DISTIilBUTION. 175 



species through gradual variations. Thus the distribution of 

 animals constitutes an extremely complicated problem, the solu- 

 tion of which necessitates comprehensive preliminary work. It 

 must be known with certainty how the connexions between the 

 continents and the climates have changed, 2iiirticularly in the 

 later geological periods; further, we must stady, not only how 

 animals are distributed over the earth's surface at the present time, 

 but also how they were distributed in earlier times. Finally, by 

 means of comparative anatomy and embryology we must have clear 

 and detailed ideas of the relationships and interrelationships of 

 animals. 



It will be an extremely long task to solve all the problems of 

 the subject here sketched in outline. What has been investigated 

 thus far can only be regarded as a preliminary proof that zoology 

 with its prevailing views of the changes of animals and of the earth 

 is on the right track. It would be a test of the correctness of this 

 view if it were proved that the faunal resemblances of two countries 

 depends, in the first place, upon how long they have been in close 

 connexion with each other, consequently allowing an interchange 

 of the animals inhabiting them. Two regions, separated early in 

 the earth's history and never again connected, must have greater 

 differences in faunal characters than two lands still connected or 

 only recently sef)arated. It is instructive when we travel in the 

 northern hemisphere and find in widely separated regions strik- 

 ingly similar faunse, while under the equator or in the southern 

 hemisphere under the same conditions striking differences are seen. 

 This is explained on the hypothesis that in all past periods as now 

 the land masses of the northern hemisphere have been closely con- 

 nected, while the parts of the continents extending to the south — 

 aside from hypothetical temjiorary connexions between South 

 America, Africa, and Australia — have been separated through 

 most of the earth's history. 



In carrying out more closely the jioints of view mentioned, 

 students of distribution have attempted to mark off the great 

 faunal areas of the earth, the faunal jirovinces or regions, and 

 within these again less important divisions, subregions. These 

 provinces have been based chiefly upon the distribution of 

 mammals, less upon that of Ijirds and other animals; for the dis- 

 tribution of mammals is chiefly determined by those changes of 

 the earth's surface which are best known geologically and possess 

 most interest. Elevation or depression of the earth's surface often 

 opposes impassable barriers to most mammals : rising, if it lead to 



