1907. Moffat. — The Problems of an Island Fattna. 139 



" oceanic islands," or, at least, very few, in the old Darwinian 

 sense of the word — that is to say which have never been con- 

 nected by land with any continent. The old oceanic island 

 has, in short, gone the way of the Ice Age, and been swept 

 into the dust-heap of once venerable thiugs no longer worth 

 thinking about. We are told that wherever we see a fauna — 

 unless, perhaps, it consists entirely of seals or sea-haunting 

 birds — we must presume that the land on which that fauna is 

 found was once - however long ago — continental. 



This revolutionises the difficulty altogether. We no longer 

 have to ask ourselves, how did lizards come, but how did frogs 

 disappear? What has become of all the mammalian life, all 

 the batrachiau life, and a very great proportion of the reptile, 

 insect, and plant life that undoubtedly must have inhabited 

 these oceanic islands in former times, if they were parts of 

 great continental areas now submerged ? " I have not found," 

 says Darwin, " a single instance, free from doubt, of a terres- 

 trial mammal (excluding domesticated animals kept by the 

 natives) inhabiting an island situated above 300 miles from a 

 continent or great continental island." The absence of frogs, 

 toads, and newts from such islands he found, after careful in- 

 quiry, to be almost equally universal. These facts, together 

 with the general fewness of the species that do occur, un- 

 doubtedly need to be explained by some law that Darwin was 

 not called on to suggest, if we are now to believe with Dr. 

 ScharfT (who gives strong grounds for his opinion) that " ac- 

 cidental introduction cannot play an important part in the 

 making of the fauna of any country," and that consequently 

 every island which has a fauna at all, no matter how limited, 

 inherits that fauna from an age when it formed part of a 

 continent. 



Such a conclusion may be too sweeping ; but I think there 

 can be no doubt that the mere conversion of a piece of land 

 into an island does increase, to every species inhabiting that 

 land, the chance of becoming extinct. We have only to re- 

 flect that any local devastation which sweeps over an area may 

 for a time enormously diminish the numbers of a species any- 

 where ; if very severe, it may for a time exterminate it. But 

 the extermination which would be temporary when the locality 

 was part of a large continuous area would be permanent if it 



