'Demand and supply are truly correlative and convertible terms. The supply of one set of commodities constitutes the demand for another. Thus, there is a demand for a given quantity of agricultural produce, when a quantity of wrought goods equal thereto in productive cost is offered in exchange for it; and conversely, there is an effectual demand for this quantity of wrought goods, when the supply of agricultural produce which it required the same expense to raise, is presented as its equivalent.'[180] The Ricardian's dodge is obvious: he has chosen to ignore the circulation of money and to pretend that commodities are immediately bought and paid for by commodities. From the conditions of highly developed capitalist production, we are thus suddenly taken to a stage of primitive barter such as we might find still flourishing at present in Central Africa. There is a distant element of truth in this trick since money, in a simple circulation of commodities, plays merely the part of an agent. But of course, it is just the intervention of an agent which separates the two transactions of circulation, sale and purchase, and makes them independent of one another in respect of both time and place. That is why a further purchase need not follow hard upon a sale for one thing; and secondly, sale and purchase are by no means bound up with the same people: in fact, they will involve the same performers only in rare and exceptional cases. MacCulloch, however, makes just this baseless assumption by confronting, as buyer and seller, industry on the one hand and agriculture on the other. The universality of these categories, _qua_ total categories of exchange, obscures the actual splitting up of this social division of labour which results in innumerable private exchange transactions where the sale and purchase of two commodities rarely come to the same thing. MacCulloch's simplified conception of commodity exchange in general which immediately turns the commodity into money and pretends that it can be directly exchanged, makes it impossible to understand the economic significance of money, its historical appearance. Sismondi's answer to this is regrettably clumsy. In order to show that MacCulloch's explanation of commodity exchange has no application for capitalist production, he takes recourse to the Leipsic Book Fair.[181]