On the universality of microlensing in quadruple gravitational lenses

It is well known that multiply imaged quasars are likely to be affected by microlensing. Quadruply imaged systems are especially useful laboratories for studying microlensing because their macrolens models are relatively well constrained. We begin with analytical results for a simple family of galaxy models. These results can be used to estimate the magnifications and time delays for the quadruple systems. We compute expected brightness fluctuations due to microlensing in several such systems for a range of source sizes. Among these we treat for the first time the limiting case of a completely unresolved source. For the case of MG 0414+0534 we interpret the discrepant optical-to-radio flux ratios for the different components as the result of microlensing. On the assumption that the lensing galaxy is roughly isothermal this gives a 95% confidence upper limit on the optical source size of 10(exp 16) cm x (mean value of M)/0.1 solar mass(exp 1/2) where (M) is the average stellar mass of the lensing galaxy. For more centrally concentrated macromodels the upper limit is greater. Sufficiently long strings of photometric observations can in principle constrain the degree of concentration.