The second is an inference from the diversion of the Protonesian population from New Guinea to the Micronesian line, since the best reason that can be assigned for the Protonesians not having taken possession of the Papuan isles, is to be found in the assumption that they were previously inhabited. This brings us to the third question, as to the import of the darker coloured populations in areas more especially belonging to the brown and olive-coloured tribes.--I do not see how we can consider these as aught else but the lighter-coloured populations in a ruder stage of society; since unless we take this view we must look upon them as the representatives of a separate section of the human kind; a supposition against which there are the two following objections. _a._ That the difficulties respecting the population of the Polynesian area are just doubled by such an assumption; since instead of having to account for the undoubted Polynesians alone (a matter quite difficult enough of itself) we should then have to account for an earlier migration of Negritos as well. _b._ That if such a previous migration had taken place, we should expect to find--considering the vast number of Polynesian islands--_at least one_ island where the blacker race remained unmixed, and (as such) speaking the original non-polynesian language, which is implied in the assumed independence of origin; since it is exceedingly unlikely that a second migration should have so nearly coincided with a former one as to people and leave unpeopled exactly the same areas. Now out of all the isles of the South Sea none presents the phenomenon of a pure black population, as determined by the double test of colour and of language. On the other hand, it may be urged--_a._ That, although it may be a matter of doubt with competent judges whether improved physical and social conditions have so great an influence upon the colour of the skin and the texture of the hair as is imagined by some extreme thinkers on the point, it is generally admitted that they have _some_ influence. _b._ That in some groups (and sometimes in particular islands) the identity of the darker and lighter-coloured population is beyond a doubt; coinciding, as it does, with such differences. _c._ That transitional forms occur where it is wholly gratuitous to assume the influence of intermixture. With this opinion our view of the relations between the continuous Kelænonesian areas and the areas of the mixed population would be as follows:-- _a._ That at a period anterior to the development of the proper Malay and Polynesian characters of the typical Protonesians, New Guinea and Australia were peopled from the Moluccas and Timor respectively; the immigrants having a type which might lose or gain Kelænonesian characters according to circumstances. _b._ That the conditions of Protonesia and Polynesia favoured the change from dark to fair; those of New Guinea and Australia from fair to dark.