EVOLUTION BY CHOICE 461 
tevant and others, the direct effects of external agencies like 
the effects of selection are confined to modifying types rather 
than originating them. 
Those of my readers who have played with a kaleidoscope 
will remember that as the cylinder is moved slowly forward 
or back gradual changes in the design take place, and any © 
favorite arrangement may be recovered by simply moving 
the cylinder back to the place where that arrangement ap- 
peared,—all this being possible so long as the cylinder does 
not move beyond a certain point; for if it gets ever so little 
beyond that point there is a sudden rearrangement of the 
elements thereby forming an entirely new design, which may 
in turn be modified as before by restricted changes of position. 
The gradual modification of the design within definite limits 
is like the modification of a type as effected by fluctuating 
variations or acquired characters; the sudden change is like 
a mutation upsetting the previous equilibrium and estab- 
lishing a new equilibrium which is not at all disturbed by 
vacillating modifications. 
170. Evolution by choice. To make the foregoing anal- 
ogy complete we should have to imagine a kaleidoscope with 
the power of self-movement; for whatever may be the factors 
which bring about mutations, the process is somehow influ- 
enced from within. A living thing is active as well as passive. 
The idea is thus suggested that organic evolution may have 
as its controlling factor some power of choice, essentially like 
our own, residing in all living organisms—a will as truly free, 
although apparently very different because exercised under 
very different conditions. This is a hard saying, but perhaps 
we shall find it to contain important truth. 
Doubtless to many readers the idea of plants willing or 
choosing in any way whatever will appear quite absurd. 
“‘ How is it possible,” they will urge, ‘‘to conceive of voluntary 
action in vegetable life?’’ Let us try to consider the matter 
without prejudice. Surely, as we watch plants they seem to 
act spontaneously, to improve opportunities, and, some of 
them at least, appear to have gained experience. All ob- 
servers would agree that a climbing shoot or a root-tip acts 
almost as if it were intelligent. If the reader will admit that 
