Mark Elkington wrote:
We left a party the other night and drove away up a hill. The moon was
full and low on the horizon, viewable through the branches of trees
and over the rooftops of houses at the top of the hill.

It looked huge, as big as the distant trees and houses.

As we drove toward the top of the hill, the moon shrank! By the time
we reached the top, the moon was only the size a soccer ball in the
branches of the now close trees.

So there you have it. Closeness to the horizon was not the cause, but
rather relative distances to terrestrial reference objects.

In my opinion, this experiment demonstrates only that you found the
Moon smaller in relation to the nearby branches, not that this is the
only conceivable explanation. (That seemed to be what you meant by
"definitive.")

For instance, any satisfactory theory of the Moon illusion will have
to explain why it seems to work even at sea, when the horizon is often
devoid of *any* points of reference. For me, it even works on my own
planetarium program. The first time I tried out my horizon view, I got
a rather startling impression of increased size at the exact moment my
program drew the horizon line (the very last thing it draws in that
view, by the way).

So although the effect you describe is undoubtedly in play in many
cases, it can't be the whole story.

Brian Tung
The Astronomy Corner at http://astro.isi.edu/
Unofficial C5+ Home Page at http://astro.isi.edu/c5plus/
The PleiadAtlas Home Page at http://astro.isi.edu/pleiadatlas/
My Own Personal FAQ (SAA) at http://astro.isi.edu/reference/faq.txt