to recover it. But the greatest difficulty is that when the systematic method of setting out the complex material has been proceeded with and seems about to lead to good results, then at once objections of the most serious kind arise and almost forbid the attempt to make types. With regard to the differences between the sexes, for instance, the most useful theory that has been put forward is the existence of a kind of polarity, two extremes separated by a multitude of intermediate conditions. The characterological differences appear to follow this rule in a fashion not dissimilar to the suggestion of the Pythagorean, Alcmæon of Kroton, and recalling the recent chemical resurrection of Schelling's "Natur-philosophie."
But even if we are able to determine the exact point occupied by an individual on the line between two extremes, and multiply this determination by discovering it for a great many characters, would this complex system of co-ordinate lines really give us a conception of the individual? Would it not be a relapse to the dogmatic scepticism of Mach and Hume, were we to expect that an analysis could be a full description of the human individual? And in a fashion it would be a sort of Weismannistic doctrine of particulate determinants, a mosaic psychology.
This brings us in a new way directly against the old, overripe problem. Is there in a man a single and simple existence, and, if so, in what relation does that stand to the complex psychical phenomena? Has man, indeed, a soul? It is easy to understand why there has never been a science of character. The object of such a science, the character itself, is problematical. The problem of all metaphysics and theories of knowledge, the fundamental problem of psychology, is also the problem of characterology. At the least, characterology will have to take into account the theory of knowledge itself with regard to its postulates, claims, and objects, and will have to attempt to obtain information as to all the differences in the nature of men.
This unlimited science of character will be something more than the "psychology of individual differences," the
