Correlation 107 



the region just back of this and the primordia of flowers and leaves that 

 originate there are considerably larger in the large-fruited type. 



In this general category of correlations are many of those described 

 by J. Arthur Harris. In Nothoscordum and Allium (1909), for example, 

 he found that flower clusters with relatively large numbers of flowers are 

 borne on relatively long peduncles. Size changes are not always propor- 

 tional, however, for in Ficaria (1918) he found that flowers with a large 

 number of sporophylls have relatively more pistils than stamens. 



The relation between cell size and body size belongs to the part-to- 

 whole category. This has been discussed in a previous chapter (p. 32) 

 and is the basis of a very considerable literature. In a single organ 

 there are often marked differences among the various tissues in the 

 strength of the correlation between cell size and tissue size (Sinnott, 

 1930). In general, it is clear that body size usually does not depend on 

 cell size but on cell number. In many cases, however, it has been shown 

 that in organs of limited growth, such as fruits, large cell size is associated 

 with large organ size, though the range of the former is much less than 

 that of the latter. 



Correlations between Different Parts. There are many genetic correla- 

 tions which do not involve part-to-whole relationship but one between 

 different parts and are thus less obvious as to origin. Sometimes these 

 parts grow at the same rate but more frequently they do not. 



One of the most conspicuous of such growth relationships in higher 

 plants is that between the two main organ systems of the body, the root 

 and the shoot. Its nutritional aspect has already been discussed. The rela- 

 tion is often so precise, however, as to suggest that it has a basis in the 

 genetic constitution of the plant. Its value differs in different plants, at 

 different stages of development, and under different environments. In 

 most cases the root is relatively large in the seedling but grows less 

 rapidly than the shoot. One increases at a rate which maintains a con- 

 stant proportion to that of the other. Pearsall (1927) plotted the dry 

 weight of the root against that of the shoot, both logarithmically, in a 

 series of growing plants in various species. In most cases the allometric 

 constant k was greater than 1, though its value differed in different 

 species. In other words, the shoot grew more rapidly than the root. In 

 etiolated plants, however, it was much greater than 1, and in those with 

 storage roots, it was much less. 



Tammes ( 1903 ) made a study of the growth relationships between an 

 internode and the leaf above it. She found that removal of a leaf would 

 shorten the length to which the internode below would grow but would 

 not reduce the number of its cells. This relationship does not hold in 

 climbing plants, where internode length is usually attained before the 

 leaf above becomes very large. In Ipomoea the excision of leaves on the 



