. : http://jet.sao.ru/hq/balega/PUBL_BAL/PUB_2005/2005ARep_49_217.ps
: Wed Feb 24 16:01:51 2010
: Tue Oct 2 07:03:15 2012
:
: earth orbit
Astronomy
Re ports, Vol. 49, No. 3, 2005, pp. 217--225.
Translate d from
Astronomiche ski Zhurnal, Vol. 82, No. 3, 2005, pp. 247--255.
Original Russian
Te xt Copyright c
# 2005 by
Bale ga,
Le ushin,
We ige lt.
Atmospheri c Elemental Abundances for the Components
of the
Multi ple System ADS 11061. 41
Draconi s
Yu. Yu. Balega 1 , V. V.
Leushi n 1 , and G.
Wei gelt 2
1 Special
As trophys ical
Obs ervatory,
Rus s ian Academy of
Sciences , Nizhni

Arkhyz,
Karacha

Cherkes s ian Republic, 357147
Rus s ia
2
MaxPlanckIns titut f

ur
Radioas tronomie, Auf dem H

ugel 69, Bonn, 53121 Germany
Received
March 25, 2004; in final form, September 20, 2004
Abstract---We obtained speckle interferometric and spectroscopic observations of
th e system 41 Dra
during its periastron passage in 2001.
Th e components' lines are resolved in
th e spectral interval
3700--9200

A.
Th e observed
wavelength dependence of
th e
brigh tness di#erence between
th e compo
nents is used to estimate
th e B - V indices separately for
each of
th e components: B - V = 0.511
for component a and B - V = 0.502 for component b. We derived improved e#ective temperatures
of
th e components from
th eir B - V values
andh ydrogenline profiles.
Th e observations can be de
scribed
with th e parameters for
th e components T a
eff = 6370 K,
lo g g a = 4.05 and T b
eff = 6410 K,
lo g g b = 4.20.
Th e iron, carbon, nitrogen, and oxygen abundances in
th e
atmosph eres of
th e compo
nents are
lo g N (Fe) a = 7.55,
lo g N (Fe) b = 7.60,
lo g N (C) a = 8.52,
lo g N (C) b = 8.58,
lo g N (N) a = 8.05,
lo g N (N) b = 7.99,
lo g N (O) a = 8.73,
lo g N (O) b = 8.76. c
# 2005
Pleiades Publis hing, Inc.
1. INTRODUCTION
Th e quadruple system ADS 11061 contains two
spectroscopic binaries, 40 Dra and 41 Dra, orbiting
th eir common center of mass
with a period of several
tens of
th ousands of years and an apparent orbital
semiaxis
th at varies from 10 3 to 10 4 AU [1].
Th e
periods of
th e inner orbits of
th e 40 Dra and 41 Dra
pairs are 10.5 and 1246.7 days [2].
Th e system is
h ierarch ical and probably dynamically stable. All its
componentsh ave spectral types of F5--F7.
Th e iron
abundance estimate of [3] indicates
th at ADS 11061
belongs to
th e young disk population, and
th at its
age does not exceed 2.5 billion years. In late evolu
tionary stages,
such stars become giants
with lumi
nosities only
sligh tly di#erent from
th ose of
th e main
sequence stars.
Th e system is of special interest due to
th e very
h igh orbital eccentricity of
th e 41 Dra pair: e =
0.9754 0.0001 [2].
Th eir
motionh as been stud
ied using
both spectroscopic and interferometric
meth ods, and
th e combined orbit's parameters are
accurately known. For example, despite
th e system's
long period,
th e
epoch of periastron passage is known
to
with in about 10 minutes. At periastron,
th e com
ponents of
th e pair
approach each oth er to a distance
of about 10 stellar radii.
Th eh igh eccentricity raises
th e question
ofh ow
such a pair could be formed
and
wh y its
orbith as not been circularized by
th e
action of tidal forces over
th e system's lifetime, as
h ash appened for 40 Dra. One possible scenario of
th e origin and dynamic evolution of ADS 11061
was suggested in [2], but
th e situation is far from
completely clear. In
th e recent periastron passage
of 41 Dra in May--June 2001, it was possible to
obtain integrated spectra in
wh ich th e components'
spectra were fully resolved during a comparatively
sh ort time interval (15 days from periastron).
Th is
made it possible to obtain reliable information on
th e
atmosph eric
ch emical compositions of
each of
th e
stars.
Analyses of
th e orbit of
th e 41 Dra pair yielded
th e
components' masses, M a = (1.39 0.15)M# and
M b = (1.30 0.14)M# , and
th e system's dynamic
parallax [2]. An accurate parallax is crucial for deter
mining
th e star's luminosity and, provided
th e tem
peratures are known,
th e components' radii. Essen
tially all estimates of # are close to # = 0.023 ## [1--6],
wh ereas
th e parallax of 41 Dra in
th e HIPPARCOS
catalog is 0.01884 ## [7],
much lower
th an
th is value.
Th e HIPPARCOS value is probably wrong, since
th e
measurements did not take into account
th e system's
binarity.
Th us,
th e probable distance to
th e system is
44.6 1.0 pc.
Below, we present
th e results of our estimates
of
th e components' temperatures and of
th e iron,
carbon, nitrogen, and oxygen abundances in
th eir
atmosph eres based on spectra taken
with th e
ech elle
spectrograph s of
th e 6 m and Zeiss1000 telescopes
10637729/05/49030217$26.00 c
# 2005 Pleiades
Publish ing, Inc.

