648 THE PHYLOGENY OF THE FUNGI 



The fact that the asymmetrical manner of attachment of the basidio- 

 spore to the sterigma is found throughout the Heterobasidiae (except for 

 a few "angiocarpous" forms) and in the Polyporales and Agaricales 

 probably indicates that they all have a common origin, for it is unlikely 

 that so intricate a mechanism for spore discharge has been polyphyletic 

 in origin. Since the same structure occurs in the Uredinales and in the 

 Tilletiaceae these too must be descended from the same ancestors. There 

 is considerable evidence for the surmise that the Basidiomyceteae with 

 symmetrically attached basidiospores which are not shot off from the 

 tips of the sterigmata are secondary modifications from the asymmet- 

 rical types, associated with the angiocarpous development of the spore 

 fruit (but see below for discussion of the Agaricales-Gasteromyceteae 

 relationships). 



If then the Uredinales (and Ustilaginales) are of common phylogenetic 

 origin with the Heterobasideae and the Eubasidial Hymenomyceteae we 

 must decide whether the heterobasidium or eubasidium is the more primi- 

 tive. Linder's suggestion has been mentioned above that from the Ured- 

 inales arose all other types of Basidiomycetous fungi. Rogers, on the 

 contrary, would derive the Uredinales from Tulasnella, Tremella, and 

 Auriculariales, while deriving the Eubasidiae in another direction from 

 the same primitive genera. If either of these hypotheses is correct then 

 the objects that we call basidiospores (or sporidia) are not homologous 

 with ascospores but are secondary spores produced from them to provide 

 for aerial distribution, the homologues of the ascospores being the indi- 

 vidual cells of the transversely or longitudinally septate basidium. If 

 Rogers is right in postulating the change in position of the four septa in 

 Tulasnella so as to divide the ''probasidium" or "hypobasidium" ver- 

 tically as in Tremella and that then by a further shifting of the septa 

 (intermediate forms seem to occur) they become transverse, as in Auricu- 

 laria, we still must insist upon the homology of the external basidiospores 

 in all these cases. If the Uredinales arose from the Auriculariales by the 

 emphasis on the development of the "probasidium" into a thick-walled 

 resting spore, as occurs in many species of Septobasidium, the sporidia of 

 these too are homologous to the foregoing basidiospores, i.e., are second- 

 ary spores. The close relationship of the Auriculariales, Tremellales, 

 Dacrymycetales, Tulasnellales, Uredinales, and Ustilaginales would seem 

 to be emphasized by the wide-spread occurrence of the production of 

 secondary spores from the basidiospores (sporidia) by budding. The 

 germination of these spores directly by single germ tubes may occur but 

 in water or damp air they may give rise to yeast-like or to allantoid buds, 

 often accompanied by transverse divisions of the basidiospore. Even in 

 the Uredinales (e.g., as observed by the author in Kunkelia nitens) the 



