Genetic Factors 421 



4. In a considerable series of crosses between races genetically more 

 complex and differing in both the size and the shape of their fruits, a 

 positive correlation was observed in each case between fruit length 

 and fruit width in the parents and the Fj (where presumably all size 

 differences are caused by environmental factors) but a negative one in 

 F 2 where segregation occurs. This again can be explained by assuming 

 that shape is inherited independently of size. A certain amount of ma- 

 terial is genetically available for every fruit, and if its shape genotype 

 tends to produce an elongate one, this will be relatively narrow, and 

 hence the negative correlation. Maximum parental length is never com- 

 bined in F 2 with maximum width, or minimum length with minimum 

 width, as they should be if dimensions are directly determined genetically 

 and recombined independently. 



5. In the F 2 the coefficient of variation for length is twice as large as 

 that for width, which is to be expected in a radially symmetrical organ 

 where shape and size are genetically independent, for if the amount of 

 material for growth is fixed, a unitary change in width (equatorial 

 diameter) should produce a much greater change in length, since volume 

 is essentially width X width X length. 



Further evidence that genes for shape are actually operative is found 

 in the fact that in a number of plants, such as the tomato (Lindstrom, 

 1928; Butler, 1952), genes for fruit shape are linked with others and can 

 be definitely located on chromosome maps (Fig. 19-2). If it were dimen- 

 sions that are directly controlled, presumably their genes would occupy 

 different loci. 



If genes determining shape actually exist, the difficult problem arises as 

 to how they produce their effects. The ultimate mechanisms are by no 

 means clear but the visible steps in the process can be described. In 

 some cases the shape of an organ, such as a fruit, is established when it 

 is a very small primordium. After this, growth rates in the various dimen- 

 sions are equal, and as the structure grows its shape remains constant. The 

 critical step in establishing growth relationships here is taken very early. 

 This is what happens in fruits of pepper, tomato, some races of Cucurbita 

 pepo, and various other plants. It is usual in organs that are nearly iso- 

 diametric at maturity. 



In many fruits and in most leaves, especially where the dimensions are 

 markedly unlike, the primordium is a roundish mass of cells and the shape 

 of the organ is produced by differential growth among the dimensions 

 (Sinnott, 1936b). In various races of the gourd, Lagenaria siceraria, for 

 example, mature fruit shape varies from long and narrow through round 

 to flattish. All are alike in early primordia. In elongate types (like the 

 "Hercules club") length increases faster than width and as the organ 

 grows in size it becomes progressively more elongate. Conversely, the 