218 BALEGA et al.
Table 1.
Brigh tness di#erences between
th e components
of 41 Dra
# ,

A/## , A #m uncertainty Date
5000 0.40 m
0.04 m 2001.2713
5450/300 0.48 0.03 1998.7769
6000 0.46 0.05 2001.2713
6050/240 0.20 0.10 1993.3492
6050/240 0.30 0.10 1993.7646
6050/240 0.30 0.10 1993.8437
6560/300 0.38 0.05 1994.7129
7000 0.44 0.10 2001.2713
8500/300 0.41 0.14 2001.2713
12390/1380 0.47 0.20 2000.7800
12390/1380 0.55 0.20 2001.1920
16480/3170 0.46 0.20 2000.7800
21910/4110 0.47 0.10 1996.2667
of
th e Special
Astroph ysical Observatory (Russian
Academy of Sciences) and
th e Zeiss2000 tele
scope of
th e International Center for Astronomical,
Medical, and Ecological
Research (Mt. Terskol) [8--
11].
Th e
spectrograph sh ave resolutions from 15 000
to 100 000 and cover
th e spectral range from 3700

A
to 9200

A. We also analyze
th e
brigh tness di#erences
of
th e components of 41 Dra estimated over a wide
wavelength range via speckle interferometry and
compare
th ese to
th e spectroscopic data.
2. EFFECTIVE TEMPERATURES
OF THE COMPONENTS OF 41 DRA
Earlier [12], we determined
th e e#ective tempera
tures of
th e system's components from
th e continuum
spectral energy distribution of
th e system [13] and
th e equivalent
width s and profiles of
th e observed
h ydrogen lines. However, we noted certain discrep
ancies in
th e data.
Th us,
th ese earlier results need
to be verified and refined. For
th is purpose, we use
color indices in
th e UBV system, estimates of
th e
brigh tness di#erences of
th e components obtained
from speckle interferometry, and ratios of line
depth s
in
th e components' spectra measured
wh en
th e lines
for
th e two components can be resolved.
Th ough th e B - V values for 41 Dra in various
catalogs [14, 15] di#er
sligh tly, it is reasonable to
adopt
th e combined color index for
th e system B -
V = 0.507 0.001.
We noted in [12]
th at
th e speckleinterferometric
brigh tness di#erence between
th e components was,
on average, #m = 0.426 0.028. However, it is very
important to find
th e
wavelength dependence of
th is
di#erence in order to determine
th e individual color
indices and temperatures of
th e components. For
th is
purpose, we collected all
th e speckleinterferometric
estimates of #m obtained
with th e 6m telescope,
sh own in Table 1.
Th ese lead to
th e following wave
length dependence for #m:
#m = 0.330 + 0.863 10 -5 [

A -1 ]
# [

A].
Th is means
th at, at
th e center of
th e B band
(# 4400

A), #m B = 0.368, and at
th e center of
th e
V band
(# 5500

A), #m V = 0.377.
We measured
th e ratios of
th e central
depth s of
unblended lines of
th e system's components using
spectra covering
th e
wavelength range from 3700 to
9200

A obtained at
ph ases close to periastron.
Th e
central
depth s of absorption lines in
th e combined
spectrum of
th e system are related to
th e components'
luminosities as
R a,obs
R b,obs = E a
#
E b
#
R a
R b ,
wh ere R a , R b and R a,obs , R b,obs are
th e actual and
observed central line
depth s, respectively. If
th e
ph ys
ical
ch aracteristics and
ch emical compositions of
th e
components are similar,
th en
R a
# R b and E b /E a = R b,obs /R a,obs .
Our measurements of some 300 lines
sh ow
th at
th ese ratios are
wavelength dependent:
E b
# /E a
# = 0.73 - 0.54 10 -5 [

