ss only with
difficulty with related varieties, are not varied by accidental crosses,
and persist through geological periods. Varieties belong to feral nature
rather than to culture; they can assume all possible modifications
without injury to their specific characteristics, but can show no
distinctions of races, for all beginnings of race formation are
destroyed by free intercrossing. They differ from species only in that
they are to be designated as more closely related species, or species as
more remotely related varieties. Every other distinguishing
characteristic is wanting.

_Races_ arise from gamogenic or pathological variations of the
idioplasm. In the former case they presuppose crossing between related
varieties or species, in the latter case an increased sensibility and
weakening of the idioplasm. Very often both causes co-operate, since
crossing follows more easily when the idioplasm is weakened by hurtful
influences and since the irritability and weakening of the idioplasm
increases if crossing has preceded. Race formation begins in single
individuals. Among several individuals it begins in various directions
because the causes are different and hence may display a great
multiformity. Races are distinguished by more or less abnormal
characteristics; they arise quickly--often in a single generation--and
present various degrees of stability. This stability is insured to some
extent only by the strictest in-and-in breeding. All races disappear
through crossing, likewise many races that have arisen from pathological
variations disappear even in sexual reproduction (in self-fecundation).
Races belong exclusively to cultivation, where they can develop and
exist protected from free intercrossing.

While varieties and races arise by progressional or stationary variation
of the idioplasm, _modifications_ are produced by such influences of
nutrition and climate as act only on the soma-plasm and the non-plasmic
substances, and hence do not give rise to inheritable characters in the
organism. Modifications persist only so long as their causes, and under
other environments immediately pass over into the modifications
corresponding to them. The transition is completed in the lowest plants
during a limited number of cell generations; in an individual of the
higher plants on the same stem during the growth of a single year. Each
variety and each race appears clothed in a definite modification, and
can change it within a