The Origin of Chrysanthcinum Scgctum Plcninn. 193 



This was apparently no longer the case in the follow- 

 ing summer. The following are the records of five 

 plants taken in late summer : 



D' and E' were counted 6 weeks later on the same 

 plants as D and E. The plant E is the seed-parent in 

 Fig. ?)2, p. 181, under 1899, marked with a X- The 

 figures may be regarded as the expression of a tendency 

 to fall back on the secondary maximum at 47 (= 13 + 

 34), and the same result was reached by the other count- 

 ings, which it is not worth while to reproduce here. 



In the following year (1900) the maximum of the 

 lateral inflorescences was still higher. I give the data 

 derived from three plants which were "double" and con- 

 sequently sterile, and of the four next best which were 

 chosen as seedparents. 



The curve of the ''double" race thus seems to have 

 its maximum at about 55 (=21 +34). The possi- 

 bility of attaining higher mean numbers seems to be 



