OTTO G. FRANZ1 Lowell Observatory, 1400 West Mars Hill Road, Flagsta, AZ 86001; ogf=lowell.edu CARLOS E. LO PEZ

Transcription

1 THE ASTRONOMICAL JOURNAL, 121:1597È1606, 2001 March ( The American Astronomical Society. All rights reserved. Printed in U.S.A. SPECKLE INTERFEROMETRY OF SOUTHERN DOUBLE STARS. II. MEASURES FROM THE CASLEO 2.15 METER TELESCOPE, 1995È1996 ELLIOTT HORCH1 Chester F. Carlson Center for Imaging Science, Rochester Institute of Technology, 54 Lomb Memorial Drive, Rochester, NY ; horch=cis.rit.edu WILLIAM F. VAN ALTENA1 AND TERRENCE M. GIRARD Department of Astronomy, Yale University, P. O. Box , New Haven, CT ; vanalten=astro.yale.edu; girard=astro.yale.edu OTTO G. FRANZ1 Lowell Observatory, 1400 West Mars Hill Road, Flagsta, AZ 86001; ogf=lowell.edu CARLOS E. LO PEZ Observatorio Astrono mico Fe lix Aguilar, Universidad Nacional de San Juan, Avenida Benav dez 8175 Oeste, 5407 Marquesado, San Juan, Argentina; lopez2283=infovia.com.ar AND J. GETHYN TIMOTHY Nightsen, Inc., 30 Fresno Road, Warwick, RI ; gethyn=nightsen.com Received 2000 October 24; accepted 2000 November 20 ABSTRACT Relative astrometry is presented for 198 observations of 160 double stars. The data were obtained at the 2.15 m telescope at the Complejo Astrono mico El Leoncito (Argentina) with a multianode microchannel array (MAMA) detector system. Five high-quality nondetections are also reported. When judged against ephemeris positions for binaries with very well determined orbits, the separation residuals exhibit a root mean square deviation of 13.2 `3.4 mas and the position angle residuals exhibit rms deviation of ~ `0].8. Factors a ecting the measurement precision are discussed. ~0].5 Key words: astrometry È binaries: close È binaries: visual È techniques: interferometric 1. INTRODUCTION The Ðrst paper in this series (Horch et al. 1996, hereafter Paper I) discussed the relative lack of speckle observations of binary stars taken in the southern hemisphere and our attempt to help alleviate this disparity with a speckle observing program based at El Leoncito, Argentina. This e ort was a collaboration between the Yale Southern Observatory and the Universidad Nacional de San Juan, Argentina. Such observations are necessary to conðrm the orbital motion of many southern doubles and to provide high-quality relative astrometry of known binaries in preparation for orbit determinations or orbit reðnements. These orbits are thus the Ðrst step in determining masses. The Yale-San Juan program was active from 1994 July until 1996 November and utilized two telescopes at the El Leoncito site: the 76 cm reñector at the Carlos U. Cesco Observatory (run jointly by the Universidad Nacional de San Juan and Yale Southern Observatory) and the 2.15 m telescope at the Complejo Astrono mico El Leoncito (CASLEO, the Argentine National Observatory). This paper presents all measures obtained from data taken at the CASLEO telescope and not previously published in Paper I; a forthcoming publication will address the data taken at the smaller aperture. 2. OBSERVATIONS AND DATA REDUCTION Observations were obtained at the CASLEO 2.15 m telescope on Ðve separate runs between 1995 September and 1996 November. Measurements taken while observing ÈÈÈÈÈÈÈÈÈÈÈÈÈÈÈ 1 Visiting Astronomer, Complejo Astrono mico El Leoncito (CASLEO), operated under agreement between the Consejo Nacional de Investigaciones Cient Ðcas y Te cnicas de la Repu blica Argentina and the National Universities of La Plata, Co rdoba, and San Juan show that the average seeing was well over 2A on a majority of our nights at the telescope during Many of these observations did not successfully reduce. However, during this time it was noticed that the temperature of the primary mirror was often several degrees above the ambient temperature, a situation known to adversely a ect the