TJic Origin of Chrysanthctuiiin Scgctum Plciiuiii. 183 



in this year was offered by a plant with 48 rays which 

 was healthy enough to be chosen as a seed-parent. But 

 this hgin-e lies very close to the next figure in the series 

 (13 + 34 = 47). The maximum at 21 has disappeared, 

 but the form of the curve clearly indicates its participa- 

 tion in the composition of the 

 whole. 



In the following year the 

 advance was much less consid- 

 erable. The maxima at 26 and 

 34 and that near 47 became 

 more distinct, but the maxi- 

 mum number of rays increased 

 to 67. At the same time an- 

 other still more important dif- 

 ference appeared since now for 

 the first time ligulate florets 

 appeared between the tubular 

 florets of the disc. This only 

 occurred on a single plant and 

 not till the beginning of Sep- 

 tember. This plant had 66 

 rays in its terminal inflores- 

 cence, and was one of those 

 which had been selected as seed- 



l^arents, and accordingly isolated at the beginning of the 

 flowering period. On account of its possession of this 

 first sign of real doubling it was chosen for the continua- 

 tion of the experiment in 1900, to the exclusion of all 

 the rest. 



Tt is well known that in other species of this genus 

 (e. g., Chrysaiiflicniuiii ludicum and C. iiiodoniiii) tlie 

 doubling consists in exactly the same phenomenon. In 



Fig- 33- Chrysanthemum se- 

 gefiim t>lcnum. One of 

 the six inflorescences 

 which in 1899 first exhib- 

 ited true ''doubHng." The 

 figure represents the par- 

 ent plant of the "double" 

 variety. 



