There are, therefore, no innate ideas, says Cudworth in explanation of Plato, if these ideas mean formed and completed products of knowledge. All ideas in this sense begin and grow within the range of experience, and the history of their growth or development in literature and art can be at least approximately traced. We can trace, that is, the successions and connexions of the various types of beauty, or goodness: can show how the idea at one time dwelt in one of its aspects, at another in a different one. We can observe the variation, and it may be the progress, in men's conception of God. But it is another matter when we seek to explain these ideas themselves out of other elements, heterogeneous to them. When that question is asked, then with Plato we seem, in the absence of any theory of origins, obliged to own that it is by the Beautiful that beautiful things come to be beautiful. The μετάβασις εἰς ἄλλο γένοs--the crossing of essential boundaries--which Aristotle forbids to science, still raises its eternal barrier in the logical, if it cease to hold good (as has been suggested) in the physical sphere. In the totality which we call the world and experience of reality there are, so to say, ultimate and irreducible provinces. The utmost that philosophy, i. e. science, can do with these is to co-ordinate them,--to show their mutual filiations, adaptations, and harmonies,--to note their inadequacies and discrepancies. They are not all of equal rank, perhaps; they have to yield to each other, it may be in turn: but none of them can be arbitrarily expunged from the totality, and none of them shown to be a mere phase of others. To do that is to strip the universe of its variety and--it may be added--of its beauty and its interest. If it be a false philosophy that does it, there is a good deal of false philosophy abroad. There is a lust of explanation which is never content till it has found an equation for everything, till it has expressed everything in terms of the common-place, till it has emptied everything of all that made it individual and real, and turned it into an abstract, identical (as only abstracts can be) with some other abstract. Such abstractions are of course useful, and therefore need no excuse, when restricted to a special sphere. So long, that is, as we remember that it is an abstraction we are making, and that we are arbitrarily simplifying the real natural problem, no harm is done by these artificial constructions; and they are important steps in a larger process. But what is correct and useful within a range whose limits we can define, becomes dangerous when carried beyond all bounds. Its approximate truth then becomes misleading error.