DACRYOMYCETALES 539 



pression of the septa and subsequent terminal insertion of the sterigmata, 

 from the Auriculariaceae with which they show a remarkable correspond- 

 ence in gel secretion, in method of basidiospore germination, in longi- 

 tudinal orientation of the nuclear spindles and in the four basidial nuclei. 

 Consequently, they would be the most primitive stichobasidial Auto- 

 basidiomycetes and a link between the Auriculariales and the Cantharell- 

 ales. The structure and relationships of the basidia present the only 

 objection to this concept. It is difficult to explain why such a rearrange- 

 ment of the phragmobasidium should have occurred, as in the Auricular- 

 iaceae with compact gelatinous hymenia the problem of spore discharge is 

 satisfactorily solved by the elongation of the sterigmata. Also in the 

 Tulostomataceae, the only family in which the basidium could have 

 arisen directly from the phragmobasidium by the suppression of a septum, 

 the lateral insertion of the spores is retained in spite of this suppression. 

 It is impossible to say how this rearrangement could have taken place, 

 for no transitional forms between the phragobasidium and the Dacryo- 

 myces basidium are known. Similarly, there is difficulty in the extension 

 of this assumed developmental line to the eight-spored stichobasidium 

 of the Cantharellales, which one would have to consider as formed de 

 novo from the four-spored phragmobasidium and subsequently two- 

 spored basidium of the Dacryomycetales. 



The latter difficulty was raised by the attempt on page 423 to derive 

 the stichobasidial Holobasidiomycetes directly from the Ascomycetes, 

 not from the sticho-Phragmobasidiomycetes. In this case, the Dacryo- 

 mycetales would be primitive only in the structure of their fructifications, 

 while their basidia would indicate an end stage like the two-spored Canth- 

 rellales. Their long sterigmata might then be considered an adaptation 

 for penetrating the gel and, in this sense, a convergence to the sterigmata 

 of the Tremellales, which also are broad at the base and taper upwards. 

 A certain relationship with the Auriculariales need not be rejected; thus 

 one may consider that both these stichobasidial orders (as indicated on 

 p. 530) have their roots in the same Ascomycetous line and from it have 

 retained their common primitive characteristics. 



