----- Original Message ----- From: "Mr. D. Hunt" <[EMAIL PROTECTED]>
<snip>. > If anyone wants to visit one of our layouts in a public location, and then > 'reverse engineer' it - I am perfectly happy for them to do so. </snip> Oooh, there's nothing like a challenge! Unfortunately, in spite of their apparent common-ness, I have never come across any of these dials in Nova Scotia. Luckily, some pictures can be found on the web, at Mr. Hunt's website. Take a look at the example of one from a school in the USA. The direct URL is http://www.argonet.co.uk/education/sunclocks/pics/fs-015.htm It looks like a close copy of the design in Waugh. By my estimate, from the spacing of the hour marks, the smiling girl is sitting right over the East marker. Connect her bum to the West marker and you have the EW axis of the dial. Note that the declination scale is made of bricks which I assume to be of equal width and equally spaced. I reckon there are 11 to the North and 14 to the South of the EW axis. So either the dial is badly laid out, or the upper part is deliberately shortened - which is consistent with having a smaller inner ellipse for the summer hours. By the way, 11/14 is close to 3/4, which is my estimate of the ratio of the sizes of the two rings, based on the relative radii of matching pairs of hour points. Paint the inner summer ring in a nice warm red, and likewise for the corresponding part of the declination scale. Paint the outer winter ring in a icy cool blue, along with its part of the scale, and you have a intuitive way to know which ring to read from. Only thing is, in the example they used black and blue so its not particularly intuitive, but at least they thought to distinguish them. The problem with this "re-sizing" approach is that the start and finish of Daylight saving time don't coincide with the equinox. During October, the user would be on the 'big' declination scale but reading the time from the inner ring. How would the design cater for that? My guess is that it doesn't. I suspect the resultant error would be relatively small and would be swamped by the error from user not standing in the exact place for the day, EoT etc., and the claim is only for 15 minute accuracy anyway. Anyone care to improve on my analysis? The bit I don't have an answer for is "why bother?". Why not simply have a small hour mark, double labelled, and read one label set in summer and the other in winter based on colour coding, or typeface or any other distinguishing property you can think of? Steve -
