We present submillimetre (SCUBA) photometry of a sample of eight high-redshift (2.5 < z < 3.5) radio quasars from two redshift surveys: the TEXOX-1000, or TOOT, survey; and the 7C Redshift Survey (7CRS). Unlike the powerful high-redshift radio sources observed previously in the submillimetre, these radio sources are typical of those dominating the radio luminosity density of the population. We detect just two of the TOOT/7CRS targets at 850 microns, and one of these detections is probably due to synchrotron emission rather than dust. The population represented by the other six objects is detected in a statistical sense with their average 850-micron flux density implying that they are similar to low-redshift, far-infrared-luminous quasars undergoing at most moderate (< 200 solar masses per year) starbursts. By considering all the SCUBA data available for radio sources, we conclude that positive correlations between rest-frame far-infrared luminosity, 151-MHz luminosity and redshift, although likely to be present, are hard to interpret because of subtle selection and classification biases, small number statistics and uncertainties concerning synchrotron contamination and K-correction. We argue that there is not yet any compelling evidence for significant differences in the submillimetre properties of radio-loud and radio-quiet quasars at high redshift.

We present maps, source lists, and number counts from the largest, unbiassed, extragalactic sub-mm survey so far undertaken with the SCUBA camera on the JCMT. Our maps cover 260 sq. arcmin, to a noise level S(850)=2.5 mJy/beam. We have reduced the data using both SURF, and our own pipeline which produces zero-footprint maps and noise images. The uncorrelated noise maps produced by the latter approach have allowed application of a maximum-likelihood method to measure the statistical significance of each peak, leading to properly quantified flux-density errors for all potential sources. We detect 19 sources with S/N > 4, 38 with S/N > 3.5, and 72 with S/N > 3. To assess completeness and the impact of source confusion we have applied our source extraction algorithm to a series of simulated images. The result is a new estimate of the sub-mm source counts in the flux-density range S(850)=5-15mJy, which we compare with other estimates, and with model predictions. Our estimate of the cumulative source count at S(850) > 8 mJy is 320 (+80,-100) per square degree. Assuming that the majority of sources have z > 1.5, the co-moving number density of high-z galaxies forming stars at a rate >1000 solar masses per year is 10^-5 per Mpc^3, with only a weak dependence on the precise redshift distribution. This number density corresponds to that of massive ellipticals with L > 3-4 L* at low redshift and is also the same as the co-moving number density of comparably massive, passively-evolving objects in the redshift band 1<z<2 inferred from recent surveys of extremely red objects. Thus the bright sub-mm sources can plausibly account for the formation of all present-day massive ellipticals. Improved z constraints, and a proper measurement of sub-mm clustering can refine or refute this picture.

We use a simple model of spheroid formation to explore the relationship between the creation of stars and dust in a massive proto galaxy and the growth of its central black hole. This model predicts that sub-mm luminosity peaks after only ~ 0.2 Gyr. However, without a very massive seed black hole, Eddington-limited growth means that a black hole mass of 10^9 solar masses, and hence very luminous AGN activity, cannot be produced until > 0.5 Gyr after the formation of the first massive stars in the halo. The model thus predicts a time-lag between the peak of sub-mm luminosity and AGN luminosity in a massive proto-elliptical of a few times 10^8 years. For a formation redshift z = 5, this means that powerful AGN activity is delayed until z = 3.5, by which time star formation in the host is 90% complete, and sub-mm luminosity has declined to ~ 25% of its peak value. This provides a natural explanation for why successful sub-mm detections of luminous radio galaxies are largely confined to z > 2.5. Conversely the model also predicts that while all high-redshift luminous sub-mm-selected sources should contain an active (and growing) black hole, the typical luminosity of the AGN in such objects is ~1000 times smaller than that of the most powerful AGN. This is consistent with the almost complete failure to detect sub-mm selected galaxies with existing X-ray surveys. Finally the model yields a black-hole:spheroid mass ratio which evolves rapidly in the first Gyr, but asymptotes to ~ 0.001-0.003 in agreement with results at low redshift. This ratio arises not because the AGN terminates star formation, but because fueling of the massive black hole is linked to the total mass of gas available for star formation in the host.

