6.5. Source counts at 175 µm
Deep counts with ISOPHOT at 175 µm in the Lockman Hole and in the Marano field ([Kawara et al. 1999]; [Puget et al. 1999]), down to 120 mJy, also give a log N - log S curve with a steep slope and a source density which is higher by a factor of order 10 than extrapolations from IRAS or predictions of no evolution models. For these sources, identifications are uncertain, given the large error boxes. Because of the large negative K correction at this wavelength, it is likely that some of the sources seen are at z in the range 1 to 2. Indeed, using the ISOCAM results, one can show that they cannot be a population of galaxies with average redshift 0.7 and a spectrum like M82 or Arp 220, since in that case ISOCAM would see many more sources in the range > 1 mJy ([Elbaz et al. 2000]). Thus, while a fraction of the sources are in common, on the average ISOCAM and ISOPHOT sample different populations of sources. While ISOCAM pinpoints those of the B and K sources, at z < 1, that harbor activity hidden by dust, ISOPHOT misses many of them but can find more distant far-infrared beacons.
A conclusion on such a broad subject is hard to write, when a large number of our colleagues are devoting most of their time to a finer analysis of the ISO data. Let us instead remark that the areas where new infrared data turn out to be of key importance are growing day by day, strongly supporting future infrared missions.
We are grateful to the organizers for exerting the right pressure resulting in the timely delivery of the manuscript, and to the conference sponsors, the Anglo-American Chairman's fund and SASSOL, for allowing this very interesting meeting to take place.