A little more than a century and a half has passed since F.G.W. Struve began to provide the astronomical world with visual double star measures of both quantity and quality. Since that time, nearly a million individual measures have been made by all techniques, and most of these are now available in machine-readable form in the punch-card Observation Catalog maintained at the Naval Observatory. As of 1 March 1981, this Catalog contained 877242 measures grouped into 398818 means. The data file is now virtually complete for all 20th century observations, and perhaps 80% complete for the 19th century material. Additions are being made steadily and we can look forward to having a complete set of observations for every known double star in a few more years.

When a project finds enough material from the brighter stars, without high demands on instruments and observing effort, so much the better. But the very red dwarf stars certainly are not in that category. To show how limited the knowledge still is: Assume optimistically that the star count to 5 pc is complete. Extend the distance merely to 7 or 8 pc, i.e., scan a 3 or 4 times larger volume, and the present count is already short by several dozens of stars—practically the entire dM5–M8 range.

At these objects—with very little luminosity in the visible range—the observationally accessed relationship between mass and luminosity ends, and the peak in the luminosity function need not mean a peak of the mass-frequency function.

The discovery of new pairs in a given class of separation normally brings new orbits; the problem is to know how the number of such orbits will increase with time from the discoveries. A study of the Lick stars (A and Hu) leads to an estimate of the future of Nice pairs in this respect and could also give some idea of what to do with discoveries made from space vehicles.

E. Hertzsprung developed the multiple exposure technique for observing double stars photographically around 1914 at Potsdam (Hertzsprung, 1920). The technique consists in taking one or more rows of 17 to 35 exposures of a small field which normally includes only the double star itself. The total number of exposures on a plate can vary from 17 to 140, depending primarily on the separation of the double star and the anticipated systematic effects in the observational system. An important feature of the technique is the use of objective gratings to substantially reduce the problem of magnitude error. The large number of exposures serve to reduce the random error, while the objective gratings, together with other innovations introduced by Hertzsprung, help to minimize the systematic errors. In this way Hertzsprung provided us with an observational technique of considerably higher accuracy than the traditional visual micrometer work.

Attempts to detect double stars are being made with the Two-Color Refractometer (TCR). The refractometer is an instrument developed by the Quantum Electronics Group of the University of Maryland to measure atmospheric refraction.

During check-out observations using the U.S. Naval Observatory’s (USNO’s) 24“ reflecting telescope, the refractometer is being used to detect the astrometric separation of the ultraviolet and red photocenters of the combined light from a double star, differentially referenced to nearby stars. The refractometer contains two dispersive wedges which null the incoming dispersion; the positions of the wedges then provide a measure of the dispersion of the incoming light. The angular part of this measure immediately gives the position angle of the double. The angular distance between the two components of the pair is a combined function of the measured quantity and of the magnitude and color differences of the components of the double star, and is therefore obtainable only with additional information.

This paper discusses the application of apodization to reflecting telescopes. The diffraction pattern of a telescope, which is the image of a star, can be changed considerably by using different aperture shapes in combination with appropriately shaped occulting masks on the optical axis. Aperture shapes studied were the circular, square, and hexagonal. Polaris (α-UMin) was used as the test system.

Preliminary results of a new photographic technique applied to visual double stars are given.

This technique consists in the determination of the ϱ, θ and Δ m from measurements made on two pairs of trails left by the components of a binary on a photographic plate, obtained through an adequate movement of the telescope in R.A. and declination.

Photocentric orbital motions from 60-year Sproul plate series are shown for three visual binaries, ζ Her, 85 Peg and Ho 296. Mass ratios of the first two pairs, which have large-amplitude blended orbits, carry internal errors at the level of one percent. Orbital elements computed independently from the blended photographic observations agree closely with the well-determined visual orbits. Ho 296 serves as a test of the Sproul plate series to detect a 20-year orbit with an amplitude of two, or one, micron. Simulations of orbital motion in single star residual series confirm this detection capability, which corresponds to planetary mass for dark objects around nearby stars. A mass-luminosity diagram of the lower main sequence is shown, including one for invisible astrometric companions.