The calibration of ground-based submillimetre observations has always been a difficult process. We discuss how to overcome the limitations imposed by the submillimetre atmosphere. Novel ways to improve line-of-sight opacity estimates are presented, resulting in tight relations between opacities at different wavelengths. The submillimetre camera SCUBA, mounted on the JCMT, is the first large-scale submillimetre array, and as such is ideal for combatting the effects of the atmosphere. For example, we find that the off-source pixels are crucial for removing sky-noise. Benefitting from several years of SCUBA operation, a database of deep SCUBA observations has been constructed to better understand the nature of sky-noise and the effects of the atmosphere on instrument sensitivity. This has revealed several results. Firstly, there is evidence for positive correlations between sky-noise and seeing and sky-noise and sky opacity. Furthermore, 850-micron and 450-micron sky-noise are clearly correlated, suggesting that 450-micron data may be used to correct 850-micron observations for sky-noise. Perhaps most important of all: if off-source bolometers are used for sky-noise removal, there is no correlation between instrument sensitivity and chop throw, for chop throws out to 180 arcsec. Understanding the effects of submillimetre seeing is also important, and we find that the JCMT beam is not significantly broadened by seeing, nor is there an obvious correlation between seeing and pointing excursions.

The Submillimetre Common User Bolometer Array (SCUBA) instrument has been operating on the James Clerk Maxwell Telescope (JCMT) since 1997. The data archive is now sufficiently large that it can be used to investigate instrumental properties and the variability of astronomical sources. This paper describes the automated calibration and reduction scheme used to process the archive data with particular emphasis on `jiggle-map' observations of compact sources. We demonstrate the validity of our automated approach at both 850- and 450-microns and apply it to several of the JCMT secondary flux calibrators. We determine light curves for the variable sources IRC+10216 and OH231.8. This automation is made possible by using the ORAC-DR data reduction pipeline, a flexible and extensible data reduction pipeline that is used on UKIRT and the JCMT.

We present submillimetre photometry of 11 3CR and 6CE radio quasars of similar 151-MHz radio luminosity and redshifts to the radio galaxies studied at 1.3<z<2 by Archibald et al. (2001). We detect all bar one quasar at 850 microns, and five quasars are confirmed as dusty hyperluminous objects by detections at shorter wavelengths. Our observations reveal a clear difference between the submillimetre luminosity distributions of the radio quasars and a matched sample of radio galaxies: the quasars are ~5-times brighter than the radio galaxies and >~2-times brighter accounting for possible synchrotron contamination. This difference is in quantitative agreement with a receding-torus unified scheme for radio sources in which the torus opening angle depends on quasar optical luminosity, provided that there is a close relationship between the optical and submillimetre luminosities. The implication is that quasar-heated dust dominates the submillimetre emission for powerful radio quasars at z~1.5. We find a significant anti-correlation between submillimetre/far-infrared luminosity L_FIR and radio source age in the sense that hyperluminous quasars tend to be associated with young (<10^7 yr) sources. This suggests that the processes controlling L_FIR are synchronized with the jet-triggering event and that L_FIR declines on a ~10^7 yr timescale. There is evidence for a small amount of obscuration in the hyperluminous quasars from reddening and associated or broad absorption lines. We speculate that shortly after jet triggering, dust is more widely distributed than at later times and is heated by the quasar nucleus and/or a short-lived synchronized starburst.

