Conclusions

MOND, an alternative theory of gravity (or inertia), has been able to explain the dynamics of massive bodies in terms of visible matter where Newtonian physics requires large amounts of dark matter. However MOND has no relativistic extension and so makes no predictions about gravitational lensing. Here the opposite approach has been taken, using observations of lensing to constrain the form of MONDian light deflection. The cleanest way of doing this is to use galaxy-galaxy lensing data to show that the deflection of photons is simply twice the deflection that would be experienced by a massive particle moving at the speed of light. Thus, a relativistic MONDian theory must have a great deal in common with GR.

With one more part of a complete theory in place a number of new observational tests become available. The clearest relate to ``simple'' lenses: situations in which there is a single, isolated deflector that has no internal structure on the scales of interest. Microlensing within the local group fulfills these criteria, although the typical impact parameters are so small that it is only the low-magnification tails of the light-curves that are of interest. However microlensing by cosmologically distant deflectors offers a good opportunity to confirm or reject MOND. Several programmes to measure low optical depth microlensing of high-redshift quasars are underway (Walker 1999; Tadros, Warren & Hewett 2001), but thus far no events have been observed. The only (probable) example of cosmological microlensing by an isolated deflector observed to date is the serendipitous detection of a peak in the light curve of gamma ray burst 000301C (Garnavich, Loeb & Stanek 2000), but the photometry is not of sufficient quality to differentiate between MOND and GR.

Further observations of galaxy-galaxy lensing offer the most likely tests of MONDian lensing in the near future. Any asymmetry in the distortions (c.f. Natarajan & Refregier 2000) would be at odds with a no-dark matter theory, and there is also the possibility of measuring an outer cut-off in the shear signal. This must eventually come from the influence of secondary deflectors along the line-of-sight, but could also signify a putative return to a Newtonian regime in the ultra-weak acceleration limit. If MOND is not contradicted by any of the above observations, more complex situations should be investigated. Multiple or extended deflectors must also be able to explain observed shear fields and cases of multiple imaging without recourse to invisible mass.

© Copyright Astronomical Society of Australia 1997