I'm probably missing the point here, but this does seem to be overcomplicating the issue. Take a look at an active site - http://www.sandgatecemetery.org.au/ - which represents a large cemetery in Australia. Each grave has a lat and a long (a combination of hand-checking and Google Maps, I believe) and a QR code that makes it easy to transport data. That's a modern thing, in that it is there and can be measured in today's terms. If you are discussing pllaces whose 'today' location is not known or cannot be determined, then I can appreciate the problem. For my use, my own material is at a lesser resolution, but works the same way.

pblair wrote: > > I'm probably missing the point here, but this does seem to be overcomplicating the issue. > > Take a look at an active site - http://www.sandgatecemetery.org.au/ - which represents a large cemetery in Australia. Each grave has a lat and a long (a combination of hand-checking and Google Maps, I believe) and a QR code that makes it easy to transport data. > > That's a modern thing, in that it is there and can be measured in today's terms. If you are discussing pllaces whose 'today' location is not known or cannot be determined, then I can appreciate the problem. > > For my use, my own material is at a lesser resolution, but works the same way. My whole training & experience tells me to to simply accept a number whether it be a grid reference or anything else. My first job as a research assistant was to be asked to proof-read a paper based on someone's PhD thesis. I think these were final page-proofs, this being in the days of hot metal type. I quickly realised that all the heights had been incorrectly converted from imperial to metric. This had been missed by the author's supervisor (my new boss), the external examiner, the author when he came to write up for publication, the prestigious journal's editorial staff and referees and the proof-reading of the galleys. As the subject of the paper was relative sea level changes this was not a minor matter. You get the message: numbers don't get taken as read, especially in a field to which I'm new. And I'm new dealing with GIS. I'm looking to take historic maps and see how man-made features such as roads, fields and buildings have changed. I'm hoping that this will eventually help sort out genealogical problems. Advice on GIS forums seems to be that the best approach to georeferencing maps in the UK is to take my base map, which is a concatenation of several whole and part 6" to the mile C19th sheets, convert it to the modern OS grid and then georeference the tithe maps, which might well not be as well drawn, to this. The first step is far from straightforward. Ideally one would take the known coordinates of an old map and, using a knowledge of the parameters of the original & OS grid projections, re-project the old map onto the grid. The coordinates are not problem; the map is framed with a scale of lat/long on each side. The projection is, however unknown. The alternative is to identify a spread of features which have remained unchanged in a century & a half, obtain accurate and sufficiently precise OS coordinates for them and feed them into the georeferencing S/W. And then check the results - but against what? In the course of this I was using the Google maps hybrid layer reprojected to the OS geometry as a source of coordinates. As an onging check I also overlaid the OS vector maps onto the Google map. (The vector map is about 1/6th of the scale of the historic map so not directly useable). Everything appeared to be going well until I noticed something wrong with this village http://maps.google.co.uk/maps?q=Briestfield&hl=en&ll=53.652215,-1.649972&spn=0.001993,0.006507&sll=53.559099,-1.817636&sspn=0.011981,0.035963&oq=briestfield&t=h&hnear=Briestfield,+West+Yorkshire,+United+Kingdom&z=18 I'd planned to use the corner of the pub at the top of the picture as a control point. There's a house at the bottom of the picture which has two garden ponds. One of them is shown on the surface water vector map except that it was in the wrong position, offset to the NW by about the diameter of the pond. When I started looking more critically I noticed that in a lot of places the relationship between features on the vector & Google versions changes with scale and even with panning. For instance the vector line can start on one side of the road and be moved over to the other by either zooming or panning and the vector of the river beside the road can be moved on top of it. I have two versions of the same map, both capable of being represented by numbers but subtly different. So here I am, more than 40 years on, still checking numbers and finding them wrong. So, the graves in the graveyard you cite can be individually measured "in today's terms" as you put it. But have you checked they're right? To go back to your first point, yes, it does overcomplicate things. It overcomplicates things to the point where I can't really get a map I wholly trust at the scale I'd like. Which is a scale adequate to answer questions such as was the route of Thick Hollins Lane here:http://maps.google.co.uk/maps?q=Meltham&hl=en&ll=53.578796,-1.831144&spn=0.000998,0.003254&sll=53.652215,-1.649972&sspn=0.001993,0.006507&t=h&hnear=Meltham,+West+Yorkshire,+United+Kingdom&z=19 where it is now or was it where the belt of trees is now? Were the thorns on the field side of the belt (flowering white) originally a roadside hedge? -- Ian The Hotmail address is my spam-bin. Real mail address is iang at austonley org uk