The milliJansky 12‐μm population: first follow‐up

We present the first results of our follow‐up programme of optical imaging and spectroscopy of a deep 12‐μm survey conducted with the ISO satellite. We find that the objects are typically of fairly low redshift (z∼0.1–0.2) , but with a tail that extends to high redshifts. The highest redshift object is a previously unknown quasar at z=1.2 . The sample of objects for which spectroscopy has been obtained forms a complete subsample of the original survey with an R ‐band magnitude limit of 19.6. We are thus able to use accessible volume methods to determine the luminosity function. We find that the luminosity function is close to the low‐redshift 12‐μm luminosity function determined by IRAS , but with a possible excess at the highest luminosities, which also correspond to the highest redshifts. This excess is compatible with the rapid luminosity evolution of L ∝ (1+z) 4.5 claimed by Xu and suggested by Elbaz et al. on the basis of number counts. These conclusions remain tentative because of small number statistics at the highest redshifts. Almost all of the galaxies in this survey are emission line objects. We use the spectra to determine emission line diagnostics for the underlying ionization source. We find that the majority of objects have H  ii region‐like spectra (roughly 2/3) but there is a significant fraction (about 1/3) that contain an active galactic nucleus (AGN). This confirms suspicions that deep mid‐IR surveys will be a powerful way to find obscured AGN. Part of the survey area has been imaged deeply by the INT Wide Field Camera. Nearly all 12‐μm sources in this area, down to the 3 σ sensitivity limit of the original ISO survey, have good optical identifications, the faintest being at R=23 . The prospects are thus good for obtaining complete spectroscopy for this sample. With such data we should be able to make a clear measurement of the evolution rate of this population and determine the role of obscured AGN in the mid‐infrared.