The β CMa stars are short-period pulsating variable stars lying slightly above the upper main sequence. Although they have been known and studied for 70 years, little is known about the nature and cause of their pulsation.

STOTHERS and SIMON (1969) have recently advanced an hypothesis – hereinafter known as the hypothesis – which proposes that all β CMa stars are the former secondary components of massive binary systems. The former primary components, being more massive, have evolved first, have shed their envelopes by rapid mass loss, and have deposited helium-enriched material from their cores upon the surfaces of the secondary components. This causes the secondary components (now the primary components) to become unstable, via the so-called „μ-mechanism“, to nuclear-energized radial pulsations.

A discussion of radial velocities determined at the Radcliffe Observatory for 20 southern Cepheids and those found in the literature for another 20 stars lead to the conclusion that 15% of Cepheids are members of long period binary systems (1). The observational data were heterogeneous, few stars having been observed by a single observer or with the same instrument over a sufficient length of time to detect long period variations in the mean velocity.

High dispersion spectra in the blue violet region and OAO II spectral scans in the UV of the star a Centauri are presented. The variation of equivalent widths and radial velocity are presented. The appearance and disappearance of the forbidden lines of helium, the change in the wing structure of λ 4471 of He I, the strengthening of the UV resonance lines at phase 0.5 and the presence of the UV feature of λ 1720 indicate that the star’s atmosphere is experiencing drastic density changes.

A model is proposed for BM Ori to explain why DOREMUS saw only the early B star in the spectrum, whereas the light curve seemed to predict an equally bright cooler star which should have been seen alone at supposed totality. The cool star is assumed to be a disk-shaped object, similar in shape to that proposed by HUANG for ε Aur, seen nearly edge-on. By means of various restraints, it is estimated that the disk covers about 55% of the B star at primary eclipse. This leads to the relative luminosities (Lh = 0.74, 0.87 and 0.92 in V, B, and U), the color of the disk (the opposite side of the disk has the color of a G or F giant), the relative geometrical dimensions (rB = 0.09, rD = 0.25, and the height of the disk h≈0.1), and the orbital inclination (i ≈ 90°). The model with these parameters predicts a theoretical light curve which represents the observations as well as did the solution of HALL and GARRISON, which was based on spherical stars.

A method of photoelectric photometry of the components of visual double stars is briefly described. Observations of ADS 9701 = δ Ser and of UV Aur = ADS 3934 B are presented to demonstrate the usefulness of this technique for the accurate photometry of variables in close visual double stars.

The existence of eclipsing binaries, though extremely rare, in stellar systems like globular clusters, should attract the observers, because one could obtain fundamental data on highly evolved stars. The variable star V 78 in the globular cluster NGC 5139, which was found by BAILEY (1901), is one of these rare cases. MARTIN (1938) derived the period P = 1ḍ168118, and showed by means of an extensive photographic photometry that it is an Algol type eclipsing binary. Recently SISTERO et al. (1968, 1969) rediscussed the observations of MARTIN, and with additional measurements derived the following light elements:

The present investigation is to determine what apparent period changes if any, would result from binary motions. In particular it is important to know whether or not such apparent period changes could be mistaken for intrinsic period changes caused by evolution.

An extremely important new frontier in variable star research lies in the application of image tube techniques to the study of the spectroscopic phenomena in extremely rapid intrinsic variables and in short-period binary systems. Within limits, the use of image tubes for the spectroscopic observation of constant or slowly varying stars can be regarded as a convenience — as a means of increasing the efficiency of use of time at the telescope; the same results could, in principle, be obtained by observing for a longer time by conventional means. However, when one wishes to observe rapid changes, there is no substitute for the higher information collection rate provided by the image tube. Electronographic image tubes are especially well suited to this type of investigation owing to their high quantum efficiency, high resolution, large storage capacity, and linearity of response. Typically, image tubes of this type may have quantum efficiencies of 10% to 20%, an information-rate gain of 10 to 20 over the fastest photographic plate (baked Kodak IIaO at 4000 A), and resolutions between 80 and 100 lp/mm. With proper choice of photographic emulsion to record the electronic images, a linear relationship exists between photographic density and intensity of the incident light up to densities of about six, making electronographic image tubes especially valuable for spectrophotometric observations. Despite the great potential of this technique, very little work of this type has so far been done. In this paper I shall outline some of the results that have been obtained at Mount Hamilton to illustrate the possibilities of this type of observational research.

RW Arietis was observed photoelectrically in the UBV-system during fall and winter 1966 with the 70 cm telescope at the Lunar and Planetary Laboratory Catalina station, University of Arizona, Tucson, on 19 nights. Peculiar variations of the usually stable RRc light curve observed on three nights can be interpreted as a superposition of the RR Lyrae type variation and an eclipsing type variation.

