Conclusions

We have presented time resolved spectroscopic observations of the pre-cataclysmic binary PG1026+002. The analysis of the spectra confirms the model elaborated by [Saffer et al. (1993)]. In particular, the phase dependence of the variations of the equivalent width of the Halpha emission components definitely proves that the emission (at least the modulated part) is not due to illumination of the secondary star by the white dwarf, but is intrinsic to the M dwarf itself. At the epoch of the present observations two activity centres appear to be present on opposite sides of the star and perpendicular to the line connecting the system components. Our observations increase the total time base over which radial velocity measurements are available by a factor of 6. In consequence a correspondingly more accurate value for the orbital period could be derived. Revised ephemeris are given.

In addition to the spectroscopy the first time resolved photometric observations of PG1026+002 are presented. Very slight quasi-sinusoidal variations at the orbital period are observed. The half-amplitude is only of the order of 0.006 mag. We interpret them as due to light of the primary reflected off the red dwarf. Although the scatter in the light curve is too large to permit a determination of the system parameters via an adjustment of a model light curve to the data, a light curve calculated with the Wilson-Devinney code based on system parameters taken from [Saffer et al. (1993)] and on additional information from our spectroscopic observations fits the observed light curve excellently, showing that the adopted system parameters are reasonable.

We are grateful to Dr. L.P.R. Vaz for making his version of the Wilson-Devinney code available to us. This work was partially supported by grants of the Conselho Nacional de Pesquisa (CNPq; grants Nos. 301029 and 301784), and the Fundação Amparo a Pesquisa de Minas Gerais (FAPEMIG; grant No. 205096).