— 145 — 



1. Leaf structure of the investigated plants. 



Concerning the anatomy of the leaves the following remarks 

 taken from the literature may be quoted: 



The leaf of Empetrum is dealt with by several investigators 

 (for instance Warming 1888, Mentz 1909). The edges of the 

 leaf bend so that the edges almost meet, and a cavity is formed. 

 The groove is closed with long hairs. The outer walls of the epi- 

 dermal cells are thick and strongly cuticularized; the inner walls 

 are mucilaginous. The stomata occur only on the epidermis in the 

 cavity. 



In Vaccinium Vitts idaea (Petersen 1908) the leaf is cori- 

 aceous and not deciduous. The cuticle is thick, the outer and lateral 

 walls of the epidermal cells are thickened and partly cuticularized. 

 The stomata occur only on the lower surface. 



The leaf of Vaccinium uliginosum (Petersen 1908) is deciduous 

 and thin with a waxy covering. The cuticle is thin and the wails of 

 the epidermal cells are only slightly thickened and not cuticularized. 

 The stomata are confined to the lower surface. 



In Betula (B o u b i e r 1896) the leaf is also deciduous and 

 thin ; the cuticle and the walls of the epidermal cells are thin ; the 

 inner walls possess a mucilaginous layer. The greater side nerves 

 are surrounded by water conducting cells. The stomata occur 

 only on the lower surface. 



2. Methodic. 



Of methods to determine the rate of transpiration the weighing 

 method is to be considered as the most accurate. Yet the method 

 can be varied in different ways. Frequently the plant, whose 

 transpiration is to be determined, is planted in a flowerpot in soil 

 and the rate of transpiration is determined through the loss of 

 weight of the flowerpot. In this form the method was not suit- 

 able for my experiments. It is always dangerous to put plants 

 living under conditions so special as is the case with high-moor 

 plants, in a flowerpot, where the surrounding conditions may 

 easily be quite different from the natural conditions. 



I have therefore preferred to use another method. The branches 

 whose transpiration should be estimated were cut off and put 

 through the hole of a rubber stopper in a glass bottle (diam 2 cm, 

 length 10 cm) filled with water. The weight of the bottle was 

 estimated with a hand balance ; it was about 50 g. The danger 

 that the transpiration should be altered during the experiment for 



