PREFACE TO THE REPRINT OP 1884. ix 



of gynodioeciousness. Three different forms occur, but 

 on any given plant the flowers are of one kind. There 

 is a normal hermaphrodite form, and two divergent 

 forms which are practically male and female, and which 

 are distinguished from the hermaphrodite form by hav- 

 ing conspicuous sexless ray-florets ; of the two, the male 

 flowers are the more conspicuous. The female florets 

 have shrunken anthers devoid of pollen; the male florets 

 have pistils which do not open, and are therefore fune- 

 tionless. Numerous gradational forms exist which ren- 

 der the whole case especially instructive, and it was a 

 study of these gradations which induced Miiller to give 

 up his theory of gynodioecious plants. Miiller formerly 

 explained the origin of gynodioeciousness by supposing 

 that those flowers which are smaller and less conspicu- 

 ous than the average tend to be visited last by insects, 

 so that their pollen is useless. In Centaurea the reduc- 

 tion of anthers is found beginning in flower-heads which 

 are not less conspicuous than the average. Miiller there- 

 fore gives up his former theory and agrees with the view 

 proposed by my father.* 



Potoni^ (' Sitzb. d. Ges. Naturforsch. Freunde zu 

 Berlin,' 1880, p. 85, quoted in the ' Bot. Zeitung,' 1880, 

 p. 749) believes that in the gynodicecious Salvia praten- 

 sis the existence of a female form serves to ensure fer- 

 tilisation by a distinct plant. 



But H. MuUer ('Bot. Zeitung,' 1880, p. 749) shows 

 that in the hermaphrodites, bees commonly visit the 

 lower and temporarily female flowers before passing on 

 to the upper male flowers, and that this ensures cross- 

 fertilisation between different plants. 



Solms-Laubach ('Abhand. K. Gesell. Wiss. Got- 

 tingen,' xxviii., and 'Kosmos,' 1881) has given in his 



* A short paper by H. Muller on gynodiosciousness in the genns 

 Dianthus, appeared in 'Nature,' 1881, xxiv. 



