FORMER LAND CONNECTIONS. 135 



hypothesis submitted. In the South Atlantic at opposite ends of 

 this swelling stand the Islands of Ascension and Bouvet, in the 

 middle the Tristan da Cunha group. From the latter there extends 

 in a north-easterly direction to the neighbourhood of Mossamedes 

 an extremely narrow ridge dividing the eastern half of the ocean 

 into two large basins. In the opposite direction lies another, but 

 less regular and lower, ridge crossing to the southern end of Brazil. 



Elevation of the ocean floor to the extent of from 10,000 to 

 15,000 feet would bring about the union of the two continents and 

 also develop a mid- Atlantic land. That along the Atlantic border 

 subsidences have occurred, recent yet of surprising magnitude, is 

 indicated by the remarkable submarine extension of the mouth of 

 the Congo River, soundings having proved a nearly straight trough 

 in the even and regularly dropping ocean floor traceable fully 130 

 miles out to sea, where the depth is no less than 7,500 feet. 



Realising the magnitude of the changes that might have 

 occurred along the coasts, one would not put too much weight upon 

 the extraordinary resemblances in outline between the opposite 

 coast-lines of Africa and South America, were it not that the 

 geological peculiarities of the two areas are so amazingly similar; 

 not only does this apply to the South but also to the North 

 Atlantic. 



Throughout the whole length of this ocean, almost as though 

 uninterrupted by it, one finds on the opposite sides the same fold- 

 systems, geological series, fossil floras and faunas, intrusive and 

 volcanic rocks. Allowing for some late Tertiary deposition and 

 upheaval and for some coastal erosion, the nearest existing 

 exposures are such that they could well have been only a few 

 hundreds of miles apart originally. Some of the geological 

 parallels have already been remarked upon, but many others exist, 

 for example the unique alkaline igneous rocks of the opposite coasts 

 as noted by Brouwer, while just recently Dr. P. A. Wagner has 

 pointed out to me that the crystalline forms among the Brazilian 

 diamonds resemble those of S.W. Africa more than the latter do 

 those of the Union. 



Our hypothesis starts with the assumption that the Continent 

 was first of all severed by great tear-lines and that these fragments 

 then started to move apart, just as though driven asunder by the 

 centrifugal forces set up through spinning around the polar axis 

 The action is to be conceived as a slipping of the outer part of 

 • the crust upon its yielding foundation and the phenomenon could 

 indeed be compared to the gradual opening out of cracks developed 

 in a sloping asphalt pavement. 



The Fold Ranges encircling the Fragments of Gondwanaland. 



The widening clefts just mentioned are regarded as becoming 

 occupied by the oceans and as having continued to broaden down 

 to the present day. Around the periphery of Gondwanaland strata 

 had been accumulating over a truly vast period — from the 

 Carboniferous to the Tertiary — and it may be surmised that it was 

 due to this marginal weakening by which the Continent became 



