Historically, our conception of metaphysical method as fundamentally analytical and directed to the detection and removal of internal contradictions in the categories of ordinary thought, is perhaps nearer to the view of Herbart than to that of any other great philosopher of the past. In our insistence upon the non-empirical and, in a sense, _a priori_ character of Metaphysics, we are again, of course, largely in agreement with the position of Kant. There is, however, a most important difference between our own and the Kantian conception of the _a priori_ upon which it is essential to insist. A-priority, as we have used the term, stands merely for a peculiarity of the _method_ of Metaphysics; by an _a priori_ method we understood one which is confined to the internal analysis of a datum and independent of external reference to outside facts. With Kant the _a priori_ is a name for certain forms of perception and thought which, because revealed by analysis as present in every experience, are supposed to be given _independently of all experience_ whatsoever, and so come to be identified by him as “the work of the mind,” in opposition to the empirical factor in experience, which is held to be the product of an external system of “things-in-themselves.” Hence Kant’s whole discussion of the _a priori_ is vitiated by a constant confusion between what is metaphysically _necessary_ (_i.e._ implied in the existence of knowledge) and what is psychologically _primitive_. This confusion, perplexing enough in Kant, reaches a climax in the works of writers like Mr. Spencer, who appear to think that the whole question of the presence of a non-empirical factor in knowledge can be decided by an appeal to genetic Psychology. It is clear that, from our point of view, the identification of the _a priori_ with the “work of the mind” would involve a metaphysical theory as to the constitution of experience which we are not entitled to adopt without proof.[27]