Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/~inr/drafts/twhyd.ps Äàòà èçìåíåíèÿ: Fri Jun 8 23:31:35 2007 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 04:18:59 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: molecular cloud

Submitted version: February 25, 2003
Activity and Kinematics of members of the TW Hydrae
Association 1
I. Neill Reid
Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218
inr@stsci.edu
ABSTRACT
We present high­resolution echelle spectroscopy of twenty stars in 16 systems
catalogued as members of the TW Hydrae association, and 16 stars identified
as possible new members. We have calibrated the range of coronal and chromo­
spheric activity expected for such young stars as a function of spectral type by
combining our observations with literature data for field and open cluster stars.
We also compute space motions for TWA members and candidate members with
proper motion measurements, using two techniques to estimate distances to stars
lacking direct trigonometric parallax measurements. The mean space motion of
the four TWA members with known parallaxes is (U,V,W:­10.0,­17.8,­4.6)km s -1
. Fourteen of the candidates have properties inconsistent with cluster member­
ship; the remaining two are potential new members, although further observations
are required to confirm this possibility.
Subject headings: Galaxy: solar neighbourhood, (Galaxy:) open clusters and
associations: individual (TW Hydrae)
1. Introduction
The TW Hydrae Association is a loose aggregate of 15 to 20 stellar systems with ages of
10 to 15 Myrs, lying at distances of only #60 parsecs from the Sun. The archetype, TW Hy­
drae, was identified as a T Tauri star by Herbig (1978). Unlike most T Tauris, TW Hydrae
1 Based partly on observations made at the Keck Observatory which is operated by the Californian As­
sociation for Research in Astronomy, and was made possible by generous grants from the W. M. Keck
Foundation.

-- 2 --
is not associated with a parent molecular cloud, but the presence of a substantial gas­ and
dust­rich circumstellar disk, imaged at near­infrared wavelengths through both NICMOS
(Krist et al., 2000; Weinberger et al., 2002) and ground­based (Trilling et al., 2001) obser­
vations, clearly confirms its young age. Several other isolated T Tauri stars were identified
within the same region by de la Reza et al.(1989) and Gregorio­Hetem et al.(1992). Kastner
et al.(1997) proposed that the similar properties exhibited by these systems, particularly at
X­ray wavelengths, indicated a physical association. Following up on the latter proposal,
Webb et al.(1999 ­ WZP99) combined ROSAT data with photographic astrometry and low­
resolution spectroscopy to identify seven additional T Tauri stars in five systems within a
200 square degree region centred on the putative cluster. Further observations by Sterzik et
al.(1999 ­ SACP99), Webb (1999­ W99) and Zuckerman et al.(2001 ­ ZWSB) have expanded
the sample to include as many as 32 stars and brown dwarfs in twenty­one systems.
While it is clear that all of the stars in the TW Hydrae association have similar ages,
it has not yet been established whether they all originated within the same star formation
region. In particular, Jensen et al.(1998) have argued that backtracking 10 Myrs using the
measured proper motions for TW Hya, HD 98800 and CD ­36 o 7429 does not place those stars
in su#cient proximity for membership of the same star­forming cloud. Accurate distance
estimates and full (U, V, W) space motions are required to test whether distributed or single­
origin star formation provides a better model for the association as a whole, and it is partly
to that end that we have obtained echelle observations of other candidate members.
High resolution spectroscopy can also be used to address another issue ­ multiplicity. The
proximity and consequent bright apparent magnitudes of the TW Hydrae association stars
provides a unique opportunity to examine the relative frequency of formation of single and
multiple systems in a low­density star­forming system. It is clear even from the statistical
summary given above that the TW Hydrae stars include a substantial number of binary
and multiple systems. Most of the known companions have been identified through direct
imaging and lie at separations exceeding 10 AU. Spectroscopy samples a di#erent r’egime,
and can reveal systems at much smaller separations of only a few stellar radii.
Our paper is organised as follows: Section 2 presents new spectroscopic and astrometric
observations of both previously identified TWA members and new candidates. Section 3
considers the spectral characteristics, particularly the emission line activity, of the latter
objects. Section 4 considers the space motions and identifies possible new members of TW
Hydrae Association. Finally, Section 5 presents our conclusions.

-- 3 --
2. Observations
2.1. The sample
The stars observed in this program include 18 previously­identified TW Hydrae asso­
ciation members, two of which (TWA 2AB and 3AB) are known to be unresolved (< 1 ##
separation) binaries, and 16 new candidate members. All of the latter were selected as can­
didate members (W99) based on their having relatively high X­ray fluxes, consistent with
the activity expected in extremely young stellar systems. Several are included as likely mem­
bers of the Lower Centaurus­Crux (LCC) Association in Mamajek, Meyer & Liebert's (2002)
survey of post­T Tauri stars.
Since our obtaining those observations, a number of additional candidates have been
proposed based on proper motion characteristics coupled with X­ray activity (Makarov &
Fabricius, 2001 ­ MF01), and on spectral characteristics suggestive of low gravity and sig­
nificant chromospheric activity (Gizis, 2002 ­ G02). Song et al.(2002) have discussed the
spectral properties of the former sample, and show that only three stars have H# and Li
6708 š A line­strengths consistent with ages of 10 to 30 Myrs. Those three stars are probably
members of the more distant LCC Association, but we include them in the present analysis
for comparative purposes.
Table 1 collects photometry and positions for both the previously­identified TW Hydrae
members, and for the new candidates considered in this paper. The naming conventions are
from W99. One of the ROSAT sources, XACT 164, has three possible optical counterparts,
labelled A, B and C in Figure 1. As discussed further below, the brightest optical candidate,
star A, is clearly not related to the X­ray source.
Many of the stars listed in Table 1, particularly the new candidates, have relatively
sparse photometric observations, but almost all have spectral type information. We have
therefore used mean spectral­type/colour relations to estimate (B­V) and (V­I) colours (and
hence bolometric corrections) and (V­R) colours (hence H# continuum fluxes) where such
data are lacking. Fitzgerald (1970) lists (B­V) colours as a function of spectral type, while
Ducati et al.(2001) provide similar data for the Johnson RI system. We have used the
equations given by Bessell (1979) to transform the latter to the Cousins system, and Figure
2 compares the mean relations against observational data for stars within 8­parsecs of the
Sun (from Reid & Gizis, 1997). The Ducati et al.relations provide a reasonable match at
earlier spectral types, but tend to underestimate the colours for mid­G and later dwarfs.
We therefore adopt the median value from the nearby­star dataset as the appropriate colour
for spectral types later than G5. A self­consistent set of optical photometry of all of these
systems would clearly be useful.

-- 4 --
2.2. Echelle spectroscopy
The majority of the stars listed in Table 1 were observed with the HIRES spectrograph
(Vogt et al., 1994) on the Keck­I 10­metre telescope. Most observations were made using
the blue cross­disperser and collimator, covering the wavelength range 3770 to 5280 š A with
minimal gaps between orders. The setup is identical to that described by Zuckerman & Reid
(1998), save that the B2 slit (0 ## 57) was used, giving a resolution of 68,000 or 4.4 kms -1 .
These spectra were obtained on the nights of June 23 and 24 1998 and April 19 and 20 1999
(UT dates). The remaining stars were observed on December 29 and 30 1999 (UT), using
the red cross­disperser and collimator. Those observations cover the wavelength range 6200
to 8500 š A and match those described by Reid & Mahoney (2000 ­ RM2000). We used the
C1 slit (0 ## 86) to give a resolution of 45,000 or 6.67 kms -1 .
All of the observations were reduced and analysed using the same combination of proce­
dures described by RM2000. The mckee suite of programs written by T. Barlow was used to
extract the echelle orders, subtracting the bias level and sky background as well as dividing
the data by tungsten flat­field images. The extracted spectra were wavelength calibrated
using standard iraf routines, with the calibration defined by a single Th­Ar image taken at
the start of each series of observations. As discussed in RM2000, tests show that the wave­
length calibration of HIRES is stable to 2 kms -1 over the course of a full night's observations.
Stability should be higher for the present datasets, since all of the targets are bright, with
integration times of less than 300 seconds, and each set of observations took less than 90
minutes.
We also obtained observations of several early­type M dwarfs as radial velocity stan­
dards. Gliese 507.1 (M1.5), Gl 526 (M1.5) and Gl 638 (K5) were observed in June 1998; Gl
526 in April 1999; and Gl 447 (M4) in December 1999. The spectral types are taken from
Reid, Hawley & Gizis (1995a). All of these stars have accurate radial velocity determinations
by Marcy & Benitz (1989), and the spectral types are reasonably matched to the majority
of the TW Hydrae members. We have used these stars as cross­correlation templates in the
iraf routine fxcor to determine radial velocities for the later­type stars in the present sample.
In the case of earlier­type stars, such as HD 110817 and HD 102458, the radial velocities are
determined directly from the measured wavelength of the Ca II K line. The apparent radial
velocities are adjusted to heliocentric values using corrections derived from the iraf routine
rvcorrect. We estimate that these results are accurate to better than ±2 kms -1 , while the
cross­correlation measurements are accurate to 1 to 1.5 kms -1 for narrow­lined spectra. In
support of this assertion, our results agree to better than 1 kms -1 with recent observations
of several TWA members by Torres et al.(2003). The derived radial velocities are listed in
Tables 2 and 3.

