IX 



HEREDITY 



153 



tliis book we cannot compress a full set of F2 data,, 

 like the set given for fluctuation (Fig. 48), but the three 

 main features are as follows : — 



A. The internode-length of the parents reappears in 

 F2 forming modes in the curve. 



B. The early rate of growth varies in normal plants. 



June 20th. 1910 & 1911 



1 — I — I — I — I — 1 — I — I — I — |— 1 — I ' r 1 — I — I — r 

 20 30 40 50 60 70 80 cm. 



Fig. 57. — Height op Stem in June. 



Individual plants. 



even if we dissect the curve of " growth -rate in June " 

 into any component groups we may choose. Thus, taking 

 only those plants which afterwards showed the habit of 

 continuous growth, and omitting all but the normal 

 seedlings, we find the growth-rate ranging symmetrically in 

 thirty-five plants from 3 '6 to 17"8 mm. per day. More- 

 over, on plotting these growth-rates against the ultimate 

 height of these plants at the beginning of October — which 

 is legitimate, since their growth curves were all " con- 

 tinuous " — we find a close correlation, with a value for " r" 



