68 



the Alpine specimen. We may be justified in connecting this 

 tendency toward self-poUination with the scarcity of insects in 

 the country in question. It will facilitate self-pollination if the 

 interior of the corolla by its covering of hairs is able to gather 

 up the pollen grains and retain them as mentioned above. 

 Some species appear to me to be less well-adapted for 

 self-pollination, especially of Ledum and Rhododendron. The 

 former with its richly-flowering inflorescence recalls somewhat 

 that of the Umbelliferæ, and geitonogamy must evidently be 

 able to take place easily in it. 



All the flowers are for the most part homogamous for a 

 long period, but some begin by being protogynous for a short 

 time [Arctostaphylos alpina, Cassiope hypnoides, Ledum, Phyl- 

 lodoce, Loiseleuria, Rhododendron) and others protandrous, also 

 for a short time [Cassiope tetragona, Vaccinium Myrtillus, V. 

 Oxycoccos, V. uliginosum, V. Vitis-idœa). 



Probably self-pollination sometimes takes place before the 

 flower opens, so that a kind of cleistogamy occurs, because 

 as has been mentioned, the anthers generally open while still 

 in the bud, and in some cases, I even found the pollen shed 

 (Arctostaphylos TJva-ursi). 1 found the anthers dehisced, and 

 fully developed pollen grains lying loose in the buds in this 

 species and in the following; Cassiope tetragona, С hypnoides, 

 Andromeda polifolia, Ledum palustre, Vaccinium Vitis-idœa, 

 V. uliginosum, V. Oxycoccos, Lyonia, as also in Erica Tetralix, 

 E. cinerea and Calluna vidgaris. In the greater number of 

 these species I also found that the stigma in the bud was 

 viscid and capable of retaining the pollen, or was at least able 

 to do so in flowers which had just expanded. In a greenhouse 

 in Copenhagen, in a tTower of Lyonia not yet expanded, I found 

 pollen upon the stigma. In Finmark, in a bud of Cassiope 

 tetragona^ I found the stigma viscid and the anthers open ; the 

 pollen could not however fall out, but did so immediately that 

 the anthers became dry after the expansion of the bud. This 



