There are two immediate predictions of the intermittency model. First,
Fig. 2b suggests that statistically-complete samples of GPSs and small CSOs
will show them to be very much more radio luminous than suggested by
extrapolation of MSO/FR-II properties. Assuming that the (optically thin)
radio luminosity depends on cocoon pressure in some way, this prediction
should hold for any physically reasonable prescription relating p, 
and the radio luminosity, Q. It must be noted, however, that there are
observational complications involved in testing this prediction. As well
as being comparatively rare, these small sources are often absorbed at
typical radio frequencies. This may be due to free-free absorption in an
inhomogeneous foreground screen or synchrotron self-absorption in the
source itself. This absorption must be corrected for before comparisons of
the type discussed above can be made.
A second prediction is that there should be a large number of medium size
objects (a few hundred to a few thousand parsecs across) which are in a
`coasting' phase and have faded below the flux limits of current radio
surveys. There are several methods that could be employed to search for
such sources. Deep, low-frequency radio maps might reveal the coasting
cocoons of such sources. Alternatively, we might hope to observe the
ISM/ICM shock either through X-ray signatures (using the high spatial
resolution of AXAF) or via the H 
emission that it surely
excites (e.g., see Bicknell & Begelman 1996).