I9 2 HEMIBASIDIOMYCETES [CH. 



Basidiospores are budded off in abundance from all three cells, and in turn 

 give rise to conidia. In the meantime the basidium has separated from its 

 parent brand-spore, and the spore, after nuclear division, may produce another 

 basidium, and others in- succession at the same spot, so that free basidia 

 accumulate in the culture. If cultures in nutrient solution are allowed to 

 starve, association now takes place between basidial cells, basidiospores or 

 conidia by means of conjugating tubes (fig. 1 58 ). The paired cells increase 

 markedly in volume, but no interchange of cytoplasm takes place and the 

 nuclei remain in their respective cells without visible change. Harper observed 

 that when fresh beerwort was supplied to his cultures at this stage the produc- 

 tion of conidia began again. They are produced from one or both of the con- 

 jugating cells, but only a single nucleus is concerned in the development of 

 each conidium, the other remains quiescent and the conidia are uninucleate. 

 In a host plant, the name of which is not recorded, Werth and Ludwig 

 failed to find binucleate elements, the youngest cells in which they could iden- 

 tify the nuclei being uninucleate (fig. I59). They infer that in this species 



nuclear association fails to take 



place> and " binucleate sta g e 

 exists. This hypothesis accords 



we ^ w i tn Harper's observations, 

 on the saprophytic phase which 



r 'g- ' 59 Ustilago antherantm P r. ; a. young brand- 

 spores b older brand-spores; c. basidia ; d. basidio- he Studied ill material grown On 

 spores ; after Werth and Ludwig T i IL r\ i LI f. j 



Lychnis alba. On the other hand, 



a binucleate stage was identified by him in the sporogenous cells of U. an- 

 therarum on Saponaria, and by Dangeard on Lychnis dioica. These facts 

 suggest the possibility of two or more varieties of U. antherarum on different 

 hosts and differing in their cytological behaviour; the forms studied by 

 Dangeard and Harper point to a condition comparable to that of U. Maydis, 

 while Werth and Ludwig's observations indicate the possibility of a truly 

 apogamous strain. A complete life-history of this fungus, in material obtained 

 from a single host, would probably prove of interest. 



Another possibly apogamous form is Ustilago levis, a common 'smut 

 on oats. According to Lutman the basidio- 

 spores give rise to considerable numbers of 

 conidia. These are multinucleate if formed in 

 crowded masses, uninucleate when comparatively 

 isolated. Germinated conidia found on the epi- 



dermis of infected seedlings usually contain two 



or three nuclei. The parasitic hyphae are multi- 

 Fig. .160. Ustilago levis (K. and nucleate and their swollen ends, when Spore- 

 Is.) Magn.; mycelium with r , 



mu'tii uc'eate and binucleate lormation is about to take place, contain ten to 

 :Us; after Lutman fi f tee n nuclei. The final segments however are 



