RIVAL THEORIES OF DEVELOPMENT 455 



embryonic cells, which almost amounts to assuming the differen- 

 tiation which the theory was intended to interpret. 



Weismann tries to avoid this difficulty by assuming an auto- 

 nomic dissolution of the determinant complexes, though he does 

 not reject the view that the differently related vital areas or 

 cells in which the determinants find themselves may serve as 

 liberating stimuli. In a marching army the differently related 

 localities serve as liberating stimuli to the diverse kinds of men 

 composing the army ; here the sappers and miners go to work, 

 there the commissariat erects a depot, in a third place a heliograph 

 is set up, and so on. 



Herbert Spencer's Theory of Physiological Units. — Spencer 

 postulated " physiological units," ultimate life-bearing elements, 

 intermediate between the chemical molecules and the cell. 

 Just as the same kinds and even the same number of atoms may 

 compose, by different arrangements, numerous quite different 

 chemical molecules- — e.g. in the protein-group — so out of similar 

 molecules diversely grouped an immense variety of " physio- 

 logical units " may be evolved, like the variety of patterns in 

 a kaleidoscope. But for each kind of living creature Spencer 

 postulated " physiological units " or " constitutional units " of 

 one kind. 



Spencer credited his " constitutional units " with much. 



1. They carry within them the traits of the species, and even 

 some of the traits of the ancestors of the species ; the traits of the 

 parents, and even some of the traits of their immediate ancestors ; 

 and the inborn idiosyncrasies of the individual organism itself. 



2. They " must be at once in some respects fixed and in other 

 respects plastic ; ' while their fundamental traits, expressing the 

 structure of the type, must be unchangeable, their superficial traits 

 must admit of modification without much dif&culty ; and the 

 modified traits, expressing variations in the parents and imme- 

 diate ancestors, though unstable, must be considered as capable o f 

 becoming stable in course of time." 



3. Moreover, " We have to think of these physiological units 



