For a discussion of the large-scale structure of spiral nebulae, we should like, of course, to choose a stellar system which we are certain is similar to our own. Since we know that our own Galaxy is a spiral galaxy, this narrows our choice to the spiral galaxies and, of those, we can immediately eliminate the so-called barred spirals, as we know that our galaxy is not a barred spiral. Hence, we need only consider spirals of types Sa, Sb, and Sc on Hubble's system. NGC 4594 is an Sa system with a large central spheroidal system typical of these early-type spirals. Undoubtedly, our Galaxy does not have such a large spheroidal system since it would be obvious as a large bulge which simply has not been observed. Next, M 81 is typical of the Sb spirals in which the central system has shrunk considerably. Finally, M 33 is typical of the Sc spirals in which the central system has shrunk until it actually approaches a semi-stellar point.

Within each cluster of galaxies, the ratio of the number of multiple galaxies to the number of single galaxies is much higher than the ratio calculated under the assumption of statistical equilibrium. This fact leads, without any supplementary assumption, to the conclusion that the components of any given multiple galaxy have a common origin. In this respect the multiple galaxies are similar to the multiple stars of our stellar system.

A progress report on work on (a) spectral types and (b) radial velocities of twenty-eight members of the Large Cloud (mostly mpg = 11 or brighter) will appear shortly in The Observatory. The commonest spectral types found lie between B 5 and A 2. A star with Mpg= − 10·1 has slightly narrower H-lines than the galactic A 2 super-giant HD 92207. [Fe 11] emission appears in four of the stars, including recent spectra of S Dor.

If ω is the angular velocity of circular motions in a stellar system depending on the distance R from the centre, and κ is the frequency of oscillation in the radius vector for an orbit in the equatorial plane which differs slightly from a circular orbit, we may show that in a coordinate system of angular speed the orbit in question is a closed oval, in the first approximation an ellipse, with two opposite vertices and with its centre in the centre of the system. The conditions are illustrated in Fig. 1, when AA is the apsidal line of an orbit. The expression for κ is where A is Oort's constant of differential rotation. If the distribution of matter in the stellar system may be approximated by a sum of concentric spheroids with one and the same plane of symmetry, the density increasing with decreasing dimensions of the spheroids, we should expect κ = 2ω close to the centre, whereas in the outermost regions we must expect that κ approaches to ω. It is therefore likely that ω – ½κ does not change appreciably within the system. Fig. 2 gives the run of κ, ω, and ω – ½κ for the galactic system according to the values of ω given by Kwee, Muller, and Westerhout [2]. In this case ω – ½κ changes very slowly with R. A relative orbit has a quite important property, if ω – ½κ is constant in the interval of R covered by the orbit. In this case, if the centre of gravity of a cloud of free particles follows the orbit, and if the velocity dispersion is small, the particles will pursue orbits which differ only slightly from each other and from the central orbit, but on the other hand there will in general be considerable differential motions along the central orbit. This orbit is therefore the curve in space along which a cloud of free particles tends to disperse. The process is similar to the dispersion of the meteoric particles of a comet along its orbit.

It is my privilege to report here about the investigations of the distribution of neutral hydrogen in the Galaxy. These observations, conducted by the Netherlands Foundation for Radio Astronomy, were made at Kootwijk and reduced at the Leiden Observatory. More than a dozen persons have participated in this work during the past year. Among them are Prof. Oort and myself, but I should mention three in particular: C. A. Muller has perfected the instruments even further and has supervised the observations that were made during 24 hr a day. G. Westerhout has supervised most of the reductions at Leiden and has prepared the map and model of the outer parts of the Galaxy. M. Schmidt has unravelled the situation in the inner parts of the Galaxy, on the basis of his observations made for this purpose.

Six years ago, in a joint paper with W. Baade, I presented certain spectrographic data bearing on the distribution and motions of gaseous masses in spiral nebulae. Since then progress in this special field has been unspectacular and relatively slow by conventional optical means, mainly because of the considerable faintness of the sources. Nevertheless, the optical data now seem sufficiently representative to warrant a review directed toward an eventual comparison of optical and radio results. For there can be little doubt that the new radio astronomy techniques give every indication of revolutionary new advances. Highly significant results have been obtained for the Magellanic Clouds, a beginning has been made in the local group spirals such as the Andromeda nebula, and it is reasonable to expect, as more powerful installations come into operation, that the nearer field and great cluster nebulae will come under detailed radio observation. But these matters have been considered at the recent Jodrell Bank symposium on radio astronomy, in Manchester. For this reason and because I am not a radio astronomer I shall proceed, without further discussion of these exciting new developments, to a description of the more prosaic optical data obtained from spectrographic observations.

