I 62 



THE COTTON PLANT IN EGYPT chap. 



some approaching the long narrow boll of the Sea Island 



type. The form is expressed as t. A cross of 



Upland with a mean form of 075 and a width of 31 mm., 

 upon an Egyptian with a form of 0"58 and a width of 27 

 mm., gives an Fi with a form of 0"60 and a width of 32 

 mm. In other words, the long narrow form is dominant 



though very much 

 larger than the 

 long narrow parent. 

 (Figs. 61 and 62). 



The Fa of such a 

 cross ranges up to 

 the Fi width and 

 form, but includes 

 plants the boUs of 

 which may be as 

 narrow as 0'44 and 

 22 mm. So far as 

 mere width is con- 

 cerned, this might 

 well be due to auto- 

 genous fluctuation, 

 but the new form of the extremely narrow boUs cannot be 

 attributed to this cause. As in the case of leaf-form, we 

 are probably dealing with a compound inheritance, and 

 the new forms are the result of recombinations of 

 allelomorphs. 



A side-light on this interpretation is provided by plotting 

 the correlation table of form against width in F^. While 

 the general trend of the diagram indicates positive correla- 

 tion between wider boll and more spherical boll, yet on 

 comparing such a diagram with that from a pure strain, it 

 becomes clear that the Fj family is heterogeneous. 



Again, families have been raised in F4 which showed no 

 more fluctuation in width than the parent strain, while the 



20 25 30 35 mm. 



Fig. 61. — Width of the Boll. 



