48 MK. A. MURBAT ON THE OEOaBAPHICAL BELATIONS OF 



Under my theory tlie Pacific Islands lead to no such inference. 

 We must bear in mind that we have two distinct kinds of islands 

 to deal with in the Pacific — the coral islands, like Keeling Island, 

 the mountain-islands, like Tahiti. The coral islands are in the 

 case described by my friend ; the mountain-islands are not. What 

 Darwin says of Keeling Island is no doubt true of all the coral 

 islands :■ — " As these islands consist entirely of coral, and at one 

 time probably existed as a jnere water-washed reef, all the pro- 

 ductions now living here must have been transported by the waves 

 of the sea. In accordance with this, the flora has quite the cha- 

 racter of a refuge for the destitute ; Professor Henslow informs 

 me that of the twenty species, nineteen belong to different genera, 

 and these again to no less thaa sixteen orders "; and these nearly 

 all " common littoral species in the East-India archipelago " *. The 

 animals on that island were a rat, one or two wading or sea-birds 

 (obviously stragglers), a small lizard, some spiders, and thirteen spe- 

 cies of insects of different orders, — viz. of beetles, one minute Elater 

 (the reader will remember that of the three beetles introduced in 

 the Azores from Brazil, two were Elaters, the mode of their lar- 

 val life in timber and the hard wire-like skin of these larvae, as un- 

 susceptible to wet as a duck's back, seeming favourable to their 

 ' chances of successful transit) ; of Orthoptera, a Gryllus and a 

 Blatta ; Hemiptera, one ; Homoptera, two ; of Neuroptera, a 

 Ohrysopa ; of Hymenoptera, two ants ; of Lepidoptera, a Diopcea 

 and a Pterophorus ; of Diptera, two ; and of these, without attach- 

 ing much importance to it (as I admit that they are the product 

 of chance introduction), I may still observe that, with the excep- 

 tion of Blatta (doubtless a naval cadeau), the whole, so far as 

 named, are microtypal genera, a circumstance which ceases to be 

 surprising if I am right in considering the non- coral isles of Poly- 

 nesia as microtypal. I abandon the coral isles as " no man's land," 

 but I claim as microtypal the islands which are composed of more 

 solid stuff, especially those lying between Australia and America, 

 and which are furthest from the influence of the New Guinea and 

 Malayan subfaunas. It is from Tahiti and the Marquesas that 

 the species of Coleoptera which we know from the eastern 

 part of Polynesia have chiefly been obtained, from the New 

 Hebrides and New Caledonia that we have received those from its 

 western half. I shall take the eastern part first. Its microtypal 

 character is thus well, . although unconsciously, depicted by Pair- 



* Darwin's ' Journal,' pp. 541-543. 



