Galactic X-ray sources studied by UHURU can be classified into several categories, including supernova remnants, transient or nova-like sources, Sco X-1 like sources, and pulsating sources. The latter are discussed in some detail with particular emphasis upon the characteristics of the periodic binary X-ray sources, Cen X-3 and Her X-1. It is suggested that all galactic X-ray sources (except S/N remnants) may be binary systems.

The Seventh Orbiting Solar Observatory, OSO-7, carries on board a 1–60 keV X-ray detector for the purpose of measuring the positions, spectra and time variations of X-ray sources. The instrument is described herein, and several topics under investigation and results in publication are mentioned.

The observational properties of Sco X-1, Cyg X-2, and Cen X-3 are reviewed in connection with the hypothesis that X-ray power is derived from gravitational energy released when matter is accreted onto the surface of one component in a mass transfer binary star. Evolutionary mechanisms for producing suitable types of binaries are considered. The following boundary conditions on possible evolutionary models are also treated briefly: (1) a quite significant fraction of hard X-ray sources are associated with the nuclear bulge of the galaxy; (2) mass-transfer binaries such as U Gem stars are not hard X-ray sources; (3) counts of binary stars lead to a considerably larger number of X-ray source candidates than are actually observed.

A review of the observational status of X-ray sources detected in the 20 ⋍ 500 keV range is presented. Of the approximately 115 sources listed in the March 1972 edition of the UHURU 2–6 keV sky survey catalog, about 15 sources have been studied in hard X-rays. Most of the data have been obtained from balloons, although the OSO-3, and more recently the OSO-7, have contributed. With the exception of CEN A, the SMC, and possibly M-87, all the sources detected at higher energies are galactic and heavily concentrated in the galactic plane. The Crab Nebula has been measured to about 500 keV in continuous emission and a component at the ⋍ 33 ms pulsar period comprising about 20% of the total emission has been detected to ∼10 MeV. Objects such as SCO-1 and CYG-2 are characterized by an exponential spectrum, which varies over a 10 min. time scale about a factor of two, and a flatter spectrum extending to above 40 keV which exhibits independent variability. Objects such as CYG-1 and possibly CYG-3 have a multi-component power law spectrum extending to over 100 keV, and may vary many factors over a period of weeks. Other sources generally not yet identified with optical or radio candidates, located in the Galactic Center and the Centaurus/Crux region also show considerable variability, and in one case may have been detected to nearly 500 keV. Only upper limits at about 2 × 10−4 photon (cm2 s keV)−1 in the 20–50 keV range exist for most supernova remnants and extragalactic sources.

Simultaneous observations of the X-ray source Sco X-1 are reviewed and discussed. Several conclusions can be drawn from the observations. (a) There is no correlation of the radio intensities with either the optical or X-ray intensities. (b) The X-ray intensity is high and variable only when the blue optical magnitude is brighter than 12.6–12.7. (c) There is often correlation of X-ray and optical flares, but the relationship is not one to one.

Attention is called to the paucity of simultaneous observations of other identified X-ray sources.

Observational work on radio emission from galactic X-ray sources is reviewed. Some problems associated with the detection of these weak radio sources are mentioned and our present knowledge of radio identifications is summarized. The radio properties of the most studied objects are discussed in detail.

Some of the implications of the data on radio counterparts of X-ray sources are discussed. Data on the radio counterpart of Cyg X-3 is compared with the data on the radio star β Persei (Algol), to show that similar radio emission mechanisms are involved. The data for such radio stars not yet known as X-ray sources imply that such objects should be transient X-ray sources.

Available X-ray observations of supernova remnants are reviewed. The number of SNR seen above 2 keV remains small after inclusion of the UHURU results and for only the Crab Nebula is the data adequate to clearly indicate the radiation mechanism. The increasing importance of low energy X-ray studies (below 1 keV) of older and relatively nearby remnants is noted. Brief discussion is given of the relation of the X-ray data to current ideas of the evolution of SNR.

Stellar X-ray spectroscopy and polarimetry are discussed in terms of the source parameters that can be determined through such studies and in terms of the constraints that these studies will place on theoretical models. The spectroscopic and polarimetric results that have been obtained to date are reviewed. These include the recent discovery of X-ray polarization in the Crab Nebula and the recent evidence for X-ray coronal line emission in the Cygnus Loop. Finally, the properties and predicted performance of a number of satellite-borne spectrometers and polarimeters are presented.

A statistical analysis of the UHURU catalogue of X-ray sources leads to the following conclusions. If the weak sources are omitted, there is a strong concentration to low galactic latitude but the absence of a strong background and the presence of some strong sources at low galactic longitude indicate an appreciable number of sources of luminosity L ∼ 104L⊙ in the vicinity of the ‘nuclear bulge’. This region generally suggests ‘stellar population II’ and therefore stars of small mass. However, there is some suggestion of a second class of sources, distributed in the galactic plane like ‘stellar population I’ and suggesting large stellar masses.

