108 THE GASTEROMYCETES OF THE UNITED STATES AND CANADA 



clean of trash except at base, mycelium attached only at a central point below; rays 

 reflexed, the long tapering tips usually revolute; outer layer not cracked into strips or 

 flakes, but separating as a rule from the central region and the proximal part of the rays, 

 and remaining convex below while the inner layer arches upward and elevates the spore 

 sac (pseudofornicate). Spore sac subglobose with a more or less obvious apophysis, 

 brown, nearly glabrous and with a short, thick, often flattened stalk; peristome definite, 

 silky, often paler than the sac, broadly conical, the mouth fimbriate. Fleshy layer 

 brown, rather thick, mostly adnate and not cracking. Columella subspherical, about 

 4—5 mm. thick. 



Spores (of plant from Wisconsin) spherical, distinctly warted, 3.7-4.8/*, most 

 about 4fi thick. Capillitium threads straight, not branched, concolorous with the 

 spores, up to ly. thick. 



This plant is evidently nearly related to G. triplex as shown by long-pointed lobes 

 (indicating a pointed button), absence of flocculence on the outer surface, which is clean 

 and glabrous in large part, and in the definite mouth and similar spores. It differs 

 from G. triplex in the splitting away of the outer layer over the central region and the 

 arching up of the middle layer. The stalk is probably no more obvious than it would be 

 in G. triplex if it were to assume such a position. 



There is much confusion among authors and in herbaria as to the identity of G. 

 limbatus. This is the only American plant labelled G. limbatus in American herbaria 

 which cannot be referred to G. triplex or G, rufescens or G. pectinatus. Most of the 

 collections at the New York Botanical Garden labelled G. limbatus are G. rufescens, e.g., 

 from Alabama, Kentucky, Ohio, Missouri, Indiana, Colorado and Canada. All the 

 material in the Path, and Myc. Herb, at Washington as G. limbatus is G. rufescens, as 

 Ellis, N. Amer. Fungi, No. 1309, and others from Ohio and Colorado. Morgan figures a 

 plant as G. limbatus (N. Amer. Geasters, fig. 6) which is probably G. rufescens, as noted 

 by Lloyd and Hollos. A plant in the Curtis Herbarium from Alabama as G. limbatus 

 (Peters, No. 934) is G. rufescens with a short stalk. We are adopting the name of G. 

 limbatus for the present species as it is so determined by Bresadola, as shown by a 

 European specimen from him in the Lloyd Herbarium. This has exactly the habit of 

 the American plant described above, the outer surface clean and separated from the 

 fibrous layer over the center and proximal part of the rays. Another plant from Bresa- 

 dola so labelled has the outer surface earthy and not separated in part. The mouth is 

 distinct in both. Our plants described above are also the G. limbatus of Lloyd, at least 

 in so far as the American specimens are concerned, as shown by a number of plants in 

 his herbarium. Among Lloyd's illustrations of G. limbatus (The Geastrae, figs. 42^16), 

 figs. 45 and 46 represent the species as we have it. His three other figures are probably 

 not the same. Figure 42 of a plant from Rea (England) and fig. 44 of a plant from 

 Hollos (Hungary) apparently represent the dark species noted below and considered 

 G. limbatus by English authors and by Hollos. 



It is certain that the G. limbatus of Hollos and of Massee and Berkeley and to all 

 appearances that of Elias Fries* is not our plant, and we have never seen anything like 

 it in this country. It is a large, usually very dark plant (blackish brown; some forms 

 are paler and glaucous) with a short thick stalk, a large, definite, fimbriate mouth of 



* In the Curtis Herbarium are two plants from E. P. Fries (Upsala) labelled by him G. limbatus, but they are really of 

 pectinatus. Worthington G. Smith gives a figure (Grevillea 2: pi. 17, fig. 2) as G. limbatus which has all the appearance G. 

 G. pectinatus except that grooves are not shown on the peristome. 



