342 The American Naturalist. [April, 



Contrary to my surmise from the published d< • 

 is not invested with a thin frail cuticle emitting from all points of the 

 surface slender filaments which bind it to the soil, as in Geaster lim- 

 batus, but a fibrous rooting mycelium proceeds from the lowest central 

 point of the base, as in Geaster saccatus; hence the cuticle is persistent, 

 A stout fibrillose layer succeeds forming the main structure of the 

 plant and connecting with the inner peridium at the base by the num- 

 erous pedicels. These are angular or prismatic in shape, sometimes 

 rather flattened and just about 2 mm. in length. A thin fleshy 

 layer lines the inner surface, which is not at all hygrometric as in 

 Astrazus. The inner peridium is depressed-globose and has the silver- 

 gray lustre mentioned by the old botanists ; the surface also is 

 roughened with minute pointed warts, as noticed by Plowright, and 

 this fact leads me to infer that the inner and outer peridium are at 

 first united by what De Bary terms a split-layer, as is the case in 

 rometricus. 



Now as to the internal structure. There is no columella such as we 

 find in all true Geasters. The inner peridium has a soft fleecy lining 

 of fine slender threads, in the dry state much curled and entangled. 

 By careful manipulation, I found each thread to be a long simple 

 structure, brownish in color by transmitted light, 3-4 n in thickness 

 in the middle and tapering to a fine point at each extremity ; it is 

 attached to the membranous wall by one end and of course is free at 

 the other ; so far the agreement of the threads is with those of Geaster, 

 but a peculiarity now occurs. 



The pedicels fuse at once into the wall of the peridium and are not 

 erected in any way into one or more columellas. Instead the threads 

 of the fleecy lining concentrate at numerous points upon the base, 

 elongate and becomes compacted together forming several irregu- 

 lar branched processes, which attain half the height of the perid- 

 ium. These processes remind me forcibly of the numerous dendroid 

 columellas which are erected from the base in the Myxomycete, Rett- 

 cularia lycoperdon ; indeed the silvery surface of the peridium in the 

 two is altogether similar. 



Mingled with the spores, I find numerous free threads similar to 

 those which grow upon the inner face of the wall but much shorter; 

 I suspect that these grow upon the traraa along with the spores and are 

 left free by deliquescence as in Bovista. If this is true it is a unique 

 feature. But I am not able to assert this positively from the mature 



