The technique found by Jon Gibbs is the correct way to accurately produce a known short taper. You run the compound slide along the hypotenuse and check the rise at the end.
However you do not need a setting bar between centres and in fact this can introduce errors. In a good chuck or soft jaws a long cutter, silver steel or dowel pin will do.
Also you can reverse the process and eliminate any need for a parallel.
Put the clock in the chuck and a button or small bearing in the toolpost. Then rock the saddle so the clock finds the peak whilst at the bottom and top of the hypotenuse. The difference must agree with the calculated rise.
It was an apprentice piece for me to make mating tapers that "locked" like a morse taper. A taper fit with less than 5 degrees included will lock if the finish is good. Our test piece had to carry two kilos by fit alone 
A toolmaker's lathe with a vernier taper turning attachment was nice too.