Mr. Herbert Spencer has contended that, in certain modifications, natural selection is excluded on the grounds of the extreme complexity of the changes, and adduces the case of the Irish "elk" with its huge antlers, and the giraffe with its specially modified structure. He points out that in either case the conspicuous modification--the gigantic antlers or the long neck--involves a multitude of changes affecting many and sometimes distant parts of the body. Not only have the enormous antlers involved changes in the skull, the bones of the neck, the muscles, blood-vessels, and nerves of this region, but changes also in the fore limbs; while the long neck of the giraffe has brought with it a complete change of gait, the co-ordinated movements of the hind limbs sharing in the general modification. Mr. Spencer, therefore, argues that it is difficult to believe that these multitudinous co-ordinated modifications are the result of fortuitous variations seized upon by natural selection. For natural selection would have to wait for the fortunate coincidence of a great number of distinct parts, all happening to vary just in the particular way required. That natural selection should seize upon the favourable modification of a particular part is comprehensible enough; that two organs should coincidently vary in favourable directions we can understand; that half a dozen parts should, in a few individuals among the thousands born, by a happy coincidence, vary each independently in the right way is conceivable; but that the whole organization should be remodelled by fortunately coincident and fortuitously favourable variations is not readily comprehensible. It may be answered--Notwithstanding all this, we know that such happy coincidences have occurred, for there is the resulting giraffe. The question, however, is not whether these modifications have occurred or not, but whether they are due to fortuitous variation alone, or have been guided by functional use. The argument seems to me to have weight.[DI] Still, we should remember that among neuter ants--for example, in the Sauba ant of South America (_Oecodoma cephalotes_)--there are certain so-called soldiers with relatively enormous heads and mandibles. The possession of these parts so inordinately developed must necessitate many correlated changes. But these cannot be due to inherited use, since such soldiers are sterile.