284 CAEBOXTFEEOrS PLANTS TEOM PEEr. [vol. Ixxviii, 



whole coast of South America was now definitely shown to be 

 incorrect. 



■ This isolated Palaeozoic fragment of Paracas must clearly be 

 assigned to the coastal Cordillera, as distinct from the Mesozoic 

 Western Cordillera, and its existence led to the hope that other 

 fossiliferous Palaeozoic rocks still awaited discovery among the 

 crystalline rocks of the coast of Southern Peru. Further, the 

 evidence afforded by these fossil plants seemed to support the view 

 that the Andean Cordilleras, although subjected to repeated oscil- 

 lation, had their origin at a very much earlier date than Avas 

 generally recognized. 



On stratigraphical grounds, it could be shown that an uplift 

 took place dming the interval between Mid-Devonian and' Permo- 

 Carboniferous times ; and the surmise that the altitude of the chain 

 thus formed fell little short of that attained by the existing range 

 was sus^orested bv the dissimilaritv between the flora of the Pacific 

 region and that of Gondwanaland on the east. At the present 

 day the snowclad chain forms an effective barrier to plant- 

 migration. Was it too much to suggest that similar conditions 

 prevailed in late Palaeozoic times ? At any rate, it seemed to the 

 speaker that the argument in favour of a land-barrier separating 

 the two regions was more plausible than one in favom* of a barrier 

 formed by the sea. 