seeing. In 1996, the observatory installed a cooling system for the primary, and thereafter we obtained much better average seeing conditions. As a result, our three 1996 runs were much more productive than the two 1995 runs. This is illustrated in Figure 1, which shows a histogram of the seeing measures for all observations presented in this paper. Though the number of nights that we observed in 1995 and 1996 was comparable, about 3 times as many successful measures were obtained in the latter year. As the astrometric quality of speckle observations is known to deteriorate in seeing greater than 2A (see, e.g., Horch et al. 1997), we rejected any observation having seeing worse than 2A.3 whether it successfully reduced or not, to ensure reasonable quality for our measures. As in Paper I, speckle interferograms were obtained with a multianode microchannel array (MAMA) detector system operating in time-tag mode. With only a few exceptions, the data presented here were taken in high-resolution ÏÏ mode (Kasle & Morgan 1991), where the device e ectively has 7 km square pixels. The MAMA detector had a bialkali photocathode with quantum efficiency of approximately 1% in the visible range. Trialkali (S-20) photocathodes have been more typically used in speckle cameras and exhibit detective quantum efficiencies of 5% to 10%. The low quantum efficiency of our camera is another factor that limits the quality of the data obtained, but that camera was the only device available for the project. A full-aperture, four-slit mask was constructed for the telescope in order to measure the pixel scale precisely.

2 1598 HORCH ET AL. FIG. 1.ÈSeeing histogram for all observations for which measures appear in Table 2. The darker histogram is for 1995 observations, and the lighter histogram is for 1996 observations. Many 1995 observations failed to reduce because of poor seeing and other factors and are not represented here. Because of the relatively large size of the primary aperture, the mask consisted of an aluminum ring bolted to the same surface as the spider holding the secondary mirror. Sheet metal strips with 1 cm slits cut into them were then attached to the ring. Other sections of the mask were covered in black cloth to keep the weight as low as possible. With the mask attached, observations of bright unresolved stars were recorded. These exhibited a fringe pattern where the fringe spacing together with the spacing between the di erent slits and the observation wavelength provide the necessary information to deduce the pixel scale. Although the wavelength passband used for these observations was small (40 nm), we nonetheless computed a correction to the e ective wavelength based on the starïs spectral type and the known shape of the Ðlter passband. Stellar spectra from Jacoby, Hunter, & Christian (1984) were used in this calculation. With the MAMA detector, we have the ability to take time-tag observations of stars with the telescope tracking o. These star trails give an estimate not only of the detector orientation relative to celestial north, but also of the pixel scale, since in time-tag mode the centroid position of the star can be measured as a function of time. When this measurement is compared with the diurnal rate of the star based on its declination, the pixel scale is obtained. The pixel format of the detector (1920 ] 448) helps to increase the precision of the star-trail method, since we aligned the detectorïs long axis with right ascension. Table 1 contains Ðnal results for scale and orientation using both the slitmask method and the star-trail method. Our speckle camera is known to exhibit a small amount of optical Ðeld angle distortion (OFAD); this was Ðrst TABLE 1 DETECTOR SCALE AND ORIENTATION RESULTS Zero Point Angle Scale Type of Measure (deg) (mas pixelv1) Slit mask ^ 0.09 Star trails... [0.30 ^ ^ 0.04 Final adopted value... [0.30 ^ ^ 0.09 detected after the camera was returned to the United States in 1997 (Horch et al. 1999). In that work, the authors corrected position angles and separation of double stars based on the location of the observation relative to the optical axis. The presence of the OFAD implies that a star trail taken with the MAMA will appear to have a nonconstant velocity as the image crosses the detectorïs active area. A trail will also appear to curve if it does not pass through the optical axis. For the star-trail data analyzed here, a correction factor was calculated based on the 1997 OFAD measurements and applied for both the detector orientation angle and the pixel scale. Likewise, the fringe spacing obtained with slit-mask observations will be a ected by OFAD, and we have also calculated correction factors for these data. As is apparent from Table 1, the agreement in scale between the two methods is very good. However, the star trails necessarily include data on the edges of the detector (where the OFAD correction is most signiðcant), far from the optical axis, whereas the mask Ðles are taken in the portion of the detector used for most double-star observations, near the optical axis. We therefore consider the star-trail method to be more susceptible to systematic error propagated from the uncertainty in the determination of the OFAD constants. For this reason, we have adopted the slit-mask result for scale and, there being no alternative, the star-trail measure for the detector orientation, as shown in Table 1. Observations of unresolved bright stars were obtained near in time and sky position to the double stars observed in order to have an estimate of the speckle transfer function for the double-star observations. These point-source calibrators, chosen from the Bright Star Catalogue (Hoffleit & Jaschek 1982), were then used in the reduction phase to deconvolve the speckle transfer function from the binarystar power spectrum, as described in Paper I. Our basic reduction approach is fully described there; brieñy, it consists of a weighted least-squares Ðt to the binary spatial frequency power spectrum. Once instrumental position angles and separations are obtained from this reduction software, the scale and detector orientation angle numbers discussed above are used to convert these numbers to the Ðnal results discussed in the next section. 3. RESULTS 3.1. Measures Table 2 contains the main set of astrometric results from the CASLEO observations. The columns give (1) the Aitken Double Star (ADS) Catalogue number, or, if none, the Bright Star Catalogue (HR) number, or, if none, the Durchmusterung (BD, CP, or CD) number; (2) the discoverer designation; (3) the HD number; (4) the Hipparcos Catalogue (ESA 1997) number; (5) the right ascension and declination in J coordinates, which is the same as the identiðcation number in the Washington Double Star Catalog (WDS; Worley & Douglass 1997) for all objects that have WDS entries; (6) the observation date in fraction of the Besselian year; (7) the observed position angle (h), in degrees, with north through east deðning the positive sense of h; (8) the observed separation (o) in arcseconds; (9) the center wavelength of the Ðlter used to make the observation, in nanometers; and (10) the full width at half-maximum of the Ðlter passband, also in nanometers. The position angles have not been corrected for precession and are appropriate

