72 THEORY OF EVOLUTION 



acteristic of wild types and which at the same 

 time appear to have no survival value, are only 

 hy-products of factors whose most important 

 effect is on another j^art of the organism where 

 their influence is of vital importance. 



It is well known that systematists make use 

 of characters that are constant for groups of 

 species, hut which do not appear in themselves 

 to have an adaptive significance. If we may 

 suppose that tlie constancy of such characters 

 may be only an index of the presence of a 

 factor whose chief influence is in some other 

 direction or directions, some physiological in- 

 fluence, for example, we can give at least a 

 reasonable explanation of the constancy of 

 such characters. 



I am inclined to think that an overstatement 

 to tlie efl^ect that each factor may affect the en- 

 tire body, is less likely to do liarm than to state 

 that each factor affects only a particular char- 

 acter. The reckless use of the phi-ase "miit 

 character" has done much to mislead the imini- 

 tiated as to the effects that a single change in 

 the germ plasm may produce on the organism. 

 Fortunately, the exj^ression "unit character" 



