Radio galaxies are known to exhibit a variety of scales in their structure. First we have the nuclear sources, which so far have not been completely resolved even on the scale of 1/10 milliarcsecond with VLBI observing methods (e.g. Preuss, 1981). Then we have the ‘jets’ (which at some stage break up into ‘blobs’) which are considered to transfer energy from the ‘nuclear engine’ to the outer heads. The latter appear to be the sites of transfer of the collimated jet energy into a diffuse emission region. Despite their usually low brightness these diffuse emission regions dominate the internal energy content in particles and fields, even for the collimated doubles. Note that only 1% of the total energy in Cyg A is in the hot spots (Perola, 1981).

We know almost nothing about the evolutionary tracks of extragalactic radio sources but those tracks are, however, strongly constrained by the distribution of sources in the radio luminosity, P, overall physical size, D, diagram. This is the radioastronomer's H-R diagram, an analogy which two lines of algebra shows is exact. Fig. 1 is the P-D diagram for the 3CR 166 source sample of Jenkins et al. (1977) with later additions. Most of the sources are identified and have known redshifts. It is a flux density limited sample so that the numbers at any P are weighted relative to the true space density by P3/2 because of the differing volumes of space sampled. The important feature of the diagram is the lack of sources greater than 1 Mpc in size. Because of doubts about the completeness of the sample in this region, we have made searches in the 6C 151MHz survey for sources having surface brightnesses lying between the two lines of slope 2 on the right of Fig.1. The numbers found to a limiting flux density of 1–2 Jy suggest that there is no serious underestimate of the numbers in the 166 source sample.

The power for a strong extragalactic radio source comes from deep within the nucleus, but the extended radio structure is clearly related to the larger-scale properties of the galaxy in which it lives. Very large sources are found in elliptical rather than spiral galaxies, and big galaxies have stronger radio sources than small ones. The narrow jets mapped in weaker radio galaxies do not expand with a constant opening angle, but become better focussed along their length, suggesting that they are confined by an external pressure. This paper discusses how the rotation of a radio galaxy affects the distribution of gas within it, and consequently the radio structure in elliptical and Seyfert galaxies. A model is proposed which leads to a specific prediction, relating the width of radio jets to the rotation speed of the galaxy in which they lie.

Recent observations of Cygnus A with the Hat Creek interferometer at 86.2 GHz limit the spectral curvature of the hotspots and show that the diffuse lobe emission has a spectral index of about 1.5. The central component is, at most, weakly variable and its spectrum shows a distinct break to a spectral index near 0.65 at about 20 GHz.

High dynamic range maps of Cygnus A at 2.7 and 5 GHz have been used to investigate the variation of spectral index over the extended parts of the source. Although both components show a steepening of spectral index away from the hotspots there is a marked asymmetry between the two components. This is interpreted as being due to a higher value of magnetic field in the Sf component.

We shall tell you briefly about the main observational program connected with radio galaxies. Different theories of radio galaxies predict different types of spectral index variations across the main body of the source. One would expect that the best solution of the problem is the construction of two-dimensional maps at a number of frequencies. However, we suggest that in some cases (i.e., for standard well-aligned structures) one-dimensional images with filled aperture may be much more accurate in determination of the variations of the spectral index along the major axes of radio galaxies. We now have 47 one-dimensional multifrequency images of all sources brighter than 1 Jy at centimeter wavelengths in the declination range −43° −+53° resolvable with our beam. Up to 7 frequencies were used (1.35, 2.08, 3.9, 6.5, 8.2, 13, and 31 cm). Cyg A is the best example showing structures of different scale: nuclear sources, bridges, main bubbles, and hot spots.

A recent analysis (Slee, 1981) of the spectra of extragalactic sources derived by Slee et al. (1981) has revealed a relationship between spectral index and redshift in the sense that the spectra of identified radio galaxies steepen with increasing redshift; this behaviour is consistent with an effect first reported by Kellermann et al. (1969) that the average spectral index of identified radio galaxies is lower than that of sources in empty optical fields, which are now generally regarded as distant radio galaxies (rather than QSOs). However, because of the selection effects present in the sample, Slee was not able to decide whether the basic correlation is between spectral index and redshift (implying an evolutionary origin), between spectral index and radio power, or between spectral index and linear dimension.

Most quasars with extended radio structures have the classical linear double structure (Class II) which is also characteristic of the most luminous radio galaxies. A very small fraction of quasars do, however, show deviations from the classical double structure which are similar to the distortions observed in some low luminosity radio galaxies, e.g., bent double sources. Among radio galaxies, structural distortions seem to be associated with membership in a cluster of galaxies. Indeed, Scott and Hintzen, 1978, have suggested that such distortions might be indicators of cluster membership for quasars as well. To test this hypothesis and examine the physics of distorted quasars, we have mapped the quasars 3C 270.1 and 3C 275.1 in both I and P using the VLA A-Array at 20 cm and 6 cm. Our high resolution (0.1 arc sec at 6 cm) and high dynamic range (200:1) observations demonstrate that the displaced components reported previously for these sources are indeed connected with the quasars themselves and, therefore, the distortions are not simply projection effects. We also report the discovery of a very narrow northwesterly pointing jet in the quasar 3C 275.1.