3 TABLE 2 DOUBLE STAR SPECKLE MEASURES FROM CASLEO Discoverer WDS Date h o j *j HR, ADS, DM, etc. Designation HD HIP(a, d J2000.0) (1900]) (deg) (arcsec) (nm) (nm) CD [ I [ c CD [ [ \ CD [ [ \ CD [ [ \ CD [ [ \ HR HdO [ ADS Bu 391 AB [ c b CD [ Hu [ a HR I 43 AB 661[ [ CP [ I [ b BD [ A [ CP [ I [ ADS Ho 212 AB [ HR HdO [ a a CD [ I [ c CD [ HJ [ HR Slr 1 AB [ CD [ See [ CD [ Bu [ b ADS A [ ADS StF 113 A, BC [ CP [ I [ c ADS Bu [ c c CD [ Rst [ a CP [ I 264 AB [ c ADS A [ b CD [ HJ [ a ADS StF [ ADS A [ b BD [ Bu6AB [ CD [ I [ a ADS Hu [ ADS Bu [ c BD [ Don [ ADS StF ] c CD [ I [ BD [ See [ c CD [ Stn [ BD ] Bu ] CD [ Bu [ b CP [ Daw 1 AB [ b HR Kui ] a ADS Bu [ CD [ B [ HR Fin [ c CP [ I 386 BC [ a ADS See [ ADS Stt ] HR StF 299 AB ] BD [ Bu [ ADS A ] HR JC 8 AB [ BD [ AC2AB [ ADS Hu [ a CD [ Bu 1004 AB [

4 TABLE 2ÈContinued Discoverer WDS Date h o j *j HR, ADS, DM, etc. Designation HD HIP(a, d J2000.0) (1900]) (deg) (arcsec) (nm) (nm) b b c ADS Daw [ ADS Hu [ b CD [ I [ b CD [ Rst 98 AB [ ADS A [ c ADS A [ CD [ Hwe [ ADS Bu 547 AB ] CD [ Rst [ HR Gle [ b BD [ Rst [ CD [ Rst [ ADS Bu 744 AB [ b b b BD [ Bu [ ADS Bu [ c b ADS Hu ] a BD [ Bu [ HR Kui [ BD [ Rst [ HR HJ 3683 AB [ ADS B [ a CP [ I [ ADS Bu 883 AB ] b ADS Bu [ c BD ] Stt90AB ] ADS Stt ] ADS Bu 314 AB [ ADS Stt ] ADS A [ CD [ CoO [ BD [ A [ ADS Stt ] ADS Bu [ CD [ I [ c BD [ Rst [ a ADS Bu 190 AB [ ADS A [ CD [ I [ BD [ H 68 A [ \ CP [ I [ b CD [ Hu [ c HR Fin [ c CD [ Daw [ b CD [ I 63 AB [ ADS A ] a CD [ CpO [ BD [ Bu [ ADS A 2715 AB ] c HR Dun 23 AB [ CD [ I [ ADS B [ ADS A ] c ADS B [ CD [ R 65 AB [

5 TABLE 2ÈContinued Discoverer WDS Date h o j *j HR, ADS, DM, etc. Designation HD HIP(a, d J2000.0) (1900]) (deg) (arcsec) (nm) (nm) ADS Ho [ a CD [ Bu [ CD [ Bu 755 AB [ CD [ I [ CD [ Fin [ BD [ B [ a ADS A [ BD [ StF971AB [ ADS I [ HR I [ a BD [ Bu327AB [ ADS A 3042 AB [ ADS Bu [ HR Dun [ ADS A [ c ADS A [ a CD [ B [ CD [ Rst [ ADS A ] CD [ I [ CP [ HJ [ HR Fin 324 AB, C [ CD [ I [ CD [ I [ a ADS Hu [ a ADS Bu [ HR See 96 Aa, B [ a ADS StF 1670 AB È [ b b ADS Bu [ c CD [ B [ CD [ HJ [ a a CP [ I [ ADS Bu 148 AB [ ADS StF [ b ADS See [ CP [ I [ a CD [ Bu [ ADS Stt 535 AB ] a ADS See [ a CD [ Bu [ b HR Fin 330 AB [ CP [ I [ ADS Ho 296 AB ] a b HR HdO [ a ADS A 417 AB [ c a HR JC 20 AB [ CD [ I [ ADS A [ ADS Hu [

6 1602 HORCH ET AL. Vol. 121 TABLE 2ÈContinued Discoverer WDS Date h o j *j HR, ADS, DM, etc. Designation HD HIP(a, d J2000.0) (1900]) (deg) (arcsec) (nm) (nm) CD [ See 486 AB [ ADS Hu ] a CD [ See [ CD [ B [ CD [ I [ a b CP [ Slr [ a ADS A ] b a Quadrant ambiguous, but consistent with previous measures in the CHARA 3d Catalog (Hartkopf et al. 1997). b Quadrant ambiguous, but inconsistent with previous measures in the CHARA 3d Catalog. c Quadrant unambiguous, but inconsistent with previous measures in the CHARA 3d Catalog. for the epoch of observation shown. Position angles and separations are given without uncertainty estimates, but reasonable uncertainties may be obtained by combining the measurement precision given in the next subsection with the uncertainty in the scale and detector orientation given in Table 1 in quadrature using standard error formulas. Five objects in Table 2 were unresolved under excellent observing conditions. Four of these (HIP328, 345, 349, and 389) are thought to be members of the open cluster Blanco 1 (Platais et al. 1998). We conclude that, to the limit of our detection capabilities, these objects have no companions. The Ðfth nondetection is HIP25327 (HD 35502); this is the brighter component of the wide double H 68 (WDS 05250[0249) and is known to be a chemically peculiar star (Leone, Catalano, & Malaroda 1997). To the limit of our detection capabilities, we conclude that the irregularities present in the spectrum are not due to a companion Measurement Capability In order to characterize the astrometric accuracy and precision of the measures in Table 2, Ðve subsets of objects were studied in detail. These are (1) those objects for which a very high quality orbit, i.e., an orbit calculated with the inclusion of a substantial quantity of speckle data, already exists, referred to below as speckle orbit ÏÏ objects; (2) those objects for which three or more measures appear in