106 ECOLOGICAL FOUNDATIONS OF ZOOGEOGRAPHY 



separated by seas, are less convincing. Palaeontology, from the identity 

 or differentiation of fossil faunae, is able to conclude with a high degree 

 of probability on the nature of the contemporary connections or 

 separations. The data of palaeontology, however, are for the most part 

 meager, so that only rarely, as in the question of the former connection 

 of North America and Eurasia, is this kind of evidence fully available. 

 The principal and often the only source of evidence for the existence 

 of former land connections remains in the data of zoogeography, but 

 this evidence is least reliable and becomes progressively less useful as 

 the period of the supposed connection becomes remote. As accessory 

 evidence, to support and illustrate the conclusions of geology and 

 palaeontology, especially with regard to geologically recent changes, 

 the phenomena of animal distribution have a high value. When em- 

 ployed alone, they are of doubtful importance, and even a large amount 

 of evidence must be interpreted carefully and critically. 



Such critical care has only too often been wanting, and this field of 

 zoogeography has become a clearing-house for fantastic combinations. 

 From Forbes to von Ihering and Scharff, the tendency to explain facts 

 of animal and plant distribution by assuming the presence of land 

 bridges has been uncontrolled. The unbridled hypotheses concerning 

 the rise, displacement, and connection of land masses have left scarcely 

 a spot which has not at some time been involved in a land bridge. 21 In 

 spite of the repeated warning of conservative investigators, some 

 zoogeographers continue to "make continents as easily as a cook makes 

 pancakes" (Darwin) , 22 



The first step in the comparison of related faunae, whose compo- 

 nents are "homologous," i.e., taxonomically comparable, is to deter- 

 mine the degree of relationship between the animals in question. This 

 is not always an easy problem, and the views of specialists on the rela- 

 tions within the same group may be divergent. Every phylogenetic 

 conclusion, unless supported by abundant palaeontological evidence, is 

 dependent on assumptions, and these become the less dependable the 

 greater the degree of difference between the forms which are compared. 

 The presence of large flightless birds in the southern hemisphere, os- 

 triches in Africa, rheas in South America, emus and cassowaries in 

 Australia, and the recently extinct moas in New Zealand, was regarded 

 as important evidence of the connection of these regions by an antarc- 

 tic continent by Hutton 23 and others. The assumption, however, that 

 these forms belong to a natural group is highly dubious, and they are 

 now placed in distinct families. Their resemblances, consisting in re- 

 duced wings with the loss of the keel on the sternum, strongly devel- 

 oped legs, with a reduced number of toes, loosening of the plumage, 