Two periods can be distinguished in the past history of our subject. The first period, which begun about 100 years ago, was dominated by intensive studies of novae and U Gem type stars at their outbursts and led to a fairly detailed, photometric and — particularly — spectroscopic description of these phenomena. The faintness of all these objects at minimum was a seriously hampering circumstance for observing them spectroscopically outside of their outbursts and it was only about 30 years ago that two pioneering surveys — of novae by HUMASON (1938) and of U Gem type stars by ELVEY and BABCOCK (1943) — begun to unravel their true nature.

Nova Delphini 1967 is outstanding among other recent novae since it belongs to the rare class of extremely slow objects. Another unusual characteristic is the pronounced pattern of the emission lines during the nebular stage.

Some characteristics of the small-amplitude, quasi-periodic light fluctuations common to novalike variables are described. There appears to be a weak correlation for the majority of the objects between the dispersion about the mean magnitude and the interval between outbursts such that systems with infrequent outbursts exhibit the largest dispersions.

High resolution spectral scans of some Be stars, taken with an Isocon television camera, show changes in the Ha line profile in times of the order of a minute or less.

The well-known problem one has in constructing zero age contact binaries stems from the fact that given two mass values, the ratio of the radii for zero age stellar models differs from that derived from the Roche model (KUIPER 1941). Therefore one cannot achieve contact by adjusting just the distance of the two stars, since if for instance the distance would be such that the critical equipotential surface is of the right volume for the primary to fill it completely, then the secondary would be smaller than its critical equipotential surface and one would end up with a semidetached system.

Test calculations have been made to study the effects of central hydrogen depletion, luminosity exchange and mass exchange on the evolution of a contact binary. The Munich Henyey programme was adapted to allow the study of both components in a time-dependent way.

The first model considered was that of LUCY (Ap. J. 151, 1126) in which entropy is assumed to be at all times uniform throughout the adiabatic portion of a common convective envelope. This model was found to be unstable, in the sense that mass flows from the primary to the secondary in a time short compared with the nuclear time-scale.

The question of whether W UMa stars are contact binaries or not is still a problem. It is promising to see that the theoretical investigators are on the way to explain that and how contact binaries can really exist, even at age zero and with different masses. Let us assume — as is generally done — that the surfaces of the two components of a W UMa star can be set identical to the equipotential surfaces of the Roche model of two mass points. If these surfaces coincide with the limiting Roche lobe or even exceed it for a little amount, than we are justified to call the figure a contact system. However, it may well be that one star fills its critical lobe and the other one is a little bit below it, or even both components may be below their limiting lobes by a small amount. The uncertainty of the elements that can be derived from W UMa light curves for the time being does not allow for a definite decision. We have to look, therefore, for the possible differences between the light-curves of very close, not contact binaries and the variability of contact pairs.

The selection of close binaries which compose the usual tabulations of solutions has informed us in detail on main sequence stellar atmospheric and interior parameters, and on the theories of the tidal and radiative interactions, and has permitted the creation of the theory of binary evolution. As we have seen, the eruptive doubles are beginning to yield to modern instrumentation. Adding more pairs to those already „understood“ will bring continuity into our appreciations of these processes and undoubtedly change them sowewhat. The origins of close pairs, however, remain obscure and one can only say that the theories which describe fission of a proto-star after its core de-couples from the envelope seem promising. The latest presentation of this idea by BODENHEIMER and OSTRIKER (1970) can be viewed as the successor to the indicative theories of ROXBURGH (1966) and CAMERON (1969). In this way of looking at binary origins it is simply a question of when fission occurred in the proto-stellar interval. Observationally, candidate stars are anything but numerous and it is for this reason that we have begun the program described here.

AH Virginis was discovered by GUTHNICK and PRAGER (1929) and since that time the system was observed photographically, visually and, more recently, photoelectrically. The variable belongs to the class of eclipsing binaries of the W UMa type. It has a period of 0.41 days and spectrographic as well as photometric elements have been derived by CHANG (1948) and BINNENDIJK (1960), respectively. The most interesting result of discussions by KWEE (1958), BINNENDIJK (1960), HERCZEG (1962), and PURGATHOFER and PROCHAZKA (1967) is the fact that the residuals between the observed and the computed epochs of minima cannot be represented by a linear relation and show the tendency of producing sudden variations in the period of the system.

The existence of hot spots on components of semi-detached Algol systems has been suspected for a long time, the reason being the distortions observed in many light curves outside of eclipse. However, the question whether such a distortion is caused by absorption or by additional emission and, in the latter case, from which component the emission may come from, could be answered only recently, now that we have a good idea of the trajectories which gas streams follow.

Ultraviolet photometric observations by OAO-A2 were made of α2 CVn covering the entire 5ḍ5 period of this magnetic Ap variable. The light curves ranging from 1250A to 3330 A indicate the dominant role of rare-earth line-blanketing in redistributing flux. In a broad depression of the continuum covering 2300 A to 2600 A scanner observations identify strong lines of Eu III as major contributors to this feature. At maximum intensity of the rare-earth lines the ultraviolet continuum shortward of 2900 A is greatly diminished while the longer wavelength regions into the visual become brighter. Thus, the light variations in α2 CVn are: due to the variable strong line-blanketing by the abundant rare-earth elements.