NavList:

Did you compare the time from level to toppled to the time from start-up to level? David, because the bearing lies above the centre of mass of the rotor, it cannot topple. I cannot determine at what speed of rotation it ceases to maintain the direction of its axis for practical reasons. My rev counter senses reflected light from a white patch. The collimating lens also reflects light, but not in a predictable way, and the top half of the rotor is shiny. I have to wrap black tape right around it and add a white patch to get any sort of stable reading. Thus I can measure speed of rotation or behaviour of the rotor as revealed by the movement of the graticule, but not both at the same time.

Setting my collimator horizontal to within ten seconds, I was able to meaure the index error at 18 minutes. The Admiralty Laboratory in 1945 did not meaure this so that the mean error of 10.9 minutes for fifty observations of the sun was of no significance. However, the standard deviation was nearly nine arc-minutes for a hand-held instrument, implying that about two thirds of the observations had errors spread over 18arc- minutes. Perhaps one day I will do a similar exercise, but will probably stop at thirty observations...

Bill Morris

Pukenui

New Zealand