Table 2; (3) several objects listed in the catalog of Worley & Heintz (1983); (4) wide pairs (o º 0A.8) that show very little or no motion over a long history of observations from the WDS; and (5) wide pairs (o º 1A.0) that have contemporaneous measures by other observers in the Third Catalog of Interferometric Measurements of Binary Stars (Hartkopf, McAlister, & Mason 1997). The set of speckle orbit objects consists of 10 stars with 17 measures, for which the orbital information was taken from the series of speckle orbit papers by the group at the Center for High Angular Resolution Astronomy (CHARA), in particular Hartkopf, McAlister, & Franz (1989), Hartkopf, Mason, & McAlister (1996), W. I. Hartkopf (1998, private communication), and Mason (1997). These authors have published uncertainty estimates as well as the orbital elements of each system, and we have used these to compute the ephemeris positions and their uncertainties for the epochs of observation listed in Table 2. In order to compare with the ephemeris positions, our measures were precessed to the J equinox, though corrections in all cases were very small. Figure 2 shows the observed minus calculated residuals in separation (Fig. 2a) and position angle (Fig. 2b). There appears to be no signiðcant systematic o set in the determination of the separation, but the average residual in position angle is approximately [1, a 1.1 p o set. Table 3 shows both the average value and rms deviation obtained FIG.2a FIG. 2.ÈResiduals as a function of ephemeris separation. (a) Separation residuals for all objects in Table 2 having speckle orbit determinations. (b) Position-angle residuals for the same set of observations, excluding two objects with very large position-angle uncertainties. In both plots, open circles are measures obtained in 1995 and Ðlled circles are measures obtained in The error bars are the ephemeris uncertainties as calculated from the published orbital parameters. The gray band marks the region below the di raction limit of the telescope at 503 nm. FIG.2b

7 No. 3, 2001 SOUTHERN DOUBLE STARS. II TABLE 3 SUMMARY OF RESIDUALS, SPECKLE ORBITS Number of Average Residual rms Deviation from Orbit Requirement Sample Measures (obs[eph) Average Residual dh ¹ *h \[0.64^1.25 p \ 3.31 `1].27 eph h ~0] *h \[1.77^1.34 p \ 2.33 `1].59 h ~0].52 All 12 *h \[1.02^0.89 p \ 2.95 `0].85 h ~045 do ¹ 5.0 mas *o \[2.2 ^ 5.9 mas p \ 15.7 `6.0 mas eph o ~ *o \]3.4 ^ 4.2 mas p \ 9.4 `4.5 mas o ~1.8 All 14 *o \]0.2 ^ 3.7 mas p \ 13.2 `3.4 mas o ~1.9 for both coordinates using objects from the sample, subject to the orbit quality criteria listed. The values have been computed for 1995 and 1996 observations separately, as well as for the entire data set. As the orbit criteria are very stringent, we may assume that the rms deviations are good estimates of the measurement precision of the telescopedetector combination. Note that the precision achieved during 1996 appears to be signiðcantly better than that during Error bars for the average residuals in Table 3 are simply the rms deviations divided by (N [ 1)1@2, where N is the number of measures, while the uneven error bars for the rms deviations are derived from a s-squared analysis. Six objects in Table 2 have three or more observations that were taken close together in time (often two on the same night). By studying the standard deviations of the separation and position angle values, further information regarding the accidental measurement errors can be obtained. Figure 3 shows the standard deviations in separation (Fig. 3a) and position angle (Fig. 3b). Dashed lines are drawn in each plot indicating a random linear measurement error of 13.2 mas (equivalent to 0.76/o degrees in position angle); these show that the random errors are fully consistent with the measurement precision obtained in the speckle orbit