The main program of the Sydney 19.7-Mc/s cross is devoted to the observation of the brightness distribution in a strip, extending some 10 degrees on either side of the galactic equator, the particular interest being the detection of absorption in HII regions. This program is well advanced, but since observing time is limited by ionospheric absorption and interference, records are not yet available for a number of galactic crossings. The detailed analysis of the observations has just begun, and will take some time. In the meantime a rough analysis of the data has been made in order to get a general picture of what might be expected, and this paper gives an outline of the results. The application of various instrumental corrections and the detailed analysis of ionospheric effects may subsequently necessitate some changes in the numbers involved in the calculations, but it is expected that the general conclusions will still hold.

Two surveys of the galactic background radiation at low frequencies have recently been made at Cambridge. At 38 Mc/s, a survey of the sky between declinations −20° and +70° has been made by Blythe [1] using a new type of pencil-beam aerial system. This instrument, consisting of an array 1200 feet long in an east-west direction together with a small movable aerial, used the principle of aperture synthesis to provide a beam 2°.2 × 2°.3 at the zenith increasing to 2°.2 × 7°.4 at an angle of 70 degrees from the zenith.

In 1956 and 1958 Heeschen [1, 2] reported hydrogen-line radiation at 21 cm from the Coma cluster of galaxies. He found a maximum antenna temperature of 2 °K and gave a profile with a width of about 4 Mc/s. From his details, it seems clear that these results depend mainly on a large number of drift curves on a frequency of 1386.9 Mc/s, since the accuracy of other points of the profile was much lower. Because the intensity is low and near the limit of what can be observed with present-day receivers we decided to repeat some of the observations with the 21-cm equipment at Dwingeloo.

Results have been presented from continuum observations in the range from 19.7 to 1390 Mc/s (λ = 15 to 0.22 m).

At the present stage of radio astronomy it is convenient to divide the observed radio sources into two classes: Class I sources, which show a concentration into the galactic plane; and Class II sources, which are distributed more isotropically. In the present session we shall be concerned principally with the Class II sources.

Dr. Pawsey has asked me to make some general remarks on the problem of confusion when the traces of adjacent radio sources overlap. This is a question that affects all surveys in which the observations are limited by resolving power and not by sensitivity; there is perhaps some misunderstanding about the relative performance of interferometers and pencil-beam systems of similar sizes, and there seems to be an impression that the former are more seriously affected than the latter.

A fundamental limitation to the number of radio sources observable with a given aerial system is set by the finite solid angle of the aerial beam's reception. In any survey, whether it be a total-power survey or an interferometer survey, errors tend to occur whenever two sources are present simultaneously in the aerial beam. If the receiving system is a total-power system the contributions from the two sources will add, and either they will be interpreted as a single source of greater intensity than each of the individual sources, or the sources may be resolved but their intensities and positions may be subject to considerable errors.

Two years ago the Cambridge 4-aerial interferometer [1] was adapted to work at a frequency of 159 Mc/s, the resolving power thereby being increased by a factor of four over that at the previous frequency, 81.5 Mc/s. The overall beamwidth at 159 Mc/s is 1.2 degrees by 7 degrees but the beam contains interference fringes in two planes at right angles, so that, by phase-switching between the east pair and the west pair of aerials, sources with angular diameters greater than about 7 minutes of arc are eliminated. A survey using this technique has been carried out by Edge and Shakeshaft and may be called a “small diameter” survey. If the phase-switching receiver is connected between the north pair and the south pair of aerials it is possible to record sources with diameters up to about 1 degree and to measure diameters between 2 and 8 minutes of arc. Archer and Baldwin have used the aerial in this way to make a “large diameter” survey. These two surveys will be referred to collectively as the 3C survey.

Un des problèmes majeurs de l'étude des radiosources est leur identification avec des objets optiques. Cette identification est presque toujours rendue très difficile par l'imprécision des mesures de position des radiosources.

To discover the possibly large number of weak discrete radio sources and to determine their positions accurately is a very important and, at the same time, a difficult task. Radio telescopes with a large resolving power and a high-sensitivity receiver are necessary for this purpose.

