In a constant-returns-to-scale-technology sans joint products, relative goods and factor prices must in general depend competitively on the two legs of demand composition and cost-technology. Sraffa's 1926 classic article, besides validly contending the need for imperfect-competition paradigms in increasing-returns situations, wrongly emphasized the primacy of the special one-leg case of constant costs. Here it is shown that Sraffa's leitmotiv involved overstress on this nongeneric case: his 1925 Italian paper, his late-1920s exploration of input-output, his editing of the classical writings, and his 1960 classic book involve the same bias. Copyright 1991 by Royal Economic Society.