276 H. MARSHALL WARD. 



If the lower Ustilaginese are allied to the Chytrideee by 

 means of Entyloraa and Protorayces, we seem to have a 

 satisfactory position for the former group. Of course, if this 

 view be accepted, the resting spores of the Ustilaginese are 

 the homologues of oogonia, which become developed apo- 

 gamously. The preceding facts may be summarised in the 

 following diagram : 



Ustilagiuese 



/ 

 Ascomycetes, &c. Protomyces, &c. 



/ 

 Olpidiopsis, &c. 



/ r N.B.— If these forms produce 



"Rrvsinhpfp PolvDha^us &c 5 zygospores, they may be allied 



Jl^rysipliese roiypliagus, ffic. < ^o the Zygomycetes also; this 



V^could not be shown on one plane. 

 Ancylisttse 



/ 

 Peronosporese 



/ \ 



Zygomycetes Saprolegnise. 



Continuing our survey of De Bary's memoir, we may pass 

 over the opinion respecting the Saccharomyces, and proceed 

 to the part dealing with a much more difficult aud important 

 series of forms. As the author showed in 1879/ the Tremel- 

 lini may well be looked upon as derived from Uredinese and 

 allies; while those Uredinese which form secidia resemble 

 the Ascomycetes in so many points of structure and develop- 

 ment, that we may regard them as closely allied. The formation 

 of three forms of conidia (uredospores, teleutospores, 

 and sporidia) may be in part due to specialisation; but it 

 must be remembered that the Ascomycetes are also in the 

 habit of forming many and various conidia. It is, however, in 

 the many points of resemblance between the secidia and 

 perithecia, and the spermagonia and spermatia of both groups, 

 that the alliance appears most clearly. True, no observer has 

 found a trace of sexual organs in the young secidia; but the 

 same is certainly true for the perithecia of many Ascomycetes. 



» 'Bot. Zeir..; 1S79, p. S25, &c. 



