GRAMINEAE 



527 



a short time the glumes close again and grip the stamens and stigmas in doing so, 

 the former with the emptied anthers at the tip, and the latter on both sides at the 

 base of the flower. The stigmas seem to be still receptive when the anthers have 

 fallen. Automatic self-pollination is excluded. 



Warnstorf says that this species has dehisced by 6 a.m. Besides plants in which 

 the glumes are closed by this time and the empty yellow anthers hang far out, there 

 are others with widely opened glumes, and stigmatic branches and stamens situated 

 within them ; self-pollination is inevitable in this case. 



3037. G. aquatica Wahlenb. — This species is homogamous, and the anthers 

 dehisce between 5 and 6 p.m. (probably for the second time). The pollen-grains 

 are whitish in colour, pyramidal in form, up to 50 /x long and 35 /n broad (Warnstorf, 

 Verh. bot. Ver., Berlin, xxxviii, 1896). 



3038. G. distans (= Festuca distans Kunlh). — This species is homogamous. 

 When the glumes diverge the anthers are situated on relatively short, stiff filaments, 

 and project little beyond the long stigmatic branches, autogamy thus being easily 

 possible. The anthers dehisce at 6 a.m. (Warnstorf, Verh. bot. Ver., Berlin, 

 xxxviii^ 1896). 



1006. Molinia Schrank. 



3039. M. caerulea Moench. — For the interesting observations on this species 

 by Ludwig, see p. 516. 



1007. Dactylis L. 



3040. D. glomerata L. — Kirchner (' Beitrage,' p. 9) describes this species as 

 freely protogynous with persistent stigmas. The filaments have not yet completely 

 elongated nor the anthers dehisced when the stigmas are receptive and protrude 

 laterally between the divergent glumes. When the filaments have reached their full 

 length they do not bend downwards but remain stretched out rather stiffly, so that 

 automatic self-pollination can now easily take place. The stigmas appear to be still 

 receptive after the anthers have emptied their pollen. Hildebrand (' Bestaubungsverh. 

 d. Gramineen,' p. 756) describes the flowers as homogamous, but states that cross- 

 pollination is favoured. Kerner says that the flowers open between 6 and 7 a.m., 

 but according to Warnstorf this takes place between 6 and 9 a.m. 



Visitors. — Ludwig observed flies. 



1008. Cynosurus L. 



3041. C. cristatus L. — Hildebrand ('Bestaubungsverh. d. Gramineen,' p. 758) 

 describes this species as homogamous. Cross- and self-poUination are favoured to 

 about the same extent. The yellow or violet anthers are at first situated on long, stiff 

 filaments, but soon become pendulous ; the stigmatic branches project widely. The 

 anthers dehisce between 6 and 7 a.m. The pollen-grains are pyramidal in form, 

 irregularly whitish in colour, up to 37 /x long and 31 yu. broad (Warnstorf, Verh. bot. 

 Ver., Berlin, xxxviii, 1896). 



Kirchner ('Flora v. Stuttgart,' p. 143) describes the anthers as partly red, partly 

 yellow in colour. 



