SECTION III 



THE RACE 



AT intervals throughout the preceding pages we have had 

 to resort to evidence drawn from the average behaviour of 

 groups of plants, in default of "precise data from in- 

 dividuals. In other words, we have considered a group as 

 representing an average individual. These groups have 

 now to be considered for their intrinsic interest. 



Such groups may be derived from a "pure strain," 

 being all of identical gametic composition ; in this case we 

 find material to study in the fluctuation, which is the 

 result of slight irregularities in a normally uniform 

 environment. Again, the groups may be derived from 

 the crossing of two pure strains, in which case their 

 gametic composition may not be uniform ; we have then 

 to apply corrections for fluctuation, and having done this 

 we search in the residuum for evidence of gametic 

 differences wherewith to formulate expressions for the 

 inheritance of each characteristic. 



Lastly though it will be more convenient to discuss 

 this subject in the second place the group may consist of 

 many different pure strains, mingled with hybrids between 

 them of every generation imaginable. Such is the com- 

 position of the commercial " varieties." Within limits, 

 and for certain purposes, we may consider such varieties 



89 



