19t 



stigmas, as yet not fully developed, are able to retain the pollen- 

 grains; self-pollination ^vill even at that time be able to take 

 place. 



In other cases the flowers were more highly protandrous. 

 In cultivated specimens (Hortus Hafniensis) the terminal flower 

 was so protandrous that self-pollination was impossible; here 

 the corolla had attained a diameter of as much as 15 — 20mm.^ 

 From Spitzbergen Ekstam reports "homogamy, with marked 

 inclination to protandry." 



Fig. 16. Saxifraga groenlandica. 



A — D, Pistillate flower from Iceland (Dyrefjord; June 10, 1895; C. H. Ostenfeld) ; A, flower 

 in longitudinal section (with the accompanying scale) : B, part of an inflorescence (^/i) ; 



C, D, petal and stamens of Fig. A. 



E — K, forma cryptopttala from West Greenland (Egedesminde ; leg. N. Hartz). E, A flower 



(about '^li) ; the petals are considerably smaller than the sepals. P, A sepal with antisepa- 



lous stamens. 0, H, Petals and an antipetalous stamen, J, Style. Ä, Pollen-grains appear 



to be quite normal. (E. W.) 



Protogyny occurs also, at any rate in Spitzbergen (War- 

 ming, 1886). According to Andersson and Hesselman the f. uni- 

 flora is "always protogynous" in Spitzbergen. In Nova Zembla 

 the flowers were "in numerous cases protogynous-homogamous" 

 (Ekstam, 1897). 



Homogamy. Lindman reports that in Norway (Dovre) the 

 flower is homogamous; in Nova Zembla Ekstam found some 

 flowers which were "almost homogamous." With regard to 



' GüXTHART found cultivated specimens in Switzerland so highly pro- 

 tandrous that "between the staminate and the pistillate stage a short 

 neutral, 1. e. sexless intermediate condition occurred." 



