This is a near approach to a worse fallacy than the Wages Fund. The archaic reasoning is the more unhappy, because the reasoner proceeds to use it in a good cause. Jean Baptiste Say[661] had taught that all increased or diminished demand depended on increased or diminished supply, and argued thence, with James Mill and Ricardo, the impossibility of general or rather universal gluts. Goods[662] being always meant to be exchanged with goods, one-half will furnish a market for the other half; and thus, as production (which gives the means of buying) is the sole source of demand (so far as demand is effective), an excess in the supply of one article merely proves a deficiency in the supply of another, and is improperly called over-production. Indeed, whereas consumption takes an article away from the market, production brings one into it, and thereby increases, _pro tanto_, the demand by increasing the means of buying. James Mill’s neat demonstration of this doctrine[663] would be quite conclusive if we, first of all, defined demand and supply so as to include each other, and, second, supposed general to mean universal. The reply of Malthus himself is that goods are not always exchanged for goods, but frequently, perhaps most frequently, for labour. Say rejoins that he for his part used a term (“products”) which includes both goods and services, and that the latter are always the real object of an exchange.[664] Malthus makes a better point when he accuses his opponents of treating goods as if they were mathematical symbols, instead of objects of human consumption owing their whole character to human wants.[665] But his case could be made convincing even on his opponents’ premises. Division of labour, all admitted, is limited by the extent of the market;[666] allow that the most satisfactory cure for the limitation is to widen the market, not to lessen the division of the labour—still, _given_ the limitation of the market, the extension of the division of the labour will cause an over-production. All that Malthus maintained was that this might happen in a great many cases as well as in a few; Say went as near as he dared to the assertion that it could not happen at all.