Spectral Analysis and Classification of Herbig Ae/Be Stars

SPECTRAL ANALYSIS AND CLASSIFICATION OF HERBIG Ae/Be STARS

Jesus Hernandez,1,2,3

Nuria Calvet,4,2

Cesar Briceno,1,2,4

Lee Hartmann,4and Perry Berlind

4

Received 2003 September 16; accepted 2003 November 19

ABSTRACT

We present an analysis of the optical spectra of 75 early-type emission-line stars, many of which have beenclassified previously as Herbig Ae/Be (HAeBe) stars. Accurate spectral types were derived for 58 members of thesample; high continuum veiling, contamination by nonphotospheric absorption features, or a composite binaryspectrum prevented accurate spectral typing for the rest. Approximately half of our sample exhibited [O i] k6300forbidden-line emission down to our detection limit of 0.1 A equivalent width; a third of the sample exhibitedFe ii emission (multiplet 42). A subset of 11 of the HAeBe sample showed abnormally strong Fe ii absorption;75% of this subset are confirmed UX Ori objects. Combining our spectral typing results with photometry fromthe literature, we confirm previous findings of high values of total-to-selective extinction (RV � 5) in our largersample, suggesting significant grain growth in the environments of HAeBe stars. With this high value of RV , thevast majority of HAeBe stars appear younger than with the standard RV ¼ 3:1 extinction law and are moreconsistent with being pre–main-sequence objects.

Key words: Hertzsprung-Russell diagram — stars: emission-line, Be — stars: pre–main-sequence —techniques: spectroscopic

1. INTRODUCTION

The Herbig Ae/Be (HAeBe) stars are emission-line stars ofspectral types B, A, and in a few cases F, in most instancesspatially correlated with dark clouds or bright nebulosities(Herbig 1960; Finkenzeller & Jankovics 1984; Waters &Waelkens 1998). Comparison of their effective temperature(TeA) and luminosities with theoretical evolutionary tracks(Strom et al. 1972; Cohen & Kuhi 1979; van den Ancker et al.1998; Palla & Stahler 1991) indicates that these objects areyoung, still approaching the main sequence. The HAeBe starsexhibit IR excesses, which are attributed to dust emission fromcircumstellar disks (Finkenzeller & Mundt 1984; Lorenzettiet al. 1983; Davies et al. 1990; Hillenbrand et al. 1992; vanden Ancker et al. 1997; Malfait, Bogaert, & Walkens 1998).Millimeter observations confirm the existence of dusty disksof substantial mass around some of these objects (Mannings &Sargent 1997, 2000; Natta et al. 2000, 2001). In some casesthe emission lines seen in HAeBe stars exhibit P Cygni pro-files, suggesting formation in winds; more symmetric linesmight arise in hot, extended chromospheres (Herbig 1960;Finkenzeller 1985; Hamann & Persson 1992). Sometimesinverse P Cygni profiles are observed, leading Sorelli, Grinin,& Natta (1996) and Muzerolle et al. (2004) to argue that themagnetospheric infall paradigm that has been applied to low-mass, accreting T Tauri stars (Muzerolle, Calvet, & Hartmann2001) may also hold in these systems.

However, despite the significant progress that has beenmade toward understanding HAeBe stars, problems still re-main. One of these basic issues is deriving reliable estimates

of luminosities and Teff , because it is not always straightfor-ward to determine accurate spectral types and extinction forthese objects. This results in considerable uncertainty in thelocation of HAeBe stars in the Hertzsprung-Russell (H-R)diagram. The presence of continua and emission lines formedoutside the photosphere complicate traditional spectral clas-sification schemes used for early-type stars. Several effortshave been made in the past to classify HAeBe stars, applyingqualitative and quantitative spectral classification schemes(Strom et al. 1972; Cohen & Kuhi 1979; Finkenzeller &Mundt 1984; Finkenzeller 1985; Hillenbrand et al. 1992;Hillenbrand 1995; Mora et al. 2001). However, differences ofseveral subclasses and even classes can be found betweenthese various works.The discrepancies probably arise because of the different

methods used. Strom et al. (1972) used solely the K line ofCa ii and He i lines to derive Teff for 18 HAeBe stars; sincethese lines can sometimes be found in emission, methods thatrely heavily on these features cannot always be used forspectral classification. Cohen & Kuhi (1979) classified 71 H�emission stars earlier than G0 using several spectral indices inthe range 4270–6710 A at a resolution of 7 A. However, someof the features they used, such as He i kk4922, 5016 and Na ikk5890, 5896, can be affected by emission or anomalousabsorption (xx 4 and 5). Finkenzeller (1985) used a schemebased on nine spectral indices in the range 3500–5000 A, buttheir sample consists of only a few stars. Hillenbrand (1995)applied a quantitative spectral classification scheme to 33HAeBe stars using features in the R and I photometric bands;however, these wavelength regions contain few useful spectralindices for classifying stars earlier than F0. Recently, in dataobtained during spectroscopic campaigns carried out by theEXPORT (EXoPlanetary Observational Research Team) con-sortium, Mora et al. (2001) determined spectral types andluminosity classes for 29 HAeBe stars. They selected photo-spheric lines that did not vary on multiepoch spectra, and usedhigh-resolution spectra to correct absorption features for ro-tational broadening. While this investigation yielded more

1 Centro de Investigaciones de Astronomıa (CIDA), Apartado Postal 264,Merida 5101-A, Venezuela; [email protected], [email protected].

2 Postgrado de Fısica Fundamental, Universidad de Los Andes (ULA),Merida 5101-A, Venezuela.

3 Visiting student, Harvard-Smithsonian Center for Astrophysics.4 Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, Cam-

bridge, MA 02138; [email protected], [email protected],[email protected].

1682

The Astronomical Journal, 127:1682–1701, 2004 March

# 2004. The American Astronomical Society. All rights reserved. Printed in U.S.A.

reliable results, the extensive observational effort needed isdifficult to apply to a large number of objects.

Reliable spectral types are important in determiningobservables that can yield valuable information about thephysical environment surrounding these young stars, like thevalue of total-to-selective extinction (RV) and infrared ex-cesses. Some authors (Strom et al. 1972; The et al. 1981;Herbst et al. 1982; Sorelli, Grinin, & Natta 1996; Bibo et al.1992; Gorti & Bhatt 1993; Waters & Waelkens 1998; Whittetet al. 2001) have suggested that HAeBe stars have highervalues of RV than given by the standard interstellar extinctionlaw (RV ¼ 3:1) frequently used to calculated the visual line-of-sight absorption (AV) toward these stars (Hillenbrand et al.1992; Testi et al. 1998; Oudmaijer et al. 2001; Mora et al.2001). The value of RV can be used to infer grain properties ofthe dust surrounding HAeBe stars. In addition, knowledge ofthe reddening law is essential in placing these objects in theH-R diagram and thus deriving masses and ages by compar-ison with evolutionary tracks.

In this contribution we obtain spectral types for a large setof HAeBe stars. We use spectral indices constructed to min-imize the effects of nonphotospheric emission as far as pos-sible. In x 2 we present the observations and data reduction.The spectral classification method is described in x 3. Wediscuss the resulting spectral types and details of specificobjects in x 4. In x 5 we discuss the anomalous features seen inour spectra. In x 6 we discuss the determination of reddeningfor our sample and explore the nature of the interstellar ex-tinction law toward these objects. In x 7 we locate the stars inthe H-R diagram and derive their ages and masses. A sum-mary and conclusions are presented in x 8.

2. OBSERVATIONS

Optical spectra were obtained for 75 of the 99 stars in theHerbig and Bell Catalog (HBC; Herbig & Bell 1988) havingspectral types B, A, and F. Observations were made during1999 July and 2000 January using the 1.5 m telescope of theWhipple Observatory with the FAST Spectrograph (Fabricantet al. 1998), equipped with the Loral 512� 2688 CCD. Thespectrograph was set up in the standard configuration used for‘‘FAST COMBO’’ projects, a 300 groove mm�1 grating anda 300 wide slit. This combination offers 3400 A of spectralcoverage centered at 5500 A, with a resolution of 6 A. We alsoobserved 59 main-sequence and 16 giant and subgiant stan-dard stars covering a spectral range from O8 to M6, using thesame setup (see x 3). Several spectra were obtained for mostprogram and standard stars. The spectra were reduced at theCfA using software developed specifically for FAST COMBOobservations. All individual spectra were wavelength cali-brated and combined using standard IRAF routines.5 The ef-fective exposure times for the combined spectra ranged from afew seconds to 1200 s. The signal-to-noise ratio (S/N) of ourcombined spectra are typically k12 at the central wavelengthregion of the spectra. The spectra were corrected for the rel-ative system response using the IRAF SENSFUNC task andobservations of spectrophotometric standard stars. In Figure 1we show four examples of typical FAST spectra: one standardstar, two HAeBe stars, and one ‘‘continuum’’ star (see x 4.4).

Spectra for the entire sample, plus additional informationfor each star, is available on-line.6

3. SPECTRAL CLASSIFICATION METHOD

Spectral classification of early-type stars (B, A, and F) reliesmainly on the strength of atomic absorption lines, such as thehydrogen Balmer series and He i and Fe i lines, which aresensitive to changes in TeA. In cooler objects like G stars,metallic lines start to increase in strength as a function of TeA,hence the usefulness of features such as Mg ii, Ca ii, Ca i, andthe G band (CH k4300) for spectral typing.

The classification scheme we present here is based on 33spectral features that are sensitive to changes in TeA, listed inTable 1. Column (1) of this table gives an ID number for eachfeature band (FB), column (2) lists the main atomic/molecularspecies contributing to each FB, column (3) gives the centralwavelength of the FB, and column (4) lists the spectral typerange over which the index is useful. These spectral featureswere selected from previous spectral classification studies fornormal stars (Morgan, Keenan, & Kellman 1943; Stock &Stock 1999; Coluzzi 1999; Gray et al. 2001; Pritchet & vanden Bergh 1977; Reid et al. 1995) and studies related toHAeBe stars (Strom et al. 1972; Cohen & Kuhi 1979;Finkenzeller 1985; Waters & Waelkens 1998; Hillenbrand1995). Following Hillenbrand (1995), the equivalent width(Wk) for each spectral feature is obtained by measuring thedecrease in flux due to line absorption from the continuum thatis expected when interpolating between two adjacent bands,

5 IRAF is distributed by the National Optical Astronomy Observatory,which is operated by the Association of Universities for Research in Astron-omy, Inc., under cooperative agreement with the National Science Foundation.

Fig. 1.—Examples of FAST Spectra of our sample. Two HAeBe stars (UXOri and AB Aur), one continuum star (MWC 137), and a standard A0 star (HD140775) are shown. Besides H� , UX Ori does not show other emission lines;however, it exhibits anomalous absorption in the Fe ii (42) multiplet. AB Aurhas H� in emission, in addition to Fe ii (42), He i k5876, He i k6678, and theforbidden line [O i] k6300; there is an emission component in H�. MWC 137exhibits the entire Balmer series in emission, as well as most of the other lines.The absence of absorption features precludes the determination of a spectraltype for this object.

6 See http://cfa-www.harvard.edu/youngstars/jhernand/haebe/principal.htmlor the mirror site at http://www.cida.ve/~jesush/haebe/principal.html.

SPECTRAL ANALYSIS OF HERBIG Ae/Be STARS 1683

TABLE 1

Selected Features Sensitive to Spectral Type

ID

(1)

Features

(2)

kFBa

(A)

(3)

Spectral

Range

(4)

Correlation

CoeDcient

(5)

Fit

Errorb

(6)

1................ Ca ii (K) 3933 A0–G0 0.99 1.6

2................ He i 4026 O8–B3 0.82 2.6

He i 4026 B3–A0 �0.97 1.6

3................ Fe i + Sc i 4047 F2–K1 0.91 2.7

4................ H� 4102 O8–A1 0.97 1.4

H� 4102 A1–F9 �0.99 1.4

5................ He i + Fe i 4144 O8-B3 0.84 2.3

He i + Fe i 4144 B3–A1 �0.97 1.7

He i + Fe i 4144 F5–K3 0.88 3.2

6................ CN + Fe i 4175 F5–G9 0.85 2.3

7................ Ca i 4226 F2–K3 0.94 2.6

8................ Fe i 4271 F2–K5 0.95 2.6

9................ CH (G band) 4305 F2–G2 0.98 0.8

10.............. H� 4349 O8–A1 0.98 1.1

H� 4349 A1–K6 �0.99 1.1

11.............. He i + Fe i 4387 O8–B3 0.50 3.0

He i + Fe i 4387 B3–A1 �0.95 1.5

He i + Fe i 4387 F2–K4 0.99 1.4

12.............. Mn i + Fe i 4458 F2–K4 0.96 2.1

13.............. He i + Fe i + Mn ii 4471 O8–B2 0.87 2.0

He i + Fe i + Mn ii 4471 B2–A1 –0.99 0.9

He i + Fe i + Mn ii 4471 A7–K1 �0.96 2.1

14.............. Mn ii 4481 B5–A1 0.98 0.9

15.............. Fe i + Mn ii 4490 B5–A1 0.98 0.8

16.............. Fe i 4532 A0–G5 0.98 2.8

17.............. He ii + Fe i 4669 O8–B3 0.93 1.2

18.............. Fe i 4787 A5–K3 0.97 3.1

19.............. He i + Fe i 4922 O8–B2 0.93 1.0

He i + Fe i 4922 B2–A1 �0.98 1.5

He i + Fe i 4922 A7–K4 0.98 1.7

20.............. He i + Fe i + Ti i 5016 A0–K5 0.95 3.8

21.............. Fe i + Ti i + Cr i 5079 A0–K3 0.98 2.8

Fe i + Ti i + Cr i 5079 K5–M5 �0.92 2.0

22.............. Fe ii + Mg i 5173 A0–G0 0.96 3.2

23.............. Ca i + Fe i 5270 A0–K0 0.96 2.9

24.............. Fe i 5329 F2–K5 0.96 2.1

25.............. Fe i 5404 O8–B2 0.90 1.3

Fe i 5404 F5–K5 0.97 1.8

26.............. Ca i + TiO 5589 A0–K3 0.95 3.7

Ca i + TiO 5589 M0–M6 0.98 0.6

27.............. Fe i + Mg i + V i 5711 A5–K5 0.96 2.7

Fe i + Mg i + V i 5711 K5–M6 �0.96 1.1

28.............. He i + Na i + TiO 5876 O8–A0 �0.95 1.7

He i + Na i + TiO 5876 G9–M0 �0.96 1.0

29.............. Na i + Ti i 5890 F2–G2 0.89 2.2

Na i + Ti i 5890 G9–K7 0.97 1.1

30.............. Mn i 6015 F2–K5 0.94 2.8

31.............. Ca i + TiO 6162 F5–K3 0.97 1.8

Ca i + TiO 6162 K0–K7 0.96 1.1

32.............. He i 6678 B2–A0 �0.96 1.4

33.............. He i + TiO 7066 O8–A1 �0.96 2.6

He i + TiO 7066 M0–M6 0.97 0.7

a Central wavelength of the feature band.b Error obtained from our first-order fit to the index, in spectral subtypes.

