a minimal order which cannot be broken up. Each monad, if it is to be anything at all, must be a continuing finite representation of the universe, and to be that it must have a body, that is to say, it must have other monads in a permanent relation of mutual correspondence with it. And if you said to Leibniz, 'But surely any physical body can be broken up, and this must mean the dissolution of the organic relation between its monadical constituents,' he would take refuge in the infinitesimal. The wonders revealed by that new miracle, the microscope, suggested what the intrinsic divisibility of space itself suggests—whatever organization is broken up, there will still be a minute organization within each of the fragments which remains unbroken—and so ad infinitum. You will never come down to loose monads, monads out of all organization. You will never disembody the monads, and so remove their representative power; you will only reduce their bodies and so impoverish their representative power. In this sense no animal dies and no animal is generated. Death is the reduction and generation the enrichment of some existing monad's body; and, by being that, is the enrichment or the reduction of the monad's mental life.

'But,' our common sense protests, 'it is too great a strain on our credulity to make the real nature of things so utterly different from what sense and science make of them. If the real universe is what you say it is, why do our minds represent it to us as they do?' The philosopher's answer is, 'Because they represent it. According to the truth of things, each monad is simply its own mental life, its own world-view, its own thoughts and desires. To know things as they are would be simultaneously to live over, as though from

within and by a miracle of sympathy, the biographies of an infinite number of distinct monads. This is absolutely impossible. Our senses represent the coexistent families of monads in the gross, and therefore conventionally; what is in fact the mutual representation of monads in ordered systems, is represented as the mechanical interaction of spatially extended and material parts.' This does not mean that science is overthrown. The physical world-view is in terms of the convention of representation, but it is not, for all that, illusory. It can, ideally, be made as true as it is capable of being. There is no reason whatever for confusing the 'well-grounded seemings' of the apparent physical world with the fantastic seemings of dream and hallucination.

So far the argument seems to draw whatever cogency it has from the simplicity and naturalness of the notion of representation. The nature of idea, it is assumed, is to represent plurality in a unified view. If idea did not represent, it would not be idea. And since there is idea (for our minds at least exist and are made up of idea) there is representation. It belongs to idea to represent, and since the whole world has now been interpreted as a system of mutually representing ideations, or ideators, it might seem that all their mutual relations are perfectly natural, a harmony of agreement which could not be other than it is. But if so, why does Leibniz keep saying that the harmony is pre-established, by special and infinitely elaborate divine decrees?

Leibniz himself says that the very nature of representation excludes interaction. By representing environment a mind does not do anything to environment, that is plain. But it is no less plain that environment does nothing to it, either. The act of representing is simply the act of the mind; it represents in view of environment, of course, but not under the causal influence of environment. Representation is a business carried on by the mind on its own account, and in virtue of its innate power to represent.

Very well; but does this consideration really drive us into theology? Is not Leibniz the victim of a familiar fallacy, that of incompletely stated alternatives? 'Either finite beings interact or else they do not directly condition one another. Monads do not interact, therefore they do not directly condition one another. How then explain the actual conformity of their mutual representation, without recourse to divine fore-ordaining?' It seems sufficient to introduce a further alternative in the first line of the argument,

and we are rid of the theology. Things may condition the action of a further thing, without acting upon it. It acts of itself, but it acts in view of what they are. We are tempted to conclude that Leibniz has introduced the Deus ex machina with the fatal facility of his age. 'Where a little further meditation on the characters in the play would furnish a natural dénouement, he swings divine intervention on to the scene by wires from the ceiling. It is easy for us to reconstruct for him the end of the piece without recourse to stage-machines.'

Is it? No, I fear it is not. There is really no avoiding the pre-established harmony. And so we shall discover, if we pursue our train of reflexion a little further. It is natural, we were saying, than an idea should represent an environment; indeed, it is the representation of one. Given no environment to represent, it would be empty, a mere capacity for representation. Then every idea or ideator, taken merely in itself, is an empty capacity. But of what is the environment of each made up? According to the Leibnitian theory, of further ideas or ideators: of empty capacities, therefore. Then no idea will either be anything in itself, or find anything in its neighbours to represent. An unhappy predicament, like that of a literary clique in which all the members are adepts at discussing one another's ideas—only that unfortunately none of them are provided with any; or like the shaky economics of the fabled Irish village where they all lived by taking in one another's washing.

It is useless, then, to conceive representations as simply coming into existence in response to environment, and modelling themselves on environment. They must all mutually reflect environment or they would not be representations; but they must also exist as themselves and in their own right or there would be no environment for them mutually to represent. Since the world is infinitely various, each representor must have its own distinct character or nature, as our minds have: that is to say, it must represent in its own individual way; and all these endlessly various representations must be so constituted as to form a mutually reflecting harmony. Considered as a representation, each monadical existence simply reflects the universe after its own manner. But considered as something to be represented by the others, it is a self-existent mental life, or world of ideas. Now when we are considering the fact of representation, that which is to be represented comes first and the representation follows upon it.

Thus in considering the Leibnitian universe, we must begin with the monads as self-existent mental lives, or worlds of ideas; their representation of one another comes second. Nothing surely, then, but omnipotent creative wisdom could have pre-established between so many distinct given mental worlds that harmony which constitutes their mutual representation.

Our common-sense pluralistic thinking escapes from the need of the pre-established harmony by distinguishing what we are from what we do. Let the world be made up of a plurality of agents in a 'loose' order, with room to manoeuvre and to adjust themselves to one another. Then, by good luck or good management, through friction and disaster, by trial and error, by accident or invention, they may work out for themselves a harmony of action. There is no need for divine preordaining here. But on Leibniz's view what the monads do is to represent, and what they are is