ALPINE RHOPALOCERA FAUNAS. 159 



five or six. There are no species whatever inhabiting the range 

 of the Atlas in Northern Morocco and Algeria. 



Turning now to the remainder of the region: in Arctic 

 Siberia to the north about five species occur ; south of the 

 central ranges their numbers diminish to two in the Persian 

 Highlands, and a similar number in the vast chain of the 

 Himalayas. 



In the Nearctic Eegion, four species only are found in the 

 Rocky Mountains, and five in the tundras of the north ; while 

 no species of this genus are found at all in the Alleghany or 

 Appalachian system in the east. 



Now what do the foregoing facts prove ? In my estimation 

 they furnish us with an approximate estimate of the extreme 

 distance south to which the alpine Rhopalocera fauna were 

 driven during the climax of the glacial period, at least in the 

 PaliTearctic Piegion. 



In preglacial times these butterflies were probably found in 

 company with their pabula throughout the whole of the northern 

 portion of the Europasian Area, their range extending to well 

 within the Arctic Circle. Upon the advance of the ice-cap they 

 retreated south until they reached the six great chains of moun- 

 tains which I have already described. Many of the species were 

 probably unable to surmount the barriers which they furnished, 

 and consequently perished ; while the majority perhaps forced 

 their way through them by means of transverse passes and 

 valleys, and survived in the sheltered and more hospitable areas 

 which they would provide immediately to the south. That the 

 butterflies did not retreat much further equatorwards I am 

 thoroughly convinced from the fact of so few species being found 

 upon the mountain chains to their front Moreover, we have 

 every reason to suppose that the climatal and phyto-geographical 

 conditions must both have been favourable to their survival in 

 the South of Europe especially, even during the climax of the 

 glacial epoch ; so that I cannot agree with Hofmann's hypothesis 

 that the alpine Rhopalocera fauna was entirely driven out of 

 Europe, the greater portion into Asia, and a few into Africa, 

 from whence he supposes they returned when the climate again 

 became warmer. 



On the other hand, I think we might fairly assume with a 

 certain degree of safety that a few species even survived in 

 certain favourable spots to the north of the Alps, the Pyrenees, 

 and the Carpathians, where the land was not submerged beneath 

 the shroud of snow and ice with which most of the north of 

 Europe was enveloped. At any rate there are some powerful 

 reasons for supposing that the remainder of Europe south of the 

 mountain chains indicated were capable of affording a safe refuge 

 to the alpine butterflies, contrary to the views expressed by 

 Hofmann, before mentioned. 



o2 