defined here as the blue continuum band [BCB] and the redcontinuum band [RCB]. The equivalent width Wk is defined as

Wk ¼ �kFB

��1� FFB

FBCB þ kFB � kBCB=kRCB � kBCBð ÞðFRCB � FBCBÞ

�;

ð1Þ

where FFB, FBCB, and FRCB are the fluxes at the centralwavelengths (kFB, kBCB, and kRCB) of the feature band andcontinuum bands, respectively, and �kFB is the width of theFB. Figure 2 shows schematically the definition of thesequantities. Equivalent widths or ‘‘indices’’ measured by thisprocedure are largely insensitive to reddening as long as thewavelength coverage of each band is relatively small. Indicesconstructed in this way should also be quite independent ofthe S/N as long as the sidebands are chosen to be next to themeasured feature and are wide enough to obtain a good fluxestimate. Thus, judicious selection of the width of each bandrelies on a compromise between minimizing reddening effectsand maximizing the S/N.

In order to calibrate our set of indices as a function ofspectral type, we selected O8–M6 main-sequence standardstars from various lists (Garcia 1989; Gray et al. 2001; Keenan& Barnbaum 1999; Jaschek 1978; Buscombe 2001) for whichwe had FAST spectra. Although our sample selection wasaimed at HAeBe stars, with spectral types spanning from B toF, we included later spectral types because some indices ex-hibit ‘‘degeneracies’’ (e.g., the G band in Fig. 3); that is, forsome indices one value does not yield a unique spectral type.This approach allows us to study stars that could have largeerrors in their published spectral types.

We measured the spectral indices in our standard starspectra and plotted them against spectral types (which were

assigned a numerical scale between 18 for spectral type O8and 75 for spectral type M5). For each index, we changed thewidth of the BCB, RCB, and FB (from 6 to 30 A, in steps of2 A) and shifted the central wavelength of the BCB and RCBuntil we found the best correlation coefficient between thespectral indices and spectral types. The final value of thecorrelation coefficient obtained for each feature is shown incolumn (5) of Table 1. The width of the bands typically rangefrom 6 to 30 A. Once the optimum widths and centralwavelengths for each band (RCB, BCB, and FB) were fixedby the best correlation coefficient, we fitted straight lines tothe values of Wk as a function of spectral type within vari-ous spectral type intervals. Column (6) of Table 1 shows theerror in spectral subtypes derived from these piece-wise first-order fits for each index.

Illustrative plots for four indices used in our classificationscheme are shown in Figure 3. The top left panel shows thecalibration for index 19 (related to He i + Fe i k4922); thisindex increases up to spectral type B2, where the He i k4922line has maximum absorption; then the index decreases downto a minimum at a spectral type of A0, and increases againfrom F0 to roughly G9 as the Fe i k4921 line becomes strong.Index 9, corresponding to the G band, and index 23, related toCa i k5270 (see Table 1) are shown in the top right and bottomleft panels, respectively; both have a monotonic behavior. Inthe bottom right panel we plot the calibration for index 4,related to the H� line; this index has a change in slope atspectral type A0, where the absorption in the Balmer lines is atits maximum.

Our method of classification can be summarized as follows.First, we used strong, conspicuous features like the G band,Fe i, and He i lines to establish whether the star is earlier orlater than A0. Then the Balmer lines H� and H� (Balmerindices) are used as a first guess to further narrow down thespectral type range. However, because the Balmer indices maybe contaminated by emission lines and affected by luminosityeffects, especially at spectral types near A0 (Morgan et al.1943; Gray et al. 2001), these indices are not given any weightin our final determination of spectral type. Using the spectraltype range guessed from the Balmer indices, we determinedwhich other indices (not affected by effects like line/continuaemission) are useful to classify the particular object.

Once we have determined a specific spectral type range, wedetermine the spectral type for each object by computing aweighted average of the individual spectral types calculatedfrom each index. The weights are estimated from the com-puted error for each index. This error has two contributions,the error from the fit to the standard main sequence as spec-ified in column (6) of Table 1 and the error in the measuredWk. The latter is calculated by assuming Poisson statistics(Gray 1992, p. 81), such that the error in each band (FB, BCB,and RCB) is the square root of the number of counts. We thenpropagated the error in each band to obtain the combinedmeasurement error in the Wk. Finally, we rejected spectralindices that yielded spectral types deviating more than 3 �from the weighted average or that have an error larger than sixspectral subtypes. In this way we minimized possible con-tamination of the indices by artifacts, emission lines, or ano-malous absorption features (x 4 and x 5).

This classification scheme is largely independent of lumi-nosity because most of the indices selected are not sensitive tothe surface gravity of the star. In Figure 4 we plot the spec-tral type determined with our method against the publishedspectral type for our set of main-sequence and giant standard

Fig. 2.—Definition of the bands used to calculate the equivalent width. Thecontinuum is established by interpolating at the central feature band (FB)between two adjacent continuum bands (BCB and RBC). The dashed linesindicate the boundary of each band. The dotted line shows the projectedcontinuum.


stars; the overall error in our calibration does not changesignificantly when we include giant stars. It can be seen thatuncertainties due to differences in luminosity class are small incomparison to the measurement errors.

4. SPECTRAL CLASSIFICATION

4.1. General Considerations

The presence of conspicuous signs of stellar and circum-stellar activity in the optical spectra of HAeBe stars dis-tinguishes them in general from their older nonpeculiar main-sequence counterparts. One of the most characteristic featuresis the presence of H� in emission, though some stars alsoexhibit an emission component in higher Balmer lines. Inaddition, other features can also be seen in emission among anumber of HAeBe stars, such as [O i] k6300, He i kk5876,6678, the Na i doublet located at 5890 A and the kk4924,5018, and 5169 lines of the multiplet 42 of Fe ii (e.g., MWC137 in Fig. 1).

Some HAeBe stars show absorption features in their spectrathat appear anomalous when compared with main-sequencestars of the same spectral type. Examples are the Na i kk5890,5896 doublet and the multiplet 42 of Fe ii in absorption (e.g.,UX Ori in Fig. 1). These anomalous absorption features arebelieved to be caused by material surrounding the star. Be-cause the emission as well as the anomalous absorption lines(x 5) are thought to originate outside the stellar photosphere,we collectively call them nonphotospheric features.Continuum radiation generated outside the stellar photo-

sphere can affect the stellar flux in the bands used to determinea spectral type (Herbig 1960; Hamann & Persson 1992;Corcoran & Ray 1997; Bohm & Catala 1994). The superpo-sition of a nonphotospheric continuum on the stellar spectrum(veiling) is an effect that has to be taken into account whenattempting to classify pre–main-sequence (PMS) stars. Veilingreduces the depth of absorption features and is wavelengthdependent, affecting the spectral type determination. Some ofthe stars in the HAeBe sample show almost all features in

Fig. 3.—Calibration for selected spectral indices. The dashed lines represent the first-order fit for each spectral index. Top left: He i + Fe i k4922; the indexreaches a maximum at a spectral type B2 because of the absorption of the He i k4922 line and then diminishes up to A0 and changes slope with the the onset of Fe ik4925 absorption. Top right: Ca i + Fe i k5270; the index has a monotonic behavior from B5 to K5. Bottom left: G band; the index has a monotonic behavior in thespectral range F0–K0. Bottom right: H�; the index shows a bimodal behavior. It has a positive slope from B0 to A0, following the increase in absorption in theBalmer lines, then decreases for later spectral types up to late G.

HERNANDEZ ET AL.1686 Vol. 127

emission (Herbig 1960; Hamann & Persson 1992; Waters &Waelkens 1998). These objects (sometimes called continuumstars because of the presence of strong, nonphotosphericcontinuum emission) cannot be classified because of the al-most complete absence of photospheric absorption features(e.g., MWC 137 in Fig. 1).

The spectral classification scheme described in x 3 is de-signed to largely avoid problems caused by nonphotospheric(emission and absorption line) contributions. We achieve thisby relying on many indices that are sensitive to TeA and re-quiring that the various spectral types calculated from eachindex agree with the others; wildly discrepant values arerejected, and a weighted mean spectral type is obtained. Inthis way, anomalous values for indices such as Ca ii k3933(index 1), He i + Fe i k4922 (index 19), He i + Fe i + Ti i k5016(index 20), and Na i + Ti i k5890 (index 29), which could beaffected by a nonphotospheric contribution, can be detectedand not included in the calculation of the final spectral type.

Although stellar rotation can in principle affect line indicesover narrow bandpasses (Mora et al. 2001; Gray et al. 2001), itis unlikely to bias our determinations because our bandpassesare wide; our spectral resolution �300 km s�1 at 6000 A islarger than the typical rotational velocities of HAeBe stars,�225 km s�1 (Finkenzeller 1985; Bohm & Catala 1995).

A further complication can arise when a star is a member ofa spectroscopic binary system. The spectroscopic binary fre-quency for HAeBe stars is larger than 35% (Corporon &Lagrange 1999), so there is a finite possibility of observinga combined spectrum in our sample. However, the primarycomponent tends to be more luminous than its companion(Corporon & Lagrange 1999), so we expect that the indiceswe use to classify the stars are probably dominated by featuresof the primary star in most cases. Still, there are a few likelycases of composite spectra.

Out of the 75 stars for which we have spectra, we havedetermined spectral types for 58 objects, or 77% of the sam-ple. Of these 58 stars, 46 have a spectral type earlier than F7.Within this early spectral type sample, 39 are cataloged asHAeBe (see x 4.2); the remainder have an uncertain evolu-tionary status and are discussed in x 4.3. There are 12 stars thatwe classified as F7 or later and discuss in x 4.4. The spectra ofseven stars were too veiled for spectral classification; these arediscussed in x 4.5. Finally, we could find no consistent spectraltype from the various indices for 10 objects of the sample,which are discussed in x 4.6.

4.2. Herbig Ae/Be Stars

In Table 2 we show the spectral types determined for the 39HAeBe stars in our sample. About 80% of the stars have anerror of P2.5 spectral subtypes. Stars with high reddeningtend to have larger errors because high extinction makes itdifficult to measure indices at short wavelengths.

In Figure 5 we compare spectral types derived in this workwith those published previously for different subsets of ourentire sample (Cohen & Kuhi 1979; Finkenzeller 1985;Hillenbrand 1995; Mora et al. 2001). Our determinationscorrelate best with those from high-resolution spectra by Moraet al. (2001; x 1). Only the HAeBe star VV Ser (HBC 282)differs by more than three subclasses in this comparison. Thepresence of lines of He i kk4026, 4144, 4387, 4471, and 5876favors our determination. Mora et al. (2001) cite a large un-certainty in their determination (more than five subclasses),perhaps in part due to difficulties in typing at high resolutionfor a rapidly rotating star (Vrot > 229 km s �1).

There are large discrepancies in the spectral types deter-mined for some stars. The spectral types published for LkH�208 (HBC 193) range from B5 to F0; our result based on theCa i kk5270, 5589, Fe i k5079, and Mg i k5711 lines isA7� 3 subclasses. We find in our spectra that some indicesare contaminated by nonphotospheric Fe ii (42)7 absorptionfeatures. This could explain the large discrepancies betweenvarious authors. The star LkH� 234 (HBC 309), cataloged byCohen & Kuhi (1979) as O9, shows enough neutral heliumlines in its spectrum to favor a spectral type B7� 3.5 sub-classes, which is in better agreement with Mora et al. (2001)and Finkenzeller (1985). Stars LkH� 338 (HBC 196) andLkH� 339 (HBC 197) show the largest discrepancies betweenthe spectral types given by Cohen & Kuhi (1979) and ourdetermination. Cohen & Kuhi (1979) classified LkH� 338 andLkH� 339 as F2, but our spectra show no evidence of the Gband or metallic lines expected if this were the appropriatespectral type.

