Using a new technique, we have determined a value for the constant of proportionality between submillimetre emission and dust mass, the dust mass–absorption coefficient (κd) at 850μm. Our method has an advantage over previous methods in that we avoid assumptions about the properties of dust in the interstellar medium. Our only assumption is that the fraction of metals incorporated in the dust (ɛ) in galaxies is a universal constant. To implement our method, we require objects that have submillimetre and far‐infrared flux measurements as well as gas mass and metallicity estimates. We present data for all the galaxies with suitable measurements, including new submillimetre maps for five galaxies. We find κ850= 0.07 ± 0.02 m2 kg−1. We have also been able to use our sample to investigate our assumption that ɛ is a universal constant. We find no evidence that ɛ is different for dwarf and giant galaxies, and show that the scatter in ɛ from galaxy to galaxy is apparently quite small.

We present optical, near-infrared (IR) and radio observations of the 3-hour field of the Canada-UK Deep Submillimetre Survey (CUDSS). Of the 27 submillimetre sources in the field, nine have secure identifications with either a radio source or a near-IR source. We show that the percentage of sources with secure identifications in the CUDSS is consistent with that found for the bright '8-mJy' submillimetre survey, once allowance is made for the different submillimetre and radio flux limits. Of the 14 secure identifications in the two CUDSS fields, eight are very red objects (VROs) or extremely red objects (EROs), five have colours typical of normal galaxies and one is a radio source that has not yet been detected at optical/near-IR wavelengths. 11 of the identifications have optical/near-IR structures which are either disturbed or have some peculiarity that suggests that the host galaxy is part of an interacting system. One difference between the CUDSS results and the results from the 8-mJy survey is the large number of lowredshift objects in the CUDSS. We give several arguments why these are genuine low-redshift submillimetre sources rather than being gravitational lenses that are gravitationally amplifying a high-z submillimetre source. We construct a K-z diagram for various classes of high-redshift galaxy and show that the SCUBA galaxies are on average less luminous than classical radio galaxies, but are very similar in both their optical/IR luminosities and their colours to the host galaxies of the radio sources detected in µJy radio surveys.

A B S T R A C TWe present SCUBA 850-mm, JCMT COðJ ¼ 2 ! 1Þ, B-band imaging and VLA H I observations of the NGC 7465/4/3 group of galaxies. The 850-mm emission associated with NGC 7465 extends to at least , 2R 25 and is well correlated with the H I. We investigate a range of possible mechanisms by which dust beyond R 25 may be heated to give the observed extended submillimetre emission. By modelling the dust heating by stars in two extreme geometries, we fail to find any reasonable star formation scenario that is consistent with both the 850-mm and optical data. Furthermore, we do not detect any COðJ ¼ 2 ! 1Þ emission coincident with the extended dust and atomic gas as would be expected if significant star formation were occurring. We show that shock-heating of dust via cloud -cloud collisions in the stripped interstellar medium of NGC 7465 could be sufficient to explain the extended 850-mm emission and lack of optical emission in the stripped gas, and suggest that cloudcloud collisions may be an important dust heating mechanism in gas-rich systems.

Galaxy mergers play a crucial role in the formation of massive galaxies and the buildup of their bulges. An important aspect of the merging process is the in-spiral of the supermassive black-holes (SMBHs) to the centre of the merger remnant and the eventual formation of a SMBH binary. If both the SMBHs are accreting they will form a dual or binary active galactic nucleus (DAGN). The final merger remnant is usually very bright and shows enhanced star formation. In this paper we summarise the current sample of DAGN from previous studies and describe methods that can be used to identify strong DAGN candidates from optical and spectroscopic surveys. These methods depend on the Doppler separation of the double peaked AGN emission lines, the nuclear velocity dispersion of the galaxies and their optical/UV colours. We describe two high resolution, radio observations of DAGN candidates that have been selected based on their double peaked optical emission lines (DPAGN). We also examine whether DAGN host galaxies have higher star formation rates (SFRs) compared to merging galaxies that do not appear to have DAGN. We find that the SFR is not higher for DAGN host galaxies. This suggests that the SFRs in DAGN host galaxies is due to the merging process itself and not related to the presence of two AGN in the system.