33 



always precedes the external one, i. c., the external fungus lias been 

 found only on plants wliicli have beeu killed by the internal one, and 

 it almost always follows the latter, so re<iularly, indeed, as to be at 

 once suspected of forming a part of the life cycle of the parasite. Sev- 

 eral hundred plants each of taelon, cotton, and cowpea in early stages 

 of the disease, in different years and from different places, have beeu 

 examined microscopically, and in every case the fungus was found 

 plugging the vessels in quantity sufficient to account for the disease, 

 and it liad not yet invaded the parenchyma. In cowpeas, prior to 

 their death, it is usual to find the fungus occu])ying tlie vessels of the 

 xyleni and browning their walls the whole length of tlie plant and not 

 simply near the earth. In ohl, wilting melon vines (still alive and not 

 showing any surface fungus), the internal fungus has been found in 

 vessels of the stem 2 to 3 meters from the root, but, in the earliest 

 stages of infection (the earliest yet observed), the fungus is found only 

 in the vessels of the root and extreme base of the stem (hypocotyl). 

 Many cotton, okra, melon, and cowpea plants further advanced in dis- 

 ease have also been examined microscopically. In the interior of these 

 plants, both in the vessels and in parenchyma cells, there was an 

 abundance of mycelium bearing great numbers of the small, colorless 

 (white), elliptical microconidia, and at the same time, on the outside, 

 on the salmon-colored conidia beds, enormous numbers of the big, 

 lunulate, 3-to 5-septate macroconidia, the hyphre of the fungus being 

 readily traceable outward from the plugged or partially occupied ves- 

 sels of the plant to the conidia beds on the surface. In early stages of 

 the disease no external spores or hyph;e of this fungus are ever 

 present, and the internal fungus is restricted to the vessels and con- 

 nective tissue of the bundles, i. e., the non-lignified living i^arenchyma 

 between the bundles is not yet invaded (PI. I, 10 and PI. lY). The 

 arrangement of the external conidia beds in rows is due to the fact that 

 the fungus comes to the surface of the dying plant along the lines of 

 least resistance, that is, between the strands of thick-walled bast fibers. 

 If the external fungus were an independent parasite or saprophyte due 

 to aerial infections, the sporodochia would be scattered irregularly 

 over the surface and not arranged in parallel rows, alternating with 

 the strands of stereorae, and always preceded by the internal fungus. 

 The same conclusion respecting the genetic relationship of the exter- 

 nal and internal conidia has been reached as the result of plant-infection 

 experiments, the watermelon being used for this purpose. More than 

 one hundred and seventy-five typical cases of the melon wilt have been 

 obtained in hothouses in Washington by simply burying a little of the 

 melon fungus in the soil of the pots. On microscopic examination the 

 fruiting fungus was found in the vessels of nearly every one of these 

 plants (all examined) in such quantity as to readily explain the wilt. 

 The soil and water were not sterilized, but under the circumstances 

 4133— No. 17 3 