4.3. Objects with Uncertain Evolutionary Status

In this subsection we analyze objects that need further studyin order to clarify their evolutionary status. In Table 3 andFigure 5 we compare spectral types derived in this work withthose published previously for different subsets of our entiresample (Cohen & Kuhi 1979; Finkenzeller 1985; Hillenbrand1995; Mora et al. 2001). Some stars have already beenrejected by some authors as members of the HAeBe class.LkH� 341 (HBC 201) was rejected by The et al. (1994) be-cause of the absence of excess at far-infrared (FIR) bands.This star was classified by Cohen & Kuhi (1979) as B3, butin our spectra we clearly detect metallic lines including

7 The multiplet number of the element is given in parentheses.

Fig. 4.—Comparison of spectral types determined in this work with thosefrom the literature for main-sequence (dots) and giant (triangles) standardstars. The points follow well a straight line of slope unity, with small scatter,indicating that our calibration is largely independent of surface gravity (forluminosity classes V and III.) Error bars indicate the uncertainty obtained fromour classification scheme.

SPECTRAL ANALYSIS OF HERBIG Ae/Be STARS 1687No. 3, 2004

TABLE 2

Stars Classified as HAeBe

HBC Name Spectral Type Error Str72 CK79 Fink84 Fink85 Hill95 Mora01

3.......... V633 Cas B9 2.5 . . . B3 a Ae A5 . . .

78........ AB Aur A1 1.5 B9 A0 A0 B9 A0 a

154...... T Ori A0 2.5 . . . A5 A5 A3 A2 A3 IV

170...... RR Tau A0 2.0 A3 A6 a A3–A5 A3 A0 IV

192...... HD 250550 B9 1.5 . . . B6 A0 B9 B9 . . .

193...... LkH� 208 A7 3.0 B8 F0 a B5–B9 A2 . . .

196...... LkH� 338 B9 3.5 . . . F2 . . . . . . . . . . . .197...... LkH� 339 A1 3.0 . . . F2 . . . . . . . . . . . .

219...... V590 Mon B7 2.0 A2 B9 a B8 B7 . . .

282...... VV Ser B6 2.0 . . . a a B1–B3 B9 A0 V

284...... AS310 NW B1 2.0 . . . B0 a B–A . . . . . .293...... PX Vul F3 1.5 . . . F5 . . . . . . . . . F3 V

305...... LkH� 324 B8 2.5 . . . B5 . . . . . . . . . . . .

309...... LkH� 234 B7 3.5 B5 O9 a B5–B7 B3 B5 V

310...... BD +46�3471 A0 1.0 A2 . . . A4 A0 A0 . . .

313...... LkH� 233 A4 3.0 A7 A7 a A7 A5 . . .

324...... MC 1 A7 2.5 a . . . A5 . . . . . . . . .

329...... VX Cas A0 1.5 . . . . . . . . . . . . . . . A0 V

334...... RNO 6 B3 2.5 . . . . . . . . . . . . . . . . . .

348...... IP Per A6 2.0 . . . . . . . . . . . . . . . . . .

350...... XY Per EW A5 1.5 . . . . . . a A2II–B6 . . . A2 IV

373...... V892 Tau B8 3.0 . . . . . . a A0 A6 . . .430...... UX Ori A3 2.5 . . . . . . . . . . . . A3 A4 IV

451...... HD 245185 A1 2.0 . . . . . . A0 A5 A1 . . .

464...... CQ Tau F3 2.0 . . . . . . . . . . . . . . . F5 IV

492...... p26887 A6 3.0 . . . . . . . . . . . . . . . . . .

493...... V350 Ori A1 2.5 . . . . . . . . . . . . . . . A2 IV

528...... LkH� 215 B6 2.5 B7 B1 a B7–B8 B7 . . .

529...... HD 259431 B6 2.5 B5 A0 B2 B6 B5 . . .548...... LkH� 218 A0 2.0 . . . B6 a B6 A0 . . .

551...... LkH� 220 B8 2.0 . . . B5 a B5 . . . . . .

686...... WW Vul A3 2.0 . . . . . . . . . . . . . . . A2 IV

689...... V1685 Cyg B3 2.0 B2 . . . B3 B2 B2 B2

705...... LkH� 147 B2 3.5 . . . a . . . . . . . . . . . .

726...... HD 200775 B3 1.0 B5 . . . B2.5 B3 B3 . . .

730...... BD +65�1637 B4 1.0 B2 . . . B3 B5 B3 . . .734...... BH Cep F5 2.0 . . . . . . . . . . . . . . . F5 III

735...... BO Cep F4 1.0 . . . . . . . . . . . . . . . F5 V

736...... SV Cep A0 1.5 . . . . . . . . . . . . . . . A2 IV

Notes.—(Str72) spectral type from Strom et al. 1972; (CK79) spectral type from Cohen & Kuhi 1979; (Fink84) spectral type from Finkenzeller &Mundt 1984; (Fink85) spectral type from Finkenzeller 1985; (Hill95) spectral type from Hillenbrand 1995; (Mora01) spectral type from Mora et al.2001.

a Star observed but spectral type not assigned.

TABLE 3

Objects with Uncertain Evolutionary Status


7.......... LkH� 201 B2 2.5 . . . B3 . . . . . . . . . . . .160...... PQ Ori F3 1.5 . . . F5 . . . . . . . . . . . .

201...... LkH� 341 F3 2.5 . . . B3 . . . . . . . . . . . .

281...... LkH� 118 B1 2.0 . . . a a B5 . . . . . .

297...... V751 Cyg A0 2.5 . . . A5 . . . . . . . . . . . .314...... LkH� 350 B8 3.0 B5 . . . . . . . . . . . . . . .

482...... BN Ori F4 2.0 . . . . . . . . . . . . . . . . . .

Notes.—(Str72) spectral type from Strom et al. 1972; (CK79) spectral type from Cohen & Kuhi 1979; (Fink84) spectral type fromFinkenzeller & Mundt 1984; (Fink85) spectral type from Finkenzeller 1985; (Hill95) spectral type from Hillenbrand 1995; (Mora01)spectral type from Mora et al. 2001.

a Star observed but spectral type not assigned.

Ca i kk4226, 5270, 5589, Fe i kk4387, 4922, and Mg i k5711,in addition to the G band, indicating a later spectral type. Ouranalysis from all these indices yield the same spectral typeF3� 2.5 subclasses. Using near-IR photometry from Cohen &Kuhi (1979), we derive for LkH� 341 a small near-IR excessusing a standard extinction law (RV ¼ 3:1), but this excessdisappears if we instead use an extinction law with RV ¼ 5:0(see x 6). This behavior is also observed in the star LkH� 118(HBC 281) when using JHKLmagnitudes from de Winter et al.(2001). This object was also rejected by The et al. (1994) asHAeBe. The spectral type for LkH� 118 differs by more thanfour subclasses from that given by Finkenzeller (1985). Webased our result on the He i and He ii lines.

LkH� 201 (HBC 7) and LkH� 350 (HBC 314) werecataloged by Herbig & Bell (1988) as possible background Bestars and rejected by The et al. (1994) as HAeBe because ofthe absence of excess at FIR bands. However, these stars showsome characteristics typical of PMS stars. We found emissionat H� , H�, and Fe ii (37, 38, 40, 42, 49, and 74) in LkH� 201(x 5). Similarly, LkH� 350 exhibits H� and H� in emission,

in addition to some abnormal absorption features due to thediffuse interstellar bands (DIB), located at 5780, 5796, and6283 A, and to the Na i doublet (k5890; Miroshnichenko et al.2001). However, emission in the Balmer lines and Fe ii canalso be found in more evolved Be stars (Miroshnichenko et al.2003). Both stars exhibit high reddening (AV > 5) and ananomalous extinction law (RV > 3:1, x 6), which would beexpected if these stars are embedded in a molecular cloud.

It is not clear if BN Ori (HBC 482) is a PMS object. It wasrejected by The et al. (1994) as an HAeBe object because ofthe lack of FIR excess. We do not detect excess at J, H, andK bands, using Two Micron All Sky Survey (2MASS) pho-tometry. However, some studies of this star propose a PMSstatus, suggesting that it could be a UX Ori object (Marconiet al. 2001) or that it has experienced an FU Ori type outburst(Shevchenko et al. 1997).

PQ Ori (HBC 160) does not show significant differencesfrom an F3 main-sequence star. In addition, it lacks emissionlines and near-IR excess, as determined from 2MASS pho-tometry. V751 Cyg (HBC 297) has been cataloged as a

Fig. 5.—Comparison of spectral types determined in this work for HAeBe with previously published values. References are shown in Tables 2 and 3. Verticalerror bars are the uncertainties derived from our spectral-type determination as explained in the text. For comparison, we show in each panel the line with slope 1.The largest scatter is observed when comparing our results with Cohen & Kuhi (1979). The best correspondence is obtained when comparing our spectral types withthose of Mora et al. (2001).


cataclysmic variable by Robinson (1973), Downes et al.(1995), and Echevarrıa et al. (2002). This star falls below themain sequence in the H-R diagram (x 7).

4.4. Stars with Spectral Types Later than F

In Table 4 we list the 12 stars of our sample with spectraltypes F7 or later. Although they appear in the HBC (Herbig &Bell 1988) as earlier than G0, our classification schemeyielded types as late as G4 for some of them.

For V1686 Cyg there is no agreement on the spectral typeassigned by different authors, the values ranging from B2 (Theet al. 1994) to F2 (Terranegra et al. 1994). In particular, Moraet al. (2001) assigned a spectral type A4 with more than fivesubclasses of spectral type error. In their multiepoch spectra,kindly provided to us by B. Merin, absorption features tend tovary significantly in time, which could explain why a reliablespectral type is rather difficult to determine. The strongestDIBs (kk 5780, 5797, 6284, and 6614) are clearly seen in theEXPORT spectra. However, our spectra look very different,the presence of the G band and metallic lines (Fe i, Ca i, Ca ii,and Mn i) are more consistent with a spectral type F9. Thisstar exhibits large photometric variations. The brightness ofV1686 Cyg decreased progressively by more than 4.5 mag ina period of 7.5 yr, then it brightened by 4 mag in about 4 yr.The decrease in brightness was accompanied by a reddeningof the star (in V�I ), suggesting that it could be caused bydusty material not too far from the star. In addition to this long-term variation, V1686 Cyg shows changes in brightness ofmore than 2 mag on timescales of roughly 2 months. Theseshorter term photometric variations could be related to thespectroscopic variability, but this remains to be investigated;this object deserves further study to clarify the physicsinvolved in its behavior.

The spectra of the stars LkH� 349 (HBC 308) and RNO 63(HBC 518) show P Cygni profiles at H� , indicating ejection ofmaterial at velocities larger than 300 km s �1 (see also Hessmanet al. 1995). In VSB 2 (HBC 531) H� is seen in absorption, butitsWk is smaller than in a standard star of the same spectral type,which may be the result of unresolved emission.

Among our sample of later type objects, only W84 (HBC217) and V360 Mon (HBC 231) show Wk of H� in emissionk10 A and emission in H�. The [O i] k6300 and Fe ii (42)lines were not detected in any of the objects listed in Table 4.

One characteristic of these later-type objects is that the Li ik6708 absorption line seems to be present in most of thesestars. Li i has been used in the past as an indicator of youth inintermediate- and late-type stars (Strom et al. 1989). However,we caution that Li i k6708 in absorption cannot be taken as anindicator of the PMS nature of stars earlier than mid-K, be-cause the shallow depth of the convective zone in these starscan allow them to reach the main sequence with a non-negligible amount of their primordial lithium content (Bricenoet al. 1997). Therefore, the presence of lithium in absorption(Wk > 0:1 A) in this spectral type range is only evidence thatthese objects are not old disk stars.

4.5. Continuum Stars

In a subset of objects we found essentially no absorptionfeatures at our resolution, so they could not be assigned aspectral type. Most of the lines appear in emission. Thesestars are the continuum stars, and they are listed in Table 5.In Table 6 we present measurements of the Wk of emissionlines seen in these stars.Previous attempts to assign spectral types to these stars are

given in Table 5, but they should be treated with caution giventheir high degree of veiling. The stars MWC 1080 and PV Cepshow strong P Cygni profiles in several Balmer lines that areresolved even at our low resolution; this suggests the presenceof strong winds or outflows. High-resolution spectroscopy ofsome of these stars (Fernandez 1995; Corcoran & Ray 1997;Parsamian et al. 1996; Magakian & Movsesian 2001) confirmthe P Cygni nature of the line profiles. In addition, they tend tobe associated with strong molecular outflows and/or opticaljets (Wu et al. 1996; Arce & Goodman 2002; Magakian &Movsesian 2001; Gomez et al. 1997) pointing to the youth ofthese objects. Recent high-resolution observations of LkH�101 show extremely peculiar double line profiles, unlike thosefound in any other HAeBe star; this suggests that LkH� 101may not belong to the HAeBe class (G. Herbig 2003, privatecommunication.)

4.6. Stars with Unknown or Uncertain Spectral Types

We found cases in which it was impossible to assign aunique spectral type to the object. These objects are discussedbelow individually and are listed in Table 7, together withprevious spectral type determinations that have appeared in

TABLE 4

Stars with Spectral Types Later than F7


217............ W84 F7 2.0 . . . F8 . . . . . . . . . . . .

