es were seen visiting the flower;
by way of getting at the honey, they set to work to gnaw off the ridges
of the lid above alluded to; in doing this they pushed one another into
the bucket, and had to crawl out by the spout. As they passed out by
this narrow aperture, they had to rub against the anthers and so carried
off the pollen. When a bee so charged gets into another bucket, or into
the same bucket a second time, and has to crawl out, he brushes against
the stigma, and leaves the pollen on it. I might well have adduced this
plant as another instance of the first objection, since it may well be
asked, How could such a development, resulting in a structure which
presents the greatest difficulty in the way of fertilization, be
beneficial to the plant? But here the point is that, even if any one
could assert the utility of such an elaborate and complicated
development, and suppose it self-caused by accident or effect of
environment, it certainly goes against the idea that all forms are due
to an _accumulation of small changes_. For these curious contrivances in
the case of _Salvia, Coryanthes_, and other plants, would in any case
have been no use to the plant till the whole machinery _was complete_.
Now, on the theory of slow changes gradually accumulating till the
complete result was attained, there must have been generation after
generation of plants, in which the machinery was as yet imperfect and
only partly built up. But in such incomplete stages, fertilization would
have been impossible, and therefore the plant must have died out. Just
the same with the curious fly-trap in _Dionoea_. Whatever may be its
benefit to the plant, till the whole apparatus as it now is, was
_complete_, it would have been of no use. In the animal kingdom also,
instances might be given: the giraffe has a long neck which is an
advantage in getting food that other animals cannot reach; but what
would have been the use of a neck which was becoming--and had not yet
become--long? here intermediate stages would not have been useful, and
therefore could not have been preserved.[2] In flat fishes it is curious
that, though they are born with eyes on different sides of the head, the
lower eye gradually grows round to the upper-side. As remarked by Mr.
Mivart, natural selection could not have produced this change, since the
_first steps towards it_ could have been of no possible use, and could
not therefore have occurred, at least not without direct