[CraigS]

Hi Rob;

I can't agree with some bits and my reasons are below. .. Perhaps we just have different wording interpretations ?
The different principles I'm basing mine on, are as follows.
(If I've misinterpreted these, please feel free to let me know):

Originally Posted by Robh

Craig,
Now consider several distant galaxies lensed by a massive foreground object. They appear distorted but aligned in an annular region surrounding the massive foreground object. The light from these is seen from the Earth only because it is sent in our direction.

The light, clearly, is radiated spherically from the source. We are on the lens corrected light path, otherwise would see nothing of the background image.

The shape of the distortion in the image we see, is a function of:

i) the mass distribution of the lens and, (of course);
ii) the shape of the object we're looking at and;
iii) the position of the background object relative to the position of the intervening lens.

Interestingly, there are different effects moving from the centre of mass, (of the lens), outwards, varying from rings, to arcs to multiple images.

Time delays also appear between multiple images, (in the case of multiple images), from the same source, because of optical path length variations and the Shapiro effect. If the source spectrum varies over time, variations will be observed in the multiple images, also.

A typical grav lens, can appear to magnify parts of, and distort very distant sources, moreso than sources closer to the lens.

Originally Posted by Robh

You have to remember that lensed galaxies are not single points and the light from all the constituent points of each galaxy must be sent our way. If we were any closer or further away, this annular region could most likely bypass us and not be seen. In this sense, we are at the focal point of observation.
We can say that the light from the galactic images, converge towards the Earth from around the site of the massive body.

I would say the only reason we see the object in the first place, is because Earth coincidentally, lies along the 'focal line' of the lens (as I mentioned before). There is no 'focal point'. 'Focal point' in optics has a very specific meaning. It is where light rays converge. There is no convergence. We merely lie along the path of propagation of the 'lensed' light. There is no 'convergence' and thus, there is no 'divergence'.

This is actually the distinguishing feature of a gravitational lens, and differentiates it from an optical lens.

(Actually, unlike an optical lens, the maximum bending occurs closest to the centre of the grav lens, and the minimum bending occurs furthest from the the centre of mass of the lens, also).

Originally Posted by Robh

Conversely, if anti-matter has the effect of repulsing "normal" matter, then the light would diverge away from the observer's line of sight.

As Steven says (and I agree with him):

Originally Posted by sjastro

Because of CPT symmetry, an observer would see the same lensing effect irrespective if the lens is made from matter or antimatter.
The null geodesic of the photon (or antiphoton) must remain the same for the laws of physics to remain invariant otherwise the symmetry is violated.

Originally Posted by Robh

If geodesic paths can be thought of in geometrical terms, I see no reason that these paths can't be thought of as convergent or divergent from a distant point.

See Steven's words above.

Cheers