of the transformation of species. Certainly those who read Mr.
Darwin's own works attentively will find no lack of fluctuation in
his case; and reflection will show them that a theory of evolution
which relies mainly on the accumulation of accidental variations
comes very close to not entering upon the causes or means of the
transformation of species. {33b}

I have shown, however, in "Evolution, Old and New," that the
assertion that Buffon does not enter on the causes or means of the
transformation of species is absolutely without foundation, and that,
on the contrary, he is continually dealing with this very matter, and
devotes to it one of his longest and most important chapters, {33c}
but I admit that he is less satisfactory on this head than either Dr.
Erasmus Darwin or Lamarck.

As a matter of fact, Buffon is much more of a Neo-Darwinian than
either Dr. Erasmus Darwin or Lamarck, for with him the variations are
sometimes fortuitous. In the case of the dog, he speaks of them as
making their appearance "BY SOME CHANCE common enough with Nature,"
{33d} and being perpetuated by man's selection. This is exactly the
"if any slight favourable variation HAPPEN to arise" of Mr. Charles
Darwin. Buffon also speaks of the variations among pigeons arising
"par hasard." But these expressions are only ships; his main cause
of variation is the direct action of changed conditions of existence,
while with Dr. Erasmus Darwin and Lamarck the action of the
conditions of existence is indirect, the direct action being that of
the animals or plants themselves, in consequence of changed sense of
need under changed conditions.

I should say that the sketch so often referred to is at first sight