How can an apparently cut-and-dried textbook matter such as orbits become the topic for an international conference? The answer is found in statements from the reports of presidents of Commission 26 which criticize the existing state of affairs and is further elaborated upon by van den Bos (1962) in his well-known query “Is this orbit really necessary?” The objections refer to useless repetitions or multiple solutions and results that are distinctly inferior to observational accuracy. Unanimous agreement on what constitutes a useful orbit, or revision, is unlikely to be achieved, but considerable published work which is based more on computer experience than on double-star knowledge is clearly of inferior quality. To some extent, errors of observation (including poor distribution of data, etc.) are unavoidable, but published material contains errors due to the method of analysis which could have been avoided.

In connection with statements made at the preceding session, some results presented at the Nice Conference on systematic and random errors affecting the elements a and P as well as a3/P2 may be recalled. If many orbits have been calculated for one binary, it is instructive to plot log P, log a, and log a3/P2 versus log Δt/P where Δt is the time interval covered by observations. As a general rule, when Δt ≃ P the systematic trends vanish, and the scatter of the plotted elements reaches a low value, not considerably changing thereafter. It may well be of interest, then, to have several orbits based on the same observations but computed by different authors and methods.

The number of visual binaries for which orbits have been computed now exceeds 600, but this laboriously accumulated wealth of data remains relatively sterile owing to the paucity of reliable parallaxes. Hertzsprung (1964) had this in mind when he revived the idea he first voiced in 1911 and proposed a new luminosity parameter defined as the absolute magnitude of a star reduced to unit mass without changing its density or surface brightness. The great advantage of this parameter is the fact that it is readily derived from the period and semi-major axis of a visual binary without recourse to the parallax. It therefore permits immediate use of most of the computed orbits for the construction of a luminosity-colour diagram analogous to the classical H-R diagram but entirely free of the scatter resulting from errors in the parallaxes.

A substantial part of our understanding of the physical nature of stars rests upon observations of visual binaries. In order to secure a complete interpretation of a binary system we require these data: semi-major axis, parallax, period, magnitudes, colours, spectral types and velocities. It goes without saying that for very few systems are all these desiderata available, largely through technical difficulties of observing two images in close proximity to one another - and these difficulties are particularly serious in carrying out photometry. Nevertheless, the usefulness of photometry of binaries is so great that no effort should be spared in resolving the observational and technical problems.

I hope that you will excuse a relative newcomer to this field and a spectroscopist for presenting his views, which may be somewhat different than those of others. I would like to enumerate the aims of research on visual systems and the needs of the field, particularly for spectroscopic data, as I see them.

I. Aims of Multiple-Star Astronomy. (1). Statistics of multiple systems. We would like to know the fraction of stars that are visually double, triple, etc. and in each case as a function of spectral type. For instance, is the frequency of visual doubles the same for Population II stars as for Population I stars? Or are the frequencies of visual doubles the same for stars off the main sequence as for their antecedents on the main sequence? For such information a negative result on duplicity is as important as a positive result, i.e. we wish to know which stars are not seen to be visually double as well as which ones are seen as doubles.

La notion d'étoiles faibles concerne deux catégories d'objets: une étoile peut être de faible éclat parce qu'elle est loin, ou bien parce qu'elle est de faible luminosité. D'ailleurs ces deux facteurs se combinent entr'eux, selon les distances au soleil, de sorte que beaucoup d'étoiles géantes sont des étoiles faibles.

Il est évident qu'une étoile double faible n'est intéressante que dans la mesure où son mouvement est appréciable, ce qui élimine les géantes. Notre rôle est done de dissocier parmi les étoiles faibles, celles qui le sont intrinsèquement.

Avant d'entamer le vif du sujet, il est bon de rappeler que si nous étudions les étoiles doubles depuis trois siècles, ce n'est que depuis trente ans que nous avons réellement des connaissances sur les étoiles à faible luminosité. Mais faut-il encore définir ce que nous entendons par là.

When IAU Symposium No. 17, the most recent international conference on visual double stars to take place in the United States, was held in Berkeley in 1961, there was little discussion on observational techniques and no introductory remarks were presented on the subject. The reasons for this are obvious. The classical techniques of double-star observation, those that provided virtually all the data available on visual double stars, were too well known to require introduction or discussions; other methods, with the possible exception of early image tube and electronic camera experiments, did not exist.

Today, more than ten years later, most double-star work continues to be carried out by the same classical techniques. However, several methods have come into use or have recently been developed that are capable of contributing importantly to visual double-star research.