Astron. Astrophys. 320, 74-78 (1997)

3. Results

3.1. Membership

Radial velocity and proper motion are strong membership criteria, but photometry is nevertheless needed to define the final membership. As shown in paper I, the large dimension of the cluster over the sky induces a geometric effect due to the projection of the space velocity along the line of sight. We have taken this effect into account to determine the membership from radial velocities. With the value for the cluster parameters adopted in paper I (, , and d = 130 pc) and the convergent point given by Rosvick et al. (1992b) ( = 92 97, = -48 32 (1950)) we computed the expected radial velocity according to the position of the stars within the cluster field. The computed velocities range from 3.3 to 7.3 km s^-1. The apparent velocity dispersion produced by the geometric projection is thus much larger than the expected internal velocity dispersion. We retained those stars which have a radial velocity within 4 km s^-1 of the expected value. The computed values of the radial-velocity () and the differences in the sense observed minus computed () are given in Tables 1 and 2.

The stars were then plotted in a colour-magnitude diagram (Fig. 1) with different symbols to distinguish between the candidate members (filled squares) and non-members (crosses). The main sequence defined by the members is clearly visible, as well as several stars with a radial velocity appropriate for membership, but located well below the main sequence. To prove that the position of the sequence corresponds to that of the Pleiades, we have plotted in a similar diagram (Fig. 2) the Geneva colours (Rufener 1989) of the stars in the central part of the Pleiades (open squares) and the present sample (filled squares). The new members fall on the same sequence, with a somewhat larger scatter. The latter may arise from observational errors, but also, as stated in paper I, from the depth effect. The Pleiades are some 23 pc across which produces magnitude differences up to 0.3 mag. This diagram shows that most of the stars selected for having similar proper motions and a radial velocity close to the cluster velocity are located at the same distance as the Pleiades. We therefore assume that they are bona fide members of the Pleiades and populate the outer part of the cluster, i.e. the corona.

Fig. 1. Colour-magnitude diagram for the present sample. Filled squares indicate radial velocity members, while crosses stand for non-members.

Fig. 2. Colour-magnitude diagram for the present sample and stars in the central region. Filled squares indicate radial velocity members, crosses stand for non-members and open squares denote the Hertzsprung (1947) stars.

Confirmation of our analysis is afforded by the proper-motion membership probabilities published by Stauffer et al. (1991). They unfortunately covered the central part only, i.e. zones Ia and Ib, which limits the number of star in common. Our members have predominently high membership probabilities, and, conversely, most stars rejected as non-members have low (well below 50%) membership probabilities.

Stars AK II-293, III-391 and III-679 are well located on the colour-magnitude diagram, but their velocities differs by 3.7, 3.1 and 3.4 km s^-1 respectively from the computed value. AK III-391 and III-679 have small values of P() and are possible binaries. All three stars are considered as possible members (remark M?).

Five stars could not be observed successfully with Coravel because they are too hot (AK III-79, III-153, III-419, III-909 and V-151). They all fall nicely on the Pleiades main sequence and are considered as probable members. It would however be useful to get radial velocities for them to more firmly confirm their membership.

Five stars (AK Ib-148a, Ib-174, Ib-574, II-46 and III-31) have a radial velocity in close agreement to be classified as members, but are rejected on the basis of the colour-magnitude diagram. They are listed in Table 3. The first three are located at least one magnitude below the ZAMS, while III-31 is about 1.5 mag above the ZAMS. Additional photometric observations obtained in January 1996 confirm the identification and the previous photometry within a few percents. In the case of Ib-174 and Ib-574, both the radial velocities and the proper motion membership probabilities (.98 and .92 respectively) are in favour of the membership to the Pleiades and it would be interesting to make independant identifications and observations of these stars to confirm their status. The only possibility to have member stars below the main sequence is the presence of a bright white dwarf companion that makes the B-V colour bluer. A few cases exist in the Hyades, all are known spectroscopic binaries.

3.2. Binarity

With two or three radial velocity observation in three years we cannot pretend to have detected all spectroscopic binaries. We have nevertheless detected 4 double-lined and 6 single-lined binaries, most of them showed up already at the second measurements. Orbital elements have been obtained for two of them (Table 4). AK III-031 (Fig. 3) has the right velocity to be a member, but lies rather far above the main sequence band. AK III-416 (Fig. 4) is clearly a non-member.

Fig. 3. Radial-velocity curve for AK III-031

Fig. 4. Radial-velocity curve for AK III-416

A double dip has been seen in the correlation function for AK Ib-146 on the third observation only which is not enough to determine its systemic velocity. It is considered as a possible member and a few more observations are needed to definitively settle its membership. AK Ib-288 has been observed as a double-lined system on several occasions and the systemic velocity quoted in Table 1 has been determined by Wilson's (1941) method and is around 4.1 km s^-1. In addition, it is located about 0.75 mag. above the ZAMS in the colour-magnitude diagram (Fig. 1), which is another strong argument for its membership. AK V-198 is another double-lined binary. Its systemic velocity, also determined by Wilson's (1941) method, is close to 6. km s^-1. There is a slight discrepancy between the mass ratio (close to 0.9), the dip depth and the position within the main sequence. If both components are of nearly equal masses, the star should be brighter than observed.

Table 1. Observational data for new members of the Pleiades

Among the other spectroscopic binaries, AK Ia-351 is a triple system, with a fixed component around +32 km s^-1 and a variable component. This system is therefore non-member.

3.3. Statistics

These 25 new members extend the previous sample of 56 corona members identified in paper I. The total number of F5-K0 members (the domain of completeness is limited to 0.45 0.90) identified in the corona amounts now to 81 and represents nearly 50% of the total number of Pleiades stars already known (88 + 81) in this spectral interval. We are approaching the figures estimated by Artjukhina & Kholopov (1966) who stated that the corona of the Pleiades and Praesepe clusters could contain as much as 60% of the total number of stars. There is a possibility that when the degree of completeness of the population in the outer part of the Pleiades will be higher, more late-type dwarfs will be found in the outer part than in Hertzsprung's (1947) area.

The study of the structure of the Pleiades and discussion of mass segregation in light of the observations accumulated for the inner and outer parts of the Pleiades will be presented in the next paper (IV) of this series (Raboud et al. 1996).

© European Southern Observatory (ESO) 1997

Online publication: July 3, 1998
helpdesk.link@springer.de