Photometric monitoring (1987 to 1994) of the gravitational lens candidate UM 425.

We present the results of a 7 year long photometric monitoring of two components (A and B) of UM 425, thought to be images, separated by 6.5", of the same z=1.47 quasar. These components have been imaged through an R filter in order to obtain their light curves. The photometry was obtained by simultaneously fitting a stellar two-dimensional profile on each component. The brightest image (component A, m_R_=15.7) shows a slow and smooth increase in brightness of 0.2 magnitude in seven years, while the faintest one (component B, m_R_=20.1) displays an outburst of 0.4 magnitude which lasts approximately two years. The variation of component B may be interpreted in two ways, assuming UM 425 is gravitationally lensed. If it is due to an intrinsic variation of the quasar, we derive a lower limit of 3 years on the time delay from the fact that it is not observed in component A. On the other hand, if it is a microlensing "High Amplification Event", we estimate the size of the source to be ~10^-3^pc, in agreement with standard models of AGNs. These observations are consistent with the gravitational lens interpretation of the object. Furthermore, all the CCD frames obtained under the best seeing conditions have been co-added, in an attempt to detect the deflector. The final R image reveals a rich field of faint galaxies in the magnitude range m_R_~22-24. No obvious deflector, nor any system of arcs or arclets is detected, down to a limiting magnitude of m_R_~24