4. The composite HR diagram
The Hipparcos mission photometry (ESA, 1997, van Leeuwen et al., 1997) was used in the composite (reddening corrected) HR diagram for the 9 clusters, thus avoiding influences from photometric system transformations. Fig. 2 shows the composite diagram, resolved both according to cluster and according to age (where for comparison also data on the Hyades cluster has been included). In this diagram the Pleiades main sequence coincides with the main sequences of 5 other clusters, all of which have estimated ages similar to the Pleiades. Three older clusters appear to be brighter for the early F and A type stars, and coincide with the average brightness of solar neighbourhood stars in the same colour region. Thus, it appears that the subluminous characteristic of the Pleiades stars is a general age-related feature rather than a local discrepancy in the Hipparcos parallaxes as was suggested by Pinsonneault et al. (1998). Such local discrepancies can also be excluded on grounds of various statistical tests on the Hipparcos parallaxes (van Leeuwen, 1999).
Soderblom et al. (1998) addressed the issue of age relation through comparing the luminosities of chromospherically active (assumed to be young) stars with inactive (assumed to be old) stars, but concluded that they could not find any of these young stars in the region where the Pleiades stars are found. However, the age estimate from a homogeneous group of stars contained in an open cluster should be much more reliable than the inference that certain types of activity can only be associated with young stars. There are some differences between known young stars with chromospheric activity in open clusters and some of the field stars of similar type, most noticeably in the relation between X-ray activity, rotational velocities and the amplitudes of the photometric variations (see Micela et al., 1985).
The conclusion by Pinsonneault et al. (1998) and by Soderblom et al. (1998) that one has to assume that the Hipparcos parallax for the Pleiades is unreliable, and affected by local problems in the Hipparcos parallax determinations can be rejected on the basis of statistical arguments derived from the Hipparcos data (van Leeuwen, 1999). In addition, to consider that 5 other young clusters would be affected in a similar way, while three older clusters are not affected at all seems unlikely. Of all 9 clusters, it is Praesepe that has the poorest coverage (shown by the correlation coefficients in the combined solution), but fits the expected position quite well.
© European Southern Observatory (ESO) 1999
Online publication: December 16, 1998