The emission nebulae, as more or less compact gaseous or gas-and-dust formations, belong to a rather widespread type of population in our Galaxy. Though, for well-known reasons, we are able to penetrate into the depth of the Galaxy no further than 2–3 kpc for H 11 regions, several hundred nebulae have already been recorded (predominantly in the spiral arms). Sometimes the emission nebulae almost entirely occupy large regions of the sky, as, for instance, in the region of ρ, γ and o1 Cygni. In addition, in the area within latitudes ± 5°, we observe a nearly continuous faint emission background.

At wave-lengths greater than about one metre the majority of the radio emission which is observed from the Galaxy cannot be explained in terms of thermal emission from ionized interstellar gas. This conclusion is widely accepted and is based on observations of the equivalent temperature of the sky and the spectrum of the radiation. The spectrum at metre wave-lengths is of the general form: where TA is the equivalent black-body temperature of a region of sky and A is the wave-length. The exponent n varies with direction but lies between about 2·5 and 2·8, and is thus significantly greater than the value of 2·0 which is the maximum to be expected for thermal emission from an ionized gas. Furthermore the value of TA is about 1050 K at 15 m and thus greatly exceeds the electron temperature expected in H 11 regions.

At centimetre wave-lengths it is likely that the majority of the radiation observed originates in thermal emission from ionized gas; however, the present discussion is limited to a range of wave-lengths from about 1 m to 10 m where the ionized gas in the Galaxy is believed to be substantially transparent and where the origin of most of the radiation is believed to be non-thermal.

Although the sources responsible for the radio emission from the Galaxy are unknown, it may nevertheless be valuable to make a comparison between the magnitude and distribution of the emission observed in our Galaxy and that of other nebulae. Analyses have already been made to relate the total emitted power from nearby nebulae with that from the Galaxy, and by considering the integrated radiation from well-defined clusters attempts were made to extend the comparison to the average emission from fainter nebulae.

The general structure of the galaxies is extremely diverse and can be described by Hubble's classification only in the first approximation. In fact, the differences in the distribution of gas and dust in some galaxies which have been established recently, the evident dissimilarity of the intensity of the radio emission in individual galaxies, as well as other facts, suggest definite differences even among galaxies belonging to the same type according to Hubble. Hubble's classification gives a sufficiently exhaustive description of the diversity of the optical forms of galaxies and unites them in a single sequence, which, in the case of spirals, branches into two parallel sequences (ordinary and barred spirals). The interpretation of Hubble's sequence as an evolutionary one became rather popular. In this case usually either the Eo form is considered as the initial stage (Jeans, Hubble) or, conversely, the irregular galaxies and late spirals are considered as the initial stage with the elliptical galaxies being the late stage of evolution (Shapley and many others).

This paper is concerned with the space distribution in our own Galaxy of the most luminous stars of spectral types A5 and later which have been classified on the basis of low-dispersion slit spectrograms. Seventy stars of luminosity classes Ia and Iab on the Morgan–Keenan system are considered, of which the majority have been previously noted as of high luminosity at the Yerkes, Harvard, Mount Wilson, and Warner and Swasey Observatories. The Cepheid variables have not been considered.

A method is described for the discovery and observation of early-type giants which are too faint for ordinary spectroscopic observation in the blue region.

In 1945 McLaughlin found that novae form a comparatively flat sub-system in the Galaxy with an appreciable concentration towards the centre. B. V. Kukarkin came to the same conclusion in 1949. However, it was found later that this result, which contradicts our ideas of galactic structure, was a result of poor absolute magnitudes and the consequent underestimate of the distances of the novae.

There is no general agreement about either the location or the direction of winding of the spiral arms in our Galaxy. Since the measurement of the distances of gaseous nebulae is based on their association with hot stars, the locations of arms based on these objects are not independent.

The colours and brightnesses of various details about 12″ or larger in diameter were measured on photographs taken in the two wave-lengths 4360 Å and 6100 Å, of the spiral galaxies: NGC 224, 598, 3031, 3623, 3627, 3628, 4192, 4217, 4254, 4258, 4321, 4565, 4631, 5194, and 5457. The plates were taken with the 500 mm meniscus astrograph at the Kazakhstan High Altitude Observatory near Alma-Ata. Extra-focal stellar images taken with a ‘two storied plate holder’ were used for standardizing.

A comparison of the Galactic System with other stellar systems might be based on:

1. (a) the regular aspect (the population 11);

2. (b) what we might call the semi-chaotic aspect, presented by the gas and the other population 1 objects;

3. (c) the spherical corona of continuous radio-emission.

As far as (a) is concerned we are not yet in a position to say very much. The only type of objects of extreme population 11 that we have recognized with certainty in the Galactic System as well as in other comparable systems are the globular clusters. They are very incompletely known in the Galaxy, perhaps somewhat more completely in the Andromeda nebula, but few systematic studies have been made in other systems. So far as we can tell, the Andromeda nebula and the Galactic System are comparable in the number and general distribution of globular clusters.