COTTON SELECTION ON THE FARM. 19 



more effective because all the individuals of a bad progeny can be 

 rejected, whereas some of them that happened to be favorably sit- 

 uated might be saved iu a mixed planting. 



The use of the progeny rows enables an additional precaution to 

 be taken to guard the ])urity of a good stock by holding over a part 

 of the seed from which each of the progeny rows has been planted. 

 If any of the rows should prove to be of exceptional merit it is possi- 

 ble to go back to the reserved seed of the parent of the best row and 

 sow it in a separate isolated plat in the next season as the founda- 

 tion of a special strain descended from a single superior plant. 



Planters of Sea Island cotton are accustomed to the plan of nar- 

 rowing their selection down to a single superior ])lant. They multi- 

 ply the seed from this plant for two or three years in separate seed 

 ])lats to secure enough for field planting. The very high quality and 

 unusual uniformity of the Sea Island cotton are to be ascribed 

 largely to the method of selection that has been followed. 



The location of plantations on islands where no Upland varieties 

 are grown and where consequently the bees could not cross-fertilize 

 the Sea Island cotton with Upland pollen has undoubtedly assisted 

 in maintaining uniformity. Planters of Upland cotton might secure 

 the same advantage if whole communities would organize and agree 

 to plant onl}' one kind of cotton. This is especially necessary in 

 districts where long-staple varieties are grown and the uniformity of 

 the lint directly determines its market value. 



The use of progeny rows in no wa}^ avoids the necessity of thorough 

 familiarity with the variety that is being selected. Unless the 

 breeder is skillful enough to limit his selection to one type of plant, 

 mixture can occur in progen}^ rows as well as in general plantings. 

 Only one type of plant should be grown in the same seed plat, whether 

 planted in progeny rows or not. It is just as important to rogue 

 out degenerate plants from progeny rows as from any other plats 

 or fields where selection is being maintained. However skillful the 

 selection that may be given at the end of the season, its value is 

 alreadv diminished if the select plants have been cross-fertilized 

 by the pollen of degenerate neighbors. It is plain from the results 

 that have been obtained that a farmer who knows his t3'pe well 

 enough can maintain a more uniform stock by general field selection 

 without progeny rows than a careless breeder with progeny rows. 



Planting difi'erent varieties side by side is a good method of learning 

 their differences or testing their yields, but seed shoukl not be saved 

 from such mixed plantings with any idea of further breeding unless 

 the object is to experiment with hybrids. The danger of cross- 

 fertilization forbids anv close combination of the work of breeding 

 with that of testing varieties. A seed plat consisting of progeny 



[Cir. 66] 



