The comment endorses and reinforces Daniel W. Graham’s highly original and attractive proposal that early Greek cosmology develops in two stages. In what Graham calls the “meteorological stage” of the sixth century BCE, celestial objects are explained as formations either from fire or from watery exhalations in a roughly planar model of the cosmos. In the “lithic stage” of the mid- and late fifth century introduced by Anaxagoras, the model is that of a central earth around which solid stone-like celestial objects revolve held aloft in a vortex. The change to the lithic stage comes about, according to Graham, as the implications of Parmenides’ epoch-making discovery that the moon is illuminated by the sun (heliophotism) come to be understood and are then theoretically exploited. The present comment also proposes that the false explanations of lunar phases and lunar and solar eclipses in the meteorological-stage cosmologies, respectively, of Xenophanes and of Heraclitus may have played a helpful heuristic role in the theoretical breakthrough to heliophotism.