-- 5 --
Most of the stars exhibit significant emission from the hydrogen Balmer series and the
Ca II H & K lines. Equivalent widths for H#, H#, H# and the Ca II K line are listed in Table
2, where we have supplemented our HIRES data with lower­resolution measurements from
W99. Typical uncertainties are 0.1 to 0.25 š A, set primarily by uncertainties in defining the
continuum level. Moreover, these emission lines are likely to be time variable by up to 30%.
TWA 6 was observed on three occasions, and there are clear variations in line strength; for
example, the H# equivalent widths vary from 3.3 š A (24/6/98) to 2.45 š A (19/4/99).
A number of the stars have significant rotational velocities. We have estimated v sin i
using the method outlined in RM2000: calibrating the width of the cross­correlation peak as
a function of v sin i. Taking Gl 526 (v sin i =< 2.9 kms -1 ; Delfosse et al., 1998) and Gl 447
(v sin i =< 2 kms -1 ; Delfosse et al., 1998) as templates, we selected 3 orders with no emission
features and few strong atomic lines, and applied rotational broadening using the line­profile
prescription given by Gray (1982). The broadened spectra were cross­correlated against the
original data, using fxcor in iraf, and the width of the cross­correlation peak recorded. Ex­
perience has shown that this technique is reliable for low and moderate rotational velocities,
v sin i < 30 kms -1 , giving results which agree to within 3 kms -1 with independent measure­
ments (Reid et al., 2002). The uncertainties are larger at higher rotation rates, reaching 5
to 10 kms -1 at v sin i > 50 kms -1 . However, at those velocities the absolute value of the
rotation is less significant than simply identifying the star as a rapid rotator.
Rotational velocity estimates for all candidate TWA members are listed in Table 2.
Several stars have very rapid rotational velocities, as expected for ages of less than 100
Myrs. Torres et al.(2003) have measured v sin i for four stars in our sample: TW Hydrae,
Hen 600A (TWA 3A), ­33:7795A (TWA 5A) and ­36:7429A (TWA 9A). The agreement is
reasonable, with our v sin i tending to be lower than the Torres et al.values by 2­3 kms -1 . The
exception is Hen 600A, where Torres et al.measure v sin i = 20 kms -1 , while our spectrum
shows a sharp cross­correlation profile and narrow lines. The origin of this discrepancy is
not clear.
2.3. Spectroscopic binaries
The TW Hydrae association includes a high fraction of binary and multiple systems.
Using some of the observations discussed in this paper, three stars in Table 1, Hen 600A,
TWA 5A and TWA 6 were reported as possible spectroscopic binaries by Webb et al.(1999).
High­resolution spectroscopy by Muzerolle et al.(2000) confirms that the first mentioned
star is double­lined, with a velocity di#erence of >40 km s -1 . In contrast, more detailed
analysis of repeated observations of the last­mentioned star, TWA 6, indicates that the

-- 6 --
initial suggestion of binarity is probably incorrect. It seems likely that the anomalous cross­
correlation profiles, which prompted that suggestion, stem from e#ects introduced by the
relatively high rotational velocity.
The evidence in favour of binarity of the remaining star, TWA 5A, is given by the
morphology of cross­correlation spectra derived by matching data for that star against our
observations of TWA 10. The latter star has narrow lines, with no indication of binarity,
and was also used as a template by Muzerolle et al.. The cross correlation peak is clearly
double, with the bluer component slightly stronger and o#set by 30 km s -1 from the redder
component. The radial velocity listed in Table 2 is the average of the two values. None of the
other stars in the present sample, either known members or new candidates, show evidence
for binarity. With the addition of TWA 5Aab, the TW Hydrae system now includes at least
one quadruple (HD 98800), two triple systems (TWA 5, Hen 600), seven binaries and eleven
single stars.
3. Activity, radial velocities and TWA membership
Chromospheric and coronal activity are known to be age­dependent phenomena, both
declining with increasing age for main­sequence stars. While the original models of coronal
activity assumed acoustic heating (e.g. Ulmschneider, 1967), both phenomena are now
regarded as manifestations of the stellar magnetic field. Chromospherically active stars
exhibit emission in the alkaline lines (Ca, Na, K) and the Balmer series, while coronal activity
is evident at X­ray wavelengths. Thus, an e#ective means of searching for young, nearby stars
is correlating X­ray and optical imaging data, and identifying sources with potentially high
activity. Positional accuracy can be an issue with X­ray survey data, leading to potential
misidentification between bright stars and background X­ray sources, but direct spectroscopy
can confirm whether the observed chromospheric activity is consistent with expectations.
The majority of candidate TW Hydrae members have been identified from ROSAT data.
In this section, we compare the derived activity levels against measurements of field stars
and cluster members, and consider whether the results are consistent with expectations for
young stars. In addition, as described below, we use our radial velocity measurements as a
check that the space motions are consistent with the known TWA members.

-- 7 --
3.1. Radial velocities for TWA members
The members of a young stellar association are expected to have nearly identical space
motions. As a consequence, the individual proper motions should be aligned, allowing the
determination of a convergent point. The expected radial velocity, V pred , for a cluster member
is simply
V pred. = V S cos # i
where V S is the total space motion and # i the angular distance from the convergent point.
Makarov & Fabricius (MF01) have undertaken a convergent point analysis of the TW
Hydrae Association. They find that the standard analysis yields a convergent point at
(# = 5 h 23 m , # = -19 o 57 # ), but predicts radial velocities in the range +3.2 to +5.8 km s -1 as
opposed to the observed +10 to +13 km s -1 . MF01 suggest that the observed discrepancies
stem from a superimposed expansion of cluster members. Their model places the association
centroid at a distance of 73 pc. from the Sun and at (# = 11 h 1.6 m , # = -32 o 6 # ), # 16pc.
beyond TW Hydrae, with an expansion term of 0.12 km s -1 pc -1 (where the distance is
measured from the cluster centroid). Based on this model, MF01 identify 23 new candidate
members, including the last three stars listed in Table 2. As discussed in §2.1, however, Song
et al.(2002) find that few of these stars have the characteristics expected for an age of 10
Myrs. Thus, it remains unclear whether the expansion term introduced by MF01 has any
validity.
For our present purposes, we have adopted a simpler, empirical approach. We as­
sume that the direction of the convergent point derived by MF01 is valid, that TW Hydrae
(V S =22.1km s -1 ) marks the cluster centroid, and adjust the angular separation to match
the predicted and observed radial velocities for that star. The last step requires an angular
separation of 55.5 o , rather than the 74.4 o derived by MF01, and places the convergent point
at (# = 6 h 37 m , # = -28.6 o ). The resultant values of V pred. are listed in Table 2. The pre­
dicted and observed velocities of previously­identified TW Hydrae members agree to within
2­3 km s -1 . The comparison suggests that this ad hoc technique allows us to weed out
candidates with highly discrepant (#V > 6km s -1 ) radial velocities.
3.2. Activity as a function of spectral type in main­sequence stars
The original demonstration of the age­activity correlation rests on observations of Ca
II H&K emission by Wilson (1966) and Kraft (1967), codified by Skumanich (1972) as a
# - 1
2 relation, where # is age. Soderblom et al.(1991) have since shown that the underlying
relation is more complex. Nonetheless, Skumanich showed that a broadly similar correlation

