40 



Williams, of Monetta, S. C, without finding any cases of the cotton 

 wilt, and Mr. Williams stated that there had never been any in it. This 

 cotton was planted on land where watermelons had wilted badly in 

 1893. It was therefore on a large scale an experiment similar to those 

 described above. In 1895 the field was again planted to cotton, alter- 

 nating with cowpeas — i. e., 3 rows of cotton and 3 rows of cowpeas. 

 In September of that year the writer again examined the field quite 

 carefully, and with t*he same negative result. No cases of cotton wilt 

 were to be found, but there were hundreds of wilting and dead cow- 

 pea vines, and the latter bore the external conidia beds of the luuulate- 

 spored Fnsarium quite regularly, and these were arranged in the 

 manner shown in PI. I, 11, 12. The melon fungus was undoubtedly 

 present im the soil, but there were no melon vines on the field by means 

 of which to establish this fact conclusively. At my request Mr. Orton 

 reexamined this district in the summer of 1899. He found an abun- 

 dance of melon and cowpea wilt, but none of the cotton disease. 



Morphologically, on the contrary, so far as I have' been able to 

 determine from the careful examination of a great many specimens, 

 this fungus is specifically identical on all of these plants, the only 

 doubt being whether it may not have varied enough physiologically 

 not to be transmissible from one host to the other. There are slight 

 variations in the length of the beak and in the size and color of the 

 perithecia on the different host plants, but these are not constant. The 

 ascospore variations are also not constant. As already stated, the 

 ascospores of the form on watermelon are usually smaller and more 

 decidedly elliptical, and a larger proportion are smooth (perhaps 

 because small). When this was discovered an earnest attempt was 

 made to distinguish two or more species, but further studies developed 

 the existence of all sorts of intermediate forms and sizes and left no 

 morphological standing ground for any such conception. For example, 

 perithecia were found on the watermelon at Monetta, S. C, in Septem- 

 ber, 1895, which contained globose wrinkled ascospores 12// in diam- 

 eter; ascospores 12 by 13// and 12 by 14 /< were also observed. Other 

 perithecia on the same root bore elliptical smooth spores. Indeed, 

 ascospores from the same perithecium may be globose or elliptical, 

 wrinkled or smooth. Ked perithecia taken from the roots of wilted, 

 dead cowpeas, at Monetta, September 10, 1895, were indistinguishable 

 in color, size, and shape from those found on watermelon roots, and 

 bore ascospores of the following sizes: 9 by 10, 9 by 14, 10 by 10, 10 

 by 12 (wrinkled), ]0 by 13, 10 by 14 (smooth), 12 by 12 (wrinkled). 

 Ascospores from perithecia which grew on the roots of sea-island cot- 

 ton on James Island were 10 by 10 (smooth and not many so small), 

 10 by 12 (smooth and wrinkled), 12 by 12 (wrinkled), 13 by 14 (wrinkled). 

 Owing to variability, it is likewise impossible to distiugnish the fun- 

 gus from the various hosts either by means of the mycelium or by the 

 external or internal conidia. At most the differences can be scarcely 



