MACLOSKIE: CHARACTER AND ORIGIN OF PATAGONIAN FLORA. 957 



and Eastern North America, and that characteristic North American 

 genera have one or more closely representative species in the Himalayan 

 region ; and he quotes an unpublished opinion of Bentham, as the natural 

 conclusion from all these data, that the interchange between the temper- 

 ate floras even of the western part of the Old World and the New World 

 took place mainly via Asia. Mr. Bentham also refers (Journ. Linn. Soc., 

 ii, p. 34), to how frequently large American genera are represented in 

 Eastern Asia by a small number of species, which gradually diminish or 

 disappear westwards towards the Atlantic limits of Europe. There are 

 few distinctively European forms in the eastern part of America, thus tes- 

 tifying that the European region was not directly in the line of passage 

 between the Old and New Worlds. 



The distribution of animals reinforces this doctrine: Dr. C. W. An- 

 drews has proved that the proboscidians have originated in Africa, and 

 the presence of remains of more than a score of species of elephants and 

 allied forms near the Himalayas points out the probable route by which 

 they travelled from Africa to Northeastern Asia and to America. Whilst 

 it is as a center of distribution rather than of origin, that we are speak- 

 ing, probably every region which is favored with suitable climate and 

 soil is in some degree a center of origin. Mountain chains have char- 

 acteristic advantages and limitations in both ways. How much have 

 the Andes of America and the Blue Mountains of Australia done for 

 distribution at least in the countries beyond them, as high roads for 

 species which could not have travelled by valleys or plains ! Shelter- 

 ing from excessive heat and from drought, and intercepting winds which 

 convey seeds from afar, they afford friendly resting-places, with variety of 

 soil and exposure so as to suit every comer ; and improving the stock, by 

 new variations, they send them out again by various channels into the 

 general circulation. This has been the function of the American chain, as 

 Patagonian plants abundantly testify ; and the evidence points to it unmis- 

 takably as the function of the great tablelands and complex mountain 

 system which surround the "roof of the world" in central Asia. 



Most significant in this respect are those species and genera which are 

 in large measure peculiar to and characteristic of the Region of Patagonia 

 and Southern Chili, and which secure to the area the rank of a proper 

 botanical region. Some of those are related to forms existing in Brazil 

 and Texas; some allied to Australian or even South African forms; 



