BASIDIOMYCETES 423 



consider, as Brefeld originally did, that this stabilization occurred first 

 in the "protobasidium," hence its name, and that the holobasidium had 

 arisen from it by suppression of the septa. As evidence for the deriva- 

 tion of basidia from conidiophores, one may compare the Auricularia 

 basidium with many earlier described conidiophores of the Ascomycetes. 

 Furthermore, one may point to the conidiophore of several Polyporales 

 (Hirsutella and Fomes) which apparently resemble basidia. Unfortu- 

 nately these conidiophores have not been cytologically investigated, so 

 that it is unknown whether in the conditions of the nucleus, such transi- 

 tions appear and whether the conidiophores in question actually belong 

 to the diploid and not rather to the haploid phase, in which case the 

 whole argument fails. 



In any case, it is difficult to understand (and in this connection 

 one has no indication of the essential facts) why the asci should be lost 

 without a trace. Either caryogamy and meiosis have been retained 

 as in the higher Ascomycetes and then the reason for the disappearance 

 cannot be found, or they are actually parallel (with the sexuality sup- 

 pressed) and then it is not clear why, after transfer into conidiophores, 

 they can transform the latter to basidiospores while they retain such 

 a fixed form throughout the Basidiomycetes. Consequently the deriva- 

 tion of basidia from conidiophores meets with great difficulties. 



With the derivation of a basidium from an ascus, one may best start 

 with a stichobasidium. Its young stages and its nuclear divisions agree 

 closely with a young ascus and in the octonucleate stage it may 

 not be distinguished from an octonucleate ascus. This correspond- 

 ence becomes even more striking if one considers forms in which 

 the clamp formation extends into the hymenium and in which also 

 the last hyphal septum is provided with a clamp. The first difference 

 appears in spore formation; the spores are not individualized within 

 the mother cell but formed exogenously on sterigmata. We have a 

 parallel to this transference of spore formation from the interior of a 

 sporangium to the surface, in the Choanephora-Piptocephalis series 

 of the Mucoraceae where we may follow step by step a transition, 

 in part limited by nourishment, from endogenous to exogenous spore 

 formation. In both cases it is a question of analogy, deferring of spore 

 formation. 



The transference of spore formation to the exterior offers no difficul- 

 ties. It is the retardation of a process which we often meet in fungi. 

 According to this, a basidium would be an ascus with exogenous spore 

 formation; or, as Vuillemin wrote more than thirty years ago (1893), 

 in a much overlooked work: "La baside est an asque dont chaque cellule- 

 fille, avant de passer a Vetat de spore, fait saillie au transport par le vent." 

 From this primitive eight-spored basidium the four-spored and two- 

 spored basidia would be produced by reduction of the spore number. 



