THE CONTINENTAL FLORA OF SOUTH SWEDEN 359 



days a comparatively abundant distribution in these districts. The species begin 

 to appear in the layers of marine sediments that were formed approximately 

 during the maximum extension of the Ancylus Lake. In the bays and lagoons 

 of brackish water which were formctl in tlie j)roccss of land-upheaval after 

 that time they liavc also been found, but at lower levels they begin to disappear. 

 Thus, the species during a certain period of time moved outwards from their 

 first localities as the coast-line was pushed outwards in consequence of the 

 upheaval of the land. 



In the regions wiiere the land has been elevated on the mainland of South 

 Sweden there are a number of occurrences of halophilous species, which for that 

 matter also grow on our sea-shores (as regards Uppland, see Sernander 1905). A 

 description of an Uppland locality (Skensta salt spring at Torstuna) with a 

 number of such species has been given by Sernander (1920 a, pp. 330 ff.). An 

 examination of the ground water at this spot showed that the salinity was quite 

 insignificant. Sernander considers that the seashore plants here are survivals 

 from a time when the locality lay on the seashore. As a rule, the species 

 have followed the seashore as it pushed its way outwards, but in some few 

 localities they have for some reason held their own. 



In Finland a number of species have their main distribution on the southern 

 coasts, but at the same time a number of isolated habitats in or near lakes. 

 Lindberg (e. g., Svensk Bot. Tidskr., 1915, p. 467) has tried to explain these 

 inland occurrences as relics from the lie of the coast of the Ancylus Lake. 



In this connection should be mentioned the theory put forward by Selander 

 (1914) as to another connection between the post-glacial land-upheavel of the 

 south-east of Sweden and the distribution of certain species. In the Stockholm 

 skerries the species which Selander (in part erroneously) has cited as southern 

 or south-easterly xerothermic species would not have had the capacity to spread 

 to such islands, islets and skerries as fall below a certain altitude above sea-level, 

 namely the altitude corresponding to the position of the shore-line at the close 

 of the dry and warm Sub-boreal period. The theory may have a certain suggestive 

 value, but the material of investigation on which Selander's account rests is quite 

 inadequate (see, for Romell's criticism, Romell 191 5). 



Something must be very briefly said about one or two species which are rare 

 in the south of Sweden. Vicia tenuifolia, a characteristic species of the scrub- 

 steppe (»steppae fruticosae», Korshinsky 1898) in South Russia, probably, in South 

 Sweden is most abundant on Oland, where it occurs in scrubs, most frequently, 

 however, as a weed-plant, especially in barley-fields. Viola elatior is found fairly 

 frequently in the southern and central parts of Oland, where it has its favourite 



