Rosaceæ. 59 



theory as against Haberlandt's, as regards the signification 

 of the so-called Trichtercellen" in the foliage-leaves; a theory 

 which should more properly be called the \\ a rming-Rywosch's 

 theory, as it corresponds with the views of the former investi- 

 gator 1 ) (see also Alchimilla vulgaris p. 108). 



The polygonal or slightly branching cells of the spongy 

 parenchyma are arranged in about four layers and include inter- 

 cellular spaces which are usually smaller than the adjacent cells 

 (Fig. 22, A,G). The bundles are not accompanied by stereom; 

 the smaller branches are surrounded with a sheath while 

 the larger ones are in addition accompanied by the usual 

 vein-parenchyma, which compensates for the assimilatory 

 tissue, up to both the upper and lower epidermis. In fresh 

 leaves from the Botanic Garden in Copenhagen tannic acid 

 occurred especially in the tissue of the veins and in both the 

 upper and lower epidermis. 



The insignificant flowers are collected in 6 — 12 — flow- 

 ered heads which consist of several small cymes. The 

 petals are pale yellow and smaller than the inner sepals, 

 The flowers are 7 — 8 mm. in diameter (Warming's notes). 

 The flowering period begins in July; Norman records July 9 

 for Arctic Norway, and at Scoresby Sound Hartz observed 

 the first flower on July 7. It no doubt usually sets ripe fruit. 

 Miller is of opinion that it has lost the power of spontaneous 

 self-fertilization (Alpenbl., p. 222); Warming, on the other 

 hand, writes in 1886, that this cannot apply to the Green- 

 land specimens because he found the anthers placed so close 

 to the stigmas that self-pollination must no doubt easily 

 take place (Fig. 23, B), and in his notes he remarks that the 

 species is homogamous and that he has seen the anthers in 

 contact with the stigmas. Similar observations are recorded 

 by Lindman (1887) who investigated the species on Dovre 



1 H alofyt- Studier. Kgl. Danske Vidensk. Selsk. Skr., 6. Række, 

 naturvidensk. og math. Afd. VIII, 4, p. 245. 18!»7. 



