290 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1928 



The theory of the ancient land bridges, created by geologists and 

 more especially by zoologists and botanists, has, however, been car- 

 ried too far. It is often ea,sy to explain any real or fancied organic 

 interrelationships by such a hypothesis, but sound science demands 

 that there should be a careful examination to see if the facts to be 

 explained can not be harmonized without recourse to bridging and 

 rebridging deep oceans. Matthew,^- for example, has shown that 

 the mammalian faunas of the Cenozoic era do not necessarily imply 

 the existence of Gondwana at this time, and Chamberlin has argued 

 against the existence of this land bridge at any time. 



Nevertheless most European and some American geologists hold 

 that a large number of organic relationships exist which establish 

 the former existence of the above-mentioned east and west continents, 

 and that the configuration of the present continents has come about 

 since Cretaceous time. 



ARGUMENTS AGAINST CONTINENTAL FRAGMENTATION 



The arguments against the acceptance of continental fragmenta- 

 tion, as the writer comprehends them, may be summarized as follows : 



First. There is no direct evidence of the existence of former 

 lands in the present ocean basins. Moreover, some of the supposed 

 paleontologic evidence calling for land bridges may perhaps be ex- 

 plained as well by existing routes, and organic evolution under 

 similar climatic conditions may have brought about likenesses in 

 form where no close relationships actually exist. 



Second. If the Paleozoic continents stretched east and west across 

 the Atlantic Ocean, there should still be recognizable east-west 

 trends in the relief of the ocean floor or in the bounding lands; the 

 former land bridge's, even if submerged, should still be distinguish- 

 able. It is, of course, true that submerged plateaus do connect 

 Australia with Asia and Antarctica, and in the North Atlantic there 

 is the evidence of the Wyville-Thomson Kidge connecting Scotland 

 and Iceland. On the other hand, the mid-Atlantic plateau runs 

 north and south, and the ocean is as deep over the site of Gondwana 

 as it is elsewhere. 



Third. The continental platforms as a whole are known to be in 

 rather close adjustment to the density of the outer 50 to 100 miles 

 of the crust. To suppose then that an ancient continent has partially 

 foundered requires that the crust must here have become on the 

 average from 3 to 5 per cent denser. There is as yet no known geo- 

 logic process which could bring about such an increase in density. 

 Unless such a process is shown to be possible, the other evidence must 



12 W. D. Matthew, Cllninto and Evolution. Ann. N. Y. Acad. Sci., 24, 171-318. 1015. 