A -1 ]
# [

A].
Transforming
th e luminosity ratios into magnitude
di#erences, we obtain
#m B,obs = 0.378 0.008
and #m V,obs = 0.387 0.008,
in fairly close agreement
with th e speckleinterfero
metric #m estimates.
Th e assumption
th at
th e B and V magnitudes for
th e system's components are related as
B b = B a +#m B,obs and V b = V a +#m V,obs
is a good approximation. In
th is case,
th e system's
observed B - V color index will be
B - V = (B a
- V a ) - 2.5
lo g 1 + 10 -0.4#m B,obs
1 + 10 -0.4#m V,obs ,
and B b
- V b = B a
- V a + (#m B,obs
-#m V,obs ).
Substituting
th e observed values into
th ese rela
tions, we can derive
th e color indices for
each of
th e
components in
th e 41 Dra system:
component a: B - V = B a
- V a = 0.511,
component b: B - V = B b
- V b = 0.502.
ASTRONOMY REPORTS Vol. 49 No. 3 2005

ATMOSPHERIC ELEMENTAL ABUNDANCES OF 41 DRACONIS 219
Using models
with normal
ch emical composition
and an
atmosph eric turbulent velocity of 2 km/s of
Kurucz [16], we obtain for
th e components' e#ective
temperatures
component a: T a
eff
= 6370 20 K,
component b: T b
eff
= 6410 20 K.
Th ese temperatures are
sligh tly lower
th an
th ose
we derived in [12], and close to
th e estimates of T eff
and B - V obtained from evolutionary tracks [2].
Using
th e masses and luminosities derived from
th e orbital motion,
th e parallax presented in [2], and
our T eff estimates, we find
th e components' surface
gravities to be
component a:
lo g g a = 4.05 0.10,
component b:
lo g g b = 4.20 0.10.
3. RADIAL VELOCITIES
OF THE COMPONENTS OF 46 DRA
To model
th e system's spectra and compare
th e
synth esized spectra
with observations at various or
bital
ph ases, we must obtain measurements of
th e
components' radial velocities. We used stellar lines
near H#, using
Earth atmosph eric lines as a com
parison spectrum. A list of
th e measured lines is
presented in Table 2.
We calculated
th e radial velocities and reduced
th em to
th e Sun using
th e
Dech 20 code of Galazutdi
nov [17].
Th e measured radial velocities of
th e compo
nents relative to
th e
Earth , V
ar and V
br , are presented
in Table 3,
togeth er
with th e correction to
th e Sun
V a .
Th e
ph ases were calculated
with th e elements
T = JD 2449571.037 and P = 1246.680 d .
Th e derived radial velocities are in good agreement
with th ose of [2], confirming
th e correctness of
th e
orbital elements of 41 Dra presented in
th at paper.
4. HYDROGEN LINE PROFILES
Synth etic spectra of 41 Dra near H# were calcu
lated by summing
th eoretical profiles derived using
th e SintVa code developed by Tsimbal [18]. We
calculated
th e
th eoretical profiles using models
with T eff = 6370 K and
lo g g = 4.05 for component a and
T eff = 6410 K and
lo g g = 4.20 for component b.
Th e
models were derived by interpolating
th e model grid
of Kurucz [16].
Th e combined
synth etic H# spectrum
was calculated for four
ph ases before and after
th e
periastron passage in 2001. We used
th e components'
radial velocities from Table 3 to take into account
wavelength sh ifts in
th e observed spectrum.
Th e
rotational speeds of
th e components (v sin i) needed
to calculate
th e
synth etic spectrum were determined
from absorption lines of metals.
Th e rotational speeds
Table 2. Lines used to measure
th e components' radial
velocities
Stellar lines
Atmosph eric lines
6526.653 SiI 6575.037 FeI 6532.359 H 2 O
6546.239 FeI 6583.710 SiI 6536.720 H 2 O
6554.223 TiI 6586.308 NiI 6542.313 H 2 O
6555.463 SiI 6587.610 C I 6547.705 H 2 O
6559.588 TiII 6592.926 FeI 6548.622 H 2 O
6562.808 H# 6593.884 FeI 6552.629 H 2 O
6569.216 FeI 6597.571 FeI 6557.171 H 2 O
6571.174 FeI 6604.600 ScII 6572.086 H 2 O
6572.779 CaI 6608.044 FeI 6574.852 H 2 O
6574.228 FeI 6609.118 FeI 6599.324 H 2 O
of
both components are close to 8.5 km/s.
Th e
turbulent velocities in
th e components'
atmosph eres
were determined in [3] to be V a
t
= 2.15 km/s and
V b
t
= 1.70 km/s.
Wh en summing
th e spectra, we
assumed a
brigh tness ratio for
th e components of
E b /E a = 0.68.
Th e figure
sh ows a comparison of
th e calculated
and observed profiles,
togeth er
with th eoretical pro
files we calculated for a binary
wh ose components
both h ave T eff = 6500 K and
lo g g = 4.00,
with th e
same
brigh tness ratios, turbulent velocities, and ra
dial velocities as for
th e first version. At all
ph ases,
th e
observed H# profiles are confined
with in
th e range of
th e
th eoretical results. However,
th e agreement
with th e
modelh aving T eff = 6370 K and
lo g g = 4.05 for
component a and T eff = 6410 K and
lo g g = 4.20 for
component b is appreciably better.
Th is confirms
th e
correctness of our
ch oice of
atmosph eric parameters
for
th e components of
th e 41 Dra system.
5. ATMOSPHERIC IRON ABUNDANCES
OF THE COMPONENTS OF 41 DRA
We determined
th e iron abundances from
th e
equivalent
width s of 50 FeI lines and 38 FeII lines
using spectra taken near periastron,
wh en
th e lines
of components a and b were clearly resolved.
Th e
equivalent
width s of
th ese lines, along
with th e
atomic parameters and
lo g N (Fe) values for
each line
and at various
ph ases, are presented in [3],
wh ere
we also describe
th e
tech nique used to determine
lo g N (Fe).
Th e calculations were performed
with th e
KONTUR code [19]. Using models
with T eff = 6575
and
lo g g = 4.08 for component a and T eff = 6600
and
lo g g = 4.26 for component b yielded
ASTRONOMY REPORTS Vol. 49 No. 3 2005

