20 COTTON SELECTION ON THE FARM. 



rows of the same type of cotton may be looked upon as a test of the 

 vigor and fertility of the different lines of descent inside the variety, 

 but this is not the same as testing different varieties. And even in 

 dealing with progeny rows of one t3^pe the best results are to be 

 expected by going back to the seed that remains from the parent 

 of the best row and planting it under conditions of as complete isola- 

 tion as possible. This sj^stem requires an increase plat for multiply- 

 ing the seed of the best selection in addition to the breeding plat 

 where select progenies are grown side by side for comparison. 



The need of reserving a part of the seed of each of the parent plants 

 until a progeny row has been grown is rendered especially necessary 

 if the conditions do not provide complete isolation from other varie- 

 ties. If crosses with other strains or varieties are taking place, 

 much of the work with the progeny rows may be wasted. 



The greater vigor and fertility of plants that result from cross- 

 polHnation may give such plants an apparent superiority and induce 

 the breeder to select them as parents of progenies. The next genera- 

 tion is likely to show diversity and allow the hybrid nature of the 

 parent to be detected, but perhaps not in time to protect the rest of 

 the breeding plat from crossing with the pollen of the hybrids. The 

 value of seed raised in progeny rows depends largely upon the degree 

 of isolation that can be maintained, but some of the danger can 

 undoubtedly be avoided if a part of the seed of the parent plants is 

 held in reserve until the progeny rows have been compared. 



The chief object to be gained by selection, whether applied to 

 progeny rows or to individual plants, is to maintain the uniformity 

 and productiveness of the variety. It is a serious mistake from the 

 agricultural standpoint to suppose that selection is of no value 

 because the variety does not contmue to improve and become more 

 and more excellent as a result of persistent selection. The old idea 

 of selection as an active agency that could compel plants to change 

 their characters in any desired direction is being given up. It is not 

 to be assumed that varieties selected for many years are necessarily 

 better than those of recent origin. New variations are often as uniform 

 as the parent stock from the time of their first appearance. The 

 present tendency is to believe that the value of selection as an agency 

 of improvement lies chiefly, if not wholly, in the fact that it separates 

 and preserves the good strains or lines of descent and throws out 

 lines that tend to produce inferior individuals. 



This kind of improvement may be considered as finished when a 

 variety has become pure and uniform, so that it contains only one 

 type of plant. Selection might stop if the purity and uniformity 

 would continue without it, but the testing of different lines of descent, 

 even in the best and purest varieties, alwaj^s shows that there are 



[Cii-. 00] 