-- 8 --
holds for age and rotation for upper main­sequence (AFGK) stars, while observations with
the Einstein satellite showed qualitatively similar behaviour at X­ray wavelengths (Vaiana
et al., 1981).
This behaviour is generally consistent with a model where magnetic activity is generated
by an
## dynamo (Parker, 1955), driven by shear between the inner radiative core and the
outer convective envelope. The level of activity generally increases with decreasing mass
along the main sequence. Theoretical stellar models predict that stars should become fully
convective at e#ective temperatures of #3000K, corresponding to spectral type #M3/M4.
At that point, the absence of a radiative core should eliminate the
## dynamo. Thus,
expectations were that magnetic activity should decline significantly for mid­ and late­type M
dwarfs. As is now well known, such is not the case. Nearby M dwarfs have substantial X­ray
fluxes, detectable even by Einstein (Vaiana et al., 1981), while extensive spectroscopic surveys
have shown that chromospheric activity, as measured by Balmer emission, is approximately
constant to spectral type M7 (Hawley et al., 1996), declining only at later types (Gizis et
al., 2000). It seems likely that activity is generated by a turbulent dynamo mechanism in
these fully­convective, later­type dwarfs (Hawley et al., 1999).
Coronal activity is typically measured using the ratio between the observed X­ray flux
and the total bolometric flux, f X
f bol
. We have used an analogous relation to characterise
chromospheric activity, f#
f bol
, where f # is the observed flux in the H# emission line (Reid et
al., 1995b). In our previous analyses, centred on K and M dwarfs, all of the targets have
known distances, allowing the use of absolute bolometric magnitude as a surrogate for mass.
That is not possible in the present study. However, spectral type serves as a ready alternative.
We have therefore compiled literature data on chromospheric and coronal activity to provide
a broader context for assessing measurements of the candidate TWA members.
Figure 3 plots coronal activity as a function of spectral type for nearby field stars,
members of the Hyades and Pleiades clusters and the TW Hydrae Association stars. The
field star X­ray data are taken from Huensch et al.(1999), with the spectral types taken from
either the PMSU survey (Reid et al., 1995a) or the preliminary CNS3 catalogue (Gliese &
Jahreiss, 1991). The Hyades X­ray data are from Stern et al.(1981) for upper main­sequence
stars and Reid et al.(1995b) and references therein for K and M dwarfs; the Pleiades data
are from Micela et al.(1999) and Stau#er et al.(1994). Spectral types for the cluster stars
are derived by combining either (B­V) or (V­I) colours with the relations discussed in §2.1.
Data for the TW Hyd Association members are taken from WZP99, SACP99 and ZWSB.
Most of the X­ray data for the new candidates are from W99 while the spectral types are
taken from SIMBAD or derived from either (B­V) colours or our spectra.
Figure 4 plots chromospheric activity. The uppermost panel plots the H# relative fluxes,

-- 9 --
where data are taken from the PMSU survey for field stars, from Reid et al.(1995b) for Hyades
and Pleiades stars, and from the sources listed above for TWA members and candidate
members. Since H# emission is largely restricted to mid­K and later­type stars, we also
show our measurements of equivalent widths of the Ca II K­line and H# for TWA members
and candidates.
Both Figures 3 and 4 show a general progression of increasing activity with decreasing
age at a given spectral type. The Pleiades stars are typically more active than the Hyades
stars, which in turn lie along the upper edge of the field star distribution. Both cluster
sequences show significant dispersion in activity at a given spectral type. The known TWA
members are clearly significantly more active, both chromospherically and coronally, than
the # 120­Myr Pleiades stars, and provide a clear reference for potential new members, as
discussed further in the following section.
3.3. TWA membership
Section §3.1 and 3.2 and Figures 3 and 4 provide the baseline data which permit a
preliminary assessment of membership probability for TWA candidates. In this section, we
measure each of the stars listed in Table 1 against that baseline.
TWA 19 Originally selected by W99 based on the high X­ray flux of the G­type primary, the
presence of lithium absorption and significant chromospheric emission clearly indicates
a young system. The Hipparcos parallax, however, places the system at a distance
of over 100 parsecs, well beyond the four TWA members with trigonometric parallax
data, and Song et al.(2002) suggest that the binary may be a member of the more
distant (# 130 pc.) Lower Centaurus Crux system.
XACT 164 (also listed as c99a* in W99) As noted above, there are three candidates for the optical
counterpart for this ROSAT source. The brightest, star A in Figure 1, can be ruled
out since it shows no Ca II emission. XACT 164 B is an early­type K dwarf with
significant Ca II emission, while XACT 164 C is a chromospherically­active M dwarf
with both Ca II and H# emission. The f X
f bol
values listed in Table 2 are both derived
from the same ROSAT flux measurement; either star could have an activity level
consistent with an age of 10 Myrs. However, the Ca II emission lines in XACT 164 C
are significantly weaker than in the early­type M dwarf TWA 2, while XACT 164 B is
not only chromospherically active, but also a rapid rotator. Thus, XACT 164 B is a
more likely candidate for the X­ray source, but its radial velocity is inconsistent with

-- 10 --
the value expected for a member of the TW Hydrae association. We exclude all three
as likely TWA members.
HD 110817 This early­type K dwarf has moderate Ca II emission and relatively rapid rotation,
although both properties may be less extreme than expected for an age of 10 Myrs.
The radial velocity is (barely) consistent with TWA membership, although Mamajek
et al. (2002) link HD 110817 with the LCC Association, with a high probability of
membership (94­99%). We consider this star further below.
HD 112227 This G dwarf has moderate Ca II emission, significant rotation and coronal activity
consistent with TWA membership. The observed radial velocity, however, disagrees
by # 6km s -1 with our predicted value. This is clearly a relatively young star, but,
like TWA 19, it may be a member of the LCC Association, rather than TWA (see
Mamajek et al.). We consider HD 112227 further in the following section.
HD 107434 This F6 dwarf was suggested as a common proper motion companion of HR 4796
by Jura et al. (1998). There is little Ca II emission, the star is not detected in
the ROSAT survey and the negative radial velocity rules out membership of the TW
Hydrae association.
HD 114335 As with HD 107434, Jura et al. (1998) identified this star as a possible cpm companion
of HR 4796. There is no Ca II emission, the star is not detected in the ROSAT survey
and the radial velocity di#ers by # 7km s -1 from the value predicted by our convergent
point model. We therefore exclude this star from TWA membership.
XACT 10 (also listed as c99a1 in W99) A chromospherically­active early­type K dwarf with rela­
tively low rotation. The X­ray flux (assuming correct identification) indicates coronal
activity consistent with the known TWA members. Webb finds no H# emission, but
measures an equivalent width of 0.22 š A for the Li I 6708 š A line (W99). We consider this
star further below.
XACT 61 (also listed as c99a5 in W99) Similar in spectral type to the preceding star, the x­ray
flux indicates a lower level of coronal activity, but still significantly above Pleiades stars
of similar spectral type. The v sin i measurement indicates moderate rotation and the
radial velocity is consistent with TWA membership. There is no emission at H# and
relatively weak (0.13 š A) lithium absorption (W99).
XACT 136 (also listed as c99a6 in W99) Another active K dwarf, the coronal activity matches
that of XACT 61. The lower equivalent width measured for the Ca II line may be
a#ected by the substantial rotational velocity. Given the significant broadening, the
radial velocity is not inconsistent with TWA membership. This star was observed by

