f
dz; the two are related by a factor of order unity. If
f is Gaussian then r_0' = 
(2
)
r_0 so the limit translates to r_0' < 40 kpc.
Even for this relatively weak limit on the scale length of the field
the magnetic field in the Coma cluster is still required to have an rms
strength exceeding 1
.No polarized emission has been detected from the halo. Kim et al. (1990) report that the polarization at 20 cm and 1 arcmin resolution varies from less than 1% at the centre to less than 30% in the outer halo. As discussed in Section 2.1, the simplest models predict that the halo polarization should be about the same as the intensity contrast, so that polarization at about the 5% level might be expected. This is not observed and one might be tempted to conclude that the intensity contrast is really only about 1%. Unfortunately this conclusion is not very robust, as both galactic wakes and dynamo fields give low levels of polarization. In the case of the dynamo field, the field is wound into ropes from which the polarization cancels by symmetry---the same would be true of a single turbulent vortex.