306 



THE CROSSING OF FLOWERS. 



Fig. 290— Flower of the Kiie (iiuta gra- 

 veolens) x 3. (After Baillou.) 



elongate, and thus assume a position overhanging the honey receptacles, where they 

 are inevitably brushed by insects. 



For the tenth group we may select the Rue {Ruta; see fig. 290) as an example. 

 The flower contains ten anthers supported by stiff filaments, arranged in the form 

 of a star. First, one of these filaments bends up and places its antlier in the middle 

 of the flower, just in the way to the fleshy ring which secretes nectar at the base 

 of the pistil: it remains there about a day, and then bends back and resumes its 

 former position. While the first stamen is bending back, a second rises up and 

 undergoes the same movements. And so it goes on until the ten anthers have all 

 stood in the centre of the flower in turn and discharged their pollen there. When, 

 finally, the last stamen has bent back again, the stigma, which has meanwhile 

 matured, is seen in the centre of the flower where the anthers have successively 



shed their pollen. 



A process which is closely connected with the 

 interchange of position of anthers and stigmas, in 

 bringing about cross-fertilization between herma- 

 phrodite flowers, is the separation and subsequent 

 falling away of the stigmas when the surrounding 

 anthers begin to dehisce. We may take one of 

 the Urticaceae, the Wall Pellitory {Parietaria; see 

 figs. 2912'3''*), as a type of this. The stigma is 

 always developed before the flowers open in this 

 plant, and at the commencement of the flowering period it may be seen projecting 

 like a dusting-brush from the green flower-bud (fig. 291^). The curved anther- 

 filaments are at this time coiled like watch-springs and covered over by the as yet 

 unopened floral-leaves. Before these filaments jerk up and scatter their pollen the 

 stigma withers and shrivels up and the style becomes detached from the ovary. 

 It falls off with the dried-up stigma, so that, when the pollen is liberated from the 

 anthers, the ovary is terminated merely by a small stump which is really the 

 withered remnant of the fallen style (fig. 291 ^). 



The falling of the anthers and stamens at the time when the adjoining stigmas 

 become mature is of much more frequent occurrence than the detachment and 

 falling of the stigmas befoi-e the pollen is shed. In the flowers of Balsams {Im- 

 patiens glanduloaa, Nolitangere, tricornis, &c.) the anthers are united into a kind 

 of cap which arches over the stigma. As soon as the flower has opened and become 

 accessible to flying insects, the anthers dehisce and only the cap which they form 

 can be seen at the entrance to the flower. Later, the filaments of the anthers 

 become detached, and the anther-cap falls out of the flower; only the stigma, 

 which has meanwhile matured, is now visible in the middle of the flower. The 

 large-flowered species of Crane's-bill (e.g. Geranium argenteuvi, pratense, sylvati- 

 cum; see fig. 291 ') have a similar arrangement. Two of the anthers which have 

 hitlierto been covered by the petals dehisce almost simultaneously with tlie opening 

 of the flower; the others then open in a certain order and expose their pollen in 



