Geographische Verbreitung, Reisen. 195 



non. Those which are wyiely migrant and constantly arriving in the Islands are 

 of little interest, but there are others probaLly derived irom one or very few 

 specimens tbat arrived by cliance long ago. These exhibit great variability, in 

 some cases much greater than in their native homes. For instance Bactra stra- 

 minea, a New Zealand moth, is excessively variable and some of its Hawaiian 

 forms may prove to be distinct species. Such species seem to be a "hall'-way- 

 house" between species quite newly arrived and the endemic fauna. They have 

 been long enough in the Islands to become extremely variable, but not long 

 enougli for the parent species to become lost in a number of endemic descen- 

 dant species. 



How then have these ancient imniigrants, progenitors of the present fauna, 

 reached the Islands — one of the most remotely isolated archipelagoes on the 

 face of the globe and probably never in connection with any continental land — 

 and whence have they come? Perkins gives an interesting section on distri- 

 bution of animals by natural agencies, but it must suffice here to say that 

 lie attaches particular importance to driftwood as one among the various agen- 

 cies by which immigrants may arrive, and gives his reasons for doing so. He 

 does not consider the immigrants to have arrived from one direction much more 

 than from another. Many groups are so isolated that they seem to have no very 

 close allies in any other part of the world, and it is then almost impossible to 

 say from what countries they may have been derived. As a result of his 20 years' 

 study of this fauna Perkins writes: — "for my part, as I understand it, the 

 Hawaiian fauna is derived from waifs and strays from all directions. At rare 

 intervals from the Eocene tili now chance immigrants have arrived. Some have 

 been able to establish themselves, many more probably, even after a landing has 

 been effected, have failed. Those that have been successful and have found con- 

 genial conditions have often thriven amazingly, giving rise to hosts of descendant 

 species as they have become adapted to, or become modified by, diverse condi- 

 tions". The fauna cannot be said to belong to any of the great faunistic regions. 

 It has received important Clements from the Oriental, Australian, and Neotro- 

 pical, but cannot be said to belong to any one of these more than to the others: 

 and so far as at present known it is certainly not Polynesian. 



Supposing the ancestral Immigrant species to have arrived and become 

 established, Perkins passes to a consideration of species-formation within 

 the islands. He lays great stress on geographica! Isolation as a factor in 

 producing species. When the fauna is viewed as a whole, one cannot entertain 

 the idea that the separate islands have ever been joined to one another during 

 the period in which they have been colonised by the progenitors of the present 

 fauna. The narrow inter-island Channels appear to act as most efficient barriers 

 to the spread of very many species. Of course there are certain species of known 

 migratory habits which ränge throughout the archipelago and probably receive 

 recruits from distant lands from time to time, and these are never truly isolated 

 and are most unlikely to produce new forms. An example is the butterfly Pyra- 

 meis cardui: but even so strongly-flying an insect as Pyrameis atalanta, which be- 

 came established in the Island of Hawaii over 20 years back, appears never to 

 have spread beyond that Island, though separated from the next Island only by 

 a Channel comparable in width to that between Continental Europe and Britain. 



In very many of the large endemic genera scarcely any species occur in 

 more than one island. Thus the bird-genus Oreomyza has a different species in 

 each island. Each species has a very varied dietary and occurs in very diverse 

 climates and localities in its own island, yet it remains constant in that island 



13* 



