TAXONOMIC DIFFERENTIATION 89 



The Poa laxa group or Poo sect. Oreinos forms an excellent example. This 

 group has its centre in central Europe and has certainly originated there. 

 It is totally absent from the Arctic. To be sure, Poa laxa has repeatedly been 

 reported from Greenland and Arctic America, but in so far as I have seen the 

 voucher specimens they have been misnamed. Hulten (1942) claims that closely 

 related plants occur in Alaska and Kamtchatka but as far as I understand 

 they have nothing to do with sect. Oreinos. Now, some 25 years ago I (Nann- 

 feldt, 1935) was able to show that the populations of Scandinavia. Iceland and 

 Scotland are very homogeneous both intra and inter se and that they show 

 sharp though small differences from the populations of the Alps and the other 

 southern mountains. Some years later, Nygren (1950, 1955) showed that the 

 northern taxon {Poaflexiiosa or P. laxa subsp. Jiexuosa) is hexaploid (2/z = 42) 

 in contrast to the true southern P. laxa, which is tetraploid (2n = 28) and 

 perhaps also diploid. A very isolated population occurs in eastern North 

 America. I found it to represent a distinct taxon {Poa fernaldiana or P. laxa 

 subsp. fernaldiana) and Nygren (1955) to be hexaploid. All these facts seem to 

 indicate that the northern populations have been isolated from the southern 

 for a very long time, and this conclusion is strengthened by Poa jemtlanclica 

 (Nannfeldt, 1937; Nygren, 1950). This viviparous taxon has certainly arisen 

 as a hybrid between P. flexuosa and P. alpina and is so uniform in all mor- 

 phological characteristics that it must be of monophyletic origin. It propagates 

 exclusively by vivipary. It occurs in Scandinavia and Scotland but is unknown 

 from Iceland. Its Scandinavian (Figs. 1 and 2) and Scottish part-areas are 

 smaller than those of P. flexuosa but fall completely within them. Its present 

 distribution affords thus an additional proof of a connection between 

 Scotland and the Scandes and is strongly indicative both of its own high 

 age and of the high age of P. flexuosa in the North Atlantic area. 



Trisetwn spicatuni is a grass with an unusually large, almost world-wide 

 Arctic-montane distribution. This species is certainly very old. Hulten 

 (1959a) has recently treated its racial differentiation, distinguishing no less 

 than 14 subspecies. For my purpose it is sufficient to mention that the type 

 subspecies has an Arctic -Subarctic distribution reaching farther south in 

 the central Asiatic mountains and in the western American mountains, whereas 

 the Pyrenees, the Alps and the Caucasus are inhabited by a separate taxon, 

 subsp. ovatipaniculatum. It may further be mentioned that Iceland and south 

 Greenland are reached by a northeast American taxon, subsp. pilosiglume. 

 The cytological features of this complex are still very imperfectly known. 

 The subsp. ovatipaniculatum is unknown cytologically. The type subspecies is 

 tetraploid, and a hexaploid is known from Greenland. 



Several additional examples could be given even if those just discussed are 

 the most instructive. It should not be concealed that there are species, in 

 which, at least so far, no significant morphological differences have been 

 found between the populations of Scandinavia and the Alps. Such is the case 



