RANUNCULACEAE 41 



the third day these also have moved down, and are replaced by a third set of eight, 

 and so on in succession, until in six days' time all the stamens are withered and 

 prostrate. 



Meanwhile the styles are erect, but they gradually turn over in a somewhat 

 spiral fashion, curving outwards and downwards so as to be almost horizontal when 

 all the anthers have withered. The stigma, which is in the form of a longitudinal 

 seam extending from the base of the style to its tip, must now come into contact 

 with insects engaged in sucking nectar, and will be dusted with pollen should these 

 happen to come from a flower of the same species in the first condition. After 

 the styles have occupied this position for three or four days, they again become erect. 

 I observed in garden plants that automatic self-pollination sometimes took place 

 by a spiral downward movement of the styles, in cases where insects had not brought 

 about cross-fertilization. Terraciano states that the pollen of the lower stamens is 

 unable to effect self-fertilization, and that automatic self-pollination usually occurs 

 by the stigmas coming into contact with the anthers of the upper stamens. This 

 investigator observed no insect visitors on wild plants : on the contrary, he found 

 that this species, and also N. sativa Z., N. damascena Z., N. Bourgaei ford., N. 

 foeniculacea DC, and N. gallica Jord., produced numerous seeds capable of 

 germinating without the aid of pollinating agents, i.e. as the results of automatic 

 self-pollination. 



Visitors. Sprengel long ago observed bees which were precisely adapted to 

 the dimensions of the flowers. Bees alone are clever enough to lift the nectar-covers. 

 I have seen Apis, and also Bombus lapidarius L. , skilfully opening the nectaries, 

 and regularly effecting cross-pollination. I also saw Vanessa Io L., as a useless 

 flower-guest. 



Friese observed in Hungary, the silk-bee Colletes punctatus Mocs. t> not 

 infrequent, 5 very infrequent, and its parasite the cuckoo-bee, Epeolus fasciatus Friese 

 (= E. transitorius Friese), not infrequent. 



88. N. sativa L. (Knuth, 'Bloemenbiol. Bijdragen') agrees in the structure 

 of its flowers with the last species, but the nectaries are slightly different. 



Visitors. I saw in the Kiel Botanic Garden the honey-bee and 2 humble- 

 bees (Bombus terrester Z. tjf, and B. lapidarius Z. ^ and J) on this species and N. 

 damascena. These opened the nectaries one after the other and sucked nectar. 

 While doing so, the upper side of the thorax was in contact with the pollen-covered 

 anthers, or the receptive stigmas, thus effecting crossing. 



89. N. damascena L. (Knuth, ' Bloemenbiol. Bijdragen,' ' Blutenbiol. Notizen '; 

 Herm. Miiller, 'Weit. Beob.,' I, p. 322.) The flowers agree in structure with those 

 of the last two species. Here automatic self-pollination regularly takes place by 

 spiral twisting downwards of the styles, and, as the flowers always set fruits, it would 

 appear to be effective. (Cf. Fig. 14,^.) 



Visitors. Vide supra, No. 88. 



Buddeberg observed the two following bees in Nassau. 



1. Ceratina callosa F. J, busy among the stamens; 2. Prosopis confusa 

 Nyl t, skg. 



