A comprehensive analysis of previously published observations of Algol has been submitted to Astronomy&Astrophysics under the title “Geometry and Dynamics of the Algol System”. The main results can be summarized as follows:

The masses of the three components are mA = 3.6±0.1, mB=0.79±0.01, mC=1.6±0.1 solar masses.

The orbit of Algol C is co-planar with the AB-orbit within some 3°, but polarimetry of the eclipsing pair indicates that the senses of revolution are opposite.

The close orbit is circular, thus the 32 year period in the times of minima can not be explained by apsidal motion.

The Na D1 and D2 lines of Algol A and B observed by Tomkin and Lambert(1978) were analyzed to obtain sodium abundance. A mild deficiency of sodium [Na/H]∼-0.5 is found in Algol B, while a slight overabundance [Na/H]∼+0.3 is found in Algol A.

1.8 A° segment of the Algol A spectrum centered at 1075.55 A° line of Ni II was scanned repeatedely with the 0.05 A° resolution from the Copernicus satellite. The numerical model is described in which 37 out of the total 55 scans obtained during two eclipses were combined in a least-squares solution, to determine the period of rotation and degree of non-solid rotation of Algol A. The period of rotation suggests full synchronism of rotational and orbital motions; the equatorial velocity is Ve = 53 ± 3 km/s. The non-solid rotation parameter ‘s’ measuring the distribution of angular velocity versus stellar latitude in ω = ωe (1 - s + s cos2 ϑ) has been found equal s = 0.07 ± 0.25 indicating the rotation law not far from solid body.

We have investigated variations in the strength and profile of the Hα emission line in the spectrum of Algol using 145 16 A° mm−1 spectrograms obtained during the period September, 1976 to December, 1977. The Hα emission line profile has been extracted by subtracting the absorption contributions of the three stars in the system. The resulting data set has been analyzed to look for variations related to orbital phase as well as shorter and longer term variations.

The radio behaviour of Algol at centimeter wavelengths has been reviewed by Hjellming (1976, 1977), and by Gibson (1976). The observed radio emission can be roughly divided into two types: a quiescent type and a flare-like or outburst type. The quiescent emission is characterized by a low flux density, usually between 20 − 50 mJy, and a nearly flat spectral index. The latter means that the value of the flux density is nearly independent of wavelength. The variations in flux density, when present, are gradual and have a time scale on the order of days. The flare-like behaviour shows rapid and large changes in flux density. The amplitude of an outburst is wavelength dependent, the shorter wavelengths having larger amplitudes. Woodsworth and Hughes (1976, hereafter WH) attributed the two types of behaviour to two physically distinct sources. The quiescent type of emission is produced by a very large, thermally emitting cloud which surrounds the eclipsing system. The flare-like variations are due to a synchrotron source that is associated with mass loss. Their model is inconsistent with optical and X-ray information concerning the Algol system. However, a model with two radio components can be modified to agree with other types of data.

Time-resolved spectroscopy of the gaseous ring of RW Tau has revealed characteristics of its structure. Observations of H-alpha emission during one eclipse show a partial eclipse of the ring which is brighter at second contact than at third contact. The radial velocities became nearly constant after mideclipse. During two other eclipses the emission continued to strengthen after second contact. X Per has shown variability on a time scale of minutes. On one occasion H-alpha emission decreased in strength by a factor of two in 90 seconds.

The eclipsing binary U Cephei has proven to be of great interest in the study of stellar evolution in close binary systems. Batten (1974), Hall and Walter (1974), Rhombs and Fix (1976), Markworth (1977), and Olson (1978), among others, have recently reported on their intensive ground based studies of U Cephei. Kondo, McCluskey and Wu (1978) have investigated the ultraviolet light curves of U Cephei obtained with Astronomical Netherlands Satellite (ANS). Kondo, McCluskey and Stencel (1979) have discussed the International Ultraviolet Explorer (IUE) spectra of U Cephei. This paper discusses results incorporating additional IUE high resolution spectra of U Cephei obtained in both far-ultraviolet and mid-ultraviolet spectral regions.

Conspicuous, variable photometric contamination was observed during primary eclipses of the totally eclipsing Algol-like system U Cep from late-1974 to late-1977. Excess light was particularly prominent between second and third contacts, and originated mainly in optically thick accretion disks surrounding the B-type hot component. Observations of such disks were obtained with uvby (I) photometry during 10 eclipses, using the 1.0 M and 0.4 M reflectors at Prairie and Kitt Peak National Observatories. Figure 1 shows ultraviolet observations between internal contacts for active (October, 1975) and quiescent eclipses. The light excess over quiescent levels was converted to flux distributions Lλc relative to the quiescent visual flux of. the G-star.

