314 Nutrition and Selection, 



whorls. This character of the stem, which must be defi- 

 nitely determined by microscopical examination, can, how- 

 ever, be predicted with sufficient accuracy by the thick- 

 ness of the shoot. Of course, the same relation obtains 

 in the summer when the contrast strikes the eye far more 

 forcibly, and then we see that the lateral branches of 

 ternary and quaternary stems have almost without ex- 

 ception decussate leaves, obviously because they are the 

 weakest of all. 



If the plant is taken into cultivation it is easy to mod- 

 ify the proportions of the various sorts of stems by suit- 

 able treatment, although vegetative methods of propaga- 

 tion alone are employed ; but it must be remembered that 

 the buds for the following year are already completely 

 laid down in October, and that the disposition of their 

 leaves is therefore finally decided at that time. The ac- 

 tual disposition in any given year is therefore deter- 

 mined by external conditions which prevailed in the pre- 

 vious year. 



The curves in Fig. 58 are plotted from observations 

 which gave the following results. In March 1890, 1 planted 

 out a group of rhizomes in a favorable position in my 

 garden, and in June I noted the two-, three- and four- 

 whorled stems, which therefore had been laid down under 

 the influence of the unfavorable conditions which ob- 

 tained in 1899. In the following summer I repeated the 

 observations and thus formed an estimate of the efTect 

 of growth under more favorable circumstances (better 

 soil and more light). The results were: 



The apex is seen to be shifted from stems with oppo- 



