Hello Frank,
Thanks to you (and John Carmichael, and others) for the very fascinating discussion on nodus design.
You wrote...
BALL NODI There is something else your experiments showed up that I hadn't really appreciated before... If you are going to use a ball nodus, then the supporting stick should go RIGHT THROUGH THE BALL so that it sticks out a little bit, perhaps half a ball diameter. This thought struck me when I tried estimating the centres of the shadows in printouts of your photographs. The point on the shadow where the stick meets the ball is not matched by a corresponding point on the far side. Once again, the lack of symmetry makes estimation a little harder.
In October, 1999 I received the following private note from William S. Maddux: *** begin quote *** Date: Fri, 15 Oct 1999 14:18:51 -0400 From: "Wm. S. Maddux" <[EMAIL PROTECTED]> Subject: Re: Vertical Wall Declination program Sender: "Wm. S. Maddux" <[EMAIL PROTECTED]> To: Mac Oglesby <[EMAIL PROTECTED]> Mac,
Now a question - what treatment of the gnomon's tip would be likely to give a shadow most easily and accurately measured?
I'm a bug on symmetry, so I might try something like the attached GIF. The Idea is that the center- lines of all three elements cross at the same point, which is the "virtual tip." You construct that point in the shadow by symmetrical interpolation, which is quite intuitive. The only thing to watch out for is a case where the thicker post casts its shadow on top of that of one of the cross bars. I suppose it might be good if all three rods were the same diameter, but that would be harder to make. The cross bars need not be at perfect right angles, but they might as well be so far as ease of construction is concerned. Bill *** end quote *** He sent a small, rough drawing, which I attach.Bill passed away in December, 2004, but I don't believe he would mind my sharing this with the Sundial Mailing List.
Best wishes, Mac Oglesby
MASSE-GN.GIF
Description: GIF image
--------------------------------------------------- https://lists.uni-koeln.de/mailman/listinfo/sundial
