C. W. SOAL 
I 7 8 
whatever of large mutations that give rise at a single step to new 
and efficient types of physiological economy. 
What therefore is the relation between these two distinct classes 
of specific variations? It is suggested that the relatively large and 
discontinuous changes in superficial characters may have little or no 
evolutionary significance in themselves, but that they are correlated 
with the minute progressive changes in the more stable organic 
functions. As an illustration, may be mentioned the effects of slight 
changes in the amount or composition of the internal glandular 
secretions upon the morphological structure of animals, a factor 
which Keith considers to have been very important in the differentia¬ 
tion of the human species. It is possible of course that such correlated 
variations may, in some cases, be themselves of the orthogenetic 
type. 
To recapitulate, there are four main classes of heritable variation: 
(1) Minute cumulative changes in the more stable organic func¬ 
tions which are mutually compensative and tend generally to increase 
the efficiency of the organism, as an energy-system. 
(2) Correlated with the former are variations in the physio¬ 
logically less important characters, often relatively large and dis¬ 
continuous, which account chiefly for morphological differentiation 
in allied species and genera. 
(3) Sporadic individual mutations affecting unimportant char¬ 
acters. These may be selected and so give rise to specific adaptations 
and domestic varieties. 
(4) Abnormal sporadic variations in important and phylo- 
genetically stable characters which are always harmful and generally 
eliminated by natural selection. 
Variations of the first class must be regarded as by far the most 
important factor in evolution, since they set the physiological type 
and maintain its stability. They play the same role in phylogeny 
that functional adaptation plays in a single life. The primary environ¬ 
ment is in both cases very important, but whereas functional 
adaptations are compensative readjustments with respect to ephe¬ 
meral changes in external conditions, the trend of germinal variation 
can be influenced only by changes in average conditions that persist 
for a considerable period and so permanently modify the con¬ 
ditioning factors in organic equilibrium. The rate at which such 
specific readjustments can be effected must, however, be regarded 
as a property of the type; consequently when external conditions 
change too rapidly, a phylogenetic readjustment may become im- 
