1 Introduction

The picture that we formed of the Galaxy suffered many modifications since the first analysis of the count of stars in the eighteenth century. After the discovery of the existence of large amounts of gas and dust, which causes light extinction and makes the observations of the center region of the Galaxy very difficult in several bands, the conclusion was that the Galaxy is of a spiral-type, with arms composed by metal rich young stars and a halo and the central region (bulge), constituted by metal poor old stars.

It was later discovered that the irregular distribution of dust and gas presents some "holes" where the extinction is considerably lower. The best known of these so-called windows has been found by Baade (1951) and extensively studied since then. Generally speaking, these low-extinction windows offer a unique opportunity to obtain information about the bulge in the optical range and compare this part of the Galaxy with the halo and disk (and with other galaxies). Great progress has been obtained by the observation of these regions and our picture of the Galaxy changed. For example, with the advent of direct measures of metal abundances of the K giants in the Baade window (Rich 1988), it was possible to conclude that the Galactic bulge is not a metal-poor structure: its mean metallicity is twice the solar neighborhood value. This means that the halo and the bulge are quite distinct structures. Today we know that in the galactic bulge reside stars showing a large range of metallicities.

However, many questions still remain open. The knowledge of the age of each galactic component is one of these and it is important to decide between the several formation scenarios: would the bulge be older than the halo, and if so by how much? How are these structures formed? Many authors have worked on these subjects. Terndrup (1988) estimated the bulge age using the color-color diagrams of four low-extinction windows, analyzing the position of the main sequence strip. He concluded that the bulge is between 11 and 14 Gyr old. In another work, Lee (1992) calculated the relative age of the RR Lyrae in the halo and in the bulge and found that bulge stars would be older than the halo RR Lyrae by 1.3 Gyr. The discovery of very old stars in the bulge does not necessarily mean that the bulge is the first structure to be formed, since the possibility of a merger population exists.

Recently, several works have shown the presence of a galactic bar in the bulge. Stanek et al. (1994), using the stars of the red clump, found an asymmetry in galactic longitude. Similar results were obtained with the hydrogen 21 cm line (Liszt & Burton 1980) and in 2.4 $\mu$m (Blitz & Spergel 1991). Whitelock & Catchpole (1992) detected a longitudinal asymmetry in the distribution of the Mira variables in the bulge. The actual extension and inclination of the bar are problems yet to be solved.

Many works help to solve this and other problems related to the galactic bulge may also be solved, provided adequate data is available. Besides their intrinsic interest as representatives of specific evolutionary phases, variable stars are one of the most valuable ones since they can be used as distance indicators, tracers of metallicity gradients and several other applications, being a very useful tool for galactic structure investigation.

Some massive monitoring programmes, as examples MACHO and OGLE groups, have been recently monitoring the bulge in search of microlensing events (see Paczynski 1996, for a review) with well-known success. The secondary products, complete catalogs of the monitored regions and the discovery of new variable stars, proved to be as interesting as the microlensing events themselves. However, most of these projects are indeed monitoring a limited area of the galactic bulge (in their first phases) in order to maximize the microlensing detection probability. Therefore, a wider coverage of stellar fields for other relevant regions is not attempted.

With these limitation in mind, we have developed and conducted a small monitoring project of 12 selected low-extinction windows from April 1997 until August 1998 (see Table 1). The windows were taken from the works of Blanco & Terndrup (1989) and Blanco (1988). In two successive campaigns we have used the recently refurbished Meridian Circle of the Abrahão de Moraes Observatory (operated by the IAG/USP) to observe these fields (see below). The main objective of the project was the discovery and classification of variable stars (in principle, for variations greater than 0.3 magnitudes), with the construction of a database that can be used by the astronomy community for several researches, as discussed above. The final goal will be to have an on-line, real time processing data to stimulate the study of potentially interesting events by other observing facilities. The first scientific result of the project has been a high-quality extension of the Tycho catalog for the observed regions (Dominici et al. 1999, hereafter Paper I), a necessary step to perform the data reduction. We should remark that this project was not appropriate for the search of microlensing events since the meridian circle allows the observation stars up to V=16 approximately and the probability to detect this kind of effect is low.

The use of the meridian circle for photometry purposes was attempted previously by Henden & Stone (1998), who published a catalog of 1602 variable stars, of which only 85 were found in the literature. The observations which originated this work were performed by FASTT, which is very similar to our instrumental setup.

Many types of variables were expected to show up in our project. Since we were working with a small refractor instrument, only the brighter stars of the bulge, like the Miras, could be observed. Stars that are present in the disc, like classical cepheids are expected and the intermediate population (the transition between disc and bulge and between halo and bulge), like RR Lyrae and W Virginis, should be also present in the sample. Binary systems can be found in all galactic regions along the sight of view, having a variety of magnitudes, periodicities and stellar components. Cataclysmic variables should be relatively rare but fully detectable with our instrumentation (Dominici et al. 1998a).

We shall describe in the next section the selected fields. Section 3 is devoted to a presentation of the instrumental facilities and data reduction method. The observational program and the differential photometry with the Meridian Circle are described in the Sect. 4. The development and tests of the program to organize and search for variable objects (Class32) is discussed in Sect. 5. The variable stars found and their classification and analysis methods are detailed in Sect. 6. Conclusions and future perspectives are presented in Sect. 7. The light curves examples in Appendix A, the catalogue of variable stars in Appendix B and a brief description of the effects of the chromatic aberration in the meridian circle in Appendix C close the present paper.