402 ZOOLOGICAL GEOGRAPHY. [PART III, 
not only on account of what we know of the permanence of 
continents and deep oceans, but because such a connection must 
have led to much more numerous and important cases of simi- 
larity of natural productions than we actually find. And if 
within the life of species such interchange may have taken 
place across seas of greater or less extent, still more easy is it 
to understand, how, within the life of genera and families, a num- 
ber of such interchanges may have occurred ; yet always limited 
to those groups whose conditions of life render transmission 
possible. Had an actual land connection existed within the 
temperate zone, or during a period of warmth in the Antarctic 
regions, there would have been no such strict limitations to the 
inter-migration of animals. It may be held to support the view 
that floating ice has had some share in the transmission of fish 
and amphibia, when we find that in the case of the narrow 
tropical sea dividing Borneo from Celebes and the Moluccas, no 
proportionate amount of transmission has taken place, but 
numerous species, genera, and whole families, terminate abruptly 
at what we have other reasons for believing to be the furthest 
limits of an ancient continent. We can hardly suppose, how- 
ever, that this mode of transmission would have sufficed for 
such groups as tree-frogs, which are inhabitants of the more 
temperate or even warm portions of the two southern lands. 
Some of these cases may perhaps be explained by the supposi- 
tion of a considerable extent of land in the South-Temperate and 
Antarctic regions now submerged, and by a warm or temperate 
climate analogous to that which prevailed in the Arctic regions 
during some part of the Miocene epoch; while others may be 
due to cases of survival in the two areas of once wide-spread 
groups, a view supported in the case of the Amphibia by the 
erratic manner in which many of the groups are spread over 
the globe. 
From an examination of the facts presented by the vari- 
ous classes of vertebrates, we are, then, led to the conclusion, 
that there is no evidence of a former land-connection be_ 
tween the Australian and Neotropical regions ; but that the 
various scattered resemblances in their natural productions 