Polarimetric and photometric observations of U Cephei during the 1974-75 outburst are discussed. An optically thin envelope or disk cannot explain the observations. The polarization too small by a factor of 10, when compared with the amount of additional light during the totality of the eclipse of the B-star. This suggests that most of the additional light came from optically thick matter around the B-star, i.e. the effective radius of the photosphere of the B-star, rB, was increased by 15-20 % during the period of high mass transfer. The polarization observations could then be explained by a simple model containing an optically thin spherical envelope of outer radius 1.1 rB and an equatorial disk of outer radius 1.7 rB surrounding the expanded B-star. The mean electron density is about 1.4 × 1012 e/cm3 in the spherical envelope and 1.6 × 1012 e/cm3 in the preceding and 0.9 × 1012 e/cm3 in the trailing side of the disk, the thickness of which is assumed to be 0.1 rB normal to the orbital plane. The variations in the position angle can be explained if the inclination of the orbit is about 83°. The details will be published elsewhere.

It has long been thought that β Lyrae is a unique system, by virtue of its UV spectrum and its nature. I will argue that a whole class of interacting long-period binaries exists, similar to β Lyrae. According to IUE observations made in 1978-79 by myself and R. H. Koch, this group comprises: RX Cas, SX Cas, V 367 Cyg, W Cru, β Lyr, and W Ser. AR Pav is a transition case linking them with the symbiotics. I also suggest that HD 218393 (KX And), HD 72754, and HD 51480 are their non-eclipsing counterparts. I will call the whole group the W Serpentis stars.

These systems are mass-transfering binaries (case B) in which the mass transfer rate is relatively high, probably on the order 10−6 to 10−4 M⊙/year. They display an ultraviolet continuum with a color temperature definitely higher than the one observed in the optical region. Even more characteristical is the presence of strong emission lines of N V, C IV, Si IV, Fe III, Al III, and lower ions of C and Si. I will attempt to discuss these phenomena on the assumption that they are due to accretion onto non-degenerate stars.

We present observational evidence relevant to the existence of binaries of the morphological type proposed by Wilson (1979), called double-contact binaries. Consider normal evolution with Roche lobe overflow, in the rapid phase of mass exchange. It is well known that such mass-exchange spins-up the accreting star (most of whose mass is transferred material by the end of the rapid phase). Many examples are known (Table 1, for example) of Algol-types with fast-spinning primaries (original secondaries). Suppose the accreting star has been spun-up so much that it cannot accept further high-angular momentum material. It will then be filling its limiting rotational lobe (but not its Roche lobe). The secondary (original primary) will still be filling its Roche lobe. The overall result is that both stars fill their limiting lobes accurately, but are not in contact with one another. Since it is common to speak of “the contact component” of a semi-detached system - meaning the one in contact with its limiting lobe - it seems natural to call members of the new morphological type “double-contact” systems. Of course, there must be a repository for the mass lost by the Roche lobe filling and rotational lobe filling components. This would naturally be the circumstellar disks around the primary components of such systems as β Lyr and V356 Sgr (Wilson and Caldwell, 1978). The end of the double-contact phase would occur when the rate of transfer of rotational angular momentum back into the orbit, by tidal braking, begins to exceed the rate of transfer of angular momentum the opposite way by mass exchange.

RX Cas appers to be a long-period Algol-like binary with very fast mass exchange (Struve 1944). In the period 1975-1977, this star was observed photometrically by Arsenijevič, Grygar, Harmanec, Horn, Koubský, Kříž and Pavlovski at Hvar Observatory (Yugoslavia) and by Zverko at Skalnaté Pleso Observatory (Czechoslovakia). The resulting UBV light curves seem to be rather peculiar, for the following reasons:

1) Combining our observations with the older ones, we obtain the following formula for the times of minima:

The period increase is surprisingly high.

2) The scatter of observed points is considerable (about 0.2 mag.) and cannot be caused by observational errors. It corresponds to short period light variations on the time scale of several days.

