In 1924 M. Wolf published a short notice in Astronomische Nachrichten announcing the discovery of a new variable star in the constellation Hercules, which much later received the GCVS name AM Herculis. In the many decades that followed, AM Her remained a Sleeping Beauty. Only in the early days of X-ray astronomy, did AM Her again attract attention, as it turned out to be the optical counter part of the bright X-ray source 3U 1809+50 detected with the Uhuru satellite. Strong flickering observed both at optical and X-ray wavelengths led to the suggestion that AM Her is a cataclysmic variable (CV) of the U Geminorum type. A little later, the detection of linearly and circularly polarised light from AM Her (Tapia 1976, IAU Circ 2984, 2987, 2994) uncovered the so far unique nature of this star: a CV containing a strongly magnetised white dwarf. It is because of the strong polarisation found in AM Her and similar stars that they were dubbed "polars".

The vast majority of CVs are made up of a non-magnetic white dwarf and a main-sequence donor. The material lost from the donor forms an accretion disc around the white dwarf, slowly spiralling inwards, and it is these discs that are the source of the dwarf nova outbursts that attract a great deal of attention by visual observers. In polars, the strong magnetic field of the white dwarf, typically exceeding 1000 Tesla and reaching 20000 Tesla in AR UMa, results in a number of fundamental differences. Firstly, the magnetic field keeps the rotation of the white dwarf synchronised with the orbital period of the binary star. In other words, the same hemisphere of the white dwarf always faces the donor star (in dwarf novae, or the weakly intermediate polars, the period of the white dwarf rotation is much shorter than the orbital period). Secondly, the strong field suppresses the formation of an accretion disc. Instead, the material lost from the donor star locks on to the magnetic field lines, and impacts near the magnetic poles of the white dwarf (see Figure 1). It is in those impact regions that the accreting matter is heated to about 100 million degrees, and produces both the observed strong X-ray emission and the polarised optical and infrared light.

The absence of an accretion disc has a very important consequence: polars do not exhibit dwarf nova outbursts. However, they do exhibit long-term variability, typically with an amplitude of 2 to 3 magnitudes. The best-documented case of this variability is the prototype AM Herculis itself (see Figure 2). The light curve of AM Her shows that the system moves between low states (V=15) and high states (V=12.5-13), but can spend some amount of time at any intermediate level. The timescales on which the system switches from one state into the other varies dramatically, some transitions occur in a few days, i.e. the short high state near JD=2448900, and the short low state near JD=2449100, or gradually over months, as the decline from a high state to a low state at JD=2448400. The only possible explanation for this variability is that the mass transfer from the donor star undergoes large variations. In fact, during the low state, the mass transfer decreases to a trickle, or may even cease totally, and AM Her looks very much like a detached white dwarf plus a main sequence binary. The exact cause of these mass transfer variations are still unknown, but most likely they are due to stellar activity on the donor star, possibly the coming and going of star spots that temporarilycover the inner Lagrange point, i.e. the point of the donor star that is closest to the white dwarf, and through which the donor loses material onto the white dwarf (e.g. Livio & Pringle 1994, ApJ 427, 956).

In non-magnetic CVs such as dwarf novae, and the high-mass transfer novalike variables, the donor stars are likely to undergo similar changes in stellar activity, modulating the rate at which material is dumped towards the white dwarf. However, in the non-magnetic systems, the accretion disc acts as a buffer, largely smoothing out any variation in the rate at which material is supplied to the disc. Only a few dwarf novae are known to exhibit "low states", i.e. phases where they are substantially fainter than their normal quiescent level, examples are HT Cas, and RX And. It is, however, likely that some of the variations in the outburst activity that we see in dwarf novae are related to changes in the mass transfer rate from the donor.

As the hallmark of polars is strong X-ray emission, satellite missions such as ROSAT and EUVE have discovered a large number of new polars, with the total number of known AM Her stars being close to 100. As mentioned, the only polar with a well-documented optical light curve covering several decades is AM Her itself, for a simple reason: it is the only polar which even during the low state is visible to visual observers. For all other polars, knowledge of their long-term variability is very poor, but the little data that is available suggests that the systems display a huge variety of variability. Just to name two examples: EF Eri has been found during all pointed observations by ground or space-based observatories in a high state for nearly twenty years, until, in the mid-nineties, it plunged into a deep low state, in which it remained until early this year! QQ Vul, in contrast, has, to my knowledge, never been found in a deep low state where accretion activity dropped close to zero.

With the advent of sensitive and relatively cheap CCD cameras, a large number of observers are now in the position to detect a few dozen polars even in their low states. However, so far the interest in these stars has remained very feeble. It would be of great scientific interest if more CCD observers would add polars to their regular monitoring targets, so that within a few years high-quality light curves such as that of AM Her would be the norm, rather than the exception, for a substantial number of the known polars. Only at that point would we have a chance to cast some light on the activity cycles of their donor stars, which is an important task for our general understanding not only of polars, but of CVs in general.

Figure 1, Diagram of a polar, showing how the material lost from the donor star locks on to the magnetic field lines, and impacts near the magnetic poles of the white dwarf. Thanks to Gavin Williams who produced this figure, and gave permission for it to be reproduced here.