PLANT BREEDING 



367 



375. Mutations. If a variety selected as just described 

 would never change .that is, if its seeds grew year after 

 year into plants which, on the average, when raised under 

 similar conditions, were like the plants that bore the seeds 

 then the work of selection for any particular quality might be 

 done once for all. But as a matter of 

 fact, varieties, no matter how pure they 

 may be at the start, do not remain un- 

 changed. In each generation, the indi- 

 vidual plants of a variety differ from one 

 another in many ways. Most of the 

 differences result from differences in the 

 conditions under which the plants grow ; 

 such differences are not passed on through 

 the seeds to the offspring, and so do not 

 affect the purity of the variety. But 

 now and then a plant differs from its 

 parents or from any of its immediate 

 ancestors in a way that is not caused, 

 at least directly, by external conditions. 

 Such a mutation, as a difference of this 

 kind is called, is due to causes that we 

 do not understand ; unlike the differ- 

 ences that result from external condi- 

 tions, it is likely to be passed on to the 

 offspring. A mutation may therefore be 

 the starting point of a new race. By the 

 appearance of new races in this way, a pure variety in the 

 course of time becomes a mixed variety. Some of the races 

 resulting from mutations are sure to be undesirable. There 

 may occur, for instance, in a large-headed variety of wheat a 

 plant that not only itself bears small heads, but which, if 



reference to certain important qualities. Such a race, while not strictly speaking 

 pure, is thoroughbred in the same sense that a closely inbred race of cattle or horses 

 is thoroughbred. 



FIG. 207. A stool 

 of rye grown at the 

 Wisconsin Experi- 

 ment Station from a 

 single seed of an im- 

 proved variety. 