study. Figure 4 shows the standard deviations in separation plotted (Fig. 4a) as a function of the total V magnitude of the system and (Fig. 4b) as a function of the magnitude di erence of the system (as listed in the Hipparcos Catalogue; ESA 1997). The error bars shown for the magnitude di erences are 0.14 mag; this is the mean Ðgure obtained by Mignard et al. (1995) for Hipparcos data. A survey of the speckle literature shows that measurement precision is related to the aperture size of the telescope used. For example, Hartkopf et al. (1989) obtained mean rms errors of 3.5 mas when comparing their observations taken at 4 mèclass telescopes with well-determined orbit predictions, while the US Naval Observatory results at the 66 cm refractor have typically yielded rms errors of 14 to 18 mas (Douglass et al. 1999; Germain, Douglass, & Worley 1999). These groups used similar instrumentation and reduction techniques. One would therefore expect results from a 2 mèclass telescope such as CASLEO to lie between these two extremes, though the results presented here are much closer to the small-aperture case. A potential limitation of measurement precision is that the lens system in the speckle camera exhibits other aberrations, in addition to OFAD, that limit the overall optical performance. This issue is currently under investigation. It is also possible that there is some degradation of the di raction-limited pointspread function due to the telescope optics. However, we believe that at least a large part of our precision limitation may be explained by the combination of the average seeing conditions during our observing runs and the low quantum efficiency of the MAMA detector used. A similar loss of precision was seen during inferior observing conditions when the system was used at the Wisconsin-Indiana-Yale- FIG.3a FIG.3b FIG. 3.È(a) Separation standard deviations as a function of average separation (from Table 2) for objects observed three or more times close in time. The line marks the rms deviation obtained from the speckle orbit study. (b) Position angle standard deviations for the same set of objects. The line is the function p \ 0.76/o, where the units of p are degrees and for o are arcseconds. In both plots, open circles indicate measures obtained in 1995 and Ðlled circles indicate measures h obtained in The gray h band marks the region below the di raction limit of the telescope (at 503 nm).

8 1604 HORCH ET AL. FIG.4a FIG. 4.È(a) Separation standard deviations as a function of total magnitude for the objects in Table 2 observed three or more times close in time. (b) Separation standard deviations as a function of magnitude di erence (as given in the Hipparcos Catalogue) for the same objects. The line marks the rms deviation obtained from the speckle orbit study. In both plots, open circles indicate measures obtained in 1995 and Ðlled circles indicate measures obtained in FIG.4b NOAO (WIYN) 3.5 m telescope during 1997 (Horch et al. 1999), whereas under good observing conditions, the system gave results comparable to those of Hartkopf et al. (1989). Many of the objects appearing in Table 2 have orbit determinations listed in the catalog of Worley & Heintz (1983), and while the quality of these orbits is in general not high enough to draw meaningful conclusions about the precision of the measures presented here, they can be used to examine the possibility of systematic o sets in the position angle and separation measures at larger separations than is possible with the speckle orbit objects (which all have separations smaller than D1A). Six such objects are shown in Figure 5, spanning a wide range of separations. The plots indicate that, where there are systematic o sets between the visual orbit and the speckle observations of other observers, our measures are in good agreement with the other