Notes on Marsilca llllosa. 49 



The locality was not revisited until May of 19 17, at which time 

 the water had not entirely dried up. Many fruitinq,- specimens were 

 collected wliere the water liad disappeared, particularly along the 

 edge of a road, where the rhizomes were covered witli a loose vol- 

 canic ash. The sporocarps were covered with rust-colored silky 

 hairs, and in many specimens were hidden in the dense rust-colored 

 wool of the rhizome. It is notahle that practically all fruiting plants 

 were found either at the edge of the water — not in it — or in places 

 that had hecome dry early in the season (see 1'. XI\ ). Plants 

 growing in dee])er water generally witlier u]) and die l)efore pro- 

 ducing sporocarps. The locality was visited several times during the 

 latter part of 1918, and finally July i. 1919. when a portion of the 

 area was found to he full of ruhhisli and dr^•. ()nly a few fruiting 

 plants and no living sterile plants were seen. 



Lender culture in an aquarium, glabrous plants of Marsilca 7'll- 

 losa l)ehave(l exactly as in nature, in respect to the development of 

 pubescence, while the water was evaporating. All specimens experi- 

 mented upon produced a vigorous vegetative growth, but died before 

 producing sporocarps. 



Following IJrackenridge' it is m_\' ()])ini()n that M. villosa Kaulf. 

 and M. 'c'cstita 1 Ik. and Grev. are the same species. Differences in 

 regard to the pubescence, lengtli of petiole, and compactness of root- 

 stalk of diff'erent herbarium s])ecimens of these species are merely 

 the result of environmental conditions at the time the specimens were 

 collected. Differences in size and widtli of the leaves of different 

 specimens are merely individual variations. 



It is probable that .1/. I'ilhsa was brought from the \\'est Coast 

 of North America to Hawaii by natural means. 



' Bracken rid.s^e, Wm. D., op. cit. 



is 



