198 The American Naturalist. [Marca 
ately examine how far these premises accord with a// the phe- 
nomena which we can actually observe in Nature, and then espouse 
the side which is most favored by probabilities. Now I have no 
hesitation in saying that neither side is showing the disposition to 
test their premises by all the observed phenomena, and this is one 
of the most hopeless features of the present situation. 
Variation, Repetition, Regression —A\ll the factors of Evolution 
interact. Variation and Repetition” in inheritance are in con- 
stant relation with every other factor. Thus we can accumulate 
facts as to variations fer se, but if our observation and induction 
enable us to formulate certain laws, these will always involve at 
least two factors, —'i. e., Variation as related to Environment, Vari- 
ation as related to the life-history of individual organisms, Vari- 
ation as related to Inheritance, Variation as related to Natural 
Selection. 
Variability is, of course, exhibited in organisms as a whole, and 
in groups of characters as well as in single characters. All would 
be diversely affected by the two diverse principles of inheritance 
under discussion, but we are to examine the variable tendency as 
exhibited in single characters. Repetition is the conservative or 
static condition wherein a character in the new individual most 
closely resembles the average development presented by the fra- 
ternity,® co-fraternity, race, variety, and species to which it belongs; 
let us adopt Galton’s term “ mediocrity ” for this state of average 
development. . Variation is the unstable or fluctuating condition 
in which a character deviates to either side of mediocrity, either 
in the plus or minus direction,—z. e., to greater or less develop- 
ment. Regression is the tendency” to revert to “ mediocrity”; 
and according to Galton’s statistics we can imagine this law of 
regression as acting like gravitation upon the pendulum of vari- 
ation : when the pendulum swings in one direction it may repre- 
15 Weismann, or his translators, uses the terms Variability and Heredity, as tendencies 
equivalent to ae But it seems to me clearer to use Heredity in the 1 sense, So as 
to includeVariation = the act of Varying, and Repetition = the act of repeating, ancestral 
ers. Variability = ™ tendency to we 
16 Galton. “ Natural Inheritance,” P- 94. e offsrpring of the same mid-parne 
(= male and female) form a a aii All a seia of a Aiari of mid-parentt 
a co- ity. 
1 Op. cit., p. 95. 
Se 
