THE CONTINENTAL FLORA OF SOUTH SWEDEN 245 



ticed that the temperature of the ground is generally rather different tVom 

 that of the air (see, for instance, Kraus 191 1). 



3. The fact that the distribution-limit coincides with a border-line of a climato- 

 logical factor does not prove the existence of a causal connection between the 

 distribution-limit of a plant and that factor. Only in a few cases has it been 

 established that the species suffers from the climate by crossing the limit. 

 Grisebach says (1872, p. 98): the eastern limits of Atlantic species are »nicht 

 bloss durch die Lage der Vegctationslinien sondern zugleich durch Unter- 

 suchungen iiber die Lebensbedingungen der einzelnen Arten zu erledigen». 



4. The use of a daily niea?i temperature may in a high degree point in a wrong 

 direction. A plant can obtain a sufficient quantity of heat at a lower mean 

 temperature, if the daily temperature amplitude is great, than if it is small 

 (compare De Candolle 1855, I, p. 202; Brockman-Jerosch, 19 13). 



5. It has been established that plants have different demands as to heat for the 

 commencement of different manifestations of life (e. g. Schimper 1908). Hence 

 the somehow or other calculated quantity of heat offered to the plant during 

 a longish time before the beginning of certain manifestations of life cannot 

 express the demands for heat of the plant in this respect. Besides, it has 

 been found that plants are in their periodical manifestations of life to a certain 

 degree tied by hereditary dispositions. These allow a plant to suit itself to 

 external conditions with regard to the time for the commencement of a certain 

 manifestation of life only within a certain boundary. (See especially the good 

 critique of Bos 1907). 



6. It is preposterous to suppose that the distribution-limit of a species should be 

 determined by only 07ie climatic factor. A cooperation of all factors must 

 take place. In many cases it seems more possible and suitable to follow 

 Koeppen (1900) in letting certain plant limits or flora limits characterize the 

 climate than to try to distinguish in the climate the various factors that may 

 determine the distribution of plants (compare Grisebach I. c, p. 76 and Brock- 

 mann-Jerosch 1. c). 



7. It must be taken into consideration that in most cases a species has no 

 opportunity of freely reacting against external conditions in its occurrence in nature. 

 Species may be more or less connected with each other; and generally there 

 is rivalry between the species in the struggle for space (compare Grisebach 

 1. c, p. 74). 



8. There may exist purel}- edaphic reasons for a distribution-limit: the more a 

 species is differentiated in its choice of habitat, the more important may tiiis 

 circumstance be. 



The causes of the distribution-limits of species being so various and also able 

 to work into one another in different ways, it would seem justified to question 



