and becomes so full of it that part rises into the wide, half-conical * 
part of the spur which is open on the inner side (@’, 5). When 
both petals lie close to one another, they form together a hollow 
cone, which splits at the end into two points filled with honey, 
and guides the insect’s proboscis, if long enough, safely to the 
honey, while by its length it prohibits the access of insects with 
shorter tongues. The anterior portions of the same petals 
produce the upper part of this hollow cone further forwards, and 
serve to give the bee’s proboscis a convenient entrance and more 
certain path to the honey. Since these anterior portions of the 
upper petals separate easily on slight pressure from within, the 
whole head of a humble-bee may be thrust in between them, 
whereby the distance to the honey is shortened by 6 to7 mm. The 
length of the hollow cone from its entrance to the anterior end of 
the honey-bearing prolongation is about 20 mm., and to the apex of 
the latter 26 to 28 mm.; so that, when the whole bee’s head is 
thrust into the aperture, a proboscis 13 to 14 mm. long is needed 
to reach the honey, and one 19 to 22 mm. long to suck it all up. 
Hence, of all our native bees, Anthophora pilipes, F., and Bonbus 
hortorum, L., can alone exhaust the honey. 
The two lower petals are of service in several ways. Their 
anterior surfaces point by bundles of upright yellow hairs towards 
the entrance to the honey, that is to say, they serve as path- 
finders; and as they stand close together and bound this entrance 
below, they leave the bee no choice but to thrust its proboscis into 
the only proper place, viz., the mterval between the two pairs of 
petals. Those parts of them immediately behind the entrance to 
the tube stand, on the other hand, so widely apart (*, 2) that they 
leave free space for the stamens and for the carpels (after the 
stamens have withered and bent back) to erect themselves in that 
part of the hollow cone lying close behind the entrance, where they 
inevitably come in contact with the under surface of the body and 
head of the bee. 
With sufficient insect visits, which this handsome plant never fails 
to receive, cross-fertilisation is completely insured by the proter- 
androus dichogamy, and by the movements of stamens and earpels 
} 7.e. forming the longitudinal half section of a cone, — 
