306 ANGIOSPERMAEDICOTYLEDONES 



the pressure is removed. The upper filament is free distally, but its proximal half 

 is fused with the other nine, so as to leave no access to the base of the inner 

 surface of the staminal tube. This is in correspondence with the absence of nectar. 

 The anthers shed their reddish-yellow pollen before the flower opens. 





213. Colutea L. 



Nectar-yielding bee flowers, mostly yellow in colour ; with a brush arrangement. 

 714. C. arborescens L. (Kirchner, ' Beitrage,' pp. 42-3 ; Loew, Bliitenbiol. 

 Floristik,' p. 395 ; Knuth, ' Weit. Beob. ii. Bl. u. Insekt. a. d. nordfr. Ins./ p. 233, 

 ' Bloemenbiol. Bijdragen.') In this species, according to Kirchner, the thick-walled 

 calyx and the strong elastic claws keep the petals in place, and cause them to return 

 to their original position when the pressure exerted by an insect visitor is removed. 

 The vexillum is erect, and possesses a faintly marked nectar-guide. At the base 

 of its lamina there are two swellings, closely apposed to the alae. The latter are 

 small, and not united with the carina. Each possesses a finger-shaped process, 

 carried obliquely downwards and grasping the sexual column behind. The large 

 strong carina possesses two posteriorly directed triangular lappets which serve the 

 same purpose : its front part is thickened, so that the upper edges of the carinal petals 

 are kept close together. 



The style projects about 3 mm. beyond the anthers, and is rolled at the end 

 in such a way that its apex points downwards. On its inner side it bears a brush 

 about 5 mm. long directed obliquely upwards. The end of the style is obliquely 

 truncated, presenting a surface from the middle of which the small papilliform stigma 

 projects. This is surrounded by hairs which prevent self-pollination from 

 taking place. 



Considerable pressure is required to depress the carina. When this is effected, 

 the style with its adherent pollen first protrudes, and then the anthers. These 

 dehisce shortly before the flower opens, and part of the pollen which covers them 

 gets transferred to the hairs on the style. 



Kirchner observed a great many honey-bees visiting this species. Some of 

 them sucked legitimately, having settled on the middle of the flowers and succeeded 

 in depressing the alae and strong carina, so that the stigma and anthers protruded. 

 In doing so, they frequently effected self-pollination, though many of them dusted 

 the stigma with pollen brought on their legs from other flowers. Most of the 

 visitors, however, preferred to thrust in their proboscis laterally between vexillum and 

 alae, without causing protrusion of the sexual organs from the carina. Sometimes 

 all the bees visiting one particular shrub sucked legitimately, while all those which 

 settled on another stole the nectar. Humble-bees were also seen inserting the 

 proboscis obliquely into the bases of the flowers without depressing the carina. 

 My own observations essentially agree with those of Kirchner. 



Visitors. Apis mellifica Z. ^ appears frequently as a visitor. At Kiel (17. 6. '96) 

 and Sonderburg, in the island of Alsen (1. 8. '96), I almost always observed it to 

 settle on the side of the blossoms, skg. laterally without operating the flower 

 mechanism. Sometimes a bee would try to suck legitimately, but was too weak to 

 force its way into the almost closed flowers. A humble-bee (Bombus lapidarius 



