KUNGL. SV. VET. AKADEMIENS HANDLINGAR. BAND 50. N:0 3. 



83 



The rainfall is rather equally distributed. There is nearly always the same, 

 regular supply of moisture in the soil, but there is a maximum in summer and 

 autumn, and it is of importance that the spring gets less rainfall than do other 

 seasons. At the same time the force of the wind is greatest in the spring, certainly 

 an unfavourable combination, making a xerofytic structure necessary. 



Thus everything seems to co-operate in making the plants protect themselves 

 against drought: low temperature of air and soil throughout the summer, strong winds 

 and very few calm drys, and the distribution of the rainfall. There are no observations 



E W 



Fig. 10. GratcBgus sp., sliowing shape of head above tlie sheltering stone-wall. 



l'liutu by author. 



to show, how large the annual amount of snow is. But I liave some experience of 

 the Falkland winter. From what was said above on the temperature, we easily under- 

 stand, that one cannot count upon a protective snow-cover; the snow never finds 

 time to accumulate, there is perhaps a week, or even a fortnight, with snow and 

 frost, but generally the winds prevent the equal distribution of the snow, any day it 

 may be transformed into water, and there follows a period of rain (see 14, p. 7). 

 The plant formation to be expected in this kind of climate is the cold-temperate, 

 oceanic heath. Both Birger and I have used this word. For South Georgia {36) I 



