THE AMERICAN NATURALIST [Vol. LV 



With apples, Shaw (15, 16, 18) and Stewart (19) have 

 pointed out that there is little relation between the yield 

 of fruit on a tree and the average size of the fruit except 

 in very heavy yields, when fruit size decreases. Small, 

 young trees may have slightly larger fruits than large, 

 old trees. Many investigators maintain, however, that 

 under most conditions thinning will increase the size of 

 the fruit and thus imply that there is usually a negative 

 correlation between yield and fruit size. 



In tobacco, Hayes (5) found that there is probably no 

 significant correlation between number of leaves and 

 average leaf area. 



In beans, Harris (3, 4) found a small positive correla- 

 tion between number of pods per plant and (1) number 

 of ovules per pod and (2) average weight of seed per 

 plant, the higher yielding plants thus having somewhat 

 larger pods and somewhat heavier seeds. 



In peas, Shaw (17) found that a positive but small cor- 

 relation exists between length of vine and average weight 

 of seed produced, and that this correlation is much 

 greater in small plants than in large ones. 



The total evidence is therefore conflicting. A majority 

 of the workers report positive correlations between plant 

 size and the size of the various organs produced. These 

 correlations, however, even when significant, are in most 

 cases so small, and there are so many instances where the 

 coefficients are clearly not significant or are even nega- 

 tive, that no general conclusion, supported by the whole 

 body of evidence, can well be drawn. 



Materials and Methods 

 The present paper is the result of a study of a group of 

 562 bean plants grown during the summer of 1918 as a 

 part of a larger investigation. The beans were Red Kid- 

 neys, nnd jilllioimli thoy wcro not Tnoiiibers of a pure line, 

 they \v<'i'(' similar in r.ll cliaractcrs studied as to in- 

 dicate tliat no wide genetic differences existed among 



