1.1. Deducing the large-scale structure of the Galaxy

When viewed on a moonless night from a place far away from the pollution of city light, the ‘band’ of the Milky Way stands out clearly and in great detail. With binoculars, a small telescope or even the naked eye we may discern not only numerous individual stars, but also star clusters, gas nebulae shining red in the light of ionized hydrogen (H II regions), dust lanes, some of which break up into fine filaments, and dark clouds. Comparing the view from the northern and southern hemispheres, we find that the density of stars, clusters but also that of dark clouds is highest toward the constellations Sagittarius and Scorpius. With our present-day knowledge about Galactic structure, it is quite possible to visualize our place within the disk of the Milky Way and even to get some feeling of being located within a three-dimensional structure. However, it is also clear that we are at some disadvantage when trying to deduce the true structure of the Milky Way, due to our viewpoint inside it, resulting in an ‘edge-on’ viewing geometry imposed by the location of the Sun within the Galactic Disk.

            Indeed, while our understanding of the general appearance of the Milky Way has made great progress over the last century, progress which has become as for almost all other astronomical phenomena ever faster in recent decades, there are still considerable blanks to fill in. The Galactic Center itself as well as the far side of the Milky Way, beyond the center, are almost entirely inaccessible to optical astronomy, making research on the Galactic Center region and, to some degree, on large-scale Galactic structure a domain for other wavelength ranges. Of these, only the radio and part of the infrared domain are accessible to ground based telescopes, and even for these, much of the necessary technology has only been developed recently.

Sometimes, it seems easier to discern the structure of external galaxies than that of our own Milky Way, at least on a large scale that does not require high resolution. A number of external galaxies have indeed been put forward as ‘templates’ for the Milky Way. NGC 891 or NGC 4565 may present us with an edge on view of the Milky Way (e.g. van der Kruit 1984), while NGC 1232 (Mollenhoff et al 1999, see figure 1.1) may look similar to our own galaxy when seen face-on. Several components or constituents (which we will examine more closely in later sections) are seen readily in these external galaxies: in face-on views, spiral arms, inter-arm regions and integral color changes when moving in from the disk to the central region are obvious, while edge-on views show the thinness of the disk, with an even thinner dust lane cutting through its central part, as well as the oval bulge region in the center.

However, studying external galaxies does not really solve the problem of the structure of our Galaxy: there are many types of galaxy, presumably in many evolutionary stages, and even within one class, e.g. spiral galaxies, the members show a great variety of more or less obvious differences. In fact, no two galaxies are exactly alike. Thus, using external galaxies as Milky Way templates requires much knowledge about the structure of the Milky Way just to select the right galaxies for companion. Ideally, high-resolution studies of the constituents of the Milky Way and, by necessity, lower-resolution work on (many) external galaxies should and can complement each other, shedding light on the more general problems of galactic structure and evolution. However, to solve specific questions on the structure of the Milky Way as a unique object, we have little choice but to turn to our Galaxy itself and try to overcome the problems due to our position within it.

1.2. Unveiling Galactic structure: history

The awareness that the main constituents of the Milky Way are stars came with the invention of the telescope. Galilei stated in 1610 that ‘ the Galaxy is nothing else but a mass of innumerable stars planted together in clusters ’ (quoted from Weaver 1975; the material in this section is largely adapted from his articles

Figure 1.1. The famous ESO VLT image of NGC 1232, a possibly Milky Way template galaxy.

(Weaver 1975a, b) and Hoskin (1985)). The shape and configuration of the stars proved, however, difficult to determine.

In 1750, Wright published what is widely regarded as the first disklike picture of the Milky Way. A plate in his ‘An Original Theory or New Hypothesis of the Universe’ depicts what appears to be a stellar disk, with the sun within it and lines-of-sight drawn across that clearly and correctly explain why the Milky Way is perceived as a bright band we see many more stars when we look along a line-of-sight within the disk than when we look perpendicularly to the plane of the disk. However, Wright’s view only seems ‘modern’: he thought of the ‘dis ’ as part of a very large spherical shell, with a radius so large that the curvature was hardly perceptible. In the center of the sphere was ‘Heaven, the Abode of God’, while Wright assumed the far side to be ‘the Shades of Darkness and Dispare, the Desolate Regions of ye Damned’. This certainly appears to be a rather unconventional view of both the Galactic Center and extragalactic space from a modern perspective, but Wright was indeed the first to assume that the Sun was rotating around some central object. However, Wright, and his contemporaries, while trying to develop world views consistent with observations, did so with philosophical or theological reasoning rather than experiments. Thus, the insights they arrived at belong to the realm of natural philosophy more than empirical science.

