Hi David,
A brave attempt - there seem to be as many classification systems for dials as there are diallists. Your scheme is not unreasonable though I think you have missed a few things: 1. The type of "hours" that the dial is intended to indicate is significant eg equal, temporary, Babylonian, Italian etc. 2. The dial plate may be movable, as in most of the card-dials etc. BTW, I note that you classify the shephers' (or cylinder) dial as having a movable gnomon, whereas I would say the gnomon is fixed - it always points south - whilst the dial plate rotates with the date behind it. 3. Gnomons can be multiple segments, as in bifilar dials, not to mention the shadow edge devices. I believe that the professionals writing museum catalogues have their own classification system but I've not found out where it's written up - anyone know? Regards, John Davis > Dear All, > > Some time ago, a list member gave the following definition of a 'singleton > dial' - "it's a horizontal dial with polar-pointing style and showing > standard time on 12 date rings" Also, in the BSS glossary of dials, key > characteristics of dials physical construction are highlighted to explain > the set of dials that belong to one dial type. > > Classification of dials based on physical characteristics is a way of > recognising basic types of dials. > > I am considering the question - > "How many dimensions are necessary to classify the types of sundials?" > > The following 4 dimensions (domains) come to mind quickly: > > *dial principle > shadow > transmission (light beam through a hole) > reflective > diffraction > > *dial plate > -planar > - horizontal > - inclined > - vertical > -cylinder > - polar > - vertical > -hemisphere > -arbitary hollowed out surface > > *gnomon > -point > - fixed > - movable > -segment > - polar > - vertical > - fixed > - movable > > *hour markers > -point > -segment > - curve > -analemma > -figure 8 > - half > - equation of time > -unfolded "analemma" > > Example classifications using the above 4 dimensions > > Horizontal Dial: principle:shadow, dial:planar-horizontal, gnomon: > segment-polar, hour markers: segment > Analemmatic Dial: principle:shadow, dial:planar-horizontal, gnomon: > segment-vertical(movable), hour markers: point > Shepherd's Dial: principle:shadow, dial:cylinder-vertical, gnomon: > point-movable, hour markers: solar altitude curve > > Is 4 dimensions enough to separate all the various types of dial? Almost > certainly not. What is the minimum set of other (semi) orthogonal dimensions > can be added to classify known dial types? > > Has any work been done on this by others? > > In this endeavour, careful attention to separating 'classifying dimensions' > from 'dial attributes'. Attributes 'flow from' key classifying > characteristics of the dial construction. Eg. 'self aligning' is probably > not a classifying dimension, as it arises as a consequence of the nature of > construction, rather than being an observable classification dimension. > > Thanks > > David Pratten > www.sunlitdesign.com > ==================================== > David Pratten > > Phone: 7 (3272) 46 04 20 > Fax (auto forwarded): 1 (603) 676-1231 > Postal: PO Box 21, Almaty 480057, KAZAKHSTAN > >
