182 Observation of the Origin of Varieties. 



The progress was uniform and regular and the sim- 

 plest index of it is the series of successive seed-parents. 

 The number of ray-florets in the terminal inflorescences 

 of my selected plants in successive years was as shown 

 in the table on page 179. 



The progress can be seen still better from the curves 

 which I have plotted of the terminal inflorescences in the 

 \arious generations. The reader is referred to the table 

 on page 180 and Fig, 32 on page 181. The original ])lant 

 of 1895 referred to as No. 1 arose from a seed which, 

 as already stated, was obtained by exchange from a 

 botanical garden. Indeed this particular lot of seed came 

 from Groningen but produced a mixture the curve of 

 which would obviously not have any special interest. The 

 cultures of the subsequent years were every time the 

 offspring of a single individual whose fertilization by its 

 like had been insured as much as possible. 



The following considerations arise out of an inspec- 

 tion of Fig. 32. 



The curve of 1897 was monomorphic like that of the 

 typical examples of the 21-rayed race (Fig. 31 B, p. 169) ; 

 but it was markedly asymmetrical, a fact which afforded 

 a pretty strong indication that the race could be improved 

 by selection in the plus direction. It confirmed the expec- 

 tation based on the part-curve of the parent of this 

 culture. 



The curve of 1898 relates to the offspring of the 34- 

 rayed plant of 1897. In it new maxima appear. These 

 conform with Ludwig's law, for the}^ lie on the figures 

 of the well-known Braun-Schimper or Fibonacci se- 

 ries. One of them is at 34 (= 13 + 21) which belongs 

 to the primary series: the other is at 26 (= 5 + 8+ 13) 

 which is one of the subsidiary numbers. The maximum 



