EASEMAN, GEOGRAPHICAL DISTRIB UTION IN SO VTH AMERICA 49 



Part II. Distribution of the South American Fishes and its 

 Bearing upon Alleged Connections between South 

 America and the Eastern Hemisphere 



distribution of the fishes 



Introduction 



In attempting to explain the distribution of South American fishes, I 

 have been handicapped by lack of data bearing upon 



(a) The question whether all of the species concerned are real species. 

 Experimental work is entirely lacking which would demonstrate whether 

 the species are real, composed of several elementary species, varieties or 

 only individual and local variations (ontogenetic species of Jordan) 

 which have responded to different external conditions, i. e., affecting the 

 somaplasm and not the germplasm. 



(b) The nature of the ancestral species. That is to say, which species 

 of a given genus is the nearest to the ancestral one that became widely 

 distributed. 



(c) Actual ancestral form. Fossils are exceptionably rare and frag- 

 mentary. 



If I followed the old static method, not only such data but also the 

 changes wrought by the environment would almost be neglected. In this 

 case, it would be relatively easy to compile the exact localities of all of 

 the species in a given family, and by comparing the long list so obtained, 

 divide the world into as many faunal regions as these data would warrant: 



In strong contrast to this older method used by the most of the writers 

 on geographical distribution of South American animals is the study on 

 Leptinotarsa by Tower. 24 I cannot improve on his statement (p. 52) 

 concerning the two viewpoints of geographical distribution, which is as 

 follows : 



"The geographical distribution of animals, or animal geography, is usually 

 considered from one of two viewpoints, the static or the dynamic. Considered 

 from the static standpoint, the facts of distribution are taken and arranged 

 according to some empirically chosen standard, and zones, sub-zones or other 

 unnatural areas of distribution are established. The study of animal distribu- 

 tion from this standpoint is a dead and profitless pursuit. Dynamically con- 

 sidered, animal geography seeks to explain the facts of animal distribution as 

 we now find them in terms of the relation of the animals to each other and to 

 their environmental complexes." 



24 W. L. Tower : "An Investigation of Evolution in Chrvsonielid Beetles of the Genus 

 Leptinotarsa." Carnegie Inst, of Washington. 1906. 



