98 MASSACHUSETTS HORTICULTURAL SOCIETY. 



or mutations which may be isolated and tested, and if found to 

 breed true may be introduced as new varieties or races. 



The Columbia cotton, bred by the writer, illustrates this t\^e of 

 breeding. While the writer was engaged in cotton breeding in- 

 vestigations conducted for the U. S. Department of Agriculture, he 

 cultivated in the season of 1902 a large number of varieties of 

 cotton at Columbia, South Carolina. While walking through this 

 variety patch one day, exhibiting the varieties to some growers, he 

 stopped before a certain plant of Russell Big Boll cotton to call the 

 attention of the party to the characters of this variety. The Russell 

 cotton ordinarily has lint about one inch to one and one-eighth 

 inches in length and has a very characteristic boll, seed, and leaf. 

 On pulling out the lint of this particular plant instead of finding 

 the ordinary lint it was found to have lint about one and three- 

 eighths inches long. In all other regards, the plant was like the 

 typical Russell. In this plat there were about four hundred plants 

 of Russell and these were later all examined to discover, if possible, 

 any tendencies to produce long lint. About a dozen plants were 

 found which had slightly longer lint than ordinarily shown by the 

 variety. These ranged from one and three-sixteenths to one 

 and one-quarter inches in length. The next season, 1903, the 

 seeds from, these dozen selected plants were planted in an isolated 

 plat by themselves, the progeny of each plant being planted 

 separately. A careful examination of the progenies of each plant 

 in the fall showed that only the one plant, that first noticed with lint 

 one and three-eighths inches long, showed any marked tendency 

 to produce a progeny with markedly longer lint than the ordinary 

 variety. The progeny of this one plant was very variable, but a 

 considerable number of the plants had long lint like the original 

 plant. Some of the poor plants in the progeny were doubtless 

 due to crossing with other inferior plants in the preceding season. 

 Again, in this season, 1903, a number of the best plants Avith long lint 

 was selected all of them being taken from the progeny of the one 

 superior plant selected in 1902. They were planted in an isolated 

 patch in 1904, the progenies of different plants being planted 

 separately. 



In 1904, a very large number of the plants had the long lint but 

 there Avere still some short-linted ])lants present, probably due to 



