The VLBI observations show that the position angle of the pc-scale jet aligns well with the ENELR but not with the north-east radio lobe. The position angle connecting the radio core and the north-east radio lobe is 27 
(adopting the error on the lobe-lobe position angle given by Tadhunter et al. [1988] as the error on the core-lobe position angle). Tadhunter et al. [1988] give the relative positions of the optical centroid and the radio lobes. Reynolds et al. [1994] and de Vaucouleurs et al. [1991] show the positions of the radio core and the optical centroid to be coincident ( 
, 
. These radio/optical coordinate offsets are typical for sources at this declination, Jauncey et al. 1989). For comparison, the position angle connecting the radio core and the south-west radio lobe is 193 
2 
. The position angles of the ENELR and the pc-scale radio jet from the core are 42 7 and 44 5 respectively.
The suggestion of interaction between the radio jet and an extra-nuclear cloud is strengthened since the pc-scale jet does not simply align with the north-east radio lobe. The strong alignment of the pc-scale jet and the ENELR naturally identifies the ENELR as the site of interaction. If this is the case, the radio jet must be deflected through an angle of 20 
so as to reach the radio lobe (see Figure 8.2). Depending upon the three dimensional orientation of the radio source, with respect to us, the true deflection angle will not be the same as the projected deflection angle which we observe. Since no information on the orientation is available, possible projection effects will be ignored here.
Figure: Schematic of jet/cloud interaction in PKS 2152-699
It is possible to characterise the deflection of a collimated jet with oblique shocks (see Bicknell [1994], section 4.2, equations and references therein) which can be set up when a low-density jet encounters an interface with a much denser region. The deflection of a relativistic jet by a planar shock which lies obliquely to the direction of flow in the jet has been modelled (Figure 8.3a). In particular, this model has been used with the value for the jet deflection estimated above for PKS 2152-699 to derive some parameters of the relativistic jet. In this case the relativistic jet is the radio jet and the interface with the dense region is the surface of the extra-nuclear cloud.
Figure: Results of oblique shock model for the PKS 2152-699 data
The model constrains the shock strength (defined as the reciprocal of the compression caused by the shock, 1/k) and the Lorentz factors of the jet material in the pre-shock and post-shock regions, 
and 
respectively. The free parameter is the shock obliqueness 
and the jet deflection is 
. The relationship between 
, 
, 
and k is
Following Landau and Lifshitz [1984] and Bicknell [1994], the x and y components of the jet speed with respect to the shock can be calculated. The x direction is perpendicular to the shock and the y direction is parallel to the shock:
where 
and 
. For each value of the free parameter, , a unique value can be found for the compression caused by the shock, 1/k, and consequently unique values of 
, 
, and 
can be calculated. 
and 
can therefore be calculated.
The results of the model can be seen in Figures 8.3b, 8.3c and 8.3d.
The plots show that the required deflection angle of 20 can be achieved for the values of shock strength and Lorentz factor shown by the solid lines as a function of shock obliqueness.
Notably, the required deflection can be achieved over a wide range in obliqueness ( 
30 ) with shock strengths of <10 and the Lorentz factor of a relativistic pre-shock jet ( 
3) . These values are reasonable since higher Lorentz factors have been inferred for the pc-scale radio jets in quasars (eg Vermeulen and Cohen 1994). It may be that in high power radio sources, such as PKS 2152-699, the radio jet is still relativistic on kpc-scales (eg Bridle and Perley 1984).
Finally, these observations have relevance for the ionisation of the extra-nuclear cloud. Since evidence for a jet/cloud interaction has now been strengthened it appears plausible that some excitation of the extra-nuclear cloud may occur in the interaction. However, the alternative hypothesis that postulates ionisation by a photon beam from the nucleus is also supported by the VLBI observations since in all reasonable models the photon beam and the pc-scale radio jet share the same directionality. Hence the extra-nuclear cloud will lie in the path of any existing 'blazar' beam from the nucleus.
It is not unlikely that both mechanisms may play a role in the excitation of the extra-nuclear cloud.