38 



few insects can be astir; and those we may assume will be 

 humble-bees (in Lappmark Bombus lapponicus Fabr. and B. 

 nivalis Dhlb.; Poppics). Also, the flowers are so well hidden by 

 branches and leaves that they can scarcely be seen. Therefore, 

 the visit of insects cannot be regarded as playing any pro- 

 minent part in pollination. This may be correlated with the fact 

 that the appendages of the anthers are so unusually small, or 

 even entirely absent; they appear to be a diminishing structure 

 on the point of becoming rudimentary. The fact that the pores 

 of the anthers are larger than in the other Ericaceae also sug- 

 gests self-pollination, for the pollen may be shed very easily 

 (Fig. 24 F, G); even in quite young flowers I have found the anthers 

 completely emptied of their contents, and pollen-grains occur in 

 quantities all over the corolla, adhering to its hairs, which are 

 denser than in Л. Uva-ursi and reach as far as to the margin 

 of the corolla {Fig.24 B, D, Ж, L). The function of the hairs 

 appears to be especially that of retaining the pollen-grains so 

 that they may fall upon the stigma when opportunity offers. 

 The narrowness of the throat is another contributory feature 

 which tends to ensure self-pollination. I have hardly ever 

 examined any open flower without finding pollen upon its stigma, 

 and as a natural consequence of this, fruit is abundantly pro- 

 duced. 



Л. alpitia appears therefore to be particularly adapted to 

 self-pollination, while A. Uva-ursi — by the brighter colour and 

 greater conspicuousness of its corolla, by the more scattered 

 hairs in the interior of the same, and by the roughness and 

 length of the appendages of its anthers — appears to be better 

 adapted to cross-pollination by insects. 



The drupes are spherical, black, and about 1 cm. in dia- 

 meter, in the north of Scandinavia they ripen as early as 

 August. According to Haglund and Ekstam the seeds are dis- 

 persed by frugivorous birds. 



