52 



In other respects I see no difference between the European 

 form and the Arctic form pumilum ^. 



The filaments in both the forms are covered with long, 

 rough hairs (Fig. 33 F, A etc.; Fig. 34 E) which perhaps serve 

 to retain the honey, but most probably are a substitute for the 

 appendages which occur on the anthers in V. uliginosum. 



The anthers lie close to the style and are rough with small 

 pointed protuberances (Fig. 33 Æ*, F) which doubtless cause the 

 insects to touch the anthers more violently and thus increase 

 the resulting shock. 



Fig. 34. Vaccinium Vitis-idæa. 

 A, Flower-bud (Stockholm); length 9mm. B, Open flower (Finmark; July 15, 1885); length 

 8 mm. from the stalk to the mouth of the corolla. C, Pollen tetrad. D, Flower (from 

 West Greenland, Auleitsivik Fjord); only a few pollen tetrads are left in the anthers. 

 E, Hair from the filament. F, Portion of a runner with a branch and a root in the axil 

 of a scale-leaf. (E. W.) 



Slight protandry occurs (H. Müller, p. 38, records homo- 

 gamy; Ekstam reports homogamy or slight protandry in Nova 

 Zembla). I have found the pores of the anthers open a long 

 time before the flowers expand, consequently, the pollen may 

 fall out immediately the flowers open. 



Self-pollination appears to be able to take place, owing to 



^ Ekstam writes, I.e. p. 427: "der Stempel ist in der Knospe hakenförmig 

 gebogen ; beim Eröffnen der Blüthe richtet er sich doch nicht ganz auf." 

 This I have not observed. 



