6 COTTON SELECTION ON THE FARM. 



somewhat increased, if the greater vigor of the first generation of 

 hybrid phmts rendered them more productive. But it often happens 

 that the more vigorous plants are not more productive, because they 

 begin by putting out too many vegetative hmbs instead of fruiting 

 branches. The crop is not set so early and is more likely to be 

 destroyed by the boll weevils. The admixture of the varieties by 

 cross-pollmation might show no bad results until the second j^ear 

 after the seed is mixed and plants of different kinds are cross-fertilized. 



If the effects of the crossing of varieties w^ere limited to the produc- 

 tion of vigorous intermediate hybrids, like those of the first generation, 

 the yield might never be seriously affected. In the second and later 

 generations of hybrids another factor comes into play, the factor of 

 increased diversity that results from the crossing of different types 

 of cotton. Even when the first generation of hybrids is closely similar 

 to one of the parents or merely intermediate between the parents, the 

 second and later generations are usually much more diverse. In addi- 

 tion to many different degrees and combinations between the charac- 

 ters of the original parents that were crossed, many other characters 

 quite outside of those of the selected parent varieties are likely to 

 appear. 



To permit indiscriminate crossing among the second and later gen- 

 erations of hybrid plants is the way to undo the work of selection and 

 return toward the original condition of indiscriminate diversity found 

 in unselected stocks. The result of indiscriminate crossing seems to 

 be that the characters of inferior ancestors are brought back into 

 expression and may completely submerge the characters of the spe- 

 cially selected superior ancestors. In a series of experiments that 

 has been made with hybrids between the Egyptian and Upland 

 types of cotton the first generation hybrids had longer and better lint 

 than either of the parent types. The later generations not only failed 

 to maintain the promise of the first generation, but gradually receded. 

 The fourth and fifth generations failed to show any plants with lint 

 equal to the Egyptian parent, and very few were equal to the Upland. 

 No matter how superior the varieties may be at the beginning, indis- 

 criminate mixing can only be expected to lead back to inferior mon- 

 grels, or "scrubs," the same kinds of inferior plants that can be found 

 in every field of ordinar}^ unselected cotton. 



DIVERSITY IN UNSELECTED FIELDS OF COTTON. 



In a neglected stock of cotton that has not been receiving any selec- 

 tion at all the plants are not all equally inferior, but each individual 

 plant is likely to be difl'erent from any of its neighbors. The differ- 

 ences between the individual plants of an unselected field correspond 

 to the differences between selected varieties. Each plant of an unse- 



[Clr. 6G] 



