changed conditions of life. The character would then be thrown over from
the old state of equilibrium into a new one.

Characters which happen to be in this state of unstable equilibrium are
called mutable. They may be either latent or active, being in the former
case derived from old active ones or produced as new ones (by the process,
designated premutation). They may be inherited in this mutable condition
during a long series of generations. I have shown that in the case of the
evening primrose of Lamarck this state of mutability must have existed for
at least half a century, for this species was introduced from Texas into
England about the year 1860, and since then all the strains derived from
its first distribution over the several countries of Europe show the same
phenomena in producing new forms. The production of the dwarf evening
primrose, or Oenothera nanella, is assumed to be due to one of the factors,
which determines the tall stature of the parent form, becoming latent; this
would, therefore, afford an example of retrogressive mutation. Most of the
other types of my new mutants, on the other hand, seem to be due to
progressive mutability.

The external causes of this curious period of mutability are as yet wholly
unknown and can hardly be guessed at, since the origin of the Oenothera
Lamarckiana is veiled in mystery. The seeds, introduced into England about
1860, were said to have come from Texas, but whether from wild or from
cultivated plants we do not know. Nor has the species been recorded as
having been observed in the wild condition. This, however, is nothing
peculiar. The European types of Oenothera biennis and O. muricata are in
the same condition. The first is said to have been introduced from
Virginia, and the second from Canada, but both probably from plants
cultivated in the gardens of these countries. Whether the same elementary
species are still growing on those spots is unknown, mainly because the