222............ W108 F7 2.0 . . . F9 . . . . . . . . . . . .

231............ V360 Mon G4 2.5 . . . F8 . . . . . . . . . . . .308............ LkH� 349 F9 2.0 . . . F8 . . . . . . . . . . . .

432............ P102 F7 2.0 . . . . . . . . . . . . . . . . . .

436............ RY Ori F7 2.5 . . . . . . . . . . . . . . . F5 IV

442............ P1394 F8 1.5 . . . . . . . . . . . . . . . . . .460............ MV Ori G1 3.0 . . . . . . . . . . . . . . . . . .

518............ RNO 63 F7 2.5 . . . F6 . . . . . . . . . . . .

531............ VSB 2 G1 2.0 . . . . . . . . . . . . . . . . . .

535............ W121 G2 2.0 . . . . . . . . . . . . . . . . . .690............ V1686 Cyg F9 3.5 . . . . . . . . . . . . B5 A4 V

Notes.—(Str72) spectral type from Strom et al. 1972; (CK79) spectral type from Cohen & Kuhi 1979; (Fink84) spectral type from Finkenzeller &Mundt 1984; (Fink85) spectral type from Finkenzeller 1985; (Hill95) spectral type from Hillenbrand 1995; (Mora01) spectral type from Mora et al.2001.

HERNANDEZ ET AL.1690

TABLE 5

Continuum Stars

HBC Name Str72 CK79 Fink84 Fink85 Hill95 Mora01

40........... LkH� 101 . . . C . . . . . . . . . . . .

164......... V380 Ori B8 B9(C) A1e . . . B9 . . .

199......... MWC 137 . . . . . . Cont + e Cont + e B0 . . .207......... R Mon . . . B0 e + s . . . B0 B8 IIIev

317......... MWC 1080 . . . B0(C) eq . . . B0 . . .

330......... V594 Cas B8 O9.5 B8, B9eq B8eq B8 . . .

696......... PV Cep . . . . . . . . . . . . . . . . . .

Notes.—(Str72) spectral type from Strom et al. 1972; (CK79) spectral type from Cohen & Kuhi 1979; (Fink84)spectral type from Finkenzeller & Mundt 1984; (Fink85) spectral type from Finkenzeller 1985; (Hill95) spectral typefrom Hillenbrand 1995; (Mora01) spectral type from Mora et al. 2001.

TABLE 6

Emission Lines in Continuum Stars

Element

k(A) LkH� 101 V380 Ori MWC 137 R Mon MWC 1080 V594 Cas PV Cep

Ca ii (1) .................................. 3934 . . . �6.5 . . . �1.8 �0.7a �0.9 . . .

H� + Ca ii (1) ......................... 3969 . . . �4.6 �5.6 . . . �0.9a �1.5 . . .He i (18)................................. 4026 . . . . . . �1.4 . . . . . . . . . . . .

H� ........................................... 4102 �6.7 �0.7 �7.9 . . . �1.9a �1.5a . . .

Fe ii (27, 28) .......................... 4176 . . . �4.8 �1.6 �2.1 �3.0 �1.0 . . .

Fe ii (27) ................................ 4233 . . . �2.7 �1.1 �2.0 �2.0 �1.0 . . .Fe ii (27) + Ti ii (41) ............. 4301 . . . �6.9 . . . �3.7 �3.0 �1.8 . . .

Ti ii (41) ................................. 4313 . . . . . . . . . . . . �0.8 . . . . . .

H� .......................................... 4340 �9.0 �3.6 �14.9 �1.7a �4.5a �2.9a . . .

Fe ii (27) ................................ 4352 . . . . . . . . . �2.0 �1.9 . . . . . .Fe ii (27) + He i (51) ............. 4385 . . . �1.29 �0.9 �0.6 �1.3 �0.7 . . .

Ti ii (19) ................................. 4395 . . . �1.5 . . . �0.8 �0.9 �0.7 . . .

Fe ii (27) ................................ 4417 . . . �2.0 . . . �1.3 �0.9 . . . . . .

Ti ii (19) ................................. 4445 . . . �1.4 . . . . . . . . . . . . . . .Ti ii (31) ................................. 4468 . . . �1.1 . . . . . . �0.6 . . . . . .

He i (14)................................. 4472 . . . . . . �1.8 . . . . . . . . . . . .

Fe ii (37) ................................ 4491 . . . �1.6 �1.0 �1.0 �1.3 �0.3 . . .Fe ii (38) ................................ 4508 . . . . . . . . . . . . �0.7 �0.4 . . .

Fe ii (37, 38) .......................... 4521 �2.5 �5.0 �1.3 �2.4 �3.0 �1.2 . . .

Fe ii (37) ................................ 4534 . . . . . . . . . . . . �0.4 . . . . . .

Fe ii (38) ................................ 4549 �2.3 . . . . . . . . . �1.8 �1.3 . . .Fe ii (37) ................................ 4556 �2.6 �5.8 �2.1 �2.7 �2.9 �1.0 . . .

Fe ii (38) ................................ 4576 . . . . . . . . . . . . �0.7 �0.2 . . .

Fe ii (37, 38) .......................... 4584 �3.7 �4.9 �1.7 �2.5 �3.3 �1.6 . . .

Fe ii (38) ................................ 4621 . . . �1.9 . . . �0.5 �0.9 �0.2 . . .Fe ii (37) ................................ 4629 �2.3 �3.7 �1.5 �1.6 �2.2 �0.8 . . .

Fe i (37) ................................. 4667 . . . �1.1 . . . �0.3 �0.4 . . . . . .

Fe i (43) ................................. 4731 . . . �0.9 . . . �0.6 �0.4 . . . . . .Cr ii (30) ................................ 4824 . . . �0.9 . . . . . . �0.5 . . . . . .

H� .......................................... 4861 �37.7 �12.1 �54.4 �12.7 �20.9a �7.5a �15.2

Fe ii (42) ................................ 4924 �3.1 �6.8 �2.4 �3.5 �4.0a �2.8 �2.5

Fe ii (42) ................................ 5018 �4.3 �8.0 �3.4 �3.9 �4.9a �3.8 �3.0

Fe ii (42) ................................ 5169 �1.8 �9.1 �1.7 �4.5 �5.0a �4.2 �5.0

Ti ii (70) ................................. 5189 . . . �1.7 . . . �1.0 �0.8 �0.4 . . .

Fe ii (49) ................................ 5198 �2.2 �3.0 �0.9 �1.7 �2.0 �0.7 �3.2

Fe ii (49) ................................ 5235 �1.4 �4.1 �1.0 �2.1 �2.5 �1.28 . . .Fe ii (49) ................................ 5255 . . . . . . . . . . . . �0.4 . . . . . .

Fe ii (48) ................................ 5264 . . . . . . . . . �1.1 �0.5 . . . �2.0

Fe ii (49) + Cr ii (43)............. 5276 �3.2 �5.3 �1.6 �2.9 �3.0 �1.4 . . .Fe ii (48, 49) + Cr ii (43) ...... 5317 �4.7 �6.7 �2.4 �3.1 �4.2 �2.1 . . .

Fe ii (49) ................................ 5326 . . . �0.6 . . . . . . . . . . . . �3.4

Fe ii (48) ................................ 5338 . . . �1.0 . . . �0.4 �0.5 . . . . . .

Fe ii (48) ................................ 5363 . . . �2.3 �0.8 �1.0 �1.1 �0.6 . . .Fe ii (49) ................................ 5425 . . . �1.7 . . . �0.5 �0.6 �0.3 . . .

Fe ii (55) ................................ 5535 �1.6 �2.6 �0.7 �1.4 �1.1 �0.6 . . .

[N ii] (3) ................................. 5755 . . . . . . �0.7 . . . . . . . . . . . .

the literature. Short comments about each star as well as arough estimate of spectral type, based on any absorptionfeatures visible, are given in the last column of the table.

MacC H12.—No reliable spectral type could be obtainedfor this star. The high reddening seen toward this star pre-cludes the use of the blue region of the spectrum for spectraltyping, especially at wavelengths below 5500 A. However, theG band and Fe i k5329 seem to be present, which suggests aspectral type around F4.

HK Ori.—This star has multiple emission lines, includingFe ii (27, 37, 67, 42, 48, and 49), Ti ii (41 and 69), Cr i (31),and Cr ii (43), which contaminate most of the indices locatedat wavelengths below 5500 A. In contrast, between 5500 and6300 A the spectrum seems to be free of emission features.The spectral type we derive from this wavelength region is F2,

based on the Ca i k5589, Fe i + Mg i k5711, and Mn i k6015indices. However, Ca i k6162 is more consistent with a G0star, while the Ca ii K line is similar to that expected in an A2main-sequence star. This behavior was already reported byStrom (1983), who found a variation of spectral type rangingfrom early A near 4000 A to late F around 6500 A. Onepossibility is that we are observing a combined spectrum(A star + F star); the multiplicity of this star is well known(Leinert et al. 1997; Pirzkal et al. 1997; Corporon & Lagrange1999), and the presence of Li i k6707 suggests that thecompanion is likely to be a star with a spectral type later thanF7 (Wk [Li i] = 0.2 A)BF Ori.—When we attempt to derive a spectral type for this

star using the indices in Table 1, we obtain an unreasonablylarge error, given the quality of the spectrum. The indices

TABLE 6—Continued

Element

k(A) LkH� 101 V380 Ori MWC 137 R Mon MWC 1080 V594 Cas PV Cep

Ca i (47)? ............................... 5857 �1.3 �0.4 . . . . . . . . . . . . . . .He i (11)................................. 5876 �3.8 �0.6 �5.6 . . . . . . . . . �0.8

Na i (1)................................... 5892 . . . �2.9 . . . �0.6 . . . �0.6 . . .

Cr i (7) + Si ii (4)?................. 5979 �1.6 . . . �0.3 . . . . . . . . . . . .

Cr i (7)?.................................. 5992 . . . �0.8 . . . �0.2 �0.3 . . . . . .Fe ii (74) ................................ 6149 . . . �2.2 . . . �0.8 �1.0 �0.4 . . .

Fe ii (74) ................................ 6238 . . . �2.1 . . . . . . �0.8 �0.3 . . .

Fe ii (74) + Si ii (2)................ 6248 �2.4 �3.0 �1.2 �1.3 �1.3 �0.6 �2.6

[O i] (1) .................................. 6300 �3.9 . . . �0.8 �3.4 �0.4 �0.2 �66.4

Fe i (1016)?............................ 6317 �4.4 �0.6 �1.1 �0.4 �0.7 �0.3 . . .

Si ii (2) ................................... 6347 �2.8 �3.0 �0.8 . . . . . . �0.4 . . .

[O i] (1) .................................. 6363 �1.2 . . . . . . �1.6 . . . . . . �21.6

Fe ii (40) + Si ii (2)................ 6370 �1.6 �1.3 �0.6 . . . . . . �0.3 . . .

[V i] (13)? .............................. 6382 �4.1 �0.4 . . . �0.5 �1.0 �0.3 . . .

Fe ii (74) ................................ 6417 . . . �1.0 �0.2 �0.4 �0.5 �0.2 �1.4

Fe ii (40) + [V ii] (13) ? ........ 6433 �0.3 �1.3 �0.3 �0.4 �0.6 �0.3 �2.2

Fe ii (74) + [V ii] (13) ? ........ 6456 �2.2 �3.64 �0.6 �1.1 �1.6 �0.8 �1.5

Ca i (18)? ............................... 6492 �2.2 �0.7 �0.6 . . . . . . . . . �1.5

Fe ii (40) ................................ 6516 �1.2 �2.7 �0.9 �0.6 �1.0 �0.3 �2.1

[N ii] (1) ................................. 6548 . . . . . . . . . . . . �6.2 . . . �4.5

H� .......................................... 6563 �464.1 �75.3 �397.0 �106.8 �135.9 �67.9a �125.0

He i (46)................................. 6678 �1.5 �0.5 �2.2 . . . . . . . . . �1.0

[S ii] (2).................................. 6717 . . . . . . . . . �0.6 . . . . . . �7.7

[S ii] (2).................................. 6731 . . . . . . . . . �1.1 . . . . . . �14.4

He i (10)................................. 7066 �2.8 . . . �2.7 . . . . . . . . . �1.0

a P Cygni profile.

TABLE 7

Stars Not Classified

HBC Name Str72 CK79 Fink84 Fink85 Hill95 Mora01 Comments

1................. MacC H12 . . . A5–F . . . . . . . . . . . . High reddening; F4?

94............... HK Ori B7 A4 A–F A4 A3 G1 V Emission lines; binary system; A2–G0

169............. BF Ori . . . A0 . . . A–F A7 A2 IV Nonphotospheric absorption lines; A0�A5

202............. VY Mon . . . O9 . . . . . . . . . A5 Vep Emission lines + High reddening ; F2 ?

273............. KK Oph . . . . . . A5–A7 A5–A7 A3 A8 V Emission lines; binary system; A0–F0

321............. MacC H4 . . . A9 . . . . . . . . . . . . High reddening; earlier than A5

325............. V376 Cas . . . B5 . . . . . . F0 . . . High reddening; A3–F2

716............. V1493 Cyg . . . A2 . . . . . . . . . . . . High reddening; Fe ii (42) in absorption; A1–A9

717............. LkH� 168 . . . F2 . . . . . . . . . . . . High reddening; F0 ?

