OKDER HYMENOGASTRALES 539 



the hymenial layer begins to appear it forms the roof of a circular cavity 

 instead of a depressed spherical one. From all sides of this cavity, except 

 the columella, centripetally directed folds begin to form and become 

 covered by the mature hymenium. They may stop short of the columella 

 leaving the circular cavity not completely divided but with numerous 

 radial lobes, or some or all of these folds may grow to the columella in 

 which case separate radial cavities are produced. Eventually the peridial 

 tissues in contact with the basal portion of the columella dissolve, thus 

 forming a basal circumscissile opening and then the remainder of the 

 peridium and finally of the tramal tissues dissolve so that a single cavity 

 filled with spores is left. Its outer wall then is the remains of the sub- 

 hymenial layer. The spores are like those of Gasterella and are produced by 

 twos or fours on the basidia. No cystidia were observed. Routien sug- 

 gested that this represents a further step in complication from Gasterella, 

 and perhaps should be placed in the same family with it, since it starts as 

 a unilocular fungus and often remains so. On the other hand it shows 

 great similarities, although it is much simpler in structure, to the Se- 

 cotiaceae. Rhizopogon is a genus of thirty or more species with subter- 

 ranean basidiocarps whose surface is covered with numerous loose or 

 adherent branching fibrils which lead into rhizomorphs. Its spores are 

 more or less ellipsoidal and smooth. The young spore fruit has a central 

 portion of loosely branching coralloid structure with the interconnecting 

 open spaces Hned with hymenium. The basidia are two- to eight-spored. 

 If a much enlarged Protogaster should develop invaginating and branched 

 ridges and lobes it would show many of the characteristics of Rhizopogon. 



Pilat (1934) discussed the genus Gastrosporium and based upon it the 

 family Gastrosporiaceae. In the latter the peridium is double while in the 

 Hymenogastraceae it is, according to him, simple. 



Family Melanogastraceae. This is an assemblage of several more 

 or less related genera that differ from the Hymenogastraceae in having 

 their lacunar hymenial cavities more or less filled or obliterated by a 

 gelatinous mass which in Leucogasier appears, according to Zeller and 

 Dodge (1924), to be the product of the gelification of conidia or chlamydo- 

 spores which were produced prior to basidial development. Into these jelly 

 filled cavities long, slender basidia push their way, partially filling them 

 with a crisscross tangle. The basidiospores are often coated with a gelati- 

 nous layer and arise two to eight perbasidium. They are almost colorless in 

 Leucogasier and dark brown in Melanogaster. The spore fruits are sub- 

 terranean or partially emerging at maturity, and without a stalk. The 

 gleba is traversed by veins or sheets of tramal tissue that divide it into 

 polyhedric or rounded units each of which is a "basidial nest" as some 

 authors call it. Fischer (1933) places Alpova tentatively in this family. It 

 has been reported only from the United States so far. It is partially sub- 



