12 COTTON SELECTION ON THE FAKM. 



viduals. The work of selection is largely a matter of comparing a 

 plant with its neighbors in the same field instead of with any arbi- 

 trary standard brought in from the outside. 



With a plant as variable as the cotton and grown under such a 

 vast range of differing conditions of climate and soil, it is worse than 

 useless to rely upon any particular feature as an absolutely fixed re- 

 quirement in selection. The Triumph cotton, for example, often 

 has a characteristic shortening or semicluster condition of a few of the 

 fruiting branches at the base of the plant. As the variety is bred by 

 Mr. Alexander Mebane, at Lockhart, Tex., shortened lower branches 

 are shown with remarkable uniformity over many acres. In other 

 locahties, and especially under conditions of more luxuriant growth, 

 the pecuhar habit of growth of the Triumph cotton may disappear 

 more or less completely, so that whole fields may be found with no 

 plants that would correspond to a standard carried over from Lock- 

 hart. 



Any peculiar character, like that of the method of brandling of the 

 Triumph cotton, though it may have no direct economic importance 

 in itself, may be very useful as a mark for distinguishing genuine 

 examples of a variety. Yet it is doubtful whether any such character 

 can always be trusted to appear under all conditions. Even a com- 

 pletely clustered variety Hke the McCall cotton produced no cluster 

 plants in an experiment made in Guatemala a few years ago. 



The need of selection does not appear to be the same every year, 

 for the number of degenerate variations differs in different seasons as 

 well as in different localities. Fields of a carefully selected variety 

 grown under continuously favorable conditions that enable the plants 

 to develop without being checked are sometimes extremely uniform. 

 Each plant appears to be an exact counterpart of its neighbor. Yet 

 in the most uniform fields an occasional individual may show pro- 

 nounced differences. It is worth while, therefore, to go over the 

 whole seed plat every year, even though a general inspection may 

 give an idea that the plants are uniform. In another place or in a 

 less favorable season the same stock may show many pronounced 

 variations and may then degenerate very rapidly unless the diverse 

 forms are removed. 



SELECTION BY CHARACTERS OF STALKS AND LEAVES. 



Young plants of a well-selected, unifotm variety growing under 

 the same conditions follow very closely the same course of develop- 

 ment. They have the same kind of leaves; the joints of the stalks 

 are of the same length and the branches develop at about the same 

 rate and at the same height above the ground. A definite difference 

 in any of these features is warrant for suspecting a plant and giving 

 it closer examination. Another useful mark of distinction may be 



[Cir. 66] 



