114 



SCIENCE 



[N. S. Vol. XLIII. No. 1100 



tions from continent to continent have 

 played a highly significant part in bring- 

 ing about the present geographical ar- 

 rangement of animals and plants. This 

 conception was first suggested by Cuvier, 

 who, however, would seem not to have at- 

 tached great importance to it, and it fell 

 into neglect together with his theory of 

 Catastrophism. Present geographical dis- 

 tribution is, when well understood, in itself 

 a partial record of those past changes, par- 

 tial, because of the extinction of many 

 forms which, in every region, once existed, 

 but have completely vanished. Such mi- 

 grations from continent to continent were, 

 it should be distinctly understood, radically 

 different in character from the annual 

 migrations of birds, and it is unfortunate 

 that the same term should be used to desig- 

 nate such very distinct classes of facts. 

 The so-called migrations with which we 

 have to deal, such as those of mammals, 

 are a purely unconscious and uninten- 

 tional spread into new areas, as the in- 

 creasing number of individuals of a given 

 species begin to press upon the sources of 

 food-supply. This spread will continue 

 from generation to generation until in- 

 superable barriers are encountered and 

 there the spread must cease, unless some 

 geographical or climatic change should re- 

 move the barrier, when the spread will con- 

 tinue. For nearly all land animals the 

 most impassable of barriers is the sea, even 

 in narrow arms, though the climatic fac- 

 tors of temperature and, in somewhat less 

 degree, moisture are of almost equal im- 

 portance. In the long course of geological 

 time and even in its later portion, with 

 which we are here more particularly con- 

 cerned, both the relations of the various 

 continents to one another and climatic 

 conditions have undergone great and re- 

 peated changes, and it is those changes, 

 with their consequently varying possibili- 



ties of intermigration, which are registered 

 in the geographically composite fauna of 

 almost every great land area of the earth. 

 Enough has been already learned regard- 

 ing the history and development of the 

 various mammalian groups to make it 

 plain that the mammals of each different 

 continent form a sort of mosaic, the pairts 

 of which are of the most diverse places of 

 origin and dates of immigration. The 

 geological date largely determines the 

 amount and kind of modification which 

 the creatures have undergone in their new 

 homes. 



In attempting to estimate the signif- 

 icance of these facts, one assumption must 

 be made, an assumption for which there is 

 a large and increasing body of evidence, 

 namely, that, in the higher animals at least, 

 the same group never originated indepen- 

 dently, from ancestors either similar or 

 different, in two disconnected regions. It 

 is perfectly true that parallel and con- 

 vergent modes of development have always 

 been important factors in the evolutionary 

 process, and no one is more firmly per- 

 suaded of this than the paleontologist, but 

 there is no reason to believe that these 

 modes of development ever went so far as 

 to produce substantially identical results 

 in separate land areas. These principles 

 are best illustrated by the mammals, simply 

 because the past geological history of that 

 group has been ascertained more fully and 

 continuously than that of any other class 

 of animals. 



The scheme for dividing the land sur- 

 face of the globe into zoological regions, in 

 accordance with the distribution of ani- 

 mals and especially of mammals, has been 

 the subject of much controversy, but now 

 a general agreement has been reached. 

 Thus, the very long isolation of Australia 

 is recognized by setting off that continent 

 and its adjoining islands, in contradistinc- 



