626 



development. Usually thc different forms have not a wide geo- 

 graphical distribution. I am therefore of opinion thai the studv of 

 the Hieracia of a single district in its relation lo the neighbouring 

 Iloral distriels oughl lo be particularly useful as a contribution to- 

 wards the solving of various plant -geographical questions. This 

 has become still more clear to me by studying the Hieracium-Rora 

 of Scandinavia. I think thai il will more particularly be of great 

 help lo us in determining the ways by which the flora of a land 

 immigrates, and also Ihe relative time for its immigration. The 

 composition of the Hiercicium-flora of the Færoes confirms the opin- 

 ion expressed by C. H. Ostenfeld regarding the origin of the rest 

 of the phanerogamous flora. Its composition proves to be rather 

 peculiar. As I said before, 21 species and 2 varielies are known. 

 These belong to the following groups of the genus: - 



Vulgata, sub-group Subcæsia 2 



— Vulgata genuina 4 



Rigida 3 



Cerinthoidea 12 (of which two are varieties) 



Alpestria 2 



The forms in question stand in the following relation to the 

 flora of the neighbouring countries: — Of the two forms of Sub- 

 cæsia, H. sarcophylloides is closely allied to a formå H. sarcoj)luilluni 

 Stenstr. which occurs in Norway and the south of Sweden, while 

 //. ardisodon has, as far as I know, no closely related forms in 

 Scandinavia. H. sarcophyllum has otherwise a distribution in Scan- 

 dinavia which suggests that it belongs to a western floral element. 

 Of the forms re ferred to Vulgata genuina, H. cordifrons is extremely 

 closely allied to the Icelandic H. holopleurum Dahlst., and H. holo- 

 pleuroides Dahlst., as also to H. ciliolatum. H. constrictiforme is un- 

 questionably very closely connected with the Scandinavian forms: — 

 H. constringens Norrl. and H. subalpestre Norrl. — which have a north- 

 ern distribution — but it comes perhaps nearer to //. protalatuin 

 K. Joh. from the south of Sweden. All these belong beyond doubt 

 lo the same group of forms, comprising both lowland and sub- 

 alpine forms. H. epileucuin occupies a more isolated position, and 

 as yet I know nothing regarding its relation to Ihe forms from 

 the surrounding countries. 



Of the Færoese Rigida, II. Simmonsianum belongs to a group 

 of forms which is ivpresented in Scandinavia by //. lapponicum Fr., 