We present the results of the first major systematic submillimetre survey of radio galaxies spanning the redshift range 1 < z < 5. The primary aim of this work is to elucidate the star-formation history of this sub-class of elliptical galaxies by tracing the cosmological evolution of dust mass. Using SCUBA on the JCMT we have obtained 850-micron photometry of 47 radio galaxies to a consistent rms depth of 1 mJy, and have detected dust emission in 14 cases. The radio galaxy targets have been selected from a series of low-frequency radio surveys of increasing depth (3CRR, 6CE, etc), in order to allow us to separate the effects of increasing redshift and increasing radio power on submillimetre luminosity. Although the dynamic range of our study is inevitably small, we find clear evidence that the typical submillimetre luminosity (and hence dust mass) of a powerful radio galaxy is a strongly increasing function of redshift; the detection rate rises from 15 per cent at z < 2.5 to 75 per cent at z > 2.5, and the average submillimetre luminosity rises as (1+z)^3 out to z~4. Moreover our extensive sample allows us to argue that this behaviour is not driven by underlying correlations with other radio galaxy properties such as radio power, radio spectral index, or radio source size/age. Although radio selection may introduce other more subtle biases, the redshift distribution of our detected objects is in fact consistent with the most recent estimates of the redshift distribution of comparably bright submillimetre sources discovered in blank field surveys. The evolution of submillimetre luminosity found here for radio galaxies may thus be representative of massive ellipticals in general.

A systematic survey measuring the submillimetre continuum luminosity in radio galaxies between redshifts z ~ 0.1 - 5 is currently in progress. The first results from observations with the bolometer array SCUBA on the JCMT suggest a trend of increasing submillimetre luminosity with redshift out to z ~ 4. Assuming the continuum emission at 850um is dominated by thermal radiation from dust heated by young, massive stars, the straightforward interpretation of these data implies that the host galaxies of powerful radio sources, presumably ellipticals or their progenitors, exhibit increased star formation activity with increasing redshift. However a severe bias may distort the true picture since only a subset of 30 galaxies have been observed, which represent the most luminous radio sources (P_(151 MHz) > 10^27.5 W Hz^-1 sr^-1), whilst the complete sample covers ~4 decades in radio power (10^25 - 10^29 W Hz^-1 sr^-1). This on-going observational programme continues to improve coverage of the radio power-redshift plane in an attempt to remove this potential bias towards higher AGN activity and hence understand more fully the relationship between high-z AGN, their level of starformation, and the evolutionary status of their host galaxies.

We present the results of new rest-frame far-IR observations of the z = 4.25 radio galaxy, 8C1435+635, which not only confirm that it contains an enormous quantity of dust (as first inferred from its mm-wave detection by Ivison 1995) but also allow the first meaningful constraints to be placed on the mass of this dust and associated gas. The new measurements consist of: (i) clear detections of sub-mm continuum emission at 450 and 850um obtained with the new sub-mm bolometer array, SCUBA, on the JCMT, (ii) continuum upper limits at 350, 750 and 175um obtained with SCUBA and the PHT far-IR camera aboard ISO, and (iii) a sensitive upper limit on the CO(4-3) line flux obtained with the IRAM 30-m MRT. The resulting rest-frame 33- to 238-um continuum coverage allows us to deduce that 2E8 Msun of dust at a temperature of 40 +/- 5 K is responsible for the observed mm/sub-mm emission. Using our CO upper limit, which constrains M(H2)/M(dust) to <950, we go on to calculate robust limits on the total gas reserves (H2 + HI) which are thereby constrained to between 4E10 and 1.2E12 Msun. The sub-mm properties of 8C1435+635 are thus strikingly similar to those of the z = 3.80 radio galaxy, 4C41.17, the only other high-redshift galaxy detected to date at sub-mm wavelengths whose properties appear not to be exaggerated by gravitational lensing (Dunlop et al. 1994; Hughes, Dunlop & Rawlings 1997). The inferred gas masses of both objects are sufficiently large to suggest that the formative starbursts of massive elliptical galaxies are still in progress at around z = 4. Observations of complete samples of radio galaxies spanning a range of redshifts and radio luminosities will be required to determine if the spectacular far-IR properties of 8C1435+635 and 4C41.17 are primarily due to their extreme redshifts or their extreme radio luminosities.