-- 11 --
Mamajek et al., but is not included in their final sample of candidate LCC Association
members. Webb (W99) finds no H# emission and an equivalent width of 0.17 š Afor Li
I 6708 š A.
R4­42 AB Neither component in this late­type G­dwarf binary shows substantial chromospheric
activity, and we exclude this system from TWA membership.
XACT 186 (also listed as c99a11 in W99) Another active K dwarf, the radial velocity is barely
consistent with our predicted value, and we retain this star for further discussion.
HD 111265 A relatively weak X­ray source, this early­type G dwarf shows little evidence for chro­
mospheric activity, and the apparent magnitude is consistent with a distance exceeding
100 parsecs. We therefore exclude this star as a possible TWA member.
CD ­39 7118B The parallax listed in Table 2 is for the brighter star in this system. The level of chro­
mospheric activity is relatively low for a 10­Myr­old K dwarf, although the rotational
velocity is relatively high. There is significant lithium absorption (0.35 š A equivalent
width). This star may be a member of the LCC association, but it is unlikely to be a
member of the nearer TWA cluster.
XACT 158 (also listed as c99b1 in W99) A late­type G dwarf, the coronal activity is consistent
with an age comparable to that of the Pleiades. There is no evidence for Ca II emission
and the H# line is in absorption (W99). We regard this star as very unlikely to be a
TWA member.
2M1139­31 One of the late­type M dwarfs identified as a possible TWA member by Gizis (2002)
based on the presence of weak CaH and K I absorption, suggestive of low gravity. Gizis
noted that the proper motion is inconsistent with other TWA members, and the level
of chromospheric activity is consistent with Pleiades stars of the same spectral type.
While this is likely to be a relatively young object, possibly a brown dwarf, it appears
unlikely to be TWA member.
2M1207­39 The second M8 candidate member identified by Gizis has substantially higher H#
emission and proper motions consistent with other TWA members.
CD­40 7164 Catalogued as 7760­0836­1 by MF01, this late­type F dwarf has coronal activity consis­
tent with Pleiades­age stars, although the lithium equivalent width measured by Song
et al. (2002) suggests an age younger than IC 2602 (30 Myrs). On this basis, it is
unlikely to be a TWA member, but we consider it further in the following section.

-- 12 --
CD­49 7027 8238­1462­1 in MF01, the coronal activity and lithium equivalent width (Song et al.,
2002) lies near the upper extreme of the Pleiades sequence. Mamajek et al. identify
the star as a member of the LCC association.
8234­2856­1 The third star from MF01, both the coronal and chromospheric activity levels lie nearer
to 120­Myr Pleiades sequence than the known TW Hydrae members. The lithium
equivalent width is consistent with data for IC 2602 stars (Song et al., 2002). This is
another star listed as a high­probability (86­98%) member of the LCC Association by
Mamajek et al..
As noted in §2.1, the last three stars are all identified as probable members of the LCC
Association by Song et al., with photometric distances estimated as 130 parsecs.
4. Space motions of the TW Hydrae Association stars
Radial velocities provide a one­dimensional test of whether candidate TWA members
have kinematics consistent with cluster membership; full space motions provide a more strin­
gent test, but require not only proper motion measurements but also reliable distance esti­
mates. Proper motions are available for ten TWA members and for all of the surviving new
candidates (Table 3). Distance measurements, in the form of Hipparcos parallaxes accurate
to # 10%, are available for only a handful of the brighter stars amongst the previously­known
TWA members: TW Hydrae, at 56.5 parsecs; HD 98800 at 46.7 pc.; CD ­36:7429 at 50.3
parsecs 2 ; and HR 4796 at 67 parsecs. Figure 5 plots the (#, #) distribution of the known
TWA members and the candidates, flagging stars with independent distance estimates. The
current TWA Association spans at least 40 degrees on the sky, corresponding to # 40 parsecs
at the distance of TW Hydrae.
Given the observed distances and spatial extent of the known members, we have used
two models to estimate distances,and hence space motions, for stars in Table 3 which lack
parallax data. First, we assigne distances of 56±15 parsecs; second, following the arguments
outlined by ZWSB, we assume that the mean distance of the TW Hydrae association varies
with position, and calculate individual distances using
d = 1.17# - 137 parsecs
2 This system is a moderately­close binary, separation 6 arcseconds, but with a component flux ratio
exceeding ten at blue/visual wavelengths, the Hipparcos astrometry should not be a#ected.

-- 13 --
where # is Right Ascension, in degrees. This linear relation approximately matches the
trigonometric distance to TW Hydrae, and places the association centroid approximately
midway between HR 4796 and HD 102458 at the eastern extreme. These models allow at
least a preliminary assessment of whether the space motions of the new candidates are likely
to be consistent with cluster membership.
The resulting (U, V, W) velocities 3 are listed in Table 3 and plotted in Figure 6. In
addition to the TW Hydrae stars, we plot data for CD­39 7118B, HD 102458 and the three
MF01 Tycho stars, representing the space motion of the LCC Association, and we show the
location of the Local Standard of Rest (LSR), using the Solar Motion derived by Dehnen
& Binney (1998). The mean motion of the four TWA members with trigonometric parallax
data is
(#U# = -10.0 ± 2.6; #V # = -17.8 ± 2.1; #W # = -4.6 ± 1.1km s -1
corresponding to a velocity of 13 km s -1 with respect to the LSR. The five background LCC
stars have a similar mean motion,
(#U# = -11.6 ± 2.6; #V # = -20.8 ± 1.4; #W # = -5.4 ± 1.0km s -1
suggesting that the TW Hydrae Association is related to the larger LCC complex.
Both distance models produce a relatively elongated distribution of velocities for the
known TWA members. The seven surviving new candidates lie on the boundaries of this
distribution. One star, XACT 61, is well separated from all other stars in the latter diagram,
and is therefore least likely to be associated with TW Hydrae. Of the remaining six stars,
XACT 10 and HD 111265, are closest to the TWA grouping, under either distance model,
and we conclude that the derived space motions do not exclude the possibility that they are
members. On the other hand, assigning distances of 130 parsecs to all six stars (but not
XACT 61) gives space motions consistent with the LCC system.
Considering the TWA members themselves, it is clear that current uncertainties in the
distances, and hence space motions, preclude a reliable determination of whether all of the
current TWA members originated in the same star­forming cloud. Yet the similarities in
motion with respect to the LCC suggests that all are likely to have originated in that com­
plex. Further observations, particularly trigonometric parallax measurements, are required
to confirm or deny this hypothesis.
3 U is positive toward the Galactic Centre, V is positive in the direction of Galactic rotation, and W is
positive toward the North Galactic Pole.

-- 14 --
5. Summary and conclusions
We have presented high­resolution spectroscopic observations of twenty stars in 16 sys­
tems which have been catalogued as members of the TW Hydrae association, and 16 stars
identified as possible new members. Combining our observations with literature data for
field and cluster stars, we have calibrated the range of coronal and chromospheric activity
expected for such young stars as a function of spectral type. Based on that calibration,
together with our radial velocity measurements, nine of the sixteen candidates can be ex­
cluded as possible new members of the TW Hydrae Association. We also suggest that the
observed activity levels argue against membership for one of the M8 dwarfs identified as a
possible TWA member by Gizis (2002) and for all three Tycho stars which survive Song et
al.'s (2002) re­analysis of Makarov & Fabricius' (2001) proper motion study.
Five further stars are unlikely to be TWA members based on the (U, V, W) space
motions deduced for an assumed distance of 56±15 parsecs. The remaining two stars, XACT
10 and HD 111265, have coronal/chromospheric properties and space motions which are not
inconsistent with cluster membership, although further observations are required to verify
this possibility. In the HR diagram, these stars lie between HR 4796 and the late­K and M
dwarfs which comprise the bulk of the current sample. Even with the addition of our more
accurate radial velocity measurements, however, it is clear that the current data do not
have su#cient accuracy to determine whether all of the stars in the TW Hydrae Association
formed within a single molecular cloud, or whether they have their origin in a more dispersed
star­forming region. Accurate trigonometric parallax data, combined with a consistent set
of reliable optical and near­infrared photometry, can go some way towards addressing this
issue.
Thanks to Ben Zuckerman for originally promting these observations, and to Ben and
Inseok Song for helpful comments on an earlier version of the paper. This research was
supported partially by a grant under the NASA/NSF NStars initiative, administered by the
Jet Propulsion Laboratory, Pasadena. We have made extensive use of the Simbad database,
maintained by Strasbourg Observatory, and of the ADS bibliographic service.
REFERENCES
Bessell, M.S. 1979, PASP, 91, 589
de la Reza, R., Torres, C.A.O., Quast, G., Castillo, B.V., Viera, G.L. 1989, ApJ, 346, L61
Dehnen, W., Binney, J.J. 1998, MNRAS, 298, 387

