5 Discussion and conclusions

We have obtained ISOCAM images in the LW3 filter, centered at $\lambda_{\rm eff} = 14.3\,\mu$m, of the $3\hbox{$.\mkern-4mu^\prime$ }2\times 3\hbox{$.\mkern-4mu^\prime$ }2$ fields containing 94 out of the 98 galaxies comprising the complete $60\,\mu$m IRAS deep survey (IDS) sample in the north ecliptic polar region. In addition, we observed a source detected by IRAS at $25\,\mu$m and found to have particularly interesting properties. Charts for all observed fields are given.

We have detected at $\geq 3\sigma$ the likely $14.3\,\mu$m counterparts of 65 IDS sources; 10 more possible $\geq 3\sigma$ identifications are doubtful because of the relatively large difference between nominal ISO and IRAS positions; 6 additional IRAS sources may have been detected at a level between 2 and $3\sigma$. In 4 further cases, we found indications that IRAS fluxes may be affected by confusion, with two or more sources contributing at a comparable level to the observed $60\,\mu$m fluxes.

On the whole, our observations confirm the reality of 69-90% of IDS sources. The 9 IRAS sources with no (even doubtful) $14.3\,\mu$m counterpart at $\geq 2\sigma$ are all relatively faint ( $S_{60~\mu{\rm m}} < 90\,$mJy); they correspond to 17% of IDS sources fainter than this limit (our sample comprises 53 such sources). Hacking & Houck (1987) estimated that the reliability of sources at the survey limit is about 80%; our findings are consistent with their estimate.

Appropriate statistical corrections for the bias affecting faint flux estimates were applied to ISOCAM data.

The areal density of serendipitous sources (stars + galaxies) detected at $\geq 5\sigma$ in our fields, $N(> 3.5\,{\rm mJy}) \simeq 1.3\ 10^{-2}\,\hbox{arcmin}^{-2}$, is in good agreement with results of ISOCAM surveys at the same frequency.

The substantially improved positional accuracy of ISOCAM, compared to IRAS, makes much easier the identification of optical counterparts, resolving ambiguities pointed out by Ashby et al. (1996). Photometric measurements and redshift determinations of candidate identifications of ISOCAM sources are in progress.

Acknowledgements

We gratefully acknowledge the contributions by Matt Ashby, Perry Hacking and L. Danese during the preparation of the ISO proposal. This work was supported in part by ASI, MURST and the TMR Network "Galaxy Formation" under contract No. ERBFMRX -CT96-0086. The ISOCAM data presented in this paper were analyzed with the CAM Interactive Analysis (CIA) software, a joint development by the ESA Astrophysics Division and the ISOCAM Consortium led by the ISOCAM PI, C. Cesarsky, Direction des Sciences de la Matière, C.E.A., France.