1897] BEES AND THE DEVELOPMENT OF FLOWERS 103 
shapes, from the simplest to the most complex, does not seem to 
imply that they have the requisite selective tendency as regards 
form. This is further illustrated by the fact that bees may at times 
be seen visiting flowers which have lost their corollas wholly or in 
part. Thus I have seen them visit petal-less flowers of wild 
geranium, bramble, and cistus. Darwin relates his observation of 
the same fact. More recently Professor Plateau of Ghent has 
removed the corollas of certain flowers, and found that this proceed- 
ing made little difference to the insects Visiting them. Bees may 
also be seen to visit abnormally developed, as well as half-faded 
flowers. In many cases, again, bees instead of using the form of the 
flower supposed to be specially fitted for their convenience, and the 
outcome of taste, will bite a hole near the base of the corolla, and 
get the honey through it. These holes may frequently be seen in 
heath and the bush vetch. 
Again, if our native flowers are the result of the selective action 
of our native bees, and those which they have specially chosen for 
countless generations, how is it that bees take so readily to many 
flowers of very different forms introduced into our gardens from 
abroad ? For such introduced plants are in many cases freely visited 
by native bees. 
In order to evolve and keep distinct new species bees would 
have to be extremely constant in their visits to flowers: in a single 
journey from the nest, or until they got rid. of all the pollen 
adhering to their bodies, they would have to visit only a single 
variety. If they did not do so they would not merely be unable to 
develop and differentiate new varieties; they would even retard by 
intercrossing, varieties developing into species by any other means. 
It is pretty generally believed that the bee is very constant in its 
visits to flowers, and that when it begins with any particular species 
it keeps to that until it has obtained its load. So long ago as the 
time of Aristotle, indeed, the constancy of the bee was noted as a 
fact in natural history. But while it is true that bees do show a 
considerable amount of constancy and often visit a large number of 
flowers of the same species in succession, they are far from possessing 
that amount of constancy required by the theory. For this they 
would require to restrict themselves, not merely to a single species, 
but to one variety of that species. This is obvious, since all species 
are supposed to have begun as varieties ; and it is even more import- 
ant that they should restrict themselves to one variety than to one 
species, since such varieties will be more readily crossed by transfer- 
ence of pollen. But it is a well-established fact that bees pass freely 
from variety to variety of the same species in our gardens. Darwin 
has observed this, and it is one of the most firmly established results of 
my own observations, They do not even confine themselves in a single 
