LIMITATION BY BARRIERS. 185 



the ocean, their distribution seems at first an impos- 

 sibility. But this is far from being true ; for every 

 one knows that fresh-water species are not thus 

 limited, and few would have the hardihood to 

 claim that each system of rivers had created for it a 

 pair of parents for every species. Indeed, the actual 

 means of dispersal from one body of fresh water to 

 another have been discovered in many cases. 

 Changes of level in the land cause rivers once sepa- 

 rated to unite either permanently or temporarily ; 

 and floods have from time to time had the same 

 effect Again, the eggs of many fresh-water species 

 can survive exposure to the air for a long time, and 

 clinging as they do to weeds and grass, may fre- 

 quently be carried from place to place by birds or 

 shaggy-haired animals. Nor does the sea always 

 prove an impassable barrier, for many cases are 

 on record of fresh-water species enduring without 

 injury, for a long time, immersion in salt water. 

 Floating ice has been known to carry fresh-water 

 organisms across bodies of salt water, and here is a 

 means of communication between different conti- 

 nents. Particularly significant is this suggestion when 

 we bear in mind that instances of fresh-water animals 

 being found on both sides of a large body of salt 

 water, occur almost exclusively in the colder regions 

 of the world, where this factor could come into 

 play and where the continents approach each other. 

 Again, some fresh-water animals can become accus- 

 tomed to salt water, and vice versa ; so that actual 

 migration through the sea may take place in some 

 instances. And the significance of these facts is 



