ONAGRARIEAE 



445 



are usually on a level with the stigmas, autogamy is the rule, and Kerner says it 

 may take place as early as the first day of anthesis. It is excluded in flowers where 

 the long stamens do not reach the stigmas. 



Visitors. — These are few. Herm. MUller observed flies (Anthomyia j, po-cltg.), 

 and a butterfly (Pieris napi L., skg.) (' Weit. Beob.,' II, p. 237). 



Schletterer records the humble-bee Bombus pomorum for the Tyrol. MacLeod 

 saw a hover-fly and a beetle in Flanders (Bot. Jaarb. Dodonaea, Ghent, vi, 1894, 

 p. 298). 



In Dumfriesshire, 2 Muscids and 2 Syrphids were recorded (Scott-Elliot, ' Flora 

 of Dumfriesshire,' p. 65). 



1046. E, collinum C. C. Gmel. — In this species again, according to Kerner, 

 automatic self-pollination may take place on the first day of anthesis, the stamens 

 being long enough for the anthers to touch the stigmas. 



Fig. 152. Epilobiuni alsiiii/olium^ ^'///. (after Herm. Miil'.er). A. Young flower, shortly after 



opening : the upper parts of the sepals and petals have been removed. B. Older flower, after removal 



of two sepals and petals (x 7). a' and a^, anthers of long and short stamens ; gr^ style; ;;, nectar : 

 tTz/, ovary; ^, petal ; /o, pollen ; .r, sepal ; ji/, nectar-cover ; j/, stigma. 



Visitors. — Herm. Miiller saw 2 short-tongued bees in the Alps (' Alpenblumen,' 

 p. 213). 



1047. E. roseum Retz. — Schulz says that automatic self-pollination is usually 

 inevitable in the homogamous flowers of this species, the anthers of the long stamens 

 reaching the level of the non-divergent stigmas and coming into contact with them. 



Visitors. — MacLeod saw the butterfly Pieris napi L. in Flanders (Bot. Jaarb. 

 Dodonaea, Ghent, vi, 1894, p. 296). 



