I'm looking to prove circular convolution using the Discrete Hartley
Transform.

I saw an old thread on this forum where someone had got it working:

http://www.dsprelated.com/showmessage/130417/1.php

I'm using the same formulations given by wiki:

http://en.wikipedia.org/wiki/Discrete_Hartley_transform
http://en.wikipedia.org/wiki/Hartley_transform

(see the "properties" section in each case)

There's some confusion about what X(-w) or Y(-w) means - is it just a
flipped transform result sequence, or do you flip the sequence before the
transform.

Basically, I have two short real number sequences which I circularly
convolve and correlate.  I then calculate the DHT of these two sequences
and check them against the DHT result using the wiki equations that combine
source DHTs.  I can't get a match with the DHT of either the convolution
result or correlation result...

Can anyone provide a worked example (Matlab / Octave?) so I can check where
I'm going wrong?