158 The Scottish Naturalist. 



All the same, slowly but surely the changes were taking place. 

 The species at one time abundant in the suppositious locality 

 under observation would become rarer and rarer, till they had 

 ceased to live in that locality at all ; but their places would be 

 supplied by other — more northern or mountain — species, which, 

 mingling with the others, would gradually supplant them, but 

 would in their turn be supplanted by still more alpine or arctic 

 species. And what would happen in one locality would happen 

 in all to which the phenomena of the glacial period extended — 

 the more northerly or more alpine suffering first, the southerly 

 and low grounds at a later period. 



Moreover, there were times in which the change in the climate 

 (and consequent alteration in the fauna and flora) would not 

 only not alter for the worse, but, to a greater or less extent, for 

 the better. And in these periods ("inter-glacial periods") the 

 species driven out would return, though perhaps not all of them, 

 and perhaps others would come which had not previously in- 

 habited the locality, all however to be again driven out or 

 supplanted when that inter-glacial period came to an end. 



At last, however, the great glacial period would reach its 

 maximum intensity, and the climate would gradually (but, as 

 before, interruptedly) become better, and would be accompanied 

 by similar phenomena of a changing fauna and flora. 



In short, there were during all this long period — extending 

 over many thousand years — great but gradual oscillations of 

 climate, and consequent shiftings in the component parts of the 

 fauna and flora, which could not fail to make immense alterations 

 in the species. 



From all this we may gather that it is probable that the species 

 whose distribution at the present day is widest are the more 

 ancient, and that those of a less wide range have possibly had a 

 later evolution. 



Of the species discussed in this paper, Ajiarta cordigera is the 

 most widely distributed, and A, mcla)wpa ranges nearly as far. 

 These are the only species (amongst our mountain Lepidoptera) 

 which occur in America, and therefore probably existed as arctic 

 species before the glacial period. In America theyjiave been 

 reported from Labrador only, so that apparently they are not 

 circumpolar. 



European (or palaearctic) species had two routes by which they 

 might have gone to America (the nearctic region). One was by 

 a broad land-connection, reaching south of Greenland, that united 



