the last paragraph of the book all this was changed, and the variations were ascribed to the conditions of existence, and to use and disuse, but a concluding paragraph cannot be allowed to override a whole book throughout which the variations have been kept to hand as accidental. Mr. Romanes is perfectly correct when he says {232a} that "natural selection" (meaning the Charles-Darwinian natural selection) "trusts to the chapter of accidents in the matter of variation" this is all that Mr. Darwin can tell us; whether they come from directly transforming agents or no he neither knows nor says. Those who accept Lamarck will know that the agencies are not, as a rule, directly transforming, but the followers of Mr. Darwin cannot.

"But showing themselves," continues Professor Ray Lankester, "at each new act of reproduction, as part of the phenomena of heredity such minute 'sports' or 'variations' are due to constitutional disturbance" (No doubt. The difference, however, between Mr. Darwin and Lamarck consists in the fact that Lamarck believes he knows what it is that so disturbs the constitution as generally to induce variation, whereas Mr. Darwin says he does not know), "and appear not in individuals subjected to new conditions" (What organism can pass through life without being subjected to more or less new conditions? What life is ever the exact fac-simile of another? And in a matter of such extreme delicacy as the adjustment of psychical and physical relations, who can say how small a disturbance of established equilibrium may not involve how great a rearrangement?), "but in the offspring of all, though more freely in the offspring of those subjected to special causes of constitutional disturbance. Mr. Darwin has further proved that these slight variations can be transmitted and intensified by selective breeding."

Mr. Darwin did, indeed, follow Buffon and Lamarck in at once turning to animals and plants under domestication in order to bring the plasticity of organic forms more easily home to his readers, but the fact that variations can be transmitted and intensified by selective breeding had been so well established and was so widely known long before Mr. Darwin was born, that he can no more be said to have proved it than Newton can be said to have proved the revolution of the earth on its own axis. Every breeder throughout the world had known it for centuries. I believe even Virgil knew it.

"They have," continues Professor Ray Lankester, "in reference to breeding, a remarkably tenacious, persistent character, as might be expected from their origin in connection with the reproductive process."

The variations do not normally "originate in connection with the reproductive process," though it is during this process that they receive organic expression. They originate mainly, so far as anything originates anywhere, in the life of the parent or parents. Without going so far as to say that no variation can arise in connection with the reproductive system--for, doubtless, striking and successful sports do occasionally so arise--it is more probable that the majority originate earlier. Professor Ray Lankester proceeds:-

"On the other hand, mutilations and other effects of directly transforming agents are rarely, if ever, transmitted." Professor Ray Lankester ought to know the facts better than to say that the effects of mutilation are rarely, if ever, transmitted. The rule is, that they will not be transmitted unless they have been followed by disease, but that where disease has supervened they not uncommonly descend to offspring. {234a} I know Brown-Sequard considered it to be the morbid state of the nervous system consequent upon the mutilation that is transmitted, rather than the immediate effects of the mutilation, but this distinction is somewhat finely drawn.

When Professor Ray Lankester talks about the "other effects of directly transforming agents" being rarely transmitted, he should