A Frozkn World 205 



fresh-water mussels ca«i ascend tlie river into tlie 

 brook, and pass by the brook into the pond, 

 wliich has thus a direct hne of communication 

 witli all waters elsewhere, including even the s^reat 

 oceans. But the pond without an outlet cannot 

 thus be peopled. Whatever inhabitants it possesses 

 have come to it much more by pure chance. They 

 are not able to walk overland from one pond to 

 another ; they must be brou<4ht there somehow, 

 by insi^nihcant accidents. Kej^arded in this lij^ht, 

 the orij^inal peopling of every pond in Enj^land is 

 a problem in itself a problem analo<4()us in its 

 own petty way to the problem of the peoplini^ of 

 oceanic islands. 



That j^reat and accomplished and ingenious 

 naturalist, Mr. Alfred Russel Wallace, workinjf 

 in part upon lines lon^ since laid down by 

 Darwin, has shown us in detail how oceanic 

 islands have in each case come to be peopled. 

 He has shown us how tliey never contain any 

 larj^e indigenous land animals belonging to tlie 

 ^reat ,qroup of manunals — any deer or elephants 

 or pi^s or horses ; because mammals, bein*^ born 

 alive, cannot, of course, be transported in the 

 egg, and because the adult beasts could seldom 

 be carried across great stretches of ocean by 

 accident without perishing on the way of cold, 

 hunger, or drowning. One can hardly imagine 

 an antelope or a buffalo conveyed safely over 

 sea by natural causes from Africa to the Cape 

 Verdes, or from America to the Bermudas. As 

 a matter of fact, therefore, the natural population 