Kant knew of Wright's ideas, and took them one step further. In his ‘Allgemeine Naturgeschichte und Theorie des Himmels’ (1755), he arrived at a ‘true’ disk picture, extending the hierarchical structure by deducing the existence of ‘Welteninseln’ or ‘Island Universes’ external galaxies.

The viewpoint of natural philosophy was changed fundamentally to one of empirical science when W Herschel, telescope builder and tireless observer, introduced not only observational but also statistical methods into the study of Galactic structure. He and his sister performed star counts for many lines-of-sight along a great circle, and, inventing the methodology of stellar statistics for this purpose, arrived at a picture of the Milky Way as a flattened, irregularly shaped object with the Sun close to the center. Herschel had to make two assumptions both of which later proved to be incorrect: he assumed an identical space density for the stars everywhere within the Milky Way and that stellar brightness was roughly indicative of distance, allowing his telescope to reach the edge of the system.

Herschel himself realized later in life that these assumptions were flawed, but his and other models based on this method were reproduced and refined for many decades after his death. In 1922, Kapteyn published a summary of all these efforts: his ‘Kapteyn Universe’ showed a circular, lens-shaped galaxy about 15 kpc in size, with the Sun again close to the center.

Even before Kapteyn's model was published, it was under what proved later to be a decisive attack. In 1915, Shapley had started to pin down the location of globular clusters, the distances of which he could determine by a method based on variable stars (Shapley 1918). Globular clusters are (today) known to be old halo objects and their distribution is not confined to the disk of the Galaxy. Thus they can be seen optically at very large distances. Shapley found that the center of the globular cluster system was located outside the Kapteyn Milky Way, a situation that is dynamically impossible. Thus, he arrived at a picture of a Galaxy that was much larger than before (in fact, too large), with the Sun relegated to a position closer to the edge than to the center.

Uncertainty about the nature of the spiral nebulae added to a confusing situation: while most of the supporters of a Kapteyn-like universe believed that nebulae like M 31 or M 51 were galaxies similar to our own, from the beginning of measurements of radial motions and the detection of novae in spirals, partisans of Shapley's view held to the opinion that spiral nebulae were part of the larger Milky Way. The ‘new star’ S Andromedae in M 31 seemed to support their view, being inexplicably bright if located in another galaxy. Another complication was the alleged measurement of rotation in spirals on photographic plates, by van Maanen, known to be a meticulous observer. While van Maanen's error was never quite explained, S Andromedae turned out to be a supernova, intrinsically much brighter than any ‘new star’ observed before.

Shapley and Curtis, an adherent of a small (Kapteyn) Galaxy and the extragalactic nature of spirals, met in a ‘Great Debate’ in Washington in 1920, where arguments were exchanged, and no resolution reached. In hindsight, we know that the views of both groups were partly true: Shapley's assessment of the size of the Milky Way and the Sun's location was close to being correct, while the spiral nebulae are indeed external galaxies.

This question was settled only a few years later, when Hubble found Cepheid variables in spiral nebulae, determining their distances, and a little later their general recession, which later became famous as the ‘Hubble law’ of galaxy redshift. Oort was the first to analyze Galactic rotation in 1927, finding a position for the center that roughly agreed with Shapley's determination.

Finally, in 1930 the reason for much of the disagreement and confusion became clear when Trumpler demonstrated the existence of Galactic extinction in his investigations of photometric distances, linear scales and reddening of Galactic open clusters. Thus, an absorbing dust component for the interstellar medium (ISM) was established as an important constituent of the Milky Way, even in regions where its presence was not obvious as dark clouds or filaments. These were only now realized to be absorbing layers of material, and not starless voids or ‘holes in the sky’ (as the astrophotography pioneer E E Barnard thought). It became clear that in the optical wavelength range our view is limited to a few kpc, and the Galactic structure at large cannot be inferred from star counts (though the local disk structure can still be investigated by stellar statistics).

For this reason the center of the Milky Way, as officially adopted by the International Astronomical Union in 1959 as the origin for the Galactic coordinate system, was eventually based on the detection of strong radio emission from the nucleus of our Galaxy (Piddington and Minnett 1951).

