HO ECOLOGICAL FOUNDATIONS OF ZOOGEOGRAPHY 



method to zoogeographic problems, the fact that gaps in distribution 

 are frequently reduced by fossil forms indicates that a similar explana- 

 tion of discontinuity of range is often to be invoked, even if the 

 palaeontological evidence is negative. 



The opinion of geologists should carry the greatest weight in the 

 reconstruction of former land connections. Unfortunately, however, 

 geologists are not at all agreed on the fundamental question of the 

 permanence or impermanence of the continental land masses. In view 

 of the numerous changes in the shore line of present land areas in 

 former times, and of the widespread presence of marine deposits on 

 the continents, there was at first no objection to the assumption that 

 any desired part of the ocean might have been occupied by land, 

 and land connections were invoked to explain similarity of geologic 

 structure or faunal resemblance even when they involved broad oceanic 

 areas. Dana 32 was the first to propose the theory that the continental 

 and oceanic areas were essentially permanent. A. R. Wallace supported 

 this position from the zoogeographic side, with the statement that the 

 distribution of plants and animals could be explained without sup- 

 posing radical changes in the extent of land and sea. The controversy 

 over this question is by no means at an end. Diener, 33 Matthew, 34 

 and Soergel 35 defend the permanence of the continents (with the inclu- 

 sion of the continental shelf) and of the oceans. Suess 36 and Haug 37 

 represent the opposing camp. Dacque, 38 after weighing the evidence on 

 both sides, leaves the question undecided. Wallace assumes a fixed 

 oceanic depth of 100 fathoms as the maximum over which former 

 land connections were possible, and this is plainly an inadequate view. 

 The whole geological structure of the area, the direction of submarine 

 ridges, and the width of the intervening ocean, must be taken into 

 consideration, together with the faunal relations, in the reconstruction 

 of former land connections. 39 



With the exception of small coral islands, such as those of Micro- 

 nesia and Bermuda, and volcanic islands such as St. Paul and those 

 of the Hawaiian and Polynesian archipelagoes, there are perhaps few 

 land areas which have been continuously isolated from all others. This 

 is undisputed for islands near the coasts like those of the Mediter- 

 ranean, Ceylon, Formosa, Tasmania, and New Guinea and even Japan. 

 The West Indies, also, were probably united with each other and with 

 the mainland, and Madagascar was probably attached to Africa, 

 though Matthew in 1915 34 defended the opposite view, that Madagas- 

 car and the West Indies are strictly oceanic islands. The Indo- 

 Australian Archipelago and the Melanesian islands extending from 

 New Guinea to New Zealand were probably mutually connected and 



