knowable, yet not to him, not through any of the forms of knowledge, the enabling attitudes, at his command. And nothing, I say once more, could be simpler or more natural. But, properly and professionally scientific as it may be to give to agnosticism this turn, it is very [p.121] decidedly an excellent example of professional blindness, being a sort of _reductio ad absurdum_, of the scientific point of view, for plainly it treats the unknowable as a matter, first, of knowledge--the scientist's knowledge of its unknowability, and as a matter, second, for knowledge--the knowledge of the creature with the different faculties. Surely such treatment is not honestly agnostic. Science, therefore, if it would be honest as well as scientific, must forget its professionalism and take the negative of the unknowable in another way. In what way? In making reply to this question I must resort to a distinction, which I have frequently found useful, between the dogmatic and the merely instrumental. Thus agnosticism may be dogmatic, as the conventional scientist would hold it, flatly declaring for an unknowable, or it may be instrumental, esteeming the unknowability in things, not merely as relative to the existing conditions of knowledge, but also as a constant demand upon science that it never rest in itself, that it for ever treat its results as only a means to some end. So viewed an instrumental agnosticism is also teleological, but not in any sense of a fixed and static telos. Telic character or purposiveness and fixity are like oil and water. Whatever the traditional theologian may think or say, they simply will not mix.