782 A CENTURY OF PROGRESS IN THE NATURAL SCIENCES 



the idea of continental drift as a necessary alternative to the idea of transoceanic 

 bridges; but he then finds himself forced to elaborate the drift theory to that 

 of an "accordion movement back and forth" of the continental areas. Equally 

 oblivious of the problems introduced by land connection theories is the French 

 entomologist Rene Jeannel, who believes in a mid-Tertiary connection of the 

 Mediterranean region with the West Indies as necessary to account for the dis- 

 tribution of certain beetles (1935-1937) and later (1941) becomes a supporter 

 of the ideas of continental drift as indicated by evidence from the faunas of the 

 subantarctic islands like Kerguelen and the Crozets, as of continental drift in 

 general from the viewpoint of entomology. 



It is quite evident that there are still numerous believers in the former 

 existence of continents where the great oceans now are ; of movement of the conti- 

 nents to their present positions from an original single continent; and of back 

 and forth movement of "accordion-type" continents. There is a strong opposing 

 school of conservatives, who hold to the belief that the continental platforms, 

 though obviously often flooded by epicontinental seas, have been stable throughout 

 the geological ages in which life has existed on land. Much of this controversy 

 is primarily geological, and only secondarily zoogeographic. My concern in this 

 matter has been lest the geologists base arguments on those of zoogeographers 

 and that these then complete an argument in a circle by triumphantly pointing 

 to the fact that the geologists support them. If the geological theories involved 

 were restricted to pre-Paleozoic or even to Paleozoic times, zoogeographers could 

 have little to say regarding them. 



Ecology and Animal Distribution 



Historical animal geography becomes properly scientific only when there is 

 adequate positive evidence from paleontology as to the history in question, and 

 negative evidence is doubly to be discounted because the fossil record is in itself 

 incomplete, while our exploration of the world for what fossils have been pre- 

 served is far from finished. The major factors producing disjunct distributions 

 are the events of geological history, which are nonrecurrent. These can be recon- 

 structed in convincing terms only when there is a quite exceptional wealth of 

 fossil evidence. I have shown how wide is the divergence of opinion among zoo- 

 geographers in this field. 



When we turn to contemporary animal distributions and examine them against 

 the background of the existing environment, we enter a sharply contrasting 

 realm of animal geography. Ecology, summing up existing environmental rela- 

 tions, supplies the guiding principles in our studies when we turn to the geog- 

 raphy of existing forms, especially at the species and infra-species level, and 

 equally when we attend to the minor biotic geographic subdivisions of a continent. 

 A new hypothesis can be tested by further observation, or even by experiment, 

 and the whole field is subject to steady and logical growth with the advance of 

 knowledge. The problems directly involved become ecological instead of historical. 

 We still must face the enormous complexity of the total environment, but at 

 least we can explore it at first hand. 



The ecological factors in animal distribution were appreciated more than a 

 century ago, for we find Schmarda's division of the earth into regions based in 



