1 Introduction

Intrinsic variability of central stars of Planetary Nebulae (CSPN) is widely believed to be a rare phenomenon. However, recently a number of such variables have been discovered. Among the hottest CSPN (of the PG 1159 spectral class), several exhibit pulsations with time scales of a few minutes (e.g. see Ciardullo & Bond 1996), while others pulsate with multiple periods of 30 min or longer (e.g. Handler et al. 1998). One at this point enigmatic object is the central star of PRTM 1 (PN G243.8-37.1, Peña & Ruiz 1998).

Furthermore, several cooler CSPN are variables. First, there are the Abell 35-type objects, which have binary nuclei. The optically dominating components are non post-AGB objects whose variations are interpreted to result from rotational modulation of starspots (e.g. see Jasniewicz et al. 1996). Second, photometric and radial velocity variations have been discovered in a number of post-AGB central stars. The best studied of those is HD 35914, the central star of IC 418 (Handler et al. 1997, hereafter Paper I, and references therein). HD 35914 varies on a time scale of about 6.5 hours; the observations can only be explained by either pulsations or variations in the stellar mass loss (spots are ruled out and binarity is improbable). Furthermore, non-periodic variability with a time scale of several days is superposed on the light curves.

To distinguish between the two hypotheses, Handler (1998) carried out a survey for photometric variability among bright Northern Hemisphere CSPN. He observed 25 objects and found that more than 30% of CSPN with effective temperatures of $25\,000 - 50\,000$ K indeed exhibit luminosity variations similar to those of HD 35914.

An outstanding object among these variables is LSIII+51 42, the central star of M 2-54 (we will use the name of the nebula for the central star throughout the remainder of this paper). M 2-54 is also known as PK 104-06 1 or PN G104.8-06.7 (see Acker et al. 1992) and its variability was discovered by Handler (1996, 1998). It is the highest amplitude variable of the latter group found so far: its light variations reach 0.3 mag in Johnson V without removal of the nebular contribution to the measurements, i.e. the intrinsic amplitude is even higher. For this reason and since M 2-54 is significantly cooler than HD 35914, it was considered to be a promising target for a more detailed study. The results of such an effort are reported below. Furthermore, variability of M 2-54 was also detected in HIPPARCOS photometry (ESA 1997). These measurements are analysed as well.