-- 15 --
Delfosse, X., Forveille, T., Perrier, C., Mayor, M. 1998, A& A, 331, 581
Ducati, J.R., Bevilacqua, C.M., Rembold, S.B., Ribeiro, D. 2001, ApJ, 558, 309
Fitzgerald, M.P. 1970, A&A, 4, 234
Flower, P.J., 1996, ApJ, 469, 355
Gliese, W., & Jahreiú, H. 1991, Preliminary Version of the Third Catalog of Nearby Stars,
ADC/CDS Catalog V/70 (pCNS3)
Gizis, J.E., Monet, D.G., Reid, I.N., Kirkpatrick, J.D., Liebert, J. & Williams, R.J. 2000,
AJ, 120, 1085
Gizis, J.E., 2002, AJ, in press (G02)
Gray, D.F. 1982, The observation and analysis of stellar photospheres Cambridge University
Press, (New York, Cambridge)
Gregorio­Hetem, J., Lepine, J.R.D., Quast, G., Torres, C.A.O., de la Reza, R. 1992, AJ,
103, 549
Hawley, S.L., Gizis, J.E., & Reid, I.N. 1996, AJ, 112, 2799
Hawley S.L., Reid I.N., Gizis J.E. & Byrne P.B. 1999, ``Chromospheric Activity in Low Mass
Stars: Observational Results from Clusters and the Field'', in Solar and Stellar Ac­
tivity: Similarities and Di#erences, ed. C.J. Butler and J.G. Doyle, (ASP Conference
Series No. 158), 63.
Herbig, G. H. 1978, Origin of the Solar System, 219
Huensch M., Schmitt J.H.M.M., Sterzik M.F., Voges W. 1999, A&AS, 135, 319
Jensen, E.L.N, Cohen, D.H., Neuh˜auser, R. 1998, AJ, 116, 414
Jura, M., Malkan, M., White, R., Telesco, C., Pina, R., & Fisher, R. S. 1998, ApJ, 505, 897
Kastner, J.H., Zuckerman, B., Weintraub, D.A., Forveille, T. 1997, Science, 277, 67
Kraft, R.P., 1967, ApJ, 150, 551
Krist, J.E., Stapelfeldt, K.R., M’enard, F., Padgett, D.L., Burrows, C.J. 2000, ApJ, 538, 793
Makarov, V.V., Fabricius, C. 2001, A&A, 368, 866 (MF01)

-- 16 --
Mamajek, E.E., Meyer, M.R., Liebert, J. 2002, ApJ, in press
Marcy, G.W., Benitz, K.J. 1989, ApJ, 344, 441
Micela, G., Sciortino, S., Harnden, F.R., Kashyap, V., Prosser, C.F., Damiani, F., Stau#er,
J., Caillault, J.­P. 1999, A&A, 341, 751
Muzerolle, J., Calvet, N., Brice”no, C., Hartmann, L., & Hillenbrand, L. 2000, ApJ, 535, L47
Parker, E.N. 1955, AJ, 122, 293
Reid, I.N., Gizis, J.E. 1997, AJ, 113, 2246
Reid, I.N., Hawley, S.L., Gizis, J.E. 1995a, AJ, 110, 1838
Reid, I.N., Hawley, S.L., Mateo, M. 1995b, MNRAS, 272, 828
Reid, I. N. & Mahoney, S. 2000, MNRAS, 316, 827
Reid, I.N., Kirkpatrick, J.D., Liebert, J., Gizis, J.E., Dahn, C.C., Monet, D.G. 2002b, AJ,
in press
Skumanich, A. 1972, ApJ, 171, 565
Soderblom, R.R, Duncan, D.K., Johnson, D.R.H. 1991, ApJ, 375, 722
Song, I., Bessell, M.S., Zuckerman, B. 2002, A&A, 385, 862 (SBZ02)
Stau#er, J.R., Caillault, J.­P., Gagn’e, M., Prosser, C.F., Hartmann, l., 1994, ApJS, 91, 625
Stern, R.A., Zolcinski, M.­C., Antiochos, S.K., Underwood, J. 1981, ApJ, 249, 647
Sterzik, M.E., Alcala, J.M., Corvino, E., Petr, M.G. 1999, A&A, 346, L41 (SACP99)
Torres, G., Guenther, E.W., Marschall, L.A., Neuhauser, R., Latham, D.W., Stefanik, R.P.
2003, AJ, in press
Trilling, D.E., Koerner, D.W., Barnes, J.W., Ftaclas, C., Brown, R.H. 2001, ApJ, 552, L151
Ulmschneider, p. 1967, Zh. Ap., 67, 193
Urban, S.E., Corbin, T.E., Wyco#, G. L. 1998, AJ, 115, 2161
Vaiana, G. et al., 1981, ApJ, 245, 163
Vogt, S.S. et al., 1994, S.P.I.E., 2198, 362.

-- 17 --
Webb, R.A., 1999, Ph.d. thesis, University of California, Los Angeles (W99)
Webb, R.A., Zuckerman, B., Platais, I., Patience, J., White, R.J., Schwatrz, M.J., McCarthy,
C. 1999, ApJ, 512, L63 (WZP99)
Weinberger, A.J., Becklin, E.E., Schneider, G., Chiang, E.I., Lowrance, P.J., Silverstone,
M., Zuckerman, B., Hines, D.C., Smith, B.A. 2002, ApJ, 566, 409
Wilson, O.C., 1966, ApJ, 144, 695
Zuckerman, B., Reid, I.N. 1998, ApJ, 505, L143
Zuckerman, B., Webb, R.A., Schwartz, M., Becklin, E.E. 2001, ApJ, 549, L233 (ZWSB)
This preprint was prepared with the AAS L A T E X macros v5.0.

