THE AMERICAN NATURALIST [Vol. LV 



question is also of considerable theoretical significance in 

 that it bears directly on the perplexing problems of in- 

 dividuality and organization. The aim of the present 

 paper is to contribute to the solution of this problem by 

 undertaking a careful analysis and interpretation of size 

 relations in a series of bean plants. 



HiSTOEICAL 



The problem, at least in certain of its aspects, has re- 

 ceived attention at the hands of workers in several fields. 

 Students of the cereal grains, in particular, have been 

 interested in determining whether those plants which are 

 large in the sense of having tall or many stems or a high 

 yield are plants which produce large heads and seeds. 

 This question is of importance in seed selection, since 

 if high yield and large seed size are correlated, it will be 

 comparatively easy to pick out from a mixture those 

 seeds which have been produced by high-yielding plants. 

 Scattered papers on other crops than the cereals also 

 provide facts of interest. 



Most work has been done with the small grains, par- 

 ticularly wheat and oats. Lyon (11) in 1906, although 

 not using biometrical methods, observed that in wheat 

 the weight of the average kernel is not correlated with 

 the number of kernels per head or with the number of 

 kernels per plant. He states that the highest yielding 

 plants have medium-sized spikes and medium-sized 

 kernels. 



Waldron (20) in 1910, working with oats, reported 

 substantial negative correlations (—.4 to —.6 approxi- 

 mately) between average weight of seed per plant and 

 (1) number of seeds, (2) length of head, and (3) length 

 of culm, thus indicating that the larger the plant, the 

 smaller wore its seeds. 



Results at variance with those of Waldron were 

 recorded in 1911 l)y Love (9), Roberts (14) and Myers 

 (12), working with wheat, and by Leighty (7), with oats. 



