organisation or on certain parts alone and in directly by affecting the
reproductive system. With respect to the direct action, we must bear in
mind that in every case, as Professor Weismann has lately insisted, and as
I have incidently shown in my work on "Variation under Domestication,"
there are two factors: namely, the nature of the organism and the nature
of the conditions. The former seems to be much the more important; for
nearly similar variations sometimes arise under, as far as we can judge,
dissimilar conditions; and, on the other hand, dissimilar variations arise
under conditions which appear to be nearly uniform. The effects on the
offspring are either definite or in definite. They may be considered as
definite when all or nearly all the offspring of individuals exposed to
certain conditions during several generations are modified in the same
manner. It is extremely difficult to come to any conclusion in regard to
the extent of the changes which have been thus definitely induced. There
can, however, be little doubt about many slight changes, such as size from
the amount of food, colour from the nature of the food, thickness of the
skin and hair from climate, etc. Each of the endless variations which we
see in the plumage of our fowls must have had some efficient cause; and if
the same cause were to act uniformly during a long series of generations on
many individuals, all probably would be modified in the same manner. Such
facts as the complex and extraordinary out growths which variably follow
from the insertion of a minute drop of poison by a gall-producing insect,
shows us what singular modifications might result in the case of plants
from a chemical change in the nature of the sap.

In definite variability is a much more common result of changed conditions
than definite variability, and has probably played a more important part in
the formation of our domestic races. We see in definite variability in the
endless slight peculiarities which distinguish the individuals of the same
species, and which cannot be accounted for by inheritance from either