2 Observations

2.1 Optical data

2.1.1 Spectroscopy

We observed HD 174005 twice in the course of an ESO Key Program on June 9, 1992 and September 7, 1993. The observations were made with the ESO 1.5 m telescope equipped with the Echelec spectrograph; the spectra covered 300 Å  centred on H$_{\gamma }$ line with a resolution of about 28000; the chosen aperture of 320 $\mu$m corresponds to 1 $.\!\!^{\prime\prime}$52 on the sky. Details about the original frames and on their reduction and calibration are given in Gerbaldi et al. (1999). These spectra have already been analysed for the measure of stellar radial velocity by Grenier et al. (1999) and the duplicity of this star was then detected.

The two spectra were obtained at significantly different phases; the duplicity is obvious in the first one, taken in 1992, while it is not so obvious in the 1993 observation (see Fig. 1). A large difference in the metal line profiles is easily noticed when these two spectra are compared to each other and can be better quantified by comparing the observations with theoretical predictions.

We looked for more observations from the bibliography. We were able to find two contradictory stellar classifications: the recent one by Gray et al. (2001) referring to a metal-weak object and the one by Abt (1988) who classified HD 174005 as an Am star.

The spectra by Gray et al. (2001) on which the classification of HD 174005 as a classic $\lambda$ Boostar is based, were obtained with a resolution of 1.8 Å  and cover the wavelength region from 3800 to 4600 Å; the number of spectra collected by these authors is not specified. The spectra by Abt (1988) have a resolution of 1.0 Å. In a previous paper, Abt (1985) classified this star as an A2 II, with spectra of 2.6 Å resolution.

According to our interpretation, these different classifications are likely to be the consequence of the spectroscopic variability of this object.

Figure 1: The cross correlation curves for the two observed spectra (HD 174005a taken on June 9, 1992 and HD 174005b taken on Sep. 7, 1993); the template is the synthetic spectrum computed by using the model described in the text.

2.1.2 Photometry

The information we could retrieve from the literature on this star are quite scanty; the star has been observed photometrically only by Oblak E. (1978) in $uvby\beta$ (two observations) and by Hall & Mallama (1974) and by Eggen (1968) once and twice respectively in UBV, so that a possible photometric variability could have passed unobserved.

The photometric observations by the Tycho experiment on board of the Hipparcos space experiment span over 3 years; no variability was detected from the 83 observed values, according to the Hipparcos Main Catalogue (ESA 1997); in fact the star is labelled "C''= constant. HD 174005 (= BD $-06^\circ 4913$ = TYC 5126-2404) has a nearby visual companion HD 174005B (= BD $-06^\circ 4912$ = TYC 5126-2381), the similar proper motion of this pair, as given in the Tycho-2 Catalogue (Hog et al. 2000) suggests that they are physically linked. However since the parallax has not been measured for HD 174005B a doubt remains on the physical association of these two objects.

From the $uvby\beta$ colour indices we derived the interstellar absorption by using the Moon (1985) code. The resulting E(b-y)=0.050 is slightly larger than that obtained for HD 174005 B by the same procedure: E(b-y)=0.030. By assuming the hypothesis that the two stars are physically associated, i.e. have the same distance, the latter E(b-y)value is coherent with the results of the analysis by Vergely et al. (1998) of the interstellar extinction in the solar neighbourhood.

2.2 UV data

HD 174005 has not been observed by IUE, but only by the S2/68 on board the ESRO TD1 satellite. The measured fluxes at 2740, 2365, 1965 and 1565 Å  allow us to determine the stellar reddening on the basis of the UV colours only, as discussed in the Catalogue of stellar UV fluxes (Thompson et al. 1978). The approximate relation for A-type stars:

has been used to derive an independent reddening value.

The extinction E(B-V)=0.152 so obtained is much higher than that derived from the visual $uvby\beta$ photometry, E(B-V)=0.069. We shall discuss these UV data in the next section with theoretical computations of the UV fluxes.