THE CONTINENTAL ELORA OF SOUTH SWEDEN 271 



The distribution of a species may be determined by historical reasons even so 

 far that its isolated occurrences within a district are remains of a formerly existing, 

 wider and more even distribution. The species has been able to hold on only 

 in rare, specially ciualified localities, where the habitat conditions have remained 

 favourable to it, while they have been changed in other parts of the district, 

 thereby causing the species to disappear. 



The isolated occurrences of a species may, however, have yet another reason. 

 The species may never have had any wider distribution in the district than the 

 isolated occurrences of the present day. These may be the first ones in an 

 immigration now in progress; or they may correspond, nowadays as in times past, 

 to the only localities suitable to the species. They are in any case the result 

 of a dispersal from afar which uiay have taken place quite recently. 



In most cases it would seem very difficult to determine the nature of the iso- 

 lated occurrences. The capacity of the species for dispersal from afar should of 

 course be examined, but it is often difficult to ascertain with certainty the pre- 

 sence or absence of such capacity, for in this case a rare mere chance may be 

 supposed to be sufficient. In the history of the flora, however, we have another 

 starting-point. If, on another basis, it can be proved that the flora as a whole 

 must have been subjected to corresponding general changes, it is of course more 

 probable that the isolated occurrences are real relics. 



It is characteristic of many species which are markedly continental in their 

 distribution that their distribution-area gets broken up into small areas that 

 become more and more isolated and diminished towards the western limit of the 

 species. In South Scandinavia there are several examples of such isolated occur- 

 rences that are the furthest outposts towards the west or north-west of the species. 

 These occurrences have long been regarded as relics from an ancient wider and 

 more equal distribution. What reasons are there for such an opinion? 



It must be regarded as proved that, during the Quarternary era, Middle Europe 

 passed through one or more periods with a climate of a more continental cha- 

 racter than the present one. That this is the case with South Scandinavia is 

 beyond doubt. With regard to the number and date of those periods, as well as 

 the degree of continentality of the climate, however, there are different opinions. 

 It does not seem necessary to give an account of these dififerences here. In the 

 sequel I shall accept the theory which seems to me the one at present most 

 generally accepted by quarternary geologists, namely that expounded by Sernander 

 and von Post. 



According to this theory, Sweden has had two such climatic periods: the Bo- 

 real Period and the Sub-boreal Period. The latter is placed at about 3 000 or 

 2500 — 500 B. C. See further Sernander, e. g. 1908, 1910, 1916; von Post 1920 

 (in this work the continentality of the Boreal Period is restricted to hold good 



