38 



they came up badly. Ou March 1, 75 seedlings of the Rattlesnake 

 melon were up and 42 seedlings of the Seminole. On March 5 two 

 plants showed drooping cotyledons, and on making sections of the tap 

 root the melon fungus was found fruiting in the vessels of each. Dur- 

 ing the next two weeks 24 melon plants wilted, and the characteristic 

 fruiting fungus was found plugging the vessels of the tai^root or hypo- 

 cotyl of each one. Subsequently many additional plants were attacked 

 and destroyed, showing clearly that the melon fungus must have been 

 alive in the soil during the whole time that the cotton and cowpeas 

 grew in it unmolested. The presence of the typical conidia-bearing 

 fungus in the vessels of the wilting plants was determined, in nearly 

 all cases, by a microscopic examination, although this became rather 

 monotonous toward the close of the experiment. It was on this bed 

 that the tomatoes grew unmolested from February to midsummer, and 

 this, too, although many of the wilted melon i)lants were buried from 

 time to time close to the roots of these tomatoes. Twenty cowpea 

 plants which came up by accident (from seeds in the buried refuse) also 

 remained healthy until they were removed in July. 



(3) On April 12, 1897, another attempt (the fifth or sixth) was made 

 to infect cotton and cowpeas with the melon fungus. For this purpose 

 m one of the hothouses there was prepared a bed of good well-rotted 

 potting earth. This bed was G feet long, 3 feet wide, and about 8 

 inches deep. As soon as this bed was ready, cultures of the melon fun- 

 gus were buried in it, in rows, at right angles to the long axis of the 

 bed, and 4 inches apart. Twenty-six test-tube cultures of the fungus 

 were used for this purpose. These were vigorous growths on slant 

 agar, potato, etc., which had been started for this purpose April G. 

 Grooves about an inch deep were made in the soil. The fungus was then 

 uuiforndy distributed in these furrows and covered with the loose earth. 



On April 13, 400 seeds of the Georgia Rattlesnake melon were planted 

 in this bed, in 15 rows, alternating with the rows of the buried fungus, 

 and consequently only 2 inches either way from the fungous masses. 

 The germinating capacity of these seeds had already been tested and 

 found to be high, but in this bed only about half of them came up. 



On April 29 about 200 melon plants had come up, and none of them 

 were unhealthy. The first case of melon wilt appeared April 30 and 

 the second May 3. 



On May 3 — i. e., as soon as it had become apparent that the melon 

 fungus was active in the soil — 200 seeds of cowpea and 175 seeds of sea- 

 island cotton were planted between the rows of melons — i. e., on top of 

 the rows of buried fungus. To help on the disease and add interest to 

 the experiment, the spores of TMelavia hasicola were also planted in 

 the bed in one or two places. This fungus makes wounds in cotton 

 and cowpea plants at or beneath the surface of the earth from the sur- 

 face inward, and it was thought that possibly such injuries might favor 

 the entrance of the Fusarium. 



