102 NATURAL SCIENCE [August 
actually do in the way of visiting flowers is of more importance to 
the theory than any such experiments; and, as we shall see pre- 
sently, bees in visiting flowers do not show any marked preference 
for blue ones, 
The analogy between man’s selection resulting in artificial breeds, 
and natural selection resulting in species, was often and strongly in- 
sisted on by Darwin; while the action of the insect in evolving a 
new species of flower specially adapted to its requirements in form 
and colour, has been compared to that of the gardener in the pro- 
duction by selection of his special varieties. The extreme care re- 
quired from the gardener, not only in the selection of the requisite 
variation, but also in its isolation and preservation is well known. 
If the new variety in the process of its evolution be allowed to cross 
with the parent stock it is lost. Darwin even went so far as to 
warn gardeners against allowing crossing between different indivi- 
duals of the new variety, as such tends to produce reversion. It is 
a little difficult to understand how the bee, even if it possessed all 
that nice discrimination of form and colour, and all that constancy 
required by the theory, could accomplish that which demands such 
care and patience on the part of the gardener. In order to evolve a 
race of blue flowers from those normally of another colour, bees 
would have to select those among the slight natural variations which 
had a tinge of blue. If they really preferred to gather honey from — 
blue flowers—they would not have merely to prefer blue in the ab- 
stract—they would do so, and these bluish flowers would receive the 
benefits of cross-fertilisation. But the few bluish flowers among the 
many normal ones would not suffice: the bees, from the necessity of 
obtaining honey, would be obliged to visit the normally-coloured 
ones also. Thus the new variety would be blended with the parent 
type and lost, for pollen would be carried from the one to the other. 
In order to isolate the bluish variety, the bee would not only have 
to prefer blue, but also steadily refuse to visit any other colour ; 
and if it did so when the bluish flowers were first appearing it 
would starve for want of honey. But we have no proof that the 
bee even possesses the taste. In all my observations of bees I 
have met with nothing to support the view that they prefer to take 
their honey out of blue flowers. Some blue flowers they visit fre- 
quently, others they visit very seldom. No blue flowers are more 
frequently visited than others which are yellowish-green, pink, and 
various other shades. Some uncompromisingly green flowers—as 
the plane tree and red currant—are frequently visited. As regards 
colour, then, the bee seems to have neither the taste to select, nor 
the ability—through the necessity of obtaining sufficient honey—to 
restrict itself in the manner required. Moreover, the fact that the 
same species of bee may be seen visiting flowers of the most diverse 
