224 ANGIOSPERMAE—DICOTYLEDONES 



off after they have shed their pollen, and the filaments bend back again to their 

 original position. After all the anthers have dropped off, the styles — which at 

 first project only about 7 mm. from the flower, and are close together — gradually 

 grow to a length of lo-ii mm. and spread out, the five stigmatic branches being 

 placed at the level which the anthers first occupied. After being pollinated, they 

 come together again. Besides these distinctly protandrous flowers, Schulz — on 

 garden plants — observed female ones, distributed gynodioeciously. 



Visitors. — Kirchner noticed numerous honey-bees at Hohenheim (Wurtem- 

 berg). I myself observed the same visitors at Kiel ; they hang upon the stamens 

 or styles in the manner described by Jordan. 



MacLeod saw 4 humble-bees in the Pyrenees (Bot. Jaarb. Dodonaea, Ghent, iii, 

 1 89 1, pp. 405-6). Darwin mentions humble-bees; Ricca humble-bees and Apis; 

 Plateau Eucera longicomis L. ; and Errera 29 Hymenoptera. 



Loew observed the following bees in the Berlin Botanic Garden. — 



I. Apis mellifica L. 5 (also the variety lividum), very freq., skg. ; 2. Bombus 

 hortorum L. j, skg. ; 3. B. lapidarius L. 5, do. ; 4. B. rajellus K. j, do. ; 5. Coelioxys 

 elongata Lep. 5, do. ; 6. Halictus albipes F. 5, do. 



565. G. macrorrhizum L. — This species also, according to Jordan, has 

 pendulous flowers that are visited in the same way as those of the last species. 

 Hildebrand (Bot. Ztg., Leipzig, xxvii, 1869, pp. 479-81) states that it is equally 

 protandrous ; at first purely female flowers appear. 



566. G. dissectum L. (Herm. Miiller, ' Ferlihsation,' p. 156, ' Weit. Beob.,' 

 IL pp. 217-18 ; MacLeod, Bot. Jaarb. Dodonaea, Ghent, vi, 1894, p. 233 ; Kirchner, 

 'Flora V. Stuttgart,' p. 338; Warnstorf, Verb. bot. Ver., Berlin, xxxviii, 1896.) — This 

 species, unlike those hitherto dealt with, bears protogynous flowers, with very 

 persistent stigmas. In the sunshine they open only far enough to be funnel-shaped, 

 with an entrance 6-8 mm. The stigmas are mature, and their branches expanded 

 when the flower opens, while the anthers which closely surround them are still unripe. 

 Later on the anthers dehisce successively, covering the stigmas with pollen. Hermann 

 Muller's researches show that this automatic self-pollination is eflFective. Insects 

 visiting the flowers may eff'ect self- as well as cross-polhnation, but the visitors 

 are very few. 



Warnstorf describes the flowers as homogamous and autogamous ; the stigmatic 

 papillae are mature at the time when the anthers dehisce. The latter are blue in 

 colour, and closely applied to the stigmatic branches, so that self-pollination is 

 inevitable. The pollen-grains are bluish-white, spherical, densely tuberculated, 

 cohering, with an average diameter of 63 /x. 



Visitors. — Herm. Miiller observed in Thuringia only a bee — Andrena gwynana 

 K. 5 and S, skg. — and 2 Diptera — Occemyia atra F. and Merodon aeneus Mg., 

 both skg. 



Schletterer saw the beautiful bee Osmia versicolor Lir., and the saw-fly Amasis 

 laeta F., at Pola. 



MacLeod noticed a Lepidopterid and a fly in the Pyrenees (Bot. Jaarb. Dodonaea, 

 Ghent, iii, 1891, p. 402). 



In Dumfriesshire a hover-fly and 2 Muscids have been recorded (Scott-Elliot, 

 ' Flora of Dumfriesshire,' p. 39). 