742............. MacC H1 . . . B8:e . . . . . . . . . . . . High reddening + emission lines; B2�A0

Notes.—(Str72) spectral type from Strom et al. 1972; (CK79) spectral type from Cohen & Kuhi 1979; (Fink84) spectral type from Finkenzeller & Mundt 1984;(Fink85) spectral type from Finkenzeller 1985; (Hill95) spectral type from Hillenbrand 1995; (Mora01) spectral type from Mora et al. 2001.


Fe i + Ti i k5079, Ca i + Fe i k5270, and Ca i k5589 indicatethat BF Ori has a spectral type A0–A9. However, whencompared with an A7 standard, this star shows strong ab-normal absorption in some features, which contaminate mostof the spectral indices (x 5).

VY Mon.—The large reddening (AV > 7:0; Casey &Harper 1990; ), together with the presence of emission lines,results in a large uncertainty in the spectral type we obtainfrom indices at wavelengths less than 5500 A. The indicesCa i k5589, Mg i k5711, and Mn i k6015 yield a spectraltype F2� 5 subclasses, which is consistent with the pres-ence of the G band.

KK Oph.—The emission lines present in this object are sonumerous that they affect most of the indices we have defined inTable 1 (x 5). Binarity is reported in this star by Bailey (1998),Leinert et al. (1997), and Pirzkal et al. (1997). The companion isprobably a T Tauri star; however, in contrast to HK Ori, no Li iabsorption is seen in our spectrum in spite of a good S/N. Theabsence of the G band, and of the Ca i k6162, Ca i k5589, Mn ik6015, Mg i k5711, He i k6678, and He i k7066 lines suggest aspectral type between A0 and F0.

MacC H4.—This star has a large extinction, which results ina poor S/N at the blue end of our spectrum, precluding anaccurate determination of spectral type. Some helium lines areobserved marginally at wavelengths less than 5500 A. The Ca i,Fe i, and Mg i are clearly absent in the red part of the spectrum;this could indicate a spectral type earlier than A5. Emissioncomponents are observed in the lines He i kk5876, 6678, [O i]k6300, and the Balmer lines, H� , H�, and H� (x 5).

V376 Cas.—Indices Fe i k4532, He i + Fe i k4922, Mg i

k5173, Ca i k5270, Ca i k5589, and Fe i + Mg i k5711 lead to aspectral type A8, with a large uncertainty of five subtypes. TheWk of H� and H� are characteristic of stars with spectral typesaround A5. Indices with wavelengths below 4500 A could notbe used because of the large reddening in our spectrum. Thisobject has the largest linear polarization observed so far in anyPMS star. Although some authors have suggested that thereflected light can be produced by a circumstellar disk observednearly edge-on (Asselin et al. 1996; Hajjar & Bastien 2000), themorphology of the reflection nebula is completely differentfrom other edge-on disk systems, resembling instead thatof objects with outflow holes, as discussed by Whitney &Hartmann (1993). Indeed, Hajjar & Bastien (2000) argue thatthis object is an extreme Class I object, i.e., a protostar with anopaque infalling envelope. Because the star is not observeddirectly, but only in scattered light, reddening corrections andthus luminosity estimates are extremely uncertain. This objectand the HAeBe star V633 Cas are the brightest objects in theisolated molecular cloud L1265.

V1493 Cyg.—The high reddening of this star and the pres-ence of nonphotospheric absorption features (x 5) complicateattempts to determine a reliable spectral type. Still, a spectraltype A1–A9 is derived from the indices Ca ii k3933, Fe i k4787,Fe i k5079, Ca i + Fe i k5270, and Ca i k5589. Weak emissionis detected in the forbidden line [O i] k6300.

LkH� 168.—Published spectral types range from A3(Fernandez et al. 1995) to F6 (Terranegra et al. 1994). Becauseof the high reddening, the blue part of the spectrum is noisy,but the absence of the G band and the weakness of the metalliclines in the red part of the spectra indicate that the spectraltype is probably earlier than F0. According to Herbig & Bell(1988) this object may be a background Be star.

MacC H1.—Lines He i kk4387, 4471, 5876, and 7066 in-dicate a spectral type between B2 and A0, and the absence of

the G band seems consistent with this estimate, but the highreddening and the presence of numerous emission lines do notallow us to obtain a more reliable spectral type. This objectwas included by The et al. (1994) as an emission-line star butnot considered as HAeBe. However, we observe emission inthe Balmer and Fe ii (38, 37, 42, and 49) lines in addition toP Cygni profiles in H� and H�. More data are necessary tostudy the evolutionary status of this object.

5. NONPHOTOSPHERIC FEATURES

As already mentioned, HAeBe stars exhibit a number ofspectral features in emission or absorption, not seen in stan-dard stars of the same spectral type, that suggest their origin isoutside the stellar photosphere. In Table 8 we list these non-photospheric features measured in our sample, together withtheir Wk. In this table we include a footnote indicating theform of the H� profile in our spectra. When asymmetries areseen, they could be produced by material moving at velocitieslarger than our spectral resolution (300 km s�1). However,given our low spectral resolution, we cannot say anythingconclusive about the shape of the lines. All stars, by definitionof the class, show H� in emission. The distribution of the Wk(H� ) is shown in the top panel of Figure 6, where HAeBestars, continuum stars, and stars with spectral types later thanF are shown separately. It can be seen that the continuum starshave the largest Wk (H� ) suggesting that they are the youngestof the sample, since activity, powered either by disk accretionor by stellar dynamos, is expected to decrease with age(Hartmann et al. 1998; Skumanich 1972). Among the sample,53% show emission in H� and 15% in H�, although these arelower limits, since at our resolution we could not detect anemission component superimposed on an absorption profile.However, we find that among the 39 HAeBe stars in Table 2,95% have Wk (H�) and 56% have Wk (H�) smaller than thatcorresponding to their spectral types, indicating that theselines are being filled in to some degree.

Other emission lines present in the spectra are forbiddenlines of [O i] and [S ii]. These lines are thought to be formed inextended, low-density, collisionally excited gas (Finkenzeller1985). We find that 31 stars out of a total of 63 (late F stars arenot included) exhibit Wk of [O i] k6300 in emission largerthan the typical lowest emission we can detect in our thesample, 0.1 A. The bottom panel of Figure 6 shows the dis-tribution of Wk of [O i] k6300. For a subset of 17 stars, the[O i] k6363 line could be measured. The mean ratio of[O i] k6300/k6363 for HAeBe and continuum stars is 2.7,close to the optically thin ratio (Osterbrock 1989).

Emission in multiplets of Fe ii, most conspicuously in mul-tiplets 42 and 49, is found in 25 objects, 33% of the sample.When the emission is present, it is related to the Balmer emis-sion. As shown in Figure 7, a correlation exists between theWkof Fe ii k5169 and the Wk of H� . The correlation coefficient is0.74. In general, Fe ii (42) is observed in emission only if [O i]k6300 is also present, except for BD +46�3471, MacC H1, andLkH� 218. However, Bohm&Catala (1994) detected emissionat [O i] k6300 in two of these stars: in BD +46�3471 with aWk ¼ �0:1 A, and in LkH� 218 with a Wk ¼ �0:2 A. Still,no clear correlation between the strengths of the lines isfound; the correlation coefficient between the equivalent widthsis 0.43.

A fraction of stars show emission in He i lines. Comparisonof Tables 6 and 8 shows that this emission is more frequentamong the continuum stars than in normal HAeBe stars (57%for the continuum stars, 18% for HAeBe stars).


TABLE 8

Nonphotospheric Spectral Features

HBC Name Typesa

H�

m = 1

k0 = 6563

H�

m = 1

k0 = 4861

[O i]

m = 1

k0 = 6300

[O i]

m = 1

k0 = 6363

[S ii]

m = 2

k0 = 6717

[S ii]

m = 2

k0 = 6731

[N ii]

m = 1

k0 = 6583

Ca ii

m = 1

k0 = 3934

Fe ii

m = 42

k0 = 4924

Fe ii

m = 42

k0 = 5018

Fe ii

m = 42

k0 = 5169

Fe ii

m = 49

k0 = 5198

Fe ii

m = 49

k0 = 5235

Fe ii

m = 49

k0 = 5276

Fe ii

m = 48, 49

k0 = 5317

He i

m = 11

k0 = 5876

He i

m = 46

k0 = 6678

1........ MacC H12 n �31.5 �2.3 �3.9 �1.4 �1.6 �2.5 �2.2 . . . . . . . . . . . . . . . . . . . . . . . . �0.9 . . .

3........ V633 Cas h �56.2 �4.0 �1.9 �0.6 �0.3 �0.4 . . . . . . �1.3 �1.7 �1.8 �0.5 �0.4 �0.8 �0.8 . . . . . .

7........ LkH� 201 u �41.6 �3.5 . . . . . . . . . . . . . . . . . . . . . �0.4 �0.5 0 �0.3 �0.2 �0.4 . . . . . .78...... AB Aur h �28.2 �1.6 �0.1 . . . . . . . . . . . . . . . �0.3 �0.4 �0.4 . . . . . . . . . . . . �0.6 �0.30

94...... HK Ori n �49.0 �4.1 �1.2 �0.5 . . . . . . . . . . . . . . . �0.6 �1.7 . . . . . . �0.4 �0.6 . . . . . .

154.... T Ori h �21.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

160.... PQ Ori u . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .169.... BF Ori n �6.7 . . . . . . . . . . . . . . . . . . . . . +1.4 +1.7 +2.4 +0.3 +0.7 +1.0 +0.7 +0.9 . . .

170.... RR Tau h �25.7 . . . �0.3 . . . . . . . . . . . . . . . +0.6 +0.7 +1.0 . . . . . . . . . . . . . . . . . .

192.... HD 250550 h �24.8b �3.3b �0.1 . . . . . . . . . . . . . . . �0.2 �0.4 �0.5 . . . . . . . . . . . . . . . . . .193.... LkH� 208 h �4.9 . . . . . . . . . . . . . . . . . . . . . +0.5 +0.5 +0.9 . . . . . . . . . . . . . . . . . .

196.... LkH� 338 h �51.0 �3.4 �0.3 . . . . . . . . . . . . . . . �0.30 �0.75 �0.53 . . . . . . . . . . . . . . . . . .

197.... LkH� 339 h �19.4 �0.80 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . �0.3 . . .

201.... LkH� 341 u �23.5 �3.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .202.... VY Mon n �28.0 �0.6b �1.4 �0.4 �0.2 �0.4 . . . . . . �0.7 . . . �1.2 . . . �0.4 �0.6 . . . . . . . . .

217.... W84 f �10.3 �0.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . �0.4 . . .

219.... V590 Mon h �47.3 �0.5 �1.4 �0.5 . . . . . . . . . . . . . . . . . . �0.4 . . . . . . . . . . . . . . . . . .

222.... W108 f �1.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .231.... V360 Mon f �20.7 �1.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

273.... KK Oph n �59.4 �1.1 �2.2 �0.5 �0.2 �0.3 . . . �0.6 . . . �0.4 �1.4 . . . . . . . . . . . . . . . . . .

281.... LkH� 118 u �19.1 �2.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . �1.3 �0.6

282.... VV Ser h �61.1 �1.3 �0.6 . . . . . . . . . . . . . . . +0.2 +0.2 +0.3 . . . . . . . . . . . . . . . . . .

284.... AS 310 NW h �7.7 . . . . . . . . . �0.6 �0.9 �3.7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

293.... PX Vul h �6.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

297.... V751 Cyg u �7.9 �1.8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . �0.2 �0.70

305.... LkH� 324 h �15.3 �0.5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

308.... LkH� 349 f �0.3b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

309.... LkH� 234 h �68.9 �5.2 �0.9 �0.3 . . . . . . . . . . . . �1.2 �1.8 �2.1 �0.7 �0.7 �1.0 �1.3 . . . . . .

310.... BD +46�3471 h �18.6 �0.6 . . . . . . . . . . . . . . . . . . �0.1 �0.2 �0.2 . . . . . . . . . . . . . . . . . .313.... LkH� 233 h �20.5 �1.1 �0.7 �0.3 �0.4 �0.5 �0.5 . . . . . . . . . . . . . . . . . . . . . . . . �0.3 . . .

314.... LkH� 350 u �29.0 �3.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

321.... MacC H4 n �24.0 �4.3 �0.7 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . �0.6 �0.7

324.... MC1 h �16.9 . . . �0.6 �0.2 �0.2 �0.3 . . . . . . +0.6 +0.6 +1.2 . . . . . . . . . . . . . . . . . .325.... V376 Cas n �22.0 . . . �1.7 �0.8 �0.9 �1.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

329.... VX Cas h �19.2 . . . �0.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

334.... RNO 6 h �0.2 . . . . . . . . . �0.2 �0.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .348.... IP Per h �21.4 �0.9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . �0.5 . . .

350.... XY Per EW h <�4.7c . . . . . . . . . . . . . . . . . . . . . +1.0 +1.2 +1.8 . . . . . . . . . . . . . . . . . .

