Abstract. Infrared to millimetre spectral energy distributions (SEDs) have been obtained for 41 bright ultraluminous infrared galaxies (ULIRGs). The observations were carried out with ISOPHOT between 10 and 200 µm and supplemented for 16 sources with JCMT/SCUBA at 450 and 850 µm and with SEST at 1.3 mm. In addition, seven sources were observed at 1.2 and 2.2 µm with the 2.2 m telescope on Calar Alto. These new SEDs represent the most complete set of infrared photometric templates obtained so far on ULIRGs in the local universe. The SEDs peak at 60-100 µm and show often a quite shallow Rayleigh-Jeans tail. Fits with one single modified blackbody yield a high FIR opacity and small dust emissivity exponent β < 2. However, this concept leads to conflicts with several other observational constraints, like the low PAH extinction or the extended filamentary optical morphology. A more consistent picture is obtained using several dust components with β = 2, low to moderate FIR opacity and cool (50 K > T > 30 K) to cold (30 K > T > 10 K) temperatures. This provides evidence for two dust stages, the cool starburst dominated one and the cold cirrus-like one. The third stage with several hundred Kelvin warm dust is identified in the AGN dominated ULIRGs, showing up as a NIR-MIR powerlaw flux increase. While AGNs and SBs appear indistinguishable at FIR and submm wavelengths, they differ in the NIR-MIR. This suggests that the cool FIR emitting dust is not related to the AGN, and that the AGN only powers the warm and hot dust. In comparison with optical and MIR spectroscopy, a criterion based on the SED shapes and the NIR colours is established to reveal AGNs among ULIRGs. Also the possibility of recognising evolutionary trends among the ULIRGs via the relative amounts of cold, cool and warm dust components is investigated.

Abstract.Mining the ISO data archive we provide the complete ISO view of PG quasars containing 64 infrared spectral energy distributions between 5 and 200 µm. About half of the sample was supplemented by MAMBO and SCUBA (sub-)millimetre data. Since the PG quasars were selected optically, the high infrared detection rate of more than 80% suggests that every quasar possesses luminous to hyperluminous dust emission with dust masses comparable to Seyferts and ultraluminous IR galaxies (ULIRGs). The gas-to-dust mass ratio (of those sources where CO measurements are available in the literature) is consistent with the galactic value providing further evidence for the thermal nature of the IR emission of radio quiet quasars. The SEDs represent templates of unprecedented detail and sensitivity. The power-law like near-to mid-IR SEDs (F ν ∝ ν α ) are smooth up to far-infrared wavelengths, favouring dust heating by the central AGN, and we conclude that, in particular for our hyperluminous quasars at z = 1, starbursts play only a minor role for powering the dust emission, even in the FIR. The IR spectral slopes α 1−10 µm range from -0.9 to -2.2 with a mean of −1.3 ± 0.3. They neither correlate with the optical spectral slope α 0.3−1 µm , nor with the IR luminosity, nor with the FIR/MIR luminosity ratio, nor with inclination-dependent extinction effects in the picture of a dusty torus. We suggest that the diversity of the SEDs reflects largely the evolution of the dust distribution, and we propose a classification of the SED shapes as well as an evolutionary scheme in which this variety can be understood. During the evolution the surrounding dust redistributes, settling more and more into a torus/disk like configuration, while the SEDs show an initial FIR bump, then an increasing MIR emission and a steeper near-to mid-infrared slope, both of which finally also decrease. Strikingly, based on the sensitive ISO data now we do not only see the coarse IR differences between ULIRGs and quasars, but also the details and a possible evolution of the dust distribution and emission even among the optically selected PG sample. Regarding cosmic evolution, our hyperluminous quasars in the "local" universe at z = 1 do not show the hyperluminous (L FIR > ∼ 10 13 L ) starburst activity inferred for z = 4 quasars detected in several (sub-)millimetre surveys. In view of several caveats this difference should be established further, but it already suggests that in the early dense universe stronger merger events led to more powerful starbursts accompanying the quasar phenomenon, while at later cosmic epochs any coeval starbursts obviously do not reach that high power and are outshone by the AGN.

Abstract. In order to test the unified scheme for luminous radio galaxies and quasars we observed 10 galaxy/quasar pairs from the 3CR catalogue with ISOPHOT at infrared wavelengths between 5 and 180 µm. Each pair was selected such that both the 178 MHz luminosity and the redshift match as close as possible between the radio galaxy and the quasar in order to minimize effects of cosmic evolution. 13 of the 20 sources were detected in at least one waveband. 12 sources show clear evidence of a thermal bump at FIR wavelength, while in the remaining 7 sources the upper limits are still compatible with the presence of luminous dust emission. In agreement with the predictions of the unified scheme, the quasars and galaxies in our sample cannot be distinguished by their observed mid-and far-infrared properties. This is in contrast to the findings on the basis of the IRAS scans which indicated that radio galaxies radiate significantly less mid-to far-infrared emission than quasars. However, the IRAS samples are dominated by low-redshift sources (z < 0.5), while our sample contains several of the most luminous radio galaxies at redshift z 1. The latter have already been suspected to contain a hidden quasar for other reasons, e.g. an extended emission line region aligned with the radio axis. From the ratio between FIR luminosity emitted by dust and the radio power at 178 MHz, we conclude that the radio galaxy/quasar unification might be perfectly valid for the most luminous 3C sources at high redshift (z > ∼ 0.8). At lower redshifts (z < 0.5), however, some of the lobe-dominated FRII radio galaxies contain active nuclei which emit less UV-optical continuum than the quasars of similar radio power. As this division is mainly a function of redshift and less one of absolute radio power, we suggest that it is caused by the evolution of the nuclear fueling rate with cosmic epoch. In order to quantify the deviation from the purely aspect-dependent unified scheme at low redshifts a larger fraction of 3C radio galaxies has to be observed at mid-to far-infrared wavelengths with sensitivities which suffice to yield secure detections rather than upper limits.