Multifrequency observations with large single dishes are the ideal tool to examine the variation of the shape of the radio spectrum across extended extragalactic radio sources. Three complex low luminosity sources with angular extent 20′ < θ < 30′ have been mapped with the 100-m telescope of the MPIfR at frequencies 2.7, 4.9 and 10.7 GHz (HPBW = 4!4, 2!6 and 1!2 resp.). To extend the frequency range we used published low frequency maps for the spectral comparison, too. Thus at least four maps with angular resolution ≦ 4!4 were available for each source. All maps were (if necessary) corrected for sidelobes, then cleaned from obvious background sources and finally smoothed to the same beam of 4!4 HPBW. To look for spectral curvature a spectrum of the form In S = a1 + a2 In ν + a3 (In ν)2 was fitted to the brightness data for each sampling point of the map. We chose two parameters to characterize the spectral shape. The first is the mean spectral index, , defined as the slope of the fit curve at the geometric mean of the lowest and highest observing frequency. As a measure of the spectral curvature we derived the change of spectral index, Δα, along the fit curve between the lowest and highest observing frequency, i.e. Δα positive for a concave (= flattening) spectrum and vice versa. In the following the results are briefly summarized (see Andernach, 1981, for details).

Observations are reported of the remarkable object PKS 0812+020, the first quasar found to have optical emission in one of its radio lobes. A distinct radio jet is also seen, and there is radio and optical evidence that the quasar is near the center of a galaxy cluster.

It has been generally assumed that wide-angle tailed (WAT) sources like 3C465 are formed in a manner similar to that of the more strongly bent U-shaped sources such as NGC 1265, i.e., by ram pressure arising from galaxy motion through a dense intracluster medium (ICM). The WAT sources were thought to be less strongly bent because of the smaller ratio of tail plasma flow momentum flux to galaxy velocity. However, as noted recently by Burns (1981), there is a serious discrepancy between the ram pressure model requirements for bending WATs and the dynamics of the associated radio galaxy. To bend the tails, we calculate that the galaxy must typically move at velocities of 0.7–1×103 km s−1 for distances comparable to the length of the radio tails (∼200 kpc for 3C465). This implied galaxy motion is inconsistent with the nature of the massive cD galaxies generally associated with WATs. Cluster galaxy velocity data, X-ray observations, and recent models suggest that these giant galaxies are nearly at rest at the bottoms of cluster potential wells, at most moving ∼200 km s−1 in an oscillatory motion of small amplitude (<0.3 of a core radius, Malumuth, 1981, private communication). Thus it appears that some other mechanism is responsible for bending WAT sources.

Examination of the structures of extragalactic radio sources shows a distinct asymmetry in addition to the more general symmetries which are well known. The most likely explanation for the observed asymmetries is that ejection from the active nucleus occurs in only one direction at a time. This direction then switches back and forth to form the large scale double structures. Implications of this picture for the nuclear engine and the radio source environment are discussed.

The fact that radio jets, which are often one-sided, are nevertheless associated with extended lobes of nearly equal luminosity indicates when analyzed in detail (1) that the energy supply by the nuclear engine to the lobes is strongly time-dependent. Time dependent ejection from the nucleus of a parent galaxy produces low density channels which are ploughed in the background medium by the passage of intermittently ejected radio emitting plasmons (2), (3), (4) and (5). An analysis of the dynamics of radio emitting plasmons in such channels leads to unique morphological features consistent with observations of both narrow jets (resulting from splashback or reflections from the channel ends) and conical lobes (resulting from slower hydrodynamic deceleration).

This is a brief report on some of the work in progress at Cambridge, with emphasis on studies of extragalactic radio sources using the One-mile and 5-km telescopes; together with the 6C survey and a new instrument now being commisioned at 151 MHz, we have a frequency range of 100:1 available for high-resolution mapping. The 6C survey has resulted in the discovery of a number of giant radio galaxies, most notably NGC 6251. At 151 MHz, the beautiful jet is not prominent; we have mapped it with the One-mile and 5-km instruments and a paper (by Saunders et al.) will shortly appear in Monthly Notices. Dr Willis will speak later on the structure of this source.

We report here some preliminary results of a multi-wavelength study of a complete sample of radio galaxies. The sample is comprised of 93 radio sources from the Parkes 11 cm catalog which are identified with galaxies of 17th magnitude or brighter in the declination zone −17° to −40°. Our objective is to cross-correlate the radio, infrared, optical, and other properties of a properly defined sample of radio galaxies.

We have compared measurements of several hundred 3C and 4C radio sources at large redshifts to investigate how radio-source structure changes over a factor of 5–10 in luminosity. Our results show that for z ≳ 0.6:

1. (i) most sources (both 3C and 4C) have hotspots about 3.5 kpc in size (Ho = 50 km s−1 Mpc−1, Ω = 1);

2. (ii) lower-luminosity sources (bottom of 4C) have less-extended outer lobes.