220 BALEGA et al.
Table 3. Radial velocities of
th e components of 41 Dra
Spectrum Date JD
Ph ase V a
r , km/s V b
r , km/s V a , km/s
z3818 14.11.2000 2451862.6090 0.8381 11.86 1.52 1.85 1.12 4.27
z3819 14.11.2000 2451862.6924 0.8382 11.13 1.32 1.92 1.32 4.28
t05311 20.01.2001 2451929.6299 0.8916 13.89 1.67 -2.15 1.22 -3.45
t05312 20.01.2001 2451929.6715 0.8917 13.54 1.77 -2.07 1.35 -3.46
t05413 21.01.2001 2451930.5882 0.8924 14.12 1.85 -2.43 1.56 -3.55
o05a 09.06.2001 2452070.4215 0.0046 -27.12 1.09 44.85 1.38 -1.33
t11406 26.11.2027 2452605.4215 0.4337 5.92 1.37 5.48 1.37 2.94
lo g N (Fe) = 7.66 (component a) and
lo g N (Fe) =
7.72 (component b),
wh ich are
sligh tlyh igh er
th an
th e solar iron abundance
(lo g N (Fe)
# = 7.50).
Th is
paper's refined e#ective temperatures and surface
gravities for
th e components (T a
eff
= 6370 K and
lo g g a = 4.05, T b
eff
= 6410 K and
lo g g b = 4.20) lead
to abundances
much closer to
th e solar value.
Th e lines of neutral iron, FeI, gave
th e
atmosph eric
abundances for
th e components
component a:
lo g N (Fe) = 7.582 0.044,
component b:
lo g N (Fe) = 7.593 0.032.
Th e abundances derived from lines of ionized iron,
FeII, are
component a:
lo g N (Fe) = 7.530 0.039,
component b:
lo g N (Fe) = 7.613 0.045.
Th e rms errors of
th e average values describe
th e
internal uncertainties of
th e results,
wh ich coincide
with in
th e errors.
Th us,
th e system's iron overabun
dance relative to
th e Sun, if present, does not exceed
+0.1dex.
Note
th at any
furth er decrease in
th e e#ective
temperatures for
th e model
atmosph eres would lead
to poorer coincidence of
th e
lo g N (Fe) values derived
from ions
with di#erent ionization stages. For a model
temperature of 6100 K,
th e di#erence in
th e iron
abundances derived for component a from
th e FeI and
FeII lines is 0.13dex, and
th at for component b is
0.27dex,
wh ich are
much larger
th an
th e uncertain
ties.
Th us, we consider our
atmosph eric parameters
to be confirmed by
th e ionization relations.
6. ATMOSPHERIC CARBON, NITROGEN,
AND OXYGEN ABUNDANCES
OF THE COMPONENTS OF 46 DRA
Th e abundances of carbon, nitrogen, and oxygen
are largely determined by
th e evolutionary status of
th e star, and are an important source of information
about
th e star's origin
andh istory.
Th e amount of
carbon and nitrogen in stars more massive
th an
th e
Sun determines
th e rate
ofh ydrogenburning nuclear
reactions in
th e CN cycle, and
th us
th e rate of
th e
star's evolution. In addition, carbon, nitrogen, and
oxygen,
th e most abundant elements in normal stars
h eavier
th anh ydrogen
andh elium, determine
th e in
ternal structure of
th e
atmosph ere and of
th e star as a
wh ole. At
th e same time,
th e comparative poorness of
th ese elements' line spectra introduces considerable
di#culties in determining
th eir abundances. For
th is
reason,
th e vast majority of papers analyzing C, N,
and O abundances in stellar
atmosph eres are based
on measurements of one or two lines of
each element
in
th e stellar spectra.
Th is makes it important to mea
sure
each carbon, nitrogen, and oxygen line present
in
th e spectrum in order to more accurately derive
th e abundances. Correctly taking into account
th e
ph ysical conditions under
wh ich th e analyzed spectral
lines are formed is also important.
Our
ech elle spectra (3700--9200

