Leon Walras (1874) was the first author who derives the demand function from an utility func-tion, which should be maximazed under a budgetary restriction. It is only a piece in a complete model of competitive equilibrium model. That is the ‘modern approach’ of the theory of demand. Alfred Marshall followed suit but avoided to settle the issue as a constrained maximization problem. This way, he provided a more handy and realistic tool for solving the question of the adjustment of compe-titive markets. Marshall built its demand theory based on two assumptions: 1) the individual assigns a different utility function to each good s/he consumes; 2) the marginal utility of money is a constant. That makes it easy to build demand functions because the Marshallian utility functions are not ‘perfect’ represen-tations of the individual preferences, in contrast with that of the modern economic theory. Besides the fact that the Marshallian demand function neither depends on income nor on the prices of the other goods, an important difference remains, which is stressed in this paper: the MDF, in contrast with the Walrasian one, reflects the individual marginal valuation, or societal mar¬ginal valua-tion, if speaking in aggregate terms, of every additional unit of good ‘x’. The ordinary demand func-tion (Walrasian) only gives information about demanded quantities along the whole range of prices, provided that all the units are paid at the same price.