4<H 



ANGIOSPERMAEDICOTYLEDONES 



bend over the matured stigmas in such a way that autogamy would seem to 

 inevitable. Andersson and Hesselman observed an apetalous variety, in which 

 petals were modified into stamens, the pollen of which was partially abnormal. 



Visitors. Numerous flies were observed by Lindman on the Dovrefjeld. As 

 frequent visitors in Spitzbergen Holmgren noticed the Ichneumonids Hemiteles 

 septentrionalis Holmgr., and Orthocentrus pedestris Holmgr.; also a Muscid, Aricia 

 (Chortophila) megastoma Bohem. On seven different days of July and August 

 Ekstam saw many small Diptera in Spitzbergen. 



957. S. rivularis L. (Warming, Bot. Tids., Kjobenhavn, xvi, 1888, pp. 7- 

 10.) The inconspicuous flowers of this boreal and arctic species are, according 

 to Lindman and Warming, at first slightly protogynous, and afterwards homogamous, 

 self-pollination being easily possible. Fruits are set early and rapidly. Wanning 

 observed purely female plants with vestigial stamens in Spitzbergen. 



Andersson and Hesselman describe the species as flowering in Spitzbergen from 

 the beginning of July to the end of August ; in Beeren Island open blossoms were 

 noticed in favourable spots as early as June 18 (' Bidrag till Kanned. om Spetsbergens 

 o. Beeren Eil. Karlvaxtflora,' pp. 29-30). The pollen is normally developed, and 

 fruits appear to ripen regularly. Ekstam says that in Spitzbergen the flowers are 

 odourless, white or pale red in colour (bright red in some Beeren Island specimens), 



Fig. 130. Saxifraga stenopetala, Gaud, (after Herm. Mailer). A. Flower at the beginning of the 

 first (male) stage. B. Flower towards the. end of the same stage. C. Flower in the second (female) 

 stage, in section, a, outer stamen ; n, nectary ; ov, ovary. 



and 6-8 mm. in diameter, or sometimes 5-10 mm. (' Bliitenbiol. Beob. a. Spitz- 

 bergen,' pp. 17-18). In still incompletely open flowers, the anthers are beginning to 

 dehisce, and the stigmas are papillose ; homogamy may therefore be inferred. 



According to Abromeit, several varieties have been recorded for Greenland, 

 including purpurascens Lange with reddish petals (' Bot. Ergeb. von Drygalski s 

 Gronlandsexped./ pp. 34-5). 



Visitors. Ekstam noticed large and small flies on two days of July in 

 Spitzbergen. 



958. S. stenopetala Gaud. ( = S. aphylla Slernb). (Herm. Mailer, 'Fertilisa- 

 tion,' p. 244, ' Alpenblumen,' pp. 108-9.) I n trns markedly protandrous species 

 self-pollination is excluded (cf. Fig. 130). The nectar is exposed, and the flower 

 mechanism is adapted to flies. 



Visitors. Flies have been observed. 