373.... V892 Tau h �17.8 . . . �0.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

430.... UX Ori h �2.3 . . . . . . . . . . . . . . . . . . . . . +0.8 +0.8 +1.1 . . . . . . . . . . . . . . . . . .432.... Par 102 f �6.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

436.... RY Ori f �7.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

442.... P1394 f �1.3b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

451.... HD 245185 h �21.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . �0.2 . . .460.... MV Ori f . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

464.... CQ Tau h �6.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1694

TABLE 8—Continued

HBC Name Typesa

H�m = 1

k0 = 6563

H�m = 1

k0 = 4861

[O i]

m = 1

k0 = 6300

[O i]

m = 1

k0 = 6363

[S ii]

m = 2

k0 = 6717

[S ii]

m = 2

k0 = 6731

[N ii]

m = 1

k0 = 6583

Ca ii

m = 1

k0 = 3934

Fe ii

m = 42

k0 = 4924

Fe ii

m = 42

k0 = 5018

Fe ii

m = 42

k0 = 5169

Fe ii

m = 49

k0 = 5198

Fe ii

m = 49

k0 = 5235

Fe ii

m = 49

k0 = 5276

Fe ii

m = 48, 49

k0 = 5317

He i

m = 11

k0 = 5876

He i

m = 46

k0 = 6678

482.... BN Ori u �1.3b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

492.... BD +26�887 h �3.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .493.... V350 Ori h �29.9 . . . �0.4 �0.2 . . . . . . . . . . . . +0.3 +0.5 +0.9 . . . . . . . . . . . . . . . . . .

518.... RNO 63 f �2.3b . . . . . . . . . . . . . . . . . . . . . . . . � . . . . . . . . . . . . . . . . . . . . .

528.... LkH� 215 h �25.7 �0.3 . . . . . . . . . . . . . . . . . . +0.7 +0.3 +0.3 . . . . . . . . . . . . . . . . . .

529.... HD 259431 h �57.5 �2.9 �0.5 �0.2 . . . . . . . . . �0.8 �0.3 �0.5 �0.9 �0.2 �0.4 �0.5 �0.6 . . . . . .531.... VSB 2 f <0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

535.... W121 f <0c . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

548.... LkH� 218 h �32.3 �1.4 . . . . . . . . . . . . . . . . . . �0.5 �0.6 �0.4 . . . . . . . . . . . . . . . . . .

551.... LkH� 222 h �54.5 �2.9 �0.3 �0.2 . . . . . . . . . . . . �0.5 �0.6 �0.5 . . . . . . . . . . . . . . . . . .686.... WW Vul h �14.4 . . . . . . . . . . . . . . . . . . . . . +0.6 +0.6 +0.7 . . . . . . . . . . . . . . . . . .

689.... V1685 Cyg h �108 �8.9 �0.9 �0.3 . . . . . . . . . . . . �0.3 �0.8 �1.22 �0.4 �0.5 �0.7 �0.7 . . . . . .

690.... V1686 Cyg f �3.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

705.... LkH� 147 h �26.9 �1.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .716.... V1493 Cyg n �9.5 . . . �0.2 . . . . . . . . . . . . . . . +0.7 +0.7 +1.0 . . . . . . . . . . . . . . . . . .

717.... LkH� 168 n �19.0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

726.... HD 200775 h �59.3 �2.7 �0.1 . . . . . . . . . . . . . . . . . . �0.1 �0.7 . . . . . . . . . . . . . . . . . .730.... BD +65�1637 h �28.0 �2.2 . . . . . . . . . . . . . . . . . . . . . �0.5 0.7 �0.3 �0.4 �0.4 �0.5 . . . . . .

734.... BH Cep h <�6.2c . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

735.... BO Cep h <�2.3c . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

736.... SV Cep h �12.1 . . . . . . . . . . . . . . . . . . . . . +0.4 +0.5 +0.6 . . . . . . . . . . . . . . . . . .742.... MacC H1 n �34.2b �4.1b . . . . . . . . . . . . . . . . . . �0.6 �1.0 �1.0 �0.5 �0.6 �0.4 �0.6 . . . . . .

Notes.—Here ‘‘m’’ indicates the multiplet of the element and k0 is the wavelength of the feature in angstroms.a Types are ‘‘h’’ HAeBe star; ‘‘f’’ late F star; ‘‘n’’ no spectral type could be assigned to this star; and ‘‘u’’ stars with uncertain evolutionary status.b P Cygni profile.c Double-peaked profile.

1695

While emission features are more easily distinguished, thenonphotospheric absorption features, sometimes referred to as‘‘shell’’ features, are more difficult to single out. However, inour classification scheme based on multiple indices, theseshell features tend to stand out as yielding spectral types in-congruent with those obtained from the rest of the indices. Themore readily identified shell features at our spectral resolutioncorrespond to lines of multiplet 42 of Fe ii. Other lines thatmay be affected by absorption external to the photosphere arethe Na i D line and He i k5876; the last when observed at highdispersion often has blueshifted components (e.g., MWC 1080He i profile in Hartmann, Kenyon, & Calvet 1993), arising inexpanding material around the star. However, at our resolutionindividual shell components cannot be picked out.

We have identified nonphotospheric absorption in the 42multiplet of Fe ii in 11 HAeBe stars and the star BF Ori(without spectral type assigned in this work), with the Wkshown in Table 8. Among the objects with Fe ii nonphoto-spheric absorption, the stars BF Ori (HBC 169), RR Tau (HBC170), LkH� 208 (HBC 193), VV Ser (HBC 282), UX Ori(HBC 430), WW Vul (HBC 686), and SV Cep (HBC 736) arereported as belonging to the UX Ori class by Natta et al. (1997)and Grinin (1994). Moreover, the star XY Per (HBC 350) alsoshows photometric properties characteristic of UX Ori objects(Shevchenko et al. 1993; Chkhikvadze 2002), and V350 Ori(HBC 493) shows UX Ori type photometric and polarimetricproperties (Yudin & Evans 1998). The stars V1493 Cyg(HBC 716) and MC 1 (HBC 324) do not have enough data todecide whether their photometric properties fall in the UX Origroup. On the other hand, the quasi-periodic light curvereported for LkH� 215 (HBC 528; Shevchenko et al.1993) and its small photometric variability range (Herbst &Shevchenko 1999) tend to argue against a UX Ori nature.Overall, 75% of the stars with anomalous Fe ii absorption havebeen reported as UX Ori, which strongly suggests that the two

phenomena are related. Models of UX Ori objects invoke ob-scuration of the star by circumstellar material crossing the lineof sight; this could be produced by large orbiting circumstellarclouds, infalling cometary bodies, or instabilities in a flareddisk (Natta & Whitney 2000; Bertout 2000; Graham 1992;Grinin 1988). Recently Dullemond et al. (2003) proposed analternative explanation for this phenomenon, in which theobscuring region is the inner rim of a truncated disk. In thismodel, the inner disk is puffed up as a result of the rim beingmuch hotter than the rest of the disk, such that this inner rimshadows the outer vicinity of the disk (Dullemond et al. 2001).The same material occulting the star could be responsible forthe anomalous absorption features. High-resolution studiesindicate variations of Fe ii (42), going from a P Cygni profile toan absorption profile (Rodgers et al. 2002; Catala et al. 1993).

6. REDDENING TOWARD HAeBe STARS

Pre–main-sequence stars have long been known to exhibitsignificant photometric variability across a wide wavelengthrange. In particular, �25% of all known HAeBe stars arereported to show strong variations in brightness and color,which can exceed 4 mag in some wavelength ranges like theJohnson V band (Finkenzeller & Mundt 1984; Herbst &Shevchenko 1999). The UX Ori objects fall within this group.(Rodgers et al. 2002; Natta et al. 2000; Natta & Whitney2000). This behavior means that great care must be exercisedwhen measuring a representative extinction (AV) toward thesestars. Variability of the spectral type could further complicatereddening determinations. Ideally one would want to havemultiepoch, simultaneous spectra and photometry for each starto estimate mean magnitudes and spectra, but this requires anobservational effort that is seldom feasible (though somecampaigns like EXPORT are aimed toward this multiepoch,multiwavelength approach).We adopt the spectral type found in this study as repre-

sentative of the photosphere of the star and use an extensive

Fig. 6.—Distribution of Wk for H� (top) and [O i] k6300 (bottom): HAeBestars (solid histogram), late F stars (shaded histogram), and continuum stars(open histogram). Continuum stars exhibit the strongest emission at H� . Thelate F stars do not show [O i] k6300 in emission.

Fig. 7.—Comparison between the Wk of H� and the Fe ii k5169 line.Although there is significant scatter, a simple linear regression yields a cor-relation index of 0.74, suggesting a trend between these lines.


photometric data set so the variability range can be readilyassessed. The optical photometry we use consists of a largedata set in the UBVR bands that Herbst & Shevchenko (1999)have amassed by monitoring a set of HAeBe stars since 1983;69% of the stars that we have classified here have UBVRmeasurements in their work. We complemented the Herbst &Shevchenko (1999) data using measurements from de Winteret al. (2001), Miroshnichenko et al. (2001), Flaccomio et al.(1999), Fernandez (1995), Hillenbrand et al. (1992), Mendoza& Gomez (1980), Herbig & Bell (1988), and MacConnell(1968). We computed mean and median UBVR magnitudes foreach star in our sample using these databases; these agreedwithin 0.1 mag or better. The most variable of the stars in oursample are the UX Ori type objects; for these, we find that themean magnitudes correspond to the bright state when the staris probably seen without the occulting screen. Thus, in ouranalysis we adopt mean magnitudes and colors as represen-tative of the brightness of the stellar photosphere.

We calculated the color excesses EV�B and EV�R using in-trinsic colors given for each spectral type by Kenyon &Hartmann (1995). With this information we obtained the valuesof the visual extinctions AV1 from EV�B and AV2 from EV�R, withdifferent values of the total-to-selective extinction RV (AV =RVEB�V). We used the relations given in Cardelli et al. (1989) tocalculate AðkÞ=AðV Þ for a specific RV . In Figure 8 we plot AV1

versus AV2 for RV ¼ 3:1, consistent with the mean interstellarmedium (left) and for RV ¼ 5 (right). The line with slope 1 isindicated in both cases. The values of the extinction derivedfrom different colors agree if RV is significantly higher than3.1; for RV ¼ 5 the correlation is AV1 ¼ 1:002AV2 � 0:136,while for RV ¼ 3:1 it is AV1 ¼ 0:779AV2 � 0:08. The signifi-cant points for determining these correlations are stars with highreddening, AV > 1:5, which constitute 73% of the sample. Inorder to determine if the reddening toward these stars is inter-stellar or circumstellar, we examine their distances, given inTable 9. We find that 85% of the highly reddened stars arelocated within 1 kpc from the Sun. Since for most lines of sight,the expected interstellar reddening is less than 1 mag for thisrange of distances (Fitzgerald 1968), we conclude that the highreddening is not interstellar in nature; in fact, this reddening ismost likely produced by a combination of circumstellar mate-rial close to the star andmaterial from the molecular clouds withwhich many of these objects are still associated. The high value

of RV then strongly suggests that the circumstellar mediumaround HAeBe stars is dominated by grains larger in size thanthe average dust grain in the diffuse interstellar medium. Otherauthors have also arrived at a similar conclusion using smallersamples of these stars (Strom et al. 1972; The et al. 1981; Herbstet al. 1982; Sorrell 1990; Bibo et al. 1992; Gorti & Bhatt 1993;Waters & Waelkens 1998; Whittet et al. 2001). Studies of sili-cate features also show strong evidence of coagulation and anincrease in average grain size (Meeus et al. 2002; Bouwmanet al. 2001; Meeus et al. 2001)

In Figure 9 we plot the ratio EV�RC=EB�V versus EV�IC=EB�V

for a subsample of HAeBe stars with measured IC from theVan Vleck Observatory public ftp server,8 Fernandez (1995),Oudmaijer et al. (2001), and de Winter et al. (2001). We alsoplot the expected color ratios for different values of RV usingthe relations of Cardelli et al. (1989). For this, extinctionsAðkÞ=AðV Þ were calculated at the effective wavelength of thefilters, using tables of effective wavelength versus V�ICkindly provided by M. Bessell. The two lines correspond tothe minimum and maximum V�IC of the sample. Again, thedata suggest an extinction law with RV > 3:1.

7. LOCATION OF THE STARS IN THE H-R DIAGRAM

Knowledge of the appropriate extinction law for HAeBestars is important in order to derive luminosities for thesestars, which in turn allow us to estimate their masses andevolutionary status. We calculated the stellar luminosity for 55out of the 58 stars shown in Tables 2, 3, and 4. Three stars,P102, MV Ori, and RNO 63, did not have enough publishedphotometric data to enable a reliable estimate of luminosity.We used the mean V magnitude given in Table 9, corrected forreddening with an AV obtained from mean colors (x 6), bolo-metric corrections from Kenyon & Hartmann (1995), anddistances from the references cited in Table 9. The extinctioncorrection was calculated by comparing the B�V colors withintrinsic colors for the spectral type from Kenyon & Hartmann(1995), using both the standard interstellar extinction lawRV ¼ 3:1 and RV ¼ 5:0. The effective temperature wasdetermined using our spectral types and the calibration ofKenyon & Hartmann (1995).

Fig. 8.—Comparison of reddening values AV determined from E(B�V ) and E(V�R), for RV ¼ 3:1 (left) and RV = 5.0 (right). The solid line represents the fit tothe data, while the dashed line has slope unity. The best correlation is observed for RV ¼ 5:0. Error bars represent the propagated error from the spectral types.

