chap, ix.] INDO-MALAY ISLANDS. 223 



these ten are found in Java and five in Sumatra, a propor- 

 tion about the same as that of the Bodents, which have no 

 direct means of migration. We learn from this fact, that the 

 seas which separate the islands from each other are wide 

 enough to prevent the passage even of flying animals, and 

 that we must look to the same causes as having led to the 

 present distribution of both groups. The only sufficient 

 cause Ave can imagine is the former connexion of all the 

 islands with the continent, and such a change is in perfect 

 harmony with what we know of the earth's past history, 

 and is rendered probable by the remarkable fact that a 

 rise of only three hundred feet would convert the wide seas 

 that separate them into an immense winding valley or plain 

 about three hundred miles wide and twelve hundred lono-. 

 It may, perhaps, be thought that birds which possess 

 the power of flight in so pre-eminent a degree, would not 

 be limited in their range by arms of the sea, and would 

 thus afford few indications of the former union or separa- 

 tion of the islands they inhabit. This, however, is not the 

 case. A very large number of birds appear to be as strictly 

 limited by watery barriers as are cpiadrupeds ; and as they 

 have been so much more attentively collected, we have 

 more complete materials to work upon, and are enabled 

 to deduce from them still more definite and satisfactory 

 results. Some groups, however, such as the aquatic birds, 

 the waders, and the birds of prey, are great wanderers ; 



