The Ricardian dynamics study the evolution of distribution when demand increases. The successive marginal methods in agriculture are determined by means of a rule which has never been stated explicitly by Ricardo's commentators. But the rule also allows us to point at two limits of Ricardo's construction, first in his attempt to get rid of rent in the analysis of distribution, second in the working of the dynamics themselves. Similarly, the identification of a productivity and a profitability criterion is at the basis of Sraffa's mistake in his reconstruction of the theory of rent, whereas post-Sraffian formalizations have abandoned Ricardo's dynamic approach.