There are three possible types of simple X-ray spectra, (a) optically thin bremsstrahlung, (b) black-body spectrum, and (c) power-law spectra. In this talk only theoretical models for type (a) are reviewed, including accretion, rotation and vibration for a white dwarf star and cocoons around a neutron star.

Cen X-3 is probably a neutron star, releasing the infall energy of accreted matter. Sufficient material for accretion will be provided by a conventional stellar wind from its more massive companion star. That star is not likely to rotate synchronously; therefore a ‘Roche lobe’ analysis of the eclipses is not valid. A ‘tidal lobe’ analysis allows the neutron star to have a mass of the order of one solar mass. Overflow of the ‘Roche lobe’ is neither necessary as a source of mass nor probable in view of the observed stellar line widths of the two identified X-ray companions.

The mass flow onto the condensed star is very small in all cases. It is limited, for an object of m solar masses by the Eddington Limiting Luminosity to Ṁac < 10–7.4m(M⊙ yr−1), which limit applies even if the accreting object contacts or traverses its companion star.

The observed 4.84 s rotation period of the Cen X-3 neutron star is very simply explained as the critical value where a centrifugal barrier regulates the rate of infall to the surface. The X-ray spectrum is understood as blackbody radiation coming from a well-defined area near each magnetic pole of the neutron star.

The outward transfer of angular momentum of accreting matter can lead to the formation of a disk around the black hole. The structure and radiation spectrum of the disk depends, in the main, on the rate of matter inflow Ṁ into the disk at its external boundary. Dependence on the efficiency of mechanisms of angular momentum transport (connected with the magnetic field and turbulence) is weaker. If Ṁ = 10−9–3 × 10−8 M⊙/yr, the disk around the black hole is a powerful source of X-radiation with hv ∼ 1–10 keV and luminosity L ∼ 1037–1038 erg s−1. If the flux of the accreting matter decreases, the effective temperature of radiation and the luminosity will drop. At the same time when Ṁ > 10−9M⊙ yr−1, the optical luminosity of the disk exceeds the solar one. The main contribution to the optical luminosity of the black hole is due to the re-radiation of that part of the X-ray and ultraviolet energy which is initially produced in the central high temperature regions of the disk and which is then absorbed by the low temperature outer regions. The optical radiation spectrum of such objects must be saturated by the broad emission recombination and resonance lines. Variability is connected with the character of the motion of the black hole and the gas flow in binary systems and possibly with eclipses. For well defined conditions, the hard radiation can evaporate the gas. This can counteract the matter inflow into the disk and lead to autoregulation of the accretion.

If M ≫ 3 × 10−8 (M/M⊙) M⊙ yr−1, the luminosity of the disk around the black hole is stabilized at the critical level of Lcr = 1038 (M/M⊙) erg s−1. A small fraction of the accreting matter falls under the gravitational radius whereas the major part of it flows out with high velocities from the central regions of the disk. The outflowing matter is opaque to the disk radiation and completely transforms its spectrum. As a consequence, a black hole in a supercritical regime of accretion can appear as a bright star with a strong outflow of matter.

We summarize some theoretical aspects of the X-ray emission from genuine pulsars and their relation to the more general problems of X-ray sources.

Data from the UHURU satellite have provided a list of more than forty high latitude sources (|b| > 20°). X-rays have been detected from among the nearest normal galaxies, giant radio galaxies, Seyferts, QSOs and clusters of galaxies. The cluster sources appear to be extended by several hundred kiloparsecs as well as being very luminous. These cluster sources have systematic differences in their X-ray spectra from individual galaxies.

About twenty sources are not reliably identified so far. A few of these are located near undistinguished 3C or MSH radio sources. The rest are either located near distant clusters or undistinguished bright galaxies, or are too far south, so that we have not sufficient optical data to allow a thorough search for possible association with clusters or unusual individual galaxies.

The luminosity function for weak, high latitude X-ray sources is determined, and the contribution of sources just below the UHURU threshold of detectability to observed fluctuations in the diffuse X-ray background is evaluated. The total contribution of all observed types of extragalactic sources to the X-ray background is estimated.

We describe the optical properties of the radio galaxy NGC 5128, the Seyfert galaxy NGC 4151 and the QSO 3C 273 all of which appear to be point sources of X-rays. We emphasize how the X-ray observations, particularly the low energy absorption cutoff, may help us to understand the detailed structure and source of energy in these diverse objects.

The clusters of galaxies in Virgo, Perseus, Coma and Centaurus, all associated with extended X-ray sources are described. They have diverse shapes, central concentrations and galactic populations, but all contain a radio galaxy and, in several cases, a low frequency radio halo around it. It is concluded that the X-ray emission is likely to be non-thermal in origin.

A review is given of the physical properties of extragalactic X-ray sources. The generation of X-rays as thermal bremsstrahlung, by the synchrotron mechanism or by Compton Scattering, is discussed. It is shown that each may be important depending on the circumstances. Only more detailed observations will enable us to decide which process dominates in any given source.