AUG. 1899] THEVSCOPE OF .NATURAL SELECTION 115 
(a) Accommodations which are the direct result of environmental 
influence. 
(6) Accommodations which result from the activity of the organism 
itself in response to its environment. 
It is obvious that these two classes, though not usually so considered, 
are in reality fundamentally distinct. Class (a) includes the only 
kind of inherited characters that can be truly called acquired. 
Class (0) includes what are in reality merely developments of already 
existing somatic tendencies, which some biologists believe may, and 
others that they may not, become germinal. In any case there must 
be an elementary something which can be developed by use or there 
would obviously be no development, but rather the formation of a 
new character, and the accommodation would then have to be classed 
under (a). In class (a) the influence of the environment in producing 
a modification is one of primary cause and effect ; in class (b), on the 
other hand, the influence of environment is secondary, it is the 
indirect cause of the degree of the response, but not of the capacity of 
responding which exists in the particular form of protoplasm itself. 
Class (@) is incompatible with selection, for in proportion as direct 
modification is able to occur, the less is the necessity of selection, and 
this direct climatic influence must obviously be also inversely pro- 
portional to the power of heredity. Class (0), on the other hand, is 
not necessarily in opposition to the selection theory because within 
certain limits the more responsive the organism the greater the rapidity 
of development, selection would become simply more rigorous, the 
selection value would be raised, the less responsive organisms being 
weeded out. 
There are thus two separate questions in this division to be 
answered :— 
1. Does a direct somatic alteration of structure ever occur as the 
result of climatic or other physical influence, and if so, how 
frequently and under what conditions? Do these altera- 
tions become germinal? or 
2. Do all, or any, somatic modifications to environment arise 
as developments of a pre-existing element in protoplasmic 
structure? If so, do somatic responses ever become ger- 
minal? For a clear statement of the Lamarckian position 
it is necessary to determine the relation, if any exists, that 
class (a) has to class (0). 
3. By the selection of organisms which possess favourable varia- 
tions, and by rejection of those which have unfavourable, the offspring 
resulting will tend to reproduce the favourable variations of their 
parents, and the selection being continued in every subsequent 
generation, so long as conditions remain fairly constant, there must 
