8 



NATURAL INFECTIONS. 



The effect of repeated infections of the small roots of the cotton is 

 very noticeable, especially when the plants are somewhat resistant to 

 the disease. Small tufts of roots ^row from each i^oint of infection, 

 doubtless on account of some stimulus exerted by the fungus. Sev- 

 eral short roots will thus start from a place which would normally 

 have produced one longer branch. (PI. Ill, fig. 1.) 



Many of these little roots are killed by the fungus and others grow 

 in their places, so that the tufted appearance of the rootlets is more 

 pronounced late in the season. The same result has been produced 

 in the laboratorj^ by inoculating seedling cotton plants with x^ure cul- 

 tures of the cotton-wilt fungus. Similar root tufts are found associ- 

 ated with the wilt diseases of okra, cowpea, watermelon, and cabbage, 

 and they are believed to be characteristic of this class of root diseases. 



In the case of cotton their presence on the roots demonstrates the 

 presence of the wilt fungus in the soil, even Avhen the amount is so 

 small that no harm is visible aside from the reduced growth of the 

 plants. (PI. Ill, fig. 2.) This dwarfing of the plants is due to the 

 killing of the small roots and is often visible over a considerable area 

 surrounding a badly infected spot. For this reason the loss in yield 

 on such a field is much greater than would appear simply from a 

 consideration of the badly diseased areas, as the dwarfing due to the 

 injuring of the small roots considerably curtails the yield. 



ARTIFICIAL INFECTIONS. 



Since the publication of Bulletin No. 17, the wilt disease has been 

 I)roduced in healthy cotton plants by inoculating the soil in which 

 the}' grew with pure cultures of conidial stages of Neocosmospora 

 rasinfecta. This removes any doubt as to the causal relation of the 

 fungus to the disease which might arise from the failure of the pre- 

 vious inoculation experiments. The plants were grown for a few 

 weeks in pots, and then a small quantity' of fungus from a pure cul- 

 ture was placed in the bottom of each one. Eight days later thej'^ 

 were transplanted to the open ground. The first case appeared after 

 about 35 days. Fourteen out of 24 plants contracted the disease. 

 The fungus was abundant in the vascular bundles of 7 i^lants and 

 tliej^ showed all the other symptoms of the disease. The other 7 

 infected plants were onl}^ slightly diseased, although the fungus was 

 found in the vessels of the stem. The check plants, 25 in number, 

 all remained healthy. It will be noted that the length of time 

 between the inoculation of the soil and the appearance of the dis- 

 ease in this experiment (35 to 50 daj's) was practically^ the same as 

 elapses in the field between the germination of the seed and the first 

 appearance of the disease. That a larger proj)ortion of the inocula- 

 tions did not succeed is believed to be due to the small amount of 