Since their discovery in 1946, the discrete sources of radio-frequency radiation have provided some of the most difficult problems in radio astronomy. For their effective study, one requires a very highly sensitive radio telescope sufficiently precise for accurate position determination and with sufficient resolution to observe separately the detectable radio sources. These requirements led to the development of the Sydney 3.5-m Cross-type radio telescope, which is a pencil-beam instrument with a beamwidth of about 50 minutes of arc and a sensitivity such that under ideal conditions a source of intensity about 1/10,000 that of the strongest known source, Cassiopeia A, is just detectable.

Previous attempts to identify any large proportion of the discrete sources discovered at meter wavelengths have met with small success. In the investigation briefly reported here an extensive search has been made on the original plates of the 48-inch Palomar—National Geographic Society Sky Survey, using the available published radio data, but more especially the as yet unpublished results of a survey between +50 and −10 degrees declination that has been made with the interferometer of the Mullard Radio Astronomy Observatory, Cambridge. This radio survey (3C) has been carried out at 159.5 Mc/s using the aerial array of the 2C survey [1] in modified form. An account of the observation and reduction of this recent survey is given by other speakers in the Symposium. The area of sky covered by the 3C survey, and the criteria for the selection and classification of the sources, are likely to undergo small extensions and modifications before the final catalog is ready for publication, but these modifications will be of a minor character and will not alter the general conclusions of the present paper.

The identification of discrete radio sources with optically observed celestial objects continues to be one of the most difficult and challenging tasks for the optical astronomer. Baade and Minkowski [1, 2] have done admirable work on the subject for the northern hemisphere and for the sections of the southern hemisphere accessible from the latitudes of Mount Wilson and Palomar Observatories. The zone south of declination −40 degrees, however, has not been studied to any great extent. As a part of our collaborative program of research with the C.S.I.R.O. Radiophysics Laboratory in Sydney, several members of the Mount Stromlo Observatory staff have initiated studies for the identification of sources with declinations south of −40 degrees.

Optical search in the positions of radio sources has led to the certain identification of some eight sources with galactic nebulosities, and nineteen with extragalactic nebulae.

The advent of radio astronomy provides a new observational tool for arriving at tests of the different theoretical cosmologies. The tests may be either wholly new, or revivals in a more cogent form of earlier work in visual astronomy.

First, I should like to say something about the use of highly specialized models in cosmology. The Einstein—de Sitter model is a relativistic model in which the cosmical constant and the space-curvature constant are both equated to zero. Likewise, the pressure is assumed to be zero throughout the history of the universe, except perhaps at the initial instant. It is well-known that the first two constants can be determined from observation, if not at present, at any rate as the data are refined in the future. Hence, I think it is a weakness to prejudge the issue and assign a priori values. Nor is it self-evident to me that the pressure must always have been zero even if it is zero now. Composite models, with nonzero pressure at first, followed by a zero-pressure condition, need to be examined.

The lively and extended discussions in this session have given to some extent an unduly pessimistic picture of the present state of the investigation of faint radio sources. It seems necessary to emphasize the fact that progress has been made, even if relatively few results are definitive and few problems have been solved.

To discuss the theories of radio emission it is convenient to divide the sources into five different classes: (1) planets, (2) the sun, (3) galactic nebulosities, (4) galaxies taken as a whole, and (5) clusters in galaxies. About (1) and (2) I shall say little. It appears that the temperatures of Venus and Jupiter obtained from the thermal radio emission measures are in reasonable agreement with the temperatures determined by the older astronomical methods. No theories explaining the nonthermal emission from Jupiter have yet been developed, though there has been discussion of the possibility that some type of discharge mechanism leading to plasma oscillations is involved.

Strong radio emissions are associated with solar flares. It seems worth while, therefore, to examine whether radio emission associated with flares and eruptions in nonstable stars could be observed. The first problem in this respect is establishing the threshold of sensitivity necessary to receive anything during an eruption or a flare; the second one is naturally fixing the duration of the radio emission, which can be much shorter in a given wavelength than the optical emission.

The swept-frequency receiver is one of the most powerful tools for studying the processes that go on in the sun's atmosphere. However, the instrumentation becomes fairly elaborate if a large frequency range is to be covered by such an apparatus. Consequently, only a few such receivers have been built, and those that are in operation are somewhat limited in their frequency coverage. It then seems worth while to examine the data from existing fixed-frequency receiving stations in order to determine the extent to which they can supplement or approximate the results of the swept-frequency apparatus. Many solar observations at fixed frequencies have been made and are still being made by numerous observers. These data contain much meaningful information if they can be interpreted successfully.