LIMITS TO THE RANGE OF PLANT-SPECIES. 211 



In Dumfriesshire there are but few localities in which these 

 Alpine plants occur. If one were to imagine the peat hags to be 

 an ocean, then the localities of these Alpine plants would be like 

 islands in an archipelago. Within these islands the range is 

 limited by the occurrence of special habitats. One form occurs 

 on bare dry rocks or screes of stones, another on steep slopes of 

 short turf, some only on wet rocks, others in rock crannies or the 

 mud of spring-heads. In the rare places in which a rich 

 accumulation of good loam has taken place, Epilohiif^m angusti- 

 /olium, Linn., will be found, and perhaps Trollius ; and so on. 

 Hence the absence of peat soil and the existence of special 

 habitats are the limiting causes for those forms in Dumfriesshire. 



The absence of any species cannot be taken to prove anything. 

 The ordinary idea of the origin of these Alpine plants is that 

 they are still, so to speak, on the fringe of the glacial ice. They 

 are supposed to have followed the ice sheet northwards to the 

 Arctic regions, and also up any convenient mountain which kept 

 the proper climate. They were driven north and upwards by 

 invading swarms of European species. But this process took a 

 considerable time, and the retreat was almost certainly very 

 gradual. Hence we should be grateful that during all that long- 

 continued period a perpetual series of appropriate habitats was 

 prepared for our Scottish Alpine plants. The curious Hieracia 

 which we find on isolated rocky places or corries in Dumfriesshire 

 must always have had a rock or corrie of the proper kind to fly to 

 when they were driven, step by step, from lower altitudes during 

 at least 30,000 years. ^ Their absence, even if appropriate 

 habitats exist at present, cannot be taken to prove anything. 



The existence of the right insect also appears to limit the range 

 of some of these forms. Many special bumble-bee and butterfly 

 flowers of the Alps do not exist in Scotland. The reason is 

 probably that it is in the highest degree unlikely that a Bombus 

 would be foolish enough to fly over several miles of peat-moss to 

 visit a little Alpine " island '"' on the chance of finding honey, 

 hence no Aquilegia alpina, Linn., and few of the characteristic 

 Leguminos?e and Gentians. This point must not be pushed too 

 far. Most flowers have a most-favoured-nation clause for one 



1 Keane, Etknologi/, 1896, p. 58. The discussion here of the glacial 

 period is the best, or at least the most recent, kno\\ni to me. 



