The situation with the multiple GSC entries is probably more complicated, but provides some insight about using the GSC, and about being skeptical toward any specific entry. As often happens with such pairs (or higher-order multiples), the GSC records the two stars separately _and_ as a combined entry with a position of the photo-center of the images and a much-too-bright magnitude. This is likely to be the case here. The northernmost entry refers to the companion star, evidently somewhat offset toward TU Aql. The very bright object is the combined entry; the southernmost object is probably TU Aql, but again ofsett somewhat to the north because of the nearby companion. USNO-A1.0 gives a better position for the companion, but because the star is grossly overexposed on the deep sky survey plates used for A1.0, the position may not be utterly reliable. However, it seems to match very well the DSS image from SkyView at large image-scale. Another position from the turn of the century is available for the companion from the Algiers zone of the Astrographic Catalogue, recently re-reduced to the J2000 system: TU Aql comp: 19 27 36.48 +2 03 26.4 (equinox 2000, epoch 1909.6) Since the AC positions are usually good to a few tenths of an arcsec, this suggests some proper motion in the companion star, and if that's the case, then the two stars are quite close together now---and the more northern GSC position may not be in error after all! Clearly, as Kato-san suggests, some current astrometry is in order to sort this out. Gordon?! \Brian