Continuing the systematic study of novae, carriedout at Asiago since 1958, two recent galactic novae, V400 Per (1974) and V373 Sct (1975) have been observed, after the announcement of the discovery, during their decline towards minimum, with the spectrographs applied to the telescopes of 122 cm and 182 cm of the Asiago astrophysical Observatory, equipped with image-tubes.

This communication gives some preliminary results of a spectroscopic survey of old novae presently carried out at Asiago Observatory. I have obtained image-tube spectra of 10 old novae in the spectral range 4000-7000 Å with dispersions from 60 to 180 Å/mm. The observed novae are: T Aur 1891, V 476 Cyg 1920, Q Cyg 1876, HR Del 1967, DN Gem 1912, DQ Her 1934, V 533 Her 1963, CP Lac 1936, V 841 Oph 1848 and GK Per 1901.

This study aims to present our photometric observations of Nova Cygni V1500 1975, and to discuss the period changes of brightness variation on the basis of all available data.

The Nova was observed from 2 September 1975 to 26 October 1976 using the single channel UBV photometer attached to the 60-cm reflector of the Ostrowik station of Warsaw University Observatory. The brightness of the Nova decreased considerably during that time and it was necessary to use a sequence of four comparison stars. The variable was usually monitored in the V filter and occasionally measured with B and U filters. The V band is not well suited for observing novae in the nebular stage, since the dominant light contribution comes from O III emission lines.

SIT Vidicon spectra (3500 - 5100 Å) with time resolutions of several minutes and spectral resolutions of 6 Å have been obtained of a sample of old novae at minimum light. For comparison, spectra of SCO X-1 (V818 Sco) are also presented. Velocity-resolved image-tube spectra of RR Pic 1925 indicate Doppler variations in the He II λ4686 emission line with a period ~ 3 hours, thus confirming the binary nature of RR Pic.

The existence of dusty envelopes around novae is shown by various authors using different methods:

We have used the Balmer decrement of novae. In the case of all novae a sudden variation of this decrement is observed a certain time after the beginning of the outburst, the Hα/Hβ ratio rising from its normal value to a maximum and then decreasing again to its normal value.

A coherent sinusoidal oscillation of approximately 88 s and amplitude has been detected at the end of the long eruption of VW Hyi in December 1975. The phenomenon was observed during five nights, when the period varied between 90 and 87 s. Some runs indicate that the amplitude depends on the phase of the orbital revolution. This result favours models in which oscillations are caused by rotation of inhomogeneities in the inner disc.

Photoelectric observations of the dwarf nova VW Hyi, obtained at the end of the December 1975 supermaximum, are presented. After decline from the outburst, the superhump period (0ḍ07622) combines with the orbital period (0ḍ07427) to a beat phenomenon: the O-C’s and the light curves of the orbital hump vary systematically with the phase of the beat period for at least one week after recovery from the supermaximum. It is suggested that the red secondary component, which rotates non-synchroneously with the superhump period, expands slightly at the beginning of a supermaximum and is heated up asymmetrically, probably due to instabilities in its convection zone. In addition, the increased mass transfer rate may trigger the long eruption in the accretion disc while short eruptions originate in the disc without participation of the secondary.

We present photometric and spectroscopic observations (intermediate dispersion in the wavelength range 3700-10900 Å) of the recently discovered variable HM Sagittae (Dokuchaeva, 1976). Also pre-outburst data, indicating a red star, are used to discuss its nature and distance.

A new class of close binary stars, proposed to be called polars, has been discovered during recent year; its prototype is AM Her. The three known polars, AM Her, AN UMa, and W Pup, are polarization, spectrum, radial velocity, optical, and infrared variables of short periods of 3.1, 1.9, and 1.7 hours, respectively (Tapia 1977 a, b, c; Michalsky et al. 1977; Krzemiński and Serkowski 1977; Stockman et al. 1978; Herbig 1960; Cowley and Crampton 1977; Priedhorsky 1977; Greenstein et al. 1977; Berg and Duthie 1977; Szkody and Brownlee 1977; Olson 1977; Mumford 1977; Priedhorsky and Krzemiński 1978; Neugebauer et al. 1977). AM Her itself has been found to have strong soft X-ray emission in the 0.1 - 0.4 keV band (Hearn and Richardson 1977) and a hard X-ray spectrum (about half the observed 2-60 keV energy lies above 28 keV, Swank et al. 1977) both modulated with a 3.1 hour period. No X-ray emissions from the other systems have been detected so far.

The different methods by which element abundances in novae have been determined are reviewed. Curve of growth studies of novae at maximum light have indicated CNO nuclei to be greatly enhanced with respect to hydrogen in certain objects. These results are questionable because they depend upon an assumed temperature distribution in the photosphere which is probably too steep to be realistic. Emission line analyses of novae, generally obtained in the period of early decline, also indicate possible heavy element enhancement, however these results are tentative because of uncertainties in the parameters of the emitting gas. It is suggested that useful abundance determinations of nova ejecta might be obtained from studies of old, extended nova shells.

The chemical composition of Nova Cygni 1975 was determined by a curve-of-growth analysis. A strong excess of helium and the CNO -group abundances and a normal Si abundance was found.

Spectral observations of N Cyg 1975 were begun at the Crimean Astrophysical Observatory on 29 August 1975, at 18h 50m UT, one day before light maximum (Boyarchuk et al. 1977). Twelve spectrograms with a dispersion of 4,6 and 12 Å/mm were taken during this night. The spectral type of the Nova was B2 Ia+ and no emissions appeared in the spectrum. During the following night, 10 spectrograms were obtained. The spectral type of the Nova was A2 Ia+ qe, and strong emission lines with P Cygni-type profiles could be detected. Since absorption lines were not observed on 30 August, only the spectrograms taken on 29 August can be used for a curve-of-growth analysis.

