46 



blossom is ;ui I'loiiLjation of the axis botweon calyx and corolla (Fi.i,'. ob'). 

 At tirst si,i,'!it it inl.L'ht ajipear that the presence ot the parasite was snffi- 

 (•i(>nt to chan.ne the sex of the plant, for the fnnsns, when present in the 

 pistillate plant, leads regularly to the development of stamens and the 

 snppression of the jiisril. However, in this case the capacity to develop 

 stamens must be assumed to be present in the pistillate plant, and the 

 parasite is able to niduce the conditions necessary to their formation and 

 the suppression of the ]^istil. and thus provide for the develoinnent of its 

 own spores. Extensi\e and elal)orate experiments bj' Strasburger upon 

 uninfected plants with tlie view of duplicating the effects produced by the 

 parasite, led to no definite results. 



Fig. 3. 



Fig. 3. Staminate aud pistillate Howers of Lychnis dioica L., halved longitudinally. 

 a. normal staminate Hower. b. normal pistillate Mower, a' staminate flower affected 

 by the smut, Ustilago violaceae; the anthers contain smut spores instead of pollen. 

 &' pistillate flower similarly affected: the pistil has remained rudimentary while anthers 

 have been developed, which, however, bear only smut spores. The presence of the 

 parasite has induced the development of anthers, the members of the flower bearing 

 male spores instead of the parts bearing the female spores,— After Strasburger. 