A high percentage of double and multiple stars among those in T-associations was first paid attention to by Prof V.A. Ambartsumian. Further at the beginning of sixties many investigators of variable stars (Kukarkin, Herbig et al.) paid attention to the necessity of studying duplicity of stars in T-associations. Just then the first lists of wide pairs in T-associations appeared.

We set a task to search for the Trapezium type multiple stars in T-associations involving the groupings of T Tau type variable stars from Kholopov’s list at a distance up to 500 pc. Such associations proved to amount 19, but out of them we succeeded in searching only in 12 ones.

Problems:

1.Size of diffraction image

2.Ross effect (shrinking of high exposed star images in the photographic emulsion)

3.Halo by diffusion of light in the photographic emulsion

4. Scintillation

Methods of improving the observation conditions are as following:

1.Deformation of diffraction image with quadratic or other formed masks

2.Diffraction gratings render possible a better measurement but don’t avoid Ross effect and halo of diffusion.

The halo of diffusion can reach much more of diameter than the original image.

The systematic search for stellar multiplicity by observations of lunar occultations began in 1969. David Evans will review a large portion of the data collected since then in the second paper of this session. Here we will outline the historical development of the technique and comment on its limitations, contributions to double star research, and future improvements.

Sir John Herschel (1865) may have been the first to suggest that “a double star, too close to be seen divided with any telescope, may yet be detected to be double by the mode of its disappearance.” However, the visual discovery of duplicity by occultation is not sufficient, for the observations are not subject to analysis after the fact. The power of the occultation technique was realized more than 100 years after the suggestion, with the application of photomultipliers and electronic digital data handling. Whitford (1939) took the first step in using a cesium photo tube, oscilloscope, and moving film to record the occultations of Beta Capricorni and Upsilon Aquarii. The resolution of the spectroscopic binary Beta Capricorni was discussed, but he concluded his data were not able to resolve it, an observation which was later accomplished at several observatories and summarized by Evans and Fekel (1979).

When systematic photoelectric observations of occultations were started some dozen years ago it was realised that they yielded a significant crop of double star discoveries and observations. The limitations of the method were well recognised: only about ten percent of the area of the sky would ever be available and one could not choose which stars to be observed. Moreover the data obtained from a successful observation of a double star are less comprehensive than the normal visual observation. The result is the vector separation or the true separation projected along a line perpendicular to the actual lunar limb, that is the position angle of the point of occultation modified by the slope of the limb at that point.

The catalogue is divided into three sections: bright stars with magnitudes less than or equal to 6.7: other stars having numbers in the Smithsonian Astrophysical Star Catalogue with magnitudes greater than 6.7: and other stars with no SAO numbers, usually faint and with often only fragmentary information concerning their magnitudes or spectra.

Each section has the following columns: SAO number, other name, magnitude, spectral type, run number of the observation, date in obvious coded form, grade — grade zero indicates no duplicity, grade 1 possible duplicity, grade 2 probable duplicity, and grade 3 certain duplicity. Grades are based on statements by observers. For grades other than zero the vector separation and its computed error are given in arc milliseconds. P.A. denotes the direction in which the vector separation is measured, that is the position angle on the lunar limb as modified by the slope deduced in favorable cases from the spacing of the diffraction bands on the occultation trace. The next column gives the limb slope and its error.

The occultations observations are traditional at Abastumani Observatory. About 3 years ago the work was started on the development of the observational equipment for measuring different characteristics of various events under investigations with as high efficiency as possible, but using relatively small aperture telescope. The program is in progress, so, at present, the following pulse-counting system is built and put into operation.

This review considers three main areas, leaving several others to be discussed in more detail in the contributed papers of this session.

1. 1. The need for spectrographs and measuring instruments of great stability for long-term projects such as radial velocity observations of visual binary stars.