This review of the radio emission properties of close binary systems and novae will be partly concerned with surveying the star systems that exhibit continuum radio emission, and partly concerned with discussing the implications of the observed radio emission. The phenomena we will encounter will range from purely thermal continuum radio emission for the case of ionized nova shells to strong, non-thermal continuum radio emission produced by relativistic particles in both ‘normal’ and X-ray emitting close binary systems. In keeping with the subject of this symposium, let me emphasize that all regions involving radio emission are from the portions of the stellar environment where Ne ≤ 1010cm-3, generally rather high in the stellar or systemic atmosphere; however, events more closely associated with the stars themselves are the root causes of the phenomena, supplying energy, mass, magnetic fields, and relativistic particles.

The optical position of the short period eclipsing binary XY UMa, which shows large scale star spot activities, coincides within 2 minutes of arc with a radio source, listed in the 4C-, 4CP-, OK-catalogues, and the position obtained with the Effelsberg radio telescope. A fainter star, centered on the radio source position, is according its U-, B-, V-values a K main sequence star. The radio source is non-thermal, and has a spectral index α ~-0.85. If the radio source is identical with the binary, it would imply important consequences about the extragalactic nature of high latitude 4C-sources, the radio background and the high radio luminosity of the binary.

Very detailed spectroscopic and photometric observations of Novae are available today. Unfortunately they do not directly tell us what suddenly makes a star a million times brighter, sometimes even transforming the familiar constellations on the sky. Twenty-five years after its publication, Mestel’s (1952) metaphor of a gigantic hydrogen bomb seems to be most widely accepted, although he at that time applied his model to a Supernova outburst. Later Giannone and Weigert (1967) as well as Rose (1968), Saslaw (1968) and others have computed hydrostatic models of such an event, using Kraft’s (1963) binary model of a late type main sequence star and a white dwarf. Detailed hydrodynamic computations were carried out by Starrfield and his collaborators (for references see Sparks, Starrfield, and Truran, 1977) and recently by Prialnik, Shara, and Shaviv (A & A 62, 339, 1978).

A periodic modulation in the optical flux from Nova Cygni 1975 has been observed since shortly after outburst (Tempesti 1975) and the period is known to have varied by at least a few percent (Semeniuk et al. 1976). We account for the modulation in terms of a simple geometrical model in which the wind emanating from the nova is shadowed by its binary companion (see Fabian and Pringle 1977 for a fuller account). This produces an azimuthal variation in the radius of the surface of last scattering, RS, which an observer sees as a periodic modulation of the. continuum with period roughly equal to the orbital period. Variation of the observed period is accounted for in terms of variation of the size of RS.

A hot spot at the outer rim of the accretion disk dominates the light of U Geminorum at minimum light. We take this as evidence that there is no accretion from the disk onto the white dwarf between the eruptions, and we assume there is no viscosity in the disk at that time. The hot spot is produced by dissipation of the radial component of velocity of stream falling from the inner Lagrangian point. Angular momentum per unit massis smaller in the stream than it is in the outer parts of the disk. This leads to angular momentum redistribution in the outer part of the disk. The difference of tangential velocity between the stream and the disk is dissipated in few orbital periods. These processes make the outer parts of the disk look like a torus. We calculated the structure of the torus in U Geminorum between the eruptions and we obtained the following oarameters: mass of the torus: 10-9 - 10-8 M⊙ (assumed), its optical thickness: 106, the large radius (i.e. the radius of the disk): 0.5 R⊙ (assumed), the small radius (i.e. the half thickness of the outer parts of the disk): 0.05 R⊙. Conditions at the surface of the torus are similar as on the solar surface.

It is shown that according to the model of continued ejection of nova envelopes by radiation pressure, discrete clouds are probably formed. Their properties are discussed.

We present results of thermal and vibrational stability analysis for 1 M⊙ white dwarf models corresponding to various accretion rates ≳ 10−11 M⊙/y. Accretion is assumed to be spherically symmetric and stationary. Thermal instability due to nuclear burning of hydrogen (at lower accretion rates) and helium (at higher rates) was found. At medium rates two growing thermal modes are simultaneously present. Vibrational instability was found for all models except those corresponding to highest accretion rates. The excitation rates for some nonradial g-modes are at least 3 orders of magnitude higher than those for radial pulsations. These rates are also higher than the excitation rates for thermal modes in certain range of accretion rates corresponding to high luminosities and effective temperatures. Among objects in which these instabilities may be important are symbiotic stars and nuclei of planetary nebulae.

We consider the usual hypothesis that the short period coherent oscillations seen in cataclysmic variables are attributable to g-modes in a slowly rotating star, for details see Papaloizou and Pringle (1977). We show that this hypothesis is untenable for three main reasons: (i) the observed periods are too short for reasonable white dwarf models, (ii) the observed variability of the oscillations is too rapid and (iii) the expected rotation of the white dwarf, due to accretion, invalidates the slow rotation assumption on which standard g-mode theory is based. We investigate the low frequency spectrum of a rotating pulsating star, taking the effects of rotation fully into account. In this case there are two sets of low frequency modes, the g-modes, and modes similar to Rossby waves in the Earth’s atmosphere and oceans, which we designate r-modes. Typical periods for such modes are 1/m times the rotation period of the white dwarfs outer layers (m is the azimuthal wave number). We conclude that non-radial oscillations of white dwarfs can account for the properties of the oscillations seen in dwarf novae.