I 



TRANSACTIONS OF SECTION D. 785 



and tlie universe, and we read him, and go on our way, and straightway forget. 

 Herodotus and Thucydides tell a plain tale in plain language, or the Curate of 

 Selborne shows us the hawk on the wing, or the snake in the grass, as he saw them 

 day by day, and, somehow, the simple story lives and moves him who reads it 

 long after the subtleties of this or that philosophical theory have had their day and 

 passed into the limbo of oblivion. But, invaluable as has been the example of 

 Gilbert "White in teaching us how to observe, his field was a very narrow one, 

 circumscribed for the most part by the boundaries of a single parish, and on the 

 subject of geographical distribution (as we know it now) he could contribute 

 nothing, a subject on which even the best explorers of that day were strangely 

 inobservant and inexact. A century and a half ago, it had not come to be recog- 

 nised that distribution is (along, of course, with morphology and physiology), a 

 most important factor in determining the facts of biology. It is difficult to esti- 

 mate what might have been gained in the case of many species, now irreparably 

 lost, had Forster and the other companions of Captain Cook, to say nothing of 

 many previous voyagers, had the slightest conception of the importance of noting 

 the exact locality of each specimen they collected. They seem scarcely to have 

 recognised the specific distinctions of the characteristic genera of the Pacific 

 Islands at all, or, if they did, to have dismissed them with the remark, ' On this 

 island was found a flycatcher, a pigeon, or a parrot similar to those found in New 

 Holland, but with white tail-feathers instead of black, an orange instead of a 

 scarlet breast, or red shoulders instead of yellow.' As we turn over the pages of 

 Latham or Shaw, how often do we find for locality ' one of the islands of the South 

 Sea,' and, even where the locality is given, subsequent research has proved it 

 erroneous, as though the specimens had been subsequently ticketed ; as Le Vaillant 

 described many of his South African birds from memory. Thus Latham, after 

 describing very accurately Wiipidiira Jlabellifera, from the south island of New 

 Zealand, remarks, apparently on Forster's authority, that it is subject to variation ; 

 that in the island of Tanna another was met with, with a diff"erent tail, &c., and 

 that there was another variety in the collection of Sir Joseph Banks. Endless per- 

 plexity has been caused by the Psittaciis pygmcens of Gmelin (of which Latham's 

 type is at Vienna) being stated in the inventory as from Botany Bay, by Latham 

 from Otaheite, and in his book as inhabiting several of the islands of the South Seas, 

 and now it proves to be the female Psittncus palmarnm from the New Hebrides. 

 These are but samples of the confusion caused by the inaccuracies of the old 

 voyagers. Had there been in the first crew who landed on the Island of Bourbon, 

 I will not say a naturalist, but even a simple-hearted Leguat, to tell the artless tale 

 of what he saw, or had there been among the Portuguese discoverers of Mauritius 

 one who could note and describe the habits of its birds with the accuracy with 

 which a Poulton could record the ways and doings of our Lepidoptera, how vastly 

 would our knowledge of a|perished fauna have been enriched ! It is only since we 

 learned from Darwin and Wallace the power of isolation in the differentiation of 

 species that special attention has been paid to the peculiarities of insular forms. 

 Here the field naturalist comes in as the helpful servant of the philosopher and 

 the systematist by illustrating the operation of isolation in the difl^ereatiation of 

 species. I may take the typical examples of two groups of oceanic islands, differ- 

 ing as widely as possible in their position on the globe — the Sandwich Islands in the 

 centre of the Pacific, thousands of miles from the nearest continent, and the Canaries, 

 within sight of the African coast — but agreeing in this, that both are truly oceanic 

 groups, of purely volcanic origin, the ocean depths close to the Canaries, and be- 

 tween the different islands, varying from 1,500 to 2,000 fathoms. In the one we 

 may study the expiring relics of an avifauna completely differentiated by isolation ; 

 in the other we have the opportunity of tracing the incipient stages of the same 

 process. 



The Sandwich Islands have long been known as possessing an avifauna not 

 surpassed in interesting peculiarity by that of New Zealand or Madagascar ; in 

 fact, it seems as though their vast distance from the continent had intensified the 

 inffuences of isolation. There is scarcely a passerine bird in its indigenous fauna 

 which can be referred to any genus known elsewhere. But, until the very recent 

 1893. 3 E 