-- 18 --
Table
1.
TW
Hydrae
members
and
candidate
members
TWA
Name
#
#
Sp.
type
V
(B­V)
R
I
Ref.
(J)
H
K
Ref.
Source
2000.0
Members
1
TW
Hydrae
B
11
01
51.9
­34
42
17
K7
10.83
0.48
9.98
9.31
O1
8.46
7.65
7.37
I1
2A
­29:8887A
B
11
09
14.0
­30
01
39
M0.5
11.60
2.21
10.25
9.20
O2,
O3
7.85
7.18
I1
2B
­29:8887B
11
09
14.2
­30
01
38
M2
13.50
12.40
10.70
O3
9.09
7.99
I1
3A
Hen
600A
B
11
10
27.9
­37
31
53
M3
12.00
10.60
9.60
O2,
O3
8.22
7.60
7.28
I1
3B
Hen
600B
11
10
27.9
­37
31
52
M3.5
13.70
11.20
10.10
O3
8.63
8.07
7.80
I1
4A
HD
98800Aab
11
22
05.3
­24
46
40
K5
9.41
1.11
8.40
7.50
O2,
O3
6.44
5.82
5.65
I1
4B
HD
98800Bab
11
22
05.3
­24
46
40
K5
9.94
1.40
8.80
7.40
O2,
O3
I1
5A
­33:7795A
B
11
31
55.4
­34
36
27
M1.5
11.72
1.72
10.40
9.10
O4,
O3
7.71
7.06
6.83
I1
5B
­33:7795B
11
31
55.4
­34
36
25
M8.5
20.40
18.10
15.80
O5,
O6
12.1
11.5
I1
6
B
10
18
28.8
­31
50
02
K7
12.00
1.90
10.81
9.77
O5
8.79
8.17
7.97
I1
7
B
10
42
30.3
­33
40
17
M1
11.06
1.54
10.10
9.10
O5,
O3
7.78
7.13
6.89
I1
8A
B
11
32
41.5
­26
51
55
M2
13.30
11.20
9.61
O5,
O6
8.37
7.72
7.44
I1
8B
B
11
32
41.5
­26
52
08
M5
13.85
11.48
O5,
O6
9.91
9.36
9.01
I1
9A
­36:7429A
B
11
48
24.2
­37
28
49
K5
11.32
1.65
10.63
9.60
O5
8.60
7.95
7.68
I1
9B
­36:7429B
B
11
48
24.0
­37
28
49
M1
14.10
12.98
11.45
O5,
O6
10.06
9.41
9.14
I1
10
B
12
35
04.3
­41
36
39
M2.5
12.70
11.80
10.50
O7
9.17
8.55
8.19
I1
11A
HR
4796A
12
36
01.0
­39
52
10
A0
5.78
0.01
O5,
O3
5.80
5.80
5.80
I1
11B
HR
4796B
12
36
01.2
­39
52
14
M2.5
12.80
11.80
10.60
O4,
O
9.32
8.57
8.36
I1
12
RXJ1121.1
11
21
05.6
­38
45
16
M2
13.60
1.50
12.60
11.35
O8
13A
RXJ1121.3A
11
21
17.3
­34
46
46
M1
12.10
1.50
11.20
10.10
O8
13B
RXJ1121.3B
11
21
17.3
­34
46
46
M2
12.40
1.50
11.20
10.10
O8
14
B
11
13
26.5
­45
23
43
M0
13.80
11.85
10.95
O4,
O6
9.85
15A
R
12
34
20.7
­48
15
15
M1.5
14.10
13.51
11.94
O4,
O6
15B
R
12
34
20.6
­48
15
20
M2
14.00
13.41
11.81
O4,
O6
16
R
12
34
56.4
­45
38
07
M1.5
12.30
11.64
10.17
O4,
O6
17
R
13
20
45.4
­46
11
38
K5
12.70
11.69
10.78
O4,
O6
18
R
13
21
37.3
­44
21
53
M0.5
12.90
12.08
10.92
O4,
O6
19A
HD
102458A
B
11
47
24.6
­49
53
03
G5
9.10
0.70
8.75
8.40
O4,
O3
19B
HD
102458B
B
11
47
20.7
­49
53
04
K7
11.90
11.06
10.21
O4,
O3
20
A2­146
B
12
31
38.1
­45
58
59
M2
13.40
12.32
11.10
O9
W99
Candidates
XACT
164
A
B
12
10
13.1
­48
55
27
G
10.70
10.50
10.25
O9
W99
XACT
164
B
B
12
10
10.7
­48
55
48
K2
11.10
10.60
10.20
O9
W99
XACT
164
C
B
12
10
9.5
­48
55
43
M2
15.30
14.00
13.00
O9
W99
HD
110817
B
12
45
06.8
­47
42
58
K1
10.40
0.80
10.00
9.30
O10
8.70
8.23
8.10
I2
W99

-- 19 --
Table
1---Continued
TWA
Name
#
#
Sp.
type
V
(B­V)
R
I
Ref.
(J)
H
K
Ref.
Source
2000.0
HD
112227
B
12
48
07.8
­44
39
17
G8
9.70
0.72
9.30
8.90
O10
8.12
7.69
7.51
I2
W99
HD
107434
B
12
21
9.6
­38
18
09
F6
8.10
0.48
7.85
7.60
O10
W99
HD
114335
B
13
10
19.5
­32
48
20
F6
8.50
0.53
8.30
8.00
O10
W99
XACT
10
B
9
47
19.9
­40
3
10
K1
10.97
0.76
10.55
10.10
O2,
O3
W99
XACT
61
B
10
32
43.9
­44
40
56
K2
9.75
0.96
9.15
8.60
O2,
O3
W99
XACT
136
B
11
35
03.8
­48
50
22
K1
10.35
0.75
9.90
9.50
O2,
O3
9.22
8.93
8.86
I2
W99
R
4­42AB
W
B
11
54
09.6
­31
59
24
G8
7.30
0.75
7.00
6.70
O10
W99
R
4­42AB
E
B
11
54
11.0
­31
59
21
G8
7.50
0.75
7.20
6.90
O10
W99
XACT
186
B
12
43
51.8
­39
46
16
K2
10.14
0.87
9.50
9.00
O2,
O3
W99
HD
111265
B
12
48
20.1
­42
37
34
G0
10.20
0.65
9.90
9.60
O2,
O3
W99
CD­39
7118B
R
11
27
28.8
­39
52
57
K2
10.20
9.60
9.10
O10
W99
XACT
158
R
10
31
24.1
­14
33
36
G8
10.83
0.80
10.40
9.90
O9
W99
2M1139­31
11
39
51.1
­31
59
21
M8
20.20
18.00
15.70
O7
12.67
11.99
11.49
I3
G02
2M1207­39
12
7
33.4
­39
32
54
M8
20.50
18.30
16.00
O7
12.98
11.40
11.95
I3
G02
CD­40
7164
12
13
07.0
­40
56
32
F8
9.80
0.51
9.55
9.25
O10
MF01
CD­49
7027
12
21
55.7
­49
46
13
G7
10.00
0.77
9.60
9.25
O10
8.49
8.05
8.02
I2
MF01
8324­2856­1
12
22
04.3
­48
41
25
K3
10.50
0.81
9.90
9.40
O10
8.78
8.27
8.17
I2
MF01
Note.
---
Column
1
gives
the
identifying
number
from
WZP99
and
ZWSB;
column
2
lists
alternative
designations,
where
XACT
refers
to
optical
identifications
form
the
ACT
Reference
Catalog
(Urban
et
al.,
1998)
­
the
superscript
B
indicates
that
the
star
was
observed
with
the
blue
HIRES
configuration
in
the
program
described
in
this
paper,
R
indicates
observations
with
the
red
HIRES
configuration;
columns
3
and
4
give
the
equinox
J2000
positions;
Column
5
lists
the
spectral
type,
taken
from
WZP99
and
ZWSB
for
known
members,
from
W99
and
SIMBAD
for
W99
candidates,
and
from
G02
and
SIMBAD
for
the
remaining
stars.
The
final
column
lists
the
source
for
the
new
candidate
members.
References
for
optical
photometry:
O1
­
Reid
et
al.(2001);
O2
­
V,
B
from
Hipparcos
catalogue
(ESA,
1997);
O3
­
R,
I
estimated
from
V+sp.
type
;
O4
­
ZWSB;
O5
­
WZP99;
O6
­
V
and/or
R
estimated
from
I+sp.
type;
O7
­
optical
colours
estimated
from
JHK+sp.
type;
O8
­
R,
I
from
SACP99,
V
from
sp.
type;
O9
­
Photometry
derived
from
GSC2.2
+
sp.
type;
O10
­
B/V
from
SIMBAD,
R/I
from
sp.
type.
References
for
infrared
photometry:
I1
­
WZP99;
I2
­
Mamajek
et
al.
(2MASS);
I3
­
G02
(2MASS)

