96 J. A. NANNFELDT 



In the northern population the reproductive features are different. Besides 

 the aposporous embryo-sacs there may now and then — though rarely — be 

 formed sexual embryo-sacs, and even the egg-cells of aposporous embryo-sacs 

 may occasionally become fertilized. 



One of the recognizable northern taxa, subsp. caespitcms, is most outstanding 

 morphologically, i.a. by the always empty anthers. It has very few localities in 

 Scandinavia, and grows there together with other very rare plants, but it has a 

 wide West Arctic distribution and seems to be common in part of its area. 

 Both normal and aposporous embryo-sacs are formed, and the latter are 

 able to form both embryo and endosperm without fertilization. Otherwise 

 the apomictic types, are as a rule, pseudogamous. The subsp. caespitans is 

 thus able to breed true and — in spite of its empty anthers — to propagate 

 even in areas where there grows no other type that can supply serviceable 

 pollen. This seems to be the situation in, for example, EUesmereland. In, 

 for example, Scandinavia the situation is different, for there both normal and 

 aposporous embryo-sacs may become fertilized. Nygren has in one of the 

 Swedish localities found specimens similar to, but not identical with, caespitans 

 and having a somatic chromosome number of 86-88. Such specimens have 

 probably arisen from caespitans by fertilization of aposporous embryo-sacs by 

 alien pollen. 



It seems clear that the populations of the south and the north of the Scandes 

 are both old and have developed independently for a long time, and that the 

 wide West Arctic distribution of caespitans proves a high age of that taxon 

 just as the occurrence of depauperata in Iceland proves a high age of that 

 taxon. 



These examples are selected from plants intensely studied, in the cases of 

 Papaver and Poa both taxonomically and cytologically, in the case of Papaver 

 also genetically. 



There are a number of species awaiting similar studies, and I am sure that 

 most of our mountain plants will repay generously the labour devoted to 

 them. It is especially important that the work is not too much concentrated 

 upon the rarest species with widely isolated, very small populations, for in 

 those we run the risk that the special features of "small populations" may 

 overshadow more general trends. Such described taxa as Artemisia norvegica 

 var. scotica, SteUaria crassipes var. dovrensis and Oxytropis deflexa subsp. 

 norvegica exemplify certainly the evolution within such small populations. 

 In several cases, the distinguishing marks of such isolated populations have 

 been found to break down, when the variability within the main area has 

 been studied more in detail. I shall take an example from another part of 

 Scandinavia. When Orchis Spitzelii was found on the Swedish island of 

 Gotland this population was described as a var. gotlandica, but later Bengt 

 Pettersson (1958, pp. 77-82) could show that closely corresponding individuals 

 occur also in the south of Europe. Also in other species such described local 



