GERANIACEAE 237 



Visitors. In the Kiel Botanic Garden I observed no bees, but repeatedly saw 

 small hover-fly, Syrphus balteatus Deg., which sucked persistently, and constantly 

 isited several flowers in succession. 



On September 10, 1897, I noticed (' Blutenbiol. Notizen') hover-flies, almost 

 exclusively Syrphus corollae F., in the Kiel Ober-Realschule Garden. This insect 

 first hovers in the sunshine in front of the flower, approaches to within a few 

 millimetres, again hovers a little further away, and repeats this sport several times 

 till finally it settles down to suck, and to devour pollen. I also observed Syrphus 

 ribesii L., the same morning, visiting the flowers several times, while Apis avoided 

 them, flying between the inflorescences and steadily visiting the flowers of Sedum 

 maximum. I therefore regard I. parviflora as a hover-fly flower, not as a bee flower 

 or humble-bee flower, like the other cultivated species. 



The insect visitors of the species are generally 

 very few, but as in spite of this all the flowers set 

 fruits, it must be assumed that the plant is self- 

 fertile. 



601. I. Balsamina L. (Sprengel, 'Entd. Geh.,' 

 p. 400; Hildebrand, Bot. Ztg., Leipzig, xxv, 1867; 

 Delpino, ' Sugli appar. d. fecondaz. nelle piante 

 autocarp.,' pp. 30-1.) The mechanism of the flower 

 of this species agrees with that of I. Noli-tangere. 

 In younger flowers the visitors are dusted with 

 pollen, which they transfer to the stigmas of older fig. 72. impatiens parviflora, 

 ones, in which the stamens have already dropped off. ^^""fnXrs F ' Wer ^ h m 



Visitors. These are bees (humble-bees). Prunet 

 (Rev. gen. Bot., Paris, iv, 1892) observed numerous insects, especially Apis, Bombus 

 hortorum Z., B. terrester L., and Polistes gallica L. 



602. I. glanduligera Lindl. (=1. Roylei Walp.). (Delpino, ' Ult. Oss.,' II ; 

 Hildebrand, Bot. Ztg., Leipzig, xxv, 1867; Stadler, 'Beitrage'; Loew, Bot. Jahrb., 

 Leipzig, xiv, 1891, pp. 166-82; Knuth, ' Bloemenbiol. Bijdragen.') This commonly 

 cultivated East Indian species bears well-marked humble-bee flowers. The large 

 markedly protandrous purple-red blossoms have a short spur, and are of such a size 

 that the body of a humble-bee just finds sufficient room in them. These insects, 

 when visiting flowers in the first stage, get their backs dusted with pollen by contact 

 with the anthers, and when visiting flowers in the second stage, they brush the stigma 

 with the same part of the body. They disappear altogether in the blossoms when 

 sucking nectar from the short blunt spurs, and leave flowers in the male condition 

 with a streak of pollen 1-3 mm. in length on their backs. 



This view of the flower mechanism, represented in my figure, is also taken 

 by Delpino and Hildebrand. Though Loew does not dispute the possibility of such 

 a mechanism, it does not appear to him to be the normal one indicated by the 

 structure of the flowers. He points out the remarkable stigma-like structure of the 

 ligules springing from the bases of the filaments, the bringing together of these 

 in a narrow space readily accessible from the front, the position of this space at 

 the most anterior point of the androecium immediately above the surface where 



