emale. But according to the evidence derived from a study of the
sex-limited descent of certain features in other animals the
conclusion seems equally clear that in them female must be regarded as
DR and male as RR. The eggs are thus each either male or female and
the spermatozoa are indifferent. How this contradictory evidence is to
be reconciled we do not yet know. The breeding work concerns fowls,
canaries, and the Currant moth (_Abraxas grossulariata_). The
accessory chromosome has been now observed in most of the great
divisions of insects,[70] except, as it happens, Lepidoptera. At first
sight it seems difficult to suppose that a feature apparently so
fundamental as sex should be differently constituted in different
animals, but that seems at present the least improbable inference. I
mention these two groups of facts as illustrating the nature and
methods of modern genetic work. We must proceed by minute and specific
analytical investigation. Wherever we look we find traces of the
operation of precise and specific rules.

In the light of present knowledge it is evident that before we can
attack the Species-problem with any hope of success there are vast
arrears to be made up. He would be a bold man who would now assert
that there was no sense in which the term Species might not have a
strict and concrete meaning in contradistinction to the term Variety.
We have been taught to regard the difference between species and
variety as one of degree. I think it unlikely that this conclusion
will bear the test of further research. To Darwin the question, What
is a variation? presented no difficulties. Any difference between
parent and offspring was a variation. Now we have to be more precise.
First we must, as de Vries has shown, distinguish real, genetic,
variation from _fluctuational_ variations, due to environmental and
other accidents, which cannot be transmitted. Having excluded these
sources of error the variations observed must be expressed in terms of
the factors to which they are due before their significance can be
understood. For example, numbers of the variations seen under
domestication, and not a few witnessed in nature, are simply the
consequence of some ingredient being in an unknown way omitted from
the composition of the varying individual. The variation may on the
contrary be due to the addition of some new element, but to prove that
it is so is by no means an easy matter. Casual observation is useless,