Correlation 



109 



below them but that the degree of this difference depends on the rela- 

 tive size of the two branches and on the angle between them. It is an 

 expression of the polar tendency of the trunk. 



There are also correlations between the parts of an organ. In leaves, 

 for example, although the length of the petiole is much more variable 



2 

 2 



cc 

 tu 



Ui 



2 

 < 



o 



2 

 2 



a. 



ID 



i- 



UJ 



2 

 < 



6 



o 

 o 



FRUIT 



_i i i i_ 



Time in Days 



Fig. 5-4. Diameter of stalk plotted against diameter of fruit in a large-fruited race of 

 Cucurbita (above) and a small one (below). Rate of growth is less for stalk than 

 for fruit but at flowering the logarithmic distance between the two is approximately 

 the same in both races. ( From Sinnott. ) 



than the dimensions of the lamina, there is a relation between them. In 

 Tropaeolum, Pearsall ( 1927 ) found that this was allometric, with lamina 

 width growing faster than petiole length (Fig. 5-3). In Acer the volume 

 of the leaf blade is much more closely correlated with the cross-sectional 

 area of the petiole than with its length. In the runner bean the area of 



