234 ANGIOSPERMAEDICOTYLEDONES 



however, was not adapted to effect fertilization, for the lazy and stupid insect 

 mistook the nectar-cover for the nectar-reservoir, and thrust its proboscis into it 

 to find, as it had been raining, only drops of water'). 



In Schleswig-Holstein, Mecklenburg, Pomerania, and Thuringia, I observed 

 Bombus hortorum Z., also the variety nigricans Schmiedekn., skg., and spending some 

 seconds in each flower. The spur is 25 mm. long, so that the proboscis of this 

 humble-bee being only 21 mm. in length would not reach its tip, but the entrance 

 to the spur is so wide, that the insect can push its head 5 mm. down the tube, 

 and thus easily exhaust its contents. The garden humble-bee is, therefore, a regular 

 fertilizer of the species. Besides this insect, I once saw an earwig (Forficula auri- 

 cularia) half-buried in the spur, and remaining there so persistently that I was able 

 to pluck the flower, and observe its behaviour. It was evidently able to get some 

 of the nectar that filled the spur to an unusual height. A third visitor that I observed 

 was Apis mellifica Z. ^- This first penetrated as far as possible into the spur, 

 and then made vain efforts to suck. When it did not succeed in getting nectar 

 in sufficient quantity, it proceeded to collect pollen, and, taught by experience, made 

 no more fruitless attempts at sucking. Alfken saw Bombus hortorum Z. $, skg., 

 at Bremen. 



594. T. minus L. (Knuth, ' Bloemenbiol. Bijdragen.') Like the last species, 

 this plant is a native of Peru, and its flower mechanism is very similar. The spur, 

 however, is usually considerably longer, i.e. 25-35 mm - None of our German 

 humble-bees or bees are able therefore to reach to the ends of the longest spurs, 

 even by thrusting in their head, which is about 5 mm. in length. One stamen after 

 another becomes erect, placing its pollen-covered anther in the entrance of the 

 flower, and after shedding its pollen again bending back against the corolla. When 

 all the anthers have dehisced, the three-lobed stigma occupies the entrance of the 

 flower, so that it is touched by nectar-sucking visitors. The three lower petals 

 are fringed on their inner surface, so as to prevent visitors from forcing their way 

 into the lower part of the flower. They are therefore compelled to seek the entrance 

 to the spur, from above the upwardly dehiscing anther, or the stigma. 



3. Tribe Oxalideae DC 

 This tribe is represented by the genus 



175. Oxalis L. 



(H. von Mohl, Bot. Ztg., Leipzig, xxi, 1863; Hildebrand, Abh. Ak. Wiss., 

 Berlin, 1866, Bot. Ztg., Leipzig, xxix, 187 1, 'Die Lebensverhaltnisse der Oxalis- 

 Arten,' Jena, 1884.) 



Flowers homogamous, with half-concealed nectar secreted in their bases. 



While the three German species appear in only one form, a large number 

 of foreign species are trimorphous or dimorphous. Experiments in artificial fer- 

 tilization made by Hildebrand with trimorphous species, confirm the law established 

 by Darwin for dimorphous plants, and for Lythrum Salicaria, i.e. that there is 

 greatest fertility with legitimate fertilization. In some species (including O. Aceto- 