1.3. ‘External’ views

More than seven decades later, data from many wavelength ranges, many of which penetrate the layer of dust extinction, are at our disposal to derive Galactic structure. Still, all our direct observational views are (and will remain for the indefinite future) internal and edge-on, with all the associated problems. By now, they cover the entire electromagnetic spectrum. This includes the radio regime, where we encounter non-thermal synchrotron emission from relativistic electrons at long cm wavelengths, emission from neutral atomic hydrogen at 21 cm and molecules, most prominently CO, which are used to trace the molecular gas component, at mm wavelengths. The far and mid-infrared region is dominated by thermal dust emission, while in the near infrared (NIR) we encounter emission from low mass, cool stars. The optical and UV bands are most affected by interstellar extinction, limiting our view to nearby stars,

Figure 1.2. Draft of the ‘external view’ of the Milky Way developed by the author and E Janssen for the exhibition ‘Seven Hills’ in Berlin (2000).

dust clouds and H II regions. X-rays, only accessible to satellite observatories, provide information on the hot component of the ISM, while the highest energy γ-rays mostly arise in collisions of cosmic rays with hydrogen atoms.

It is our task to piece together a coherent picture of Galactic structure based on these diverse sources of information on the different constituents of the Galaxy. A view of the Milky Way as it might appear to an external observer is necessarily an artist's conception, and partially based on (hopefully) educated guesses. Two attempts have been made: J Lomberg's painting at the National Air and Space Museum in Washington and a view of the Milky Way (figure 1.2) put together by the author and the artist E Janssen (European Southern Observatory) for the exhibition ‘Seven Hills Images and Perspectives for the 21st Century’ in Berlin  (2000). Both images are based on the available data. Lomberg's view focused on the Sun and its fairly well-known surroundings, thus the unknown details on the far side of the Galaxy are suitably blurred with distance. The ‘Berlin’ Milky Way is shown face-on, thus we had to invent a likely structure on the far side; we settled on an overall shape for the spiral structure, for which we chose a model with four main spiral arms, in accordance with most (but not all) the evidence.

We will examine the main constituents in more detail later, but will familiarize ourselves with the main components of the large-scale Galactic structure at this point: external views are dominated by the distinct constituents of the disk and specifically the spiral arms: young massive blue stars, reddish H II regions as the sites of stars still in the process of ionizing and dissolving their birth cloud, young clusters and dust lanes and filaments, the cradles of ongoing star formation. The disk has a diameter ≤ 30 kpc (but no sharp edge), a thickness of at most 1 kpc (depending on population), and its surface brightness falls roughly exponentially, with a scale length of 2.2-2.8 kpc. The Sun is at a distance of ∼8 kpc (7-8.5 kpc) from the center. In the vicinity of the Sun, the surface mass density is Σ_tot = (71±6) M_ּ pc⁻², a fairly certain value confirmed by a number of investigations (Kuijken and Gilmore 1991, Olling and Merrifield 2001). The volume density in the disk is far less certain: values range from ρ_tot = 0.11 to 0.076 M_ּ pc⁻³ (Creze et al 1998, Holmberg and Flynn 2000). The stellar surface density close to the Solar circle is in the range Σ_∗ = (25–50) M_ּ pc⁻². Correspondingly, the local surface density of dark matter is only very poorly known; it is estimated at Σ_DM = (10-35) M_ּ pc⁻². There is some indication of a stellar warp in the disk, and stronger evidence for a warped distribution of dust and gas.

A weak bar (or triaxial bulge) of diameter ∼3 kpc is seen in the inner part of the Galaxy, and the general color changes from whitish blue in the disk region too range or reddish in the bulge, indicative of a change in stellar population from a mix dominated, at least in luminosity, by young stars to a population made up mostly of older stars.

Beyond and above the Galactic disk extends the tenuous halo with its scattered old stars and globular clusters, the density of which is, however, strongly concentrated toward the center, and which, in addition, shows substructure and subpopulations (e.g. Zinn 1985, Burkert and Smith 1997). Dark matter, likely to be a very important halo component and decisive in structure formation scenarios, remains enigmatic, even though it is thought to dominate the overall mass budget of the Galaxy.

The Sun is located within a structure that is sometimes called the ‘Orion Arm’, but would better be named the ‘Orion Spur’. Evidence points to it not being a real spiral arm, but a short protrusion, as seen frequently in images of external galaxies. We may be fortunate not to be positioned in the midst of a ‘real’ spiral arm: while spiral galaxies, seen face-on, are largely transparent (Xilouris et al 1999, Bosma et al 1992), the enhanced opacity caused by local dust clouds likely to be encountered in a strong spiral arm might have limited the region of the Milky Way accessible to optical studies still further. It might even have rendered many or all external galaxies invisible in the optical wavelength range, greatly expanding the galactic ‘zone of avoidance’ and delaying the development of extragalactic astronomy to a time where measurements of extinction-free tracers were possible, thus profoundly changing the history of our view of the cosmos and the Galaxy retold briefly in previous sections.