202 



cimens do not appear to differ from those from Scandinavia, 

 Spitzbergen and northern Asia. 



Protandry has been observed in Greenland, Norway 

 <Lindman), Sweden (Axell), and Nova Zembla (Ekstam), and I 

 found cultivated specimens (Hortiis Hafniensis) which were 

 slightly protandrous. 



Protogyny. It appears to be the rule for the flowers to 

 be slightly protogynous (Greenland and Nova Zembla), but if so, 

 homogamy ensues very soon. The styles bend outwards but 

 slightly, and the stigmas, which in this species, as in S. hie- 

 raciifolia, are evenly rounded, glistening and without papillæ 

 (Fig. 22 B), are usually able to retain pollen before any of the 

 anthers of the still quite erect stamens are open (Fig. 21 B). 

 The plant appears, however, to be homogamous during the 

 greater part of the flowering-period, and self-pollination 

 seems to be very common, and is, at any rate, easily possible 

 in case cross-pollination does not take place. The movements 

 of the stamens are those usual in the other species. But the 

 stamens almost always stand fairly erect, or else they lean 

 slightly forward over the middle of the flower, and even touch 

 the stigmas with their anthers (Figs. 21 A and 22 C). 



In other flowers I found the stamens somewhat spreading 

 when they began to be functional (as in Fig. 21 B), so that 

 here, self-pollination by direct contact does not seem to take 

 place. The specimens from Norway appear to me to be less 

 distinctly self-pollinating than are those from Spitzbergen and 

 Greenland. 



Insects, according to Ekstam flies, visit the flowers. 



Fruit usually ripens in West Greenland, in Ellesmere Land 

 (Simmons: "fruited richly '""l, Jan Mayen, Iceland, Norway, the 

 Færoes and Siberia. Half-ripe fruit was found by Ddsén and 

 Bat in East Greenland, by Ekstam in Nova Zembla, and by 

 Andersson and Hesselman in Spitzbergen. 



Pistillate flowers. In some specimens gathered by 



