xviii INTEODUCTION. 



extinct. All the other forms which had been accounted Meliphagine presented 

 a peculiar structure of tongue forbidding that alliance, or any affinity to the 

 Prionopidce, Dicceidcv, or Nectariniidce, but revealing a distinct relationship to 

 the Ccerebidce— now known as a family characteristic of the Neotropical Region! 

 Hereby a beam of light was thrown on the origin and derivation of the ornithic 

 population of the Sandwich Islands. The distinct inference was that the first stock 

 of their existing avifauna was received from America, in days when the range of the 

 Ccerebidce extended further to the northward than it does at present, and that certain 

 cognates or ancestors of the present Ccerebidce colonized the islands, there differentiating 

 into the modern Drepanididce. The importance of this inference on views that are held 

 as to the geographical distribution of birds in North America is a subject into which 

 there is no need here to enter, for that would be a subject foreign to my present 

 remarks ; but I doubt not it will receive due attention from American ornithologists, 

 whom it most nearly concerns. 



" That these colonists, from what I have elsewhere ventured to term a ' Columbian ' 

 fauna — since it cannot literally be called a Neotropical one, and is certainly not 'Nearctic' 

 — were the earliest settlers which have left descendants one can hardly doubt, for they 

 have existed in the Sandwich Islands long enough to undergo a great amount of 

 change. Subsequently there has been a small infusion of blood from the ' Australian 

 Region.' I say subsequently, because Dr. Gadow has shewn that this immigration has 

 undergone comparatively little modification. We have (or had) the two Meliphagine 

 genera Acrulocercus and Chcetojotila — the latter, indeed, beyond anatomical examination, 

 but shewing no very great external deviation from well-known Australian types ; while 

 the former undoubtedly retains the normal Meliphagine tongue. To these may be 

 added Chasiempis, a well-marked genus ; but, without question, very nearly allied to 

 the genus Rhipidura, so widely spread over the Australian Region, and found also in 

 New Zealand. Thus three genera constitute, so far as I am able to see, the 'Australian ' 

 element in the avifauna of the Sandwich Islands — and what are they among so many 

 others % x 



" More recently than this Australian infusion has supervened an influx of Holarctic 

 types, and especially of the Fringillidce. Whether these have arrived from America or 

 Asia, I do not pretend to say ; but the long chain of islets running to the westward — 

 one of which produces a remarkable form (Telespiza cantans), the knowledge of which 

 we also owe to Mr. Wilson (Ibis, 1890, pp. 339-341, pi. ix.) — suggests the possibility 

 of an Asiatic origin, a possibility confirmed by the consideration that his fine Chloridops 

 Jcona may be the magnified descendant of the long-known Chloris kawarahiba, which 

 has already an enterprising relative, C. Mttlitzi (Ibis, 1890, p. 101), established in the 

 JBonin Islands. Still later must have been the appearance on the scene of members of 

 the genera Corvus and JButeo, both of which are, so far as is yet known, confined to 



1 " In connexion herewith may be noticed the absence of Parrots, Kingfishers, and Doves — all families that 

 are very characteristic of an ' Anstralian' fauna." 



