21 
of the latter as clouds of dust. I very often found shed pollen 
upon the inner parts of the flower. Self-pollination must 
undoubtedly take place easily because the anthers always stand 
around the style with their pores turned towards the stigma 
which is below or almost on the same level with them (Fig. 10 AK; 
Fig. 11 A). Almost all the tetrads I found on the stigma occurred 
at its edge. The normal arrangement seems to be for the 
anthers to stand immediately below the stigma or around it. 
I find that both in Greenland and in Finmark some 
variations occur in the relation between the stamens and the 
pistil, as also may be seen from the figures. The stigma may 
occur from 1 to 2 mm. above the anthers. But I have not 
found the anthers to stand as close to the stigma in the 
European as in the Greenland flowers (Fig. 10 K). These 
differences in the relative length of the organs in question do 
not seem to depend upon the age of the flower; the anthers 
appear however to stand closer to the stigma in the young 
flowers than in the older ones. Linnman has also noticed these 
variations, and has observed the stigma to be situated even 
lower than the base of the anthers (l.c. p. 72). 
The flower is also distinctly adapted for cross-pollination 
by insects. This may take place easily when an insect thrusts 
its proboscis through the narrow throat of the corolla down to 
the honey, which is secreted by a yellow, ten-lobed nectary 
(Fig. 12 E). The filaments are hairy at their bases, the hairs 
serving in this as in other cases to retain the honey in the 
drooping flower (Fig. 10 E; Fig. 12 A). 
We may assume humble bees to be among the insect- 
visitors. In Finmark | saw them visit the flowers, and on the 
Dovre Lisoman saw Bombus alpinus, in Lapland Sxorrsperc 
Bombus serimshiranus D., in Varanger Porrius Bombus lapponicus 
Fabr. and B. nivalis Dheb. do so. In Finmark | often observed 
styles and stamens which had been bitten off by ants. 
After fertilisation the corolla falls off and the calyx with 
