[9] For an account of Willis' work on the nervous system, see Sir Victor Horsley's _Fullerian Lectures_, 1891. Willis was the first, or one of the first, to recognise the fact that the cerebral ventricles are nothing more than lymph-cavities. And Gall, a century after Willis, had collected and published, in support of his system of phrenology, many cases and _post-mortem_ examinations showing the differentiation of the work of the brain. Gall is a warning for all time against the dangers of deduction; he had but one idea, and he drove it to death; but the clinical and pathological facts which he amassed, in the hope of establishing a set of doctrines out of all relation to facts, are as true now as ever; and, if he had been content to go the way of induction, and to set himself to the accumulation of facts, he might have become a great physiologist. In his knowledge of the anatomy of the brain, and in the dissection of the brain, he was far ahead of the men of his time; but he followed his own imaginings, and left nothing that could last, except those cases and pathological instances that are buried in the ruins of his system. But there they are, and are still of value. For example, Gall's case of loss of speech, after an injury involving the speech-centres, ought to have commanded the attention of all physiologists: but it came to nothing, because he used it to support his doctrine of organs and bumps, and it shared the fate of that doctrine. Phrenology is gone past recall; it died of that congenital disease, the deductive fallacy; but there was a time when it might have been turned to the service of science. The excitement that Gall aroused by the spread of his ideas shows that some belief in cerebral centres was waiting for development. All men are by nature phrenologists; the commonplace excuses that are offered for lapses of memory, venial offences, and inherited weaknesses, all appeal to the comfortable notion that the offender is not wholly perverted, and that some very small and strictly localised group of cells is at fault. And it is probable that the physiology of the central nervous system, with its present strong tendency toward psychology, will some day be back, at a far higher level, above the point where phrenology went wrong. As Mme. de Staël said, _L'esprit humain fait progrès toujours, mais c'est progrès en spirale_. But the question, whether the general desire for a rational system of psychology will ever commend itself to physiology, belongs to the future. All that is of present concern is the steady, continuous, and successful advance, by the way of induction, and by the help of experiments on animals, toward a clear and accurate statement of the departmental work of the brain.