1899] METHOD OF TREATING VARIATIONS 419 
is said that we can only tell this by the result, when the “ fittest” has 
proved itself, when the “selection” has already taken place, this simply 
means that we are unwilling or unable to understand the present. In 
reality there cannot be two opinions. A plus variation may be “useful” 
to-day, “harmful” to-morrow, and similarly for a minus one. 
It will be said, however, that “plus” and “minus” do not exist in 
nature, and that the most “useful” or most “favourable” is the mean 
or average condition. How then shall we understand the continual 
recurrence of variations from the mean? As shown by Galton, the 
tendency of successive generations is to produce offspring nearer the 
average than the parents, and hence, on the theory of the survival of 
the “ fittest,” or most “favourable” variations, we are unable to explain 
the presence of extremes. 
Natural selection is based on (1) the rate of increase of offspring, 
(2) enormous destruction and “ struggle,” (3) survival of the best fitted 
to the conditions. If we let our minds run smoothly in the train of 
thought suggested by the form of the premises, the conclusion is 
inevitable. But if we inquire more closely into the “struggle” and 
“destruction” in any particular case, say of the eggs and young of the 
herring in the sea, we find ourselves obliged to consider this destruction 
as indiscriminating and independent of struggle, and that consequently 
both the “ fit” and “unfit” survive; in other words, we cannot apply 
these latter words even “ metaphorically.” 
If natural selection thus fails to interpret present, phenomena as 
shown in the variations of single organs, its difficulties increase when 
we consider the individuals. From the study of organs we might con- 
clude that Nature was aiming at the conservation of the average, but 
when we examine many organs we find that the average of one may 
be combined with the extremes of others. Hence greater fitness in 
this or that has to make up for greater unfitness here or there. In the 
same region the individuals of a group at the same period of life are 
thus equal in the combination of their characters. This has been shown 
to be a theoretical deduction from the first proposition, and it has been 
exemplified under propositions II. and III. Any conception of greater 
or less fitness is here completely excluded. 
When we turn, however, to different regions and consider different 
groups of the same species we find that the average of the individuals 
has changed, and if we examine successive groups in successive regions 
we find intermediate stages of the averages. It might be thought, 
then, that natural selection has brought about these differences in the 
different regions. If so, then we must change our conception of 
natural selection because there is little or no “struggle” between the 
different groups, consequently no discriminating destruction and no 
“survival of the fittest.” 
It is not to the present purpose to criticise further the theory of 
natural selection, or plead for the theory of probability. This latter, 