-- 20 --
Table
2.
Emission
line
strengths,
activity
and
rotation
Name
Sp.
type
#
±#
mbol
Ca
II
K
H#
H#
log
f# f
bol
log
fX f
bol
vsin
i
v
rad
v
pred.
References
Member?
š A
š A
š A
km
s
-1
km
s
-1
km
s
-1
Members TW
Hydrae
K7
17.7±1.5
10.00
9.7
49.7
220.0
­1.58
­1.94
<
3
12.5
12.5
WZP99
Y
­29:8887A
M0.5
·
·
·
10.60
21.1
2.1
1.9
­3.51
­1.97
9
10.8
11.6
WZP99
Y
­29:8887B
M2
·
·
·
12.35
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
Y
Hen
600A
M3
·
·
·
10.10
28.1
10.1
21.8
­2.79
­2.25
3
13.4
12.1
WZP99
Y
Hen
600B
M3.5
·
·
·
10.60
·
·
·
·
·
·
7.1
­3.32
36:
·
·
·
WZP99
Y
HD
98800Aab
K5
21.4±2.3
8.85
·
·
·
·
·
·
0.0
·
·
·
­2.37
·
·
·
12.8
10.1
WZP99
Y
HD
98800Bab
K5
21.4±2.3
9.30
·
·
·
·
·
·
0.0
·
·
·
·
·
·
·
·
·
Y
­33:7795A
M1.5
·
·
·
10.00
23.0
5.7
13.4
­2.96
­1.93
>
30
9.5±5
10.5
WZP99
Y
­33:7795B
M8.5
·
·
·
5.55
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
Y
TWA
6
K7
·
·
·
10.25
20.8
2.4
4.7
­3.49
­2.18
55
16.9±5
15.0
WZP99
Y
TWA
7
M1
·
·
·
9.60
38.8
4.9
4.9
­3.44
­2.38
<
3
11.7
13.6
WZP99
Y
TWA
8A
M2
·
·
·
10.10
63.3
5.7
7.3
­3.51
­2.18
<
3
7.8
9.6
WZP99
Y
TWA
8B
M5
·
·
·
11.90
48.5
18.1
16.5
­3.49
·
·
·
5
7.8
9.1
WZP99
Y
­36:7429A
K5
19.9±2.4
10.15
15.3
0.6
2.3
­3.75
­2.15
7
10.6
9.7
WZP99
Y
­36:7429B
M1
19.9±2.4
12.00
14.1
5.3
5.0
­3.62
·
·
·
4
11.3
9.7
WZP99
Y
TWA
10
M2.5
·
·
·
11.00
45.0
3.6
8.4
­3.33
­2.25
<
3
6.6
7.2
WZP99
Y
HR
4796A
A0
14.9±0.8
5.60
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
·
6.3
6.9
WZP99
Y
HR
4796B
M2.5
14.9±0.8
11.10
·
·
·
·
·
·
·
·
·
·
·
·
­2.25
·
·
·
WZP99
Y
RXJ1121.1
M2
·
·
·
11.90
·
·
·
·
·
·
4.5
­3.56
­2.01
15:
10.9
11.6
SACP99
Y
RXJ1121.3A
M1
·
·
·
10.60
·
·
·
·
·
·
2.8
­3.72
­1.81
10
12.0
11.2±3
SACP99
Y
RXJ1121.3B
M2
·
·
·
10.60
·
·
·
·
·
·
4.0
­3.57
·
·
·
10
SACP99
Y
TWA
14
M0
·
·
·
11.45
·
·
·
·
·
·
8.2
­3.18
­2.07
·
·
·
16.0
12.4
ZWSB
Y
TWA
15A
M1.5
·
·
·
12.50
·
·
·
·
·
·
9.0
­3.38
­1.63
22
11.2
8.1
ZWSB
Y
TWA
15B
M2
·
·
·
12.35
·
·
·
·
·
·
10.5
­3.33
·
·
·
30
ZWSB
Y
TWA
16
M1.5
·
·
·
10.70
·
·
·
·
·
·
4.0
­3.70
­2.41
11
9.0
7.7
ZWSB
Y
TWA
17
K5
·
·
·
11.30
·
·
·
·
·
·
2.5
­3.69
­2.36
45
4.6±6
5.1
ZWSB
Y
TWA
18
M0.5
·
·
·
11.40
·
·
·
·
·
·
3.5
­3.66
­2.37
20
6.9±3
4.7
ZWSB
Y
HD
102458A
G5
9.6±1.4
9.00
1.3
abs
­0.6
·
·
·
­2.40
25
14.7±3
10.8
ZWSB
LCC?
HD
102458B
K7
9.6±1.4
10.70
16.6
1.3
2.8
­3.63
·
·
·
10
15.2
ZWSB
LCC?
TWA
20
M2
·
·
·
11.60
25.5
2.9
·
·
·
·
·
·
­2.26
30
8.1±4
7.9
W99
Y
Candidates XACT
164
A
G
·
·
·
10.00
abs
abs
·
·
·
·
·
·
·
·
·
<
3
34.9
9.5
W99
N
XACT
164
B
K2
·
·
·
10.80
7.6
3.3
abs
·
·
·
­2.30
50
22.5±5
9.5
W99
N
XACT
164
C
M2
·
·
·
13.50
11.7
3.7
7.1
­3.28
­1.22
<
3
13.9
9.5
W99
N?
HD
110817
K1
·
·
·
10.10
4.6
abs
0.4
­4.29
­2.33
10.0
10.9
7.4
W99
Y?

-- 21 --
Table
2---Continued
Name
Sp.
type
#
±#
mbol
Ca
II
K
H#
H#
log
f# f
bol
log
fX f
bol
vsin
i
v
rad
v
pred.
References
Member?
š A
š A
š A
km
s
-1
km
s
-1
km
s
-1
HD
112227
G8
·
·
·
9.60
4.0
abs
abs
·
·
·
­2.25
35.0
12.7±7
6.8
W99
?
HD
107434
F6
·
·
·
8.10
0.4
ans
·
·
·
·
·
·
·
·
·
<
3
­12.2
7.6
W99
N
HD
114335
F6
·
·
·
8.55
abs
abs
abs
·
·
·
·
·
·
<
3
10.1
3.4
W99
N
XACT
10
K1
·
·
·
10.70
3.4
abs
abs
·
·
·
­2.38
5
14.1
16.8
W99
Y?
XACT
61
K2
·
·
·
9.25
3.1
abs
abs
·
·
·
­2.72
15
10.6
14.6
W99
Y?
XACT
136
K1
·
·
·
10.10
1.4
abs
abs
·
·
·
­2.66
28
16.6±5
11.4
W99
?
R
4­42AB
W
G8
·
·
·
7.15
1.1
abs
·
·
·
·
·
·
·
·
·
<
3
6.7
8.7
W99
N?
R
4­42AB
E
G8
·
·
·
7.35
1.1
abs
·
·
·
·
·
·
·
·
·
<
3
7.7
W99
N?
XACT
186
K2
·
·
·
9.60
4.1
abs
abs
·
·
·
­2.56
5
11.0
6.3
W99
Y?
HD
111265
G0
·
·
·
10.10
abs
abs
abs
·
·
·
­3.69
<
3
6.3
6.5
W99
Y?
CD­39
7118B
K2
7.0±0.9
9.70
·
·
·
·
·
·
0.3
­4.46
­2.60
25
13.7±4
11.3
W99
N
XACT
158
G8
·
·
·
8.80
·
·
·
·
·
·
abs
·
·
·
­4.21
3
15.7
12.5
W99
N?
2M1139­31
M8
·
·
·
14.70
·
·
·
·
·
·
10.0
­4.25
·
·
·
·
·
·
9.7
G02
N?
2M1207­39
M8
·
·
·
15.00
·
·
·
·
·
·
300.0
­2.77
·
·
·
·
·
·
8.7
G02
Y?
CD­40
7164
F8
·
·
·
9.80
·
·
·
·
·
·
­0.8
·
·
·
­3.55
12
10.0
8.5
MF01,
SBZ
LCC?
CD­49
7027
G7
·
·
·
9.90
·
·
·
·
·
·
­0.4
·
·
·
­3.85
18
12.0
8.9
MF01,
SBZ
LCC?
8324­2856­1
K3
·
·
·
10.30
·
·
·
·
·
·
0.5
­4.07
­3.31
16
13.2
8.8
MF01,
SBZ
LCC?
Note.
---
Column
1
gives
the
common
name
of
eachTWA
Hydrae
member
or
candidatemember;
column
2
lists
the
spectral
type;
column
3
lists
Hipparcos
trigonometric
parallax
data,
if
available;
column
4
gives
the
apparent
bolometric
magnitude,
derived
from
the
photometric
data
listed
in
Table
1;
columns
5,
6
and
7
list
the
equivalent
widths
of
emission
at
the
Ca
II
K­line,
H#
and
H#
­
a
value
of
0.0
indicates
no
emission,
(
·
·
·
)
indicates
that
data
are
not
currently
available;
columns
8
and
9
list
chromospheric
activity,
as
measured
by
the
ratio
between
the
H#
flux
and
the
bolometric
flux,
and
coronal
activity,
the
ratio
between
the
X­ray
flux
measured
by
ROSAT
and
the
bolometrix
flux;
column
10
lists
the
rotational
velocity
­
uncertainties
are
±3
km
s
-1
for
v
sin
i
<
25
km
s
-1
,
rising
to
±10
km
s
-1
for
the
fastest
rotators.
Rotational
velocities
for
HD
98800Aab,
RXJ1121.1,
RXJ1121.3A
and
RXJ1121.3B
are
from
Torres
et
al.(2003).;
column
11
the
radial
velocity,
accurate
to
±2
km
s
-1
for
v
sin
i
<
15
km
s
-1
­
uncertainties
are
listed
explicitly
for
fast
rotators;
column
12
lists
the
predicted
radial
velocity
(see
§3.1);
column
13
lists
the
source
of
H#
measurements,
radial
velocities
and
rotational
velocities
for
stars
not
observed
in
the
course
of
our
HIRES
program;
finally,
column
14
indicates
whether
the
observed
data
are
consistent
with
membership
of
the
TW
Hydrae
association.

