Evidence Against a Redshift z>6 for the Galaxy STIS 123627+621755

The identification of galaxies at extreme distances provides our most direct information on the earliest phases of galaxy formation. The distance implied by redshifts z > 5 makes this a challenging endeavor for even the most luminous sources; interpretation of low signal-to-noise ratio observations of faint galaxies is difficult and occasional misidentifications will occur. Here we report on STIS 123627+621755, a galaxy with a suggested spectroscopic redshift of z = 6.68 and the most distant spectroscopically-identified object claimed. For the suggested redshift and reported spectral energy distribution, the galaxy should be essentially invisible shortward of 9300 Angstroms because the intervening intergalactic medium absorbs essentially all light energetic enough to ionize neutral hydrogen (lambda < (1 + z) x 912 Angstroms; the redshifted Lyman limit). The galaxy should be relatively bright in the near-infrared with f(nu) ~ 1 microJy. Here we report a detection of the galaxy at 6700 Angstroms, below the Lyman limit for z = 6.68. We also report a non-detection at 1.2 microns, implying the flux has dropped by a factor of three or more between rest 1216 Angstroms and rest 1560 Angstroms for z = 6.68. Our observations are inconsistent with the suggested extreme distance of STIS 123627+621755 and conservatively require z < 6.
Comment: 9 pages, 2 figures; appeared as Nature, 408, 560-562 (30 Nov 2000)