speckle data points. Because of the relatively poor seeing for many of the observations presented here, the issue of measurement accuracy at large separation is of special concern. For largeseparation systems observed in poor seeing, the secondary may begin to fall outside the isoplanatic patch, so that the speckle patterns generated by the primary and the secondary will not be identical, leading to lower signal-to-noise ratio in the autocorrelation and power spectra obtained. This in turn will a ect the ability to measure the separations and position angles accurately. In order to characterize the situation more fully, we Ðrst examined stars that have exhibited little or no motion over a decades-long history of observation in the WDS. A second group of wider doubles that have contemporaneous observations (meaning within 2 yr of our observation) by other speckle observers was also examined. Table 4 contains the results of these two studies. The most signiðcant average o set is in the position angle of the contemporaneous pairs, a 2.3 p result. Nonetheless, the very slow moving pairs, which span a similar separation range, exhibit no signiðcant o set in position angle. We conclude that, taking the results of all the studies presented in this section together, the o sets obtained here represent small-sample statistics and that overall there is no strong indication of signiðcant systematic o sets in the astrometry over the complete range of separation represented by the objects in Table 2. The rms residuals in Table 4 are, however, substantially larger than for the speckle orbit study. While some of this increase may be attributable to the relative motion of the components, it is likely that our speckle measurement precision does deteriorate for larger separations. A Ðnal issue of importance for the Ðve nondetections listed in Table 2 is the detection capability of the telescopedetector combination. Figure 6 shows a plot of magnitude di erence versus total system magnitude for all double stars in Table 2. Both the total V magnitudes and the magnitude di erences were taken from the Hipparcos Catalogue (except for the Ðve stars in Table 2 that are not listed in the Hipparcos Catalogue, in which case the values were taken from the WDS), and our measured separation is encoded in the shade of the plot symbol, with larger separations shown as lighter shades. This plot indicates that, for separations smaller than 1A, magnitude di erences of up to 2 can be TABLE 4 SUMMARY OF RESIDUALS, LARGE SEPARATION OBJECTS Separation Range Number of Average Residual rms Deviation from Object Sample (arcsec) Measures (obs[eph) Average Residual Very slow moving pairs È *h \]0.13^ 0.47 p \ 1.42 `0].46 h ~0].23 *o \[14.4 ^ 10.7 mas p \ 32.1 `10.4 mas o ~5.3 Pairs with contemporaneous speckle measures (within 2 yr) È *h \]0.59^ 0.26 p \ 0.88 `0].24 h ~0].13 *o \]6.8 ^ 8.7 mas p \ 30.2 `8.2 mas o ~4.5

9 FIG.5a FIG.5b FIG.5c FIG.5d FIG.5e FIG. 5.ÈSix objects that have orbits in the catalog of Worley & Heintz (1983). In all plots, squares indicate speckle observations by other observers, circles represent measures listed in Table 2, the cross indicates the origin, and line segments are drawn from the ephemeris positions on the orbit to the data points. North is up and east is to the left in all cases. (a) I65\ WDS 06573[3530; the grade 1 orbit drawn is that of Finsen (1963). (b) JC8AB\ WDS 03124[4425; the grade 2 orbit is that of Heintz (1979). (c) HJ 5084 \ WDS 19064[3704; the grade 2 orbit is that of Heintz (1963). (d) Bu738\ WDS 02232[2952; the grade 4 orbit is that of Finsen (1969). (e) StF 1670 AB \ WDS 12417[0127 \ c Vir. The orbit plotted in this case is the recent result of Girard et al. (2000) and not the grade 2 orbit of Strand (1937) appearing in Worley & Heintz (1983) (for clarity, only a portion of the Girard et al. orbit is plotted). ( f ) HJ 3683 AB \ WDS 04403[5857; the grade 4 orbit is that of Valbousquet (1981). FIG.5f

