itself be a dried-up form of a still earlier peach; and this will be equally true even if all the existing peaches can be shown to be descended from nut-like almonds. For the almond is a plum by family; and all the other plums have juicy fruits; while one of them, the apricot, closely approaches the almond-peach group in most of its characters. Seeing, then, that the almond must almost certainly be descended from juicy fruit-bearing ancestors, nothing is more natural than that under altered circumstances it should revert, per saltum, to a juicy peach. But to suppose that the peach type was originally developed per saltum from an almond is to suppose that it varied at once in several separate ways, all equally and correlatively adapted to a particular mode of dispersion. It is to suppose that accident could do in a minute what we have every reason to believe can only be done by infinitesimal variations and infinite selection.
But if the naturalist can not imagine the production of a peach de novo out of an almond at a single jump, how can he imagine the production of a new thinking element in a human brain? How can he suppose that the accidental introduction of one more little bit of matter into that vast organized labyrinth—a mighty maze, but not without a very definite and regular plan—can have any kind of intelligible relation to the complicated system of cross-connections and superimposed directive departments which make it up? And if it be objected that the view taken above of the constitution of the brain is wooden and mechanical, I would answer that it is certainly absurdly diagrammatic and inadequate, but that it is so far right in that it insists upon making believers in spontaneous variation try to realize their own unthinkable attitude. As to materialism, surely it is more profoundly materialistic to suppose that mere physical causes, operating on the germ, can determine minute physical and material changes in the brain, which will in turn make the individuality what it is to be, than to suppose that all brains are what they are in virtue of antecedent function. The one creed makes the man depend mainly upon the accidents of molecular physics in a colliding germ-cell and sperm-cell; the other creed makes him depend mainly upon the doings and gains of his ancestors, as modified and altered by himself.
And now let us look at this second creed, in order to see how far it surpasses its rival in comprehensibility, concinnity, and power of explaining all the phenomena. If it be true that all nerve-increment and especially all brain-increment is functionally produced, we can easily understand why each new cell or fiber should stand in its true and due relation to all the rest. It will have been evolved in the course of doing its own work, and it will be necessarily adapted to it because the act of working has brought it into being. There will be no doubt whether the new cell governs the peculiar action of the left little finger in performing that amusing conjuring trick, or is, on the contrary, connected with the perception of orange-red, because the cell