174 Observation of the Origin of Wirieties. 



niinal inflorescences. Tlie result proved my view to be 

 correct and showed the necessity of the correction which 

 it had suggested, h'or there were 22 plants which, al- 

 though their terminai inflorescences were 21 -raved, had 

 a part-curve with an apex at L3-14. The following 

 are the data as obtained at the end of August : 



L. F. 12 13 14 15 16 17 18 19 20 21 



Lateral flowerheads of 22 indiv. 2 54 58 51 28 19 19 12 2 2 



These plants therefore belonged to the 13-rayed race, 

 and were consequently eradicated. 



Besides these, there were five plants with doubtful 



curves ; tliey were also not retained. All that was left 



was a group of 6 individuals whose curves seemed to me 



sufficiently distinct and certain to justify the harvesting 



of their seed. The following line gives the sum of their 



data : 



L. F. 12 13 14 15 16 17 18 19 20 21 22 



Lateral flowerheads of 6 indiv. 1 3 5 4 6 11 21 30 29 1 



All in all there were 111 inflorescences.-^ If the 

 terminal inflorescences of these plants (5 with 21, 1 with 

 26 rays ) had been included, the maximum would have 

 been exactly at 21. Seed was saved only from these 

 six plants for the 189.6 crop. It was harvested separately 

 from each parent. 



The fertilization of these plants had not been wholly 

 pure, because the rejected plants referred to above could 

 not be recognized nor removed before the latter part 

 of August, and because flowers which bloom in Sep- 

 tember set hardly any seed with us. Kacli of the six 

 crops actually gave a curve which liad a distinct maxi- 

 mum at 21, but only one of them (No. 1) wholly lacked 



' The curve is figured in Bcv. d. d. hot. Gcs.,Yo\. XVIL Plate \\l. 

 Fig. 2^. 



