LYCOPERDACEAE 115 



Brazil. Rick. (Path, and Myc. Herb., Washington.) Also reported by Sydow (Ann. Myc. 5: 353. 

 1907). 

 Mt. Corcovado. Murrill. coll. (X. Y. B. G. Herb, and U. N. C. Herb.). 

 Africa. Uganda. 4000 ft. elevation. Diimmcr, coll. (Lloyd Herb.). These are buttons which look 



exactly like the American ones. 

 New Zealand. Cunningham, coll. (U. N. C. Herb.). Spores minutely rough, 2.5-3.6^ thick. 



Geaster radicans B. & C. 



Plates 115 and 116 



The co-types of G. radicans (Grevillea 2: 49. 1873) in the Curtis Herbarium 

 (as G. Curlisii) show it to be closely related to G. vclulinus. In so far as the appearance 

 goes, it is only a fornicate form of that species, as Lloyd says (Myc. Notes, p. 155; 

 The Geastrae, p. 31). There are two mounted plants with basal cups gone and one 

 perfect plant in a box, all Ravenel, No. 953. Ravenel's ms. notes accompany the last 

 specimen. The velutinate outer surface and the felted spore sac are exactly as in G. 

 vclulinus. A microscopic examination, however, reveals a difference in the spores, 

 which are larger than in G. vclulinus, 3.5-4m thick, and more distinctly rough. We have 

 examined the plants from Florida (Sarasota, on a cedar stump) in Washington, one of 

 which was photographed by Lloyd (The Geastrae, fig. 57), and find them exactly like 

 the types, the spores 3.4-4/* thick. The basal cup is a strong, pliable membrane, free 

 from earth, and in no w^ay like the delicate, submerged cup of G. jornicatus and G. 

 coronaius. We find at the New York Botanical Garden a specimen of this plant from 

 Bermuda, which extends its known range. The spores of this plant are spherical, 

 distinctly warted, 3.8-4.8/u, most about 4.1m thick. The species is reported as new to 

 Europe by G. Lind in "Danish Fungi as Represented in the Herbarium of E. Rostrup," 

 p. 401, 1913 (no description). Rick figures a plant as G. radicans from Brazil (Broteria 

 5: pi. 5, fig. 9) that is not fornicate and looks exactly like G. triplex with a large collar. 

 He had earlier reported this as G. triplex, and says that Bresadola referred his plant to 

 G. radicans. It may be G. vclulinus, but cannot be accepted as G. radicans. Lloyd 

 thinks that G. Wdvntsckii (from Spain) is the same as the American plant (Myc. Notes, 

 p. 315, pi. 101, fig. 3). 



The Cuban plants reported in the original description as G. radicans are represented 

 in the Curtis Herbarium as Wright, No. 873. These three plants are all buttons and 

 are not G. radicans. They were entirely submerged and are covered with humus held 

 on by a flocculence as in G. rnjescens; a single stout basal strand is attached to buttons 

 in two cases; not pointed. 



The types were sent to Berkeley by Curtis, who received them from Ravenel, who 

 also sent notes and added that if a new species then "it is G. Curlisii" wishing to honor 

 Curtis. Berkeley, however, ignored this wish of Ravenel's and named it G. radicans 

 B. & C. (See Lloyd, Myc. Notes, p. 171.) Ravenel in his notes says "with a sulcate- 

 plicate mouth," but the mouth is not truly so as we are defining that character. The 

 best specimen has the mouth pinched into a few crumples, but the others do not show 

 this. 



As the usual form and spores distinguish G. radicans from G. vehttini(s, we think 

 it well to retain the name, though it would be better to reduce it to a variety. 