A ) were acquired
near
th e periastron passage,
wh en
th e components'
spectral lines were resolved, enabling us to identify
and measure relatively many CI, NI, and OI lines in
th e spectrum of
each of
th e system's components.
Table 4 presents
th e parameters of
th e CI lines
measured in
th e spectrum of
th e 41 Dra system, to
geth er
with th e equivalent
width s, W # (in m

A), and
carbon abundances,
lo g N (C), for
each of
th e com
ponents. We determined W # for
each of
th e compo
nents from
th e W a,obs
# and W b,obs
# values measured for
th e resolved lines assuming
th e
brigh tness ratio l =
E b /E a = 0.68.
Th e actual equivalent
width s in
th e
spectra of
th e components are given by
th e relations
W a
#
= 1.68W a,obs
# and W b
#
= 2.50W b,obs
# .
Th e model
atmosph eres and
th e
tech niques used
to derive
th e abundances were
th e same as
th ose used
in our analysis of
th e iron lines.
Th e
lo g gf values were
mainly taken from
th e VALD list [20].
ASTRONOMY REPORTS Vol. 49 No. 3 2005

ATMOSPHERIC ELEMENTAL ABUNDANCES OF 41 DRACONIS 221
Table 4. Line parameters, CI equivalent
width s, and carbon abundances in
th e
atmosph eres of components a and b of
41 DRA
# ,

A # i , eV
lo g gf W a
# , m

A
lo g N (C) a
W b
# , m

A
lo g N (C) b
4762.31 7.48 -2.46 15.4 8.51 15.6 8.51
4762.53 7.48 -2.34 16.6 8.37 17.1 8.38
4792.66 7.95 -2.79 8 8.71 8 8.70
4815.22 7.95 -2.13 7.4 8.00 8 8.03
4815.48 7.95 -2.42 8.2 8.52 8.2 8.51
4817.37 7.48 -3.04 8.4 8.65 11 8.74
4926.43 8.54 -1.97 6.8 8.55 7.5 8.55
4932.05 7.68 -1.88 59.4 8.89 55.5 8.64
5039.05 7.95 -1.79 19.7 8.23 23 8.37
5039.10 7.95 -2.29 10 8.51 11 8.52
5040.12 7.95 -2.3 7 8.23 10.5 8.52
5052.17 7.68 -1.65 59 8.69 63.5 8.76
5380.34 7.68 -1.84 49.2 8.72 40 8.59
5515.55 8.85 -2.34 2.9 8.76 2.9 8.76
5793.12 7.95 -2.06 18.4 8.56 24 8.66
5800.60 7.95 -2.34 8.7 8.50 12 8.59
6586.27 9.00 -1.89 3.8 8.53 3.2 8.52
6587.61 8.54 -1.6 17.4 8.55 18.4 8.57
6588.64 9.17 -2.2 1.9 8.71 1.8 8.68
6591.46 8.85 -2.41 3 8.69 3.5 8.72
6595.24 8.85 -2.41 2.8 8.67 2.5 8.64
6602.41 8.85 -2.38 5 8.83 3.2 8.68
6605.78 8.85 -2.31 1.1 8.51 1.2 8.38
6611.35 8.85 -1.84 2.5 7.97 3.2 8.12
7116.99 8.65 -0.91 40 8.48 42 8.52
7119.66 8.64 -1.15 31.9 8.55 35 8.60
7132.11 8.65 -2.2 3.4 8.53 7 8.67
9061.43 7.48 -0.35 214 8.69 224 8.82
9062.49 7.48 -0.46 185.4 8.58 195.4 8.72
9078.29 7.48 -0.58 160.4 8.48 187.4 8.76
9088.52 7.48 -0.43 167.5 8.40 177.5 8.56
9094.83 7.49 0.15 218 8.26 228 8.44
9111.80 7.49 -0.3 184.9 8.45 225 8.78
ASTRONOMY REPORTS Vol. 49 No. 3 2005

