FRAGMENTATION BAEKELL 291 



become of compelling force and definiteness before the hypothesis of 

 fragmentation can be accepted. 



These counter arguments will now be taken up for discussion. 

 The positive evidence bearing on the first one has been given under 

 the previous topic, and is of such weight that most European and 

 some American geologists have accepted fragmentation as a funda- 

 mental phenomenon in geologic histor3^ The objections stated 

 under the last two heads, however, have never been squarely met. 

 In fact, they have gained most of their strength in America in recent 

 years through the recognition of isostasy as a real and measurable 

 factor. It is these objections that will now be more fully dealt with. 



As to the relief of the ocean floors, we really know very little 

 about it in detail. The usual method of exploration, such as that pur- 

 sued by the Challenger expedition, is to make soundings at distances 

 of many miles, along lines that are hundreds of miles apart. In 

 fact, the map of the Atlantic between Brazil and Africa published 

 by the Institut fiir Meereskunde, Berlin, gives but about 150 sound- 

 ings. If the elevations on the land were similarly determined and at 

 such distances apart, a wholly erroneous mental picture of the gentle- 

 ness of land relief would be obtained. On the other hand, the mar- 

 gins of the ocean basins must become smoothed out by the deposit of 

 land waste, but such smoothing would extend generally not more than 

 200 miles, and therefore could not obscure the larger relief of the 

 basins. Moreover, detailed surveys for cables have revealed sur- 

 prising irregularities in places, and regions are known that have to 

 be avoided. 



This sharp relief seems to be most largely due to submarine fault- 

 ing and submarine volcanic activity, but it is also possible that some 

 of it is due to the preservation of an ancient land surface, or to old 

 faults connected with subsidence. If the north-south trends on the 

 floor of the Atlantic actually exist, they may be related to the vol- 

 canic activity of Cenozoic time, but it is more probable that they are 

 due in part to horizontal pressures like those generating mountain 

 folds. However, in the submerged mountain ranges on the floor of 

 the Pacific this relation of ridges to volcanoes is quite clear. From 

 these considerations we see that the present relief of the Atlantic 

 bottom may be chiefly due to causes that have operated since the 

 breaking down of Gondwana in the later Mesozoic, and accordingly 

 would tend to obscure the original east-west trend of the lands. 



If the fragmentation of Gondwana is due to the same causes re- 

 sponsible for the original ocean basins, it might be expected that 

 subsidence under these same causes would take place to about the 

 same amount. Accordingly, the successive additions to the ocean 

 floor need not be marked by a simple steplike character. Then, too, 



