793 



single or few localities occur clearly and solely within the geographic 

 limits of one of the types, the plants will usually be associated with 

 the group to which they thus make the nearest approach. For in- 

 stance, there can be no hesitation in assigning the extremely local 

 Lychnis alpina and Oxytropis campestris to the Highland type ; and 

 not much more doubt can arise respecting the propriety of placing 

 Arenaria norvegica and Primula scotica in the Scottish type. So 

 also, the local Cicendia filiformis may go to the English type ; Vero- 

 nica verna, to the Germanic type ; Erica A'^agans, to the Atlantic type. 

 But after thus disposing of a large portion of these local species, 

 there are still some others left on hand, which cannot be so fairly as- 

 signed to any of our six principal types, Potentilla rupestris and 

 Anthericum serotinum, for instance, are peculiar to single mountains 

 in North Wales. As local western species they might be associated 

 with the Atlantic type ; but the hilly and inland nature of their loca- 

 lities, and their absence from the provinces of South Wales and the 

 Peninsula, come inconveniently in conflict with the chief characters 

 of the Atlantic type. Draba aizoides and Cotoneaster vulgaris, found 

 on the rocky coasts of Wales very locally, approximate rather nearer 

 to that type, and might be associated under it, in so far as Britain is 

 concerned ; and yet, if we should extend our views, so as to take in 

 their distribution upon the continent of Europe, this would be found 

 X a misposition. Some other less local species have also a distribution 

 which does not corres]Jond with that of any of the six types speci- 

 fied ; their localities being restricted to calcareous rocks, and occur- 

 ring in such positions as not to place them properly under one of 

 those types. Examples may be mentioned in Draba muralis and 

 Hutchinsia petraea, the distribution of which is strictly neither east- 

 ern nor western, northern nor southern ; and, though they are in 

 some degree hill plants, yet they are certainly not Highland species ; 

 while their very limited area separates them as clearly from the Bri- 

 tish or general type. Eriocaulon septangulare is another anomaly, 

 which was associated with some very few other species into the ' He- 

 bridean type ' of the former work. But as these few did not make a 

 congruous group, and were numerically too insignificant to be set up 

 against those of the other six types, the Hebridean is here discarded, 

 and its half-dozen species divided between the Scottish and Local 

 types." 



In connexion with this subject, we are bound to notice some remarks 

 comprised by Mr. Watson in an Appendix to the present volume. 

 In the ' Phytologist,' we have, on several occasions, thought it neccs- 



