Because CrA is not part of the Gould Belt, the young stars surrounding the CrA dark cloud most certainly are associated with the CrA association. They could have formed near the present locations, where the gas has dispersed since then. In that case, these young stars should be somewhat older than the on-cloud TTS. Alternatively, if the CrA cloud has not shrunk, the off-cloud young stars could have dispersed out of the cloud, either slowly (then, the outermost stars would again be older than the on-cloud stars) or, at least in some cases, have higher velocities. In the latter case, one would expect at least a few young run-away TTS ejected from the dark cloud. For checking these possibilities, we need to investigate proper motions and radial velocities.
In order to examine the kinematical state of the stars in the R CrA association, we searched for proper motions for all the stars in Tables 1 and 3 in the Hipparcos (ESA 1997), PPM (Röser & Bastian 1991; Bastian et al. 1993; Röser et al. 1994), ACT (Urban et al. 1997); TRC (Høg et al. 1998), Tycho2 (Høg et al. 2000), and STARNET (Röser 1996) proper motion catalogs. Altogether we could identify 24 stars in these catalogs, including nine stars known already before the ROSAT mission, nine TTS newly identified here, as well as four ZAMS and two dKe/dMe stars. Note that R CrA itself is not listed; it is included in the Hipparcos catalog (HIP 93449), but no meaningful solution for its astrometric parameters could be derived. Proper motions are given in Table 6. For stars present in more than one catalog we usually adopted the most precise proper motion determination, unless it was in conflict with values from other catalogs. All proper motions were transformed to the Hipparcos astrometric system (ICRS) before comparison.
![]() |
Figure 6:
Lithium versus effective temperature.
We plot the lithium equivalent width
![]() ![]() ![]() ![]() |
![]() (a) The visible double star HR 7169/HR 7170 is represented by two entries in the Hipparcos Catalogue (HIP 93368 and HIP 93371) with two individual component solutions. The solution quality is classified as fair; however the proper motion errors are much smaller in TRC, which we adopt here. (b) Hipparcos distance is ![]() (c) RXJ1921.4-3459 has an acceleration solution in the Hipparcos Catalogue. |
Positions and proper motion diagrams for stars in Table 6 are
shown in Fig. 8. The mean proper motion of the R CrA member stars seems
to be very well defined. All except maybe one (HR 7170) of the member stars
known before ROSAT (including the late B-type stars listed at the end of
Table 1, whose membership was not clear before) show very similar proper motions,
and all except two of the newly identified TTS nicely follow this trend.
In contrast to this, the stars classified as ZAMS or dKe/dMe form a
kinematically much more inhomogeneous distribution.
The mean proper motion for all
15 likely members of the association is
(
)
= (5.5,-27.0)masyr-1.
The largest part of this motion is simply the reflex of the solar motion, which is
(
)
(7.1,-23.7)masyr-1at 19
00
,
-37
00
and a distance of 130pc, and
(
)
(5.6,-24.0)masyr-1for 18
45
and the same declination and distance values as above.
Thus the slight difference between the mean proper motion in right ascension of the
eight R CrA member stars known before ROSAT
[(
)
= (7.0,-27.5)masyr-1]
and the seven newly identified member TTS
[(
)
= (3.9,-26.4)masyr-1]
is partly a projection effect reflecting the fact that the new TTS are located at
slightly lower right ascension.
We calculated space velocities for those stars with measured radial velocities (taken from Table 1 for six stars and from Table 4 for another six stars) and corrected them for the influence of galactic rotation (Fig. 9). The solar motion has also been subtracted, although this does not change the relative space velocities between stars, in contrast to galactic rotation or projection effects.
With the exception of HD 176386, which shows a discordant motion in the U-direction, all calculated space velocities are very similar. The mean (U,V,W) values are ( 4.8,-2.1,-2.7)kms-1for the stars known before ROSAT (excluding HD 176386 for taking the mean of the U-velocities) and ( 4.4,-2.2,-0.7)kms-1for the new TTS, i.e. no systematic differences seem to be present. There could still be a small difference in the W-velocities, but the number of stars is so small that this is maybe not significant.
The velocity dispersion is indeed very low,
(
)= (
1.4,1.2,1.8)kms-1
for the whole sample, which excludes the ejection mechanism
(so-called run-away TTS) as major source for off-cloud TTS,
because they should have discrepant velocities.
