That's an interesting point about the relative position of the DEMS. The test I did was with two "adjacent" DEM's which are 1:24K UTM USGS ones, and as such the edges are not square which may explain some things possibly! It did look like they were importing as individual objects. I think the larger-scale DEM's may be unprojected so they could provide a test too. It should be noted that the USGS DEM's (1:24000) issued prior to June 2001 had positional and other errors present, so make sure you get the latest ones from: http://www.gisdatadepot.com