The appearance of the Milky Way for an observer situated within our Galaxy is determined by the spatial distribution of stars and absorbing interstellar matter. Hence it may be hoped that the study of the surface brightness of the Milky Way permits to derive the spiral structure of our Galaxy.

The integrated spectrum of the Milky Way can give some information on the composition and large-scale structure of the Galaxy.

A maximum of dust density along the inner edge of the Cygnus arm is found. It seems that the same happens along the inner edges of the Perseus and the Sagittarius arms.

The Basel Observatory program of the determination of disc- and halo-density gradients for different intervals of absolute magnitude comprises in addition to Milky Way fields several directions, all pointing to Selected Areas near a plane perpendicular to the galactic equator and passing through the sun and the galactic centre. It was started with SA 51 (Becker, 1965) and continued with Sa 57, 54 and 141 (Fenkart, 1967, 1968, 1969).

The distribution of long-period Population I cepheids is studied. An age comparison of cepheids and galactic clusters shows that cepheids with periods > 11.25 d, corresponding to ages of ≲ 30 × 106 yr, should be almost as good spiral arm tracers as galactic clusters with earliest types b 2–3. The distances of cepheids are determined from a revised period-luminosity-colour relation and from colour excesses, for the determination of which a new, purely photometric method is given. The resulting distribution of cepheids shows a good correlation with the spiral arms, as traced by young clusters.

The existing data on the distribution of O and B stars, of optical and radio H II sources, of H I and of cosmic dust have been assembled for the Carina-Centaurus Section of the Milky Way, lII = 265° to 305°. The published data have been supplemented by recent photoelectric UBV data and new photographic material. Two Working Diagrams (Figures 9 and 10) of the Carina Spiral Feature have been prepared. The Feature is sharply bounded at lII = 282° and again at lII = 295° in the range of distance from 1.5 to 6 kpc from the sun. Its outer rim is observed from the sun almost tangentially to a distance of 8 kpc from the sun. The Feature is found to bend at distances greater than 9 to 10 kpc from the sun, a result shown by both radio H I and radio H II data.

Figure 9 presents our basic data for the stellar, gas and dust components of the Feature. The O and early B stars and the H II Regions are closely associated and within 6 kpc of the sun they are concentrated in the range 285° < lII < 295°. The distribution in longitude of H I is broader and spills over on both sides of the O and B and the H II peak distributions. Long period cepheids yield a concentration similar to that shown by O and B stars and H II Regions. The visual interstellar absorption between lII = 282° and 295° is represented by a value AV = 0.5 mag kpc−1, or less, applicable to distances of 4 to 5 kpc. Much higher absorption is present on the outside of the Carina Spiral Feature, 265° < lII < 280°, where total visual absorptions as great as 3.5 mags. are found at distances of the order of 2 kpc. Even heavier absorption is indicated for these longitudes at 4 kpc from the sun, thus suggesting that the heavy obscuration on the outside of the Carina Spiral Feature is a phenomenon of general structural relevance (see Figures 8 and 9). Only small values of AV are found at the inside of the Spiral Feature.

The Working Diagram (Figure 10) shows that the O and B star peak and the H II peak have a width of 800 parsecs (12°) at 4 kpc from the sun, whereas the H I width is at least 1500 parsecs at the same distance. The peak of the O and B star distribution and of the H II distribution lies at about 600 parsecs (8° at 4 kpc) within the outer edge of the spiral feature. The heaviest interstellar absorption is on the outside of the Feature.

Distances have been determined for 436 OB-type stars in Carina. A sharp outside edge is found to the OB star distribution with respect to the center of the Galaxy. The layer of OB stars follows the galactic plane to distances of the order of 3 kpc, but at distances greater than 4 kpc it appears to bend away from the plane to negative latitudes by 2° or 3° out to distances of the order of 10 kpc.

A photometric investigation of the association Car OB 2 was carried out with a photoelectric standard sequence observed with the ESO photometric telescope and plates taken with the ADH telescope of Boyden Observatory. The distance of Car OB 2 is 2.04 kpc, the color excess EB–V is 0m.38. The main sequence of this ‘OB’ association reaches at least to late A-type stars. The luminosity function was determined and compared with the luminosity functions of very young open clusters observed by Walker (1957). The distance of a group of 13 early-type stars in the area of Car OB 2 (‘group c’) was found to be 2.5 kpc, the reddening 0m.68. The relation of Car OB 2 and group c to the structure of the Carina arm is regarded.

The open cluster Stock 16 and some distant OB stars in Crux have been studied. Lack of absorption between spiral arms -I and -II is apparent. Negative residual radial velocities are found at distances of about 6 kpc at lII = 298°.

From some standpoints it should seem surprising that not all of the OB stars of our Galaxy that can be identified to a feasible limiting magnitude (say 12–13) by conventional objective prism techniques have yet been so identified, but this is in fact the state of our published data. Some people here will recall that such a systematic survey for the northern Milky Way was carried out several years ago jointly by the Hamburg and Warner and Swasey Observatories. Now at the Warner and Swasey Observatory we are extending this northern survey into the southern Milky Way that could not be reached from the north, and we are making this extension as homogeneous as possible with respect to the northern survey.

