394 



THE AMERICAN NATURALIST [Vol. LV 



the curve of means stops ascending and begins to flatten 

 out. This point is marked by a vertical line on each of 

 the curves in Fig. 1 and in the same way in Tables II, III 

 and ly. The coefficient of correlation was now deter- 

 mined both for the group below this point and for the 

 group above it — for the small plants and for the large 

 ones. The constants thus derived are shown in Table V. 

 A study of these figures makes it clear that in the smaller 

 plants there is a decided correlation between body size 

 and organ size (always exceeding ten times its probable 

 error), but in the larger plants practically none at all.^ 

 This emphasizes the conclusion drawn from the character 

 of the curve of means, namely, that np to a certain point 

 increased body size is followed by increased size of organs 

 produced, but that beyond this point there is no relation 

 between the two. 



I I I I I I I I I I I I I I I I -I 

 Shoot 5,ze 



It is of interest to note that in the case of the immature 

 plants the curve of means for leaf size rises steadily as 



sit will be noted that very similar results were obtained by Shaw (17) 

 in correlating seed size with vine length in peas. 