2. 2. The use of cross-correlation devices, both analog (radial velocity scanners) and digital, for radial velocity measurement.

3. 3. The use of comparison spectra impressed directly onto the starlight and of polarisation instruments as means to very precise radial velocities.

The frequency distribution of SB’s over apparent visual magnitude emerging from the catalogue of Batten et. al. (1978) shows a very steep decrease of the number of spectroscopically detected SB’s already for such bright stars of magnitude 7. Considering the number of all stars in the individual magnitude intervals, we find a kind of completeness parameter of the spectroscopic surveys: If we scale it somewhat optimistically at 100% between 0 and 3 mag, we see a 50% decrease of the completeness of our knowledge of stellar radial velocities already for stars fainter than 4.5 mag.

This situation is mainly due to the fact that the measurement of radial velocities with conventional slit spectrographs is extremely laborious, requiring long exposure times at large telescopes for the exposure of only one spectrum at a time. – Therefore more efficient methods for radial velocity determinations of fainter stars are urgently needed.

I will discuss one of the results of a continuing program to determine accurate masses and radii of stars in eclipsing binaries. This program actually began in 1976 with my work on the faint M-dwarf system CM Dra (Lacy 1977). By faint I mean . I needed high-dispersion spectra to get radial velocities, and an intensified solid state array detector called the Digicon, which had recently been installed on the coude spectrometer of the 2.7m reflector at McDonald Observatory, turned out to be the answer. 30 minute integrations were sufficient to get accurate radial velocities from the Hγ emission lines. The Digicon is very good at getting crumby spectra of faint objects fast. By crumby, I mean signal-to-noise ratio of less than 100.

Recent advances in detector technology as well as the launching of the International Ultraviolet Explorer (IUE) satellite have opened new spectral regions for extensive spectroscopic observation. A program to determine the feasibility of radial velocity measurement for binaries whose secondaries are of earlier spectral type has been initiated with the IUE satellite. It is expected that velocities with uncertainties as small as 2-5 km s−1 may be obtained for some stars. This observing program includes tau Per and other composite spectrum binaries as well as several early-type binaries.

For the past five years, a Reticon solid-state photodiode array detector has been used at the coudé focus of the 2.7 m McDonald Observatory reflector to obtain radial velocities of spectroscopic as well as short-period visual binaries. The high quantum efficiency, red sensitivity, and stability of this detector system permit velocities of high accuracy to be determined and previously unseen spectroscopic components to be detected.

A knowledge of the spectral types of visual components is still basic to most intensive studies of visual systems. Although there are photoelectric techniques that give good quantitative information on certain classes of stars, the only way at present to identify most classes of stars, from mercury-manganese stars to barium stars, is by obtaining their spectra. The spectra can be obtained from objective prisms, slit spectrographs, or spectrum scans, but the important things are that they (1) show a large portion of the spectra, preferrably in the blue-violet for most classes of stars, (2) have sufficient resolution and signal-to-noise to identify many spectral lines, and (3) be standardized as an independent system. The last is important because if spectral type standards are changed so that they will agree with other information, such as colors, we lose the benefits of a comparison, such as deriving interstellar reddening.

This paper is a report on a new technique for detecting close triple stars and on a current observational project with some preliminary results, using this method. The goal of the project is to discover new triple systems where the ratio between the long and the short orbital period is small. A system will be called a close triple system if its short orbital period is of several days.

In most of the known triple stars (Fekel 1981), the period ratio is very large. This can be attributed to the fact that the wide orbit has been detected visually in most cases. In the few exceptions (e.g. λ Tau, Ebbighausen and Struve 1956), the wide orbit was discovered by detecting additional periodicity in the derived velocity of the close pair mass center γ As this small modulation is imposed on the large radial velocity variation associated with the orbit of the close pair, many triple systems could have escaped this kind of detection. Consequently, the present typical or averaged period ratio might be highly overestimated.