236 PLANT-BREEDING 



tion was one supernumerary ray on each of two flower heads 

 of a single plant. Sowing the seeds of this individual, the 

 numbers were seen to increase rapidly but all the ray-florets 

 always belonged to the outer rows of the heads. After a 

 selection of three years, suddenly an individual appeared 

 which had some stray ray florets scattered among the tubular 

 florets of the center of some of its latest flowers. The seeds 

 of this plant at once gave the desired double variety. x\ll 

 the progeny produced more or less double flower heads, the 

 number of the rays ranging, in the best instances, from 

 100 to 200. In the following year, the race could be some- 

 what purified and thereby improved, but the type was not 

 essentially changed and has since remained constant. The 

 appearance of the first ray-florets among the little 

 central tubes would, of course, easily have been overlooked 

 and the increase of the number of the rays would have seemed 

 gradual throughout. The selection would have obscured 

 the mutation, as has, probably, been the case in the produc- 

 tion of many other horticultural varieties. 



