22 



YOLANDE HESLOP-HARRISON 



evidence from which we can deduce the period when this geographical isolation was last 

 complete, but there is little reason to believe that it could have been so during the post- 

 glacial period. Differentiation of the two ecospecies may have actually been completed by 

 late Tertiary, or it may perhaps have taken place during an interglacial. In any case, it is 

 likely that during one or more of the Pleistocene glacial maxima the areas of both were 

 severely compressed in southern or eastern refugia, possibly with some commingling 

 of the populations. The subsequent immigration in the post-glacial into the formerly 

 glaciated area would then see a " sorting out " of the ecospecies, N. pumila penetrating 

 further north and becoming extinct in the south except where local microclimates favoured 

 its persistence. The much- studied N. pumila populations of the Vosges and other central 

 European mountains are no doubt relics of this nature. A similar history may be postu- 

 lated for the remarkable southern colonies in the British Isles, those in Merioneth and 

 Shropshire, if indeed the latter has not resulted from a relatively recent human introduction. 



The status of N. intermedia 



On the basis of the morphological evidence given above, N. intermedia might appear 

 to be as homogeneous a unit in Europe as N. lutea or N. pumila. Two interpretations of 

 this situation are possible, namely, (a) that it forms a separate third pure-breeding species, 

 or, (h) that the plants placed under it are all hybrids of the same parentage presumably 

 having arisen in several different localities independently. The first of these possibilities 

 would seem to be ruled out decisively by the impaired fertility of the plants placed under 

 N. intermedia. No populations are known which would fit the diagnosis of N. intermedia 

 and which possess at the same time a fertility comparable with that in N. lutea or N. pumila. 



This leaves the second interpretation, long accepted by systematists, that N. inter- 

 media is the assemblage of hybrids from the cross N. lutea X N. pumila. Caspary's 

 demonstration that the artificial Fj hybrid of this parentage resembles closely plants 

 placed under N. intermedia would seem to put this interpretation beyond doubt. 



This ascription of a hybrid origin to N. intermedia does not, however, solve all the 

 problems connected with it. The remarkable apparent homogeneity of N. intermedia, 

 considered simply as a taxonomic unit, requires some explanation. The data of Tables 

 1, 3, 4 and 5 show that the intrinsic variability of the three aggregates of the taxon con- 

 sidered here, the Chartners sam.ple, and those of British and Continental European 

 herbarium material, is not appreciably greater than that found in comparable sampljes 

 of N. lutea and N. pumila. Had the herbarium samples alone been involved, this would 

 hardly have been a matter for surprise; " N. intermedia " was, after all, a taxon created 

 to cover plants falling between N. lutea and N. pumila, and it is only to be expected 

 that plants preserved in herbaria under this name should have been selected to be as 

 " intermediate " as possible in character. The case of the Chartners Lough colony is 

 somewhat different, since the sample analysed covered the full range of variation observed, 

 there being no question of selection for plants of intermediate character. 



The presumed parents of N. intermedia differ from each other mostly in " quanti- 

 tative " characters, and this fact, and the extremely intermediate nature of Caspary's 

 artificial hybrids, would seem to suggest a polygenic control of the characters involved. In 

 a hybrid colony containing generations later than the Fj, one might expect some degree 

 of segregation, and certainly a wider range of variation than in " pure " colonies of either 

 parent. The relatively narrow variation range of the Chartners colony, and others like 

 it, indicates either that subsequent generations to the Fi are not present, or that some 

 form of selection is acting to ensure the survival only of plants near the F^ mode (Fig. 5). 



The possibility that colonies like that in Chartners Lough are in fact ancient clones 

 derived from a single original Fj plant cannot be dismissed from consideration. The 



