Mac Oglesby has pointed out to me that I was wrong to say that the shadow of a sphere is a circle. It is, as he has pointed out to me, an ellipse. And my statement that you can find the hour lines by drawing lines parallel to the Italian hour lines at a distance equal to the sphere's radius is wrong as I forgot to mention a step you need to perform to get the correct distances.
The easiest way to lay out the lines for a sundial with a spherical gnomon is to use computer graphics. Without a computer you could follow the same method on pencil and paper but it would be tedious. Draw a circle equal in diameter to the sphere. Draw extended radii every 7.5 degrees for the Italian/Babylonian hours and draw tangents to the circle where the radii cut it for the conventional hour and half hour lines. Decide which radius you want to represent the meridian (12 o'clock). Now, before printing, stretch the picture by a factor equal to 1/sin(latitude) so that the circle becomes an ellipse, with the meridian as its major axis. All the lines remain straight, but their positions and angles have been changed by the stretching. The lines can be transferred to the sundial from the printed page easily enough as they are all straight. Note that the centre of the pattern of lines is NOT where the spherical gnomon is to be placed. It must be placed a distance equal to its radius x cos(latitude) along the meridian line away from the equator. The 23 radial lines to the east of the meridian line should be numbered 1 to 23 hours since dawn. Similarly for the lines to the west: they represent 1 to 23 hours to sunset. The tangential lines are the conventional hour and half hour lines. They are in pairs: you can read either side of the sphere's shadow to read the time. If anyone wishes I can send a JPEG of the unstretched image off list. All you need to do is scale it according to the size of your sphere, then stretch it according to your latitude before printing. I've labelled a few of the lines with the northern hemisphere labels to get you going. Chris Lusby Taylor --------------------------------------------------- https://lists.uni-koeln.de/mailman/listinfo/sundial
