373 
Now, in the first place, it should be observed that this 
instance is really not a case of Natural Selection—i.e., of 
selection by life and death—but selection by separation, 
which is a very different thing, for selection in nature is 
not necessarily Natural Selection. In the second place, 
the explanation assumes that the individuals composing 
a species can always be divided into two classes, one 
containing those which can see perfectly and the other 
containing those which can see so imperfectly that they 
will not move in the direction of a glimmer of light. 
That is a large assumption to start with. Then, in order 
that this difference should lead to the “ selection ” which 
is supposed to take place, it is necessary that something 
like a catastrophe should happen, sweeping a group of 
animals away from their accustomed habitat. In these 
circumstances, it is reasonable to believe that all would 
desire to return to their accustomed sphere and to retrace 
their steps, and those with imperfect vision must be almost 
blind to begin with, if a glimmer of light was not visible 
to them. There are other organs of sense besides that of 
sight, and on the hypothesis that some individuals were 
carried not only into deep sea but also into the depths 
of the sea, they would surely seek to mount to the surface 
if indeed the sudden change did not prove fatal to them 
altogether. But we will assume that the individuals with 
imperfect sight do remain in the new habitat and have 
offspring. In that case, according to the principles of 
Pure Darwinism, which Dr. Ray Lankester adopts, the off 
spring would not necessarily inherit the imperfect vision, 
but only a tendency in that direction ; and it is reasonable 
to suppose that ere that tendency were developed the 
offspring would rise to seek the light. Such are the diffi 
culties which beset this explanation, if we regard it as a 
selection in nature which is not Natural Selection. If this