8 See ftp://ftp.astro.wesleyan.edu/pub/ttauri.


TABLE 9

Ages and Masses

RV = 3.1 RV = 5.0

HBC Name

V

(Mag) Reference

Distance

(pc) Reference

Teff(K)

AV(mag)

logL

(L�)

Mass

(M�)

Age

(Myr)

AV

(mag)

logL

(L�)

Mass

(M�)

Age

(Myr)

3.......... V633 Cas 14.18 1 600 8 4.03 3.2 1.28 a a 5.1 2.06 3.2 2.59

7.......... LkH� 201 13.64 1 850 9 4.32 4.4 2.96 a a 7.1 4.04 11.0 0.09

78........ AB Aur 7.05 1 144 10 3.97 0.3 1.63 2.5 4.33 0.5 1.71 2.6 3.94

154...... T Ori 10.63 1 460 8 3.98 1.6 1.73 2.6 3.89 2.6 2.13 3.5 1.68

160...... PQ Tau 12.63 2 460 8 3.83 0.6 0.48 1.5 b 1.0 0.64 1.5 25.06

170...... RR Tau 12.08 1 800 8 3.99 2.0 1.81 2.7 3.69 3.2 2.29 3.9 1.30

192...... HD 250550 9.54 1 700 8 4.04 0.4 2.20 3.6 1.80 0.7 2.31 3.9 1.44

193...... LkH� 208 11.65 1 1000 8 3.91 0.8 1.59 2.4 4.21 1.2 1.78 2.8 2.87

196...... LkH� 338 15.12 2 830 11 4.05 3.3 1.26 a a 5.3 2.06 3.2 2.68

197...... LkH� 339 13.66 1 830 11 3.97 2.6 1.43 2.3 7.00 4.3 2.08 3.4 1.83

201...... LkH� 341 13.39 1 800 12 3.83 1.8 1.13 1.8 7.90 3.0 1.58 2.6 3.13

217...... W84 12.02 3 910 13 3.80 0.2 1.15 2.0 6.27 0.3 1.19 2.1 5.38

219...... V590 Mon 12.77 1 800 8 4.11 0.8 1.32 a a 1.3 1.52 a a

222...... W108 11.97 2 910 13 3.80 0.3 1.21 2.1 5.18 0.5 1.28 2.2 4.27

231...... V360 Mon 13.39 3 758 14 3.76 0.6 0.64 1.5 12.49 1.0 0.80 1.7 8.81

281...... LkH� 118 11.20 1 1950 15 4.43 3.5 4.55 16.2 0.04 5.7 5.42 42.6 0.01

282...... VV Ser 11.92 1 440 16 4.14 3.4 2.23 3.8 b 5.4 3.06 5.8 0.46

284...... AS 310 NW 12.45 1 2500 8 4.40 4.1 4.43 14.5 0.02 6.6 5.43 43.5 0.01

293...... PX Vul 11.49 4 420 12 3.83 1.4 1.15 1.9 7.53 2.2 1.48 2.4 3.70

297...... V751 Cyg 14.18 1 700 17 3.99 0.9 0.45 a a 1.5 0.68 a a

305...... LkH� 324 12.61 4 780 18 4.09 3.7 2.48 4.3 1.25 6.0 3.39 7.9 0.21

308...... LkH� 349 13.37 4 750 19 3.79 3.2 1.65 3.0 1.89 5.2 2.43 5.4 0.30

309...... LkH� 234 12.21 1 1000 16 4.12 3.1 2.67 4.8 0.83 5.0 3.43 7.9 0.22

310...... BD +46�3471 9.89 1 900 16 3.99 0.3 2.12 3.5 1.81 0.5 2.19 3.7 1.52

313...... LkH� 233 13.56 1 880 8 3.93 2.3 1.34 2.1 7.16 3.7 1.90 2.9 2.61

314...... LkH� 314 14.04 1 400 20 4.08 6.3 2.31 3.9 1.51 10.1 3.85 11.6 0.05

324...... MC1 10.77 5 850 9 3.90 0.4 1.67 2.5 3.75 0.7 1.78 2.8 2.83

329...... VX Cas 11.28 1 760 8 3.99 1.0 1.70 2.7 3.96 1.7 1.95 3.0 2.73

334...... RNO 6 14.52 2 1600 8 4.26 2.3 2.18 a a 3.8 2.75 5.0 b

348...... IP Per 10.47 6 350 21 3.92 0.6 1.07 1.9 14.36 0.9 1.21 2.0 7.98

350...... XY Per EW 9.21 1 120 10 3.92 1.1 0.86 a a 1.7 1.12 1.9 10.92

373...... V892 Tau 15.25 1 160 8 4.05 4.8 0.41 a a 7.8 1.60 2.8 b

430...... UX Ori 10.40 1 460 8 3.94 0.9 1.49 2.2 5.74 1.4 1.71 2.5 4.02

436...... RY Tau 11.80 7 460 8 3.80 1.2 1.05 1.8 7.62 1.9 1.34 2.3 3.86

442...... P1394 10.13 2 460 8 3.79 0.3 1.35 2.3 3.80 0.4 1.41 2.4 3.46

451...... HD 245185 9.89 1 400 8 3.97 0.2 1.31 2.1 9.95 0.3 1.36 2.2 7.97

464...... CQ Tau 10.27 1 130 22 3.83 1.2 0.55 1.5 b 2.0 0.85 1.6 13.27

482...... BN Ori 9.67 1 460 8 3.82 0.3 1.51 2.5 3.51 0.4 1.58 2.7 2.93

492...... BD +26�887 10.47 2 2000 23 3.92 0.8 2.69 5.2 0.45 1.3 2.89 6.2 0.28

493...... V350 Ori 11.47 7 460 8 3.96 1.3 1.24 2.0 b 2.1 1.55 2.3 5.64

528...... LkH� 215 10.54 1 800 21 4.14 2.0 2.75 4.8 0.80 3.2 3.24 6.6 0.30

529...... HD 259431 8.73 1 800 8 4.15 1.2 3.19 6.6 0.32 2.0 3.50 9.7 0.09

531...... VSB2 13.33 2 910 13 3.77 0.4 0.74 1.5 11.74 0.7 0.85 1.7 9.25

535...... W121 10.80 2 910 13 3.77 0.0 1.58 3.0 1.88 0.0 1.58 3.0 1.88

548...... LkH� 218 11.87 1 1150 8 3.98 1.3 1.91 3.0 2.80 2.1 2.23 3.8 1.40

551...... LkH� 222 11.81 1 1150 12 4.09 1.2 2.12 3.4 2.78 1.9 2.41 4.1 1.38

686...... WW Vul 10.74 1 550 16 3.94 1.0 1.56 2.4 4.48 1.6 1.81 2.9 2.94

689...... V1685 Cyg 10.69 1 980 22 4.27 3.0 3.58 8.8 0.13 4.9 4.33 64.1 0.11

690...... V1686 Cyg 14.06 1 980 22 3.79 2.5 1.32 2.3 3.69 4.0 1.93 3.8 0.95

705...... LkH� 147 14.46 1 800 24 4.32 5.3 2.94 a a 8.6 4.24 12.9 0.06

726...... HD 200775 7.37 1 429 10 4.27 1.8 3.73 8.9 0.13 3.0 4.17 12.5 0.06

730...... BD +65�1637 10.18 1 1250 22 4.22 1.8 3.40 7.0 0.29 2.9 3.83 25.0 0.14

734...... BH Cep 11.16 1 450 20 3.81 0.7 1.06 1.8 8.22 1.1 1.23 2.1 5.47

735...... BO Cep 11.60 1 400 20 3.82 0.5 0.72 1.5 14.41 0.8 0.85 1.6 12.48

736...... SV Cep 10.98 1 400 20 4.00 1.3 1.38 2.3 b 2.1 1.69 2.7 3.96

a The star falls below the ZAMS.b The star falls on the ZAMS.References.—(1) Herbst & Shevchenko 1999; (2) Herbig & Bell 1988; (3) Flaccomio et al. 1999; (4) Fernandez 1995; (5) MacConnell 1968; (6) Miroshnichenko

et al. 2001; (7) de Winter et al. 2001; (8) Testi et al. 1998; (9) Yonekura et al. 1997; (10) Bertout, Robichon, & Arenou 1999; (11) Herbst & Racine 1976;(12) Herbst et al. 1982; (13) Neri, Chavarria-K., & de Lara 1993; (14) Park et al. 2000; (15) Kozok 1985; (16) Pirzkal et al. 1997; (17) Chavarria et al. 1989;(18) Chavarria et al. 1983; (19) Hessman et al. 1995; (20) Kun 1998; (21) Hillenbrand et al. 1992; (22) van den Ancker et al. 1998; (23) Kawamura et al. 1998;(24) Natta et al. 2001.

Figure 10 shows the location in the H-R diagram of thesestars, for the two values of RV used. We also show the evo-lutionary tracks and isochrones of Siess et al. (2000) andBernasconi (1996). We derive masses and ages for the sampleby linear interpolation in these tracks. The derived values aregiven in Table 9 for each value of RV.

Clearly the value of RV makes a significant difference in theH-R diagram positions of the sample. When using RV ¼3:1, 12 objects fall below the zero main-sequence age(ZAMS) and many close to it. In contrast, for RV ¼ 5:0only two objects appear below the ZAMS, V751 Cyg(HBC 297) and V590 Mon (HBC 219). As discussed in seex 4.1, V751 Cyg probably is a cataclysmic variable. Testi et al.(1998) published for V590 Mon a B�V color that is 1 maglarger than any value published before. With their B�V mea-surement, V590 Mon falls well above the zero main-sequence age (ZAMS). However, a large variation in B�V isdifficult to reconcile with the fact that this star shows littlephotometric variability with a quasi-periodic light curve(Herbst & Shevchenko 1999).

8. SUMMARY AND CONCLUSIONS

We have applied a consistent spectral classification schemeaimed at early-type PMS stars. Our method relies on a largenumber of reddening-independent indices covering a widewavelength range, from 3900 to 7000 A. Our scheme is de-signed to avoid contamination by nonphotospheric contribu-tions to absorption features normally used for spectral typing.We were able to determine spectral types for 58 objects outof a total sample of 75 objects from the (Herbig & Bell1988) catalog, with an average uncertainty of 2.5 subclasses.Seven stars with spectral types have an uncertain evolu-tionary status. We could not derive spectral types for a

Fig. 9.—Ratio of color excesses EV�Rc /EB�V versus ratio of color excessesEV�IC=EB�V . The dotted lines indicate the locus of predicted color excessratios for values of RV from 2 to 8, calculated using the Cardelli et al. (1989)reddening law. The extinctions were calculated at the effective wavelength ofthe filters, for the minimum (top dotted line) and maximum (bottom dottedline) V�IC of our sample (see text). Filled circles are objects with more than fivephotometric measurements. The asterisks correspond to stars with only one ortwo measurements in each filter; for these objects variability could be affectingour estimate of color excesses.

Fig. 10.—Location of the stars in the H-R diagram. Left: Luminosities obtained using the standard value of total-to-selective extinction for the interstellarmedium, RV ¼ 3:1. Right: Results using the larger value RV ¼ 5:0. When using RV ¼ 3:1, many stars fall on or below the ZAMS. A value of RV ¼ 5 tends to yieldhigher luminosities, moving the stars upward in the H-R diagram, hence making them younger. Then almost all stars fall above the ZAMS, which is more consistentwith their pre–main-sequence nature. We show the evolutionary tracks (solid lines) and isochrones (dashed lines). Tracks represent, from top to bottom, 25, 15, and9 M� (Bernasconi 1996) and 6, 3, and 1 M� (Siess et al. 2000). The isochrones from Siess et al. (2000) are, from top to bottom, 0.1, 1, 10, and 100 Myr (which wetake as the ZAMS). Luminosity errors represent the propagated error from the spectral type.


subsample of 11 stars. In seven of these stars, no absorptionfeatures where apparent in the spectrum, which is dominatedby emission lines; these are continuum stars. The contami-nation by nonstellar spectral features was too strong for therest of this subsample, precluding spectral typing. Finally,only approximate spectral types are given for six highly red-dened stars, for which the blue end of the spectrum was toofaint to properly apply our classification scheme.

By definition, all the stars of the sample show H� inemission. However, at our resolution only 53% show H� inemission as well, and only 15% show additional emission inH�. Nonetheless, 95% and 56% of the HAeBe stars showfilling-in of H� and H�, respectively.

Almost half of the HAeBe stars classified in this work(excluding stars F7 and later) exhibit the forbidden line [O i]k6300 in emission (similar to reports by Corcoran & Ray 1997and Bohm & Catala 1994). This feature is indicative of thepresence of winds, outflows, or jets. A third of the sampleexhibits emission in multiplets of Fe ii, particularly multiplet42, and the strength of this emission is correlated with that ofH� . Emission in multiplets of Fe ii only appears if [O i] k6300is present, although their strengths do not seem to correlate.

A subset of 11 HAeBe stars, �28% of the HAeBe sample,shows lines of multiplet 42 of Fe ii with abnormally strongabsorption. Of these stars, 75% have been confirmed as UXOri objects, strongly suggesting that the anomalous Fe ii ab-sorption is produced by the same mechanism that results in theUX Ori phenomenon. In fact, if we assume that the sample ofHAeBe stars was complete, the number of objects withanomalous Fe ii absorption would be consistent with theexpected number of high-inclination (>75�) systems, which isa condition for the UX Ori phenomenon to occur (Natta &Whitney 2000).

