Variation as an Organic Function 171 
cell is indefinitely variable. There is evidently a certain elasticity in 
the hereditary type which permits fluctuations within certain limits 
in successive generations without impairing the organism’s capacity 
for reverting to a normal equilibrium position. These fluctuations 
are generally environmentally conditioned, but there is no necessity 
for assuming that this is always the case. 
In the case of determinate variations however, since they are 
specific and not individual phenomena, the extinction of individuals 
will have no evolutionary importance, and natural selection will act 
only as an eliminatory factor in the reactions between different 
species. This will be equally true whether the variations in question 
are relatively large mutations which affect several characters simul¬ 
taneously, the smaller “parvigrade” mutations confined to a single 
character, or the minute cumulative changes known as orthogenesis. 
The effect of hybridisation as an evolutionary factor (by effecting 
various combinations of different characters in bisexual inheritance) 
will also be obviously unimportant in the case of variations of this 
class. 
We shall proceed to a criticism of the theory of determinate 
variation on more general grounds. Now so long as we regard the 
organism as a kind of static aggregate of characters corresponding 
to a geometrical pattern of factors in the germ-cell, it is of course 
possible to postulate that for every change in the germinal complex 
there will be a corresponding change in the somatic complex. This 
view has been elaborated by means of a chemical analogy in the 
“ nucleus and side-chain ” theory of germinal structure, corresponding 
respectively to the more stable ordinal and generic characters in the 
organism, and the less stable specific and varietal ones. It is sup¬ 
ported also by an appeal to alleged correlations in the morphological 
structure of the germ-cell and the soma as revealed by cytological 
and genetic analysis, e.g. sex-characters and sex-chromosomes. But 
as we have seen, morphological characters are only symbols of 
persistent reactions peculiar to a particular type of physiological 
economy. The germ-cell has the dual function of initiating two distinct 
series of events—the germinal and somatic cycles—and it is the 
relation between these two series, and not between the geometrical 
configuration of the germ-cell and the soma at particular stages of 
development, that is of real importance. 
We have also seen that somatic development is always conditioned 
by the persistence of certain physical equilibria throughout ontogeny. 
The organism inherits somehow a particular type of physiological 
