THE CONTINENTAL ELORA OF SOUTH SWEDEN 361 



Thanks to human intervention, these species are in the act of spreading over 

 Scandinavia and have evidently not yet attained their full distribution there. 

 Probably the South Swedish highlands form a substantial obstacle for them. 



The following pages will show that a similar state of things can be established 

 with regard to several other continental species. (Cf. pp. 335 ff., 365.) 



Summary. 



1. The South Swedish continental species which belong to the plant com- 

 munities of thin, xerophilous or mesophilous foliferous forests of the east of 

 Europe have in the south of Sweden a mode of occurrence which is in complete 

 agreement with that in the east of Europe. — They appear on oak and grove 

 slopes, thicket slopes, rocky escarpments, and in forest-communities rich in Calama- 

 gro-stis arundinacea. 



2. Many species are widely distributed in South Sweden and are found both 

 in the eastern and the western parts, a fact which stands in agreement with an 

 extensive general distribution in Europe. The majority of these species, however, 

 have their centre of gravity in the south of Sweden in the eastern part of 

 the country, especially in the skerries and their immediate hinterlands on the 

 mainland. In Smaland some species have their distribution only or chiefly in 

 the north-eastern part. The distribution of the species would seem to be de- 

 termined by the distribiitioji of broken country and by the precipitation and soil 

 conditions. 



3. A number of species are restricted to the south-east of Sweden. Amongst 

 these those are of special interest which exclusively, or all but exclusively, occur 

 in the skerries and in certain parts of the regions lying immediately behind the 

 skerries. As to these species the following working-hypothesis may be proposed. 

 The distribution-limits of these species, which lie at different distances from the 

 coast and at different heights above sea-level for different species, may be inter- 

 preted as remains of the first distribution of the species along the coast. Erom 

 these oldest occurrences the species have spread outivards, to the new suitable 

 localities which have come into existence with the continued upheaval of the land. 

 On the other hand they have not been able, or have been able only to a very 

 slight extent, to spread inland over the old mainland. The causes for this are 

 not known, except, possibly, for Cynanchum vincctoxicum. 



In Sodermanland, and to a large extent also in Uppland, the distribution of 

 Cynanchum seems to be determined by the supply of suitable localities (rock)' 

 escarpments). In Ostergotland and Smaland this is not the case. Possibly the 



