52 THE FERTILISATION OF FLOWERS. _ [rarr it. 
In Andrena and Halictus, although the chief supplies of pollen 
are obtained by means of the hindlegs, yet in these and in all other 
hairy bees the hairy covering of the body is undoubtedly of advan- 
tage.. In many flowers this gets dusted without any direct effort 
with a considerable quantity of pollen, which is then cleared off by 
means of the tarsal brushes. In almost all bees highly specialised 
for a floral diet, we find the body more or less thickly clothed with 
long feathery hairs. 
The development of the hairy covering is of the highest im- 
portance for the fertilisation of flowers. For as the hairs easily take 
up pollen, they give it up again as easily to viscid or rough stigmas. 
It would far exceed the limits of our space to discuss the adaptive 
modifications present in all the groups of our indigenous bees. 
I can only explain the further development of the pollen-col- 
lecting apparatus and of the lower parts of the mouth by a few 
examples. 
We have seen in Sphecodes, Halitus, and Andrena, how the 
development of pollen-collecting hairs has gradually reached an 
extreme pitch on those parts of the body where the burden is 
least endangered by the movements of flying and creeping; viz. 
on the whole of the hindleg from the tarsus to the coxa, and on 
the hind part of the thorax. In forms higher than Andrena and 
Halictus, this collecting apparatus has attained still greater per- 
fection, in getting more and more restricted to those sections of 
the hindleg to which the tarsal brushes can most readily apply 
the pollen that they have collected, viz. the tarsus and tibia. In 
the highest forms, this localisation of the collecting-hairs has been 
attained without diminishing the mass of pollen, by an increased 
growth of hairs on the tibia and tarsus and a withdrawal of the 
more distal hairs: this we see most clearly in Dasypoda and 
Panurgus. 
In Dasypoda (1, Fig. 14), the collecting-hairs of the tibia and 
tarsus have become so long that they can carry a much greater 
load of pollen than the far more extensive collecting apparatus of 
Andrena pratensis, Nyl. (3, Fig. 13); but the hairs on the femur, 
trochanter, and coxa, are long and thick enough to take a large 
share inthe work. In Panurgus (2, Fig. 14), the collecting apparatus 
consists solely of the long hairs upon the tibia and tarsus. A 
further advance is seen in Hucera and Anthophora, where an increase 
in breadth of the pollen-collecting surfaces of the tibia and tarsus 
compensates for the shortening of the collecting-hairs. 
The last step in the evolution of the collecting apparatus on the 
