776 The American Naturalist. [August, 
mechanical principles; those that admit a mental source; and those 
that admit both. According to the first, instinct depends upon habit ; 
according to the second, upon selection ; according to the third, upon 
both. The common point of departure of all these theories is found in 
the generalization of habit and memory and their union in the concep- 
tion of heredity. Habit is not limited to the individual but its results 
are inherited by descendants. 
As the type of the mechanical theories, we may take that of Spencer. 
Instincts are due to complications of reflexes, and this complication is 
simply an illustration of the most general law of evolution, which in- 
volves progressive increase in heterogenity and complexity of corre- 
spondence. But this is merely a statement of a fact and not an expla- 
nation of it. We wish to know the reason why, and the method in 
which this complication takes place. 
The mental theories fall into two classes. The one, represented by 
that of Lewes, regards the instinct as a degraded form of intelligent 
act. This doctrine is discredited by the fact that it would require the 
parallel assumption that the nervous system of the lower animals is de- 
graded from a more complex form capable of manifesting the higher 
forms of intelligence. The second class, represented by that of Fouil- 
lée, merely translates into mental terms Spencer’s mechanical notions. 
Mind stuff takes the place of Force, but the details are essentially the 
same, and again the question arises, how and why can combinations of 
mind stuff bring about the new creations which we see? 
None of these theories afford any true explanation of the phenomena. - 
They bring to view the points of resemblance and difference between 
the instinct, the reflex and the voluntary act, but they do no more. 
But the most interesting of the questions that arise in connection 
with instinct is that of its mode of development. For the solution of 
this problem we are indebted to Darwin, who has shown that it is due 
to variation and selection. Yet it should be noted that this does not 
reduce the development of instinct to a purely mechanical process, 
which was Spencer’s error. The variations are not physical so much 
as mental, nor are they absolutely predetermined. The conditions that 
make them possible must be given, such as antecedent and concomi- 
tant mental states, but this does not determine their occurrence, since 
they may or may not occur. If they occur, the organism adapts itself 
to its environment and survives ; if not, it does not adapt itself and be- 
comes extinct. This introduces the last question to be considered, that 
is, what is the character of these mental variations that underlie the de- 
velopment of instinct ? 
