360 RIKARD STERNER 



haunt in the above-mentioned »Alvar forests, a mode of occurrence which would 

 seem fully to correspond to that in Eastern Europe (»in fruticetis stepposis, in 

 decliviis apricis (imprimis calcareis) stepposis vel silvosis», Korshinsky 1898, p. 

 53). Lactuca quercina and Rosa yundzillii Bess, appear in steep places on the 

 limestone cliffs in Lilla Karlso on the west coast of Gotland, to a very great 

 extent isolated from their area of distribution in Central Europe (Cf. Sernander 

 1894, p. 84). 



Here also may be mentioned the Siberian species Poteritilla fruticosa. In 

 Siberia and North America, where it is widely distributed, it seems to belong 

 to different types of vegetation. Chiefly, however, it seems to occur on shores 

 of rivers forming scrubs between the wood and the more hydrophilous vegeta- 

 tion, or it appears on rocky ground in crevices and on ledges often forming a 

 dwarf-shrub layer in thin forests. (See e. g. Printz 1921, p. 281; Marmajanova 

 1882, p. 77; Harshberger 192 1). 



In South Sweden it is common on southern Oland and has a few localities on 

 north-eastern Gotland. On Oland it occurs chiefly on the Alvar forming peculiar 

 scrub-associations, thin or closed according to the thickness of the soil on the 

 limestone rock. A couple of analyses in table 6, Appendix II may in some 

 degree show the composition of this vegetation. (Cf. Johansson 1908.) 



Finally, something must be said about three species which hold a special posi- 

 tion owing to the fact that they are greatly spread by the hand of man and 

 perhaps do not belong to the spontaneous flora of South Sweden. 



Heradeum sibirimm. This is distributed over practically the whole of Scan- 

 dinavia. In the south-west of Sweden, however, it is rare, and in the west of 

 Norway it is altogether lacking; in both these regions, on the other hand, the 

 south-west European H. sphondylium L. (= branca ursina All.) is distributed. In 

 Scandinavia H. sibiricum is most widely spread in the eastern part of Central 

 Sweden. 



Luzula pallesceyis (Wg.) Bess, and Trifolium spadicctim have a similar distribu- 

 tion. (As regards Luzula this statement is made on the strength of courteous 

 information from Docent Samuelsson, to whose memoir on the subject, which 

 will appear shortly, I must refer). Luzula, however, is less widely distributed in 

 the south-west of Sweden than Heracleum, and Trifolium spadiceum is lacking 

 altogether in a large south-westerly part of the district. 



These species, coming probably straight from the east, have immigrated to 

 Central Sweden and there attained a wide distribution. But they are rare or 

 are lacking in the south-west of Sweden, and seem thus not to have reached 

 South Sweden by a south-westerly route. Yet the species are widely distributed 

 far to the west in North Germany and also have occurrences in Southern Den- 

 mark. 



