Edith R. Saunders 225 



remain pure for a considerable period, and be taken in such areas to 

 represent the type form of the species. It is possible that the Malvern 

 Hills afford a case in point, for according to Anne Pratt the Foxglove 

 suddenly appeared in abundance on one of these hills after, for some 

 reason, the soil had been turned up\ The same argument may be 

 advanced in the analogous case of Linaria alpina mentioned above, where 

 indeed a hybrid origin seems even more improbable. 



How far a fuller knowledge of the geographical distribution of the 

 two D. purpurea forms will serve to throw light on their genealogical 

 relation remains to be seen. The home of the genus appears to be in 

 Central and Southern Europe, numerous species occurring in Germany, 

 France, Spain, Italy, Austria ; it is also represented by several forms in 

 Russia, Belgium, and Portugal, purpurea itself being found in all these 

 cuuntries, as well as in England, Denmark, Norway and Sweden where 

 it is the only recorded species. I have already drawn attention to the 

 wide distribution over Great Britain of purpurea nudicaulis in company 

 with purpurea pubescens. The further fact that herbarium material 

 examined at Kew was found to include specimens of nudicaulis as well 

 as pubescens both from Norway and Sweden, taken together with the 

 descriptions of the species in Neuman's Swedish Flora (stem cylindrical, 

 hairy above) and in Blytt's Norwegian Flora ([plant] covered with short 

 downy hairs especially above) renders it highly probable that nudicaulis 

 will very generally be found with pubescens in these countries, as in 

 England. If this should prove to be the case it strengthens still further 

 the argument against hybridisation as an explanation of the appearance 

 of nudicaulis. It may be urged no doubt that the absence of all mention 

 of glabrousness as a feature of the stem in the early descriptions of 

 D. purpurea, such as those of the older herbalists, renders the view here 

 suggested improbable. But seeing that neither in modern works do we 

 find any clear reference to the existence of two forms despite the abun- 

 dance and wide distribution of nudicaulis this objection may be put 

 aside. A similar omission is very general in the case of Linaria alpina ; 

 many botanical works dealing with the flora of regions where both forms 

 are to be found make no mention of the concolor form. And here 

 perhaps the matter must rest until further evidence throwing light on 

 the question is available. 



^ The Flowering Plants and Ferns of Great Britain, Vol. iv. 



