Decembee 2, 1921] 



SCIENCE 



531 



Much of the endless confusion and intermi- 

 nable controversies about the inheritance of 

 so-called " acquired characters " is due to the 

 neglect of this important distinction. For it 

 is quite clear that whereas factors may be 

 transmitted, characters as such never are. The 

 characters of the adult, being responses, are 

 not present as such in the fertilized ovum 

 from which it develops, they are produced 

 anew at every generation.* No distinction in 

 kind or value can be drawn between charac- 

 ters. 



If some are inherited regularly and others 

 are not, the distinction lies not in the nature 

 or mode of production of the characters them- 

 selves, but in the constancy of the factors and 

 conditions which give rise to them. Thus, 

 although there is only one kind of character, 

 there are two kinds of variation. 



Much of the confusion in evolutionary lit- 

 erature is, I think, due to the use of the word 

 variation in a loose manner. Sometimes it is 

 taken to mean the degree of divergence be- 

 tween two individuals; sometimes the char- 

 acter itself in which they differ, such as a 

 color or spot on a butterfly's wing, at other 

 times a variety or race differing from the 

 normal form of the species. If clearness of 

 thought and expression is to be attained, the 

 word variation should mean the extent or de- 

 gree of difference between two individuals or 

 between an individual and the average of the 

 species, the divergence of the new form from 

 the old; not a new character or assemblage of 

 characters, but a difference which can be 

 measured or at least estimated. We shall then 

 find that a variation is of one of two kinds 

 (which may, of course, be combined) : the 

 first kind is due to some change in the com- 

 plex of germinal factors. 



The second kind, to which the name muta- 

 tion has been applied, will, under constant con- 



* In other words, all characters are ' ' acquired 

 during the lifetime of the individual, ' ' and ' ' in- 

 herited ' ' in the sense here defined has just the 

 same meaning. Much the same view was advo- 

 cated by Professor A. Sedgwick in his address to 

 this Section at Dover in 1899, and it has also been 

 developed by Dr. Archdall Eeid and others. 



ditions, be inherited since the new complex 

 of factors will be transmitted to subsequent 

 generations. The first kind of variation, 

 which has been called a modification, will also 

 be inherited, provided, of course, the change 

 of stimulus persists. In either case, new 

 characters will result. But here, again, we 

 must be careful not to apply the terms muta- 

 tion and modification to the characters them- 

 selves, as is so often done ; ^ for we then re- 

 introduce the confusion already exposed in the 

 popular but misleading distinction between 

 " acquired " and " non-acquired " characters. 

 The characters due to mutation or modifica- 

 tion are, of course, indistinguishable by mere 

 inspection, and can only be separated by ex- 

 periment. A mutation once established should 

 give rise, under uniform conditions, to a new 

 heritable character, and may be detected by 

 crossing with normal members of the species. 



So far observations and tests have shown 

 that new characters due to modification only 

 reappear so long as the new stimulus persists. 

 The difference lies not in the value or per- 

 manence of the new character, but in the 

 causes which give rise to it." 



It is little more than a platitude to state 

 that, for the production of an organism or of 

 any of its characters, both germinal factors 

 and environmental stimuli are necessary, and 

 that if evolution is to take place there must be 

 change in one or both. Yst the changes in the 

 factors may be held to be the more important. 



5 The name ' ' mutation ' ' might be given to the 

 alteration in the factors instead of the variation 

 due to it. The latter might then be termed a muta- 

 tional variation and would be opposed to a modifi- 

 cational variation. At present the term " muta- 

 tion " ia applied to three different things: the 

 factorial change, the variation or difference, and 

 the new product response or character. 



8 We might perhaps distinguish the two cases by 

 calling them constant 'and inconstant characters, or 

 ' ' natural ' ' and ' ' acquired, "as is commonly done 

 when describing immunity. It should be meant 

 thereby that one is acquired usually (under normal 

 conditions), the other occasionally (when infec- 

 tion occurs). Error creeps in when the term 

 ' ' acquired ' ' is opposed to ' ' non-acquired " or to 

 ' ' inherited. ' ' 



