291 



in the species here investigated would be different if they were 

 cultivated in Alpine regions. When nevertheless they agree 

 within each section in Arctic regions, this proves that although 

 the structure of the mesophyll varies according to climate, yet 

 it varies correspondingly in the species belonging 

 to the same section. Leist found that S. Aizoon and S. 

 Cotyledon (from the Alps) had exactly the same structural fea- 

 tures; I, also, in my Arctic specimens, found that these two 

 species, belonging to the same section, had quite similar leaf- 

 structure, — but it must be admitted it was a structure which 

 was entirely different from that found by Leist in his Alpine 

 specimens — a good example, therefore, of corresponding va- 

 riation in closely related species. 



If we wish to define how the Arctic species of Saxifraga 

 are adapted in their leaf-anatomy to their natural surroundings, 

 emphasis must be laid upon the fact that the same characteri- 

 stic tests cannot be applied to all the species investigated, 

 taken as a whole. They cannot as a matter of fact be termed 

 either xerophytic or hygrophytic, these words having on the 

 whole only a relative value , as they do not state anything 

 about the plant's absolute relation to water -absorption and 

 transpiration. If, on the other hand, we can show that the 

 Arctic Saxifragas are either more, or else less, xerophytic 

 than their Alpine colleagues, that would be a positive result of 

 great significance, as it would bring about a comparison from 

 which, to a certain extent, we might be justified in deciding 

 how Alpine plants must adapt themselves in order to be able 

 to live under the external conditions of Arctic regions, and vice 

 versa. My endeavour has been to give as accurate a descrip- 

 tion of Arctic species as possible; future investigations must 

 prove wherein such species differ from Alpine species. Honnier 

 has already attempled something of this kind, but more ex- 

 haustive investigations are highly desirable. The variations 

 occasioned by external conditions are, in the greater number 



