656 THE PHYLOGENY OF THE FUNGI 



columella and gill-like outgrowths from the top of the chamber which may 

 divide it into several radial cavities. The spores in all three are ovoid or 

 limoniform, dark-colored and symmetrically perched on the sterigmata. 

 Many of the spores when set free bear at their base a piece of the upper 

 end of the sterigma. In their spore character they are practically identical. 

 The first two could find a place in the same family, Hymenogastraceae, 

 but the third with its percurrent columella and radial hymenial chambers 

 and basally circumscissile dehiscence shows some characters that might 

 be considered as belonging to the Secotiaceae. This is merely an example 

 to show how, possibly, the current classifications are faulty. 



Phylogenetic System of Gasteromyceteae 



In default of further studies on the anatomy and ontogeny of the very 

 young spore fruits and of intensive comparison of spore types throughout 

 the whole group the following may be suggested as a tentative phylo- 

 genetic system of the Gasteromyceteae. The primitive (or the simplest) 

 forms are minute, with one closed hymenial chamber. This might be 

 supposed to have arisen from a small Aleurodiscus-like fungus in which 

 the upward curvature at the edges continued until a closed cavity was 

 produced, lined by the hymenium. As seems to be the case when angio- 

 carpy develops the basidia come to bear sterigmata with symmetrically 

 attached spores. The genus Protogaster represents one of these small, 

 unicameral fungi. Much like this, but with an entirely different type of 

 spores, is Gasterella. In the spore fruit of this a somewhat arched hymenial 

 primordium arises angiocarpously and gradually a basidial layer appears 

 with the basidia directed downwards to line a shallow cavity which soon 

 becomes larger and nearly spherical. In the more vigorous specimens the 

 roof of this cavity may be thrown into folds and convolutions which do 

 not reach to the base so that although the hymenial surface is increased 

 the cavity is not divided. At this stage it resembles closely the early 

 developmental stages of Hymenogaster as described by Rehsteiner (see 

 Chapter 15). Probably the next step beyond the unicameral condition 

 arose by the increase in the folds and convolutions to form many cavities 

 lined by hymenium instead of just the one. In the Ilysterangiaceae the 

 coralloid development is much more marked along with the production of 

 a more or less pronounced columella. The cartilaginous and gelatinous 

 character of the gleba is possibly a modification to permit the dissolution 

 of the spore fruit at maturity. In the Lycoperdaceae the glebal tissues 

 enclosing the numerous hymenial cavities dissolve after the spores have 

 become mature, leaving one large cavity filled with the spores entangled 

 in the filamentous capillitium. By the flaking off of the outer layer of the 

 peridium and the formation of an ostiole in the inner peridium or some 

 other mode of dehiscence the spores are enabled to escape and be scat- 

 tered by air currents. 



