In a recent issue of the Journal of the Linnean Society (Zoology, No. 172) there is a short paper by my friend Dr. St. George Mivart, in which he gives numerous cases of species of Lories peculiar to various Papuan or Pacific Islands, which differ in some details of coloration from allied species in other islands, while they are usually altogether unlike the other birds inhabiting the same island. He then argues, as Captain Hutton had done with regard to similar phenomena among the fruit pigeons of the genus Ptilopus, that these various specific markings cannot be useful, and especially that they cannot be needed as "recognition-marks," because the whole coloration of the genus is so distinct that they cannot possibly be confounded with any other birds now inhabiting the same islands. He therefore concludes that these facts "are fatal to a utilitarian explanation of the origin of all specific characters." At the same time he accepts evolution and the natural biological origin of these and all other characters. These conclusions appear to me to be wholly illogical and to be reached by omitting to take account of the fundamental idea of organic evolution itself, namely, that each species has been, somehow, developed from an allied but distinct species, living or extinct. I therefore ask leave to point out how this omission affects the problem.

    It is quite clear then that each distinct species of lory or fruit pigeon now found isolated from their allies in so many of the Pacific Islands must (if evolution is admitted) have originated by modification from some other parent species. The modification may have occurred in another island (or continent) or in the island in which the modified species now exists; but, in either case during the process of differentiation, recognition-marks would be of vital importance by checking intercrossing, so much so that it is doubtful whether in many cases the required structural or physiological modifications could be brought about without them. I do not remember that this proposition has been seriously denied, and it is the omission to take account of it that invalidates the argument of Dr. Mivart and Captain Hutton, founded upon the existing distribution of the species in question.

    Perhaps these gentlemen will reply that they hold the views of Romanes and Gulick, that the specific differences in question are the direct result of the action of changed conditions on the progeny of the individuals which first reached the islands; but this theory is a pure assumption in support of which I am not aware that any adequate facts or observations have been adduced, while such changes in all the individuals exposed to the influence of the new conditions is entirely opposed to the known facts of variation. Supposing, however, that the existing species originated in the islands where they now occur by modification of some two or more original immigrants, let us consider how the change would be effected in accordance with the known facts of variation and natural selection.

    The first thing that happens on the introduction of a new form into an island well-suited to it, and with no other enemies than those to which it is already adapted, is to increase rapidly till the island is fully stocked--witness the rabbit in Australia, New Zealand, and Porto Santo, the sparrow in America, and numerous other cases. But as soon as the island is fully stocked and all future increase dies off annually, natural selection begins its work, and the least adapted to survive, in every stage from the egg to the parent birds, get destroyed by some means or other. Now, if this process of elimination is identical in character with that to which the species was subjected in its former home no specific change will take place, because the whole structure and habits which constituted "adaptation to conditions" in its former habitat are equally effective in its new abode. But if there is any difference in the environment which requires a new adaptation, whether as regards food, seasons, diseases, or enemies of other kinds, then natural selection will certainly tend to bring about that new adaptation, and as in such a limited area local segregation will be ineffective, some external indication, marking off the new and better adapted from the old less adapted type, will be of the first importance in the prevention of inter-crossing and thus hastening the process of complete adaptation; and these external indications are what I have termed "recognition-marks." When the new type is fully established and the old parent-form has died out, the work of these recognition-marks will have been done; but having been established by a severe process of selection they have become fixed and continue to form the "specific character" distinguishing the new from the old species. The repeated statement of Dr. Mivart, that in this or that case the peculiar marking cannot be a recognition-mark, or that such "recognition-marks" are quite needless, is therefore beside the question, since the very existence of the new species during the process of differentiation may have depended upon them.

    I have here confined myself strictly to the one point raised by Dr. Mivart and Captain Hutton, having already dealt with the general question of "utility" elsewhere.