I just went out and did my best to determine the actual horizon for the setting point for the Mir observation yesterday. I measure it to be approximately -0.4 deg. This was done using a level and then coming home and recreating the angle observed, then measuring and calculating by simple trig. Estimated angle of the cloud bank above the horizon (disappearance of Mir) is 0.2 degree. Thus (ignoring the errors) the observed elevation of Mir was about -0.2 degree. This means that is was below my true horizon (elevation = 0 deg) Assuming a refraction angle of 0.5 degree yields an actual elevation of -0.2 minus 0.5 equals -0.7 deg. This is consistent with the values calculated by STSPLUS and SKYMAP using post-burn elements. I am using the term "elevation" to represent vertical angle from the horizon, where the local true horizon is 0 degrees and the zenith (overhead) is 90 deg. To make this obs, I travelled to a point about 1 mile SE of my normal location to improve the horizon. It may just be a matter of semantics, but I am assuming that the elevation that would be appropriate for use in deriving element sets would be the -0.7 deg value. This obs was relatively easy due to the favorable horizon and the clear air at this altitude. There are places in the mountains to the west where gaining altitude (height above sea level) would improve the horizon (make the value more negative). Pikes Peak is approximately 7000' higher than my location. I do not know what improvement that would make. A salient point to this obs, beyond just a personal challenge, is that people should not discount low elevation passes for seeing interesting objects, especially Mir and the Shuttle. I blew off the last shuttle mission due to its 28.5 deg inclination, yet a local observer (David Brierly), did see it easily on at least one occasion. Ron Lee