2. Observations and data reductions
A first series of 127 spectra was obtained during 10 consecutive nights (from July 22 to July 31, 1985) at the Observatoire de Haute-Provence with the 1.93 m telescope and the TGR spectrograph. Two spectral regions were selected, one around the SiIII triplet 4552, 4567, 4574, and the other around the HeI 5876 line, each region being observed for 5 nights. The resolving power was very high, above 80 000, and the average exposure time was around 30 minutes, leading to a signal-to-noise ratio of the order of 25. Preliminary results are presented in Le Contel et al. (1987), and the associated radial velocity curves are represented in Fig. 1.
Another series of spectra has been obtained at the Coudé focus of the 1.93 m telescope at the Observatoire de Haute-Provence in 1987. About 60 spectra distributed over 7 consecutive nights (from July 8 to July 14, 1987) were performed with the ISIS fiber fed spectrograph equipped with a pixels CCD. The dispersion was about 8 Å.mm-1 in a domain centered on the HeI 4437 line. A thorium hollow cathode was used for wavelength calibrations and a tungsten lamp for flat field corrections. Final spectra, reduced to the continuum, were filtered with a polynomial function, the resolving power being around 10 000. The heliocentric radial velocities were derived from the average of the measurements of the following 5 lines: HeI 4387, 4437, 4471, OII 4414 and MgII 4481.
The relative stability of the instrumentation was checked on the RV standard Equ, measuring the MgII doublet. The heliocentric radial velocity of this star was observed constant within km s-1. The deduced radial velocity curves are represented in Fig. 2 The velocity curves show night-to-night variations in a range of 8 km s-1, but it is difficult to measure more than a global tendency during a given night.
More recent observations were obtained in 1995 with the spectrograph AURELIE (Gillet et al. 1994), at the Coudé focus of the 1.52 m telescope of the Observatoire de Haute-Provence. The detector was a mono-dimensional CCD. The data were obtained during 10 nearly consecutive nights between May 19 and May 30. The spectral range was centered around the SiIII triplet 4552, 4567, 4574. The resolving power was around 60 000. To ensure a signal-to-noise ratio above 200, the average exposure time was around 20 min. More than 180 spectra were obtained. The reduction processes are the same as the observations used in Paper I. The corresponding velocity curves are represented in Fig. 3. The radial velocity was computed using the SiIII 4567 line. The SiIII 4552 line is blended with SII line while the SiIII 4574 line is relatively faint. The total velocity amplitude varies within a window of approximately 5 km s-1 wide. The amplitude recorded during one individual night never exceeds 2.2 km s-1 (May 28). Note that the internal precision is better than that concerning the 1987 data. However, the velocity jumps noticed on the curves are not real, but due to a pixel drift.
In addition, Her was observed (May 17 to May 21, 1995) with the cross-dispersed spectrograph ELODIE (Baranne et al. 1996) at the Cassegrain focus of the 1.93 m telescope of the Observatoire de Haute-Provence. The detector was a elements CCD. For the whole visible range, the resolving power was about 40 000. To obtain a signal-to-noise ratio around 200, the typical exposure time was around 5 min. The spectra were reduced with the INTER-TACOS software (Baranne et al. 1996) which takes care of the offset and flat-field pixel-to-pixel corrections, as well as the wavelength calibration in the heliocentric frame. To ensure a good internal precision ( km s-1), the stellar and thorium lamp spectra were obtained concurrently. The velocities were deduced from the correlation peak obtained from more than 220 spectral lines. Fig. 4 represents the corresponding radial velocity curves. The radial velocity ranges between -19.0 and -15.6 km s-1. As can be noted, the dispersion is very small.
Our observations were obtained in 1987, from two 8 hour separated observatories: Sierra Nevada in Spain and San Pedro Màrtir in México. The v filter of Strömgren's system was used. The star HD 161569 (B9.5 V, V=6.6) was taken as a comparison and HR 6509 (A4 V, V=5.8) as a check star. No significant variations within 0.003 magnitudes were found in the differences between comparison and check stars during all the campaign.
Due to a probable instability of the instrument together with poor weather conditions, the San Pedro Màrtir observations were not considered further in this paper. In Spain, observations were collected during the second half of July and the first half of August (15 nights spanning 28 days). The observations of the 10 best nights are presented in Fig. 5. The total amplitude is about 0.028 mag, and night-to-night variations are present.
Finally, we also used Hipparcos data (ESA 1997) which consist of two sets (105 measurements each), corresponding respectively to filters and . The measurements are spread over about 1 000 days. Depending on the considered filter, the total amplitude range is 0.046 or 0.033 mag.
© European Southern Observatory (ESO) 2000
Online publication: October 30, 19100