THE CONTINENTAL FLORA OF SOUTH SWEDEN 327 



present distribution is thus, probably not its maximum distribution under the 

 present geographical conditions. 



Of the species confined to souiJiernmost Sweden (Skane, in certain cases also 

 Blekinge, Oland and Gotland) those belonging to psammophilous vegetation should 

 above all be noted: Helichrysum arenarium (Plate 3), Holosteum umbellatum (on 

 Oland also on the rock-pavements of the Alvar), Koeleria glauca, Medicago mi- 

 nima (the occurrence on Oland at Borgholm, however, being on »pavement»), and 

 Poa bulbosa (on Oland and Gotland also on the »pavement»). To these may 

 properly be added the Sarmatian species. Astragalus arenarius (Plate ig), Dian- 

 thus arenarius, and Gypsophila fastigiata (Plate 21). 



The scanty distribution of these species in South Sweden is primarily, of course, 

 to be ascribed to the distribution of large sandy districts (or limestone pave- 

 ments). However, the first-named species are in their whole P2uropean distribu- 

 tion more southerly than the steppe species that are also widely distributed in 

 Central Sweden, and should be assumed to be more heat-loving than the latter. 

 The Sarmatian species, on the other hand, reach far north in Eastern Europe, 

 which is already testified by the peculiar occurrence of Gypsophila on sandy 

 fields in Dalarna. 



Isolated Occurrences of the species in South Scandinavia. 



With the treatment of these last-named species we have come to the feature 

 that is so conspicuous in the distribution of many steppe species, i. e. the iso- 

 lated position of the luesterly occurrences. 



A comparison, with regard to the distribution in Central Europe, between 

 steppe species and continental species with another mode of occurrence, for in- 

 stance forest species, shows that this feature is particularly characteristic of the 

 distribution of the steppe species. The others have generall}' a more or less 

 continuous distribution even to the extreme limit (see, for instance, Plate 15 and 

 16). The reason is naturally in the first place that, in accordance with the strongly 

 specialized ecological demands of the steppe species localities suitable for these 

 species get more scarce the farther we get from the steppes. 



It is curious, however, that the species have been able to reach all these 

 strongh' isolated localities, and that this capacity is common to a great number 

 of species with a rather different means of dispersal. Another remarkable cir- 

 cumstance is the fact that these steppe species with isolated outposts in Central 

 or Western PLurope consist almost soleh' of perennial species which are also 

 often able to propagate vegetatively, while the steppe species include compara- 

 tively numerous annual hapaxanthous plants. 



These remarkable circumstances agree well with the current opinion about the 



