The Origin of Chrysanthemum Segetum Plenum. 169 



were removed so early that there was no danger of these 

 15 being fertihzed by them. These plants flowered abun- 

 dantly from their lateral shoots but exhibited no tendency 

 to form a curve with an apex at 21. They were there- 

 fore sufficiently pure representatives of the supposed 

 race. 



In September I harvested the seeds of the 13-rayed 

 plants which I had spared, and sowed half of them in the 



12 13 



A 



17 18 19 20 21 22 23 24 25 26 27 28 



B 



Fig. 31. A. Chrysanthemum segetum. B. Chrysanthemum 

 segetum grandiRorum (after purification). Curves of 

 the races after isolation. A, Curve of the 13-rayed race 

 in 1894. B, Curve of the 21-rayed race in 1897. The 

 ordinates give the number of individuals with like 

 number of ray-florets in the primary inflorescences of 

 the individual plants. The numbers of ray-florets them- 

 selves are given below the abscissa. 



following spring (1893). I raised 162 flowering indi- 

 viduals, and recorded the numbers of rays on their ter- 

 minal heads. The curve representing this generation 

 was steep, monomorphic and symmetrical (see Fig. 31 A 

 for 1894), and agrees satisfactorily with the curves, 

 given above, for the plants from the wild locality (p. 167 

 and Fig. 32, Vol. I, p. 152). Therefore there can be no 



