May 11, 1906.] 



SCIENCE. 



751 



how a few of them might become lost at any 

 camping place, and how they might germinate 

 there by natural means. 



It is true that all the floral districts of the 

 earth contain plants the migratory introduc- 

 tion of which has been caused by the agency 

 of man. As a rule, those plants are self- 

 perpetuated in their new habitats by their 

 unimpaired function of reproduction; and if 

 they have lost that function from any cause 

 their preservation is due to man's intervention 

 for his own benefit. In the case of the papaw 

 here mentioned its vegetative growth has ex- 

 tended far beyond its fruiting limits, a con- 

 dition which savage man could have no in- 

 terest in preserving. His gambling habit 

 seems, therefore, to have been the accidental 

 cause of that part of the dispersion of the 

 papaw which it would not have attained by 

 merely natural causes. 



Charles A. White. 



Smithsonian Institution, 

 April 21, 1906. 



the parasitism of neocosmospora. 



A wilt disease was discovered by the writer 

 in the ginseng gardens of Missouri in the 

 summer of 1904, which proved to be caused 

 by the fungus Neocosmospora vasinfecta var. 

 nivea Sm. The fungus has been studied and 

 described in a bulletin soon to be issued by 

 the Missouri Agricultural Experiment Sta- 

 tion. In the course of the investigation sev- 

 eral new facts were ascertained concerning the 

 parasitism of the fungus, which may properly 

 be mentioned in this place. 



The characters of this fungus have been 

 studied principally by Atkinson and Smith. 

 The former^ first described the fungus in 1892, 

 believing it to be a species of Fusarium. He 

 stated the belief that it was a weak parasite, 

 since it usually infected only such plants as 

 had been previously attacked by another dis- 

 ease. Smith," in 1899, published a detailed 

 account of the entire life history of the fun- 

 gus, giving it the name Neocosmospora vas- 

 infecta. He found that there were three dis- 

 tinct physiological varieties which attacked 



>Bul. 41, Alabama Agr. Exp. Sta., 1892. 



= Bul. 17, Div. Veg. Phys. and Pathol., U. S. 

 Dept. Agr., 1899. 



cotton, watermelons and cow-peas, respective- 

 ly, and that cross-inoculations always failed. 

 In contrast to Atkinson's assumption. Smith 

 stated that all three varieties were parasitic, 

 and especially the variety nivea. 



Although I failed to obtain the perithecia, 

 yet in all other respects the ginseng fungus 

 agreed with the variety nivea (the watermelon 

 fungus of Smith). The results related in 

 the bulletin soon to be issued go far toward 

 establishing Atkinson's theory that Neocos- 

 mospora is a weak parasite and only attacks 

 plants which are first weakened by the pres- 

 ence of another fungus. 



The facts upon which this conclusion is 

 based are as follows: (1) In the field the wilt 

 disease never appeared except where the gin- 

 seng plants had been previously attacked by 

 an anthracnose. (2) Plants which were 

 sprayed with the Bordeaux spraying mixture 

 (and consequently free from anthracnose) 

 were not attacked by Neocosmospora. (3) 

 Watermelon seeds were planted in crocks of 

 rich garden earth which certainly contained 

 microorganisms, but had never been infected 

 with Neocosmospora. Each crock received a 

 test-tube culture of Neocosmospora at the 

 time of planting seeds and three weeks later 

 the melon seedlings were attacked by the wilt 

 fungus. Microscopical examination of the 

 wilted seedlings showed the pink mycelium 

 and spores of Neocosmospora in the fibrovas- 

 cular bundles of the hypocotyls. Other crocks 

 filled with the same kind of soil were sterilized 

 by steam in an autoclave. When cool they 

 each received a tube culture of Neocosmospora 

 and were planted with watermelon seeds. The 

 wilt fungus grew abundantly in the sterilized 

 soil, but at the expiration of twelve weeks 

 none of the watermelon plants showed the 

 slightest indication of the wilt disease. 



These facts are intei-preted to mean that 

 Neocosmospora itself is a weak parasite, but 

 when associated (as it usually is) with other 

 fungi, e. g., Rhizoctonia,, Pythium, etc., it 

 gains entrance into the watermelon plant. In 

 the case of ginseng, its entrance seems to 

 depend upon an anthracnose caused by Yer- 

 micularia Dematium. 



