forward by Mr. Spencer with the more immediate object of proving that there is "no concomitant variation in co-operative parts," even when "formed out of the same tissue, like the crab's eye and its peduncle" (pp. 12-14, 23, 33). It escapes his notice, however, that in two out of his three cases it is disuse, or diminished use, which fails to cause concomitant variation or proportionate variation.

THE GIRAFFE, AND NECESSITY FOR CONCOMITANT VARIATION.

Having unwittingly shown that lessened use of closely-connected and co-operative parts does not cause concomitant variation in these parts, Mr. Spencer concludes that the concomitant variation requisite for evolution can only be caused by altered degrees of use or disuse. He elaborately argues that the many co-ordinated modifications of parts necessitated by each important alteration in an animal are so complex that they cannot possibly be brought about except by the inherited effect of the use and disuse of the various parts concerned. He holds, for instance, that natural selection is inadequate to effect the numerous concomitant changes necessitated by such developments as that of the long neck of the giraffe. Darwin, however, on the contrary, holds that natural selection alone "would have sufficed for the production of this remarkable quadruped."[8] He is surprised at Mr. Spencer's view that natural selection can do so little in modifying the higher animals. Thus one of the chief arguments with which Mr. Spencer supports his theory is so poorly founded as to be rejected by a far greater authority on such subjects. All that is needed is that natural selection should preserve the tallest giraffes through times of famine by their being able to reach otherwise inaccessible stores of foliage. The continual variability of all parts of the higher animals gives scope for innumerable changes, and Nature is not in a hurry. Mr. Spencer, however, says that "the chances against any adequate readjustments fortuitously arising must be infinity to one." But he has also shown that altered degree of use does not cause the needed concomitant variation of co-operative parts. So the chances against a beneficial change in an animal must be, at a liberal estimate, infinity to two. Mr. Spencer, if he has proved anything, has proved that it is practically impossible that the giraffe can have acquired a long neck, or the elk its huge horns, or that any species has ever acquired any important modification.

Mr. Wallace, in his Darwinism, answers Mr. Spencer by a collection of facts showing that "variation is the rule," that the range of variation in wild animals and plants is much greater than was supposed, and that "each part varies to a considerable extent independently" of other parts, so that "the materials constantly ready for natural selection to act upon are abundant in quantity and very varied in kind." While co-operative parts would often be more or less correlated, so that they would tend to vary together, coincident variation is not necessary. The lengthened wing might be gained in one generation, and the strengthened muscle at a subsequent period; the bird in the meanwhile drawing upon its surplus energy, aided (as I would suggest) by the strengthening effect of increased use in the individual. Seeing that artificial selection of complicated variations has modified animals in many points either simultaneously or by slow steps, as with otter-sheep, fancy pigeons, &c. (many of the characters thus obtained being clearly independent of use and disuse), natural selection must be credited with similar powers, and Mr. Wallace concludes that Mr. Spencer's insuperable difficulty is "wholly imaginary."

The extract concerning a somewhat similar "class of difficulties," which Mr. Spencer quotes from his Principles of Biology, is faulty in its reasoning,[9] though legitimate in its conclusion concerning the increasing difficulty of evolution in proportion with the increasing number and complexity of faculties to be evolved. But this increasing difficulty of complex evolution is only overcome by some favourably-varying individuals and species—not by all. And as the difficulty increases we find neglect and decay of the less-needed faculties—as with domesticated animals and civilized men, who lose in one direction while they gain in another. The increasing difficulty of complex evolution by natural selection is no proof whatever of use-inheritance[10] except to those who confound difficulty with impossibility.

ALLEGED RUINOUS EFFECTS OF NATURAL SELECTION.

Mr. Spencer further contends that natural selection, by unduly developing specially advantageous modifications without the necessary but complex secondary modifications, would render the constitution of a variety "unworkable" (p. 23). But this seems hardly feasible, seeing that natural selection must continually favour the most workable constitutions, and will only preserve organisms in proportion as they combine general workableness with the special modification. On the other hand, according to Mr. Spencer himself, use-inheritance must often disturb the balance of the constitution. Thus it tends to make the jaws and teeth unworkable through the overcrowding and decay of the teeth—there being, as his illustrations show, no simultaneous or concomitant or proportional variation in relation to altered degree of use or disuse.

ADVERSE CASE OF NEUTER INSECTS.

Mr. Spencer also holds that most mental phenomena, especially where complex or social or moral, can only be explained as arising from use-inheritance, which becomes more and more important as a factor of evolution as we advance from the vegetable world and the lower grades of animal life to the more complex activities, tastes, and habits of the higher organizations (preface, and p. 74). But there happens to be a tolerably clear proof that such changes as the evolution of complicated structures and habits and social instincts can take place independently of use-inheritance. The wonderful instincts of the working bees have apparently been evolved (at least in all their later social complications and developments) without the aid of use-inheritance—nay, in spite of its utmost opposition. Working bees, being infertile "neuters," cannot as a rule transmit their own modifications and habits. They are descended from countless generations of queen bees and drones, whose habits have been widely different from those of the workers, and whose structures are dissimilar in various respects. In many species of ants there are two, and in the leaf-cutting ants of Brazil there are three, kinds of neuters which differ from each other and from their male and female ancestors "to an almost incredible degree."[11]