472 ANGIOSPERMAE—MONOCOTYLEDONES 



9 mm., and the three shorter by i2-6 mm. In the medium-styled forms the 

 corresponding elongations are respectively 13-5 and i8-5 mm., and in the short- 

 styled ones 10 and 15 mm. In the last form automatic self-pollination takes 

 place towards the end of anthesis, not only by means of the pollen which has fallen 

 on the inner side of the perianth lobes, but also by immediate contact of the stigma 

 with the tips of the pollen-covered anthers. 



Warnstorf also found the style in the flowers of the same bulb sometimes 

 as long as, and sometimes longer than, the stamens, with a short apical, violet 

 stigma, slightly bent downwards. The anthers were easily movable on conical, 

 stiff filaments, and dehisced laterally. The pollen-grains were golden-yellow in 

 colour, oily, clinging, irregularly tetrahedral, tuberculate, up to 43 f* in diameter. 



The flowers are generally bilaterally symmetrical, one perianth leaf being longer 

 and broader than the others, with the shortest one placed opposite to it. 



Visitors. — The following were recorded as stated. — 



Herm. Miiller observed the humble-bee Bombus hortorum L. 5, skg. In doing 

 so it brushed against the stigmas with one side of its body and the anthers with the 

 other, thus effecting cross-pollination. Knuth observed the honey-bee in particular. 

 It was not content with searching for nectar in open flowers, but forced its way 

 through the narrow entrances of flowers just opening, and dusted their already 

 mature stigmas, having come completely covered with pollen from flowers with 

 dehisced anthers. I also saw Musca domestica Z., skg. (.?), and numerous minute 

 muscids i\ mm. long, as well as two butterflies (Vanessa io L. and V. urticae L.) 

 creeping about the flower. On g.g.'^'j Knuth observed numerous po-dvg. hover- 

 flies (Eristalis tenax L. ; Syritta pipiens L. ; Syrphus arcuatus Fall. ; S. coroUae L. ; 

 S. pyrastri L.). The Lepidoptera and flies did not touch the stigmas regularly 

 in alighting, but generally flew straight to the anthers of the perianth lobes; this 

 occasional touching of the stigma was enough, however, to pollinate all, for I found 

 the stigmas of all open flowers thickly covered with pollen-grains. Self-pollination is 

 also frequently brought about by the small muscids mentioned : it could not take 

 place automatically in the flowers I observed, on account of the great projection 

 of the stigmas beyond the anthers. Knuth also noticed a slug, Limax cinerea (?) 

 [cf. Vol. I, p. 79, foot-note). 



914. Uvularia L. 



Kerner describes all the species of this genus as protogynous. 



2823. U. grandiflora Sm. — This species agrees exactly with Trillium 

 grandiflorum Salisi. (p. 470). 



2824. U. perfoliata L. (=U. flava Sm.). — 



VisiTOKS. — Loew (Berlin Botanic Garden) observed the small bee Halictus 

 cylindricus F. 5, po-cltg. 



915. Tricyrtis Wall. 



2825. T. pilosa Wall. (Kerner, 'Nat. Hist. PI.,' Eng. Ed. i, II, p. 351.)— 

 Self-poUination is at first excluded in this species by protogyny, and by the distance 

 between stigma and anthers, and only cross-pollination by means of insects possible. 

 Autogamy can take place later by the bending down of the st}'le. 



