The Aristotelian criticisms which may have affected Epicurus are such as we find in _de Caelo_, 275 b 29 sqq. Aristotle there argues that, as Leukippos and Demokritos made the φύσις of the atoms one, they were bound to give them a single motion. That is just what Epicurus did, but Aristotle’s argument implies that Leukippos and Demokritos did not. Though he gave the atoms weight, Epicurus could not accept Aristotle’s view that some bodies are naturally light. The appearance of lightness is due to ἔκθλιψις, the squeezing out of the smaller atoms by the larger. Footnote 963: In dealing with Empedokles, Aristotle expressly makes this distinction. Cf. _de Caelo_, Β, 13, especially 295 a 32 sqq., where he points out that Empedokles does not account for the weight of bodies on the earth (οὐ γὰρ ἥ γε δίνη πλησιάζει πρὸς ἡμᾶς), nor for the weight of bodies before the vortex arose (πρὶν γενέσθαι τὴν δίνην). Footnote 964: Diog., _loc. cit._ (p. 390). Footnote 965: This seems to be in the main the view of Dyroff, _Demokritstudien_ (1899), pp. 31 sqq., though I should not say that lightness and weight only arose in connexion with the atoms of the _earth_ (p. 35). If we substitute “world” for “earth,” we shall be nearer the truth. There is a striking confirmation of the view just given in the atomist cosmology quoted above.[966] We are told there that the separation of the larger and smaller atoms was due to the fact that they were “no longer able to revolve in equilibrium owing to their number,” which implies that they had previously been in a state of “equilibrium” or “equipoise.” Now the word ἰσορροπία has no necessary implication of weight in Greek. A ῥοπή is a mere leaning or inclination in a certain direction, which may be caused by weight or anything else. The state of ἰσορροπία is therefore that in which the tendency in one direction is exactly equal to the tendency in any other, and such a state is more naturally described as the absence of weight than as the presence of opposite weights neutralising one another. That way of looking at it may be useful from the point of view of later science, but it is not safe to attribute it to the thinkers of the fifth century B.C. Footnote 966: See above, p. 390.