Carl Friedrich Gauss and the 'useless' FFT in 1805

Question

This is a history question, so you need to know something about math history to answer it. There's a rumour that says that Carl Friedrich Gauss knew the FFT in 1805, but he thought it was useless, because the speedup wasn't very big for small numbers of N. And at that time there were no supercomputers, so you couldn't calculate large N's anyway.

But what's is the best evidence you can find that support this rumour?

Answer 1

This is a history question, so you need to know something about math history to answer it.

Oh yeah? The first Google hit I get for "Gauss fast Fourier transform" is this paper, which says that Gauss wrote down something that looks like the FFT in an unpublished paper that appears in his collected works.