332 Eue Warming. 



land. In the Greenland specimens I found flowers formed 

 the year previous to that in which they were to open. In 

 specimens gathered on May 10th, 1887, by Lieut. Ryder, 

 near Upernivik, young flowers, about 1 / 2 mm in diameter, 

 were found far down among the leaves. (See p. 282, Fig. 24). 



Greenland. Trioecism: 5, o and $; as ? sets fruit, 

 we have here evidence of insect-pollination. I have not seen 

 g flowers, but Lange mentions them in "Conspectus Floræ 

 Groenlandiæ" on the authority of Vahl. 



S are > $ (Fig. 44, A, B, C), which is especially seen 

 in the length of the petals and in the diameter of the corolla. 

 In <$ the petals are 8 — 10 mm, in $ 5 1 / 2 — Q 1 /^ of which the 

 claws are about 4 mm long and the limb scarcely 2 mm (see 

 Fig. D, E, F). The calyx tube is generally larger in <$ (5 — 

 7 mm) than in $ (4 — 5 mm). The corolla-tube in S is 6— 

 6^2 mm, in $ 4 — h x J 2 mm, the diameter of the corolla in 3 

 is 7 — 11 mm, in ? 6 — 8 mm (see Fig. A, B, C). <? has always 

 a small pistil with styles l 1 / 2 — 2 mm in length, which I have 

 not seen grow and separate (Fig. Z)), as they do in the Alps 

 according to H. Müller. There are shrunken ovules, the inte- 

 guments of which are however formed. The cal.-st. are the 

 first to develop, then the cor.-st. The styles have at the top 

 some very short stigmatic papillæ, which never grow longer 

 than those shown in Fig. D. 



$ flowers have always rudiments of stamens, but in vari- 

 ous degrees of development; in some cases they are as much 

 as 2 mm long, and have distinctly marked-off anthers (Fig. 

 E, std), which are however shrunken, transparent and devoid 

 of pollen; in other cases they have still shorter stamens, and 

 the anthers are hardly indicated (Fig. F). Correspondingly 

 with this there appears to be a different degree of develop- 

 ment of the size of the flowers, the size of the petals, and the 

 degree of development of the stigmatic papillae, so that the 



