It's a well-known phenomenon that the moon on the horizon looks larger than the moon high in the sky. This has turned out to be an extraordinarily difficult scientific problem to solve, and is still unsolved today. We have eliminated some solutions. We know it's not actually larger or closer (it stays the same size, of course, and I believe it's actually farther away at the horizon). I believe Aristotle and Ptolemy both proposed the idea that it was an atmospheric phenomenon, in which the air magnifies it, but promising an idea as it was, it has been rigorously ruled out for quite some time. In the medieval period arguments began to be made that it was actually a psychological effect, with the most popular being that it just looked bigger because intervening objects made it look farther away.
Berkeley argued for a mix of the atmospheric theory with the psychological theory in which the moon looked fainter on the horizon, and therefore farther away. His position was that the eye receives fewer rays throught he atmosphere and thus less light from the moon itself. I notice, however, that Malebranche had already rejected this position, although he puts it in terms of the horizontal sun rather than the horizontal moon. In Dialogue XII of Dialogues on Metaphysics and on Religion he recounts an experiment with smoked glass in which he looked at the sun with and without the glass (don't try this at home!) and says that it can't be that the glass is letting in fewer rays, because the meridional sun, high in the sky, looks the same size whether the glass is used or not. Malebranche takes his experiments with the smoked glass (which obscured intervening bodies) to indicate that the intervening-body theory, in which intervening bodies make the sun look farther away, was the right one. He also suggests that this effect is made stronger because "the sky appears like a flattened spheroid" (JS 220).