-- 22 --
Table 3. Space motions
Name V rad µ# µ # Ref. d U V W Member?
km s -1 mas. mas.
TW Hydrae 12.5 ­66.9 ­12.4 1 56.4 ± 5.3 ­12.0±1.6 ­17.9±1.5 ­5.0±1.5
TWA 2AB 10.8 ­90.8 ­21.0 2 56 ± 15 ­16.6±6.5 ­19.3±2.5 ­8.4±3.0
59 ± 15 ­17.5±6.5 ­19.8±2.5 ­9.0±3.0
TWA 3AB 13.4 ­112.0 ­11.0 2 56 ± 15 ­21.8±7.5 ­22.9±2.9 ­8.7±5.5
59 ± 15 ­23.1±7.5 ­23.4±2.9 ­9.4±5.5
TWA 4ABab 12.8 ­85.5 ­33.4 1 46.7 ± 4.7 ­11.3±2.1 ­20.6±1.5 ­4.8±1.6
TWA 5AB 9.5 ­82.0 ­29.0 2 56 ± 15 ­13.5±5.8 ­19.0±2.5 ­8.9±3.0
65 ± 15 ­16.1±5.5 ­20.8±2.5 ­11.1±3.0
TWA 6 16.8 ­57.0 ­21.0 2 56 ± 15 ­9.3±4.0 ­20.3±1.6 ­6.5±3.1
44 ± 10 ­7.4±2.5 ­19.3±1.5 ­3.8±2.5
TWA 7 11.7 ­122.0 ­29.0 2 56 ± 15 ­22.4±7.8 ­21.4±2.4 ­16.9±4.3
51 ± 10­20.3±5.5 ­20.5±2.5 ­15.0±3.5
TWA 9AB 10.6 ­54.1 ­20.0 1 50.3 ± 6.0 ­6.2±1.8 ­15.9±1.4 ­3.0±1.6
TWA 10 6.6 ­67.0 ­43.0 2 56 ± 15 ­9.4±4.7 ­17.8±2.9 ­9.2±3.8
84 ± 17 ­15.7±6.0 ­24.0±3.0 ­15.0±5.5
TWA 11AB 6.3 ­55.9 ­24.0 1 67.1 ± 4.7 ­10.4±1.3 ­16.6±1.5 ­5.5±1.5
HD 102458AB 14.9 ­33.7 ­10.2 1 104 ± 15 ­7.4±2.5 ­20.7±1.5 ­5.8±1.5 LCC?
Candidates
HD 110817 10.9 ­30.3 ­14.0 2 56 ± 15 ­0.6±2.0 ­14.0±1.3 ­0.9±1.5 LCC
87 ± 17 ­4.0±2.5 ­16.8±2.0 ­3.0±1.5
HD 112227 12.7 ­34.4 ­20.2 2 56 ± 15 ­0.2±2.5 ­16.5±1.6 ­1.2±1.9 LCC
88 ± 18 ­4.0±3.0 ­20.0±2.0 ­4.0±2.8
XACT 10 14.1 ­43.7 20.1 3 56 ± 15 ­12.6±2.8 ­14.3±1.5 ­0.6±2.7 Y?
35 ± 8 ­7.9±2.0 ­14.1±1.5 0.6±2.0
XACT 61 10.6 ­97.7 59.7 56 ± 15 ­28.4±6.2 ­14.9±1.8 ­2.6±4.7 N
48 ± 10 ­2.2±4.0 ­14.3±1.5 2.5±3.0
XACT 136 16.6 ­24.7 ­7.0 56 ± 15 0.6±2.0 ­17.9±1.3 ­0.1±1.6 LCC?
66 ± 15 ­0.3±2.0 ­18.4±1.5 ­0.8±2.0
XACT 186 11.0 ­31.0 ­13.5 56 ± 15 ­0.8±2.3 ­14.2±1.8 ­0.8±1.5 LCC?
86 ± 17 ­4.1±2.0 ­17.2±3.0 ­1.1±2.5
HD 111265 6.3 ­51.4 ­13.9 2 56 ± 15 ­7.5±3.3 ­13.4±2.1 ­1.4±1.5 Y?
88 ± 18 ­13.6±4.0 ­18.2±2.5 ­3.3±2.0
CD­40 7164 10.0 ­32.2 ­12.2 2 130 ± 20 ­11.0±4.5 ­19.8±2.0 ­6.1±3.0 LCC?
CD­49 7027 12.0 ­38.9 ­17.2 2 130 ± 20 ­14.2±4.5 ­22.8±2.5 ­4.2±3.0 LCC?
8324­2856­1 13.2 ­32.4 ­8.1 2 130 ± 20 ­11.6±4.5 ­19.2±2.5 ­4.5±1.5 LCC?
Note. --- Column 1 lists the TWA identification or common name; Column 2 lists the radial velocity;
Columns 3 and 4 give the proper motion, in milliarcseconds; Column 5 lists the source of the astrometry:
1. Hipparcos (ESA, 1997); 2 ­ Tycho ; 3 ­ Web (1999); 4 ­ Gizis (2002)
Columns 6,7 and 8 list the derived (U, V, W) space motions, where U is positive toward the Galactic Centre; V, positive in
the direction of rotation; and W positive toward the North Galactic Pole. The uncertainties, #, are derived from undertainties
in proper motion, radial velocity and distance, setting # = 1.4km s -1 as a lower limit.

-- 23 --
Fig. 1.--- The 4â4 arcminute region centred on the position of XACT 164 (from the POSS II
F­plate in the Digitised Sky Survey). The three candidate optical counterparts are identified
and discussed in the text.

-- 24 --
Fig. 2.--- (V­R) and (V­I) colours as a function of spectral type: lines plot the transformed
mean relations from Ducati et al.(2001); the symbols plot data for stars within 8­parsecs
(crosses), ultracool M dwarfs (triangles) and L dwarfs (solid points). Note that the spectral
sequence runs ..K4, K5, K7, M0... We adopt the median colour derived from the observed
datapoints in estimating magnitudes for the TWA candidates.

-- 25 --
Fig. 3.--- Coronal activity as a function of spectral type. The crosses mark data for field
stars in the immediate Solar Neighbourhood (from Huensch et al., 1999); solid trinagles are
data for the Hyades cluster (Reid et al., 1995a; Stern et al., 1981); open squares are data
for the Pleiades (Stau#er et al., 1994; Micela et al., 1999). Solid points mark data for TWA
members; 5­point stars plot data for the candidates listed in Table 1.

-- 26 --
Fig. 4.--- Emission line activity of the TW Hyadrae candidate members. The upper panel
plots H# activity as a function of spectral type, with the symbols having the same meaning
as in Figure 8. The o#set between earlier­type Hyades and Pleiades stars is less pronounced
at H# than at X­rays since most of the Hyades stars with emission are binaries. The lower
two panels compare the Ca K and H# emission line widths of the new candidates against
data for previously identified members.

-- 27 --
Fig. 5.--- The (#, #) distribution of known TW Hydrae members (solid points) and candidate
members (five­point stars). The subset of the latter stars listed in Table 3 are enclosed in
circles. The numbers list distance estimates for stars with independent distance estimates.

-- 28 --
Fig. 6.--- The (U, V, W) space motions of known TWA members (solid points) and new
candidates (5­point stars), including only stars with both proper motion and radial velocity
measurements (see Table 3). The left­hand panels plot the results if we assume d=56 ± 15
parsecs for stars lacking parallax data; the right­hand panels plot results is we assume the
linear distance gradient described in the text. In both cases, the three MF01 stars, CD­39
7118B and HD 102458, representing the LCC Association, are plotted as crosses. The open
circle marks the mean velocity of the TWA members with measured trigonometric parallaxes,
and the solid square identifies the location of the Local Standard of Rest.