A few weeks ago during the last full moon, I was driving East as the moon was rising. My car has a tinted band at the top of the windshield. If I looked at the moon through the clear part of the windshield, it looked big, per the standard illusion. However, if I looked at it through the tint, it appeared small, the way it does at the zenith.
I haven’t seen this effect referred to in any of the literature, including Cecil’s column (http://www.straightdope.com/classics/a2_110.html). Here’s my initial assessment: Near the horizon, the light from the moon is smeared out and diffused more than at the zenith, and this might contribute to the way the brain calculates the size of the moon (this is NOT the same as the “refraction” theory). However, I can’t remember if the moon appears larger when it’s at or near the zenith and it’s smeared or diffused by clouds. Anybody know? Also, I don’t see why bending over and looking at the moon through your legs (which makes the moon look small) would make smearing and diffusion go away, which is a serious strike against my surmise. I should point out that the tint wasn’t enough to obscure other objects on the horizon, so I don’t think the “visual cue” argument applies (that is, that the tint eliminated any visual cues on the horizon, thereby making it look like the blank sky at the zenith).
This is sort of a question about one of Cecil’s columns. But on the other hand, it’s more of a General Question since you say Cecil didn’t address it. What it is most definitely not is a Great Debate.
I have a hunch that the tint in the windshield is the significant factor. It does, of course, screen out some light. The net effect, I think, would be a subtle reduction in its apparent size. Perhaps, for instance, the very tiny amount of light being reflected from the outer margin of the moons edge, would be filtered out entirely, resulting in a very slightly reduced apparent size. I imagine one could argue this and suggest that the moon would only look dimmer, not smaller, but I think the overall effect on the viewer would be the appearance of diminished size.
The Moon Illusion is a very complicated effect. I did a lot of research on it earlier this year for a chapter in a book I am writing (shameless plug: it’s due out in March).
Anyway, the illusion is not physical, it’s psychological. It has to do with the way our brains interpret sizes of objects coupled with the fact that the sky is not actually perceived as a hemisphere; we see it as an inverted bowl. There are lots of websites on it; try here http://www.lhup.edu/~dsimanek/3d/moonillu.htm for a particularly lucid explanation.
I suspect that when you looked at the Moon through the tinted part of your windhield. the roof of your car blocked enough of the sky to lose the effect of it being bowl-shaped. When you looked at the Moon through the lower half of your windshield, you could see more of the sky. I’m guessing, but it sounds logical.
Simple way to test this. Walk outside tonight and put on some sunglasses. I’m not sure if the tints are similar, but i imagine you should at least be able to see the effect of tint on the illusion from that.