1.     Martin Garner (personal communication, Fall, 1997) pointed out to me that Bruno Ernst (1969) wrote about an eighteenth century artist, Giovanni Piranesi, who predated Duchamp and the Penroses with his intuitive application of conflicting depth clues. Ernst, however, states that the first "conscious construction" of an "impossible object" was in 1934 by Oscar Reutersvärd who did not realize the scientific importance of what he had done with his systematic working out of various impossible cubes, configurations and "forks" until he read the Penroses' 1958 paper. Ernst quotes Roger Penrose's statement that his "own interest in impossible objects stems from the year 1954" (p. 71), when he discoverd Escher. But, as Ernst (1992) correctly points out, Escher had not done anything we "would call an impossible object" until after 1958 (pp. 71, 72)! Penrose goes on to say that he did not learn of Oscar Reutersvärd until 1984 (p. 72).
        Evidence indicates that Duchamp's "Impossible Bed" was not an intuitive appreciation of an "impossible object" but a specific exploration of one case within his general category of perceptual illusions (for example, the convex/concave illusion in both Duchamp's "fig leaf" sculpture and magazine cover; and his cheese cover for the "first Papers of Surrealism") (Schwarz, 1997). Ernst (1992, pp.28-29) discusses the convex/concave illusion. Moreover, I recently discovered that Duchamp drew both an "impossible object" and a convex/concave illusion in his 1925 chess poster with its cascading and impossible and ambiguous cubes (Schwarz, 1997). With both his impossible cubes and bed, Duchamp beats out Oscar Reutersvärd as the first artist to make a series of consciously constructed optical illusion/impossible figures in a conscious series (see Ernst, 1992, pp. 69, 70).