245 



which has been exhaustively discussed by Lazniewski (1. c.l in 

 regard to ro?ette-plants and which we shall find again in the 

 very decidedly rosetle-formiug species (S. Cotyledon^ Aizoon, etc.). 

 The 2 — 3 palisade-layers are differenced, but not very distinctly, 

 from the cells of the spongy parenchyma, which are decidedly 

 stellate (i. e. branched), long, and loosely arranged (Fig. 1. E.) 



The veins are not accompanied by mechanical tissue, but 

 are surrounded by a (usually one- layered! bundle-sheath of 

 elongated cells devoid of chlorophyll (Fig. I C\. 



All the veins at the apices of the leaves terminate in a 

 hydathode. The epithema is interwoven, and enveloped by 

 the spirally thickened tracheids of afferent veins and is covered 

 with a small-celled epidermis with water-pores. Lime-secretions 

 were totally absent from the Arctic specimens examined by me. 

 The surface of the epithema is convex and situated directly at 

 the edge of the leaf-margin; a hydathode-cavity is absent 

 (compare, <S\ oppositifolia, S. Aizoon, etc., the hydathodes of 

 which open upon the upper side of the leaf-margin and have 

 a cavity). Fig. 1 D. 



The leaves of the bulbils are morphologically somewhat 

 different; a gradual transition maybe traced from entirely blade- 

 less scale-leaves to bulb-scales with rudimentary leaf-blades 

 which have entire margins, and ultimately to bulb-scales with 

 a small three-lobed blade (the cells of which are devoid of 

 starch-grains although the latter occur in quantities in the leaf- 

 base). 



The bulb-scales are without hydathodes. The cell-walls 

 of the lower epidermis are almost straight; no stomata were 

 found by me although tliey were found by Th. Holm (I. c, p. 47). 

 The hairs are similar to those upon the lower surface of the 

 foliage-leaves. 



The cell-w-alls of the upper epidermis are straight, and 

 there are no stomata. The hairs are similar to those upon the 

 upper surface of the foliage-leaves. 