10 1606 HORCH ET AL. Juan speckle interferometry program. These measures were obtained at the 2.15 m CASLEO telescope at El Leoncito, Argentina using a MAMA detector system operating in time-tag mode. The measures of binary stars with very well known orbits exhibited residuals consistent with a linear measurement error of 13.2 `3.4 mas, a Ðgure that is higher than expected for this size ~1.9 aperture. We believe that the measurement precision is reduced by factors such as average seeing conditions at the site and low quantum efficiency of the detector. Four members of the open cluster Blanco 1 were unresolved during excellent observing conditions, and the chemically peculiar star HD was likewise unresolved. FIG. 6.ÈMagnitude di erence as a function of total V magnitude for all systems appearing in Table 2 (except nondetections). The shade inside the plot symbol is an indication of the separation of the system; white is used for systems with separations greater than 1A and as the separation decreases below 1A, the shade darkens. Black represents a system at the di raction limit of the telescope. reliably detected and measured. For larger separation systems, fainter companions can be measured, up to 3È3.5 mag fainter than the primary. 4. CONCLUSIONS We have presented a total of 203 speckle observations of double stars and other targets as a part of the Yale-San We are grateful to H. Levato of CASLEO for his vigorous support of the project, and to I. Platais and S. Malaroda for suggestions regarding observing targets. We also thank Z. Ninkov, R. Meyer, and the referee for their interest and contributions. This work was made possible through NSF support of Yale Southern Observatory (in particular grant AST ), and publication funds were made available by the Chester F. Carlson Center for Imaging Science, Rochester Institute of Technology. Douglass, G. G., Mason, B. D., Germain, M. E., & Worley, C. E. 1999, AJ, 118, 1395 ESA 1997, The Hipparcos and Tycho Catalogues (ESA SP-1200) (Noordwijk: ESA) Finsen, W. S. 1963, Republic Obs. Circ., 7, 39 ÈÈÈ. 1969, Republic Obs. Circ., 7, 190 Germain, M. E., Douglass, G. G., & Worley, C. E. 1999, AJ, 117, 1905 Girard, T. M., et al. 2000, AJ, 119, 2428 Hartkopf, W. I., Mason, B. D., & McAlister, H. A. 1996, AJ, 111, 370 Hartkopf, W. I., McAlister, H. A., & Franz, O. G. 1989, AJ, 98, 1014 Hartkopf, W. I., McAlister, H. A., & Mason, B. D. 1997, Third Catalog of Interferometric Measurements of Binary Stars (CHARA Contrib. No. 4) (Atlanta: Georgia State Univ.) Heintz, W. D. 1963, Vero. Sternw. Mu nchen, 5, 263 ÈÈÈ. 1979, PASP, 91, 356 Hoffleit, D., & Jaschek, C. 1982, The Bright Star Catalogue (4th ed., New Haven: Yale Univ. Obs.) Horch, E. P., Dinescu, D. I., Girard, T. M., van Altena, W. F., Lo pez, C. E., & Franz, O. G. 1996, AJ, 111, 1681 (Paper I) REFERENCES Horch, E. P., Girard, T. M., van Altena, W. F., Meyer, R. D., Lo pez, C. E., & Franz, O. G. 1997, in Visual Double Stars: Formation, Dynamics, and Evolutionary Tracks, ed. J. A. Docobo, A. Elipe, & H. McAlister (Dordrecht: Kluwer), 43 Horch, E. P., Ninkov, Z., van Altena, W. F., Meyer, R. D., Girard, T. M., & Timothy, J. G. 1999, AJ, 117, 548 Jacoby, G. H., Hunter D. A., & Christian, C. A. 1984, ApJS, 56, 257 Kasle, D. B., & Morgan, J. S. 1991, Proc. SPIE, 1549, 52 Leone, F., Catalano, F. A., & Malaroda, S. 1997, A&A, 325, 1125 Mason, B. D. 1997, AJ, 114, 808 Mignard, F., et al. 1995, A&A, 304, 94 Platais, I., et al. 1998, AJ, 116, 2556 Strand, K. A. 1937, Ann. Sternw. Leiden, 18, 77 Valbousquet, A. 1981, A&AS, 43, 23 Worley, C. E., & Douglass, G. G. 1997, A&AS, 125, 523 Worley, C. E., & Heintz, W. D. 1983, Fourth Catalog of Orbits of Visual Binary Stars, Pub. US Naval Obs. (2d Ser.), 24, pt. 7