222 BALEGA et al.

0.1
6540
6530 6550 6560 6570
658 0 6590
0
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.1
0
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

1
2
3
4



,

r 
Comparison of observed (dots) and
synth etic (solid c rves) spectra of 41 Dra.
Th e pper
th eoretical spectr m was calc lated
for T a
eff = 6370 K, log g a =
4.05 and T b
eff = 6410 K, log g b = 4.20.
Th e lower
th eoretical spectrum was calculated for
T eff =
65 00 K and log g = 4.00.
Th e spectra correspond to
ph ases (1) 0.83, (2) 0.89, (3) 0.00, (4) 0.43.
Our
equivalentwidth calculations included four
broadening
mech anisms for
th e absorption coe#
cient: radiation damping, Doppler broadening, broad
ening due to
th e quadratic Stark e#ect, and van der
Waals broadening.
Th ese e#ects enter
th e absorption
coe#cient as:
a =
(## R
+## st
+## w
)/## D ,
v =
(## +
d)/## D ,
wh ere
## R is
th e line
width due to radiation damping,
## st is
th e Stark
width ,
## w is
th e van der Waals
width , and d is
th e Stark displacement.
## w can be calculated using
th e approximate for
mula of Unsold:
## w
=# 2 /(4#c)C 0.4
6
34N (HeI)(8kT/#mH ) 0.3 .
ASTRONOMY REPORTS Vol. 49 No. 3 2005

ATMOSPHERIC ELEMENTAL ABUNDANCES OF 41 DRACONIS 223
Table 5. Line parameters, NI equivalent
width s, and nitrogen abundances in
th e
atmosph eres of components a and b
# ,

A # i , eV
lo g gf W a
# , m

A
lo g N (N) a W b
# , m

A
lo g N (N) b
4253.39 10.34 -1.37 7.5 8.66 2 7.99
7442.30 10.33 -0.38 9.4 7.98 10 8.00
8184.86 10.33 -0.29 7.6 7.87 5.8 7.60
8188.01 10.32 -0.29 16.8 8.15 23 8.35
8216.33 10.34 0.13 21.2 7.92 16 7.66
8680.28 10.34 0.35 30.9 7.93 38.2 8.06
8683.40 10.33 0.09 26.5 8.04 28.5 8.07
8686.15 10.33 -0.3 19.6 8.17 23.8 8.29
8703.25 10.33 -0.32 8.4 7.85 8.5 7.80
8711.71 10.33 -0.23 16.1 8.02 12.5 7.92
8718.83 10.34 -0.34 11.2 7.97 18.5 8.17
Table 6. Line parameters, OI equivalent
width s, and oxygen abundances in
th e
atmosph eres of components a and b
# ,

