On Wed, 21 Nov 2007 12:09:31 -0800, Robert Smts wrote:

I'd like to know whether that is because the
effective mean height of the antenna varies according to the shape, or
because of where the antenna is fed.

Hi Robert,

You can model this, and all variations easily with EZNEC available
online (and is within the limitations of its free use).

As for feeding it at the top, as you illustrate. That would tend to
be unbalanced with the transmission line dropping in close proximity.
Without special attention to choking this line, it will become a
radiator for some of the bands (it'll be a guess). This may not be
all that bad, but you might find the antenna tunes oddly.

You could as easily put up a dipole and skip the complexity (hint,
loops are NOT low noise).

73's
Richard Clark, KB7QHC