72 ANGIOSPERMAEDICOTYLEDONES 



and if it has come from another flower of the same species effects cross-pollination. 

 This plant according to Hildebrand's experiments is self-fertile. Kerner confirms 

 the fact of autogamy in the absence of insect-visits. 



159. C. lutea DC. (Hildebrand, op. cit; Herm. Muller, 'Fertilisation,' p. 99, 

 'Weit. Beob.,' I, p. 324.) The flower mechanism essentially agrees with that of 

 the last species. 



Visitors. Herm. Muller at Lippstadt observed Bombus agrorum F. 5, skg. 

 legitimately, and he also saw the following bees at Jena. 



1. Anthophora aestivalis Pz. 5 and J, skg.; 2. Bombus confusus Schenck 5, skg.; 

 3. B. lapidarius Z. 5, skg. ; 4. B. pomorum Pz. 5, skg. ; 5. B. rajellus K. 5, skg. ; 

 6. Eucera longicornis L. <j>, skg. ; 7. Halictus xanthopus K. 5, skg., or at least 

 attempting to do so ; 8. Osmia aurulenta Pz. 5, skg. ; 9. Psithyrus rupestris F. $>, skg. 



Schenck in Nassau observed the following bees skg.: Osmia cornuta Ltr., 

 and Podalirius acervorum L. In Dumfriesshire (Scott-Elliot, ' Flora of Dumfries- 

 shire,' p. 10) a humble-bee was noticed. 



160. C. acaulis Pers. Kerner states that the flower mechanism of this plant 

 resembles that of the last species. 



161. C. bracteata Pers., and 162. C. kolpakowskiana Regel. 

 Visitors. Loew observed in the Berlin Botanic Garden Anthophora pilipes 



F. $, skg. 



163. C. claviculata DC. (Knuth, Bot. Centralbl., Cassel, Hi, 1892, pp. 1-2; 

 Hart, Nature, London, x, 1874, p. 5.) The inconspicuous whitish flowers are only 

 6-8 mm. long and 2 mm. broad, and are arranged in racemes with few (at most six) 

 flowers. They are homogamous. The stamens are at first rather shorter than the 

 style, so that self-pollination can only occur when the pollen is discharged into 

 the dark lilac hood which ensheaths the stigma. The hood when pressed down 

 does not return to its original position : the stigma remains concealed under the 

 folded plate of the inner upper petal. The smallness of the flower makes an 

 exact study of its mechanism difficult. 



Visitors. I have not observed any, but have seen traces of the activity of 

 nectar-sucking insects. In many flowers the connection between the spurred petal 

 and the three others had been forcibly severed, so that these latter must have formed 

 a convenient platform for bees. Willis (' Fls. and Ins. in Gt. Britain,' Part 1) noticed 

 the following near the south coast of Scotland. 



Hymenoptera. Apidae: 1. Bombus agrorum F., freq., skg. ; 2. B. terrester Z., 

 skg. ; both effecting pollination. Every flower appears to set fruit. 



In Dumfriesshire (Scott-Elliot, ' Flora of Dumfriesshire,' p. 10) there have been 

 observed, Apis, 4 humble-bees, a short-tongued bee, and a Muscid. 



48. Fumaria L. 



Homogamous bee flowers. The nectar is secreted by a short process from the 

 upper filament. It is concealed in a short rounded pouch of the upper petals. The 

 other arrangements are as in Corydalis. 



164. F. officinalis L. (Hildebrand, op. cit., p. 450 ; Herm. Mailer, ' Ferti- 

 lisation,' pp. 99-100; MacLeod, Bot. Jaarb. Dodonaea, Ghent, v, 1893, pp. 188-90; 



