COTTON SELECTION ON THE FARM. 21 



(liflereiices among them. Even though they do not show any defi- 

 nite departure from the characters of the variety they are found to 

 di(fer in vigor and fertihty, so that the need of selection remains. 



The tendency of recent investigations of such questions is. to show 

 more clearly that the chief effect of selection is to maintain varieties 

 rather than to change them. Some investigators hold that no further 

 imj)rovement is to be expected after a pure, uniform type has been 

 obtained, but it is no less important to preserve such a type than to 

 select it in the first place. 



BREEDING NEW VARIETIES OF COTTON. 



The preceding discussion of selection relates to the work of main- 

 taining the uniformity of desirable varieties rather than to the 

 origination of new varieties. It is true, of course, that every special 

 selection of cotton may be considered as somewhat different from 

 every other selection, and seed dealers have a custom of giving many 

 new names to slightly difTerent strains of old varieties or even to 

 strains that do not show any appreciable differences. 



Any farmer who learns how to select the variety of cotton that he 

 is already growing will find it extremely easy to breed new varieties 

 if he wishes to take the trouble. Any plant that varies distinctly 

 from the characters of its parent stock is likely to give rise to a new 

 variety. Most of the degenerate plants that have to be pulled out 

 and thrown away might be made into a new variety. The difficulty 

 is not to get new varieties ; the trouble is that most of the variations 

 are degenerate, in the sense that they are inferior to the parent 

 stock for agricvdtural purposes. 



For those who can learn to distinguish the types and carry on the 

 selection, the chief practical difficulty in the work of breeding new 

 varieties is that of isolation; that is, of keeping the difTerent varieties 

 from crossing with each other and with the parent stock. The 

 danger is that the farmer who desires to originate new varieties will 

 turn his seed plat or his progeny rows into a collection of different 

 sorts instead of maintaining a pure culture of one kind of cotton. 

 The mixing of different types of cotton in progeny rows may be 

 expected to have the same undesirable effects as mixing in the field. 



If the farmer undertakes the breeding of new varieties there is the 

 same reason for keeping them away from the regular seed plat as 

 for keeping the seed plat away from other fields of cotton. Only 

 one type or variety of cotton ought to be planted in any plat or field 

 that is being grown for seed or for breeding purposes. The selection 

 of peculiar plants for the purpose of obtaining new varieties is an 

 entirely different line of work from the selection that has to be carried 



[Cir. 66] 