A # i , eV
lo g gf W a
# , m

A
lo g N (O) a
W b
# , m

A
lo g N (O) b
6155.96 10.74 -1.36 1.9 8.83 2.4 8.86
6155.97 10.74 -1.01 3.6 8.78 4.8 8.83
6155.98 10.74 -1.12 3 8.80 3.8 8.84
6156.74 10.74 -1.49 1.5 8.84 1.7 8.84
6156.76 10.74 -0.9 4.6 8.78 5.2 8.80
6156.78 10.74 -0.69 6.3 8.75 7.1 8.76
6158.14 10.74 -1.84 1 8.91 -- --
6158.18 10.74 -1.00 4.9 8.85 -- --
6158.19 10.74 -0.41 11.6 8.77 -- --
7771.94 9.15 0.37 143.6 9.02 131.6 8.97
7774.16 9.15 0.22 123.3 8.93 117.6 8.94
7775.39 9.15 0.00 99.1 8.85 90.5 8.81
8446.25 9.52 -0.46 35.1 8.46 42.7 8.69
8446.36 9.52 0.24 51.7 8.09 63 8.31
8446.76 9.52 0.01 48 8.24 58.5 8.46
Here, line broadening due to collisions
with neutral
h ydrogen
andh elium atoms is taken into account,
and
th e constant C 6 can
eith er be fixed (we took it
from
th e VALD list) or calculated as
C 6 = 6.5 10 -9 ((Z + 1) 2 + 13.595)/# up ) 4/5 ,
wh ere Z is
th e ion's
ch arge and # up is
th e ionization
potential for
th e upper level.
We used two
tech niques to take into account
th e quadratic Stark e#ect. For lines for
wh ich ac
curate calculations are available, we can specify
th e temperaturedependent
electroncollisionh alf
width , w,
th e
ioncollisionh alfwidth , #, and
th e
displacement, d [21].
Th e Stark line
width and dis
placement can
th en be calculated for any point in
th e
atmosph ere using
th e equations
## st = 2wN e 10 -16 (1 + #N 1/4
e A),
d = wN e 10 -16 (d/w + #N 1/4
e B),
wh ere A = 1.75(1 - 0.75r), B = 2(1 - 0.75r) for
neutral atoms; A = 1.75(1 - 1.2r), B = 2(1 - 1.2r)
ASTRONOMY REPORTS Vol. 49 No. 3 2005

224 BALEGA et al.
for ions; and r = 1.85# 1/6 N 1/6
e (e 2 /kT ) 1/2 is
th e
Debye
sh ielding radius.
If no data for w, #, and d are available, we use
th e following approximation to estimate
th e Stark
broadening:
## st
=# 2 /(4#c)C 4 N e ,
wh ere
th e constant C 4 is taken from
th e VALD list or
calculated as
C 4 = 10
-8 ((Z + 1) 2 + 13.595)/# up ) 5/2 .
Th e mean carbon abundances derived for
th e at
mosph eres of components a and b via a comparison
of
th e observed and calculated equivalent
width s are
component a:
lo g N (C) = 8.524 0.036,
component b:
lo g N (C) = 8.576 0.031.
Th ere are many fewer nitrogen lines
th an carbon
lines in
th e spectra of 41 Dra, and nearly all mea
surable lines
th at are not blended
with lines of
oth er
elementsh ave# > 8000 A. We could find only two NI
lines in
th e
sh orterwavelength part of
th e spectrum:
# 4253.39

A
and# 7442.30

A,
wh ich h ave equivalent
width s lower
th an 10 m

A. Table 5 collects for
th e
NI lines
th e same data as
th ose
sh own in Table 4
for
th e carbon lines.
Th e mean
atmosph eric nitrogen
abundances of components a and b are
component a:
lo g N (N) = 8.051 0.068,
component b:
lo g N (N) = 7.991 0.034.
We determined
th e
atmosph eric abundances of
oxygen for
th e components of 41 Dra from lines of
five triplets in
th e red, beginning
with # > 6100

