The radio colour-colour diagram is clearly a useful tool for the analysis of radio synchrotron spectra. Compared with the normal analysis of spectra, it has several advantages. It shows differences between different models far more clearly than the original spectra. It is ideally suited to the common situation of having a small number of observing frequencies. The source can be considered globally, allowing all the data to be analysed together. Much better sampling of the spectrum is obtained because different parts of the source sample different parts of the spectrum. These last two points require the additional assumption that the spectral shape is independent of position, which would need to be verified.

The colour-colour diagrams presented in this paper for realistic spectra agree well with the colour-colour plot for Cygnus A presented by Katz-Stone et al. (1993). This shows two things. First, that the model spectra presented here and in Paper I do indeed give a good description of the radio spectra in real sources. And secondly, that the radio spectrum of Cygnus A is described by a realistic ageing model. Katz-Stone et al. argued that their inability to fit the Cygnus A spectrum with standard models raised doubts about the ageing interpretation of radio spectra. While there may still be some problems with low-frequency curvature, the success of the models presented here strongly supports the standard ageing interpretation.

The model colour-colour plots presented in this paper are all of a very similar shape. We therefore expect there not to be too much variation in observed colour-colour diagrams. Two systematic effects have been identified. Increasing the importance of inverse Compton losses or diffusion moves the colour-colour locus down in the diagram, while changing the distribution of magnetic field strength can move the ``parallel'' section of the colour-colour locus relative to the diagonal.

The shape of the colour-colour locus depends on the magnetic field model. There is an indication that the field in Cygnus A is skewed, with rare regions of strong field on a predominantly low-field background. With no diffusion, the field strength in Cygnus A is about 8.7microGauss. A more likely interpretation is that this represents a lower bound on the field strength, and that electrons diffuse slowly between regions of different field strength, with a mean free path of 10 pc or less.

On to:

Up to: ___________________________________
Peter Tribble,