4 ANGIOSPERMAE—DICOTYLEDONES 



anthers dehisce, so that the flowers are in the first or male condition, in which 

 they yield pollen to insects, but are incapable of retaining on their stigmas any 

 that may be brought. Automatic self-pollination is equally impossible at this stage. 

 The outward curving and dehiscing of the stamens progresses centripetally, but 

 before the innermost ones have dehisced the stigmas have matured, and are liable 

 to be touched by such pollen-covered insects as may alight in the middle of the 

 flower. Bees collecting the abundant pollen almost invariably alight in the centre — 

 according to Hermann MuUer's observations — while pollen-devouring flies are very 

 erratic in their mode of settling and, creeping about the flowers, may effect either 

 cross- or self-pollination. Failing insect-visits, self-pollination may easily be effected 

 by contact of the stigmas — which remain receptive — with the innermost stamens. 



Visitors. — Hermann Miiller observed the following. — A. Coleoptera. Scara- 

 baeidae : i. Trichius fasciatus Z., devouring the anthers. B. Diptera. (a) Muscidae : 

 2. Prosena siberita Z'.; (h) Syrphidae : 3. Eristalis arbustorum Z.; 4. E. sepulcralis 

 Z.; 5. Helophilus floreus Z.; 6. Syrphus pyrastri Z. ; 7. Syritta pipiens Z.; 8. Xylota 

 ignava Pz.; 9. X. lenta Mg.; all po-dvg. C. Hymenoptera. (a) Apidae: 10. 

 Andrena albicans i)/M. 5; 11. A. gwynana A'. 5; 12. Apis meilifica Z. 5; iS.Bombus 

 terrester Z. §; 14. Halictus sexnotatus K. 5; 15. Osmia rufa Z. 5; 16. Prosopis 

 signata /"z. 5 ; all po-cltg. {b) Sphegidae: 17. Gorytes mystaceus Z., perhaps only 

 hunting flies; 18. Oxybelus uniglumis Z., po-dvg. (c) Vespidae: 19. Odynerus 

 parietum Z. 5, as 17. Handlirsch mentions as a visitor the fossorial wasp Gorytes 

 mystaceus Z. 



On garden plants I saw only one pollen-devouring hover-fly, Eristalis tenax Z. 



3. C. Viticella L. (Knuth, ' Bloemenbiol. Bijdragen ') is nectarless Uke the 

 preceding, despite its very large dark-violet, blue, or red pollen flowers. I have 

 only once observed the honey-bee collecting pollen on plants that were grown 

 in Kiel for the purpose of covering a bower. No observations are available from 

 the Mediterranean region, which is the home of this form. 



4. C. balearica Rich. (=C. cirrhosa Z.) is a nectar flower indigenous in the 

 Mediterranean region. The outer stamens — according to Delpino — are modified 

 into spoon-shaped nectaries. The same observer mentions Bombus and Xylocopa 

 as visitors of this species. 



5. C. integrifolia L. is also a nectar flower. According to Delpino (' Appli- 

 cazione d. teoria Darwiniana,' p. 8) the inner stamens secrete nectar. The pendulous 

 flowers — according to Kerner ('Nat. Hist. PI.,' Eng. Ed. i, II, pp. 349-50) — are 

 protogynous for a short time, and therefore adapted for cross-pollination at the 

 beginning of flowering. The stamens lie close together so as to form a short tube, 

 in the base of which the numerous still immature stigmas are situated, while the 

 outer anthers have already dehisced, thus furthering cross-pollination. The anthers 

 of the inner stamens gradually dehisce, but, owing to the pendulous position of 

 the flower, would be unable to effect self-pollination were there not an elongation 

 of the carpels during the last two days of flowering, so that if pollination has not 

 been effected by insects, the stigmas — spreading out to some extent — receive some 

 of the pollen still adhering to the stamens. 



6. C. angustifolia Jacq. — 



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

 Hymenoptera. Apidae: Bombus terrester Z. 5 , po-cltg. 



