42 HORTICULTURAL SOCIETY OF NEW YORK. 



fertilizers designed to stimulate and strengthen the plant, and with fungicides 

 applied to the soil to kill the fungus, were all without effect in checking the 

 disease. 



Areas thus infected are found scattered through the Southern States, and 

 the control of the disease is an important question, since many thousand 

 acres are affected, and the aggregate loss amounts to hundreds of thousands 

 of dollars annually. 



The possibility of using plant breeding as a means of controlling the 

 disease had been kept in mind from the beginning of the investigations, and 

 after other methods had failed experiments were undertaken, with the view 

 of discovering or originating a variety resistant to the wilt. These have 

 proved very successful in a number of instances, which will be mentioned 

 briefly. 



1. The Rivers Cotton. 



It was early observed that not all plants were equally attacked by the 

 wilt disease. Frequently one of two plants in the same hill died and the 

 other lived, while in a field where nearly everything was killed some few 

 plants would survive and show no trace of disease. The 'first attempt to 

 produce a resistant strain by selection of such plants was made on Sea Island 

 cotton about 1895 by ^Ir. E. L. Rivers, of James Island, S. C. This resulted 

 in a failure, owing to the fact that the single plant selected proved to be a 

 hybrid without desirable commercial qualities. In 1899 Mr. Rivers, who was 

 then co-operating with the Department of Agriculture in its work on the 

 disease, saved seed from another resistant stalk, which he planted in 1900 in 

 a single row through a field of his ordinary cotton. The land was badly in- 

 fected with wilt, and nearly all the cotton in the field died, while not a plant 

 in the select row was killed. This strain was planted on wilt infected land 

 the next year, and preserved its resistance well. This season (1902) fifteen 

 acres were planted on land wliich had formerly been abandoned for cotton 

 because of its infection with the wilt fungus. With the exception of a few 

 scattered plants, this variety resisted the disease completely (PI. II.), 

 while adjoining cotton of another kind, on land not so badly infected, was 

 very much injured by wilt. These three successful tests of this selection 

 indicate that it is as nearly resistant to wilt as any variety can be expected 

 to be, and that the use of this or other resistant varieties will be the solution 

 of the cotton wilt problem in the Sea Islands. 



In its other features, such as length, fineness and uniformity of staple, it 

 is above the average. It yields as much or more per acre than the ordinary 

 non-resistant kinds, thus showing that in securing resistance to disease other 

 desirable qualities have not been sacrificed. Its cultivation will be continued 

 and seed will be distributed by the Department of Agriculture for the relief 

 of the Sea Island cotton planters in Georgia and Florida. 



2. Other Resistant Selections. 



In connection with the efforts of the Department of Agriculture to pro- 

 duce wilt resistant varieties of cotton for distribution to the farmers in the 

 affected districts, a number of other selections have been made, wiiich have 

 proved very resistant to wilt. They have now been grown for two years on 

 wilt infected land, where they have remained healthy, while ordinary cotton 