The spiral structure of our Galaxy, as defined by optical evidence, shows a striking gap around lII = 270°. Intrigued by this feature, the author started, in 1955, a survey of early type stars based on objective-prism plates taken with the ADH-Schmidttelescope of the Boyden Observatory (South-Africa); dispersion of the spectra: 240 Å/mm at Hγ.

In order to find evidence for supporting or rejecting the idea of a major spiral feature in the direction of lII = 270°, as proposed by Velghe (1969), it is necessary to extend the limits of the survey both in longitude and in distance. The extension towards smaller longitudes was placed arbitrarily at lII = 257°, while in the direction of Carina, the investigation was confined to lII = 285°. In doing so, sufficient overlap was made with the studies by Velghe and by Graham and Lyngå (1965), in order to get a homogeneous and complete material in a large, but important section of the southern Milky Way. Following the same technique as outlined in the preceding paper, about 360 young type stars were selected from objective-prism plates, taken with the ADH-telescope of the Boyden Observatory. Later on (in 1966) a first series of UBV photoelectric measurements was carried out at the Boyden Observatory, however with insufficient accuracy. Before finishing the reductions, a second series was made with the 1m-photometric telescope of the ESO (La Silla, Chile) at the end of 1968. These (ESO)-values, in the Johnson UBV-system, will be available before long. In the definite form, they will be combined with the Boyden-measurements. The surface distribution of the stars in the whole section is shown in Figure 1. Many of the stars between lII = 273° and 282° are of rather late B-type, but they were included for the purpose of determining the absorption in this field. The space distribution will enable us to decide whether the edge of the Carina spiral feature is seen in this direction (Bok et al., 1969) and if there is any indication for a possible linkage from Carina to Puppis. To extend the distance limit, a search for faint blue stars, applying the idea of Bok (1966), is planned and some photoelectric sequences will be measured early in 1970 as standards for photographic photometry. As another approach to the galactic structure in this field a complete set of red objective-prism plates has been obtained since 1966, to pick out the objects with H α in emission. As for the reality of the association I Vel, some preliminary material on Radial Velocities will be collected also in 1970, using the Fehrenbach astrograph of the ESO at La Silla (Chile).

Observing has been underway for several years, at Kitt Peak and Cerro Tololo, of the bright B and A stars, and of the brighter and nearer open clusters, to supply data necessary for calibration of 4-color and Hβ indices in terms of intrinsic color and absolute magnitude, as well as for use in studies of galactic structure. To date, data are on hand – some published, most being prepared for publication – for all O- to G0-type stars brighter than V = 5m.0, for most O- to B5-type to 6m.5, and for most northern hemisphere O- to G0-type to 6m.5. In addition, the following open clusters have been studied: Hyades, Coma, Ursa Major, Praesepe, α Persei, Pleiades, IC 2602, IC 2391, h and χ Persei, NGC 6231, NGC 752, and IC 4665.

At the Prague IAU meeting (Isserstedt and Schmidt-Kaler, 1967) we presented some results of investigations on stellar rings, a new kind of stellar association and powerful spiral tracer. We have since kept looking for similar features of gas or dust which might be early stages in the evolution of stellar rings.

Stellar rings are shell type prolate ellipsoidal stellar aggregates. The most important property is the constancy of the minor diameters which allows precise geometric distance determinations up to great distances. New photoelectric and photographic UBV-observations of two stellar rings confirm the reality and the diameter constancy of these objects. All A- and F-stars of a nearby ring in Aquila exhibit an UV-excess, similar to that observed in the very young open cluster NGC 2264.

With the Fehrenbach objective prism radial velocities of about 700 stars of type B0 to A0 were determined in two fields of the Southern Milky Way (lII = 295° bII = −0.6°; lII = 320° bII = −2.5°), with a mean error of ±20 km s−1. An additional photographic UBV-photometry with plates of the ADH-telescope at Boyden Observatory was accomplished. Minimum distances for the stars resulted from absorption-corrected magnitudes and a MK-spectral classification. About 200 stars lay at distances greater than 1.5 kpc from the sun. The largest distances determined were 5 kpc. From the radial velocities and distances circular velocities were derived and plotted against galactocentric distances R. Even within the possible error limits a positive velocity gradient showed up in the range 8 kpc < R < 9.5 kpc, which French authors had already found for the region 10.5 kpc < R < 12.5 kpc. If there do not exist significant deviations from circular motion for these stars, a conformity with Schmidt's 1965 model cannot be obtained.

The results of a study of the dispersion of velocities of O-B5 stars up to a distance of 3 kpc are described.

OB stars, as the most typical members of O-associations, are of great interest for the problems of stellar evolution. Their concentration in spiral arms indicates the importance of the O-associations for the formation of the spiral arm population in galaxies.

A proposal for a cooperative effort to secure three-color photometry of all stars up to the 18th magnitude of a 10° field around the galactic equator has been made.

The basic ideas and some of the most important recent developments of the gravitational theories of spiral structure are described. A separation between linear and non linear effects is made. The linear self consistent problem consists of the problem of modes and of the initial value problem, which is discussed here in some detail. More emphasis is put on the non linear problem near resonances and in particular the inner Lindblad resonance. The linear density response to a slightly growing spiral potential (trailing or leading) near the inner Lindblad resonance is always trailing, while non linear effects form a density distribution with a roughly quadruple symmetry.