2 CORAVEL radial velocities

2.1 CORAVEL monitoring of the Henize sample

Radial-velocity monitoring of the Henize sample of S stars has been performed between 1992 and 1997 on the Danish 1.54 m telescope at the European Southern Observatory (La Silla, Chile). A full description of the CORAVEL spectro-velocimeter can be found in Baranne et al. (1979). Basically, CORAVEL measures the velocity of a star by cross-correlating its spectrum with a mask reproducing about 1500 lines of neutral and ionized iron-group species from the spectrum of Arcturus (K1-2 III). The minimum of a gaussian fitted to the cross-correlation dip (cc-dip) thus obtained yields the radial velocity of the star. Further information on the CORAVEL observation and reduction techniques can be found in Duquennoy et al. (1991). Some useful additional information can be derived from the cc-dip: in particular, the parameter Sb, defined as the width of the cc-dip corrected for the instrumental profile, is related to the average stellar line width. It is defined as $Sb = (\sigma^2 - \sigma_0^2)^{1/2}$, where $\sigma$is the observed width of the stellar cc-dip, and $\sigma_0 = 6.29~$km s^-1 is the instrumental width (i.e., the width of the cc-dip of minor planets reflecting the sun light, corrected for the solar rotational velocity and photospheric turbulence).

The radial-velocity data are listed in Table 3, as described in 9. The standard deviation of the radial velocity is listed instead of the $P(\chi^2)$, because most S stars exhibit large velocity variations due to either binarity or envelope motions. These velocity variations are typically larger than the error on the measurements, hence the $P(\chi^2)$ is most of the time close to zero and is therefore not an efficient tool to detect binary stars.

The individual measurements will be available at the Centre de Données Stellaires (CDS) in Strasbourg or on our dedicated web page http://www-astro.ulb.ac.be/.

2.2 Radial velocity curves and orbital parameters

Despite the fact that all extrinsic S stars ought to be binaries (see Jorissen et al. 1998 and Paper III), the number of radial-velocity measurements in the present survey was generally not sufficient to derive a reliable orbit, except in the cases listed in Table 1 and displayed in Fig. 1. Table 1 also provides a few very preliminary orbits derived from 6 or 7 measurements only.

   

Table 1: Orbital elements for Henize S stars. The second line of the first five entries provides the errors on the orbital elements

Hen	P	T [HJD	e	$\gamma$	$\omega$	K	a sini	f(m)	N	O-C	$\Delta T$
	[days]	-2400000]		[km s-1]	[deg]	[km s-1]	[Gm]	$M\sb{\odot}$		[km s-1]	[days]
2	1146.97	38128.45	0.21	21.06	33.40	5.57	86.00	0.019	8	0.10	4931
	2.05	16.28	0.01	0.07	5.96	0.07	1.11	0.001
108	197.24	48632.01	0.00	40.51	0.00	14.41	39.09	0.061	10	0.60	1826
	0.30	1.48	-	0.30	-	0.52	1.40	0.007
121	763.63	49280.49	0.00	-5.495	0.00	10.54	110.69	0.092	11	0.96	1778
	5.89	5.04	-	0.30	-	0.51	5.40	0.013
137	636.39	49699.02	0.44	-21.12	120.22	6.57	51.68	0.014	9	0.63	1780
	12.38	24.16	0.17	0.47	23.57	1.88	8.29	0.007
147	335.76	49559.54	0.22	-8.13	195.56	11.01	49.58	0.043	8	0.52	1579
	1.09	23.78	0.14	0.60	31.33	2.39	10.86	0.028
$\bullet$ preliminary orbits
119	1300	48612	0.14	-23	254	7	118	0.039	6	0.20	1781
124	1983	51068	0.17	6	226	6	155	0.038	7	0.69	1414
183	889	49059	0.10:	-11	173	9	114	0.076	7	0.63	1827

Figure 1: Orbits for 5 Henize S stars. The last three panels labelled "orb?" provide preliminary orbits