The velocity dispersion is highest in the W-direction, and taking
also into account that the positions of the stars form a
broader distribution in the Z-direction (
pc,
or
pc if the far off lying RXJ1917.4-3756
is excluded) than in the Y-direction (
pc)
this could be interpreted in terms of stars oscillating around the galactic plane.
It simultaneously would explain why no PMS stars were found in the opposite
direction of the R CrA cloud: the stars already reached the largest
distance from the galactic plane and are currently near their turning point,
consistent with their W-velocities being close to zero.
One complete oscillation around the galactic plane would last
yrs, but need not be finished by now for this scenario to be true.
The off-cloud stars are currently located on average at z = -35 pc,
while the cloud is at z = -39 pc. The stars which now appear to be off-cloud
and the cloud itself, given their current locations and velocities and
tracing their paths back in time, would have been at the same location
yrs ago. The stars located outside of the current
cloud borders are on average
Myrs old, i.e. older than those
inside the cloud, which supports the cloud oszillation scenario. However,
the details and exact time-scales depend on the unknown total cloud mass.
A very similar interpretation could also explain the positions
and the motions of the stars found south of the Taurus clouds
(Neuhäuser et al. 1997; Frink et al. 1997).
![]() |
Figure 7:
H-R diagram with the newly found young stars.
Bolometric luminosity versus effective temperature for all newly
found Li-rich ROSAT counterparts (from Table 4, all filled symbols)
and for Li-rich EO counterparts (Walter et al. 1997, open symbols),
compared to tracks and isochrones from D'Antona & Mazzitelli (1994)
for ages (given on a log scale) and masses (given in ![]() ![]() |
Lépine & Duvert (1994) suggested that high velocity cloud impacts could
trigger star formation, and that subsequently to the impact clouds and stars
could be separated from each other due to different friction during the passage
through the galactic plane. If this is true, the new TTS found outside the dark
cloud today could very well have been born inside the molecular cloud.
The fact that the observed velocity dispersions are so low supports this
scenario. A typical proper motion uncertainty of 3masyr-1 and
an assumed distance uncertainty of 15pc translate into a combined uncertainty
of 3kms-1 for the V and W components of the space
velocity. The observed velocity dispersions are found to be even lower, so
that this is consistent with a very small intrinsic dispersion of the velocities.
There are three stars among the sample of pre-ROSAT R CrA members and newly identified TTS with discordant proper motions from the mean.
HR 7170 belongs to a possibly quadruple system, and therefore the proper motion determination is highly problematic (see footnote to Table 6). Given the fact that the TRC proper motion of HR 7169 is consistent with kinematical membership to the R CrA association and that HR 7170 presumably belongs to the same system, it is likely that the different TRC proper motion of HR 7170 reflects orbital motions within the system. Although the kinematical membership of HR 7170 could not be proven directly, it still should be considered an R CrA association member. The stars HR 7169 and HR 7170 lie in a cavity in the CO distribution (Loren 1979). This suggests similar distances to the stars and cloud. Additionally, optical images show an extensive reflection nebulosity surrounding the pair.
The two other stars are the newly identified TTS RXJ1844.3-3541 and RXJ1901.4-3422. The case for RXJ1844.3-3541 is not clear. According to its spectral type and lithium line strength, it is clearly a young TTS, and also its radial velocity is consistent with membership. Either the proper motion from STARNET maybe in error or the star could have been ejected from the cloud. Comparing its proper motion with the association mean, it would move towards the south-west relative to the cloud, but it is located north-west of the cloud, so that it probably was not ejected from the R CrA cloud.
For RXJ1901.4-3422 the classification as TTS is doubtful. At a late-F spectral
type, even Pleiades ZAMS stars have not yet burned significantly lithium,
so that classification as either pre-MS or ZAMS stars is difficult.
Because its lithium line is stronger than its nearby calcium line,
we classify the star as TTS. The Hipparcos parallax of
pc clearly places it foreground to the R CrA cloud, which explains
its large proper motion, and it is the only star located clearly north of the
R CrA cloud. We conclude that RXJ1901.4-3422 is most likely not an
association member, but young.
Copyright The European Southern Observatory (ESO)