A.
Th e parameters of
th e measured lines and
lo g N (O)
values for
both components are presented in Table 6.
Th e lines are fairly strong and essentially unblended,
ensuring
th e reliability of our oxygen abundance de
terminations.
Th e
atmosph eric abundances for
each of
th e components are
component a:
lo g N (O) = 8.727 0.067,
component b:
lo g N (O) = 8.761 0.055.
7. CONCLUSION
Our spectroscopic and speckleinterferometric
observations of
th e binary system 41 Dra in
th e
h ierarch ical multiple system ADS 11061 obtained
close to its periastron
passageh ave enabled us to
study
th e
atmosph eres of
each of
th e components
of 41 Dra separately.
With in
th e uncertainties, all
our observational data indicate
th at
th e e#ective
temperatures of
th e system's components are 6370 K
and 6410 K, respectively, for components a and b.
Th e virtually exact coincidence of
th e elemental
abundances studied for
th e
atmosph eres of
th e two
components of 41 Dra
with th ose in
th e solar at
mosph ere and meteorites [22] is remarkable;
th e dif
ferences of
severalh undredth s of a dex are probably
due to
equivalentwidth errors and are not significant.
Th is may mean
th at
th e
ch emical composition of
th e
interstellar medium
with in a 50 pc radius of
th e Sun
did not experience any considerable
ch anges during
th e time from
th e formation of
th e solar system some
five billion years ago until
th e formation of
th e 41 Dra
system 2.5 billion years ago [2]. At
th e same time,
th e
absence of any significant deviations of
th e C : N : O
ratios in
th e
atmosph eres of
th e 41 Dra components
from
th e solar values testify to a complete absence of
mixing between
th e core and
th e
atmosph ere, despite
th e
sh ock gravitational stresses periodically experi
enced by
th e stars during
th eir periastron passages,
wh en
th ey
approach each oth er to distances of only a
few stellar radii.
ACKNOWLEDGMENTS
Th e
auth ors
th anks F.A. Musaev and
D.O. Kudryavtsev for taking
th e spectra of 41 Dra.
Th is study was supported by
th e Russian Foundation
for Basic
Research (project no. 010216563a).
REFERENCES
1. A. A. Kisilev and L. G. Romanenko, Astron.
Zh . 73,
875 (1996) [Astron. Rep. 40, 795 (1996)].
2. A. Tokovinin, Yu. Yu. Balega, E. A.
Pluzh nik, et al.,
Astron.
Astroph ys. 409, 245 (2003).
3. Yu. Yu. Balega, V. V.
Leush in, and G. Weigelt, Bull.
Spec.
Astroph ys. Obs. 54, 5 (2003).
4. A. Tokovinin, Pis'ma Astron.
Zh . 21, 286 (1995) [As
tron. Lett. 21, 250 (1995)].
5. I. I. Balega, Yu. Yu. Balega, G. Fal'ke, et al., Pis'ma
Astron.
Zh . 23, 774 (1997) [Astron. Lett. 23, 172
(1997)].
6. V. A. AlWardat, Yu. Yu. Balega, V. V.
Leush in, et al.,
IAU Symp. No. 210: Modelling of Stellar Atmo
s pheres , June 17--21, 2002,
Upps ala, Sweden.
7. The
Hypparcos and Tycho
Catalogues . ESA SP
1200, ESA, 1997.
8. F. A. Musaev, Pis'ma Astron.
Zh . 19, 776 (1993)
[Astron. Lett. 19, 315 (1993)].
9. V. E.
Panch uk, I. D. Najdenov, V. G.
Kloch kova, et al.,
Bull. Spec.
Astroph ys. Obs. 44, 127 (1998).
10. V. E.
Panch uk and S. V. Ermakov, Spets. Astrofiz.
Obs.,
Nauch noTekh .
Otch et No. 267 (1999).
11. F. A. Musaev, G. A. Galazutdinov, A. V. Sergeev,
et al., Kinematika Fiz.
Nebesnykh Tel 15, 282 (1999).
12. Yu. Yu. Balega, V. V.
Leush in, and E. A.
Pluzh nik,
Bull. Spec.
Astroph ys. Observ. 51, 61 (2001).
13. A. V.
Kh aritonov, V. M.
Teresh ch enko, and
L. N. Knyazeva, Spectrophotometric Catalogue of
Stars (Nauka, AlmaAta, 1988).
14. H. L.
Joh nson and W. W. Morgan,
Astroph ys. J. 117,
313 (1953).
ASTRONOMY REPORTS Vol. 49 No. 3 2005

ATMOSPHERIC ELEMENTAL ABUNDANCES OF 41 DRACONIS 225
15. V. G. Kornilov, I. M. Volkov, A. I.
Zakh arov, et al.,
Catalogue UBVR of Northern Sky Bright
Stars (Tr. Gos. Astron. Inst.
Sh ternberga, Moscow, 1991),
Vol. 63.
16. R.
Kurucz,h ttp://kurucz.h arvard.edu
17. G. A. Galazutdinov, Preprint No. 92, SAO (Spec.
Astroph ys. Obs., 1992).
18. V. V. Tsimbal, Magnetic
Stars (Spec.
Astroph ys.
Obs., Russ. Acad. Sci.,
Nizh ni


Arkh yz, 2003)
(in press).
19. V. V.
Leush in and G. P. Topil'skaya, Astrofiz. 22, 121
(1985)
[Astroph ys. 21, 74 (1985)].
20. F. Kupka, N. E. Piskunov, #. #.
Ryabch ikova, et al.,
Astron.
Astroph ys., Suppl. Ser. 138, 119 (1999).
21. G. Grim, Spectral Line Widening in
Plas ma (Mir,
Moscow, 1978).
22. N. Grevesse and A. J. Sauval, Solar
Compos ition
and
Its Evolution---From Core to Corona, Ed. by
C.
Frolich , M. C. E. Huber, S. K. Solanki, and
R. von Steiger (Kluwer,
Dordrech t, 1998), p. 161.
Trans lated by N.
Samus '
ASTRONOMY REPORTS Vol. 49 No. 3 2005