3) After J.D. 24422660, the brightness of the whole system decreased in all colours. This decrease is about 0.2 mag. in V colour and 0.5 mag. in U colour. Only after J.D. 2442900 the brightness increased again. Such a behaviour confirms the reality of old observations by Gaposchkin (1944) and Martynov (1950) who found a long-term light variation with a period of approximately 500 days.

4) The part of the light curve around the phase 0.75 is definitely much lower than the corresponding part around the phase 0.25. This effect is most pronounced in the ultraviolet; the depression around the phase 0.75 is much deeper than the secondary minimum. This behaviour is opposite to the behaviour of dwarf novae. Instead of a bright shoulder connected with a bright spot, we observe some darkening.

The common peculiarities of these two systems are: a) the companion is a massive object (probably m2≥10) whose spectrum is not observable; b) both systems show evidence, though in different degrees, of mass-transfer and mass-loss; c) both present, in different degrees, hydrogen deficiency; d) ultraviolet observations have shown, in both cases, the presence of lines of highly ionized elements like N V, C IV, Si IV, probably formed in an extended envelope because they do not show orbital radial velocity shifts, and cannot be explained by the effective temperature of the star whose spectrum we observe. The latter property seems to be common to several close binaries, as shown by the ultraviolet observations with IUE by Plavec and Koch (1979); e) both systems present infrared excess, suggesting the presence of an extended envelope (Gehrz et al. 1974; Lee and Nariai, 1967; Humphreys and Ney, 1974; Treffers et al. 1976).

An observational study of the Hα feature in Algol type binary systems has revealed the presence of emission in nearly all systems with periods greater than 6 days. The disks which produce this emission are generally thin and can extend out to 90% of the Roche surface of the primary star. If emission is found in the systems of shorter period, it tends to be transient, lasting only a few orbital cycles. In several systems, there is evidence of mass loss from the vicinity of the primary at phase 0.5.

In this paper, we present far ultraviolet observations of three interacting binary stars (HR 2142, HR 7084, and λ Tauri). All stars exhibit evidence of a gas stream and extensive mass outflow from the system. Physical parameters are derived for the gas stream in HR 2142.

The application of narrow band photometry to Algol type systems is considered and advantageous circumstances pointed out. Observable effects can be divided into two kinds: classical (associated essentially with differential centre to limb variation) and interactive (where an emission component may be present in the line). Observations of U Sge and β Per are discussed from this basis and deductions made about the size and shape of an emitting region. Conclusions are compared with theoretical ideas about such regions and their expected stability, both for these and other similar systems.

Seven years ago, at the Parksville Symposium, I reported on first results obtained from light curves of typical Algol systems which were discussed according to principles taking into account the presence of gas streams. According to this modified method the reflection parameter is derived from parts of the light curve outside eclipses which are least influenced by gas stream effects. Light curves rectified in this manner, or expressed more exactly, roughly released from reflection and ellipticity effects, only in parts display a horizontal course, in the surplus in intensity after primary eclipse and by part also before it. This indicates an additional light source visible during about half the orbital period. On the preceding side of the primary eclipse a loss in intensity is frequently also observed, which is caused by absorption effects of the gas stream, and often overlaps the surplus intensity.

X-ray emission has now been detected from a large number and variety of binary systems. A brief survey of the properties of these binary systems is given, primarily from an optical standpoint.

Recently, masses of component stars have been determined for many X-ray close binaries (XCBS). For relatively well determined sources the masses of X-ray components are plotted against the masses of their optical components in Figure 1 and their orbital periods are summarized in Table 1 (Cowley 1977; Bradt, Doxsey, and Jernigan 1979; Wheeler 1978; and references quoted therein). Cowley (1977) prepared a table and noticed that there are two distinct types of XCBS. The Type I XCBS consists of an X-ray star and an early type star more massive than about 12 M⊙. On the contrary, the Type II XCBS consists of an X-ray star and a star less massive than about 2 M⊙. As seen in Figure 1, there is not any distinct intermediate type for which the mass of the optical component lies in the range of about 2-12 M⊙. The aim of the present paper lies in interpreting the origin of these types of XCBS on the bases of the conditions for the formation of a neutron star and of mass exchange in close binary systems.

A simple analytical method has been used to investigate thermonuclear flashes in a degenerate envelope of an accreting neutron star. The heating source is the compression of envelope and the cooling is due to the radiative cooling of the envelope. The role of hydrogen burning is discussed. A possible evolutionary scenario for formation of preburster binaries consisting of a neutron star and a low mass main sequence star is proposed.