104 GLACIAL PERIOD AND GEOGRAPHICAL DISTRIBUTION. 



Nature/ 19. THE GLACIAL PERIOD AND GEOGRAPHICAL 



Feb. 20, 1879. DISTRIBUTION.* 



PEOF. ASA GRAY, in his very interesting lecture on the distribution of 

 the forest trees of the northern temperate region ('Nature,' vol. xix. 

 p. 327), after pointing out the remarkable differences that exist between 

 the forests of the eastern and western sides both of North America and 

 the Old World, suggests that the great poverty of the European as com- 

 pared with the Japan-Manchurian region in this respect was caused by 

 the Mediterranean cutting off the retreat of the flora which then occu- 

 pied Europe, as it retired, at the approach of the glacial epoch, before 

 the ice from the north. This explanation derives considerable support 

 from some other facts in geographical distribution. The most character- 

 istic Alpine and Arctic butterflies of the Palsearctic region belong to the 

 three genera, Parnassius, Chionobas, and Erebia. Of Parnassius, Dr. 

 Staudinger in his latest catalogue (1871) enumerates fourteen Palsearctic 

 species, of which three occur in North and Central Europe, ranging 

 as far south as the Balkans, but always in or near high lands, about 

 a dozen occur in temperate Asia, ranging as far east as the Amur, 

 and probably as many in North America, where they also are truly 

 Alpine butterflies. Of Chionobas one species (C. aello, confined to the 

 Alps) occurs in Central Europe, whilst six or seven others range from 

 Lapland over Russia and Siberia, Mongolia, &c., to the Amur, and there 

 are numerous species in Arctic and Alpine North America. Of Erebia 

 there are forty- five Palsearctic species enumerated by Staudinger, and of 

 these no less than twenty-five occur in the central Alpine chains of 

 Europe. The genus likewise ranges all over temperate Asia, going as 

 far south as the Himalayas and Moupin, and in North America is repre- 

 sented by a dozen or more species. Now, though an Erebia (E. tyndarus, 

 var.) occurs as far south in Europe as the Sierra Nevada, not a single 

 species of any of these three genera occurs in North Africa, although the 

 Atlas Mountains would seem eminently well suited for such Alpine 

 insects. In this case, then, it seems clear that the same cause the 

 barrier of the Mediterranean which in the case of the miocene flora of 

 Europe prevented any further retreat south, has operated to prevent any 

 similar southerly spread amongst the victorious invaders from the north 

 which pressed on the retiring host. 



With regard to the general similarity in facies and richness between 

 the East American and East Asiatic tree-flora, certain facts pointing in 

 the same direction will at once occur to the zoologist. Thus the Meno- 



* Nature, Feb. 20, 1879. 