We have used published photometric data together with ourderived spectral types to estimate the reddening law that bestcharacterizes the class of HAeBe. We find that a reddening lawwith a high value of RV , �5.0, yields a much better agreementbetween values of the extinction AV obtained from different

colors than the standard reddening law. Since 85% of the starswith large values of AV are located within 1 kpc from the Sun,the high extinction values are probably not due to interstellarreddening. Rather, the stars must be mostly extincted by theircircumstellar environments. Thus, the high value of RV thatcharacterizes the reddening law toward the intermediate massPMS stars indicates that dust has grown with respect to thetypical grain size in the interstellar medium.Using reddening values determined for different extinction

laws, we locate the stars in the H-R diagram. The position ofthe stars depends critically on the value chosen for RV, henceaffecting estimates of masses and most particularly of ages.With the most appropriate value of RV ¼ 5:0, objects appearsystematically younger and brighter relative to their positionscalculated with the standard law. In particular, the majority ofthe spectroscopically selected, young, bona fide HAeBe starsconsistently falls above the ZAMS.Additional information about the spectra, H� emission,

UBVRIJHK magnitudes, observed emission lines, findingcharts, and optical and near infrared spectral energy distribu-tion for each object analyzed in this work are reported on theWorld Wide Web.9

We thank Michael Bessell for sending us the color-depen-dence of the filter effective wavelengths, Bruno Merın forsending us the multiepoch spectra of the star V1686 Cyg,George Herbig for insightful conversations, and G. Meeus, thereferee, for his careful reading of the manuscript and his de-tailed and useful comments and suggestions. We also thankSusan Tokarz of the SAO Telescope Data Center for carryingout the data reduction and Michael Calkins for obtaining someof the spectra. This work was supported in part by NASAgrants NAG5-9670 and NAG10545, NSF grant AST 99-87367 and grant S1-2001001144 of FONACIT, Venezuela.

REFERENCES

Arce, H. G., & Goodman, A. A. 2002, ApJ, 575, 911Asselin, L., Menard, F., Bastien, P., Monin, J., & Rouan, D. 1996, ApJ,472, 349

Bailey, J. 1998, MNRAS, 301, 161Bernasconi, P. A. 1996, A&AS, 120, 57Bertout, C. 2000, A&A, 363, 984Bertout, C., Robichon, N., & Arenou, F. 1999, A&A, 352, 574Bibo, E. A., The, P. S., & Dawanas, D. N. 1992, A&A, 260, 293Bohm, T., & Catala, C. 1994, A&A, 290, 167———. 1995, A&A, 301, 155Bouwman, J., Meeus, G., de Koter, A., Hony, S., Dominik, C., & Waters,L. B. F. M. 2001, A&A, 375, 950

Briceno, C., Hartmann, L. W., Stauffer, J. R., Gagne, M., Stern, R. A., &Caillault, J. 1997, AJ, 113, 740

Buscombe, W. 2001, VizieR On-line Data Catalog, III/222Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, 245Casey, S. C., & Harper, D. A. 1990, ApJ, 362, 663Catala, C., Bohm, T., Donati, J.-F., & Semel, M. 1993, A&A, 278, 187Chavarria, C., de Lara, E., Finkenzeller, U., Appenzeller, I., & Cardona, O.1983, A&A, 118, 189

Chavarria-K., C., Terranegra, L., Alcala, J. M., & Neri, L. 1989, Rev. MexicanaAstron. Astrofis., 18, 178

Chkhikvadze, I. N. 2002, Astrophysics, 45, 150Cohen, M., & Kuhi, L. V. 1979, ApJS, 41, 743Coluzzi, R. 1999, VizieR On-line Data Catalog, VI/71ACorcoran, M., & Ray, T. 1997, A&A, 321, 189Corporon, P., & Lagrange, A.-M. 1999, A&AS, 136, 429Davies, J. K., Evans, A., Bode, M. F., & Whittet, D. C. B. 1990, MNRAS,247, 517

de Winter, D., van den Ancker, M. E., Maira, A., The, P. S., Djie, H. R. E. T. A.,Redondo, I., Eiroa, C., & Molster, F. J. 2001, A&A, 380, 609

Downes, R., Hoard, D. W., Szkody, P., & Wachter, S. 1995, AJ, 110, 1824Dullemond, C. P., Dominik, C., & Natta, A. 2001, ApJ, 560, 957Dullemond, C. P., van den Ancker, M. E., Acke, B., & van Boekel, R. 2003,ApJ, 594, L47

Echevarrıa, J., Costero,R., Tovmassian,G., Zharikov, S., Pineda,L.,&Michel, R.2002, Rev. Mexicana Astron. Astrofis. Ser. Conf., 12, 86

Fabricant, D., Cheimets, P., Caldwell, N., & Geary, J. 1998, PASP, 110, 79Fernandez, M. 1995, A&AS, 113, 473Fernandez, M., Ortiz, E., Eiroa, C., & Miranda, L. F. 1995, A&AS, 114, 439Finkenzeller, U. 1985, A&A, 151, 340Finkenzeller, U., & Jankovics, I. 1984, A&AS, 57, 285Finkenzeller, U., & Mundt, R. 1984, A&AS, 55, 109Fitzgerald, M. P. 1968, AJ, 73, 983Flaccomio, E., Micela, G., Sciortino, S., Favata, F., Corbally, C., & Tomaney, A.1999, A&A, 345, 521

Garcia, B. 1989, Bull. Inf. Centre Donnees Stellaires, 36, 27Gomez, M., Kenyon, S. J., & Whitney, B. A. 1997, AJ, 114, 265Gorti, U., & Bhatt, H. C. 1993, A&A, 270, 426Graham, J. A. 1992, PASP, 104, 479Gray, D. F. 1992, The Observation and Analysis of Stellar Photospheres(2d ed.; Cambridge: Cambridge Univ. Press), 81

Gray, R. O., Napier, M. G., & Winkler, L. I. 2001, AJ, 121, 2148Grinin, V. P. 1988, Soviet Astron. Lett., 14, 27———. 1994, ASP Conf. Ser. 62, The Nature and Evolutionary Status ofHerbig Ae/Be Stars, ed. P. S. The, M. R. Perez, & E. P. J. van den Heuvel(San Francisco: ASP), 63

Hajjar, R., & Bastien, P. 2000, ApJ, 531, 494

9 See http://cfa-www.harvard.edu/youngstars/jhernand/haebe/principal.html.


Hamann, F., & Persson, S. E. 1992, ApJS, 82, 285Hartmann, L., Calvet, N., Gullbring, E., & D’Alessio, P. 1998, ApJ, 495, 385Hartmann, L., Kenyon, S. J., & Calvet, N. 1993, ApJ, 407, 219Herbig, G. H. 1960, ApJS, 4, 337Herbig, G. H., & Bell, K. 1988, Lick Obs. Bull., 1111Herbst, W., & Racine, R. 1976, AJ, 81, 840Herbst, W., & Shevchenko, V. S. 1999, AJ, 118, 1043Herbst, W., Warner, J. W., Miller, D. P., & Herzog, A. 1982, AJ, 87, 98Hessman, F. V., Beckwith, S. V. W., Bender, R., Eisloeffel, J., Goetz, W., &Guenther, E. 1995, A&A, 299, 464

Hillenbrand, L. A. 1995, Ph.D. thesis, Univ. MassachusettsHillenbrand, L. A., Strom, S. E., Vrba, F. J., & Keene, J. 1992, ApJ, 397, 613Jaschek, M. 1978, Bull. Inf. Centre Donnees Stellaires, 15, 121Kawamura, A., Onishi, T., Yonekura, Y., Dobashi, K., Mizuno, A., Ogawa, H., &Fukui, Y. 1998, ApJS, 117, 387

Keenan, P. C., & Barnbaum, C. 1999, ApJ, 518, 859Kenyon, S. J., & Hartmann, L. 1995, ApJS, 101, 117Kozok, J. R. 1985, A&AS, 62, 7Kun, M. 1998, ApJS, 115, 59Leinert, C., Richichi, A., & Haas, M. 1997, A&A, 318, 472Lorenzetti, D., Saraceno, P., & Strafella, F. 1983, ApJ, 264, 554MacConnell, D. J. 1968, ApJS, 16, 275Magakian, T. Y., & Movsesian, T. A. 2001, Astrophysics, 44, 419Malfait, K., Bogaert, E., & Walkens, C. 1998, A&A, 331, 211Mannings, V., & Sargent, A. I. 1997, ApJ, 490, 792———. 2000, ApJ, 529, 391Marconi, M., Ripepi, V., Bernabei, S., Palla, F., Alcala, J. M., Covino, E., &Terranegra, L. 2001, A&A, 372, L21

Meeus, G., Bouwman, J., Dominik, C., Waters, L. B. F. M., & de Koter, A.2002, A&A, 392, 1039

Meeus, G., Waters, L. B. F. M., Bouwman, J., van den Ancker, M. E.,Waelkens, C., & Malfait, K. 2001, A&A, 365, 476

Mendoza, V. E. E., & Gomez, T. 1980, MNRAS, 190, 623Miroshnichenko, A. S., Bjorkman, K. S., Chentsov, E. L., Klochkova, V. G.,Gray, R. O., Garcıa-Lario, P., & Perea Calderon, J. V. 2001, A&A, 377, 854

Miroshnichenko, A. S., et al. 2003, A&A, 408, 305Mora, A., et al. 2001, A&A, 378, 116Morgan, W. W., Keenan, P. C., & Kellman, E. 1943, An Atlas of SpectraClassification (Chicago: Univ. Chicago Press)

Muzerolle, J., Calvet, N., & Hartmann, L. 2001, ApJ, 550, 944Muzerolle, J., Calvet, N., Hartmann, L., Briceno, C., Hillenbrand, L. A., &Hernandez J. 2004, in preparation

Natta, A., Grinin, V., & Mannings, V. 2000, in Protostars and Planets IV, ed.V. Manning, A. P. Boss, & S. S. Russell (Tucson: Univ. Arizona Press), 559

Natta, A., Grinin, V. P., Mannings, V., & Ungerechts, H. 1997, ApJ, 491, 885Natta, A., Prusti, T., Neri, R., Wooden, D., Grinin, V. P., & Mannings, V. 2001,A&A, 371, 186

Natta, A., & Whitney, B. A. 2000, A&A, 364, 633Neri, L. J., Chavarria-K., C., & de Lara, E. 1993, A&AS, 102, 201Osterbrock, D. E. 1989, Astrophysics of Gaseus Nebulae and Active GalacticNuclei (Mill Valley: Univ. Science Books)

Oudmaijer, R. D., et al. 2001, A&A, 379, 564Palla, F., & Stahler, S. W. 1991, ApJ, 375, 288Park, B., Sung, H., Bessell, M. S., & Kang, Y. H. 2000, AJ, 120, 894Parsamian, E. S., Gasparian, K. G., & Ohanian, G. B. 1996, Astrophysics,39, 121

Pirzkal, N., Spillar, E. J., & Dyck, H. M. 1997, ApJ, 481, 392Pritchet, C., & van den Bergh, S. 1977, ApJS, 34, 101Reid, I. N., Hawley, S. L., & Gizis, J. E. 1995, AJ, 110, 1838Robinson, E. L. 1973, ApJ, 180, 121Rodgers, B., Wooden, D. H., Grinin, V., Shakhovsky, D., & Natta, A. 2002,ApJ, 564, 405

Shevchenko, V. S., Ezhkova, O., Tjin A Djie, H. R. E., van den Ancker, M. E.,Blondel, P. F. C., & de Winter, D. 1997, A&AS, 124, 33

Shevchenko, V. S., Grankin, K. N., Ibragimov, M. A., Mel’Nikov, S. Y., &Yakubov, S. D. 1993, Ap&SS, 202, 121

Siess, L., Dufour, E., & Forestini, M. 2000, A&A, 358, 593Skumanich, A. 1972, ApJ, 171, 565Sorelli, C., Grinin, V. P., & Natta, A. 1996, A&A, 309, 155Sorrell, W. H. 1990, ApJ, 361, 150Stock, J., & Stock, J. M. 1999, Rev. Mexicana Astron. Astrofis., 35, 143Strom, K. M., Wilkin, F. P., Strom, S. E., & Seaman, R. L. 1989, AJ, 98, 1444Strom, S. E. 1983, Rev. Mexicana Astron. Astrofis., 7, 201Strom, S. E., Strom, K. M., Yost, J., Carrasco, L., & Grasdalen, G. 1972, ApJ,173, 353

Terranegra, L., Chavarria-K., C., Diaz, S., & Gonzalez-Patino, D. 1994, A&AS,104, 557

Testi, L., Palla, F., & Natta, A. 1998, A&AS, 133, 81The, P. S., de Winter, D., & Perez, M. R. 1994, A&AS, 104, 315The, P. S., et al. 1981, A&AS, 44, 451van den Ancker, M. E., de Winter, D., & Tjin A Djie, H. R. E. 1998, A&A,330, 145

van den Ancker, M. E., The, P. S., Feinstein, A., Vazquez, R. A., de Winter, D.,& Perez, M. R. 1997, A&AS, 123, 63

Waters, L. B. F. M., & Waelkens, C. 1998, ARA&A, 36, 233Whitney, B. A., & Hartmann, L. 1993, ApJ, 402, 605Whittet, D. C. B., Gerakines, P. A., Hough, J. H., & Shenoy, S. S. 2001, ApJ,547, 872

Wu, Y., Huang, M., & He, J. 1996, A&AS, 115, 283Yonekura, Y., Dobashi, K., Mizuno, A., Ogawa, H., & Fukui, Y. 1997, ApJS,110, 21

Yudin, R. V., & Evans, A. 1998, A&AS, 131, 401


Spectral Analysis and Classification of Herbig Ae/Be Stars

Documents